From owner-freebsd-current Sun Jul 27 03:12:57 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA25624 for current-outgoing; Sun, 27 Jul 1997 03:12:57 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA25615 for ; Sun, 27 Jul 1997 03:12:51 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id TAA00531; Sun, 27 Jul 1997 19:42:47 +0930 (CST) From: Michael Smith Message-Id: <199707271012.TAA00531@genesis.atrad.adelaide.edu.au> Subject: Re: make world failure In-Reply-To: <199707250821.SAA29154@godzilla.zeta.org.au> from Bruce Evans at "Jul 25, 97 06:21:05 pm" To: bde@zeta.org.au (Bruce Evans) Date: Sun, 27 Jul 1997 19:42:46 +0930 (CST) Cc: bde@zeta.org.au, msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG, mrcpu@cdsnet.net X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Bruce Evans stands accused of saying: > >I'm not sure I follow you here. It is not possible to run 'make > >depend' without generating the headers, as the depend will fail when > >the headers aren't found. > > Running `make depend' is not a requirement for building sources, at > least for the first build. Makefiles that require it are broken. Fair enough. > Another bugfeature in libedit/Makefile is that `.depend' instead of > `beforedepend' depends on the generated headers. .depend is a private > target in bsd.dep.mk (if the default `depend' target is used), so it > should not be used. OTOH, the default .depend target: > > depend: beforedepend .depend afterdepend _SUBDIR > > does not enforce the beforeness of `beforedepend' - it does not prevent > `make -j2' from attempting to build beforedepend and .depend concurrently. > Things work correctly when .depend depends on the generated headers. (I > don't know how to enforce the required ordering. I tried to use a > `beforeall' target in sh/Makefile to generate the headers before `all', > but this failed for `make -j2'. `make' probably just needs a way to > force building of a list of targets in list order.) depend: _SUBDIR _SUBDIR: afterdepend afterdepend: .depend .depend: beforedepend Moderately ugly, but I don't _think_ that would break. > >Given that .depend is written into the object directory, I can't see > >anything wrong with the method that libedit uses. It's certainly a > >sight less disgusting than the way that sh/Makefile does it. > > The method in sh/Makefile has the advantage of actually working. ... for some values of "working" which exclude "ease of development". > Bruce -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Sun Jul 27 03:15:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA25694 for current-outgoing; Sun, 27 Jul 1997 03:15:21 -0700 (PDT) Received: from news1.gtn.com (news1.gtn.com [194.77.0.15]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA25689 for ; Sun, 27 Jul 1997 03:15:16 -0700 (PDT) Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) with UUCP id MAA12032 for current@FreeBSD.org; Sun, 27 Jul 1997 12:00:20 +0200 (MET DST) Received: (from andreas@localhost) by klemm.gtn.com (8.8.6/8.8.6) id LAA00708; Sun, 27 Jul 1997 11:57:17 +0200 (CEST) Message-ID: <19970727115717.28963@gtn.com> Date: Sun, 27 Jul 1997 11:57:17 +0200 From: Andreas Klemm To: current@FreeBSD.org Subject: using a kernel of Jul-27, isdn connection to Internet locks up Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-current@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk Hi ! With my last -current kernel of Jul 21th, I hab no problems to connect to internet using bisdnd 0.97. Using the latest bits and bites from this morning (from cvsup2.freebsd.org) causes an established connection to internet to lock within 2 minutes or so. Killing the bisdnd daemon and starting it again doesn't even result in an dial on demand to internet. The default route to my ISP's router is still present I checked with netstat -rn. I noticed no error message in the xconsole display. No core dump, nothing. Might it be related to some changes in the last week ? This behaviour is reproduceable, I rebooted the machines two times even turning power off and on. Andreas /// -- Andreas Klemm | klemm.gtn.com - powered by Symmetric MultiProcessor FreeBSD http://www.freebsd.org/~fsmp/SMP/SMP.html http://www.freebsd.org/~fsmp/SMP/benches.html From owner-freebsd-current Sun Jul 27 04:29:23 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA27768 for current-outgoing; Sun, 27 Jul 1997 04:29:23 -0700 (PDT) Received: from isbalham.ist.co.uk (isbalham.ist.co.uk [192.31.26.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA27763 for ; Sun, 27 Jul 1997 04:29:21 -0700 (PDT) Received: from gid.co.uk (uucp@localhost) by isbalham.ist.co.uk (8.8.4/8.8.4) with UUCP id MAA16491 for freebsd.org!current; Sun, 27 Jul 1997 12:12:35 +0100 (BST) Received: from [194.32.164.2] by seagoon.gid.co.uk; Sun, 27 Jul 1997 12:18:11 +0100 X-Sender: rb@194.32.164.1 Message-Id: Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Date: Sun, 27 Jul 1997 12:14:10 +0100 To: current@freebsd.org From: Bob Bishop Subject: lockstatus() panic: a data point Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, FWIW, I used to be regularly afflicted with the lockstatus panic. I've been running with the following patch (I forget who from) since 28 May, since when I haven't had a single occurrence. Index: ufs_vnops.c =================================================================== RCS file: /home/ncvs/src/sys/ufs/ufs/ufs_vnops.c,v retrieving revision 1.50 diff -u -r1.50 ufs_vnops.c --- ufs_vnops.c 1997/05/17 18:32:53 1.50 +++ ufs_vnops.c 1997/05/27 13:11:19 @@ -1788,6 +1788,9 @@ } */ *ap; { + if (ap->a_vp->v_flag & VXLOCK) + return TRUE; + return (lockstatus(&VTOI(ap->a_vp)->i_lock)); -- Bob Bishop (0118) 977 4017 international code +44 118 rb@gid.co.uk fax (0118) 989 4254 between 0800 and 1800 UK From owner-freebsd-current Sun Jul 27 05:37:48 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA02254 for current-outgoing; Sun, 27 Jul 1997 05:37:48 -0700 (PDT) Received: from nagual.pp.ru (ache.relcom.ru [194.58.229.133]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA02249; Sun, 27 Jul 1997 05:37:43 -0700 (PDT) Received: (from ache@localhost) by nagual.pp.ru (8.8.6/8.8.5) id QAA00682; Sun, 27 Jul 1997 16:37:06 +0400 (MSD) Date: Sun, 27 Jul 1997 16:37:04 +0400 (MSD) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= To: Garrett Wollman cc: Marc Slemko , wollman@FreeBSD.ORG, FreeBSD-current Subject: Re: Another 'fetch' reset by peer: -b NOT work for this site... In-Reply-To: <199707270233.WAA24701@khavrinen.lcs.mit.edu> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 26 Jul 1997, Garrett Wollman wrote: > < said: > > > 18:00:43.330713 valis.worldgate.com.tr-rsrb-p2 > 205.180.58.12.http: SP 1382936415:1382936515(100) win 16464 (DF) > > 18:00:43.382215 205.180.58.12.http > valis.worldgate.com.tr-rsrb-p2: S 1234567:1234567(0) ack 1382936416 win 2048 > > 18:00:43.382867 valis.worldgate.com.tr-rsrb-p2 > 205.180.58.12.http: FP 101:155(54) ack 1 win 16464 (DF) > > 18:00:43.433453 205.180.58.12.http > valis.worldgate.com.tr-rsrb-p2: R 1:1(0) ack 102 win 2048 > > This is pretty clearly brain-damaged behavior. Coule you please add an option into fetch to turn off T/TCP? It is not good to turn it off globally just for fetch... -- Andrey A. Chernov http://www.nagual.pp.ru/~ache/ From owner-freebsd-current Sun Jul 27 09:36:15 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id JAA11958 for current-outgoing; Sun, 27 Jul 1997 09:36:15 -0700 (PDT) Received: from ox.ismi.net (root@ox.ismi.net [206.31.56.6]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id JAA11953 for ; Sun, 27 Jul 1997 09:36:12 -0700 (PDT) Received: from aerosmith.dyn.ml.org (mrr@pm1-25.ismi.net [206.31.56.35]) by ox.ismi.net (8.6.9/8.6.9) with SMTP id MAA15414; Sun, 27 Jul 1997 12:35:24 -0400 Date: Sat, 27 Jul 1996 12:29:10 -0400 (EDT) From: "Michael R. Rudel" To: Greg Shaffer cc: current@FreeBSD.ORG Subject: Re: Minor nit In-Reply-To: <199707260500.WAA09466@intrepid.leaky.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Fri, 25 Jul 1997, Greg Shaffer wrote: > The following output from 'w' shows that I have a process that has been idle > for > 5 days, but the system has only been up for just under 2 days. > > {intrepid}/home/gshaffer>> w > 9:51PM up 1 day, 23:37, 4 users, load averages: 0.80, 1.23, 1.38 > USER TTY FROM LOGIN@ IDLE WHAT > gshaffer v0 - Wed10PM 1day xinit > /home/gshaffer/.xinitrc -bpp 15 -- > gshaffer p0 :0.0 Wed10PM 1day -sh (csh) > gshaffer p1 :0.0 Wed10PM 5days sleep 60 > gshaffer p2 :0.0 Wed10PM - w > > System is -current (about 14 days old). This happens when your XFree system is older then your proc_fs (or something, somewhere) in your system. Best bet I can tell you is to recompile (at least) xterm. -- Michael R. Rudel -=- FreeBSD: There are no limits -=- mrr@aerosmith.dyn.ml.org FreeBSD aerosmith.dyn.ml.org 3.0-CURRENT PGP Key Block: finger mrrpgp@aerosmith.dyn.ml.org From owner-freebsd-current Sun Jul 27 11:14:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA15388 for current-outgoing; Sun, 27 Jul 1997 11:14:36 -0700 (PDT) Received: from scanner.worldgate.com (scanner.worldgate.com [198.161.84.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA15383 for ; Sun, 27 Jul 1997 11:14:34 -0700 (PDT) Received: from znep.com (uucp@localhost) by scanner.worldgate.com (8.8.5/8.8.5) with UUCP id MAA03039; Sun, 27 Jul 1997 12:14:27 -0600 (MDT) Received: from localhost (marcs@localhost) by alive.znep.com (8.7.5/8.7.3) with SMTP id MAA26342; Sun, 27 Jul 1997 12:16:15 -0600 (MDT) Date: Sun, 27 Jul 1997 12:16:15 -0600 (MDT) From: Marc Slemko To: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= cc: Garrett Wollman , FreeBSD-current Subject: Re: Another 'fetch' reset by peer: -b NOT work for this site... In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from QUOTED-PRINTABLE to 8bit by hub.freebsd.org id LAA15384 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 27 Jul 1997, [KOI8-R] Андрей Чернов wrote: > On Sat, 26 Jul 1997, Garrett Wollman wrote: > > > < said: > > > > > 18:00:43.330713 valis.worldgate.com.tr-rsrb-p2 > 205.180.58.12.http: SP 1382936415:1382936515(100) win 16464 (DF) > > > 18:00:43.382215 205.180.58.12.http > valis.worldgate.com.tr-rsrb-p2: S 1234567:1234567(0) ack 1382936416 win 2048 > > > 18:00:43.382867 valis.worldgate.com.tr-rsrb-p2 > 205.180.58.12.http: FP 101:155(54) ack 1 win 16464 (DF) > > > 18:00:43.433453 205.180.58.12.http > valis.worldgate.com.tr-rsrb-p2: R 1:1(0) ack 102 win 2048 > > > > This is pretty clearly brain-damaged behavior. > > Coule you please add an option into fetch to turn off T/TCP? It is not > good to turn it off globally just for fetch... The following code, which I have been using locally for a while, does that. Index: fetch.1 =================================================================== RCS file: /usr/cvs/src/usr.bin/fetch/fetch.1,v retrieving revision 1.18 diff -u -r1.18 fetch.1 --- fetch.1 1997/07/25 19:35:41 1.18 +++ fetch.1 1997/07/27 18:13:01 @@ -99,6 +99,10 @@ under any circumstances, even if the transfer failed or was incomplete. .It Fl r Restart a previously interrupted transfer. +.It Fl t +Ensure that the use of T/TCP is not attempted for connections. This is +used to workaround bugs in some remote OS stacks that give improper +replies to T/TCP connections. .It Fl T Ar seconds Set timeout value to .Ar seconds. Index: fetch.h =================================================================== RCS file: /usr/cvs/src/usr.bin/fetch/fetch.h,v retrieving revision 1.4 diff -u -r1.4 fetch.h --- fetch.h 1997/07/25 19:35:42 1.4 +++ fetch.h 1997/07/27 18:13:01 @@ -50,6 +50,7 @@ int fs_precious; /* -R option */ int fs_auto_retry; /* -a option */ int fs_linux_bug; /* -b option */ + int fs_use_connect; /* -t option */ time_t fs_modtime; void *fs_proto; int (*fs_retrieve)(struct fetch_state *); Index: http.c =================================================================== RCS file: /usr/cvs/src/usr.bin/fetch/http.c,v retrieving revision 1.10 diff -u -r1.10 http.c --- http.c 1997/07/26 20:18:43 1.10 +++ http.c 1997/07/27 18:13:01 @@ -581,6 +581,20 @@ fs->fs_status = "sending request message"; setup_sigalrm(); alarm(timo); + + /* some hosts do not properly handle T/TCP connections. If + * sendmsg() is used to establish the connection, the OS may + * choose to try to use one which could cause the transfer + * to fail. Doing a connect() first ensures that the OS + * does not attempt T/TCP. + */ + if (fs->fs_use_connect && (connect(s, (struct sockaddr *)&sin, + sizeof(struct sockaddr_in)) < 0)) { + warn("connect: %s", https->http_hostname); + fclose(remote); + return EX_OSERR; + } + if (sendmsg(s, &msg, fs->fs_linux_bug ? 0 : MSG_EOF) < 0) { warn("sendmsg: %s", https->http_hostname); fclose(remote); Index: main.c =================================================================== RCS file: /usr/cvs/src/usr.bin/fetch/main.c,v retrieving revision 1.41 diff -u -r1.41 main.c --- main.c 1997/07/25 19:35:44 1.41 +++ main.c 1997/07/27 18:13:01 @@ -73,7 +73,7 @@ fs.fs_verbose = 1; change_to_dir = file_to_get = hostname = 0; - while ((c = getopt(argc, argv, "abc:D:f:h:HilLmMnNo:pPqRrT:vV:")) != -1) { + while ((c = getopt(argc, argv, "abc:D:f:h:HilLmMnNo:pPqRrtT:vV:")) != -1) { switch (c) { case 'D': case 'H': case 'I': case 'N': case 'L': case 'V': break; /* ncftp compatibility */ @@ -130,6 +130,10 @@ fs.fs_precious = 1; break; + case 't': + fs.fs_use_connect = 1; + break; + case 'T': /* strtol sets errno to ERANGE in the case of overflow */ errno = 0; From owner-freebsd-current Sun Jul 27 12:19:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA17592 for current-outgoing; Sun, 27 Jul 1997 12:19:26 -0700 (PDT) Received: from helbig.informatik.ba-stuttgart.de (rvc1.informatik.ba-stuttgart.de [141.31.112.22]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA17584 for ; Sun, 27 Jul 1997 12:19:18 -0700 (PDT) Received: (from helbig@localhost) by helbig.informatik.ba-stuttgart.de (8.8.6/8.8.5) id VAA01496; Sun, 27 Jul 1997 21:19:11 +0200 (MET DST) From: Wolfgang Helbig Message-Id: <199707271919.VAA01496@helbig.informatik.ba-stuttgart.de> Subject: Re: using a kernel of Jul-27, isdn connection to Internet locks up In-Reply-To: <19970727115717.28963@gtn.com> from Andreas Klemm at "Jul 27, 97 11:57:17 am" To: andreas@klemm.gtn.com (Andreas Klemm) Date: Sun, 27 Jul 1997 21:19:10 +0200 (MET DST) Cc: current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Hi ! > > With my last -current kernel of Jul 21th, I hab no problems > to connect to internet using bisdnd 0.97. > > Using the latest bits and bites from this morning > (from cvsup2.freebsd.org) causes an established connection > to internet to lock within 2 minutes or so. > > Killing the bisdnd daemon and starting it again doesn't even > result in an dial on demand to internet. The default route > to my ISP's router is still present I checked with netstat -rn. > > I noticed no error message in the xconsole display. No core > dump, nothing. Might it be related to some changes in the > last week ? > > This behaviour is reproduceable, I rebooted the machines > two times even turning power off and on. It still works with the kernel source from yesterday, Jul 26 17:22. So either your problem is hidden on my system or the disturbing change occurred between yesterday and this morning. What does bisdntrc -i3 say? Wolfgang > > Andreas /// > > -- > Andreas Klemm | klemm.gtn.com - powered by > Symmetric MultiProcessor FreeBSD > http://www.freebsd.org/~fsmp/SMP/SMP.html > http://www.freebsd.org/~fsmp/SMP/benches.html > From owner-freebsd-current Sun Jul 27 14:01:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA21334 for current-outgoing; Sun, 27 Jul 1997 14:01:36 -0700 (PDT) Received: from helbig.informatik.ba-stuttgart.de (rvc1.informatik.ba-stuttgart.de [141.31.112.22]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA21325 for ; Sun, 27 Jul 1997 14:01:30 -0700 (PDT) Received: (from helbig@localhost) by helbig.informatik.ba-stuttgart.de (8.8.6/8.8.5) id WAA00364 for FreeBSD-current@FreeBSD.org; Sun, 27 Jul 1997 22:27:39 +0200 (MET DST) From: Wolfgang Helbig Message-Id: <199707272027.WAA00364@helbig.informatik.ba-stuttgart.de> Subject: Sound Blaster PnP in -current To: FreeBSD-current@FreeBSD.org Date: Sun, 27 Jul 1997 22:27:39 +0200 (MET DST) X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.org X-Loop: FreeBSD.org Precedence: bulk Hi, I installed Sound Blaster 16 PnP. It is not recognized by my kernel. Rumors have it that Sound Blaster 16 PnP works well in 2.2.x. So I am wondering if anyone got it to work under -current. This is /var/log/messages: Jul 27 21:31:22 helbig /kernel: npx0: INT 16 interface Jul 27 21:31:22 helbig /kernel: sb0 not found at 0x220 Jul 27 21:31:22 helbig /kernel: sbxvi0 not found Jul 27 21:31:22 helbig /kernel: sbmidi0 not found at 0x330 Jul 27 21:31:22 helbig /kernel: opl0 not found at 0x388 Strange thing is these messages are printed after the npx0-line, which used to be the last one. This is part of the kernel configuration: (straight from LINT) # Sound support controller snd0 device sb0 at isa? port 0x220 irq 5 drq 1 vector sbintr device sbxvi0 at isa? drq 5 device sbmidi0 at isa? port 0x330 device opl0 at isa? port 0x388 Finally I changed the port address of the midi driver in sound_config.h from 0x300 to LINT's 0x330. All these values are factory default, and there are no conflicts with other devices as to my knowledge. Any ideas? Thanks Wolfgang From owner-freebsd-current Sun Jul 27 14:04:10 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA21471 for current-outgoing; Sun, 27 Jul 1997 14:04:10 -0700 (PDT) Received: from alpha.xerox.com (alpha.Xerox.COM [13.1.64.93]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id OAA21466 for ; Sun, 27 Jul 1997 14:04:07 -0700 (PDT) Received: from crevenia.parc.xerox.com ([13.2.116.11]) by alpha.xerox.com with SMTP id <52607(4)>; Sun, 27 Jul 1997 14:03:36 PDT Received: from localhost by crevenia.parc.xerox.com with SMTP id <177512>; Sun, 27 Jul 1997 14:03:25 -0700 To: Marc Slemko cc: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= , Garrett Wollman , FreeBSD-current Subject: Re: Another 'fetch' reset by peer: -b NOT work for this site... In-reply-to: Your message of "Sun, 27 Jul 97 11:16:15 PDT." Date: Sun, 27 Jul 1997 14:03:24 PDT From: Bill Fenner Message-Id: <97Jul27.140325pdt.177512@crevenia.parc.xerox.com> Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk As far as I can tell, Marc's patch and the "-b" flag *should* do the same thing -- if you don't give the MSG_EOF flag to sendmsg(), it will not set the FIN bit on the first data packet. The only difference is that explicitly using connect(), it won't send data with the first packet. RFC793 section 3.9 explicitly allows data with a SYN. Bill From owner-freebsd-current Sun Jul 27 14:23:45 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA22249 for current-outgoing; Sun, 27 Jul 1997 14:23:45 -0700 (PDT) Received: from scanner.worldgate.com (scanner.worldgate.com [198.161.84.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA22244 for ; Sun, 27 Jul 1997 14:23:43 -0700 (PDT) Received: from znep.com (uucp@localhost) by scanner.worldgate.com (8.8.5/8.8.5) with UUCP id PAA14342; Sun, 27 Jul 1997 15:23:35 -0600 (MDT) Received: from localhost (marcs@localhost) by alive.znep.com (8.7.5/8.7.3) with SMTP id PAA27861; Sun, 27 Jul 1997 15:25:26 -0600 (MDT) Date: Sun, 27 Jul 1997 15:25:26 -0600 (MDT) From: Marc Slemko To: Bill Fenner cc: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= , Garrett Wollman , FreeBSD-current Subject: Re: Another 'fetch' reset by peer: -b NOT work for this site... In-Reply-To: <97Jul27.140325pdt.177512@crevenia.parc.xerox.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Sun, 27 Jul 1997, Bill Fenner wrote: > As far as I can tell, Marc's patch and the "-b" flag *should* do the > same thing -- if you don't give the MSG_EOF flag to sendmsg(), it > will not set the FIN bit on the first data packet. > > The only difference is that explicitly using connect(), it won't send > data with the first packet. RFC793 section 3.9 explicitly allows data > with a SYN. And, due to this fact, it won't send a FIN in response to the other end's SYN. Either one of these two things could be what is causing problems on the remote system. This particular problem is not caused by the early half-close that -b works around. Even though sending data with the SYN is permitted by the RFC, the fact is that there are too many broken boxes out there that don't work right with it. It could be viewed as an argument for disabling T/TCP (and the sending of data in the SYN) on your machine entirely or for allowing the few applications which do make such calls to disable it. I don't really have any preference either way. It could also be viewed as an argument for giving up and making the changes to fetch so it is less efficient and doesn't cause problems like ftp and lynx, but I don't think I would go for that. Some of the systems which I noticed being broken in this way were SunOS 4.x systems, but not all SunOS 4.x systems behave like that. From owner-freebsd-current Sun Jul 27 14:44:11 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA23109 for current-outgoing; Sun, 27 Jul 1997 14:44:11 -0700 (PDT) Received: from lsd.relcom.eu.net (lsd.relcom.eu.net [193.124.23.23]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA23104 for ; Sun, 27 Jul 1997 14:44:07 -0700 (PDT) Received: (from ache@localhost) by lsd.relcom.eu.net (8.8.6/8.8.5) id BAA28053; Mon, 28 Jul 1997 01:43:31 +0400 (MSD) Date: Mon, 28 Jul 1997 01:43:30 +0400 (MSD) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= X-Sender: ache@lsd.relcom.eu.net To: Wolfgang Helbig cc: FreeBSD-current@FreeBSD.ORG Subject: Re: Sound Blaster PnP in -current In-Reply-To: <199707272027.WAA00364@helbig.informatik.ba-stuttgart.de> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 27 Jul 1997, Wolfgang Helbig wrote: > So I am wondering if anyone got it to work under -current. I have SB 32 PnP working with -current (if it helps), my config is the same as in LINT. -- Andrey A. Chernov http://www.nagual.pp.ru/~ache/ From owner-freebsd-current Sun Jul 27 17:53:06 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA00810 for current-outgoing; Sun, 27 Jul 1997 17:53:06 -0700 (PDT) Received: from rocky.mt.sri.com (rocky.mt.sri.com [206.127.76.100]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA00805 for ; Sun, 27 Jul 1997 17:53:03 -0700 (PDT) Received: (from nate@localhost) by rocky.mt.sri.com (8.7.5/8.7.3) id SAA02793; Sun, 27 Jul 1997 18:52:56 -0600 (MDT) Date: Sun, 27 Jul 1997 18:52:56 -0600 (MDT) Message-Id: <199707280052.SAA02793@rocky.mt.sri.com> From: Nate Williams MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Poul-Henning Kamp Cc: "Jordan K. Hubbard" , hosokawa@jp.freebsd.org, current@freebsd.org Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <6019.869903071@critter.dk.tfs.com> References: <20225.869879961@time.cdrom.com> <6019.869903071@critter.dk.tfs.com> X-Mailer: VM 6.29 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > right back then, I can confidently say that to get our "removable > device" support into a decent shape will take some work. > > The Nomads work are pretty focused on pccards, but in the same framework > we should look at CardBus, docking stations, hot-plug PCI and goddamnit: > PlugAndPray since this is the mechanism that is supposed to make support > for all these things managable. > > As usual, if you start to fiddle this stuff, you hit your head on the > usual suspects: > > 1. Improved support for loadable devices. > 2. Real mode BIOS calls. > 3. A Registry. > 4. (optional:) pageable kernel modules. You forgot: 5. Device drivers that actually 'give up' resources when not in use, such as interrupts, ports, etc... You can't currently 'unload' an IRQ right now safely. Nate From owner-freebsd-current Sun Jul 27 19:13:02 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA03852 for current-outgoing; Sun, 27 Jul 1997 19:13:02 -0700 (PDT) Received: from rah.star-gate.com (rah.star-gate.com [204.188.121.18]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA03842 for ; Sun, 27 Jul 1997 19:12:58 -0700 (PDT) Received: from rah.star-gate.com (localhost.star-gate.com [127.0.0.1]) by rah.star-gate.com (8.8.5/8.8.5) with ESMTP id TAA01171; Sun, 27 Jul 1997 19:12:40 -0700 (PDT) Message-Id: <199707280212.TAA01171@rah.star-gate.com> X-Mailer: exmh version 2.0gamma 1/27/96 To: Wolfgang Helbig cc: FreeBSD-current@FreeBSD.ORG Subject: Re: Sound Blaster PnP in -current In-reply-to: Your message of "Sun, 27 Jul 1997 22:27:39 +0200." <199707272027.WAA00364@helbig.informatik.ba-stuttgart.de> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 27 Jul 1997 19:12:40 -0700 From: Amancio Hasty Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Subscribe to the multimedia mailing list: mail majordomo@freebsd.org subscribe multimedia ---- The latest sound driver for -current is available from: ftp://rah.star-gate.com/pub/guspnp12.tar.gz It supports the SB16 PnP thingies among many cards. We are in the process of stabilizing the Sound Driver Version 3.5 for cards other than the gus pnp pro and work has begun on a major rewrite for the sound driver. Luigi is leading the effort on the new sound driver and I am trying to buy some time by cleaning up the 3.5 version (guspnp series of sound drivers). The current sound driver on -current will be obsoleted soon. Amancio >From The Desk Of Wolfgang Helbig : > Hi, > > I installed Sound Blaster 16 PnP. It is not recognized by my kernel. > Rumors have it that Sound Blaster 16 PnP works well in 2.2.x. > > So I am wondering if anyone got it to work under -current. > > This is /var/log/messages: > Jul 27 21:31:22 helbig /kernel: npx0: INT 16 interface > Jul 27 21:31:22 helbig /kernel: sb0 not found at 0x220 > Jul 27 21:31:22 helbig /kernel: sbxvi0 not found > Jul 27 21:31:22 helbig /kernel: sbmidi0 not found at 0x330 > Jul 27 21:31:22 helbig /kernel: opl0 not found at 0x388 > > Strange thing is these messages are printed after the npx0-line, > which used to be the last one. > > This is part of the kernel configuration: (straight from LINT) > # Sound support > controller snd0 > device sb0 at isa? port 0x220 irq 5 drq 1 vector sbintr > device sbxvi0 at isa? drq 5 > device sbmidi0 at isa? port 0x330 > device opl0 at isa? port 0x388 > > Finally I changed the port address of the midi driver in sound_config.h > from 0x300 to LINT's 0x330. > > All these values are factory default, and there are no conflicts > with other devices as to my knowledge. > > Any ideas? > > Thanks > Wolfgang From owner-freebsd-current Sun Jul 27 19:21:08 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA04217 for current-outgoing; Sun, 27 Jul 1997 19:21:08 -0700 (PDT) Received: from dfw-ix5.ix.netcom.com (dfw-ix5.ix.netcom.com [206.214.98.5]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA04198 for ; Sun, 27 Jul 1997 19:20:57 -0700 (PDT) Received: (from smap@localhost) by dfw-ix5.ix.netcom.com (8.8.4/8.8.4) id VAA08077; Sun, 27 Jul 1997 21:20:24 -0500 (CDT) Received: from scz-ca5-19.ix.netcom.com(199.182.129.179) by dfw-ix5.ix.netcom.com via smap (V1.3) id sma007848; Sun Jul 27 21:19:31 1997 Message-ID: <33DC01B5.167EB0E7@ix.netcom.com> Date: Sun, 27 Jul 1997 19:19:33 -0700 From: Sherwin X-Mailer: Mozilla 3.01Gold (X11; U; FreeBSD 2.2.2-RELEASE i386) MIME-Version: 1.0 To: current@FreeBSD.ORG Subject: Multiple FreeBSD Systems on a Single Disk Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk I installed 2.2.2 in the first partition of my disk. It was previously a Linux partition. I have 2.2.1 installed in the second partition. It is all still there, I can mount it and access it, but I do not know how to boot it. Boot-Easy gives the same result for both F1 and F2, it boots the 2.2.2 partition. While there is no great harm, if I can no longer run 2.2.1, I would love to find a solution, so that I can experiment with 3.0 Snap, and still have 2.2.2 to fall back on. I have searched the documentation for an answer, particularily NextBoot, but nothing seems to explain this behavior. Please ket me know what I am overlooking. Thanks :-) From owner-freebsd-current Mon Jul 28 01:13:24 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA17351 for current-outgoing; Mon, 28 Jul 1997 01:13:24 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA17344 for ; Mon, 28 Jul 1997 01:13:19 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id RAA05543; Mon, 28 Jul 1997 17:36:59 +0930 (CST) From: Michael Smith Message-Id: <199707280806.RAA05543@genesis.atrad.adelaide.edu.au> Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <25979.869936845@time.cdrom.com> from "Jordan K. Hubbard" at "Jul 26, 97 10:07:25 am" To: jkh@time.cdrom.com (Jordan K. Hubbard) Date: Mon, 28 Jul 1997 17:36:58 +0930 (CST) Cc: msmith@atrad.adelaide.edu.au, nate@mt.sri.com, hosokawa@jp.freebsd.org, current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Jordan K. Hubbard stands accused of saying: > > Because the current startup code is _too_stupid_ to deal with > > transient interfaces. This was hashed out ages ago. > > > > If we accept your model where the system can't start until a network > > card is inserted to match the ifconfig_foo lines, I'd be screwed if I > > decided to boot without my 'net card to save power. > > Well, then we have a problem which really needs resolution since it's > not going to work well the way it is with the current installation > code. Understood. > I've also no objection to changing things so that, somehow, a pccard > device is somehow recognised and the behavior of adding ifconfig_foo > lines is skipped in favor of adding more specialized configuration, > but I think that we need to seriously think about our configuration > mechanisms and how "dynamism" in your device picture is going to be > handled. For a device which may enter and leave the picture at any > time, you obviously may wish to treat its configuration data > differently. The question is: How exactly? :) I think that a clear separation has to be made between the configuration information and the possible process(es) that may use this information. The rc.conf idea helps this immensely. My "vision" for looks something like this : - configuration information is kept "somewhere" (rc.conf for now) - The parameters include a list of interface names to be looked for, and corresponding address/routing parameters associated with these interfaces. - a script exists which, given an interface name, consumes the parameters for said interface and configures/deconfigures it. - at startup, the list of configured-for interfaces is scanned, and any such interfaces present are configured using the script above. ie. at bootup, fixed interfaces have effectively "arrived". - upon PCCARD arrival/departure, the card's identifier is linked to a driver name (as is currently done). The PCCARD arrival/departure handler can then use the same script as above to manage arrival and departure, using the same parameters as for fixed interfaces. - one could go one further and specify type-like wildcards for interface congfiguration, eg. "ether", so that arrival of _any_ ethernet card would result in it being configured without having to stipulate the exact driver. Yah, I know I should just get out and do it. Gimme something with PCCARDS in it that I can call "mine" 8) 8) 8) > Jordan -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Mon Jul 28 02:01:40 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id CAA19354 for current-outgoing; Mon, 28 Jul 1997 02:01:40 -0700 (PDT) Received: from karon.dynas.se (karon.dynas.se [192.71.43.4]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id CAA19348 for ; Mon, 28 Jul 1997 02:01:34 -0700 (PDT) Received: (qmail 13529 invoked from network); 28 Jul 1997 09:01:31 -0000 Received: from spirit.dynas.se (HELO host.domain) (172.16.1.10) by karon.dynas.se with SMTP; 28 Jul 1997 09:01:31 -0000 Received: by spirit (Smail3.1.28.1 #32) id m0wslgI-000JelC; Mon, 28 Jul 97 11:01:30 +0200 Date: Mon, 28 Jul 1997 11:01:30 +0200 (MET DST) From: Mikael Hybsch Reply-To: Mikael Hybsch To: freebsd-current@freebsd.org Subject: aha0: DMA beyond end Of ISA Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I have a 486 with 24M memory and a 1542B connected to a HP 4020i. Frequently when I try to access the CD or burn a new CD, I get the message aha0: DMA beyond end Of ISA: 0x1154d24 or other addresses above 0x1000000. By repeating the mount command it will eventually succeed. I'm running 3.0-current as of some days ago and have the BOUNCE_BUFFERS option in the kernel, so what could be causing this? -- Mikael Hybsch Email: micke@dynas.se DynaSoft, Dynamic Software AB Phone: +46-8-7250900 Box 10704 Fax: +46-8-6494970 S-121 29 STOCKHOLM, SWEDEN From owner-freebsd-current Mon Jul 28 05:46:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA01185 for current-outgoing; Mon, 28 Jul 1997 05:46:34 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA01145 for ; Mon, 28 Jul 1997 05:46:23 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id WAA17057; Mon, 28 Jul 1997 22:40:24 +1000 Date: Mon, 28 Jul 1997 22:40:24 +1000 From: Bruce Evans Message-Id: <199707281240.WAA17057@godzilla.zeta.org.au> To: current@FreeBSD.ORG, sherwink@ix.netcom.com Subject: Re: Multiple FreeBSD Systems on a Single Disk Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >I installed 2.2.2 in the first partition of my disk. It was previously a >Linux partition. I have 2.2.1 installed in the second partition. It is >all still there, I can mount it and access it, but I do not know how to >boot it. Boot-Easy gives the same result for both F1 and F2, it boots >the 2.2.2 partition. Only booting from the first FreeBSD partition (slice) is supported. The second one can be booted from by making it the first, e.g., by changing the partition type of the previous first one to something other than 0xa5. Multiple versions or backup versions of FreeBSD are better handled by putting a small root file partition for each in the same (first) slice. It's easiest to have this slice on the first drive. 6 or 7 versions can easily be handled like this if you have enough slices (perhaps on other drives) to hold all the different usr and var partitions. Bruce From owner-freebsd-current Mon Jul 28 07:29:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA09209 for current-outgoing; Mon, 28 Jul 1997 07:29:18 -0700 (PDT) Received: from korin.warman.org.pl (korin.warman.org.pl [148.81.160.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA09195 for ; Mon, 28 Jul 1997 07:29:02 -0700 (PDT) Received: from localhost (abial@localhost) by korin.warman.org.pl (8.8.6/8.8.5) with SMTP id QAA23282; Mon, 28 Jul 1997 16:30:23 +0200 (CEST) Date: Mon, 28 Jul 1997 16:30:23 +0200 (CEST) From: Andrzej Bialecki To: Bruce Evans cc: current@FreeBSD.ORG, sherwink@ix.netcom.com Subject: Re: Multiple FreeBSD Systems on a Single Disk In-Reply-To: <199707281240.WAA17057@godzilla.zeta.org.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Mon, 28 Jul 1997, Bruce Evans wrote: > >I installed 2.2.2 in the first partition of my disk. It was previously a > >Linux partition. I have 2.2.1 installed in the second partition. It is > >all still there, I can mount it and access it, but I do not know how to > >boot it. Boot-Easy gives the same result for both F1 and F2, it boots > >the 2.2.2 partition. > > Only booting from the first FreeBSD partition (slice) is supported. > The second one can be booted from by making it the first, e.g., Even simpler: you can install OS-BS bootloader (from /tools), which is much more configurable than boot-easy. Sincerely yours, --- Andrzej Bialecki FreeBSD: Turning PCs Into Workstations http://www.freebsd.org Research and Academic Network in Poland From owner-freebsd-current Mon Jul 28 08:33:15 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id IAA12664 for current-outgoing; Mon, 28 Jul 1997 08:33:15 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA12657 for ; Mon, 28 Jul 1997 08:33:11 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id BAA23105; Tue, 29 Jul 1997 01:28:27 +1000 Date: Tue, 29 Jul 1997 01:28:27 +1000 From: Bruce Evans Message-Id: <199707281528.BAA23105@godzilla.zeta.org.au> To: abial@korin.warman.org.pl, bde@zeta.org.au Subject: Re: Multiple FreeBSD Systems on a Single Disk Cc: current@FreeBSD.ORG, sherwink@ix.netcom.com Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> Only booting from the first FreeBSD partition (slice) is supported. >> The second one can be booted from by making it the first, e.g., > >Even simpler: you can install OS-BS bootloader (from /tools), which is >much more configurable than boot-easy. The lack of support has nothing to do with the master boot loader. The FreeBSD boot loader only supports booting from the first FreeBSD slice (on any disk that can be accessed by the BIOS). Bruce From owner-freebsd-current Mon Jul 28 10:26:27 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA19263 for current-outgoing; Mon, 28 Jul 1997 10:26:27 -0700 (PDT) Received: from bmccane.uit.net (bmccane.uit.net [208.129.189.48]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA19252 for ; Mon, 28 Jul 1997 10:26:13 -0700 (PDT) Received: from bmccane.uit.net (localhost.mccane.com [127.0.0.1]) by bmccane.uit.net (8.8.5/8.8.5) with ESMTP id MAA11701; Mon, 28 Jul 1997 12:25:34 -0500 (CDT) Message-Id: <199707281725.MAA11701@bmccane.uit.net> X-Mailer: exmh version 2.0gamma 1/27/96 To: Garrett Wollman cc: Lars Eggert , freebsd-current@FreeBSD.ORG Subject: Re: Who is working on the TCP stack? In-reply-to: Your message of "Sat, 26 Jul 1997 15:16:30 EDT." <199707261916.PAA23440@khavrinen.lcs.mit.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 28 Jul 1997 12:25:33 -0500 From: Wm Brian McCane Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > < said: > > > and sorry if this not the right place to ask. I will be implementing some > > changes to TCP as part of a research project during the fall (if you are > > interested, see RFC 2140, "TCP Control Block Interdependence." J. Touch. > > April 1997.) > > > I was wondering, if anybody else is working on the TCP stack at this time. > > Would -current be the right place to implement those changes, or should I > > stick with -stable? Is there anything else I should be aware of? > > There are a group of people who have been working on and off on > various network issues, of which I am a member. The people in the > CAIRN consortium have been doing a lot of work as well, but since they > don't talk much to me I don't know what they are up to. I don't think > that anyone is working on TCP in particular at this point (I could be > wrong); certainly if I had the time I would like to, but I haven't had > any time to spare for the past couple of months, and when I have had > the time, -current has been broken (sigh). > > Certainly current is the right platform to target; there have been and > will continue to be substantial ongoing changes in the basic > substructure of the networking code which make it unlikely that any > efforts based on 2.2 will easily carry into the main line of > development. > > If you're up for a major re-working of the TCP implementation, let me > know and I'll give you a few notions of the way we're headed. > > -GAWollman > > -- > Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same > wollman@lcs.mit.edu | O Siem / The fires of freedom > Opinions not those of| Dance in the burning flame > MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick Somewhat related.... Has anyone checked into the reported incompatiblities with MS Winsock? I read in some 'Zine (Lan Times?) that Microsoft did not follow the RFC, and that Sun has chosen to accept and work with this incompatibility rather than wait for Microsoft to correct their software. From what I read, the implementation bug causes performance problems with non-MS Web servers (and as such was probably on purpose, it wouldn't be the first time). Anyway, I was just wondering if FreeBSD is affected and/or if it has already been addressed. brian BTW> The possibility of collusion is my own opinion, not from the article. From owner-freebsd-current Mon Jul 28 10:29:39 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA19574 for current-outgoing; Mon, 28 Jul 1997 10:29:39 -0700 (PDT) Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA19558 for ; Mon, 28 Jul 1997 10:29:34 -0700 (PDT) Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.8.5/8.7.3) id KAA18261; Mon, 28 Jul 1997 10:29:04 -0700 (PDT) From: "Rodney W. Grimes" Message-Id: <199707281729.KAA18261@GndRsh.aac.dev.com> Subject: Re: Multiple FreeBSD Systems on a Single Disk In-Reply-To: <199707281528.BAA23105@godzilla.zeta.org.au> from Bruce Evans at "Jul 29, 97 01:28:27 am" To: bde@zeta.org.au (Bruce Evans) Date: Mon, 28 Jul 1997 10:29:04 -0700 (PDT) Cc: abial@korin.warman.org.pl, bde@zeta.org.au, current@FreeBSD.ORG, sherwink@ix.netcom.com X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > >> Only booting from the first FreeBSD partition (slice) is supported. > >> The second one can be booted from by making it the first, e.g., > > > >Even simpler: you can install OS-BS bootloader (from /tools), which is > >much more configurable than boot-easy. > > The lack of support has nothing to do with the master boot loader. > The FreeBSD boot loader only supports booting from the first FreeBSD > slice (on any disk that can be accessed by the BIOS). And should be fixed (I've mentioned this before) to boot from the first ACTIVE (flag 0x80) FreeBSD slice. Should be a trival amount of code, and would do what everyone else does, even Microsoft got that part right. I had patches for this once, but they got lost when one of the major changes to the boot code happened. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation, Inc. Reliable computers for FreeBSD From owner-freebsd-current Mon Jul 28 10:47:53 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA20576 for current-outgoing; Mon, 28 Jul 1997 10:47:53 -0700 (PDT) Received: from rah.star-gate.com (rah.star-gate.com [204.188.121.18]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA20568 for ; Mon, 28 Jul 1997 10:47:47 -0700 (PDT) Received: from rah.star-gate.com (localhost.star-gate.com [127.0.0.1]) by rah.star-gate.com (8.8.5/8.8.5) with ESMTP id KAA11373; Mon, 28 Jul 1997 10:47:43 -0700 (PDT) Message-Id: <199707281747.KAA11373@rah.star-gate.com> X-Mailer: exmh version 2.0gamma 1/27/96 To: lamaster@george.arc.nasa.gov cc: current@freebsd.org Subject: Re: Your message of "Fri, 25 Jul 1997 17:53:59 PDT In-reply-to: Your message of "Mon, 28 Jul 1997 10:33:22 PDT." <199707281733.KAA13863@george.arc.nasa.gov> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Mon, 28 Jul 1997 10:47:42 -0700 From: Amancio Hasty Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Well, we used to have an mrouted capable kernel by default however it got taken out for some reason... Perhaps, someone from the FreeBSD core team care to explain? Regards, Amancio >From The Desk Of lamaster@george.arc.nasa.gov : > > > My statement was in reference to NT's ip stack not being capable of > > supporting mrouted so the next logical thing is to get a > > tcp/ip stack which can support mrouted . > > I see what you mean now. I have no idea if alternative stacks > are available as they were for Windows. It seem unlikely, > because TCP/IP is supported by default, and isn't considered > an extra layer as it was under Windows 3.1, requiring the "Winsock" > solution. > > Also, as Steve Casner said, the real question is whether or not > the OS/stack combination supports multicast forwarding, not just > multicast. > > I haven't done a survey, but, AFAIK, the only system that supports > forwarding *by default* is Irix. Even with FreeBSD, you have > to recompile the kernel that way, and, with Solaris, you have > to patch the default release with one or two recompiled modules. > > SGI may have its problems, but, it is cool that they have been > in multicast for such a long time. > > > Regards, > Hugh LaMaster > > From owner-freebsd-current Mon Jul 28 10:52:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA20851 for current-outgoing; Mon, 28 Jul 1997 10:52:34 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id KAA20846 for ; Mon, 28 Jul 1997 10:52:31 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id KAA01355; Mon, 28 Jul 1997 10:48:56 -0700 From: Terry Lambert Message-Id: <199707281748.KAA01355@phaeton.artisoft.com> Subject: Re: pccard and -current; a long way to go. :-( To: msmith@atrad.adelaide.edu.au (Michael Smith) Date: Mon, 28 Jul 1997 10:48:56 -0700 (MST) Cc: jkh@time.cdrom.com, msmith@atrad.adelaide.edu.au, nate@mt.sri.com, hosokawa@jp.freebsd.org, current@FreeBSD.ORG In-Reply-To: <199707280806.RAA05543@genesis.atrad.adelaide.edu.au> from "Michael Smith" at Jul 28, 97 05:36:58 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > My "vision" for looks something like this : [ ... ] Don't forget: - A "card services" utility to "preunload" cards; many cards are not happy with an unplug (the MS docuemnts and Intel documents go over this in great, gory detail). > Yah, I know I should just get out and do it. Gimme something with > PCCARDS in it that I can call "mine" 8) 8) 8) I thought you had a Multia? My Multia have to PCMCIA slots... Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Mon Jul 28 11:56:31 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA24695 for current-outgoing; Mon, 28 Jul 1997 11:56:31 -0700 (PDT) Received: from rocky.mt.sri.com (rocky.mt.sri.com [206.127.76.100]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA24690 for ; Mon, 28 Jul 1997 11:56:28 -0700 (PDT) Received: (from nate@localhost) by rocky.mt.sri.com (8.7.5/8.7.3) id MAA06433; Mon, 28 Jul 1997 12:50:47 -0600 (MDT) Date: Mon, 28 Jul 1997 12:50:47 -0600 (MDT) Message-Id: <199707281850.MAA06433@rocky.mt.sri.com> From: Nate Williams MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Terry Lambert Cc: msmith@atrad.adelaide.edu.au (Michael Smith), jkh@time.cdrom.com, nate@mt.sri.com, hosokawa@jp.freebsd.org, current@freebsd.org Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707281748.KAA01355@phaeton.artisoft.com> References: <199707280806.RAA05543@genesis.atrad.adelaide.edu.au> <199707281748.KAA01355@phaeton.artisoft.com> X-Mailer: VM 6.29 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > Don't forget: > > - A "card services" utility to "preunload" cards; many cards are not > happy with an unplug (the MS docuemnts and Intel documents go > over this in great, gory detail). Yeah, but in reality Win95 'handles' most every card I've thrown at it fine, although it whines alot. In any case, this is already done by the Nomads (it's an X appliation). Nate From owner-freebsd-current Mon Jul 28 14:09:27 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA03227 for current-outgoing; Mon, 28 Jul 1997 14:09:27 -0700 (PDT) Received: from mercury.Sun.COM (mercury.Sun.COM [192.9.25.1]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id OAA03178; Mon, 28 Jul 1997 14:09:09 -0700 (PDT) Received: from East.Sun.COM ([129.148.1.241]) by mercury.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id OAA02357; Mon, 28 Jul 1997 14:08:33 -0700 Received: from suneast.East.Sun.COM by East.Sun.COM (SMI-8.6/SMI-5.3) id RAA22092; Mon, 28 Jul 1997 17:08:35 -0400 Received: from compound.east.sun.com by suneast.East.Sun.COM (SMI-8.6/SMI-SVR4) id RAA13058; Mon, 28 Jul 1997 17:08:35 -0400 Received: (from alk@localhost) by compound.east.sun.com (8.8.6/8.7.3) id QAA17786; Mon, 28 Jul 1997 16:09:40 -0500 (CDT) Date: Mon, 28 Jul 1997 16:09:40 -0500 (CDT) Reply-To: Anthony.Kimball@East.Sun.COM Message-Id: <199707282109.QAA17786@compound.east.sun.com> From: Tony Kimball MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: ahd@kew.com Cc: chat@FreeBSD.ORG Subject: bouncing mail from sites without a valid MX/A record References: <199707260544.BAA02854@pandora.hh.kew.com> X-Face: O9M"E%K;(f-Go/XDxL+pCxI5*gr[=FN@Y`cl1.Tn Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk [Again, I redirect to chat.] Quoth Drew Derbyshire on Sat, 26 July: : : Actually, this nukes about ~ 20 - 60 % of the SPAM off the top. : Sites don't like their good name used by spammers, so many SPAM : generators just generate random all number domains in .COM. : Hmm. I've received about 198 spam messages in the past 3 months. I don't find *any* tainted by bogus domains. Nary a one. In fact, they all have valid MX records deducible from the headers, although in many cases there is no identifiable mailbox at the corresponding smtp host. The top three sources account for 75% of all spam recieved, and they are Cyber Promotions, Juno, and HotMail. My sample may not be statistically representative, but it is at least *reality-based*. I'd like to learn of more broadly representative studies. From owner-freebsd-current Mon Jul 28 14:17:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA03576 for current-outgoing; Mon, 28 Jul 1997 14:17:03 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id OAA03570 for ; Mon, 28 Jul 1997 14:17:01 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id OAA01641; Mon, 28 Jul 1997 14:14:21 -0700 From: Terry Lambert Message-Id: <199707282114.OAA01641@phaeton.artisoft.com> Subject: Re: Who is working on the TCP stack? To: root@bmccane.uit.net (Wm Brian McCane) Date: Mon, 28 Jul 1997 14:14:21 -0700 (MST) Cc: wollman@khavrinen.lcs.mit.edu, larse@usc.edu, freebsd-current@FreeBSD.ORG In-Reply-To: <199707281725.MAA11701@bmccane.uit.net> from "Wm Brian McCane" at Jul 28, 97 12:25:33 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Has anyone checked into the reported incompatiblities with MS > Winsock? I read in some 'Zine (Lan Times?) that Microsoft did > not follow the RFC, and that Sun has chosen to accept and work > with this incompatibility rather than wait for Microsoft to > correct their software. From what I read, the implementation > bug causes performance problems with non-MS Web servers (and > as such was probably on purpose, it wouldn't be the first > time). Anyway, I was just wondering if FreeBSD is affected > and/or if it has already been addressed. The problem is that the client applications, specifically the nono-Internet Explorer applications, do not call the WinSock function "shutdown( s, 3);" and then "recv( s, ...);" all the packets until the remote side close the connection. If the WinSock client does not call "shutdown( s, [2,3]);", then the PC will not send a FIN to the server. Note that there is some bias in the specification here for Microsoft WinSock implementations; specifically, if you call "shutdown( s, 3);", you can't also call "shutdown( s, 2);", and a number of non-Microsoft WinSock implementations will not correctly close down the connection (or send the FIN) with the "3" parameter. This basically means that if you use a non-Microsoft browser or client software, where they originators were primarily UNIX geeks (like NetScape), or if you use a non-Microsoft proxy server or WinSock implementation that doesn't act correctly on a "3" parameter and a client that uses a "3" parameter instead of a "2" parameter, then you are screwed. The Sun response was to implement a timeout in "FIN WAIT", just as there is a timeout in "FIN WAIT" in FreeBSD. This is not really a big deal, and it's not a failure to implement the TCP RFC, it's a failure of the client to trigger TCP RFC compliant behaviour. Clearly, the MS web servers run on MS OS's, and expect this common bug (and deal with it at the server stack level). As to whether FreeBSD is affected: it was, but it has timeouts (controllable via sysctl) now, so it's not an issue (the same soloution that Sun implemented). As far as the implication of collusion: I think it's more an artifact of the WinSock code being implemented as a DLL than any intent to make it screwed up: DLL's are an inherently screwed up architecture for implementing OS services, but they were the only viable way to jam socket services into the Windows 3.11 framework, before the MPR implementation in Windows 95 became widespread. Now there's a legacy issue of Windows 3.11 target software to support on Windows95, and the bogosity lives on... References: Windows(tm) Sockets Network Programming Bob Quinn, Dave Shute Addison-Wesley ISBN 0-201-63372-8 Regards, Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Mon Jul 28 14:22:37 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA03939 for current-outgoing; Mon, 28 Jul 1997 14:22:37 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id OAA03921 for ; Mon, 28 Jul 1997 14:22:25 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id OAA01658; Mon, 28 Jul 1997 14:18:29 -0700 From: Terry Lambert Message-Id: <199707282118.OAA01658@phaeton.artisoft.com> Subject: Re: pccard and -current; a long way to go. :-( To: nate@mt.sri.com (Nate Williams) Date: Mon, 28 Jul 1997 14:18:29 -0700 (MST) Cc: terry@lambert.org, msmith@atrad.adelaide.edu.au, jkh@time.cdrom.com, nate@mt.sri.com, hosokawa@jp.freebsd.org, current@freebsd.org In-Reply-To: <199707281850.MAA06433@rocky.mt.sri.com> from "Nate Williams" at Jul 28, 97 12:50:47 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > > Don't forget: > > > > - A "card services" utility to "preunload" cards; many cards are not > > happy with an unplug (the MS docuemnts and Intel documents go > > over this in great, gory detail). > > Yeah, but in reality Win95 'handles' most every card I've thrown at it > fine, although it whines alot. In any case, this is already done by the > Nomads (it's an X appliation). I was more worried about non-network PCMCIA cards. Examples might be on of: o Adaptec SCSI controller PCMCIA card and all attached SCSI devices. o Flash RAM card with mounted FS on it, with some FS data in cache and needing to be flushed. o Mounted hard drives located in the dock, in the case where the machine is about to be undocked. o Etc. The network cards aren't nearly as bad; most the the network code was born being tolerant of intermittent connections. Yeah, I'm paranoid. 8-). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Mon Jul 28 14:49:24 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA05667 for current-outgoing; Mon, 28 Jul 1997 14:49:24 -0700 (PDT) Received: from labs.usn.blaze.net.au (root@labs.usn.blaze.net.au [203.17.53.30]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA05653 for ; Mon, 28 Jul 1997 14:49:11 -0700 (PDT) Received: from labs.usn.blaze.net.au (davidn@local [127.0.0.1]) by labs.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id HAA13318; Tue, 29 Jul 1997 07:48:28 +1000 (EST) Message-Id: <199707282148.HAA13318@labs.usn.blaze.net.au> X-Mailer: exmh version 2.0gamma 1/27/96 To: Wolfgang Helbig cc: FreeBSD-current@FreeBSD.ORG Subject: Re: Sound Blaster PnP in -current In-reply-to: Your message of "Sun, 27 Jul 1997 22:27:39 +0200." <199707272027.WAA00364@helbig.informatik.ba-stuttgart.de> X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Mime-Version: 1.0 Content-Type: multipart/mixed ; boundary="==_Exmh_-1001429240" Date: Tue, 29 Jul 1997 07:48:27 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk This is a multipart MIME message. --==_Exmh_-1001429240 Content-Type: text/plain; charset=us-ascii > I installed Sound Blaster 16 PnP. It is not recognized by my kernel. > Rumors have it that Sound Blaster 16 PnP works well in 2.2.x. The rumours are wrong. > So I am wondering if anyone got it to work under -current. Yes. Method 1: boot dos first and run ctcm. Method 2: use the pnp patch to configure the card (my -current-fied) version of this attached. Apply in /usr/src/sys. You'll have to edit pnp.c to insert your card's settings. As you can see, I'm using 0x220, 0x300 (usually 0x330, but I have an aha1542 at that address), 0x388, and dma 1 7 (usually 5), irq 5. Regards, David --==_Exmh_-1001429240 Content-Type: text/plain ; name="pnp.diff"; charset=us-ascii Content-Description: pnp.diff Content-Disposition: attachment; filename="pnp.diff" diff -ur --new-file OLD/i386/conf/LINT i386/conf/LINT --- OLD/i386/conf/LINT Fri May 3 04:46:25 1996 +++ i386/conf/LINT Sun May 5 22:12:50 1996 @@ -407,6 +407,11 @@ controller isa0 # +# Optional ISA Plug-n-Play device support +# +controller pnp0 + +# # Options for `isa': # # AUTO_EOI_1 enables the `automatic EOI' feature for the master 8259A diff -ur --new-file OLD/i386/conf/files.i386 i386/conf/files.i386 --- OLD/i386/conf/files.i386 Sat May 4 04:59:45 1996 +++ i386/conf/files.i386 Sun May 5 22:13:22 1996 @@ -127,6 +127,7 @@ i386/isa/pcvt/pcvt_out.c optional vt device-driver i386/isa/pcvt/pcvt_sup.c optional vt device-driver i386/isa/pcvt/pcvt_vtf.c optional vt device-driver +i386/isa/pnp.c optional pnp device-driver i386/isa/prof_machdep.c optional profiling-routine i386/isa/psm.c optional psm device-driver i386/isa/qcam.c optional qcam device-driver diff -ur --new-file OLD/i386/i386/autoconf.c i386/i386/autoconf.c --- OLD/i386/i386/autoconf.c Thu May 2 10:54:05 1996 +++ i386/i386/autoconf.c Sun May 5 22:15:04 1996 @@ -65,6 +65,11 @@ #include #endif +#include "pnp.h" +#if NPNP > 0 +#include +#endif + #include "eisa.h" #if NEISA > 0 #include @@ -186,6 +191,10 @@ #if NPCI > 0 pci_configure(); +#endif + +#if NPNP > 0 + pnp_configure(); #endif #if NISA > 0 diff -ur --new-file OLD/i386/isa/pnp.c i386/isa/pnp.c --- sys/i386/isa/pnp.c.orig Tue Jul 22 05:41:20 1997 +++ sys/i386/isa/pnp.c Tue Jul 22 06:09:13 1997 @@ -0,0 +1,290 @@ +/* + * Copyright (c) 1996, Sujal M. Patel + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * $Id: pnpcfg.c,v 1.6 1996/05/06 02:08:25 smpatel Exp smpatel $ + */ + +#include +#include +#include +#include + +#include + + +#define SEND(d, r) { outb (ADDRESS, d); outb (WRITE_DATA, r); } + + +/* The READ_DATA port that we are using currently */ +static int rd_port; + + +/* + * Hard coded for now -- Will need to use resource information from + * ISA, PCI, and EISA drivers to auto-configure the PnP-devices + */ +static struct cinfo cinfo[] = +{ + /* Configuration for a Creative Labs Soundblaster16 */ + { + 0x00006e24, /* Serial Number */ + -1, /* Logical Device Number */ + { + { 5, -1 }, /* Primary IRQ Number, Type */ + { -1, -1 } /* Second IRQ Number, Type */ + }, + { 1, 7 }, /* DRQ Number */ + { + 0x220, /* Ports 1 */ + 0x300, /* Ports 2 */ + 0x388, /* Ports 3 */ + -1, /* Ports 4 */ + -1, /* Ports 5 */ + -1, /* Ports 6 */ + -1, /* Ports 7 */ + -1, /* Ports 8 */ + }, + { + { -1, -1, -1 }, /* Memory desc0 - base, ctrl, range */ + { -1, -1, -1 }, /* Memory desc1 - base, ctrl, range */ + { -1, -1, -1 }, /* Memory desc2 - base, ctrl, range */ + { -1, -1, -1 } /* Memory desc3 - base, ctrl, range */ + } + }, + +}; + + +int power __P((int base, int exp)); +void send_Initiation_LFSR __P((void)); +int get_serial __P((unsigned char *data)); +void config_device __P((unsigned char *data, int csn)); +int isolation_protocol __P((void)); + + +int +power(base, exp) + int base, exp; +{ + if (exp <= 1) + return base; + else + return base * power(base, exp - 1); +} + + +/* + * Send Initiation LFSR as described in "Plug and Play ISA Specification, + * Intel May 94." + */ +void +send_Initiation_LFSR() +{ + int cur, i; + + /* Reset the LSFR */ + outb(ADDRESS, 0); + outb(ADDRESS, 0); + + cur = 0x6a; + outb(ADDRESS, cur); + + for (i = 1; i < 32; i++) { + cur = (cur >> 1) | (((cur ^ (cur >> 1)) << 7) & 0xff); + outb(ADDRESS, cur); + } +} + + +/* + * Get the device's serial number. Returns 1 if the serial is valid. + */ +int +get_serial(data) + unsigned char *data; +{ + int i, bit, valid = 0, sum = 0x6a; + + bzero(data, sizeof(char) * 9); + + for (i = 0; i < 72; i++) { + bit = inb((rd_port << 2) | 0x3) == 0x55; + DELAY(250); /* Delay 250 usec */ + + /* Can't Short Circuit the next evaluation, so 'and' is last */ + bit = (inb((rd_port << 2) | 0x3) == 0xaa) && bit; + DELAY(250); /* Delay 250 usec */ + + valid = valid || bit; + + if (i < 64) + sum = (sum >> 1) | + (((sum ^ (sum >> 1) ^ bit) << 7) & 0xff); + + data[i / 8] = (data[i / 8] >> 1) | (bit ? 0x80 : 0); + } + + valid = valid && (data[8] == sum); + + return valid; +} + + +/* + * Configure PnP devices + */ +void +config_device(data, csn) + unsigned char *data; + int csn; +{ + struct cinfo *ci; + int i; + + if (bootverbose) + printf("Card assigned CSN #%d\n", csn); + + printf("Board Vendor ID: %c%c%c%02x%02x", + ((data[0] & 0x7c) >> 2) + 64, + (((data[0] & 0x03) << 3) | ((data[1] & 0xe0) >> 5)) + 64, + (data[1] & 0x1f) + 64, data[2], data[3]); + printf(" Board Serial Number: %08x\n", *(int *)&(data[4])); + + SEND(SET_CSN, csn); /* Move this out of this function XXX */ + outb(ADDRESS, STATUS); + + for (ci = cinfo; (int)ci < ((int)cinfo + sizeof (cinfo)); ci++) { + if (ci->serial == *(int *)&(data[4])) { + printf (" Configuring (Logical Device %x)\n", + ci->ldn != -1 ? ci->ldn : 0); + + if (ci->ldn >= 0) + SEND (SET_LDN, ci->ldn); + + for (i = 0; i < 8; i++) + if (ci->port[i] > 0) { + SEND (IO_CONFIG_BASE + i * 2, + ci->port[i] >> 8); + SEND (IO_CONFIG_BASE + i * 2 + 1, + ci->port[i] & 0xff); + } + + for (i = 0; i < 2; i++) + if (ci->irq[i].num > 0) { + SEND (IRQ_CONFIG + i * 2, + ci->irq[i].num); + if (ci->irq[i].type >= 0) + SEND (IRQ_CONFIG + i * 2 + 1, + ci->irq[i].type); + } + + for (i = 0; i < 2; i++) + if (ci->drq[i] > 0) { + SEND (DRQ_CONFIG + i, + ci->drq[i]); + } + + for (i = 0; i < 4; i++) + if (ci->mem[i].base > 0) { + SEND (MEM_CONFIG + i * 8, + ci->mem[i].base >> 16); + SEND (MEM_CONFIG + i * 8 + 1, + (ci->mem[i].base >> 8) & + 0xff); + /* + * This needs to be handled better for + * the user's sake. XXX + */ + if (ci->mem[i].control >= 0) { + SEND (MEM_CONFIG + i * 8 + 2, + ci->mem[i].control); + } + SEND (MEM_CONFIG + i * 8 + 3, + ci->mem[i].range >> 16); + SEND (MEM_CONFIG + i * 8 + 4, + (ci->mem[i].range >> 8) & + 0xff); + } + SEND (IO_RANGE_CHECK, 0); + SEND (ACTIVATE, 1); + } + } +} + + +/* + * Run the isolation protocol. Use rd_port as the READ_DATA port value (caller + * should try multiple READ_DATA locations before giving up). Upon exiting, + * all cards are aware that they should use rd_port as the READ_DATA port; + */ +int +isolation_protocol() +{ + int csn; + unsigned char data[9]; + + send_Initiation_LFSR(); + + /* Reset CSN for All Cards */ + SEND(0x02, 0x04); + + for (csn = 1; (csn < MAX_CARDS); csn++) { + /* Wake up cards without a CSN */ + SEND(WAKE, 0); + SEND(SET_RD_DATA, rd_port); + outb(ADDRESS, SERIAL_ISOLATION); + DELAY(1000); /* Delay 1 msec */ + + if (get_serial(data)) + config_device(data, csn); + else + break; + } + return csn - 1; +} + + +void +pnp_configure() +{ + int num_pnp_devs; + + printf("Checking for Plug-n-Play devices...\n"); + + /* Try various READ_DATA ports from 0x203-0x3ff */ + for (rd_port = 0x80; (rd_port < 0xff); rd_port += 0x10) { + if (bootverbose) + printf("Trying Read_Port at %x\n", + (rd_port << 2) | 0x3); + + num_pnp_devs = isolation_protocol(); + if (num_pnp_devs) + break; + } + if (!num_pnp_devs) { + printf("No Plug-n-Play devices were found\n"); + return; + } +} --- sys/i386/isa/pnp.h.orig Tue Jul 22 05:41:20 1997 +++ sys/i386/isa/pnp.h Tue Jul 22 05:41:20 1997 @@ -0,0 +1,104 @@ +/* + * Copyright (c) 1996, Sujal M. Patel + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by Sujal M. Patel + * 4. Neither the name of the author nor the names of any co-contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * $Id: pnpcfg.h,v 1.2 1996/01/12 21:35:10 smpatel Exp smpatel $ + */ + +#ifndef _I386_ISA_PNP_H_ +#define _I386_ISA_PNP_H_ + +/* Maximum Number of PnP Devices. 8 should be plenty */ +#define MAX_CARDS 8 + + +/* Static ports */ +#define ADDRESS 0x279 +#define WRITE_DATA 0xa79 + + +/* PnP Registers. Write to ADDRESS and then use WRITE/READ_DATA */ +#define SET_RD_DATA 0x00 +#define SERIAL_ISOLATION 0x01 +#define WAKE 0x03 +#define RESOURCE_DATA 0x04 +#define STATUS 0x05 +#define SET_CSN 0x06 +#define SET_LDN 0x07 +#define ACTIVATE 0x30 +#define IO_RANGE_CHECK 0x31 +#define MEM_CONFIG 0x40 +#define IO_CONFIG_BASE 0x60 +#define IRQ_CONFIG 0x70 +#define DRQ_CONFIG 0x74 + +/* Small Resource Item names */ +#define PNP_VERSION 0x1 +#define LOG_DEVICE_ID 0x2 +#define COMP_DEVICE_ID 0x3 +#define IRQ_FORMAT 0x4 +#define DMA_FORMAT 0x5 +#define START_DEPEND_FUNC 0x6 +#define END_DEPEND_FUNC 0x7 +#define IO_PORT_DESC 0x8 +#define FIXED_IO_PORT_DESC 0x9 +#define SM_RES_RESERVED 0xa-0xd +#define SM_VENDOR_DEFINED 0xe +#define END_TAG 0xf + +/* Large Resource Item names */ +#define MEMORY_RANGE_DESC 0x1 +#define ID_STRING_ANSI 0x2 +#define ID_STRING_UNICODE 0x3 +#define LG_VENDOR_DEFINED 0x4 +#define _32BIT_MEM_RANGE_DESC 0x5 +#define _32BIT_FIXED_LOC_DESC 0x6 +#define LG_RES_RESERVED 0x7-0x7f + +/* Configuration Information */ +struct cinfo { + unsigned int serial; /* Board's Serial Number */ + int ldn; /* Log. Device #, If only 1 ld then ldn = -1 */ + struct { + int num; /* IRQ Number */ + int type; /* IRQ Type */ + } irq[2]; + int drq[2]; + int port[8]; /* The Base Address of the Port */ + struct { + int base; /* Memory Base Address */ + int control; /* Memory Control Register */ + int range; /* Memory Range *OR* Upper Limit */ + } mem[4]; +}; + +void pnp_configure __P((void)); + +#endif /* !_I386_ISA_PNP_H_ */ --==_Exmh_-1001429240 Content-Type: text/plain; charset=us-ascii David Nugent - Unique Computing Pty Ltd - Melbourne, Australia Voice +61-3-9791-9547 Data/BBS +61-3-9792-3507 3:632/348@fidonet davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ --==_Exmh_-1001429240-- From owner-freebsd-current Mon Jul 28 17:57:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA19323 for current-outgoing; Mon, 28 Jul 1997 17:57:18 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA19315 for ; Mon, 28 Jul 1997 17:57:16 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id RAA29178; Mon, 28 Jul 1997 17:57:03 -0700 (PDT) Message-Id: <199707290057.RAA29178@austin.polstra.com> To: msmith@atrad.adelaide.edu.au Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707261527.AAA28100@genesis.atrad.adelaide.edu.au> References: <199707261527.AAA28100@genesis.atrad.adelaide.edu.au> Organization: Polstra & Co., Seattle, WA Cc: current@freebsd.org Date: Mon, 28 Jul 1997 17:57:03 -0700 From: John Polstra Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In article <199707261527.AAA28100@genesis.atrad.adelaide.edu.au>, Michael Smith wrote: > Where is the ELF toolset at these days? The gcc that's in our main source tree can trivially be built for ELF by setting BINFORMAT=elf in the environment before running make. The current version of GNU binutils fully supports the configuration target "i386-unknown-freebsdelf". Our kernel, libc, and all other libraries I've tried build and work just fine for ELF. (They did the last time I checked, anyway.) The most recent work I did was to make the kernel compile, boot, and run when built as an ELF executable. That was met with a great big collective yawn when I committed it. All available evidence indicates that nobody else gives a tinker's damn about FreeBSD-ELF except as something fun to argue about passionately. (The people working on the alpha port are probably going to care about it some day, whenever they get around to that part.) I'm not complaining, but I'm also not feeling encouraged to expend any more effort on it. John -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Mon Jul 28 19:01:07 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA23949 for current-outgoing; Mon, 28 Jul 1997 19:01:07 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA23901 for ; Mon, 28 Jul 1997 19:00:16 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id LAA09348; Tue, 29 Jul 1997 11:25:50 +0930 (CST) From: Michael Smith Message-Id: <199707290155.LAA09348@genesis.atrad.adelaide.edu.au> Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707281748.KAA01355@phaeton.artisoft.com> from Terry Lambert at "Jul 28, 97 10:48:56 am" To: terry@lambert.org (Terry Lambert) Date: Tue, 29 Jul 1997 11:25:49 +0930 (CST) Cc: msmith@atrad.adelaide.edu.au, jkh@time.cdrom.com, nate@mt.sri.com, hosokawa@jp.freebsd.org, current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Terry Lambert stands accused of saying: > > My "vision" for looks something like this : > > [ ... ] > > Don't forget: > > - A "card services" utility to "preunload" cards; many cards are not > happy with an unplug (the MS docuemnts and Intel documents go > over this in great, gory detail). Understood. Unfortunately, there's no way to lock cards in, so you have to assume that the card has gone already. Certainly "revoking" the driver(s) handling a card should include an "unload" operation. > > Yah, I know I should just get out and do it. Gimme something with > > PCCARDS in it that I can call "mine" 8) 8) 8) > > I thought you had a Multia? My Multia have to PCMCIA slots... No. I looked at several, but came to the conclusion that AUD$1000 for a working Multia system was too much money, and spent it on a P6 instead. > Terry Lambert -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Mon Jul 28 20:53:32 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA04265 for current-outgoing; Mon, 28 Jul 1997 20:53:32 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA04232 for ; Mon, 28 Jul 1997 20:53:10 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id NAA10045; Tue, 29 Jul 1997 13:20:44 +0930 (CST) From: Michael Smith Message-Id: <199707290350.NAA10045@genesis.atrad.adelaide.edu.au> Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707282118.OAA01658@phaeton.artisoft.com> from Terry Lambert at "Jul 28, 97 02:18:29 pm" To: terry@lambert.org (Terry Lambert) Date: Tue, 29 Jul 1997 13:20:44 +0930 (CST) Cc: nate@mt.sri.com, terry@lambert.org, msmith@atrad.adelaide.edu.au, jkh@time.cdrom.com, hosokawa@jp.freebsd.org, current@freebsd.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Terry Lambert stands accused of saying: > > I was more worried about non-network PCMCIA cards. Examples might > be on of: > > o Adaptec SCSI controller PCMCIA card and all attached > SCSI devices. > > o Flash RAM card with mounted FS on it, with some FS data > in cache and needing to be flushed. "Please insert the card 'XXXXX' as it is needed for a critical system function" 8) Of course, that's not as simple as it sounds. > o Mounted hard drives located in the dock, in the case > where the machine is about to be undocked. Yipe. In most cases you won't get the APM/SMI event until _after_ it's been undocked. Talk about bogus design. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Mon Jul 28 21:24:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id VAA06752 for current-outgoing; Mon, 28 Jul 1997 21:24:26 -0700 (PDT) Received: from smoke.marlboro.vt.us (smoke.marlboro.vt.us [198.206.215.91]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA06744 for ; Mon, 28 Jul 1997 21:24:19 -0700 (PDT) Received: (from cgull@localhost) by smoke.marlboro.vt.us (8.8.5/8.8.5/cgull) id AAA23424; Tue, 29 Jul 1997 00:24:09 -0400 (EDT) Date: Tue, 29 Jul 1997 00:24:09 -0400 (EDT) Message-Id: <199707290424.AAA23424@smoke.marlboro.vt.us> From: john hood MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 To: SЬren Schmidt Cc: freebsd-current@FreeBSD.ORG Subject: code talks: announcing EIDE bus master patches In-Reply-To: <199707231824.UAA24871@sos.freebsd.dk> References: <199707231805.DAA10825@genesis.atrad.adelaide.edu.au> <199707231824.UAA24871@sos.freebsd.dk> X-Mailer: VM 6.31 under Emacs 19.34.2 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by hub.freebsd.org id VAA06745 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk SЬren Schmidt writes: > Well, I have two 4G EIDE Maxtors on my P6 :), when I get the bus master > DMA going, it'll be a tough bunch to beat... It's going to go a little quicker :) I've been coding up bus-mastering support for EIDE drives, and I'm at a point where I can release it to the world. I've got a patch against 2.2.2 available and various notes for the enjoyment of the curious kernel hacker. The wd driver is enough of a living fossil that the patches should apply easily to any kernel since then, though. The code is designed to support the SFF-8038 standard for busmaster controllers. It's generic enough to run on anything that nominally conforms to this (which includes every current PCI chipset I've seen); here's to hoping chipset quirks don't wreck this sublime quest. It's all very alpha yet, and it has been tested on exactly one system. but if the surreal thrill of watching a system transfer 6MB/s off an IDE disk with almost no cpu load mingled with the slight fear that it's trashing your disk at 6MB/s appeals to you, go to: http://localhost/pub/cgull/starBSD/sff8038/, download the tarball, try the code, and help me get it to a more functional state. --jh -- John Hood cgull@smoke.marlboro.vt.us Predictably, they all eventually wandered away, rubbing their bruises and brushing mud out of their hair. Some went off to work for the ESA, launching much smaller rockets into low orbits, while others elected to sit on their front porches drinking Jim Beam from the bottle and launching bottle rockets from the empties. [Jordan Hubbard] From owner-freebsd-current Mon Jul 28 22:09:48 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA09251 for current-outgoing; Mon, 28 Jul 1997 22:09:48 -0700 (PDT) Received: from smoke.marlboro.vt.us (smoke.marlboro.vt.us [198.206.215.91]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA09240 for ; Mon, 28 Jul 1997 22:09:44 -0700 (PDT) Received: (from cgull@localhost) by smoke.marlboro.vt.us (8.8.5/8.8.5/cgull) id BAA23557; Tue, 29 Jul 1997 01:09:40 -0400 (EDT) Date: Tue, 29 Jul 1997 01:09:40 -0400 (EDT) Message-Id: <199707290509.BAA23557@smoke.marlboro.vt.us> From: john hood MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: john hood Cc: SЬren Schmidt , Michael Smith , freebsd-current@FreeBSD.ORG Subject: code talks: announcing EIDE bus master patches In-Reply-To: <199707290424.AAA23424@smoke.marlboro.vt.us> References: <199707231805.DAA10825@genesis.atrad.adelaide.edu.au> <199707231824.UAA24871@sos.freebsd.dk> <199707290424.AAA23424@smoke.marlboro.vt.us> X-Mailer: VM 6.31 under Emacs 19.34.2 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk john hood writes: > http://localhost/pub/cgull/starBSD/sff8038/, ^^^^^^^^^ "but it works fine for me!" for the rest of you: http://smoke.marlboro.vt.us/pub/cgull/starBSD/sff8038/ --jh -- John Hood cgull@smoke.marlboro.vt.us Predictably, they all eventually wandered away, rubbing their bruises and brushing mud out of their hair. Some went off to work for the ESA, launching much smaller rockets into low orbits, while others elected to sit on their front porches drinking Jim Beam from the bottle and launching bottle rockets from the empties. [Jordan Hubbard] From owner-freebsd-current Mon Jul 28 22:10:41 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA09330 for current-outgoing; Mon, 28 Jul 1997 22:10:41 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA09322 for ; Mon, 28 Jul 1997 22:10:35 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id PAA18490; Tue, 29 Jul 1997 15:02:45 +1000 Date: Tue, 29 Jul 1997 15:02:45 +1000 From: Bruce Evans Message-Id: <199707290502.PAA18490@godzilla.zeta.org.au> To: bde@zeta.org.au, rgrimes@GndRsh.aac.dev.com Subject: Re: Multiple FreeBSD Systems on a Single Disk Cc: abial@korin.warman.org.pl, current@FreeBSD.ORG, sherwink@ix.netcom.com Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> The lack of support has nothing to do with the master boot loader. >> The FreeBSD boot loader only supports booting from the first FreeBSD >> slice (on any disk that can be accessed by the BIOS). > >And should be fixed (I've mentioned this before) to boot from the >first ACTIVE (flag 0x80) FreeBSD slice. Should be a trival amount That would mainly break certain configurations. The active flag should only be set for one partition, so you shouldn't have any active FreeBSD partitions if you have an active DOS partition. This should be fixed by adding slice support to the boot loader's name parser. Then you could bounce off the first FreeBSD to any other slice by putting a boot loader on the first slice and configuring it to boot from slice sN, e.g., "0:sd(0,s5,a)kernel" to boot the first logical drive on an extended slice. Bruce From owner-freebsd-current Tue Jul 29 00:24:50 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA20550 for current-outgoing; Tue, 29 Jul 1997 00:24:50 -0700 (PDT) Received: from helmholtz.salk.edu (helmholtz.salk.edu [198.202.70.34]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA20545 for ; Tue, 29 Jul 1997 00:24:45 -0700 (PDT) Received: from helmholtz (helmholtz [198.202.70.34]) by helmholtz.salk.edu (8.7.5/8.7.3) with SMTP id AAA12475 for ; Tue, 29 Jul 1997 00:24:43 -0700 (PDT) Date: Tue, 29 Jul 1997 00:24:43 -0700 (PDT) From: Tom Bartol X-Sender: bartol@helmholtz To: current@freebsd.org Subject: make world fails in tcl Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi all, Does anyone else get the following failure in make world, cvsupped a few minutes ago? cc -O2 -m486 -pipe -I/usr/src/lib/libtcl/../../contrib/tcl/generic -I/usr/src/lib/libtcl/../../contrib/tcl/unix -DNO_VALUES_H=1 -DHAVE_UNISTD_H=1 -DUSE_TERMIOS=1 -DHAVE_SYS_TIME_H=1 -DTIME_WITH_SYS_TIME=1 -DHAVE_TM_ZONE=1 -DHAVE_TM_GMTOFF=1 -DSTDC_HEADERS=1 -DNEED_MATHERR=1 -DHAVE_SIGNED_CHAR=1 -DHAVE_SYS_IOCTL_H=1 -DHAVE_SYS_FILIO_H=1 -DTCL_SHLIB_EXT=\".so\" -DTCL_LIBRARY=\"/usr/libdata/tcl\" -DTCL_PACKAGE_PATH=\"/usr/lib/tclpkg\ /usr/local/lib/tclpkg\" -c /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c -o tclMtherr.o /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c: In function `matherr': /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: dereferencing pointer to incomplete type /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: `DOMAIN' undeclared (first use this function) /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: (Each undeclared identifier is reported only once /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: for each function it appears in.) /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: dereferencing pointer to incomplete type /usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: `SING' undeclared (first use this function) *** Error code 1 Stop. The problem seems to arise from the following code in tclMtherr.c: /* * The following definitions allow matherr to compile on systems * that don't really support it. The compiled procedure is bogus, * but it will never be executed on these systems anyway. */ #ifndef NEED_MATHERR struct exception { int type; }; #define DOMAIN 0 #define SING 0 #endif Which never gets used because the Makefile in src/lib/libtcl does a -DNEED_MATHERR=1 within the CFLAGS definition. Therefore SING and DOMAIN are undefined. Tom From owner-freebsd-current Tue Jul 29 01:03:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA22422 for current-outgoing; Tue, 29 Jul 1997 01:03:26 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA22416 for ; Tue, 29 Jul 1997 01:03:18 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id RAA12011; Tue, 29 Jul 1997 17:33:05 +0930 (CST) From: Michael Smith Message-Id: <199707290803.RAA12011@genesis.atrad.adelaide.edu.au> Subject: Re: make world fails in tcl In-Reply-To: from Tom Bartol at "Jul 29, 97 00:24:43 am" To: bartol@salk.edu (Tom Bartol) Date: Tue, 29 Jul 1997 17:33:05 +0930 (CST) Cc: current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Tom Bartol stands accused of saying: > > > Hi all, > > Does anyone else get the following failure in make world, cvsupped a few > minutes ago? You have changed your math library preferences in /etc/make.conf -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Tue Jul 29 01:20:46 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA23662 for current-outgoing; Tue, 29 Jul 1997 01:20:46 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA23654 for ; Tue, 29 Jul 1997 01:20:41 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id RAA12092; Tue, 29 Jul 1997 17:50:11 +0930 (CST) From: Michael Smith Message-Id: <199707290820.RAA12092@genesis.atrad.adelaide.edu.au> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707290509.BAA23557@smoke.marlboro.vt.us> from john hood at "Jul 29, 97 01:09:40 am" To: cgull@smoke.marlboro.vt.us (john hood) Date: Tue, 29 Jul 1997 17:50:10 +0930 (CST) Cc: cgull@smoke.marlboro.vt.us, sos@sos.freebsd.dk, msmith@atrad.adelaide.edu.au, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk john hood stands accused of saying: > john hood writes: > > http://localhost/pub/cgull/starBSD/sff8038/, > ^^^^^^^^^ > > "but it works fine for me!" > > for the rest of you: > > http://smoke.marlboro.vt.us/pub/cgull/starBSD/sff8038/ *chuckle* You wouldn't be being, er, sponsored for this particular URL now would you? The real question - do you have a reference document/pointer to this standard? > Predictably, they all eventually wandered away, rubbing their bruises > and brushing mud out of their hair. Some went off to work for the > ESA, launching much smaller rockets into low orbits, while others > elected to sit on their front porches drinking Jim Beam from the > bottle and launching bottle rockets from the empties. [Jordan Hubbard] Uhh. He _does_ hang out in strange places, doesn't he? -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Tue Jul 29 01:32:29 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA24522 for current-outgoing; Tue, 29 Jul 1997 01:32:29 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA24515 for ; Tue, 29 Jul 1997 01:32:24 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id SAA25614; Tue, 29 Jul 1997 18:31:45 +1000 Date: Tue, 29 Jul 1997 18:31:45 +1000 From: Bruce Evans Message-Id: <199707290831.SAA25614@godzilla.zeta.org.au> To: bartol@salk.edu, current@FreeBSD.ORG Subject: Re: make world fails in tcl Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >/usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: `DOMAIN' >undeclared (first use this function) >... >/usr/src/lib/libtcl/../../contrib/tcl/unix/tclMtherr.c:80: `SING' >undeclared (first use this function) >*** Error code 1 >... >The problem seems to arise from the following code in tclMtherr.c: > >/* > * The following definitions allow matherr to compile on systems > * that don't really support it. The compiled procedure is bogus, > * but it will never be executed on these systems anyway. > */ > >#ifndef NEED_MATHERR >struct exception { > int type; >}; >#define DOMAIN 0 >#define SING 0 >#endif > > > >Which never gets used because the Makefile in src/lib/libtcl does a >-DNEED_MATHERR=1 within the CFLAGS definition. Therefore SING and DOMAIN >are undefined. SING and DOMAIN are defined in , provided the standard math library (src/lib/msun) is used. The old math library (src/lib/libm) has various problems and should normally not be used. The standard math library is configured to never call matherr(). To enable matherr(), recompile the library to support multiple modes and switch to SYSV mode. You shouldn't want this mode (except possibly for matherr()). Bruce From owner-freebsd-current Tue Jul 29 01:42:04 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA25021 for current-outgoing; Tue, 29 Jul 1997 01:42:04 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA25016 for ; Tue, 29 Jul 1997 01:42:00 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id KAA00348; Tue, 29 Jul 1997 10:42:05 +0200 (MEST) From: SЬren Schmidt Message-Id: <199707290842.KAA00348@sos.freebsd.dk> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707290424.AAA23424@smoke.marlboro.vt.us> from john hood at "Jul 29, 97 00:24:09 am" To: cgull@smoke.marlboro.vt.us (john hood) Date: Tue, 29 Jul 1997 10:42:05 +0200 (MEST) Cc: freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to john hood who wrote: > SЬren Schmidt writes: > > Well, I have two 4G EIDE Maxtors on my P6 :), when I get the bus master > > DMA going, it'll be a tough bunch to beat... > > It's going to go a little quicker :) > > I've been coding up bus-mastering support for EIDE drives, and I'm at > a point where I can release it to the world. I've got a patch against > 2.2.2 available and various notes for the enjoyment of the curious > kernel hacker. The wd driver is enough of a living fossil that the > patches should apply easily to any kernel since then, though. > > The code is designed to support the SFF-8038 standard for busmaster > controllers. It's generic enough to run on anything that nominally > conforms to this (which includes every current PCI chipset I've seen); > here's to hoping chipset quirks don't wreck this sublime quest. > > It's all very alpha yet, and it has been tested on exactly one system. > but if the surreal thrill of watching a system transfer 6MB/s off an > IDE disk with almost no cpu load mingled with the slight fear that > it's trashing your disk at 6MB/s appeals to you, go to: Well, doing a little surgery to it and sticking it into current, (the PCI subsystem has changed a little from 2.2.x to current) I'd say it works pretty well here too (440FX chipset Maxtor drives). I'l beat on it for a while, then commit the changes to -current. I think we should leave 2.2-STABLE out for the moment :) -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- SЬren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end .. From owner-freebsd-current Tue Jul 29 04:44:24 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA06279 for current-outgoing; Tue, 29 Jul 1997 04:44:24 -0700 (PDT) Received: from hauki.clinet.fi (root@hauki.clinet.fi [194.100.0.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA06274 for ; Tue, 29 Jul 1997 04:44:16 -0700 (PDT) Received: from katiska.clinet.fi (root@katiska.clinet.fi [194.100.0.4]) by hauki.clinet.fi (8.8.6/8.8.6) with ESMTP id OAA06317 for ; Tue, 29 Jul 1997 14:43:55 +0300 (EET DST) Received: (root@localhost) by katiska.clinet.fi (8.8.6/8.6.4) id OAA08586; Tue, 29 Jul 1997 14:43:54 +0300 (EEST) Date: Tue, 29 Jul 1997 14:43:54 +0300 (EEST) Message-Id: <199707291143.OAA08586@katiska.clinet.fi> From: Heikki Suonsivu To: freebsd-current@freebsd.org Subject: Conversion tool from sysconfig to rc.conf Reply-To: Heikki Suonsivu Organization: Clinet Ltd, Espoo, Finland Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Has anyone written a conversion tool from sysconfig to rc.conf ? Anyway, if I need to upgrade old machine /etc to 2.2-stable, are there other files besides rc*,daily,weekly,monthly I need to update ? I assume sysconfig -> rc.conf in only really necessary conversion, others I could just copy ? Some of these systems are as old as 2.0 /etc, but have been upgraded every now and then with make reinstall. -- Heikki Suonsivu, T{ysikuu 10 C 83/02210 Espoo/FINLAND, hsu@clinet.fi mobile +358-40-5519679 work +358-9-43542270 fax -4555276 From owner-freebsd-current Tue Jul 29 07:08:41 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA15210 for current-outgoing; Tue, 29 Jul 1997 07:08:41 -0700 (PDT) Received: from whqvax.picker.com (whqvax.picker.com [144.54.1.1]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id HAA15196 for ; Tue, 29 Jul 1997 07:08:37 -0700 (PDT) Received: from ct.picker.com by whqvax.picker.com with SMTP; Tue, 29 Jul 1997 10:07:24 -0400 (EDT) Received: from elmer.ct.picker.com ([144.54.57.34]) by ct.picker.com (4.1/SMI-4.1) id AA17992; Tue, 29 Jul 97 10:07:21 EDT Received: by elmer.ct.picker.com (SMI-8.6/SMI-SVR4) id KAA14699; Tue, 29 Jul 1997 10:04:55 -0400 Message-Id: <19970729100455.10977@ct.picker.com> Date: Tue, 29 Jul 1997 10:04:55 -0400 From: Randall Hopper To: "Rodney W. Grimes" Cc: Bruce Evans , current@FreeBSD.ORG Subject: Re: Multiple FreeBSD Systems on a Single Disk References: <199707281528.BAA23105@godzilla.zeta.org.au> <199707281729.KAA18261@GndRsh.aac.dev.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.76 In-Reply-To: <199707281729.KAA18261@GndRsh.aac.dev.com>; from Rodney W. Grimes on Mon, Jul 28, 1997 at 10:29:04AM -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Rodney W. Grimes: |> >> Only booting from the first FreeBSD partition (slice) is supported. |> >> The second one can be booted from by making it the first, e.g., |> > |> >Even simpler: you can install OS-BS bootloader (from /tools), which is |> >much more configurable than boot-easy. |> |> The lack of support has nothing to do with the master boot loader. |> The FreeBSD boot loader only supports booting from the first FreeBSD |> slice (on any disk that can be accessed by the BIOS). | |And should be fixed (I've mentioned this before) to boot from the |first ACTIVE (flag 0x80) FreeBSD slice. Should be a trival amount |of code, and would do what everyone else does, even Microsoft got |that part right. That's one option. Another one that'd be better is for FreeBSD to just boot the slice whose Boot Record is loaded. I.e. if OS/BS transfers control to Disk 1 Slice 2, Disk 1 Slice 2's Boot Record is loaded, which should boot FreeBSD off Slice 2. Ditto for Slice 1 if I happen to have FreeBSD in that slice as well. If I boot Slice 2, FreeBSD shouldn't try to read or touch Slice 1 during the boot loading sequence. IMO, it shouldn't go divining whether there's a valid FreeBSD on Slice 1 (that may or may not be trashed) which it can pick up a boot loader from. The MBR transferring control to Boot Record Slice 2 should mean to boot off Slice 2. Randall From owner-freebsd-current Tue Jul 29 08:19:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id IAA20653 for current-outgoing; Tue, 29 Jul 1997 08:19:26 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA20643 for ; Tue, 29 Jul 1997 08:19:23 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id BAA06551; Wed, 30 Jul 1997 01:14:05 +1000 Date: Wed, 30 Jul 1997 01:14:05 +1000 From: Bruce Evans Message-Id: <199707291514.BAA06551@godzilla.zeta.org.au> To: rgrimes@GndRsh.aac.dev.com, rhh@ct.picker.com Subject: Re: Multiple FreeBSD Systems on a Single Disk Cc: bde@zeta.org.au, current@FreeBSD.ORG Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > |And should be fixed (I've mentioned this before) to boot from the > |first ACTIVE (flag 0x80) FreeBSD slice. Should be a trival amount > |of code, and would do what everyone else does, even Microsoft got > |that part right. > >That's one option. Another one that'd be better is for FreeBSD to just >boot the slice whose Boot Record is loaded. I.e. if OS/BS transfers >control to Disk 1 Slice 2, Disk 1 Slice 2's Boot Record is loaded, which >should boot FreeBSD off Slice 2. Ditto for Slice 1 if I happen to have >FreeBSD in that slice as well. The slice number is difficult to determine. 1. If the FreeBSD bootstrap is on the MBR of a floppy or hard disk, then there is no slice number. 2. Booteasy documents passing the slice number indirectly as `ES:SI = address of partition table entry to boot. It also leaves useful garbage in registers. Even the documented interface is unportable (it isn't documented in the Phoenix BIOS book (1991) or _The Undocumented PC_ (1994), although these references document passing the drive number in DL and we know that isn't portable to floppies). The slice number could be encoded in the first sector of the FreeBSD bootstrap, but this would break device independence. Bruce From owner-freebsd-current Tue Jul 29 08:49:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id IAA23555 for current-outgoing; Tue, 29 Jul 1997 08:49:21 -0700 (PDT) Received: from helmholtz.salk.edu (helmholtz.salk.edu [198.202.70.34]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA23503 for ; Tue, 29 Jul 1997 08:49:07 -0700 (PDT) Received: from pauling.salk.edu (pauling [198.202.70.108]) by helmholtz.salk.edu (8.7.5/8.7.3) with SMTP id IAA18682; Tue, 29 Jul 1997 08:48:33 -0700 (PDT) Date: Tue, 29 Jul 1997 08:48:30 -0700 (PDT) From: Tom Bartol To: Michael Smith cc: current@FreeBSD.ORG Subject: Re: make world fails in tcl In-Reply-To: <199707290803.RAA12011@genesis.atrad.adelaide.edu.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Yup, that was it! I specified WANT_CSRG_LIBM=yes in /etc/make.conf on a whim to try it out for performance. Doctor says, "Then don't do that!" Thanks, Tom On Tue, 29 Jul 1997, Michael Smith wrote: > Tom Bartol stands accused of saying: > > > > > > Hi all, > > > > Does anyone else get the following failure in make world, cvsupped a few > > minutes ago? > > You have changed your math library preferences in /etc/make.conf > > > -- > ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ > ]] Genesis Software genesis@gsoft.com.au [[ > ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ > ]] realtime instrument control. (ph) +61-8-8267-3493 [[ > ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ > From owner-freebsd-current Tue Jul 29 09:30:49 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id JAA26849 for current-outgoing; Tue, 29 Jul 1997 09:30:49 -0700 (PDT) Received: from GndRsh.aac.dev.com (GndRsh.aac.dev.com [198.145.92.241]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id JAA26836 for ; Tue, 29 Jul 1997 09:30:45 -0700 (PDT) Received: (from rgrimes@localhost) by GndRsh.aac.dev.com (8.8.5/8.7.3) id JAA20288; Tue, 29 Jul 1997 09:30:13 -0700 (PDT) From: "Rodney W. Grimes" Message-Id: <199707291630.JAA20288@GndRsh.aac.dev.com> Subject: Re: Multiple FreeBSD Systems on a Single Disk In-Reply-To: <199707290502.PAA18490@godzilla.zeta.org.au> from Bruce Evans at "Jul 29, 97 03:02:45 pm" To: bde@zeta.org.au (Bruce Evans) Date: Tue, 29 Jul 1997 09:30:13 -0700 (PDT) Cc: bde@zeta.org.au, abial@korin.warman.org.pl, current@FreeBSD.ORG, sherwink@ix.netcom.com X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > >> The lack of support has nothing to do with the master boot loader. > >> The FreeBSD boot loader only supports booting from the first FreeBSD > >> slice (on any disk that can be accessed by the BIOS). > > > >And should be fixed (I've mentioned this before) to boot from the > >first ACTIVE (flag 0x80) FreeBSD slice. Should be a trival amount > > That would mainly break certain configurations. The active flag > should only be set for one partition, so you shouldn't have any active > FreeBSD partitions if you have an active DOS partition. Right, and if I have an active DOS partition the BIOS is going to boot the ACTIVE dos partition. It won't even load the BSD boot blocks if that is not the active partition. Any configuration that was broken by this change was broken such that the BIOS should have not even booted anything! The BIOS should have returned a ``No Active Partition'', or gone on to the next disk drive (yes, AMI bios'es as old as 1982 can boot from drive 0x81 if there are no active partitions on drive 0x80.) > This should be fixed by adding slice support to the boot loader's > name parser. Then you could bounce off the first FreeBSD to any other > slice by putting a boot loader on the first slice and configuring it > to boot from slice sN, e.g., "0:sd(0,s5,a)kernel" to boot the first > logical drive on an extended slice. Poppy cock, the active flag model should work and not require the addition of slice choosing support to the boot blocks. OSBS and other multiboot loaders will infact set the chosen partition to active, it is just the brain damaged FreeBSD boot code that ignores the active flag, and chooses the FIRST BSD parition it finds no matter what, which is _WRONG_. -- Rod Grimes rgrimes@gndrsh.aac.dev.com Accurate Automation, Inc. Reliable computers for FreeBSD From owner-freebsd-current Tue Jul 29 09:46:20 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id JAA28031 for current-outgoing; Tue, 29 Jul 1997 09:46:20 -0700 (PDT) Received: from ox.ismi.net (root@ox.ismi.net [206.31.56.6]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id JAA28017 for ; Tue, 29 Jul 1997 09:46:12 -0700 (PDT) Received: from aerosmith.dyn.ml.org (mrr@pm3-18.ismi.net [206.31.56.88]) by ox.ismi.net (8.6.9/8.6.9) with SMTP id MAA19246; Tue, 29 Jul 1997 12:45:09 -0400 Date: Tue, 29 Jul 1997 12:38:50 -0400 (EDT) From: "Michael R. Rudel" To: Minsung Kim cc: FreeBSD-CURRENT Usergroup Subject: Re: 3.0-CURRENT kernel anomaly with X In-Reply-To: <33DA8E28.E16EA8BA@mci2000.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 26 Jul 1997, Minsung Kim wrote: > I've upgraded my kernel to -current recently, and seeing these weird things > during X is running. (The symptom goes worse if Netscape 4.01b6[en] for Linux > 2.0 is running.) > > o Many instances of of shell scripts die screaming "out of swap space." That's > seen on the console like this: > > Jul 25 00:01:34 x /kernel: swap_pager: out of swap space > Jul 25 00:01:37 x /kernel: pid 2553 (bash), uid 0, was killed: out of swap space > > o Compilation of program fails occasionally with apparently absurd errors -- I > can't catch any errors in the source code -- when a simple retry succeeds. > o Outputs of compiled code varies from time to time. (I noticed it building my > kernel -- I usually builds the kernel twice and compares object files to each > other, except for vers.o.) > o The 1st characters of items on my TWM menu are garbled. (For instance, "File > Runner" sometimes looks "Nile Runner", etc. :-&) > > Does any of these ring a bell? TIA. Somebody has probably replied by now, but try a make world from /usr/src... I've had this problem (or similar to this) once or twice... I've found if your system is broke, a make world is the happy, easy, fix that takes ~8 hours. ;) > [...SNIP...] -- Michael R. Rudel -=- FreeBSD: There are no limits -=- mrr@aerosmith.dyn.ml.org FreeBSD aerosmith.dyn.ml.org 3.0-CURRENT PGP Key Block: finger mrrpgp@aerosmith.dyn.ml.org When you are born your afraid of the darkness ... Then your afraid of the light ... I'm not afraid when I dance with my shadows ... This time I'm gonna get it right ... -- Aerosmith: Taste of India From owner-freebsd-current Tue Jul 29 10:46:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA01984 for current-outgoing; Tue, 29 Jul 1997 10:46:34 -0700 (PDT) Received: from lsd.relcom.eu.net (lsd.relcom.eu.net [193.124.23.23]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA01971 for ; Tue, 29 Jul 1997 10:46:24 -0700 (PDT) Received: (from ache@localhost) by lsd.relcom.eu.net (8.8.6/8.8.5) id VAA13410; Tue, 29 Jul 1997 21:46:06 +0400 (MSD) Date: Tue, 29 Jul 1997 21:46:05 +0400 (MSD) From: =?KOI8-R?B?4c7E0sXKIP7F0s7P1w==?= X-Sender: ache@lsd.relcom.eu.net To: "Michael R. Rudel" cc: Minsung Kim , FreeBSD-CURRENT Usergroup Subject: Re: 3.0-CURRENT kernel anomaly with X In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Tue, 29 Jul 1997, Michael R. Rudel wrote: > Somebody has probably replied by now, but try a make world from > /usr/src... I've had this problem (or similar to this) once or twice... > > I've found if your system is broke, a make world is the happy, easy, fix > that takes ~8 hours. ;) I notice slightly different problem few times: X server running several days just start to eat CPU time and swap and dies... Make world not heal it. -- Andrey A. Chernov http://www.nagual.pp.ru/~ache/ From owner-freebsd-current Tue Jul 29 11:08:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA03394 for current-outgoing; Tue, 29 Jul 1997 11:08:36 -0700 (PDT) Received: from watson.grauel.com (watson.grauel.com [199.233.104.36]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA03383 for ; Tue, 29 Jul 1997 11:08:34 -0700 (PDT) Received: (from rjk@localhost) by watson.grauel.com (8.8.5/8.7.3) id NAA02326; Tue, 29 Jul 1997 13:08:42 -0500 (EST) Date: Tue, 29 Jul 1997 13:08:42 -0500 (EST) Message-Id: <199707291808.NAA02326@watson.grauel.com> From: Richard J Kuhns To: freebsd-current@freebsd.org Subject: Can't mount CD under -current. Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk With a -current kernel as of 7/28, I can no longer mount a CD. I remember seeing comments about breakages when the webnfs stuff went in; I hope that this was just overlooked. My cdrom is the one that came with my SoundBlaster 16, and has worked fine until now. When I try to mount a CD, moran:~# mount /cdrom cd9660: Invalid argument moran:~# and the following appears on the console: Jul 29 13:06:06 moran /kernel: matcd0: Bogus parameters received - kernel may be corrupted If any more info is needed, I'll be happy to supply it. Thanks... -- Rich Kuhns rjk@grauel.com PO Box 6249 Tel: (765)477-6000 \ 100 Sawmill Road x319 Lafayette, IN 47903 (800)489-4891 / From owner-freebsd-current Tue Jul 29 12:07:38 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA06333 for current-outgoing; Tue, 29 Jul 1997 12:07:38 -0700 (PDT) Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id MAA06323 for ; Tue, 29 Jul 1997 12:07:33 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr2-41.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA14415 (5.67b/IDA-1.5 for ); Tue, 29 Jul 1997 21:07:29 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.6/8.6.9) id VAA03768; Tue, 29 Jul 1997 21:07:24 +0200 (CEST) X-Face: " Date: Tue, 29 Jul 1997 21:07:23 +0200 From: Stefan Esser To: =?iso-8859-1?Q?S=F8ren_Schmidt?= Cc: freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches References: <199707290424.AAA23424@smoke.marlboro.vt.us> <199707290842.KAA00348@sos.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 X-Mailer: Mutt 0.74 In-Reply-To: =?iso-8859-1?Q?=3C199707290842=2EKAA00348=40sos=2Efreebsd=2Edk=3E=3B_fro?= =?iso-8859-1?Q?m_S=F8ren_Schmidt_on_Tue=2C_Jul_29=2C_1997_at_10=3A42=3A0?= =?iso-8859-1?Q?5AM_+0200?= X-Mime-Autoconverted: from 8bit to quoted-printable by x14.mi.uni-koeln.de id VAA03768 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by hub.freebsd.org id MAA06326 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Jul 29, SЬren Schmidt wrote: > Well, doing a little surgery to it and sticking it into current, > (the PCI subsystem has changed a little from 2.2.x to current) > I'd say it works pretty well here too (440FX chipset Maxtor drives). Are these Maxtor DiamondMax drives ? I'd love to see Bonnie results for them with the bus-master EIDE driver ... You don't have, by chance, any numbers to share ? :) Regards, STefan From owner-freebsd-current Tue Jul 29 12:50:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA08765 for current-outgoing; Tue, 29 Jul 1997 12:50:16 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA08710; Tue, 29 Jul 1997 12:49:39 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id VAA00271; Tue, 29 Jul 1997 21:49:29 +0200 (MEST) From: SЬren Schmidt Message-Id: <199707291949.VAA00271@sos.freebsd.dk> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970729210723.18104@mi.uni-koeln.de> from Stefan Esser at "Jul 29, 97 09:07:23 pm" To: se@FreeBSD.ORG (Stefan Esser) Date: Tue, 29 Jul 1997 21:49:29 +0200 (MEST) Cc: freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to Stefan Esser who wrote: > On Jul 29, SЬren Schmidt wrote: > > Well, doing a little surgery to it and sticking it into current, > > (the PCI subsystem has changed a little from 2.2.x to current) > > I'd say it works pretty well here too (440FX chipset Maxtor drives). > > Are these Maxtor DiamondMax drives ? Yep these are Maxtor DiamondMax 84000A6 drives (4G) > I'd love to see Bonnie results for them > with the bus-master EIDE driver ... > > You don't have, by chance, any numbers > to share ? :) Sure, on my dev machine (P6@233Mhz-Natoma/64MB/2*Maxtor 84000A6) First without busmatering DMA: -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU 100 4871 35.9 4738 9.4 1722 3.6 4488 37.2 6973 8.5 126.7 2.2 Then with busmastering DMA: -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU 100 5225 39.8 5278 11.1 2017 4.4 7315 60.2 8737 10.8 149.6 2.2 Not bad, for a "simple" software upgrade :) It has improved my worldstone by ~10% if that counts for real world use.. -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- SЬren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end .. From owner-freebsd-current Tue Jul 29 13:10:41 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA09931 for current-outgoing; Tue, 29 Jul 1997 13:10:41 -0700 (PDT) Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id NAA09910 for ; Tue, 29 Jul 1997 13:10:35 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr2-41.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA15117 (5.67b/IDA-1.5 for ); Tue, 29 Jul 1997 22:10:32 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.6/8.6.9) id WAA04817; Tue, 29 Jul 1997 22:10:36 +0200 (CEST) X-Face: " Date: Tue, 29 Jul 1997 22:10:36 +0200 From: Stefan Esser To: =?iso-8859-1?Q?S=F8ren_Schmidt?= Cc: freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches References: <19970729210723.18104@mi.uni-koeln.de> <199707291949.VAA00271@sos.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 X-Mailer: Mutt 0.74 In-Reply-To: =?iso-8859-1?Q?=3C199707291949=2EVAA00271=40sos=2Efreebsd=2Edk=3E=3B_fro?= =?iso-8859-1?Q?m_S=F8ren_Schmidt_on_Tue=2C_Jul_29=2C_1997_at_09=3A49=3A2?= =?iso-8859-1?Q?9PM_+0200?= X-Mime-Autoconverted: from 8bit to quoted-printable by x14.mi.uni-koeln.de id WAA04817 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by hub.freebsd.org id NAA09926 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Jul 29, SЬren Schmidt wrote: > > I'd love to see Bonnie results for them > > with the bus-master EIDE driver ... > > > > You don't have, by chance, any numbers > > to share ? :) > > Sure, on my dev machine (P6@233Mhz-Natoma/64MB/2*Maxtor 84000A6) This seems to be an OEM version of the 83840a6, if I interpret the data sheets on their web page correctly. AFAIK the fastest EIDE drive series on the market today! > First without busmatering DMA: > > -------Sequential Output-------- ---Sequential Input-- --Random-- > -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- > Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU > 100 4871 35.9 4738 9.4 1722 3.6 4488 37.2 6973 8.5 126.7 2.2 Well, as was discussed before, the PIO transfers take place mostly in an interrupt handler, which prevents accounting of the cycles to the process that caused them ... > Then with busmastering DMA: > > -------Sequential Output-------- ---Sequential Input-- --Random-- > -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- > Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU > 100 5225 39.8 5278 11.1 2017 4.4 7315 60.2 8737 10.8 149.6 2.2 And the result is apparently *higher* CPU load, at least if you (did) trust the numbers reported by Bonnie ... But the "per char" read performance improvement indicates what's really going on: CPU cycles spent in the interrupt handler probably account for the 4.5MB/s mark in the PIO case, while raw data rate of the disk is approached with the DMA driver. But writes are still significantly slower than reads, which makes me think, that one revolution of the media is lost per 64KB written ... > Not bad, for a "simple" software upgrade :) True! But there is still room for improvement ... ;) > It has improved my worldstone by ~10% if that counts for real world use.. Yes, one thing that is often underestimated is the fact, that while PIO mode transfers consume only a small fraction of the cycles of a fast CPU, but they tend to consume them exactly at the time, when the CPU is highly loaded anyway. Thanks for sending those very interesting Bonnie results! Regards, STefan From owner-freebsd-current Tue Jul 29 14:17:33 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA13504 for current-outgoing; Tue, 29 Jul 1997 14:17:33 -0700 (PDT) Received: from hda.hda.com (hda-bicnet.bicnet.net [208.220.66.37]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA13498 for ; Tue, 29 Jul 1997 14:17:29 -0700 (PDT) Received: (from dufault@localhost) by hda.hda.com (8.8.5/8.8.5) id QAA28302 for current@freebsd.org; Tue, 29 Jul 1997 16:31:53 -0400 (EDT) From: Peter Dufault Message-Id: <199707292031.QAA28302@hda.hda.com> Subject: where to put access restriction for scheduling classes To: current@freebsd.org Date: Tue, 29 Jul 1997 16:31:52 -0400 (EDT) X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I'm looking at switching from a pseudo device that sets a sysctl variable to something more mainstream to control access to sched_setscheduler. Can someone in the know explain the right thing to do - I don't want to create a new resource, so is the right thing a sysctl variable that somehow gets set during login via login_class while still root? Which of those are inherited per-process? Or do I actually add a resource and only change the login stuff? Peter -- Peter Dufault (dufault@hda.com) Realtime development, Machine control, HD Associates, Inc. Safety critical systems, Agency approval From owner-freebsd-current Tue Jul 29 15:24:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA16932 for current-outgoing; Tue, 29 Jul 1997 15:24:36 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA16922; Tue, 29 Jul 1997 15:24:24 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id AAA00624; Wed, 30 Jul 1997 00:24:36 +0200 (MEST) From: SЬren Schmidt Message-Id: <199707292224.AAA00624@sos.freebsd.dk> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970729221036.52544@mi.uni-koeln.de> from Stefan Esser at "Jul 29, 97 10:10:36 pm" To: se@FreeBSD.ORG (Stefan Esser) Date: Wed, 30 Jul 1997 00:24:36 +0200 (MEST) Cc: freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to Stefan Esser who wrote: > On Jul 29, SЬren Schmidt wrote: > > > > Sure, on my dev machine (P6@233Mhz-Natoma/64MB/2*Maxtor 84000A6) > > This seems to be an OEM version of the 83840a6, if I > interpret the data sheets on their web page correctly. > AFAIK the fastest EIDE drive series on the market today! Well, I think they are mainline drives, you can buy them on every corner here in DK land for about US$250. > And the result is apparently *higher* CPU load, at least > if you (did) trust the numbers reported by Bonnie ... Well its most visible in the "per char" mesurements, I guess time is spent processing the inividual chars through the I/O system, the increase in used time is almost linear. > But the "per char" read performance improvement indicates > what's really going on: CPU cycles spent in the interrupt > handler probably account for the 4.5MB/s mark in the PIO > case, while raw data rate of the disk is approached with > the DMA driver. Exactly! > But writes are still significantly slower than reads, which > makes me think, that one revolution of the media is lost per > 64KB written ... I think it might be some kind of interaction with the drives cache, but our driver might be guilty too.. > > Not bad, for a "simple" software upgrade :) > > True! But there is still room for improvement ... ;) Well, I don't think we are going to get much more performance out of it, at least not in an order like this, but there sure is room for cleaning up the driver, which I'll eventually get to if/when I get the time... > > It has improved my worldstone by ~10% if that counts for real world use.. > > Yes, one thing that is often underestimated is the fact, > that while PIO mode transfers consume only a small fraction > of the cycles of a fast CPU, but they tend to consume them > exactly at the time, when the CPU is highly loaded anyway. Yep. Now we only need ATA drives that can overlap commands. It seems they are getting there, but its a mess still. When that settles (and we get new drives :) ), we can get even a wee bit more performance... > Thanks for sending those very interesting Bonnie results! You're welcome :) -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- SЬren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end .. From owner-freebsd-current Tue Jul 29 16:31:24 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA21057 for current-outgoing; Tue, 29 Jul 1997 16:31:24 -0700 (PDT) Received: from internet1.mel.cybec.com.au (internet1.mel.cybec.com.au [203.103.154.130]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA21049 for ; Tue, 29 Jul 1997 16:31:16 -0700 (PDT) Received: from tech34 (tech34.mel.cybec.com.au [203.103.154.37]) by internet1.mel.cybec.com.au (post.office MTA v2.0 0813 ID# 0-14031) with ESMTP id AAA507; Wed, 30 Jul 1997 09:32:43 +1000 Message-ID: <33DE7DD8.2A404ED0@cybec.com.au> Date: Wed, 30 Jul 1997 09:33:44 +1000 From: TLiddelow@cybec.com.au (Tim Liddelow) X-Mailer: Mozilla 4.01 [en] (WinNT; I) MIME-Version: 1.0 To: "Rodney W. Grimes" CC: Bruce Evans , abial@korin.warman.org.pl, current@FreeBSD.ORG, sherwink@ix.netcom.com Subject: Re: Multiple FreeBSD Systems on a Single Disk X-Priority: 3 (Normal) References: <199707291630.JAA20288@GndRsh.aac.dev.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Rodney W. Grimes wrote: > > > Any configuration that was broken by this change was broken such > that the BIOS should have not even booted anything! The BIOS should > have returned a ``No Active Partition'', or gone on to the next > disk drive (yes, AMI bios'es as old as 1982 can boot from drive > 0x81 if there are no active partitions on drive 0x80.) > > > This should be fixed by adding slice support to the boot loader's > > name parser. Then you could bounce off the first FreeBSD to any > other > > slice by putting a boot loader on the first slice and configuring it > > to boot from slice sN, e.g., "0:sd(0,s5,a)kernel" to boot the first > > logical drive on an extended slice. > > Poppy cock, the active flag model should work and not require the > addition of slice choosing support to the boot blocks. OSBS and > other multiboot loaders will infact set the chosen partition to > active, it is just the brain damaged FreeBSD boot code that ignores > the active flag, and chooses the FIRST BSD parition it finds > no matter what, which is _WRONG_. > I agree - I have a real need for this functionality right now. It would be very nice for the BSD boot code to check the active flag - can this be put into -current ? Tim. From owner-freebsd-current Tue Jul 29 16:36:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA21506 for current-outgoing; Tue, 29 Jul 1997 16:36:43 -0700 (PDT) Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA21499 for ; Tue, 29 Jul 1997 16:36:41 -0700 (PDT) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.8.5/8.8.5) id TAA16191; Tue, 29 Jul 1997 19:36:33 -0400 (EDT) Date: Tue, 29 Jul 1997 19:36:33 -0400 (EDT) From: Garrett Wollman Message-Id: <199707292336.TAA16191@khavrinen.lcs.mit.edu> To: Peter Dufault Cc: current@FreeBSD.ORG Subject: where to put access restriction for scheduling classes In-Reply-To: <199707292031.QAA28302@hda.hda.com> References: <199707292031.QAA28302@hda.hda.com> Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk < said: > right thing a sysctl variable that somehow gets set during login > via login_class while still root? Which of those are inherited > per-process? Or do I actually add a resource and only change the > login stuff? I'm sorry, I can't parse this. sysctl is a SYStem ConTroL mechanism. Nothing it does is per-process. It sounds like what you want is either a new system call to set some flags in the proc structure, or a new sort of resource limit (if what you're trying to restrict can easily be modeled in terms of unsigned, nondecreasing integer values) (which effectively does the same thing only with a bit more pre-existing infrastructure). -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick From owner-freebsd-current Tue Jul 29 17:35:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA24817 for current-outgoing; Tue, 29 Jul 1997 17:35:22 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id RAA24812 for ; Tue, 29 Jul 1997 17:35:20 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id RAA03703; Tue, 29 Jul 1997 17:33:08 -0700 From: Terry Lambert Message-Id: <199707300033.RAA03703@phaeton.artisoft.com> Subject: Re: pccard and -current; a long way to go. :-( To: jdp@polstra.com (John Polstra) Date: Tue, 29 Jul 1997 17:33:08 -0700 (MST) Cc: msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG In-Reply-To: <199707290057.RAA29178@austin.polstra.com> from "John Polstra" at Jul 28, 97 05:57:03 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > The most recent work I did was to make the kernel compile, boot, > and run when built as an ELF executable. That was met with a great > big collective yawn when I committed it. All available evidence > indicates that nobody else gives a tinker's damn about FreeBSD-ELF > except as something fun to argue about passionately. (The people > working on the alpha port are probably going to care about it some > day, whenever they get around to that part.) I'm not complaining, > but I'm also not feeling encouraged to expend any more effort on > it. I think you need to document what needs to be done here. I would be happy to run an ELF kernel in my now (mostly) non-ELF environment. Did you use the "fast call" conventions on the ELF kernel system calls? This would imply a libc rebuild, right? (Some of us are pro-ELF for reasons other than toying with it 8-)). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Tue Jul 29 17:57:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA25705 for current-outgoing; Tue, 29 Jul 1997 17:57:42 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id RAA25700 for ; Tue, 29 Jul 1997 17:57:38 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id RAA03727; Tue, 29 Jul 1997 17:35:22 -0700 From: Terry Lambert Message-Id: <199707300035.RAA03727@phaeton.artisoft.com> Subject: Re: Multiple FreeBSD Systems on a Single Disk To: bde@zeta.org.au (Bruce Evans) Date: Tue, 29 Jul 1997 17:35:22 -0700 (MST) Cc: bde@zeta.org.au, rgrimes@GndRsh.aac.dev.com, abial@korin.warman.org.pl, current@FreeBSD.ORG, sherwink@ix.netcom.com In-Reply-To: <199707290502.PAA18490@godzilla.zeta.org.au> from "Bruce Evans" at Jul 29, 97 03:02:45 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > >And should be fixed (I've mentioned this before) to boot from the > >first ACTIVE (flag 0x80) FreeBSD slice. Should be a trival amount > > That would mainly break certain configurations. The active flag > should only be set for one partition, so you shouldn't have any active > FreeBSD partitions if you have an active DOS partition. The boot managers (at least the one he's been pushing) set the active partition to the one selected for boot. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Tue Jul 29 18:05:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA26089 for current-outgoing; Tue, 29 Jul 1997 18:05:26 -0700 (PDT) Received: from internet1.mel.cybec.com.au (internet1.mel.cybec.com.au [203.103.154.130]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA26070 for ; Tue, 29 Jul 1997 18:04:55 -0700 (PDT) Received: from tech34 (tech34.mel.cybec.com.au [203.103.154.37]) by internet1.mel.cybec.com.au (post.office MTA v2.0 0813 ID# 0-14031) with ESMTP id AAA583; Wed, 30 Jul 1997 11:05:36 +1000 Message-ID: <33DE939C.E3907FDE@cybec.com.au> Date: Wed, 30 Jul 1997 11:06:36 +1000 From: TLiddelow@cybec.com.au (Tim Liddelow) X-Mailer: Mozilla 4.01 [en] (WinNT; I) MIME-Version: 1.0 To: John Polstra CC: msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( X-Priority: 3 (Normal) References: <199707261527.AAA28100@genesis.atrad.adelaide.edu.au> <199707290057.RAA29178@austin.polstra.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk John Polstra wrote: > > > The most recent work I did was to make the kernel compile, boot, > and run when built as an ELF executable. That was met with a great > big collective yawn when I committed it. All available evidence > indicates that nobody else gives a tinker's damn about FreeBSD-ELF > except as something fun to argue about passionately. (The people > working on the alpha port are probably going to care about it some > day, whenever they get around to that part.) I'm not complaining, > but I'm also not feeling encouraged to expend any more effort on > it. Well that's a shame John; you may remember my enthusiasm a few years back regarding your efforts and I still feel that way. I just think that perhaps a little bit of PR is in order - why is ELF a good thing? (I know, but perhaps other people don't). What would be the benefits of moving to ELF ? Perhaps you've already done this ... I don't know. But it is a great effort that you've managed to get both libc and the kernel moved to ELF - especially if FreeBSD is going to move to other platforms. Keep up the good work - it's easy to get discouraged - enthusiasm is one of the best traits one can have in a project such as this. Cheers Tim. From owner-freebsd-current Tue Jul 29 18:19:02 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA26848 for current-outgoing; Tue, 29 Jul 1997 18:19:02 -0700 (PDT) Received: from hda.hda.com (hda-bicnet.bicnet.net [208.220.66.37]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA26840 for ; Tue, 29 Jul 1997 18:18:55 -0700 (PDT) Received: (from dufault@localhost) by hda.hda.com (8.8.5/8.8.5) id UAA28927; Tue, 29 Jul 1997 20:33:12 -0400 (EDT) From: Peter Dufault Message-Id: <199707300033.UAA28927@hda.hda.com> Subject: Re: where to put access restriction for scheduling classes In-Reply-To: <199707292336.TAA16191@khavrinen.lcs.mit.edu> from Garrett Wollman at "Jul 29, 97 07:36:33 pm" To: wollman@khavrinen.lcs.mit.edu (Garrett Wollman) Date: Tue, 29 Jul 1997 20:33:11 -0400 (EDT) Cc: dufault@hda.com, current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > < said: > > > right thing a sysctl variable that somehow gets set during login > > via login_class while still root? Which of those are inherited > > per-process? Or do I actually add a resource and only change the > > login stuff? > > I'm sorry, I can't parse this. Let the apologies be mine. > sysctl is a SYStem ConTroL mechanism. Nothing it does is per-process. You are right. I thought that resource limits had been inhumed by sysctl, and that some were inherited. > It sounds like what you want is either a new system call to set some > flags in the proc structure, or a new sort of resource limit (if what > you're trying to restrict can easily be modeled in terms of unsigned, > nondecreasing integer values) (which effectively does the same thing > only with a bit more pre-existing infrastructure). Yes, I want something similar to number of open files per proc, as currently set by limit openfiles, without having to modify all shells in exhistence thus missing the ones I don't know about. To help with the parsing, by "only change login stuff" I meant: "Shall I add another resource limit that requires changing all shells, thus being a big pain in the neck since there must be a new gnu shell I don't know about, or, since resource limits appear to be the appropriate mechanism for what IS after all a process group resource limit in spite of this support issue, shall I add a resource limit to FreeBSD and change only that part of the FreeBSD login mechanism that is modifying resource limits at login time while still effectively root thus ignore this portability af shells issue, in particular ignoring the other 4.4 derived OS's?" Or something like that. Peter -- Peter Dufault (dufault@hda.com) Realtime development, Machine control, HD Associates, Inc. Safety critical systems, Agency approval From owner-freebsd-current Tue Jul 29 18:27:15 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA27159 for current-outgoing; Tue, 29 Jul 1997 18:27:15 -0700 (PDT) Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA27154 for ; Tue, 29 Jul 1997 18:27:11 -0700 (PDT) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.8.5/8.8.5) id VAA16803; Tue, 29 Jul 1997 21:26:58 -0400 (EDT) Date: Tue, 29 Jul 1997 21:26:58 -0400 (EDT) From: Garrett Wollman Message-Id: <199707300126.VAA16803@khavrinen.lcs.mit.edu> To: Peter Dufault Cc: current@FreeBSD.ORG Subject: Re: where to put access restriction for scheduling classes In-Reply-To: <199707300033.UAA28927@hda.hda.com> References: <199707292336.TAA16191@khavrinen.lcs.mit.edu> <199707300033.UAA28927@hda.hda.com> Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk < said: > "Shall I add another resource limit that requires changing all > shells, [...] or [...] shall I add a resource limit to FreeBSD and > change only that part of the FreeBSD login mechanism that is > modifying resource limits at login time while still effectively root > thus ignore this portability af shells issue, in particular ignoring > the other 4.4 derived OS's?" No shell should break as a result of defining a new resource limit. Therefore, you should update setusercontext(3) and the two system shells, and let the other shells fend for themselves. -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick From owner-freebsd-current Tue Jul 29 20:14:13 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA02854 for current-outgoing; Tue, 29 Jul 1997 20:14:13 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA02845 for ; Tue, 29 Jul 1997 20:14:06 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id NAA28631; Wed, 30 Jul 1997 13:06:06 +1000 Date: Wed, 30 Jul 1997 13:06:06 +1000 From: Bruce Evans Message-Id: <199707300306.NAA28631@godzilla.zeta.org.au> To: bde@zeta.org.au, rgrimes@GndRsh.aac.dev.com Subject: Re: Multiple FreeBSD Systems on a Single Disk Cc: abial@korin.warman.org.pl, current@FreeBSD.ORG, sherwink@ix.netcom.com Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> That would mainly break certain configurations. The active flag >> should only be set for one partition, so you shouldn't have any active >> FreeBSD partitions if you have an active DOS partition. > >Right, and if I have an active DOS partition the BIOS is going to >boot the ACTIVE dos partition. It won't even load the BSD boot >blocks if that is not the active partition. > >Any configuration that was broken by this change was broken such >that the BIOS should have not even booted anything! The BIOS should Only most configurations. >have returned a ``No Active Partition'', or gone on to the next >disk drive (yes, AMI bios'es as old as 1982 can boot from drive >0x81 if there are no active partitions on drive 0x80.) Booting is mostly handled by the bytes in the MBR, which have nothing to do with the BIOS. I believe BIOSes that support booting from any disk look at the 0xAA55 boot signature and not at the active flag. There is no active flags for floppies (unless the MBR supports them). >> This should be fixed by adding slice support to the boot loader's >> name parser. Then you could bounce off the first FreeBSD to any other >> slice by putting a boot loader on the first slice and configuring it >> to boot from slice sN, e.g., "0:sd(0,s5,a)kernel" to boot the first >> logical drive on an extended slice. >Poppy cock, the active flag model should work and not require the >addition of slice choosing support to the boot blocks. OSBS and >other multiboot loaders will infact set the chosen partition to >active, it is just the brain damaged FreeBSD boot code that ignores >the active flag, and chooses the FIRST BSD parition it finds >no matter what, which is _WRONG_. The active flag model only works when the MBR supports it, and it is less flexible. In the active flag model, the default boot slice is stored as an active flag, and non-defaults are specified by typing the slice number and passing the choice to the secondary boot loader. In the configured boot lader model, the default boot drive, slice, partition and kernel name is stored in /boot.config in the first FreeBSD slice, and non-defaults are specified by typing them at the secondary boot prompt. Bruce From owner-freebsd-current Tue Jul 29 20:49:12 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA04562 for current-outgoing; Tue, 29 Jul 1997 20:49:12 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA04554 for ; Tue, 29 Jul 1997 20:49:08 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id NAA17840; Wed, 30 Jul 1997 13:18:29 +0930 (CST) From: Michael Smith Message-Id: <199707300348.NAA17840@genesis.atrad.adelaide.edu.au> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707290509.BAA23557@smoke.marlboro.vt.us> from john hood at "Jul 29, 97 01:09:40 am" To: cgull@smoke.marlboro.vt.us (john hood) Date: Wed, 30 Jul 1997 13:18:29 +0930 (CST) Cc: cgull@smoke.marlboro.vt.us, sos@sos.freebsd.dk, msmith@atrad.adelaide.edu.au, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk john hood stands accused of saying: > "but it works fine for me!" > > for the rest of you: > > http://smoke.marlboro.vt.us/pub/cgull/starBSD/sff8038/ Works pretty well here too. I picked up Soren's commit as soon as it hit the repo. CPU: Pentium Pro (199.42-MHz 686-class CPU) Origin = "GenuineIntel" Id = 0x619 Stepping=9 Features=0xfbff real memory = 16777216 (16384K bytes) avail memory = 14647296 (14304K bytes) ... ide_pci0: rev 0x00 on pci0.7.1 ... wdc0 at 0x1f0-0x1f7 irq 14 flags 0xa0ffa0ff on isa wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-16 wd0: 3098MB (6346368 sectors), 6296 cyls, 16 heads, 63 S/T, 512 B/S Without DMA enabled : # dd if=/dev/rwd0 of=/dev/null bs=1m count=200 209715200 bytes transferred in 30.751826 secs (6819602 bytes/sec) systat -vmstat excerpt : 1.2%Sys 81.0%Intr 0.2%User 0.0%Nice 17.6%Idl With DMA enabled : # dd if=/dev/rwd0 of=/dev/null bs=1m count=200 209715200 bytes transferred in 31.191724 secs (6723424 bytes/sec) systat -vmstat excerpt : 1.1%Sys 0.3%Intr 0.2%User 0.0%Nice 98.4%Idl No more throughput, but all of a sudden I have a CPU again. Lovely! -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Tue Jul 29 21:14:13 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id VAA06035 for current-outgoing; Tue, 29 Jul 1997 21:14:13 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA06029 for ; Tue, 29 Jul 1997 21:14:06 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id VAA07616; Tue, 29 Jul 1997 21:12:58 -0700 (PDT) Message-Id: <199707300412.VAA07616@austin.polstra.com> To: Terry Lambert cc: msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( In-reply-to: Your message of "Tue, 29 Jul 1997 17:33:08 PDT." <199707300033.RAA03703@phaeton.artisoft.com> Date: Tue, 29 Jul 1997 21:12:58 -0700 From: John Polstra Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > I think you need to document what needs to be done here. I would > be happy to run an ELF kernel in my now (mostly) non-ELF environment. Well, there are a zillion things that need to be done. Just off the top of my head: * Solve the booting problem. My hacked-up bootloader was within about 6 bytes of the maximum size -- way too big to commit. GRUB? An a.out wrapper for the ELF kernel? * Design a directory structure for the toolchain components so that it's reasonable to have both a.out and ELF tools installed at the same time. * Work out the various problems that stem from the fact that a.out prepends a leading underscore to global symbol names while ELF does not. (The usual reply to this is "let's just make our ELF tools prepend the underscore too," although I bet you know better than that. You're not allowed to do it because the ELF spec says you can't. If we want ELF, then let's have ELF and not elf.out or a.elf.) The underscore problem affects lots of areas: - assembly language code that references C globals - all programs that call nlist() - all programs that call dlopen() There are versions of nlist floating around that address this problem, and dlopen() could be made to do the right thing most of the time too. * Work out a build structure that will allow an a.out->ELF make world. * Consider the feasibility of mixed linking combining both ELF and a.out object modules and libraries. Likewise for dynamic linking. Some people have a.out libraries and executables which they paid money for. They don't want to throw them away. Again, the underscore problem is an obstacle here. > Did you use the "fast call" conventions on the ELF kernel system > calls? This would imply a libc rebuild, right? Are you talking about this in "src/lib/libc/i386/SYS.h"? #ifdef __ELF__ #define KERNCALL int $0x80 /* Faster */ #else #define KERNCALL LCALL(7,0) /* The old way */ #endif And by the way, can somebody explain why system calls care about the object file format of the application making the calls? It's not obvious to me. -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Tue Jul 29 21:30:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id VAA06803 for current-outgoing; Tue, 29 Jul 1997 21:30:43 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA06798 for ; Tue, 29 Jul 1997 21:30:34 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id OAA31416; Wed, 30 Jul 1997 14:26:49 +1000 Date: Wed, 30 Jul 1997 14:26:49 +1000 From: Bruce Evans Message-Id: <199707300426.OAA31416@godzilla.zeta.org.au> To: bde@zeta.org.au, terry@lambert.org Subject: Re: Multiple FreeBSD Systems on a Single Disk Cc: abial@korin.warman.org.pl, current@FreeBSD.ORG, rgrimes@GndRsh.aac.dev.com, sherwink@ix.netcom.com Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> >And should be fixed (I've mentioned this before) to boot from the >> >first ACTIVE (flag 0x80) FreeBSD slice. Should be a trival amount >> >> That would mainly break certain configurations. The active flag >> should only be set for one partition, so you shouldn't have any active >> FreeBSD partitions if you have an active DOS partition. > >The boot managers (at least the one he's been pushing) set the >active partition to the one selected for boot. Mine doesn't :-). It stores the default partition number in its code and ignores the active flags. It passes a pointer to the partition being booted in es:si (same as booteasy), but the secondary bootstrap shouldn't assume this. Bruce From owner-freebsd-current Tue Jul 29 21:30:52 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id VAA06826 for current-outgoing; Tue, 29 Jul 1997 21:30:52 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA06818 for ; Tue, 29 Jul 1997 21:30:49 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id VAA07677; Tue, 29 Jul 1997 21:29:49 -0700 (PDT) Message-Id: <199707300429.VAA07677@austin.polstra.com> To: TLiddelow@cybec.com.au (Tim Liddelow) cc: msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( In-reply-to: Your message of "Wed, 30 Jul 1997 11:06:36 +1000." <33DE939C.E3907FDE@cybec.com.au> Date: Tue, 29 Jul 1997 21:29:49 -0700 From: John Polstra Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > > I'm not complaining, > > but I'm also not feeling encouraged to expend any more effort on > > it. > > Well that's a shame John; you may remember my enthusiasm a few years > back regarding your efforts and I still feel that way. I just think > that perhaps a little bit of PR is in order - why is ELF a good thing? > (I know, but perhaps other people don't). What would be the benefits > of moving to ELF ? Perhaps you've already done this ... I don't > know. The people who recognize the benefits of moving to ELF are the ones who need to do the PR. The question is not "would ELF be better?" but rather "would ELF be _enough_ better to make it worth the pain of the transition?" Focusing on the language tools as I have, I see certain benefits: * Our linker has major design faults in its handling of shared libraries. They're probably never going to be fixed without a rewrite of the whole thing or at least the shared library code. The ELF linker doesn't have these faults. * The ELF tools are maintained by people besides us. * There are many bugs in g++ which cause the dreaded "relocation burbs" for a.out. These bugs happen to make no difference for ELF. They are in areas of the compiler that are a total mess and that are extremely complicated. I have already fixed the ones that are "easy enough" to fix. Because they affect only us, the remaining bugs are never going to be fixed, in my opinion. These benefits are real, but I'm not sure they're enough. (If you are looking for a facts-be-damned advocate of FreeBSD-ELF, I'm not your guy.) There is plenty of room for intelligent disagreement about it. I myself have about three opinions on the subject, all in conflict with one another. :-) I am sure there would be other benefits, but I don't know enough about them to lobby on their behalf. The people who do are the ones who need to do the PR. Explain why FreeBSD would be a better place if it used ELF. So far nobody's done a very good job of that. "Pageable kernel data areas?" Yawn. What does it mean to me as a user or as a developer? Would the system run faster? Use less memory? How much faster? How much less? Who's going to implement it if we switch to ELF? Enquiring minds want to know. John -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Tue Jul 29 22:02:33 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA08328 for current-outgoing; Tue, 29 Jul 1997 22:02:33 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA08323 for ; Tue, 29 Jul 1997 22:02:26 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id PAA32716; Wed, 30 Jul 1997 15:02:05 +1000 Date: Wed, 30 Jul 1997 15:02:05 +1000 From: Bruce Evans Message-Id: <199707300502.PAA32716@godzilla.zeta.org.au> To: freebsd-current@FreeBSD.ORG, rjk@grauel.com Subject: Re: Can't mount CD under -current. Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >With a -current kernel as of 7/28, I can no longer mount a CD. I remember >seeing comments about breakages when the webnfs stuff went in; I hope that >this was just overlooked. The comments gave the fix: just use current binaries with current kernels. This is sometimes incconvenient, but anything else is not current. Bruce From owner-freebsd-current Tue Jul 29 22:09:28 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA08608 for current-outgoing; Tue, 29 Jul 1997 22:09:28 -0700 (PDT) Received: from rover.village.org (rover.village.org [204.144.255.49]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id WAA08603 for ; Tue, 29 Jul 1997 22:09:25 -0700 (PDT) Received: from rover.village.org [127.0.0.1] by rover.village.org with esmtp (Exim 1.60 #1) id 0wtQzY-0000LO-00; Tue, 29 Jul 1997 23:08:08 -0600 To: John Polstra Subject: Re: pccard and -current; a long way to go. :-( Cc: TLiddelow@cybec.com.au (Tim Liddelow), msmith@atrad.adelaide.edu.au, current@freebsd.org In-reply-to: Your message of "Tue, 29 Jul 1997 21:29:49 PDT." <199707300429.VAA07677@austin.polstra.com> References: <199707300429.VAA07677@austin.polstra.com> Date: Tue, 29 Jul 1997 23:08:07 -0600 From: Warner Losh Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In message <199707300429.VAA07677@austin.polstra.com> John Polstra writes: : The people who recognize the benefits of moving to ELF are the ones : who need to do the PR. The question is not "would ELF be better?" : but rather "would ELF be _enough_ better to make it worth the pain : of the transition?" I'm curious. Does anybody have a pure ELF FreeBSD system? Warner From owner-freebsd-current Tue Jul 29 22:12:35 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA08838 for current-outgoing; Tue, 29 Jul 1997 22:12:35 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA08826 for ; Tue, 29 Jul 1997 22:12:28 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id OAA18750; Wed, 30 Jul 1997 14:41:01 +0930 (CST) From: Michael Smith Message-Id: <199707300511.OAA18750@genesis.atrad.adelaide.edu.au> Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: from Warner Losh at "Jul 29, 97 11:08:07 pm" To: imp@village.org (Warner Losh) Date: Wed, 30 Jul 1997 14:41:01 +0930 (CST) Cc: jdp@polstra.com, TLiddelow@cybec.com.au, msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Warner Losh stands accused of saying: > In message <199707300429.VAA07677@austin.polstra.com> John Polstra writes: > : The people who recognize the benefits of moving to ELF are the ones > : who need to do the PR. The question is not "would ELF be better?" > : but rather "would ELF be _enough_ better to make it worth the pain > : of the transition?" > > I'm curious. Does anybody have a pure ELF FreeBSD system? Not yet. I'm tempted to try it though, or at least go as far as possible. One of the things that would help this _greatly_ is Sataoshi's New World Order patch, which hasn't seen much coverage lately. > Warner -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Tue Jul 29 22:59:52 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA11123 for current-outgoing; Tue, 29 Jul 1997 22:59:52 -0700 (PDT) Received: from smoke.marlboro.vt.us (smoke.marlboro.vt.us [198.206.215.91]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA11115; Tue, 29 Jul 1997 22:59:44 -0700 (PDT) Received: (from cgull@localhost) by smoke.marlboro.vt.us (8.8.5/8.8.5/cgull) id BAA02894; Wed, 30 Jul 1997 01:59:35 -0400 (EDT) Date: Wed, 30 Jul 1997 01:59:35 -0400 (EDT) Message-Id: <199707300559.BAA02894@smoke.marlboro.vt.us> From: john hood MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 To: SЬren Schmidt Cc: se@FreeBSD.ORG (Stefan Esser), freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707292224.AAA00624@sos.freebsd.dk> References: <19970729221036.52544@mi.uni-koeln.de> <199707292224.AAA00624@sos.freebsd.dk> X-Mailer: VM 6.31 under Emacs 19.34.2 Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from quoted-printable to 8bit by hub.freebsd.org id WAA11117 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk SЬren Schmidt writes: > In reply to Stefan Esser who wrote: > > But the "per char" read performance improvement indicates > > what's really going on: CPU cycles spent in the interrupt > > handler probably account for the 4.5MB/s mark in the PIO > > case, while raw data rate of the disk is approached with > > the DMA driver. > > Exactly! Well, bonnie reports an increase in CPU load and some increase in I/O speed on the hack box I developed the code on, while a little spin-loop-at-idle-priority hack i coded up shows an improvement in available CPU that it can consume. My conclusion: bonnie's CPU load numbers are useless, at least for IDE drives. Here's some more useless bonnie results. These are for a system with a K5-PR90, a "VXPro" chipset (a relabeled VIA Apollo VP-1), 32MB of RAM and two Quantum FB1080As. bash-2.00$ cat bonnie.pio -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU 100 2913 66.7 4432 25.9 1541 15.5 3764 55.1 4512 31.4 70.9 9.1 bash-2.00$ cat bonnie.dma -------Sequential Output-------- ---Sequential Input-- --Random-- -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU 100 3934 98.2 4290 28.0 1606 22.3 4538 86.2 4483 48.2 74.9 9.9 > > But writes are still significantly slower than reads, which > > makes me think, that one revolution of the media is lost per > > 64KB written ... Could be; the code path is more complex than would be good for best performance. I noticed that somebody long ago added commands to enable read and write caches on IDE drives, and then somebody commented them out. What happened there? Is it worth trying that again, or shall we just have another long discussion about filesystem commits and stability and recoverability and not do anything to the code? :) > I think it might be some kind of interaction with the drives > cache, but our driver might be guilty too.. > > > > Not bad, for a "simple" software upgrade :) > > > > True! But there is still room for improvement ... ;) > > Well, I don't think we are going to get much more performance > out of it, at least not in an order like this, but there sure > is room for cleaning up the driver, which I'll eventually get > to if/when I get the time... > > > > It has improved my worldstone by ~10% if that counts for real world use.. > > > > Yes, one thing that is often underestimated is the fact, > > that while PIO mode transfers consume only a small fraction > > of the cycles of a fast CPU, but they tend to consume them > > exactly at the time, when the CPU is highly loaded anyway. But said CPU is often waiting for the disk anyway. > Yep. Now we only need ATA drives that can overlap commands. It seems they > are getting there, but its a mess still. When that settles (and we get > new drives :) ), we can get even a wee bit more performance... Eeeugh. I think it's time for a new disk i/o standard that uses a simple high-speed serial link and starts over on the whole command set architecture thing. I had some hopes for Fibre Channel and SSA, but both of those use SCSI commands if I'm not mistaken. But I don't have an electrical engineering degree, and so nobody listens to my rants about hardware. :) --jh -- John Hood cgull@smoke.marlboro.vt.us Predictably, they all eventually wandered away, rubbing their bruises and brushing mud out of their hair. Some went off to work for the ESA, launching much smaller rockets into low orbits, while others elected to sit on their front porches drinking Jim Beam from the bottle and launching bottle rockets from the empties. [Jordan Hubbard] From owner-freebsd-current Wed Jul 30 00:10:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA14956 for current-outgoing; Wed, 30 Jul 1997 00:10:16 -0700 (PDT) Received: (from hsu@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA14949 for current; Wed, 30 Jul 1997 00:10:13 -0700 (PDT) Date: Wed, 30 Jul 1997 00:10:13 -0700 (PDT) From: Jeffrey Hsu Message-Id: <199707300710.AAA14949@hub.freebsd.org> To: current Subject: Re: pccard and -current; a long way to go. :-( Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > The people who recognize the benefits of moving to ELF are the ones > who need to do the PR. The question is not "would ELF be better?" > but rather "would ELF be _enough_ better to make it worth the pain > of the transition?" Focusing on the language tools as I have, I see > certain benefits: I work on a lot of language tools and the first thing I have to do is modify them to understand our ad-hoc a.out format. This argues strongly against using a non-standard binary format when the wealth of language tools and utilities are all ELF compatible. People argue that we don't need ELF because all the tools we have work fine with a.out. But, they are neglecting to consider all the tools we don't have because we don't use ELF. Now that we have more complete support in our system for ELF (I silently cheered when jdp committed the code in discussion), I intend to make full use of it. > * Our linker has major design faults in its handling of shared > libraries. They're probably never going to be fixed without > a rewrite of the whole thing or at least the shared library > code. The ELF linker doesn't have these faults. I see this problem too in conjunction with language tools. My next JDK port which I'm working on now (not sure if I should do a stable 1.1.4 port or just skip ahead to 1.2, neither of which the Linux people have yet) would happily use ELF if it solves the dlsym(RTLD_NEXT) problem (namely, we don't have it) and it also eliminates a bunch of kludges in the Java interpreter code that I have to make because we prepend underscores and the rest of the world doesn't. From owner-freebsd-current Wed Jul 30 00:30:48 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA15808 for current-outgoing; Wed, 30 Jul 1997 00:30:48 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA15802; Wed, 30 Jul 1997 00:30:42 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id RAA19811; Wed, 30 Jul 1997 17:00:28 +0930 (CST) From: Michael Smith Message-Id: <199707300730.RAA19811@genesis.atrad.adelaide.edu.au> Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707300710.AAA14949@hub.freebsd.org> from Jeffrey Hsu at "Jul 30, 97 00:10:13 am" To: hsu@FreeBSD.ORG (Jeffrey Hsu) Date: Wed, 30 Jul 1997 17:00:27 +0930 (CST) Cc: current@hub.freebsd.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Jeffrey Hsu stands accused of saying: > fine with a.out. But, they are neglecting to consider all the > tools we don't have because we don't use ELF. Now that we have Could you enumerate a few of these tools, for the edification of those of us that are ignorant of same? > I see this problem too in conjunction with language tools. My next > JDK port which I'm working on now (not sure if I should do a stable > 1.1.4 port or just skip ahead to 1.2, neither of which the Linux > people have yet) would happily use ELF if it solves the dlsym(RTLD_NEXT) > problem (namely, we don't have it) and it also eliminates a bunch > of kludges in the Java interpreter code that I have to make because > we prepend underscores and the rest of the world doesn't. RTLD_NEXT would help the Willows TWIN stuff as well. So I guess there you have two; win16/win32 crossdevelopment and Java support would both benefit from a move to the ELF toolchain. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 01:04:00 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA17670 for current-outgoing; Wed, 30 Jul 1997 01:04:00 -0700 (PDT) Received: from forbidden-donut.anet-stl.com (forbidden-donut.anet-stl.com [206.114.203.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id BAA17665 for ; Wed, 30 Jul 1997 01:03:55 -0700 (PDT) Received: from localhost (doogie@localhost) by forbidden-donut.anet-stl.com (8.8.6/8.8.5) with SMTP id DAA27151; Wed, 30 Jul 1997 03:00:28 GMT Date: Wed, 30 Jul 1997 03:00:26 +0000 (GMT) From: Jason Young To: john hood cc: SЬren Schmidt , freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707290424.AAA23424@smoke.marlboro.vt.us> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk -----BEGIN PGP SIGNED MESSAGE----- On Tue, 29 Jul 1997, john hood wrote: > I've been coding up bus-mastering support for EIDE drives, and I'm at > a point where I can release it to the world. I've got a patch against > 2.2.2 available and various notes for the enjoyment of the curious > kernel hacker. The wd driver is enough of a living fossil that the > patches should apply easily to any kernel since then, though. I'm not sure what this driver did to my 1GB Seagate, but all I have to say is buy this man a beer. :) The machine: AMD 5x86-133, 1GB seagate and 1.2GB WD EIDE drives, 32MB RAM. I ran these stats several times on an idle system without appreciable changes each time I ran them. The 'before' picture: %dd if=/dev/rwd0 of=/dev/null bs=1024 count=20000 20000+0 records in 20000+0 records out 20480000 bytes transferred in 37.234251 secs (550031 bytes/sec) %dd if=/dev/rwd0 of=/dev/null bs=1024 count=20000 20000+0 records in 20000+0 records out 20480000 bytes transferred in 19.194747 secs (1066959 bytes/sec) The 'after' picture: %dd if=/dev/rwd0 of=/dev/null bs=1024 count=20000 20000+0 records in 20000+0 records out 20480000 bytes transferred in 8.717158 secs (2349390 bytes/sec) %dd if=/dev/rwd0 of=/dev/null bs=1024 count=20000 20000+0 records in 20000+0 records out 20480000 bytes transferred in 19.363463 secs (1057662 bytes/sec) - ----- Copyright (c) 1992-1997 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-CURRENT #0: Wed Jul 30 01:44:36 CDT 1997 doogie@toybox.anet-stl.com:/usr/src/sys/compile/TOYBOX CPU: AMD Am5x86 Write-Back (486-class CPU) Origin = "AuthenticAMD" Id = 0x4f4 Stepping=4 Features=0x1 real memory = 33554432 (32768K bytes) avail memory = 31006720 (30280K bytes) Probing for devices on the ISA bus: sc0 at 0x60-0x6f irq 1 on motherboard sc0: VGA color <16 virtual consoles, flags=0x0> ed0 at 0x300-0x31f irq 5 on isa ed0: address 00:40:95:26:a1:d0, type NE2000 (16 bit) sio0 at 0x3f8-0x3ff irq 4 on isa sio0: type 16550A sio1 at 0x2f8-0x2ff irq 3 on isa sio1: type 16550A lpt0 at 0x378-0x37f irq 7 on isa lpt0: Interrupt-driven port lp0: TCP/IP capable interface fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa fdc0: NEC 72065B fd0: 1.44MB 3.5in wdc0 at 0x1f0-0x1f7 irq 14 flags 0xa0ffa0ff on isa wdc0: unit 0 (wd0): , 32-bit, multi-block-32 wd0: 1033MB (2116800 sectors), 2100 cyls, 16 heads, 63 S/T, 512 B/S wdc0: unit 1 (wd1): , 32-bit, multi-block-16 wd1: 1222MB (2503872 sectors), 2484 cyls, 16 heads, 63 S/T, 512 B/S wdc1 not found at 0x170 npx0 on motherboard npx0: INT 16 interface I'll be happy to answer any questions or do any additional testing that anyone's interested in. Jason Young ANET Technical Staff -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQB1AwUBM96uTKInE6ybC66VAQFX5AL/UKNaLG1pr71QTYxqWIhcJQMcyp2D21iz 0vYZ0L3keCqlq0ZRU7nFJUbHexgBaSBhkOMTc6xROBOi/hBrfEAfAC+tzoDIc3Fz XxHdWSUSJ4TXW00ez8QIBFDs3VKaibek =BhEi -----END PGP SIGNATURE----- From owner-freebsd-current Wed Jul 30 01:29:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id BAA18820 for current-outgoing; Wed, 30 Jul 1997 01:29:36 -0700 (PDT) Received: from karon.dynas.se (karon.dynas.se [192.71.43.4]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id BAA18815 for ; Wed, 30 Jul 1997 01:29:33 -0700 (PDT) Received: (qmail 19021 invoked from network); 30 Jul 1997 08:29:27 -0000 Received: from spirit.dynas.se (HELO host.domain) (172.16.1.10) by karon.dynas.se with SMTP; 30 Jul 1997 08:29:27 -0000 Received: by spirit (Smail3.1.28.1 #32) id m0wtU8M-000JemC; Wed, 30 Jul 97 10:29:26 +0200 Date: Wed, 30 Jul 1997 10:29:26 +0200 (MET DST) From: Mikael Hybsch Reply-To: Mikael Hybsch To: freebsd-current@freebsd.org Subject: Re: aha0: DMA beyond end Of ISA In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk It now looks as if the problem is in the function scsi_read_capacity() in scsi_base.c. It calls scsi_scsi_cmd() with the address of a local variable (rdcap) as the data_addr argument. By printing &rdcap and vtophys(&rdcap) while trying to mount the cdrom and running programs in the background that allocate memory, the physical address of rdcap eventually exceeded 16M and the mount failed. If the kernel stack is allowed to be allocated above 16M, scsi_read_capacity() needs some special code to handle the BOUNCE_BUFFERS option. On Mon, 28 Jul 1997, I wrote: > I have a 486 with 24M memory and a 1542B connected to a HP 4020i. > > Frequently when I try to access the CD or burn a new CD, I get the message > aha0: DMA beyond end Of ISA: 0x1154d24 > or other addresses above 0x1000000. > > By repeating the mount command it will eventually succeed. > > I'm running 3.0-current as of some days ago and > have the BOUNCE_BUFFERS option in the kernel, so what > could be causing this? -- Mikael Hybsch Email: micke@dynas.se DynaSoft, Dynamic Software AB Phone: +46-8-7250900 Box 10704 Fax: +46-8-6494970 S-121 29 STOCKHOLM, SWEDEN From owner-freebsd-current Wed Jul 30 02:03:02 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id CAA20338 for current-outgoing; Wed, 30 Jul 1997 02:03:02 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id CAA20330 for ; Wed, 30 Jul 1997 02:02:56 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id SAA21195; Wed, 30 Jul 1997 18:31:21 +0930 (CST) From: Michael Smith Message-Id: <199707300901.SAA21195@genesis.atrad.adelaide.edu.au> Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707300412.VAA07616@austin.polstra.com> from John Polstra at "Jul 29, 97 09:12:58 pm" To: jdp@polstra.com (John Polstra) Date: Wed, 30 Jul 1997 18:31:21 +0930 (CST) Cc: terry@lambert.org, msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk John Polstra stands accused of saying: > * Solve the booting problem. My hacked-up bootloader was within > about 6 bytes of the maximum size -- way too big to commit. GRUB? > An a.out wrapper for the ELF kernel? Someone (please) help me get a standalone executable running? ie. like the boot3 code Bill posted last year? > * Design a directory structure for the toolchain components so that > it's reasonable to have both a.out and ELF tools installed at the > same time. Toolchain : /usr/libexec/- So we might have : /usr/libexec/i386-aout/* /usr/libexec/i386-elf/* /usr/libexec/alpha-elf/* Libraries : /usr/lib-- and a symlink /usr/lib -> the default for the running platform. eg : /usr/lib-i386-aout /usr/lib-i386-elf /usr/lib-alpha-elf > * Work out a build structure that will allow an a.out->ELF make > world. I presume it's possible to build the ELF toolchain as a.out executables? I'd hope that we could handle all the host/target combinations, actually. The above structure doesn't allow for sharing /usr between different architectures; that would seem silly given that the contents of bin and libexec wouldn't work. Can you answer me an ignorant question and explain how many things have to change in the gcc makefile to swap between generating a.out and ELF? > * Consider the feasibility of mixed linking combining both ELF and > a.out object modules and libraries. Likewise for dynamic linking. > Some people have a.out libraries and executables which they paid > money for. They don't want to throw them away. Again, the > underscore problem is an obstacle here. Is this critical? Can't the people with a.out libraries/executables continue to use a.out tools? -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 03:50:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA23891 for current-outgoing; Wed, 30 Jul 1997 03:50:16 -0700 (PDT) Received: from smoke.marlboro.vt.us (smoke.marlboro.vt.us [198.206.215.91]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA23883 for ; Wed, 30 Jul 1997 03:50:11 -0700 (PDT) Received: (from cgull@localhost) by smoke.marlboro.vt.us (8.8.5/8.8.5/cgull) id GAA03660; Wed, 30 Jul 1997 06:49:54 -0400 (EDT) Date: Wed, 30 Jul 1997 06:49:54 -0400 (EDT) Message-Id: <199707301049.GAA03660@smoke.marlboro.vt.us> From: john hood MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Jason Young Cc: john hood , SЬren Schmidt , freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: References: <199707290424.AAA23424@smoke.marlboro.vt.us> X-Mailer: VM 6.31 under Emacs 19.34.2 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Jason Young writes: > I'm not sure what this driver did to my 1GB Seagate, but all I have to say > is buy this man a beer. :) The machine: AMD 5x86-133, 1GB seagate and > 1.2GB WD EIDE drives, 32MB RAM. I ran these stats several times on an > idle system without appreciable changes each time I ran them. well, i won't pass up a free beer, but my code didn't do it. :) you have a 486-class machine that doesn't have busmastering IDE. it's not even a PCI bus machine. you just happened to turn on the already-existing multi-block and 32-bit PIO flags in your latest kernel upgrade. folks: performance with the default flags on the wd driver is *terrible*, and the flags should always be configured for your machine. see LINT for details. --jh -- John Hood cgull@smoke.marlboro.vt.us Predictably, they all eventually wandered away, rubbing their bruises and brushing mud out of their hair. Some went off to work for the ESA, launching much smaller rockets into low orbits, while others elected to sit on their front porches drinking Jim Beam from the bottle and launching bottle rockets from the empties. [Jordan Hubbard] From owner-freebsd-current Wed Jul 30 04:49:05 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA25860 for current-outgoing; Wed, 30 Jul 1997 04:49:05 -0700 (PDT) Received: from cenotaph.snafu.de (gw-deadnet.snafu.de [194.121.229.33]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id EAA25791; Wed, 30 Jul 1997 04:47:21 -0700 (PDT) Received: by cenotaph.snafu.de from deadline.snafu.de using smtp id m0wtXDi-000KB8C; Wed, 30 Jul 1997 13:47:10 +0200 (CEST) (Smail-3.2 1996-Jul-4 #1) Received: by deadline.snafu.de id m0wtXDg-0004nDC; Wed, 30 Jul 1997 13:47:08 +0200 (CEST) (Smail-3.2 1996-Jul-4 #1) Message-Id: From: root@deadline.snafu.de (Andreas S. Wetzel) Subject: COMPAT_LINUX + -current To: smp@freebsd.org Date: Wed, 30 Jul 1997 13:47:07 +0200 (CEST) Cc: current@freebsd.org Organization: A world stranger than you have ever imagined. X-Mailer: ELM [version 2.4ME+ PL13] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi! --- Some days ago I reported problems with -current and the linux emulator code. I have now updated to -current as of 30/07/97 and rebuild world and everything seems to run fine again apart from the Linux QuakeWorld client which doesn't run because of: Linux-emul(267): ioperm() not supported So I think there hasn't been a real problem with -current but a problem with my last build. Sorry for your inconvenience. Regards, Mickey -- (__) (@@) Andreas S. Wetzel Mail: mickey@deadline.snafu.de /-------\/ Utrechter Strasse 41 Web: http://cenotaph.snafu.de/ / | || 13347 Berlin Fon: <+4930> 456 066 90 * ||----|| Germany Fax: <+4930> 456 066 91/92 ~~ ~~ From owner-freebsd-current Wed Jul 30 04:51:44 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA26089 for current-outgoing; Wed, 30 Jul 1997 04:51:44 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA26084 for ; Wed, 30 Jul 1997 04:51:41 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.84]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA119; Wed, 30 Jul 1997 07:41:53 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id HAA00677; Wed, 30 Jul 1997 07:26:16 -0400 (EDT) Message-ID: <19970730072456.05185@scsn.net> Date: Wed, 30 Jul 1997 07:24:56 -0400 From: "Donald J. Maddox" To: john hood Cc: Jason Young , Sxren Schmidt , freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches Reply-To: dmaddox@scsn.net References: <199707290424.AAA23424@smoke.marlboro.vt.us> <199707301049.GAA03660@smoke.marlboro.vt.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 In-Reply-To: <199707301049.GAA03660@smoke.marlboro.vt.us>; from john hood on Wed, Jul 30, 1997 at 06:49:54AM -0400 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, Jul 30, 1997 at 06:49:54AM -0400, john hood wrote: > Jason Young writes: > > I'm not sure what this driver did to my 1GB Seagate, but all I have to say > > is buy this man a beer. :) The machine: AMD 5x86-133, 1GB seagate and > > 1.2GB WD EIDE drives, 32MB RAM. I ran these stats several times on an > > idle system without appreciable changes each time I ran them. > > well, i won't pass up a free beer, but my code didn't do it. :) > > you have a 486-class machine that doesn't have busmastering IDE. it's > not even a PCI bus machine. you just happened to turn on the > already-existing multi-block and 32-bit PIO flags in your latest > kernel upgrade. > > folks: performance with the default flags on the wd driver is > *terrible*, and the flags should always be configured for your > machine. see LINT for details. Anybody know why I'm seeing the following after building a new kernel with the EIDE bus master patches? Do I need to add something to my config to activate this, or ??? Copyright (c) 1992-1997 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-CURRENT #0: Tue Jul 29 22:29:35 EDT 1997 root@rhiannon.scsn.net:/usr/src/sys/compile/RHIANNON CPU: Pentium (166.19-MHz 586-class CPU) Origin = "GenuineIntel" Id = 0x52c Stepping=12 Features=0x1bf real memory = 33554432 (32768K bytes) avail memory = 30920704 (30196K bytes) Probing for devices on PCI bus 0: chip0: rev 0x01 on pci0.0.0 chip1: rev 0x00 on pci0.7.0 ide_pci0: rev 0x00 on pci0.7.1 ide_pci: warning, ide0:0 not configured for DMA? ide_pci: warning, ide0:1 not configured for DMA? ide_pci: warning, ide1:0 not configured for DMA? ide_pci: warning, ide1:1 not configured for DMA? ahc0: rev 0x00 int a irq 12 on pci0.10.0 ahc0: aic7880 Wide Channel, SCSI Id=7, 16/255 SCBs scbus0 at ahc0 bus 0 ahc0: target 0 Tagged Queuing Device sd0 at scbus0 target 0 lun 0 sd0: type 0 fixed SCSI 2 sd0: Direct-Access 2040MB (4178874 512 byte sectors) sd0: with 2708 cyls, 19 heads, and an average 81 sectors/track ahc0: target 1 Tagged Queuing Device sd1 at scbus0 target 1 lun 0 sd1: type 0 fixed SCSI 2 sd1: Direct-Access 507MB (1039329 512 byte sectors) sd1: with 2380 cyls, 6 heads, and an average 72 sectors/track vga0: rev 0x00 int a irq 11 on pci0.12.0 Probing for devices on the ISA bus: sc0 at 0x60-0x6f irq 1 on motherboard sc0: VGA color <16 virtual consoles, flags=0x0> lpt0 at 0x378-0x37f irq 7 on isa lpt0: Interrupt-driven port lp0: TCP/IP capable interface sio0 at 0x3f8-0x3ff irq 4 flags 0x10 on isa sio0: type 16550A sio1 at 0x2f8-0x2ff irq 3 on isa sio1: type 16550A pca0 on motherboard pca0: PC speaker audio driver wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ff80ff on isa wdc0: unit 0 (wd0): , 32-bit, multi-block-32 wd0: 1277MB (2615760 sectors), 2595 cyls, 16 heads, 63 S/T, 512 B/S wdc1 at 0x170-0x177 irq 15 flags 0x80ff80ff on isa wdc1: unit 0 (atapi): , removable, iordy atapi1.0: unknown phase fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa fdc0: NEC 72065B fd0: 1.44MB 3.5in npx0 on motherboard npx0: INT 16 interface sb0 at 0x220-0x22f irq 5 drq 1 on isa sb0: sbxvi0 drq 5 on isa sbxvi0: sbmidi0 at 0x330-0x331 on isa sbmidi0: opl0 at 0x388-0x38b on isa opl0: joy0 at 0x201 on isa joy0: joystick From owner-freebsd-current Wed Jul 30 04:58:30 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA26301 for current-outgoing; Wed, 30 Jul 1997 04:58:30 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA26292 for ; Wed, 30 Jul 1997 04:58:27 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id NAA02177; Wed, 30 Jul 1997 13:58:02 +0200 (MEST) From: SЬren Schmidt Message-Id: <199707301158.NAA02177@sos.freebsd.dk> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970730072456.05185@scsn.net> from "Donald J. Maddox" at "Jul 30, 97 07:24:56 am" To: dmaddox@scsn.net Date: Wed, 30 Jul 1997 13:58:02 +0200 (MEST) Cc: cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to Donald J. Maddox who wrote: > > Anybody know why I'm seeing the following after building a new kernel > with the EIDE bus master patches? Do I need to add something to my config > to activate this, or ??? Yes, as stated in the commit etc, you need to set the flags to 0xa0ffa0ff to get DMA support (if the drive and controller supports it). > wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ff80ff on isa ^^^^^^^^^^ > wdc1 at 0x170-0x177 irq 15 flags 0x80ff80ff on isa ^^^^^^^^^^ > wdc1: unit 0 (atapi): , removable, iordy -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- SЬren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end .. From owner-freebsd-current Wed Jul 30 05:03:24 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA26516 for current-outgoing; Wed, 30 Jul 1997 05:03:24 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA26511 for ; Wed, 30 Jul 1997 05:03:20 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.84]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA130; Wed, 30 Jul 1997 07:53:37 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id IAA00801; Wed, 30 Jul 1997 08:03:03 -0400 (EDT) Message-ID: <19970730080303.36552@scsn.net> Date: Wed, 30 Jul 1997 08:03:03 -0400 From: "Donald J. Maddox" To: =?iso-8859-1?Q?S=F8ren_Schmidt?= Cc: dmaddox@scsn.net, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches Reply-To: dmaddox@scsn.net References: <19970730072456.05185@scsn.net> <199707301158.NAA02177@sos.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Mutt 0.79 In-Reply-To: =?iso-8859-1?Q?=3C199707301158=2ENAA02177=40sos=2Efreebsd=2Edk=3E=3B_fro?= =?iso-8859-1?Q?m_S=F8ren_Schmidt_on_Wed=2C_Jul_30=2C_1997_at_01=3A58=3A0?= =?iso-8859-1?Q?2PM_+0200?= Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, Jul 30, 1997 at 01:58:02PM +0200, SЬren Schmidt wrote: > In reply to Donald J. Maddox who wrote: > > > > Anybody know why I'm seeing the following after building a new kernel > > with the EIDE bus master patches? Do I need to add something to my config > > to activate this, or ??? > > > Yes, as stated in the commit etc, you need to set the flags to 0xa0ffa0ff > to get DMA support (if the drive and controller supports it). > > > wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ff80ff on isa > ^^^^^^^^^^ > > wdc1 at 0x170-0x177 irq 15 flags 0x80ff80ff on isa > ^^^^^^^^^^ > > wdc1: unit 0 (atapi): , removable, iordy > Geez, how did I miss that? Thanks for the info :-) From owner-freebsd-current Wed Jul 30 05:18:02 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA27231 for current-outgoing; Wed, 30 Jul 1997 05:18:02 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA27222 for ; Wed, 30 Jul 1997 05:17:58 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.178]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA161; Wed, 30 Jul 1997 08:08:13 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id IAA00198; Wed, 30 Jul 1997 08:17:33 -0400 (EDT) Message-ID: <19970730081733.55871@scsn.net> Date: Wed, 30 Jul 1997 08:17:33 -0400 From: "Donald J. Maddox" To: =?iso-8859-1?Q?S=F8ren_Schmidt?= Cc: dmaddox@scsn.net, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches Reply-To: dmaddox@scsn.net References: <19970730072456.05185@scsn.net> <199707301158.NAA02177@sos.freebsd.dk> <19970730080303.36552@scsn.net> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Mutt 0.79 In-Reply-To: <19970730080303.36552@scsn.net>; from Donald J. Maddox on Wed, Jul 30, 1997 at 08:03:03AM -0400 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, Jul 30, 1997 at 08:03:03AM -0400, Donald J. Maddox wrote: > On Wed, Jul 30, 1997 at 01:58:02PM +0200, SЬren Schmidt wrote: > > In reply to Donald J. Maddox who wrote: > > > > > > Anybody know why I'm seeing the following after building a new kernel > > > with the EIDE bus master patches? Do I need to add something to my config > > > to activate this, or ??? > > > > > > Yes, as stated in the commit etc, you need to set the flags to 0xa0ffa0ff > > to get DMA support (if the drive and controller supports it). > > > > > wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ff80ff on isa > > ^^^^^^^^^^ > > > wdc1 at 0x170-0x177 irq 15 flags 0x80ff80ff on isa > > ^^^^^^^^^^ > > > wdc1: unit 0 (atapi): , removable, iordy > > > > Geez, how did I miss that? Thanks for the info :-) > Well, apparently, that's not the whole story: Copyright (c) 1992-1997 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-CURRENT #0: Wed Jul 30 08:11:06 EDT 1997 root@rhiannon.scsn.net:/usr/src/sys/compile/RHIANNON CPU: Pentium (166.19-MHz 586-class CPU) Origin = "GenuineIntel" Id = 0x52c Stepping=12 Features=0x1bf real memory = 33554432 (32768K bytes) avail memory = 30920704 (30196K bytes) Probing for devices on PCI bus 0: chip0: rev 0x01 on pci0.0.0 chip1: rev 0x00 on pci0.7.0 ide_pci0: rev 0x00 on pci0.7.1 ide_pci: warning, ide0:0 not configured for DMA? ide_pci: warning, ide0:1 not configured for DMA? ide_pci: warning, ide1:0 not configured for DMA? ide_pci: warning, ide1:1 not configured for DMA? ahc0: rev 0x00 int a irq 12 on pci0.10.0 ahc0: aic7880 Wide Channel, SCSI Id=7, 16/255 SCBs scbus0 at ahc0 bus 0 ahc0: target 0 Tagged Queuing Device sd0 at scbus0 target 0 lun 0 sd0: type 0 fixed SCSI 2 sd0: Direct-Access 2040MB (4178874 512 byte sectors) sd0: with 2708 cyls, 19 heads, and an average 81 sectors/track ahc0: target 1 Tagged Queuing Device sd1 at scbus0 target 1 lun 0 sd1: type 0 fixed SCSI 2 sd1: Direct-Access 507MB (1039329 512 byte sectors) sd1: with 2380 cyls, 6 heads, and an average 72 sectors/track vga0: rev 0x00 int a irq 11 on pci0.12.0 Probing for devices on the ISA bus: sc0 at 0x60-0x6f irq 1 on motherboard sc0: VGA color <16 virtual consoles, flags=0x0> lpt0 at 0x378-0x37f irq 7 on isa lpt0: Interrupt-driven port lp0: TCP/IP capable interface sio0 at 0x3f8-0x3ff irq 4 flags 0x10 on isa sio0: type 16550A sio1 at 0x2f8-0x2ff irq 3 on isa sio1: type 16550A pca0 on motherboard pca0: PC speaker audio driver wdc0 at 0x1f0-0x1f7 irq 14 flags 0xa0ffa0ff on isa wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-32 wd0: 1277MB (2615760 sectors), 2595 cyls, 16 heads, 63 S/T, 512 B/S wdc1 at 0x170-0x177 irq 15 flags 0xa0ffa0ff on isa wdc1: unit 0 (atapi): , removable, iordy atapi1.0: unknown phase fdc0 at 0x3f0-0x3f7 irq 6 drq 2 on isa fdc0: NEC 72065B fd0: 1.44MB 3.5in npx0 on motherboard npx0: INT 16 interface sb0 at 0x220-0x22f irq 5 drq 1 on isa sb0: sbxvi0 drq 5 on isa sbxvi0: sbmidi0 at 0x330-0x331 on isa sbmidi0: opl0 at 0x388-0x38b on isa opl0: joy0 at 0x201 on isa joy0: joystick From owner-freebsd-current Wed Jul 30 05:22:06 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA27425 for current-outgoing; Wed, 30 Jul 1997 05:22:06 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA27416 for ; Wed, 30 Jul 1997 05:22:00 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id OAA02238; Wed, 30 Jul 1997 14:20:44 +0200 (MEST) From: SЬren Schmidt Message-Id: <199707301220.OAA02238@sos.freebsd.dk> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970730081733.55871@scsn.net> from "Donald J. Maddox" at "Jul 30, 97 08:17:33 am" To: dmaddox@scsn.net Date: Wed, 30 Jul 1997 14:20:44 +0200 (MEST) Cc: dmaddox@scsn.net, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to Donald J. Maddox who wrote: > > > > > > Yes, as stated in the commit etc, you need to set the flags to 0xa0ffa0ff > > > to get DMA support (if the drive and controller supports it). > > > > > > > wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ff80ff on isa > > > ^^^^^^^^^^ > > > > wdc1 at 0x170-0x177 irq 15 flags 0x80ff80ff on isa > > > ^^^^^^^^^^ > > > > wdc1: unit 0 (atapi): , removable, iordy > > > > > Geez, how did I miss that? Thanks for the info :-) > > Well, apparently, that's not the whole story: Yes it is, and you have DMA support: > wdc0 at 0x1f0-0x1f7 irq 14 flags 0xa0ffa0ff on isa > wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-32 ^^^ > wd0: 1277MB (2615760 sectors), 2595 cyls, 16 heads, 63 S/T, 512 B/S see?? -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- SЬren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end .. From owner-freebsd-current Wed Jul 30 05:39:31 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA28038 for current-outgoing; Wed, 30 Jul 1997 05:39:31 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA28030 for ; Wed, 30 Jul 1997 05:39:29 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.178]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA158; Wed, 30 Jul 1997 08:29:46 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id IAA00279; Wed, 30 Jul 1997 08:39:13 -0400 (EDT) Message-ID: <19970730083912.16090@scsn.net> Date: Wed, 30 Jul 1997 08:39:12 -0400 From: "Donald J. Maddox" To: =?iso-8859-1?Q?S=F8ren_Schmidt?= Cc: dmaddox@scsn.net, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches Reply-To: dmaddox@scsn.net References: <19970730081733.55871@scsn.net> <199707301220.OAA02238@sos.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Mutt 0.79 In-Reply-To: =?iso-8859-1?Q?=3C199707301220=2EOAA02238=40sos=2Efreebsd=2Edk=3E=3B_fro?= =?iso-8859-1?Q?m_S=F8ren_Schmidt_on_Wed=2C_Jul_30=2C_1997_at_02=3A20=3A4?= =?iso-8859-1?Q?4PM_+0200?= Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, Jul 30, 1997 at 02:20:44PM +0200, SЬren Schmidt wrote: > In reply to Donald J. Maddox who wrote: > > > > > > > > Yes, as stated in the commit etc, you need to set the flags to 0xa0ffa0ff > > > > to get DMA support (if the drive and controller supports it). > > > > > > > > > wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ff80ff on isa > > > > ^^^^^^^^^^ > > > > > wdc1 at 0x170-0x177 irq 15 flags 0x80ff80ff on isa > > > > ^^^^^^^^^^ > > > > > wdc1: unit 0 (atapi): , removable, iordy > > > > > > > Geez, how did I miss that? Thanks for the info :-) > > > > Well, apparently, that's not the whole story: > > Yes it is, and you have DMA support: > > > wdc0 at 0x1f0-0x1f7 irq 14 flags 0xa0ffa0ff on isa > > wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-32 > ^^^ > > wd0: 1277MB (2615760 sectors), 2595 cyls, 16 heads, 63 S/T, 512 B/S > > see?? Well, no, not entirely... My original msg was basically an enquiry as to why I am seeing the following msgs at boot: Copyright (c) 1992-1997 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-CURRENT #0: Wed Jul 30 08:11:06 EDT 1997 root@rhiannon.scsn.net:/usr/src/sys/compile/RHIANNON CPU: Pentium (166.19-MHz 586-class CPU) Origin = "GenuineIntel" Id = 0x52c Stepping=12 Features=0x1bf real memory = 33554432 (32768K bytes) avail memory = 30920704 (30196K bytes) Probing for devices on PCI bus 0: chip0: rev 0x01 on pci0.0.0 chip1: rev 0x00 on pci0.7.0 ide_pci0: rev 0x00 on pci0.7.1 ide_pci: warning, ide0:0 not configured for DMA? ide_pci: warning, ide0:1 not configured for DMA? ide_pci: warning, ide1:0 not configured for DMA? ide_pci: warning, ide1:1 not configured for DMA? ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Adding the proper flag does indeed enable the DMA, but I could care less about that, since I only have Win95/NT on IDE drives. I just want to know why I am getting the messages from ide_pci, and how to shut them up. I should have been more specific in my original msg. From owner-freebsd-current Wed Jul 30 06:01:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id GAA28895 for current-outgoing; Wed, 30 Jul 1997 06:01:42 -0700 (PDT) Received: from gw.itfs.nsk.su (gw.itfs.nsk.su [193.124.36.33]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id GAA28884 for ; Wed, 30 Jul 1997 06:01:35 -0700 (PDT) Received: from itfs.UUCP (uucp@localhost) by gw.itfs.nsk.su (8.6.12/8.6.12) with UUCP id UAA21062 for current@freebsd.org; Wed, 30 Jul 1997 20:00:14 +0700 Received: by itfs.nsk.su; Wed, 30 Jul 97 19:49:00 +0700 (NST) Received: (from daemon@localhost) by news.itfs.nsk.su (8.7.5/8.6.12) id TAA00431; Wed, 30 Jul 1997 19:44:09 +0700 (NSD) From: nnd@itfs.nsk.su To: current@freebsd.org Subject: Re: code talks: announcing EIDE bus master patches Date: 30 Jul 1997 12:44:08 GMT Message-ID: <5rncuo$sp8@news.itfs.nsk.su> References: <199707290424.AAA23424@smoke.marlboro.vt.us> <199707301049.GAA03660@smoke.marlboro.vt.us> <19970730072456.05185@scsn.net> Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Donald J. Maddox wrote: > On Wed, Jul 30, 1997 at 06:49:54AM -0400, john hood wrote: > Anybody know why I'm seeing the following after building a new kernel > with the EIDE bus master patches? Do I need to add something to my config > to activate this, or ??? I have the same question (see start of my dmesg later on). My motherboard is Gateway's GA-586DX. I can not see any settings about IDE-DMA in BIOS setup :-( And the next one - when I try to set flags on wcd0 to '0xa0ffa0ff' the system can't start - I receive: panic - can not mount root (?) message. Is there any possibility to avoid this panic and fall back to PIO-mode ? N.Dudorov ------------------------------------------------------------------- Copyright (c) 1992-1997 FreeBSD Inc. Copyright (c) 1982, 1986, 1989, 1991, 1993 The Regents of the University of California. All rights reserved. FreeBSD 3.0-CURRENT #0: Wed Jul 30 18:13:59 NSS 1997 nnd@nnd.itfs.nsk.su:/arch/3.0/src/sys/compile/SMP-NND CPU: Pentium (586-class CPU) Origin = "GenuineIntel" Id = 0x52c Stepping=12 Features=0x3bf real memory = 67108864 (65536K bytes) avail memory = 63660032 (62168K bytes) FreeBSD/SMP: Multiprocessor motherboard cpu0 (BSP): apic id: 0, version: 0x00030010, at 0xfee00000 cpu1 (AP): apic id: 1, version: 0x00030010, at 0xfee00000 io0 (APIC): apic id: 2, version: 0x00170011, at 0xfec00000 Probing for devices on PCI bus 0: chip0: rev 0x01 on pci0.0.0 chip1: rev 0x01 on pci0.7.0 ide_pci0: rev 0x00 on pci0.7.1 ide_pci: warning, ide0:0 not configured for DMA? ide_pci: warning, ide0:1 not configured for DMA? ide_pci: warning, ide1:0 not configured for DMA? ide_pci: warning, ide1:1 not configured for DMA? ...... From owner-freebsd-current Wed Jul 30 06:07:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id GAA29103 for current-outgoing; Wed, 30 Jul 1997 06:07:16 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id GAA29096 for ; Wed, 30 Jul 1997 06:07:11 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id WAA22063; Wed, 30 Jul 1997 22:35:56 +0930 (CST) From: Michael Smith Message-Id: <199707301305.WAA22063@genesis.atrad.adelaide.edu.au> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970730083912.16090@scsn.net> from "Donald J. Maddox" at "Jul 30, 97 08:39:12 am" To: dmaddox@scsn.net Date: Wed, 30 Jul 1997 22:35:56 +0930 (CST) Cc: sos@sos.freebsd.dk, dmaddox@scsn.net, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Donald J. Maddox stands accused of saying: > ide_pci: warning, ide0:0 not configured for DMA? > ide_pci: warning, ide0:1 not configured for DMA? > ide_pci: warning, ide1:0 not configured for DMA? > ide_pci: warning, ide1:1 not configured for DMA? > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > Adding the proper flag does indeed enable the DMA, but I could care > less about that, since I only have Win95/NT on IDE drives. I just want to > know why I am getting the messages from ide_pci, and how to shut them up. Turn on IDE busmastering in your BIOS. Note the syntax of the warning "not configured". That's a message from the chipset driver daying "why is this channel not configured for DMA?". It should probably just list the ports on which DMA is available and enabled/not enabled, but this is, after all, only beta-level code. (Damn fine beta-level code if you ask me, but all the same..) > I should have been more specific in my original msg. Yes. 8) -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 06:23:02 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id GAA29898 for current-outgoing; Wed, 30 Jul 1997 06:23:02 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id GAA29882 for ; Wed, 30 Jul 1997 06:22:55 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id WAA22357; Wed, 30 Jul 1997 22:51:06 +0930 (CST) From: Michael Smith Message-Id: <199707301321.WAA22357@genesis.atrad.adelaide.edu.au> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <5rncuo$sp8@news.itfs.nsk.su> from "nnd@itfs.nsk.su" at "Jul 30, 97 12:44:08 pm" To: nnd@itfs.nsk.su Date: Wed, 30 Jul 1997 22:51:05 +0930 (CST) Cc: current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk nnd@itfs.nsk.su stands accused of saying: > And the next one - when I try to set flags on wcd0 to '0xa0ffa0ff' > the system can't start - I receive: > panic - can not mount root (?) message. > Is there any possibility to avoid this panic and > fall back to PIO-mode ? We would need the 'wd*' section of the boot probe to have any chance of answering that. It's also possible that you're the first SMP user to have tried this (?). > N.Dudorov > > ------------------------------------------------------------------- > Copyright (c) 1992-1997 FreeBSD Inc. > Copyright (c) 1982, 1986, 1989, 1991, 1993 > The Regents of the University of California. All rights reserved. > FreeBSD 3.0-CURRENT #0: Wed Jul 30 18:13:59 NSS 1997 > nnd@nnd.itfs.nsk.su:/arch/3.0/src/sys/compile/SMP-NND > CPU: Pentium (586-class CPU) > Origin = "GenuineIntel" Id = 0x52c Stepping=12 > Features=0x3bf > real memory = 67108864 (65536K bytes) > avail memory = 63660032 (62168K bytes) > FreeBSD/SMP: Multiprocessor motherboard > cpu0 (BSP): apic id: 0, version: 0x00030010, at 0xfee00000 > cpu1 (AP): apic id: 1, version: 0x00030010, at 0xfee00000 > io0 (APIC): apic id: 2, version: 0x00170011, at 0xfec00000 > Probing for devices on PCI bus 0: > chip0: rev 0x01 on pci0.0.0 > chip1: rev 0x01 on pci0.7.0 > ide_pci0: rev 0x00 on pci0.7.1 > ide_pci: warning, ide0:0 not configured for DMA? > ide_pci: warning, ide0:1 not configured for DMA? > ide_pci: warning, ide1:0 not configured for DMA? > ide_pci: warning, ide1:1 not configured for DMA? > ...... > -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 06:25:47 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id GAA00241 for current-outgoing; Wed, 30 Jul 1997 06:25:47 -0700 (PDT) Received: from mail11.digital.com (mail11.digital.com [192.208.46.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id GAA00225 for ; Wed, 30 Jul 1997 06:25:41 -0700 (PDT) From: garyj@frt.dec.com Received: from cssmuc.frt.dec.com (cssmuc.frt.dec.com [16.186.96.161]) by mail11.digital.com (8.7.5/UNX 1.5/1.0/WV) with SMTP id JAA10920; Wed, 30 Jul 1997 09:08:39 -0400 (EDT) Received: from localhost by cssmuc.frt.dec.com; (5.65v3.2/1.1.8.2/14Nov95-0232PM) id AA00366; Wed, 30 Jul 1997 15:08:23 +0200 Message-Id: <9707301308.AA00366@cssmuc.frt.dec.com> X-Mailer: exmh version 1.6.4 10/10/95 To: dmaddox@scsn.net Cc: current@freebsd.org In-Reply-To: Message from "Donald J. Maddox" of Wed, 30 Jul 97 08:39:12 EDT. Reply-To: gjennejohn@frt.dec.com Subject: Re: code talks: announcing EIDE bus master patches Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Wed, 30 Jul 97 15:08:23 +0200 X-Mts: smtp Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk root@scsn.net writes: > Well, no, not entirely... My original msg was basically an enquiry > as to why I am seeing the following msgs at boot: > > ide_pci0: rev 0x00 on pci0.7.1 > ide_pci: warning, ide0:0 not configured for DMA? > ide_pci: warning, ide0:1 not configured for DMA? > ide_pci: warning, ide1:0 not configured for DMA? > ide_pci: warning, ide1:1 not configured for DMA? > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > Adding the proper flag does indeed enable the DMA, but I could care > less about that, since I only have Win95/NT on IDE drives. I just want t= > o > know why I am getting the messages from ide_pci, and how to shut them up. > > I should have been more specific in my original msg. > I'm seeing these messages too. You probably need to enable DMA in the BIOS, if your BIOS supports that. Mine doesn't seem to :( --- Gary Jennejohn (work) gjennejohn@frt.dec.com (home) Gary.Jennejohn@munich.netsurf.de (play) gj@freebsd.org From owner-freebsd-current Wed Jul 30 07:00:05 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA02226 for current-outgoing; Wed, 30 Jul 1997 07:00:05 -0700 (PDT) Received: from watson.grauel.com (watson.grauel.com [199.233.104.36]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA02178 for ; Wed, 30 Jul 1997 07:00:01 -0700 (PDT) Received: (from rjk@localhost) by watson.grauel.com (8.8.5/8.7.3) id JAA04730; Wed, 30 Jul 1997 09:00:02 -0500 (EST) Date: Wed, 30 Jul 1997 09:00:02 -0500 (EST) Message-Id: <199707301400.JAA04730@watson.grauel.com> From: Richard J Kuhns MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Bruce Evans CC: current@freebsd.org Subject: Re: Can't mount CD under -current. In-Reply-To: <199707300502.PAA32716@godzilla.zeta.org.au> References: <199707300502.PAA32716@godzilla.zeta.org.au> X-Mailer: VM 6.22 under 19.15p6 XEmacs Lucid Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Bruce Evans writes: > >With a -current kernel as of 7/28, I can no longer mount a CD. I remember > >seeing comments about breakages when the webnfs stuff went in; I hope that > >this was just overlooked. > > The comments gave the fix: just use current binaries with current kernels. > This is sometimes incconvenient, but anything else is not current. > > Bruce You're right; my apologies. I thought I'd done a make world since then, but obviously I hadn't. It works fine now. Thanks. -- Rich Kuhns rjk@grauel.com PO Box 6249 Tel: (765)477-6000 \ 100 Sawmill Road x319 Lafayette, IN 47903 (800)489-4891 / From owner-freebsd-current Wed Jul 30 07:06:47 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA02774 for current-outgoing; Wed, 30 Jul 1997 07:06:47 -0700 (PDT) Received: from trifork.gu.net (trifork.gu.net [194.93.190.194]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA02761 for ; Wed, 30 Jul 1997 07:06:35 -0700 (PDT) Received: from localhost (localhost.gu.kiev.ua [127.0.0.1]) by trifork.gu.net (8.8.5/8.8.5) with SMTP id RAA01582; Wed, 30 Jul 1997 17:04:15 +0300 (EEST) Date: Wed, 30 Jul 1997 17:04:15 +0300 (EEST) From: Andrew Stesin Reply-To: stesin@gu.net To: gjennejohn@frt.dec.com cc: current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <9707301308.AA00366@cssmuc.frt.dec.com> Message-ID: X-NCC-RegID: ua.gu MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, 30 Jul 1997 garyj@frt.dec.com wrote: > I'm seeing these messages too. You probably need to enable DMA in the > BIOS, if your BIOS supports that. Mine doesn't seem to :( If your MB is sufficiently modern, the BIOS image is stored in flash memory and you can upgrade it (usually BIOS upgrades are available from manufacturer's website of FTP, together with all the instructions). I did those upgrades several times, on ASUS, PackardBell, Intel MBs, no big problem, just be careful and attentive. > --- > Gary Jennejohn (work) gjennejohn@frt.dec.com > (home) Gary.Jennejohn@munich.netsurf.de > (play) gj@freebsd.org Best regards, Andrew Stesin nic-hdl: ST73-RIPE From owner-freebsd-current Wed Jul 30 07:44:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA05135 for current-outgoing; Wed, 30 Jul 1997 07:44:16 -0700 (PDT) Received: from rocky.mt.sri.com (rocky.mt.sri.com [206.127.76.100]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA05116; Wed, 30 Jul 1997 07:44:06 -0700 (PDT) Received: (from nate@localhost) by rocky.mt.sri.com (8.7.5/8.7.3) id IAA04592; Wed, 30 Jul 1997 08:44:02 -0600 (MDT) Date: Wed, 30 Jul 1997 08:44:02 -0600 (MDT) Message-Id: <199707301444.IAA04592@rocky.mt.sri.com> From: Nate Williams MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Jeffrey Hsu Cc: current@hub.freebsd.org Subject: JDK port (was Re: pccard and -current; a long way to go. :-() In-Reply-To: <199707300710.AAA14949@hub.freebsd.org> References: <199707300710.AAA14949@hub.freebsd.org> X-Mailer: VM 6.29 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > My next JDK port which I'm working on now (not sure if I should do a > stable 1.1.4 port or just skip ahead to 1.2, neither of which the > Linux people have yet) would happily use ELF if it solves the > dlsym(RTLD_NEXT) problem (namely, we don't have it) and it also > eliminates a bunch of kludges in the Java interpreter code that I have > to make because we prepend underscores and the rest of the world > doesn't. If you can give me a JDK port, I'll be an ELF advocate. :) :) Nate ps. I'd be *really* happy to have a 1.1.4 port first, since I've found that if you jump too far ahead of the JDK curve, you end up spending most of your time debugging the tools/libraries, and a 1.1.4 port would give us a lot more 'longevity'. People are still actively using your 1.0.2 port. From owner-freebsd-current Wed Jul 30 07:45:55 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA05288 for current-outgoing; Wed, 30 Jul 1997 07:45:55 -0700 (PDT) Received: from labs.usn.blaze.net.au (root@labs.usn.blaze.net.au [203.17.53.30]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA05280 for ; Wed, 30 Jul 1997 07:45:48 -0700 (PDT) Received: from labs.usn.blaze.net.au (davidn@local [127.0.0.1]) by labs.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id AAA00985; Thu, 31 Jul 1997 00:43:47 +1000 (EST) Message-Id: <199707301443.AAA00985@labs.usn.blaze.net.au> X-Mailer: exmh version 2.0zeta 7/24/97 To: Garrett Wollman cc: Peter Dufault , current@FreeBSD.ORG Subject: Re: where to put access restriction for scheduling classes In-reply-to: Your message of "Tue, 29 Jul 1997 21:26:58 -0400." <199707300126.VAA16803@khavrinen.lcs.mit.edu> X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 31 Jul 1997 00:43:47 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > < said: > > > "Shall I add another resource limit that requires changing all > > shells, [...] or [...] shall I add a resource limit to FreeBSD and > > change only that part of the FreeBSD login mechanism that is > > modifying resource limits at login time while still effectively root > > thus ignore this portability af shells issue, in particular ignoring > > the other 4.4 derived OS's?" > > No shell should break as a result of defining a new resource limit. > Therefore, you should update setusercontext(3) and the two system > shells, and let the other shells fend for themselves. Don't forget /usr/bin/limits. :-) Just out of curiosity, what type of resource are we actually discussing here? -- David Nugent - Unique Computing Pty Ltd - Melbourne, Australia Voice +61-3-9791-9547 Data/BBS +61-3-9792-3507 3:632/348@fidonet davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Wed Jul 30 10:07:56 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA16074 for current-outgoing; Wed, 30 Jul 1997 10:07:56 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id KAA16063 for ; Wed, 30 Jul 1997 10:07:52 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id KAA05229; Wed, 30 Jul 1997 10:05:24 -0700 From: Terry Lambert Message-Id: <199707301705.KAA05229@phaeton.artisoft.com> Subject: Re: pccard and -current; a long way to go. :-( To: jdp@polstra.com (John Polstra) Date: Wed, 30 Jul 1997 10:05:24 -0700 (MST) Cc: terry@lambert.org, msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG In-Reply-To: <199707300412.VAA07616@austin.polstra.com> from "John Polstra" at Jul 29, 97 09:12:58 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > > I think you need to document what needs to be done here. I would > > be happy to run an ELF kernel in my now (mostly) non-ELF environment. ^^^^ This "non-" shouldn't be here... sorry! > Well, there are a zillion things that need to be done. Just off the > top of my head: > > * Solve the booting problem. My hacked-up bootloader was within > about 6 bytes of the maximum size -- way too big to commit. GRUB? > An a.out wrapper for the ELF kernel? Frankly, I'd be happy to live with your hacked-up boot loader. I expect the whole boot code thing to change significantly, eventually. It nearly hit critical mass in the last "bootfs" discussion; it will take one or two more discussions before someone who is an acceptable code source for commit actually does the code. > * Design a directory structure for the toolchain components so that > it's reasonable to have both a.out and ELF tools installed at the > same time. I'm not terrifically concerned with this, either. If this is an issue, then the correct respone is to support BSD a.out more fully in binutils, especially the linker. > * Work out the various problems that stem from the fact that a.out > prepends a leading underscore to global symbol names while ELF does > not. (The usual reply to this is "let's just make our ELF tools > prepend the underscore too," although I bet you know better than > that. You're not allowed to do it because the ELF spec says you > can't. If we want ELF, then let's have ELF and not elf.out or > a.elf.) Don't be so sure that I am not an "underscore advocate". There's plenty of code that distinguishes the assembly from the C symbol name space using underscores... the non-underscore functions are a means of providing duplicate symbol hiding. In many cases, the removal of the C symbol underscore is BAD(tm), in that it causes a collision with an assembly function of the same name. I could see this being a *big* problem for an in-kernel math emulator (one example). > * Work out a build structure that will allow an a.out->ELF make > world. "Just Do It" (tm). Specifically, a build world works for the a.out -> a.out update; there's no reason that, if the tool defaults change and the kernel supports ELF image activation, that the same won't "just work" for a.out -> ELF. The *only* possible stumbling block I see is the existance of tandem shared libraries perhaps taking too much space. A binutils that knew about a.out *should* fix this; the biggie is a replacement ld.so for a.out that can link ELF shared libraries instead (or in addition to) a.out shared libraries. > * Consider the feasibility of mixed linking combining both ELF and > a.out object modules and libraries. Likewise for dynamic linking. > Some people have a.out libraries and executables which they paid > money for. They don't want to throw them away. Again, the > underscore problem is an obstacle here. Binutils. And the libraries with colliding C/asm namespaces are the main argument for doing the underscore in the conversion. Generally, I'd say add a flag bit, and let the linker know about it, and have binutils set the falg on conversion. This is about as close as you could get to fully transparent operation. > > Did you use the "fast call" conventions on the ELF kernel system > > calls? This would imply a libc rebuild, right? > > Are you talking about this in "src/lib/libc/i386/SYS.h"? > > #ifdef __ELF__ > #define KERNCALL int $0x80 /* Faster */ [ ... ] yes. > And by the way, can somebody explain why system calls care about the > object file format of the application making the calls? It's not > obvious to me. Apparently, the int gate is only there if the ELF stuff is there, and it's faster. The reason for the difference is lame backward compatability support for the new feature (IMO). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Wed Jul 30 10:41:27 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA19287 for current-outgoing; Wed, 30 Jul 1997 10:41:27 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA19275; Wed, 30 Jul 1997 10:41:20 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id DAA26172; Thu, 31 Jul 1997 03:35:52 +1000 Date: Thu, 31 Jul 1997 03:35:52 +1000 From: Bruce Evans Message-Id: <199707301735.DAA26172@godzilla.zeta.org.au> To: cgull@smoke.marlboro.vt.us, sos@sos.freebsd.dk Subject: Re: code talks: announcing EIDE bus master patches Cc: freebsd-current@FreeBSD.ORG, se@FreeBSD.ORG Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >Well, bonnie reports an increase in CPU load and some increase in I/O >speed on the hack box I developed the code on, while a little >spin-loop-at-idle-priority hack i coded up shows an improvement in >available CPU that it can consume. My conclusion: bonnie's CPU load >numbers are useless, at least for IDE drives. Bonnie doesn't report interrupt overhead. FreeBSD does too good a job of separating interrupt overhead from syscall overhead, and there is no system call for reporting the interrupt overhead. The overhead is easy to see using `systat -vmstat' and easy to account for provided there is only one process causing most of the interrupts (the time for a few hundred normal interrupts is insignificant except on slow machines). This is the same setup as is required for the spinloop method. Bonnie's CPU load numbers are also fairly useless for SCSI drives :-). They measure mainly the driver, filesystem and buffer cache overheads. DMA overhead is hard to measure, but the spinloop method works well for giving the total overhead. Bruce From owner-freebsd-current Wed Jul 30 11:01:25 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA20565 for current-outgoing; Wed, 30 Jul 1997 11:01:25 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA20557 for ; Wed, 30 Jul 1997 11:01:21 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id DAA26679; Thu, 31 Jul 1997 03:56:39 +1000 Date: Thu, 31 Jul 1997 03:56:39 +1000 From: Bruce Evans Message-Id: <199707301756.DAA26679@godzilla.zeta.org.au> To: cgull@smoke.marlboro.vt.us, freebsd-current@FreeBSD.ORG Subject: problems with IDE DMA Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk After successfully testing IDE DMA for a little while (reading rwd0, fsck -n and bonnie -s 64 worked fine) I got the following messages after the write stages for bonnie -s 64: wd0: interrupt timeout: wd0: status 50 error 0 wd0: interrupt timeout: wd0: status 58 error 1 wd0: interrupt timeout: wd0: status 58 error 1 wd0: interrupt timeout: wd0: status 58 error 1 wd0: Last time I say: interrupt timeout. Probably a portable PC.: wd0: status 58 error 1 The system then drifted South: - no more interrupt timeouts, but wd seemed to get stuck in single sector mode. ls on the bonnie output file was very slow, and systat reported about 20 interrupts/sec, 10 "blks", and often "***** msps" (division by 0?). - long before the 13107 seconds required to read 128MB at 10K/sec, the system hung (the keyboard LEDs didn't work). - the test drive was dead after hard reset. - the test drive was OK after cycling the power. Boot-time log messages: ... ide_pci0: rev 0x02 on pci0.7.1 ide_pci: warning, ide0:0 not configured for DMA? ide_pci: warning, ide0:1 not configured for DMA? ide_pci: warning, ide1:0 not configured for DMA? ide_pci: warning, ide1:1 not configured for DMA? ... wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ffa0ff on isa wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-16 wd0: 76kB cache 2014MB (4124736 sectors), 4092 cyls, 16 heads, 63 S/T, 512 B/S I tested before I noticed the error messages. DMA did work. Bruce From owner-freebsd-current Wed Jul 30 11:05:09 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA20854 for current-outgoing; Wed, 30 Jul 1997 11:05:09 -0700 (PDT) Received: from forbidden-donut.anet-stl.com (forbidden-donut.anet-stl.com [206.114.203.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA20849 for ; Wed, 30 Jul 1997 11:05:06 -0700 (PDT) Received: from localhost (doogie@localhost) by forbidden-donut.anet-stl.com (8.8.6/8.8.5) with SMTP id NAA29037; Wed, 30 Jul 1997 13:01:41 GMT Date: Wed, 30 Jul 1997 13:01:39 +0000 (GMT) From: Jason Young To: john hood cc: SЬren Schmidt , freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707301049.GAA03660@smoke.marlboro.vt.us> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk -----BEGIN PGP SIGNED MESSAGE----- On Wed, 30 Jul 1997, john hood wrote: > Jason Young writes: > > I'm not sure what this driver did to my 1GB Seagate, but all I have to say > > is buy this man a beer. :) The machine: AMD 5x86-133, 1GB seagate and > > 1.2GB WD EIDE drives, 32MB RAM. I ran these stats several times on an > > idle system without appreciable changes each time I ran them. > > well, i won't pass up a free beer, but my code didn't do it. :) > > you have a 486-class machine that doesn't have busmastering IDE. it's > not even a PCI bus machine. you just happened to turn on the > already-existing multi-block and 32-bit PIO flags in your latest > kernel upgrade. > > folks: performance with the default flags on the wd driver is > *terrible*, and the flags should always be configured for your > machine. see LINT for details. Yes it's 486-class, but it has a PCI bus. Otherwise I wouldn't have bothered with the code. Jason Young ANET Technical Staff -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQB1AwUBM987NKInE6ybC66VAQHk8gL/RSH6lWWQPohVgOcemrZzIxA7F2v8CeJx 4ZTMf6+REPc6ZFF/RjZGRQhfZQqCnVPOWRCgZ5P9uDLYoaNDRU77UGnGCvzutQyb kyjppH+XeYx/vaJGQWMBauVRuN7n1KbO =blN9 -----END PGP SIGNATURE----- From owner-freebsd-current Wed Jul 30 11:34:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA22984 for current-outgoing; Wed, 30 Jul 1997 11:34:34 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id LAA22976 for ; Wed, 30 Jul 1997 11:34:31 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id LAA05510; Wed, 30 Jul 1997 11:31:43 -0700 From: Terry Lambert Message-Id: <199707301831.LAA05510@phaeton.artisoft.com> Subject: Re: pccard and -current; a long way to go. :-( To: jdp@polstra.com (John Polstra) Date: Wed, 30 Jul 1997 11:31:42 -0700 (MST) Cc: TLiddelow@cybec.com.au, msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG In-Reply-To: <199707300429.VAA07677@austin.polstra.com> from "John Polstra" at Jul 29, 97 09:29:49 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > These benefits are real, but I'm not sure they're enough. (If you > are looking for a facts-be-damned advocate of FreeBSD-ELF, I'm not > your guy.) There is plenty of room for intelligent disagreement > about it. I myself have about three opinions on the subject, all > in conflict with one another. :-) > > I am sure there would be other benefits, but I don't know enough > about them to lobby on their behalf. The people who do are the > ones who need to do the PR. Explain why FreeBSD would be a better > place if it used ELF. So far nobody's done a very good job of > that. "Pageable kernel data areas?" Yawn. What does it mean to > me as a user or as a developer? It means: o Demand loading for LKM's. o Demand loading of dependencies (LKM TCP requires LKM IP, so loading LKM TCP loads LKM IP). o The generic kernel can discard drivers it's not using. o You can page out memory and page it back in to let you rearrange the pages. This means you don't have to statically link ing your "QuickCam" or similar driver that needs a region of physically contiguous kernel memory (ie: it can only currently allocate this at boot time). o You can write glue to let BSD use NT drivers (NT drivers, by design, require section coloring) for video cards, SCSI controllers, ethernet cards, etc.. NT drivers come with most new hardware. o You can replace drivers in a kernel using a section archiver instead of rebuilding and relinking. > Would the system run faster? Yes. It would page user data less if it could reclaim the memory from unused drivers and/opr probe code for drivers that *are* used. > Use less memory? Yes. > How much faster? How much less? Depends. How many transient or unused drivers are in your current kernel? Probably a hell of a lot, if you run generic, like most people. > Who's going to implement it if we switch to ELF? I think the point is that unless we switch to ELF, no one will *ever* implement it. > Enquiring minds want to know. Gotta disagree with you there. ;-). Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Wed Jul 30 11:41:28 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA23613 for current-outgoing; Wed, 30 Jul 1997 11:41:28 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id LAA23608 for ; Wed, 30 Jul 1997 11:41:26 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id LAA05569; Wed, 30 Jul 1997 11:37:41 -0700 From: Terry Lambert Message-Id: <199707301837.LAA05569@phaeton.artisoft.com> Subject: Re: pccard and -current; a long way to go. :-( To: imp@village.org (Warner Losh) Date: Wed, 30 Jul 1997 11:37:41 -0700 (MST) Cc: jdp@polstra.com, TLiddelow@cybec.com.au, msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG In-Reply-To: from "Warner Losh" at Jul 29, 97 11:08:07 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > : The people who recognize the benefits of moving to ELF are the ones > : who need to do the PR. The question is not "would ELF be better?" > : but rather "would ELF be _enough_ better to make it worth the pain > : of the transition?" > > I'm curious. Does anybody have a pure ELF FreeBSD system? I have a machine that almost has a pure ELF user space (shells, etc.) built with John's ELFKit. The only piece I'm missing is the kernel loader, and I can't just use the Linux one (this was discussed in detail a while back), so my kernel is the missing piece. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Wed Jul 30 11:53:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA24399 for current-outgoing; Wed, 30 Jul 1997 11:53:42 -0700 (PDT) Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id LAA24373; Wed, 30 Jul 1997 11:53:34 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr3-2.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA19624 (5.67b/IDA-1.5); Wed, 30 Jul 1997 20:53:26 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.6/8.6.9) id UAA12429; Wed, 30 Jul 1997 20:53:23 +0200 (CEST) X-Face: " Date: Wed, 30 Jul 1997 20:53:22 +0200 From: Stefan Esser To: john hood Cc: =?iso-8859-1?Q?S=F8ren_Schmidt?= , freebsd-current@FreeBSD.ORG, Stefan Esser Subject: Re: code talks: announcing EIDE bus master patches References: <19970729221036.52544@mi.uni-koeln.de> <199707292224.AAA00624@sos.freebsd.dk> <199707300559.BAA02894@smoke.marlboro.vt.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.74 In-Reply-To: <199707300559.BAA02894@smoke.marlboro.vt.us>; from john hood on Wed, Jul 30, 1997 at 01:59:35AM -0400 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Jul 30, john hood wrote: > Well, bonnie reports an increase in CPU load and some increase in I/O > speed on the hack box I developed the code on, while a little > spin-loop-at-idle-priority hack i coded up shows an improvement in > available CPU that it can consume. My conclusion: bonnie's CPU load > numbers are useless, at least for IDE drives. Well, as I already wrote (and Bruce Evans explained in much detail), the PIO transfers' CPU cycles are accounted anonymously in the "interrupt" category, and can't be associated with the process that caused them ... > Here's some more useless bonnie results. These are for a system with > a K5-PR90, a "VXPro" chipset (a relabeled VIA Apollo VP-1), 32MB of > RAM and two Quantum FB1080As. > > bash-2.00$ cat bonnie.pio > -------Sequential Output-------- ---Sequential Input-- --Random-- > -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- > Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU > 100 2913 66.7 4432 25.9 1541 15.5 3764 55.1 4512 31.4 70.9 9.1 What's missing from 100% in the "per char" tests is mostly spent on PIO transfers. > bash-2.00$ cat bonnie.dma > -------Sequential Output-------- ---Sequential Input-- --Random-- > -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks--- > Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU /sec %CPU > 100 3934 98.2 4290 28.0 1606 22.3 4538 86.2 4483 48.2 74.9 9.9 > > > > It has improved my worldstone by ~10% if that counts for real world use.. > > > > > > Yes, one thing that is often underestimated is the fact, > > > that while PIO mode transfers consume only a small fraction > > > of the cycles of a fast CPU, but they tend to consume them > > > exactly at the time, when the CPU is highly loaded anyway. > > But said CPU is often waiting for the disk anyway. Yes, it is. But my point was that in many applications the CPU is **not** waiting for the disk, and then PIO can have a large impact on latencies observed by the user. On a Unix system, lots of file operations are running asynchronously (writes and file read ahead). And 10% is about the performance that can be had by going from a P-133 to a P-166 (which used to be a few hundred dollars until very recently :) > Eeeugh. I think it's time for a new disk i/o standard that uses a > simple high-speed serial link and starts over on the whole command set > architecture thing. I had some hopes for Fibre Channel and SSA, but > both of those use SCSI commands if I'm not mistaken. Pardon my ignorance, but what's wrong with SCSI commands ??? Regards, STefan From owner-freebsd-current Wed Jul 30 11:55:47 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA24557 for current-outgoing; Wed, 30 Jul 1997 11:55:47 -0700 (PDT) Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id LAA24541 for ; Wed, 30 Jul 1997 11:55:42 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr3-2.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA19641 (5.67b/IDA-1.5 for ); Wed, 30 Jul 1997 20:55:30 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.6/8.6.9) id UAA12450; Wed, 30 Jul 1997 20:55:25 +0200 (CEST) X-Face: " Date: Wed, 30 Jul 1997 20:55:24 +0200 From: Stefan Esser To: Jason Young Cc: john hood , =?iso-8859-1?Q?S=F8ren_Schmidt?= , freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches References: <199707301049.GAA03660@smoke.marlboro.vt.us> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.74 In-Reply-To: ; from Jason Young on Wed, Jul 30, 1997 at 01:01:39PM +0000 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Jul 30, Jason Young wrote: > > you have a 486-class machine that doesn't have busmastering IDE. it's > > not even a PCI bus machine. you just happened to turn on the > > already-existing multi-block and 32-bit PIO flags in your latest > > kernel upgrade. > > > > folks: performance with the default flags on the wd driver is > > *terrible*, and the flags should always be configured for your > > machine. see LINT for details. > > Yes it's 486-class, but it has a PCI bus. Otherwise I wouldn't have > bothered with the code. I don't think there is bus-master EIDE support in any 486 class PCI chip set. And besides, you also need a hard disk that knows about DMA modes. Most dind't until recently ... Regards, STefan From owner-freebsd-current Wed Jul 30 12:03:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA25175 for current-outgoing; Wed, 30 Jul 1997 12:03:34 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA25158 for ; Wed, 30 Jul 1997 12:03:28 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id VAA28368; Wed, 30 Jul 1997 21:02:54 +0200 (MEST) From: SЬren Schmidt Message-Id: <199707301902.VAA28368@sos.freebsd.dk> Subject: Re: problems with IDE DMA In-Reply-To: <199707301756.DAA26679@godzilla.zeta.org.au> from Bruce Evans at "Jul 31, 97 03:56:39 am" To: bde@zeta.org.au (Bruce Evans) Date: Wed, 30 Jul 1997 21:02:53 +0200 (MEST) Cc: cgull@smoke.marlboro.vt.us, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to Bruce Evans who wrote: > ... > wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ffa0ff on isa > wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-16 > wd0: 76kB cache 2014MB (4124736 sectors), 4092 cyls, 16 heads, 63 S/T, 512 B/S ^^^^ What's that ?? Are you by any chances fiddleling with drive cache setup ?? > I tested before I noticed the error messages. DMA did work. I've run it here now for 'bout 36 hours, did several make worlds, no problems so far. -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- SЬren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end .. From owner-freebsd-current Wed Jul 30 12:27:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA26899 for current-outgoing; Wed, 30 Jul 1997 12:27:34 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA26887; Wed, 30 Jul 1997 12:27:26 -0700 (PDT) Received: from hurricane.cs.duke.edu (hurricane.cs.duke.edu [152.3.145.1]) by duke.cs.duke.edu (8.8.5/8.8.5) with ESMTP id PAA18542; Wed, 30 Jul 1997 15:24:55 -0400 (EDT) Received: (from gallatin@localhost) by hurricane.cs.duke.edu (8.8.4/8.7.3) id PAA08936; Wed, 30 Jul 1997 15:24:54 -0400 (EDT) Date: Wed, 30 Jul 1997 15:24:54 -0400 (EDT) Message-Id: <199707301924.PAA08936@hurricane.cs.duke.edu> From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Stefan Esser CC: freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970729221036.52544@mi.uni-koeln.de> References: <19970729210723.18104@mi.uni-koeln.de> <199707291949.VAA00271@sos.freebsd.dk> <19970729221036.52544@mi.uni-koeln.de> X-Mailer: VM 6.32 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Stefan Esser writes: > > Yes, one thing that is often underestimated is the fact, > that while PIO mode transfers consume only a small fraction > of the cycles of a fast CPU, but they tend to consume them > exactly at the time, when the CPU is highly loaded anyway. Speaking of PIO vs. DMA, an interesting data point is the bandwidth differences between PIO & DMA. For example, on a 200Mhz P. Pro (Asus XP6NP5, Intel 82440FX chipset) we can use PIO to move an 8k chunks of memory to a PCI device(*) at a rate of 29.1 MB/sec, but using PIO for reads gives us bandwidth of only 8.1MB/sec. Other platforms (Intel Triton @ 13MB/sec) show similar results. However, using DMA we see b/w of 114.4MB/sec (host -> PCI device) and 121.5 MB/sec (PCI -> host memory). So.. if IDE disks can really supply data at > 8MB/sec, you may actually be running into a b/w limitation of PIO. (*) PCI device == a Myrinet M2F-PCI32 card. This is a programmable gigabit networking card. It has a 256k bank of SRAM on the card, and is very good for doing things like measuring PCI b/w. The tests were done from user space operating on mmap'ed device memory & a kernel allocated chunk of RAM to do DMA xfers to/from. It also runs IP traffic at better than 300Mb/sec. I suppose this is as good a time to ask as any: I have a FreeBSD driver 95% done for Myrinet cards such as this. Who can I talk to about the possibility of getting it committed to -current? There are some sticky issues that may prevent its inclusion (specifically GPL'ed libraries that are required to communicate w/the vendor supported "firmware" that runs on the card & which I cannot think of a clean way to build). Drew ------------------------------------------------------------------------------ Andrew Gallatin http://www.cs.duke.edu/~gallatin Duke University Email: gallatin@cs.duke.edu Department of Computer Science Phone: (919) 660-6590 From owner-freebsd-current Wed Jul 30 12:31:35 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA27362 for current-outgoing; Wed, 30 Jul 1997 12:31:35 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA27357 for ; Wed, 30 Jul 1997 12:31:32 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id FAA29164; Thu, 31 Jul 1997 05:28:58 +1000 Date: Thu, 31 Jul 1997 05:28:58 +1000 From: Bruce Evans Message-Id: <199707301928.FAA29164@godzilla.zeta.org.au> To: bde@zeta.org.au, sos@sos.freebsd.dk Subject: Re: problems with IDE DMA Cc: cgull@smoke.marlboro.vt.us, freebsd-current@FreeBSD.ORG Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ffa0ff on isa >> wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-16 >> wd0: 76kB cache 2014MB (4124736 sectors), 4092 cyls, 16 heads, 63 S/T, 512 B/S > ^^^^ > What's that ?? Left over from debugging the "cache size is 0" bug report. >Are you by any chances fiddleling with drive cache setup ?? No, I just use whatever the BIOS sets. It's an almost 2 year old BIOS on an ASUS P55TP4XE. DMA seemed to work OK under Linux :-). Bruce From owner-freebsd-current Wed Jul 30 12:33:47 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA27583 for current-outgoing; Wed, 30 Jul 1997 12:33:47 -0700 (PDT) Received: from scanner.worldgate.com (scanner.worldgate.com [198.161.84.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA27576 for ; Wed, 30 Jul 1997 12:33:43 -0700 (PDT) Received: from znep.com (uucp@localhost) by scanner.worldgate.com (8.8.5/8.8.5) with UUCP id NAA26355; Wed, 30 Jul 1997 13:32:40 -0600 (MDT) Received: from localhost (marcs@localhost) by alive.znep.com (8.7.5/8.7.3) with SMTP id NAA21104; Wed, 30 Jul 1997 13:34:17 -0600 (MDT) Date: Wed, 30 Jul 1997 13:34:16 -0600 (MDT) From: Marc Slemko To: Terry Lambert cc: freebsd-current@FreeBSD.ORG Subject: Re: Who is working on the TCP stack? In-Reply-To: <199707282114.OAA01641@phaeton.artisoft.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Mon, 28 Jul 1997, Terry Lambert wrote: > > Has anyone checked into the reported incompatiblities with MS > > Winsock? I read in some 'Zine (Lan Times?) that Microsoft did > > not follow the RFC, and that Sun has chosen to accept and work > > with this incompatibility rather than wait for Microsoft to > > correct their software. From what I read, the implementation > > bug causes performance problems with non-MS Web servers (and > > as such was probably on purpose, it wouldn't be the first > > time). Anyway, I was just wondering if FreeBSD is affected > > and/or if it has already been addressed. > > The problem is that the client applications, specifically the > nono-Internet Explorer applications, do not call the WinSock > function "shutdown( s, 3);" and then "recv( s, ...);" all > the packets until the remote side close the connection. Erm... no. That has absolutely nothing to do with this issue. This particular issue is a disagreement in the slow-start implementation. Slow-start is documented after-the-fact in RFC2001 for anyone interested. The difference revolves around who should send what when; one box is expecting the other to send something, the other is expecting the first to send something, so they deadlock until one times out and retramsmits. Don't have the exact details offhand and I don't know if FreeBSD is impacted. From owner-freebsd-current Wed Jul 30 12:52:39 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA28795 for current-outgoing; Wed, 30 Jul 1997 12:52:39 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA28786 for ; Wed, 30 Jul 1997 12:52:36 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.68]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA148; Wed, 30 Jul 1997 15:42:46 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id PAA00649; Wed, 30 Jul 1997 15:52:02 -0400 (EDT) Message-ID: <19970730155201.26891@scsn.net> Date: Wed, 30 Jul 1997 15:52:01 -0400 From: "Donald J. Maddox" To: Michael Smith Cc: dmaddox@scsn.net, sos@sos.freebsd.dk, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches Reply-To: dmaddox@scsn.net References: <19970730083912.16090@scsn.net> <199707301305.WAA22063@genesis.atrad.adelaide.edu.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 In-Reply-To: <199707301305.WAA22063@genesis.atrad.adelaide.edu.au>; from Michael Smith on Wed, Jul 30, 1997 at 10:35:56PM +0930 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, Jul 30, 1997 at 10:35:56PM +0930, Michael Smith wrote: > Donald J. Maddox stands accused of saying: > > ide_pci: warning, ide0:0 not configured for DMA? > > ide_pci: warning, ide0:1 not configured for DMA? > > ide_pci: warning, ide1:0 not configured for DMA? > > ide_pci: warning, ide1:1 not configured for DMA? > > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > > > Adding the proper flag does indeed enable the DMA, but I could care > > less about that, since I only have Win95/NT on IDE drives. I just want to > > know why I am getting the messages from ide_pci, and how to shut them up. > > Turn on IDE busmastering in your BIOS. Note the syntax of the warning > "not configured". That's a message from the chipset driver daying > "why is this channel not configured for DMA?". It should probably > just list the ports on which DMA is available and enabled/not enabled, > but this is, after all, only beta-level code. There is no such option that I can see in my BIOS setup. I am using an ASUS P/I-P55T2P4 mainboard with BIOS version 2.047, but I can't recall ever seeing any such option in a BIOS from ASUS. From owner-freebsd-current Wed Jul 30 13:00:51 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA29533 for current-outgoing; Wed, 30 Jul 1997 13:00:51 -0700 (PDT) Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id NAA29513 for ; Wed, 30 Jul 1997 13:00:43 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr3-2.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA20131 (5.67b/IDA-1.5 for ); Wed, 30 Jul 1997 22:00:40 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.6/8.6.9) id WAA12751; Wed, 30 Jul 1997 22:00:38 +0200 (CEST) X-Face: " Date: Wed, 30 Jul 1997 22:00:38 +0200 From: Stefan Esser To: Andrew Gallatin Cc: freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches References: <19970729210723.18104@mi.uni-koeln.de> <199707291949.VAA00271@sos.freebsd.dk> <19970729221036.52544@mi.uni-koeln.de> <199707301924.PAA08936@hurricane.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.74 In-Reply-To: <199707301924.PAA08936@hurricane.cs.duke.edu>; from Andrew Gallatin on Wed, Jul 30, 1997 at 03:24:54PM -0400 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Jul 30, Andrew Gallatin wrote: > Speaking of PIO vs. DMA, an interesting data point is the bandwidth > differences between PIO & DMA. > > For example, on a 200Mhz P. Pro (Asus XP6NP5, Intel 82440FX chipset) > we can use PIO to move an 8k chunks of memory to a PCI device(*) at a > rate of 29.1 MB/sec, but using PIO for reads gives us bandwidth of > only 8.1MB/sec. Other platforms (Intel Triton @ 13MB/sec) show > similar results. However, using DMA we see b/w of 114.4MB/sec (host > -> PCI device) and 121.5 MB/sec (PCI -> host memory). Is this PIO to an I/O port, or to a memory mapped port ? But in any case: In order to use PCI burst transfers, you have to write to sequential memory addresses, e.g. to some double-mapped SRAM buffer. PCI is slow, if you can't take advantage of burst transfers! > So.. if IDE disks can really supply data at > 8MB/sec, you may > actually be running into a b/w limitation of PIO. If you assume 8 PCI clocks for a non-burst transfer, the limit comes out as 16MB/s. (The minimum is 4 clocks for such a transfer, but don't assume your chip-set does support that :) > (*) PCI device == a Myrinet M2F-PCI32 card. This is a programmable > gigabit networking card. It has a 256k bank of SRAM on the card, and > is very good for doing things like measuring PCI b/w. The tests were > done from user space operating on mmap'ed device memory & a kernel > allocated chunk of RAM to do DMA xfers to/from. It also runs IP > traffic at better than 300Mb/sec. This was a memory mapped device buffer ? That's *very* different from an I/O port, and you should be able to copy data to that SRAM at a much higher rate than 8 or 13MB/s! Regards, STefan From owner-freebsd-current Wed Jul 30 13:24:41 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA01109 for current-outgoing; Wed, 30 Jul 1997 13:24:41 -0700 (PDT) Received: from smoke.marlboro.vt.us (smoke.marlboro.vt.us [198.206.215.91]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA01102 for ; Wed, 30 Jul 1997 13:24:35 -0700 (PDT) Received: (from cgull@localhost) by smoke.marlboro.vt.us (8.8.5/8.8.5/cgull) id QAA04790; Wed, 30 Jul 1997 16:24:21 -0400 (EDT) Date: Wed, 30 Jul 1997 16:24:21 -0400 (EDT) Message-Id: <199707302024.QAA04790@smoke.marlboro.vt.us> From: john hood MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Michael Smith Cc: sos@sos.freebsd.dk, dmaddox@scsn.net, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707301305.WAA22063@genesis.atrad.adelaide.edu.au> References: <19970730083912.16090@scsn.net> <199707301305.WAA22063@genesis.atrad.adelaide.edu.au> X-Mailer: VM 6.31 under Emacs 19.34.2 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk [resent with correct To: headers] Michael Smith writes: > Donald J. Maddox stands accused of saying: > > ide_pci: warning, ide0:0 not configured for DMA? > > ide_pci: warning, ide0:1 not configured for DMA? > > ide_pci: warning, ide1:0 not configured for DMA? > > ide_pci: warning, ide1:1 not configured for DMA? > > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > > > Adding the proper flag does indeed enable the DMA, but I could care > > less about that, since I only have Win95/NT on IDE drives. I just want to > > know why I am getting the messages from ide_pci, and how to shut them up. > > Turn on IDE busmastering in your BIOS. Note the syntax of the warning > "not configured". That's a message from the chipset driver daying > "why is this channel not configured for DMA?". It should probably > just list the ports on which DMA is available and enabled/not enabled, > but this is, after all, only beta-level code. > > (Damn fine beta-level code if you ask me, but all the same..) *alpha*-level, Michael, and this is part of why. :) that message is there because i still don't know whether the BIOS is responsible for doing any configuration of busmaster drives and controllers. so far, the evidence is that most BIOSes do at least some basic configuration, but don't set the bit that says "everything's set up now, it's OK to go ahead and use DMA." from what I've seen so far, most BIOSes don't have a setup knob for IDE busmastering, and it's a complete mishmash as to what BIOSes actually do. we've seen one panic caused by the code not working, and i'm still waiting for somebody's drive to get corrupted because of these setup issues. --jh -- John Hood cgull@smoke.marlboro.vt.us Predictably, they all eventually wandered away, rubbing their bruises and brushing mud out of their hair. Some went off to work for the ESA, launching much smaller rockets into low orbits, while others elected to sit on their front porches drinking Jim Beam from the bottle and launching bottle rockets from the empties. [Jordan Hubbard] From owner-freebsd-current Wed Jul 30 13:33:48 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA01638 for current-outgoing; Wed, 30 Jul 1997 13:33:48 -0700 (PDT) Received: from smoke.marlboro.vt.us (smoke.marlboro.vt.us [198.206.215.91]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA01633 for ; Wed, 30 Jul 1997 13:33:44 -0700 (PDT) Received: (from cgull@localhost) by smoke.marlboro.vt.us (8.8.5/8.8.5/cgull) id QAA04812; Wed, 30 Jul 1997 16:33:28 -0400 (EDT) Date: Wed, 30 Jul 1997 16:33:28 -0400 (EDT) Message-Id: <199707302033.QAA04812@smoke.marlboro.vt.us> From: john hood MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: nnd@itfs.nsk.su Cc: Michael Smith , current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707301321.WAA22357@genesis.atrad.adelaide.edu.au> References: <5rncuo$sp8@news.itfs.nsk.su> <199707301321.WAA22357@genesis.atrad.adelaide.edu.au> X-Mailer: VM 6.31 under Emacs 19.34.2 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Michael Smith writes: > nnd@itfs.nsk.su stands accused of saying: > > > And the next one - when I try to set flags on wcd0 to '0xa0ffa0ff' > > the system can't start - I receive: > > panic - can not mount root (?) message. > > Is there any possibility to avoid this panic and > > fall back to PIO-mode ? > > We would need the 'wd*' section of the boot probe to have any chance > of answering that. It's also possible that you're the first SMP user > to have tried this (?). Actually, send the messages from a verbose boot (unfortunately, all that text sometimes overflows the dmesg buffer and screen scrollback). Also, what motherboard do you have, and what BIOS is installed on it? Does the kernel panic instantly on trying to mount root, or does it flail for a while and spew any error messages from wd? More details will help. There's some sanity checks I forgot to include in the code and this may flush one of them out. It *should* fall back to PIO if DMA mode doesn't work. To disable DMA entirely, unset bit 0x2000 in the flags for each drive. The PCI probe still happens but the wd driver will never use any of that info. --jh -- John Hood cgull@smoke.marlboro.vt.us Predictably, they all eventually wandered away, rubbing their bruises and brushing mud out of their hair. Some went off to work for the ESA, launching much smaller rockets into low orbits, while others elected to sit on their front porches drinking Jim Beam from the bottle and launching bottle rockets from the empties. [Jordan Hubbard] From owner-freebsd-current Wed Jul 30 13:36:52 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA01918 for current-outgoing; Wed, 30 Jul 1997 13:36:52 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA01908; Wed, 30 Jul 1997 13:36:35 -0700 (PDT) Received: from hurricane.cs.duke.edu (hurricane.cs.duke.edu [152.3.145.1]) by duke.cs.duke.edu (8.8.5/8.8.5) with ESMTP id QAA19651; Wed, 30 Jul 1997 16:36:31 -0400 (EDT) Received: (from gallatin@localhost) by hurricane.cs.duke.edu (8.8.4/8.7.3) id QAA23330; Wed, 30 Jul 1997 16:36:30 -0400 (EDT) Date: Wed, 30 Jul 1997 16:36:30 -0400 (EDT) Message-Id: <199707302036.QAA23330@hurricane.cs.duke.edu> From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Stefan Esser CC: freebsd-current@freebsd.org Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970730220038.02422@mi.uni-koeln.de> References: <19970729210723.18104@mi.uni-koeln.de> <199707291949.VAA00271@sos.freebsd.dk> <19970729221036.52544@mi.uni-koeln.de> <199707301924.PAA08936@hurricane.cs.duke.edu> <19970730220038.02422@mi.uni-koeln.de> X-Mailer: VM 6.32 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Stefan Esser writes: > On Jul 30, Andrew Gallatin wrote: > > Speaking of PIO vs. DMA, an interesting data point is the bandwidth > > differences between PIO & DMA. > > > > For example, on a 200Mhz P. Pro (Asus XP6NP5, Intel 82440FX chipset) > > we can use PIO to move an 8k chunks of memory to a PCI device(*) at a > > rate of 29.1 MB/sec, but using PIO for reads gives us bandwidth of > > only 8.1MB/sec. Other platforms (Intel Triton @ 13MB/sec) show > > similar results. However, using DMA we see b/w of 114.4MB/sec (host > > -> PCI device) and 121.5 MB/sec (PCI -> host memory). > > Is this PIO to an I/O port, or to a memory mapped port ? It's memory mapped PCI memory. > But in any case: In order to use PCI burst transfers, you > have to write to sequential memory addresses, e.g. to some > double-mapped SRAM buffer. Double-mapped? What does that mean? > PCI is slow, if you can't take advantage of burst transfers! > > > So.. if IDE disks can really supply data at > 8MB/sec, you may > > actually be running into a b/w limitation of PIO. > > If you assume 8 PCI clocks for a non-burst transfer, the > limit comes out as 16MB/s. (The minimum is 4 clocks for > such a transfer, but don't assume your chip-set does > support that :) > > > (*) PCI device == a Myrinet M2F-PCI32 card. This is a programmable > > gigabit networking card. It has a 256k bank of SRAM on the card, and > > is very good for doing things like measuring PCI b/w. The tests were > > done from user space operating on mmap'ed device memory & a kernel > > allocated chunk of RAM to do DMA xfers to/from. It also runs IP > > traffic at better than 300Mb/sec. > > This was a memory mapped device buffer ? Yes. > That's *very* different from an I/O port, and you should be able > to copy data to that SRAM at a much higher rate than 8 or 13MB/s! Right. The Natoma is using burst transfers for PIO writes and getting nearly 30MB/sec. Its apparently not bursting the reads, and is getting only 8MB/sec. How do you get an Intel to burst reads? I didn't think they could support multiple outstanding reads. I know that alphas do, and there are apparently tricks you can use to make an alpha burst reads & get 50MB/sec bandwidth, but I wasn't aware of Intels that could get better than 15MB/sec for PIO reads. I'm getting most of my information from emperical results obtained with our P6's and Alphas using Myrinet cards, plus the DEC PCI Pamette paper ("Systems Performance Measurement on PCI Pamette", Laurent Moll, Mark Shand, Fccm97, ftp://pam.devinci.fr/pub/pamette/fccm97.ps.gz). > > Regards, STefan Drew ------------------------------------------------------------------------------ Andrew Gallatin http://www.cs.duke.edu/~gallatin Duke University Email: gallatin@cs.duke.edu Department of Computer Science Phone: (919) 660-6590 From owner-freebsd-current Wed Jul 30 13:49:10 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA02527 for current-outgoing; Wed, 30 Jul 1997 13:49:10 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id NAA02519 for ; Wed, 30 Jul 1997 13:49:08 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id NAA05810; Wed, 30 Jul 1997 13:47:06 -0700 From: Terry Lambert Message-Id: <199707302047.NAA05810@phaeton.artisoft.com> Subject: Re: Who is working on the TCP stack? To: marcs@znep.com (Marc Slemko) Date: Wed, 30 Jul 1997 13:47:06 -0700 (MST) Cc: terry@lambert.org, freebsd-current@FreeBSD.ORG In-Reply-To: from "Marc Slemko" at Jul 30, 97 01:34:16 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > > The problem is that the client applications, specifically the > > nono-Internet Explorer applications, do not call the WinSock > > function "shutdown( s, 3);" and then "recv( s, ...);" all > > the packets until the remote side close the connection. > > Erm... no. That has absolutely nothing to do with this issue. > This particular issue is a disagreement in the slow-start > implementation. Slow-start is documented after-the-fact in RFC2001 > for anyone interested. > > The difference revolves around who should send what when; one box > is expecting the other to send something, the other is expecting > the first to send something, so they deadlock until one times out > and retramsmits. Don't have the exact details offhand and I don't > know if FreeBSD is impacted. I have never heard of this one, actually. I do know for a fact that you can end up with thousands of sockets in "WAIT FIN" with MS clients, effectively locking up UNIX servers. I guess this is just another toy surprise in a long line of toy surprises... 8-|. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Wed Jul 30 13:52:13 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA02716 for current-outgoing; Wed, 30 Jul 1997 13:52:13 -0700 (PDT) Received: from shale.csir.co.za (shale.csir.co.za [146.64.46.5]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA02653 for ; Wed, 30 Jul 1997 13:52:00 -0700 (PDT) Received: (from reg@localhost) by shale.csir.co.za (8.8.6/8.8.5) id WAA01184; Wed, 30 Jul 1997 22:52:36 +0200 (SAT) Message-ID: <19970730225235.58179@shale.csir.co.za> Date: Wed, 30 Jul 1997 22:52:35 +0200 From: Jeremy Lea To: current@freebsd.org Subject: More bogons in Makefiles... Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.76 Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi all, I've been looking through the makefiles, and messing around a bit with the make world procedure to tickle things, and I wanted to ask a few questions before I start making suggestions that are going to attract heat... 1. What is the "general" policy with Makefile's pointing to directories like ${.OBJDIR}../../foo/bar when the library (normally) in question is installed in /usr/lib? In a number of places there are references like this, which might cause problems if the module is built without a full src tree. I percieve the purpose of 'make world' to be to make sure that these libraries get installed before the build happens. Or are these constructs needed for 'make release' or 'make reinstall' or something? 2. There are a lot (15 or so, like cc, cvs) programs which have library directories, which are not installed, but are depended apon. If you cd to one of the subdirectories (without /usr/obj populated) and do a 'make all' then it falls over in the linking stage... I know: "Don't do that". Is there a clean way of building and installing the lib in ${.OBJDIR}/.. as a first step in the Makefile, and cleaning it from the other subdirectory's Makefile if only that subdirectory was built? 3. There are a lot of tests of the form: .if exists(../../secure) && !defined(NOCRYPT) && !defined(NOSECURE) It seems that NOCRYPT could be set if NOSECURE is set, and NOSECURE set if there was no directory (../../secure), tested from a Makefile or Makefile.inc in /usr/src or elsewhere, saving a bit on tests and possible mix ups. Anyway, here are some other things which I think are a bit bogus (guess I should learn to send-pr): Some people don't cvsup /usr/src/games, since they dont want the games... *** ./share/doc/usd/Makefile.orig Thu May 22 10:18:17 1997 --- ./share/doc/usd/Makefile Wed Jul 30 01:03:20 1997 *************** *** 11,16 **** BINDIR= /usr/share/doc/usd SUBDIR= title contents 04.csh 07.mail 10.exref 11.vitut 12.vi 13.viref \ ! 18.msdiffs 19.memacros 20.meref 30.rogue 31.trek .include --- 11,20 ---- BINDIR= /usr/share/doc/usd SUBDIR= title contents 04.csh 07.mail 10.exref 11.vitut 12.vi 13.viref \ ! 18.msdiffs 19.memacros 20.meref ! ! .if exists(../../../games) && !defined(NOGAMES) ! SUBDIR+= 30.rouge 31.trek ! .endif .include The purpose of dependancies is to make sure files exist... 'make all install' fails because of this: *** ./share/zoneinfo/Makefile.orig Thu May 22 11:34:30 1997 --- ./share/zoneinfo/Makefile Wed Jul 30 01:03:24 1997 *************** *** 16,22 **** TZFILES+= backward .endif ! .if exists(${.OBJDIR}/yearistype) YEARISTYPE= ${.OBJDIR}/yearistype .else YEARISTYPE= ${.CURDIR}/yearistype --- 16,22 ---- TZFILES+= backward .endif ! .if exists(${.OBJDIR}) YEARISTYPE= ${.OBJDIR}/yearistype .else YEARISTYPE= ${.CURDIR}/yearistype Something tells me this hasn't been used for a while ;): *** ./bin/Makefile.inc.orig Wed Jul 23 16:58:56 1997 --- ./bin/Makefile.inc Wed Jul 30 01:51:06 1997 *************** *** 3,15 **** BINDIR?= /bin NOSHARED?= YES - - .if exists (${.CURDIR}/../../secure) - - .if exists(${.CURDIR}/../../secure/lib/libcipher/obj) - CIPHEROBJDIR= ${.CURDIR}/../../secure/lib/libcipher/obj - .else - CIPHEROBJDIR= ${.CURDIR}/../../secure/lib/libcipher - .endif - - .endif --- 3,5 ---- *** ./bin/ed/Makefile.orig Wed Jul 23 17:15:41 1997 --- ./bin/ed/Makefile Wed Jul 30 01:11:58 1997 *************** *** 5,15 **** LINKS= ${BINDIR}/ed ${BINDIR}/red MLINKS= ed.1 red.1 ! .if exists(../../secure) && !defined(NOCRYPT) && !defined(NOSECURE) DISTRIBUTION=des CFLAGS+=-DDES ! LDADD+= -L${CIPHEROBJDIR} -lcipher ! DPADD+= ${CIPHEROBJDIR}/libcipher.a .endif .include --- 5,15 ---- LINKS= ${BINDIR}/ed ${BINDIR}/red MLINKS= ed.1 red.1 ! .if exists(${DESTDIR}/usr/lib/libcipher.a) && !defined(NOCRYPT) && !defined(NOSECURE) DISTRIBUTION=des CFLAGS+=-DDES ! LDADD+= -lcipher ! DPADD+= ${DESTDIR}/usr/lib/libcipher.a .endif .include Why doesn't libcipher have an entry in bsd.libnames.mk? and shouldn't there be some ${DESTDIR}'s near those ${BINDIR}'s (Also in ./bin/test/Makefile)? That's what ./share/mk/bsd.README says... There's lots of those around the tree. Anyway, that's all for now. I'm still trying to understand make, *.mk and the Makefiles. Oops, one last thing: should ./usr.bin/lex/lib be built twice in the bootstrapping of 'make world' (as part of lib-tools and libraries) and should the whole of ./lib be rebuilt, including stuff that's been done a minute earlier? only libc and libm are needed, as far as I can see, to make the build-tools targets before the libraries are built anyway as part of the 'make all'. -Jeremy -- .sig.gz From owner-freebsd-current Wed Jul 30 14:28:56 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA04807 for current-outgoing; Wed, 30 Jul 1997 14:28:56 -0700 (PDT) Received: from smoke.marlboro.vt.us (smoke.marlboro.vt.us [198.206.215.91]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA04798 for ; Wed, 30 Jul 1997 14:28:51 -0700 (PDT) Received: (from cgull@localhost) by smoke.marlboro.vt.us (8.8.5/8.8.5/cgull) id RAA04913; Wed, 30 Jul 1997 17:28:31 -0400 (EDT) Date: Wed, 30 Jul 1997 17:28:31 -0400 (EDT) Message-Id: <199707302128.RAA04913@smoke.marlboro.vt.us> From: john hood MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Bruce Evans Cc: cgull@smoke.marlboro.vt.us, freebsd-current@FreeBSD.ORG Subject: problems with IDE DMA In-Reply-To: <199707301756.DAA26679@godzilla.zeta.org.au> References: <199707301756.DAA26679@godzilla.zeta.org.au> X-Mailer: VM 6.31 under Emacs 19.34.2 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Bruce Evans writes: > After successfully testing IDE DMA for a little while (reading rwd0, > fsck -n and bonnie -s 64 worked fine) I got the following messages > after the write stages for bonnie -s 64: > > wd0: interrupt timeout: > wd0: status 50 error 0 > wd0: interrupt timeout: > wd0: status 58 error 1 > wd0: interrupt timeout: > wd0: status 58 error 1 > wd0: interrupt timeout: > wd0: status 58 error 1 > wd0: Last time I say: interrupt timeout. Probably a portable PC.: > wd0: status 58 error 1 from the messages, it would *seem* as though you have a bad sector on the drive. you might try dd'ing the entire raw drive in PIO mode and seeing if any error messages crop up. > The system then drifted South: > - no more interrupt timeouts, but wd seemed to get stuck in single > sector mode. ls on the bonnie output file was very slow, and > systat reported about 20 interrupts/sec, 10 "blks", and often > "***** msps" (division by 0?). curious. that's single sector mode, along with a significant delay. > - long before the 13107 seconds required to read 128MB at 10K/sec, > the system hung (the keyboard LEDs didn't work). curiouser. well, i guess the error handling *does* need a bit of work. > - the test drive was dead after hard reset. > - the test drive was OK after cycling the power. this i wouldn't worry too much about. low-end quantums use a cheesy single-CPU/DSP design for their controller, and they seem to hang without too much provocation and with no permanent effect. > Boot-time log messages: > > ... > ide_pci0: rev 0x02 on pci0.7.1 > ide_pci: warning, ide0:0 not configured for DMA? > ide_pci: warning, ide0:1 not configured for DMA? > ide_pci: warning, ide1:0 not configured for DMA? > ide_pci: warning, ide1:1 not configured for DMA? > ... > wdc0 at 0x1f0-0x1f7 irq 14 flags 0x80ffa0ff on isa > wdc0: unit 0 (wd0): , DMA, 32-bit, multi-block-16 > wd0: 76kB cache 2014MB (4124736 sectors), 4092 cyls, 16 heads, 63 S/T, 512 B/S boot -v messages, motherboard, and BIOS, please? :) --jh -- John Hood cgull@smoke.marlboro.vt.us Predictably, they all eventually wandered away, rubbing their bruises and brushing mud out of their hair. Some went off to work for the ESA, launching much smaller rockets into low orbits, while others elected to sit on their front porches drinking Jim Beam from the bottle and launching bottle rockets from the empties. [Jordan Hubbard] From owner-freebsd-current Wed Jul 30 15:49:17 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA09986 for current-outgoing; Wed, 30 Jul 1997 15:49:17 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [206.246.122.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id PAA09981 for ; Wed, 30 Jul 1997 15:49:13 -0700 (PDT) Received: from Journey2.mat.net (journey2.mat.net [206.246.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id SAA29670 for ; Wed, 30 Jul 1997 18:49:10 -0400 Date: Wed, 30 Jul 1997 18:49:25 -0400 (EDT) From: Chuck Robey X-Sender: chuckr@Journey2.mat.net To: FreeBSD current Subject: core group topics Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk I was wondering if core has been discussing the two topics lately that can't really move without core approval? 1) Satoshi's new Makefile. His makefile has been around plenty long enough for testing, and while several people have reported good results, I've seen no negatives reported at all. Will it get ignored or committed? 2) John Polstra's ELF work. He's done all the work that anyone could possibly want, to move FreeBSD to ELF. There are several good reasons to move, and none excepting inertia not to. I know that moving FreeBSD to ELF isn't a high priority project, but all the work's been done, the benefits are real, and it's high time to consider it. I don't necessarily want immediate decisions here, but I'd like some assurance core is paying attention to these. Could someone post at least some kind of overall summary of the current feeling on these projects? I read John Polstra's comments about lack of movement, and it seems that what he said was well justified. ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 213 Lakeside Drive Apt T-1 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-current Wed Jul 30 16:19:07 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA11865 for current-outgoing; Wed, 30 Jul 1997 16:19:07 -0700 (PDT) Received: from hda.hda.com (hda-bicnet.bicnet.net [208.220.66.37]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA11856 for ; Wed, 30 Jul 1997 16:19:00 -0700 (PDT) Received: (from dufault@localhost) by hda.hda.com (8.8.5/8.8.5) id SAA02017; Wed, 30 Jul 1997 18:32:31 -0400 (EDT) From: Peter Dufault Message-Id: <199707302232.SAA02017@hda.hda.com> Subject: Re: where to put access restriction for scheduling classes In-Reply-To: <199707301443.AAA00985@labs.usn.blaze.net.au> from David Nugent at "Jul 31, 97 00:43:47 am" To: davidn@labs.usn.blaze.net.au (David Nugent) Date: Wed, 30 Jul 1997 18:32:30 -0400 (EDT) Cc: current@freebsd.org X-Mailer: ELM [version 2.4ME+ PL25 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > Just out of curiosity, what type of resource are we actually discussing > here? A limit on which process schedulers the process and descendants can select, to permit other than root to select scheduling policies that could hang the system given a bug or purposeful denial of services attack. Peter -- Peter Dufault (dufault@hda.com) Realtime development, Machine control, HD Associates, Inc. Safety critical systems, Agency approval From owner-freebsd-current Wed Jul 30 17:23:31 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA15805 for current-outgoing; Wed, 30 Jul 1997 17:23:31 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA15787 for ; Wed, 30 Jul 1997 17:23:18 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id RAA19296; Wed, 30 Jul 1997 17:21:29 -0700 (PDT) Message-Id: <199707310021.RAA19296@austin.polstra.com> To: Terry Lambert cc: msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( In-reply-to: Your message of "Wed, 30 Jul 1997 10:05:24 PDT." <199707301705.KAA05229@phaeton.artisoft.com> Date: Wed, 30 Jul 1997 17:21:29 -0700 From: John Polstra Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > > * Solve the booting problem. My hacked-up bootloader was within > > about 6 bytes of the maximum size -- way too big to commit. GRUB? > > An a.out wrapper for the ELF kernel? > > Frankly, I'd be happy to live with your hacked-up boot loader. OK: ftp://ftp.polstra.com/pub/FreeBSD/elfboot/elfboot-1.0.tar.gz Here is the README.ELF file: This is a hacked-up version of the FreeBSD bootloader that can load both ELF and a.out kernels. The modifications are based on a set of patches received from Bryan Ford . This work was done between March, 1997 and May, 1997, starting with the version of the bootloader from FreeBSD-current at that time. I have not attempted to incorporate more recent changes into it. It is possible that you will need to embed it in a "/usr/src/sys" tree of the same vintage, in order to get the right versions of the include files. This bootloader is not suitable for committing to FreeBSD, because it is too large. The limit is 14 sectors (7168 bytes). This bootloader is currently only 32 bytes under that (7136 bytes). John Polstra 30 July 1997 -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Wed Jul 30 20:12:17 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA26982 for current-outgoing; Wed, 30 Jul 1997 20:12:17 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA26974; Wed, 30 Jul 1997 20:12:08 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id MAA25330; Thu, 31 Jul 1997 12:41:49 +0930 (CST) From: Michael Smith Message-Id: <199707310311.MAA25330@genesis.atrad.adelaide.edu.au> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707301924.PAA08936@hurricane.cs.duke.edu> from Andrew Gallatin at "Jul 30, 97 03:24:54 pm" To: gallatin@CS.Duke.EDU (Andrew Gallatin) Date: Thu, 31 Jul 1997 12:41:49 +0930 (CST) Cc: se@FreeBSD.ORG, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Andrew Gallatin stands accused of saying: > > I suppose this is as good a time to ask as any: I have a FreeBSD > driver 95% done for Myrinet cards such as this. Who can I talk to > about the possibility of getting it committed to -current? There are > some sticky issues that may prevent its inclusion (specifically > GPL'ed libraries that are required to communicate w/the vendor > supported "firmware" that runs on the card & which I cannot think of a > clean way to build). The GPL'd libraries won't prevent it being included; there is a precedent for handling useful GPL kernel components. If nobody else sticks their hand up, I'd be happy to do the actual grunt work involved in committing it. > Drew -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 20:16:08 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA27239 for current-outgoing; Wed, 30 Jul 1997 20:16:08 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA27092 for ; Wed, 30 Jul 1997 20:15:23 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id MAA25339; Thu, 31 Jul 1997 12:43:24 +0930 (CST) From: Michael Smith Message-Id: <199707310313.MAA25339@genesis.atrad.adelaide.edu.au> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970730155201.26891@scsn.net> from "Donald J. Maddox" at "Jul 30, 97 03:52:01 pm" To: dmaddox@scsn.net Date: Thu, 31 Jul 1997 12:43:24 +0930 (CST) Cc: msmith@atrad.adelaide.edu.au, dmaddox@scsn.net, sos@sos.freebsd.dk, cgull@smoke.marlboro.vt.us, doogie@forbidden-donut.anet-stl.com, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Donald J. Maddox stands accused of saying: > > > > Turn on IDE busmastering in your BIOS. Note the syntax of the warning > > "not configured". That's a message from the chipset driver daying > > "why is this channel not configured for DMA?". It should probably > > just list the ports on which DMA is available and enabled/not enabled, > > but this is, after all, only beta-level code. > > There is no such option that I can see in my BIOS setup. I am using an > ASUS P/I-P55T2P4 mainboard with BIOS version 2.047, but I can't recall ever > seeing any such option in a BIOS from ASUS. Then ignore the warnings and be happy; that's all they are. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 20:17:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA27327 for current-outgoing; Wed, 30 Jul 1997 20:17:43 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA27315; Wed, 30 Jul 1997 20:17:24 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id MAA25355; Thu, 31 Jul 1997 12:47:07 +0930 (CST) From: Michael Smith Message-Id: <199707310317.MAA25355@genesis.atrad.adelaide.edu.au> Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <19970730220038.02422@mi.uni-koeln.de> from Stefan Esser at "Jul 30, 97 10:00:38 pm" To: se@FreeBSD.ORG (Stefan Esser) Date: Thu, 31 Jul 1997 12:47:07 +0930 (CST) Cc: gallatin@cs.duke.edu, freebsd-current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Stefan Esser stands accused of saying: > > (*) PCI device == a Myrinet M2F-PCI32 card. This is a programmable > > gigabit networking card. It has a 256k bank of SRAM on the card, and > > is very good for doing things like measuring PCI b/w. The tests were > > done from user space operating on mmap'ed device memory & a kernel > > allocated chunk of RAM to do DMA xfers to/from. It also runs IP > > traffic at better than 300Mb/sec. > > This was a memory mapped device buffer ? > > That's *very* different from an I/O port, and you should be able > to copy data to that SRAM at a much higher rate than 8 or 13MB/s! If I read the AMCC PCI matchmaker databook correctly, PIO to a memory-mapped buffer is likely to be severely affected by whether the PCI bridge in question supports cache line read/write operations. Is this a potential issue here? -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 20:21:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA27627 for current-outgoing; Wed, 30 Jul 1997 20:21:26 -0700 (PDT) Received: from bne16unx226.workcover.qld.gov.au (server.workcover.qld.gov.au [203.101.253.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA27607; Wed, 30 Jul 1997 20:21:17 -0700 (PDT) Received: from manila.workcover.qld.gov.au (bne16unx215 [172.22.16.215]) by bne16unx226.workcover.qld.gov.au (8.8.5/8.8.6) with SMTP id NAA15026; Thu, 31 Jul 1997 13:38:25 +1000 (EST) Received: from localhost by manila.workcover.qld.gov.au (8.6.8.1/DEVETIR-0.1) id DAA15629; Thu, 31 Jul 1997 03:22:13 GMT Message-Id: <199707310322.DAA15629@manila.workcover.qld.gov.au> X-Mailer: exmh version 2.0zeta 7/24/97 To: ports@freebsd.org, current@freebsd.org Subject: Please, can someone put a tk8.0b2 port into ports current? X-Face: 3}heU+2?b->-GSF-G4T4>jEB9~FR(V9lo&o>kAy=Pj&;oVOc<|pr%I/VSG"ZD32J>5gGC0N 7gj]^GI@M:LlqNd]|(2OxOxy@$6@/!,";-!OlucF^=jq8s57$%qXd/ieC8DhWmIy@J1AcnvSGV\|*! >Bvu7+0h4zCY^]{AxXKsDTlgA2m]fX$W@'8ev-Qi+-;%L'CcZ'NBL!@n?}q!M&Em3*eW7,093nOeV8 M)(u+6D;%B7j\XA/9j4!Gj~&jYzflG[#)E9sI&Xe9~y~Gn%fA7>F:YKr"Wx4cZU*6{^2ocZ!YyR Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 31 Jul 1997 13:22:13 +1000 From: Stephen Hocking Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I couldn't give a fig if it's beta - I think it has a place after the recent tcl8.0 merge. Stephen -- The views expressed above are not those of WorkCover Queensland, Australia. "We've heard that a million monkeys at a million keyboards could produce the Complete Works of Shakespeare; now, thanks to the Internet, we know this is not true." Robert Wilensky, University of California From owner-freebsd-current Wed Jul 30 20:31:08 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA28062 for current-outgoing; Wed, 30 Jul 1997 20:31:08 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA28056 for ; Wed, 30 Jul 1997 20:31:02 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id NAA25430; Thu, 31 Jul 1997 13:00:21 +0930 (CST) From: Michael Smith Message-Id: <199707310330.NAA25430@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: from Chuck Robey at "Jul 30, 97 06:49:25 pm" To: chuckr@glue.umd.edu (Chuck Robey) Date: Thu, 31 Jul 1997 13:00:21 +0930 (CST) Cc: freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Chuck Robey stands accused of saying: > > 1) Satoshi's new Makefile. His makefile has been around plenty long > enough for testing, and while several people have reported good results, > I've seen no negatives reported at all. Will it get ignored or committed? I'll second this; I've been using it off and on for a few days now and have noticed no ill effects. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Wed Jul 30 22:54:38 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA06128 for current-outgoing; Wed, 30 Jul 1997 22:54:38 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA06121 for ; Wed, 30 Jul 1997 22:54:34 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.6/8.6.9) id WAA14109; Wed, 30 Jul 1997 22:54:25 -0700 (PDT) Date: Wed, 30 Jul 1997 22:54:25 -0700 (PDT) Message-Id: <199707310554.WAA14109@silvia.HIP.Berkeley.EDU> To: chuckr@glue.umd.edu CC: freebsd-current@freefall.FreeBSD.org In-reply-to: (message from Chuck Robey on Wed, 30 Jul 1997 18:49:25 -0400 (EDT)) Subject: Re: core group topics From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk * I was wondering if core has been discussing the two topics lately that * can't really move without core approval? Sorry, I've been busy/away for the past few days. * 1) Satoshi's new Makefile. His makefile has been around plenty long This is not a "policy" issue anyway, I don't think core's approval is needed. (Right, the rest of core? :) * enough for testing, and while several people have reported good results, * I've seen no negatives reported at all. Will it get ignored or committed? It will be committed soon. I had the impression that Bruce was unhappy about some of the environment variables, but haven't gotten around to verify what exactly is the fix. However, I don't have a -current machine to test anymore, so I'm not sure if I can support -current. I'll send out a patch for -stable later tonight. * 2) John Polstra's ELF work. He's done all the work that anyone could * possibly want, to move FreeBSD to ELF. There are several good reasons to * move, and none excepting inertia not to. I know that moving FreeBSD to * ELF isn't a high priority project, but all the work's been done, the * benefits are real, and it's high time to consider it. This would have to be decided by core. Myself, I have no idea what ELF offers that a.out doesn't (other than what John said), but if we can get rid of the "text relocation error" that pops up all over the place during package builds, that's good news enough for me. :) Satoshi From owner-freebsd-current Wed Jul 30 22:58:23 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA06458 for current-outgoing; Wed, 30 Jul 1997 22:58:23 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA06438; Wed, 30 Jul 1997 22:58:17 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.6/8.6.9) id WAA14118; Wed, 30 Jul 1997 22:58:06 -0700 (PDT) Date: Wed, 30 Jul 1997 22:58:06 -0700 (PDT) Message-Id: <199707310558.WAA14118@silvia.HIP.Berkeley.EDU> To: sysseh@workcover.qld.gov.au CC: ports@freebsd.org, current@freebsd.org In-reply-to: <199707310322.DAA15629@manila.workcover.qld.gov.au> (message from Stephen Hocking on Thu, 31 Jul 1997 13:22:13 +1000) Subject: Re: Please, can someone put a tk8.0b2 port into ports current? From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk ports-current will be discontinued -- announcement to follow later. * I couldn't give a fig if it's beta - I think it has a place after the * recent tcl8.0 merge. Whoever does the port (Chuck, didn't you say you had them working?) please include a tcl8.0beta2 port too, so they can be compiled and tested on -stable machines. Satoshi From owner-freebsd-current Wed Jul 30 23:38:13 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA09207 for current-outgoing; Wed, 30 Jul 1997 23:38:13 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA09200 for ; Wed, 30 Jul 1997 23:38:11 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id XAA19898; Wed, 30 Jul 1997 23:37:16 -0700 (PDT) To: Jeremy Lea cc: current@FreeBSD.ORG Subject: Re: More bogons in Makefiles... In-reply-to: Your message of "Wed, 30 Jul 1997 22:52:35 +0200." <19970730225235.58179@shale.csir.co.za> Date: Wed, 30 Jul 1997 23:37:15 -0700 Message-ID: <19894.870331035@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > 1. What is the "general" policy with Makefile's pointing to directories like > ${.OBJDIR}../../foo/bar when the library (normally) in question is > installed in /usr/lib? In a number of places there are references like Ah, but maybe it's not yet, eh? Or maybe it's entirely the wrong version because you're building a source tree for a different release of FreeBSD than the build host is running. Assuming anything about the contents of the final destination directories for what may be a simple build is highly bogus and, while this is still done in a number of places in /usr/src, the eventual intent is to get rid of this and make /usr/src build in a properly self-referential fashion. > src tree. I percieve the purpose of 'make world' to be to make sure that > these libraries get installed before the build happens. Or are these It's the current kludge of `make world' to do this, but it's still not its most ideal purpose. :) > Is there a clean way of building and installing the lib in > ${.OBJDIR}/.. as a first step in the Makefile, and cleaning it from the > other subdirectory's Makefile if only that subdirectory was built? Not really, no. Not a clean way. :-) > [ numerous bogons identified and diff'd ] Yes, as you can see, things aren't always 100% consistent in there and you've certainly identified some definite bogons. I or someone will review and commit from this list fairly shortly. > Anyway, that's all for now. I'm still trying to understand make, *.mk and > the Makefiles. You seem to be doing pretty well so far. :-) Jordan From owner-freebsd-current Wed Jul 30 23:45:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA09734 for current-outgoing; Wed, 30 Jul 1997 23:45:14 -0700 (PDT) Received: from ghpc8.ihf.rwth-aachen.de (ghpc8.ihf.RWTH-Aachen.DE [134.130.90.8]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA09728; Wed, 30 Jul 1997 23:45:07 -0700 (PDT) Received: from ghpc6.ihf.rwth-aachen.de (ghpc6.ihf.rwth-aachen.de [134.130.90.6]) by ghpc8.ihf.rwth-aachen.de (8.8.5/8.8.5) with ESMTP id IAA09706; Thu, 31 Jul 1997 08:38:37 +0200 (CEST) Received: (from thomas@localhost) by ghpc6.ihf.rwth-aachen.de (8.8.6/8.8.5) id IAA05521; Thu, 31 Jul 1997 08:38:36 +0200 (CEST) To: asami@cs.berkeley.edu (Satoshi Asami) Cc: sysseh@workcover.qld.gov.au, ports@FreeBSD.ORG, current@FreeBSD.ORG Subject: Re: Please, can someone put a tk8.0b2 port into ports current? References: <199707310558.WAA14118@silvia.HIP.Berkeley.EDU> From: Thomas Gellekum Date: 31 Jul 1997 08:38:35 +0200 In-Reply-To: asami@cs.berkeley.edu's message of Wed, 30 Jul 1997 22:58:06 -0700 (PDT) Message-ID: <874t9b915g.fsf@ghpc6.ihf.rwth-aachen.de> Lines: 14 X-Mailer: Gnus v5.4.37/XEmacs 19.15 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk asami@cs.berkeley.edu (Satoshi Asami) writes: > ports-current will be discontinued -- announcement to follow later. > > * I couldn't give a fig if it's beta - I think it has a place after the > * recent tcl8.0 merge. > > Whoever does the port (Chuck, didn't you say you had them working?) > please include a tcl8.0beta2 port too, so they can be compiled and > tested on -stable machines. Yeah, and don't forget the proper dependencies, please. :-] tg From owner-freebsd-current Wed Jul 30 23:52:15 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA10237 for current-outgoing; Wed, 30 Jul 1997 23:52:15 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA10224 for ; Wed, 30 Jul 1997 23:52:09 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id XAA19994; Wed, 30 Jul 1997 23:51:54 -0700 (PDT) To: Chuck Robey cc: FreeBSD current Subject: Re: core group topics In-reply-to: Your message of "Wed, 30 Jul 1997 18:49:25 EDT." Date: Wed, 30 Jul 1997 23:51:53 -0700 Message-ID: <19990.870331913@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > 1) Satoshi's new Makefile. His makefile has been around plenty long > enough for testing, and while several people have reported good results, > I've seen no negatives reported at all. Will it get ignored or committed? I think it's hung up on the question of what to do with LDDESTDIR - all agree that it's broken as currently used now but no one wants to bell the cat, either. :-) > 2) John Polstra's ELF work. He's done all the work that anyone could > possibly want, to move FreeBSD to ELF. There are several good reasons to > move, and none excepting inertia not to. I know that moving FreeBSD to Erm, "none excepting inertia and fear of chaos, confusion, shock and increased entropy all around." :) Inflicting another ABI on our user base, one which will result in a new class of packages which all older FreeBSD releases cannot run, is NOT something to be considered lightly. Jordan From owner-freebsd-current Thu Jul 31 00:09:57 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA11401 for current-outgoing; Thu, 31 Jul 1997 00:09:57 -0700 (PDT) Received: from critter.dk.tfs.com (critter.phk.freebsd.dk [195.8.133.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA11389 for ; Thu, 31 Jul 1997 00:09:48 -0700 (PDT) Received: from critter.dk.tfs.com (localhost [127.0.0.1]) by critter.dk.tfs.com (8.8.6/8.8.5) with ESMTP id JAA00689; Thu, 31 Jul 1997 09:09:13 +0200 (CEST) To: Chuck Robey cc: FreeBSD current From: Poul-Henning Kamp Subject: Re: core group topics In-reply-to: Your message of "Wed, 30 Jul 1997 18:49:25 EDT." Date: Thu, 31 Jul 1997 09:09:12 +0200 Message-ID: <687.870332952@critter.dk.tfs.com> Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk I'm "stuck" in my beachhouse for the time being, with only my laptop, so I havn't had time to test either. Until I have, I will not endorse either of course. I belive both is just a case of testing, agreeing, committing. And I don't thing it's been dropped on the floor or anything. Things like this take time. Poul-Henning In message , Chuck Rob ey writes: >I was wondering if core has been discussing the two topics lately that >can't really move without core approval? > >1) Satoshi's new Makefile. His makefile has been around plenty long >enough for testing, and while several people have reported good results, >I've seen no negatives reported at all. Will it get ignored or committed? > >2) John Polstra's ELF work. He's done all the work that anyone could >possibly want, to move FreeBSD to ELF. There are several good reasons to >move, and none excepting inertia not to. I know that moving FreeBSD to >ELF isn't a high priority project, but all the work's been done, the >benefits are real, and it's high time to consider it. > >I don't necessarily want immediate decisions here, but I'd like some >assurance core is paying attention to these. Could someone post at least >some kind of overall summary of the current feeling on these projects? I >read John Polstra's comments about lack of movement, and it seems that >what he said was well justified. > >----------------------------+----------------------------------------------- >Chuck Robey | Interests include any kind of voice or data >chuckr@eng.umd.edu | communications topic, C programming, and Unix. >213 Lakeside Drive Apt T-1 | >Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD >(301) 220-2114 | version 3.0 current -- and great FUN! >----------------------------+----------------------------------------------- > -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. Power and ignorance is a disgusting cocktail. From owner-freebsd-current Thu Jul 31 00:14:25 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA11609 for current-outgoing; Thu, 31 Jul 1997 00:14:25 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [206.246.122.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id AAA11604 for ; Thu, 31 Jul 1997 00:14:20 -0700 (PDT) Received: from Journey2.mat.net (journey2.mat.net [206.246.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id DAA26972; Thu, 31 Jul 1997 03:14:12 -0400 Date: Thu, 31 Jul 1997 03:14:25 -0400 (EDT) From: Chuck Robey X-Sender: chuckr@Journey2.mat.net To: "Jordan K. Hubbard" cc: FreeBSD current Subject: Re: core group topics In-Reply-To: <19990.870331913@time.cdrom.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, 30 Jul 1997, Jordan K. Hubbard wrote: > > 1) Satoshi's new Makefile. His makefile has been around plenty long > > enough for testing, and while several people have reported good results, > > I've seen no negatives reported at all. Will it get ignored or committed? > > I think it's hung up on the question of what to do with LDDESTDIR - > all agree that it's broken as currently used now but no one wants to > bell the cat, either. :-) > > > 2) John Polstra's ELF work. He's done all the work that anyone could > > possibly want, to move FreeBSD to ELF. There are several good reasons to > > move, and none excepting inertia not to. I know that moving FreeBSD to > > Erm, "none excepting inertia and fear of chaos, confusion, shock and > increased entropy all around." :) Inflicting another ABI on our user > base, one which will result in a new class of packages which all older > FreeBSD releases cannot run, is NOT something to be considered > lightly. I agree. How long have we had to discuss it so far? I think you'll have to agree, it's been more than long enough. It feels more like a case of "if we ignore it long enough, maybe it'll go away." There are ways to avoid the confusion, and John's been diligent in getting the thing working compatibly. How about we move the discussion from "ignore" stage to an enumeration of possible transition problems, so that last problems can be solved? I don't honestly think they're all that bad, myself. Suggest what your worst worry is, and either John can handle it, we decide we can live with it, or it's too much, and (as John just spoke of) we just let the opportunity go away. John's work deserves an open review, not unnamed fears. > > Jordan > > ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 213 Lakeside Drive Apt T-1 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-current Thu Jul 31 00:22:32 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA11996 for current-outgoing; Thu, 31 Jul 1997 00:22:32 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA11987 for ; Thu, 31 Jul 1997 00:22:19 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id QAA26987; Thu, 31 Jul 1997 16:51:50 +0930 (CST) From: Michael Smith Message-Id: <199707310721.QAA26987@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: <19990.870331913@time.cdrom.com> from "Jordan K. Hubbard" at "Jul 30, 97 11:51:53 pm" To: jkh@time.cdrom.com (Jordan K. Hubbard) Date: Thu, 31 Jul 1997 16:51:50 +0930 (CST) Cc: chuckr@glue.umd.edu, freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Jordan K. Hubbard stands accused of saying: > > 1) Satoshi's new Makefile. His makefile has been around plenty long > > enough for testing, and while several people have reported good results, > > I've seen no negatives reported at all. Will it get ignored or committed? > > I think it's hung up on the question of what to do with LDDESTDIR - > all agree that it's broken as currently used now but no one wants to > bell the cat, either. :-) Hm, I think we just saw that one die. Three cheers to Satoshi! > > 2) John Polstra's ELF work. He's done all the work that anyone could > > possibly want, to move FreeBSD to ELF. There are several good reasons to > > move, and none excepting inertia not to. I know that moving FreeBSD to > > Erm, "none excepting inertia and fear of chaos, confusion, shock and > increased entropy all around." :) Inflicting another ABI on our user > base, one which will result in a new class of packages which all older > FreeBSD releases cannot run, is NOT something to be considered > lightly. I think "3.0" is about as big and scary as things get. If we don't do it here, it's going to be another two or thee years before we get another chance, and by then what is a minor pain now is going to look and feel like a Mack-branded suppository. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Thu Jul 31 00:35:19 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA12900 for current-outgoing; Thu, 31 Jul 1997 00:35:19 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA12895 for ; Thu, 31 Jul 1997 00:35:15 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id AAA20288; Thu, 31 Jul 1997 00:34:53 -0700 (PDT) To: Chuck Robey cc: FreeBSD current Subject: Re: core group topics In-reply-to: Your message of "Thu, 31 Jul 1997 03:14:25 EDT." Date: Thu, 31 Jul 1997 00:34:53 -0700 Message-ID: <20284.870334493@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > I agree. How long have we had to discuss it so far? I think you'll have > to agree, it's been more than long enough. It feels more like a case of > "if we ignore it long enough, maybe it'll go away." There are ways to Well, that's one way to look at it. Another way to look at it is that a significant percentage of people looking at the whole idea have so far gone "No, I'm outright OPPOSED to this idea. I do not want ELF." It's not as passive a situation as you might think. Jordan From owner-freebsd-current Thu Jul 31 00:42:40 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA13501 for current-outgoing; Thu, 31 Jul 1997 00:42:40 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA13496 for ; Thu, 31 Jul 1997 00:42:35 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id RAA27102; Thu, 31 Jul 1997 17:12:11 +0930 (CST) From: Michael Smith Message-Id: <199707310742.RAA27102@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: from Chuck Robey at "Jul 31, 97 03:14:25 am" To: chuckr@glue.umd.edu (Chuck Robey) Date: Thu, 31 Jul 1997 17:12:10 +0930 (CST) Cc: jkh@time.cdrom.com, freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Chuck Robey stands accused of saying: [ELF transition] > "if we ignore it long enough, maybe it'll go away." There are ways to > avoid the confusion, and John's been diligent in getting the thing working > compatibly. How about we move the discussion from "ignore" stage to an > enumeration of possible transition problems, so that last problems can be > solved? I don't honestly think they're all that bad, myself. Er, we've been doing this on -hackers the last couple of days. 8) Great minds and all that stuff... -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Thu Jul 31 00:55:54 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA14297 for current-outgoing; Thu, 31 Jul 1997 00:55:54 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA14292 for ; Thu, 31 Jul 1997 00:55:51 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id RAA27165; Thu, 31 Jul 1997 17:25:16 +0930 (CST) From: Michael Smith Message-Id: <199707310755.RAA27165@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: <20284.870334493@time.cdrom.com> from "Jordan K. Hubbard" at "Jul 31, 97 00:34:53 am" To: jkh@time.cdrom.com (Jordan K. Hubbard) Date: Thu, 31 Jul 1997 17:25:15 +0930 (CST) Cc: chuckr@glue.umd.edu, freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Jordan K. Hubbard stands accused of saying: > > Well, that's one way to look at it. Another way to look at it is that > a significant percentage of people looking at the whole idea have so > far gone "No, I'm outright OPPOSED to this idea. I do not want ELF." > It's not as passive a situation as you might think. Beg to differ, but most of the people that have stood up in opposition have not stated an opposition to ELF per se, but as you have previously observed to the significant transitional problems that would be bound to surface, particularly in the light of their perception of the benefits involved. John P. put it succintly when he said that we need to think about what ELF _would_ buy us, not what it might cost. IMHO Jeff Hsu's observation about Java is almost enough weight on its own; if we want to be taken seriously as a development and server platform, the JDK is going to be a mandatory component. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Thu Jul 31 00:58:46 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id AAA14522 for current-outgoing; Thu, 31 Jul 1997 00:58:46 -0700 (PDT) Received: from internet1.mel.cybec.com.au (internet1.mel.cybec.com.au [203.103.154.130]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id AAA14514 for ; Thu, 31 Jul 1997 00:58:36 -0700 (PDT) Received: from tech34 (tech34.mel.cybec.com.au [203.103.154.37]) by internet1.mel.cybec.com.au (post.office MTA v2.0 0813 ID# 0-14031) with ESMTP id AAA671; Thu, 31 Jul 1997 17:59:46 +1000 Message-ID: <33E04610.81C2C2EC@cybec.com.au> Date: Thu, 31 Jul 1997 18:00:16 +1000 From: TLiddelow@cybec.com.au (Tim Liddelow) X-Mailer: Mozilla 4.01 [en] (WinNT; I) MIME-Version: 1.0 To: "Jordan K. Hubbard" CC: Chuck Robey , FreeBSD current Subject: Re: core group topics X-Priority: 3 (Normal) References: <19990.870331913@time.cdrom.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Jordan K. Hubbard wrote: > > Erm, "none excepting inertia and fear of chaos, confusion, shock and > increased entropy all around." :) Inflicting another ABI on our user > base, one which will result in a new class of packages which all older > FreeBSD releases cannot run, is NOT something to be considered > lightly. > > Jordan It doesn't have to be _that_ hard. The Linux camp did the dual ABI thing for quite a while. Sure - the transition wasn't all beer and skittles but they did it. And they're reaping the rewards (see previous postings) for it. Another thing - so what if OLDER FreeBSD releases can't run the ELF packages ? You have symlinks for packages for a certain release, right ? There's no way if I am running FreeBSD 4.0 (ELF) that I would try to add an a.out package. And vise versa. OLDER FreeBSD releases will continue to use the a.out packages. New packages would have to be made for both 2.1.7.x release(s) and the newer ELF release. Noone is suggesting there aren't going to be hurdles. What has to be considered is the cost in remaining a.out, and the advantages of ELF in the longer term. We rely on third party (GNU) software that supports ELF "out of the box". If we continue down the a.out path for too long, we risk obscurity in terms of language support, etc. When gcc 2.8 comes out (fingers crossed) will it support a.out correctly ? C++ support on FreeBSD is something that really interests/concerns me. FreeBSD plugs itself as a great networking/internet server. What about a development box ? ELF is great for development - e.g. Java, C++, etc. Just my $0.02. Tim. From owner-freebsd-current Thu Jul 31 02:53:59 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id CAA23960 for current-outgoing; Thu, 31 Jul 1997 02:53:59 -0700 (PDT) Received: from access.kuwait.net (access.kuwait.net [194.54.234.234]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id CAA23938 for ; Thu, 31 Jul 1997 02:53:40 -0700 (PDT) Received: from flame(src addr [194.54.235.33]) (677 bytes) by access.kuwait.net via sendmail with P\:smtp/R:inet_hosts/T:smtp (sender: ) id for ; Thu, 31 Jul 1997 12:52:47 +0300 (GMT) (Smail-3.2.0.96 1997-Jun-2 #2 built 1997-Jul-14) Message-Id: Comments: Authenticated sender is From: flame@access.kuwait.net To: freebsd-current@freefall.FreeBSD.org Date: Thu, 31 Jul 1997 12:53:58 +0000 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Subject: unsubscribe Priority: normal In-reply-to: <199707310755.RAA27165@genesis.atrad.adelaide.edu.au> References: <20284.870334493@time.cdrom.com> from "Jordan K. Hubbard" at "Jul 31, 97 00:34:53 am" X-mailer: Pegasus Mail for Win32 (v2.53/R1) Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk unsubscribe From owner-freebsd-current Thu Jul 31 03:03:27 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA24541 for current-outgoing; Thu, 31 Jul 1997 03:03:27 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA24536 for ; Thu, 31 Jul 1997 03:03:21 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.6/8.6.9) id DAA13501; Thu, 31 Jul 1997 03:03:12 -0700 (PDT) Date: Thu, 31 Jul 1997 03:03:12 -0700 (PDT) Message-Id: <199707311003.DAA13501@silvia.HIP.Berkeley.EDU> To: jkh@time.cdrom.com CC: chuckr@glue.umd.edu, freebsd-current@freefall.FreeBSD.org In-reply-to: <19990.870331913@time.cdrom.com> (jkh@time.cdrom.com) Subject: Re: core group topics From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk * I think it's hung up on the question of what to do with LDDESTDIR - * all agree that it's broken as currently used now but no one wants to * bell the cat, either. :-) The LDDESTDIR problem was solved a long time ago, at least as far as my patches are concerned. Whether we go in and remove all the other LDDESTDIRs in the tree or not, is an entirely different topic. Satoshi From owner-freebsd-current Thu Jul 31 03:08:59 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA24781 for current-outgoing; Thu, 31 Jul 1997 03:08:59 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA24772 for ; Thu, 31 Jul 1997 03:08:53 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.6/8.6.9) id DAA14325; Thu, 31 Jul 1997 03:07:23 -0700 (PDT) Date: Thu, 31 Jul 1997 03:07:23 -0700 (PDT) Message-Id: <199707311007.DAA14325@silvia.HIP.Berkeley.EDU> To: reg@shale.csir.co.za CC: current@FreeBSD.ORG In-reply-to: <19894.870331035@time.cdrom.com> (jkh@time.cdrom.com) Subject: Re: More bogons in Makefiles... From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk * Anyway, that's all for now. I'm still trying to understand make, *.mk and * the Makefiles. You know, if you really want to do some work here, try calling the source tree something else than "/usr/src" and try to build world from there. You will be surprised to see just how many places that string is hard-coded in. Satoshi From owner-freebsd-current Thu Jul 31 03:13:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA25126 for current-outgoing; Thu, 31 Jul 1997 03:13:22 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA25115 for ; Thu, 31 Jul 1997 03:13:16 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id DAA01924; Thu, 31 Jul 1997 03:12:49 -0700 (PDT) To: asami@cs.berkeley.edu (Satoshi Asami) cc: chuckr@Glue.umd.edu, freebsd-current@freefall.FreeBSD.org Subject: Re: core group topics In-reply-to: Your message of "Wed, 30 Jul 1997 22:54:25 PDT." <199707310554.WAA14109@silvia.HIP.Berkeley.EDU> Date: Thu, 31 Jul 1997 03:12:48 -0700 Message-ID: <1921.870343968@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > However, I don't have a -current machine to test anymore, so I'm not > sure if I can support -current. I'll send out a patch for -stable I'm building a new thud right now. You will have your -current box very shortly. Jordan From owner-freebsd-current Thu Jul 31 03:22:35 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA25888 for current-outgoing; Thu, 31 Jul 1997 03:22:35 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA25878 for ; Thu, 31 Jul 1997 03:22:29 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id UAA26613; Thu, 31 Jul 1997 20:18:54 +1000 Date: Thu, 31 Jul 1997 20:18:54 +1000 From: Bruce Evans Message-Id: <199707311018.UAA26613@godzilla.zeta.org.au> To: current@FreeBSD.ORG, reg@shale.csir.co.za Subject: Re: More bogons in Makefiles... Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >1. What is the "general" policy with Makefile's pointing to directories like > ${.OBJDIR}../../foo/bar when the library (normally) in question is > installed in /usr/lib? In a number of places there are references like > this, which might cause problems if the module is built without a full > src tree. Paths relative to the source tree are preferred, although this causes problems with partial source trees. Paths relative to the object tree are used by many sloppy Makefiles (this works if the object tree has the same layout as the source tree, and fails messily if you do things like "cd /usr/src/gnu/usr.bin/cc/cc; mkdir obj; make COPTS='-g -static' to attempt to build a temporary debugging version of cc. This should work if the rest of cc has already been made, and used to work when the object tree was inside the source tree (perhaps actually elsewhere with a per-source directory obj symlink)). Non sloppy makefiles would use a more complicated search for the object directory that would actually be used if `make' were run in the external source directory. > I percieve the purpose of 'make world' to be to make sure that > these libraries get installed before the build happens. Or are these `make world' attempts to do this, among other things, but only for standard libraries. It fails in many cases when DESTDIR is set (the libraries get installed under ${DESTDIR} so the old libraries in /usr/lib get linked to). >2. There are a lot (15 or so, like cc, cvs) programs which have library > directories, which are not installed, but are depended apon. If you cd to > one of the subdirectories (without /usr/obj populated) and do a 'make > all' then it falls over in the linking stage... I know: "Don't do that". > Is there a clean way of building and installing the lib in > ${.OBJDIR}/.. as a first step in the Makefile, and cleaning it from the > other subdirectory's Makefile if only that subdirectory was built? There doesn't seem to be a clean way. tn3270 attempts to build things outside the current directory, but makes a mess of it. >3. There are a lot of tests of the form: > > .if exists(../../secure) && !defined(NOCRYPT) && !defined(NOSECURE) The path here should be ${.CURDIR/../../secure}. `make' chdir's to the object directory early, so most paths are relative to the object directory (relative `.include' paths are searched for first in the source directory, then in the current (object) directory). > It seems that NOCRYPT could be set if NOSECURE is set, and NOSECURE set > if there was no directory (../../secure), tested from a Makefile or > Makefile.inc in /usr/src or elsewhere, saving a bit on tests and possible > mix ups. The idea is for optional sources to automagically be used completely if they exist. >--- ./share/doc/usd/Makefile Wed Jul 30 01:03:20 1997 >... >! .if exists(../../../games) && !defined(NOGAMES) >! SUBDIR+= 30.rouge 31.trek >! .endif The path should be relative to ${.CURDIR} >*** ./share/zoneinfo/Makefile.orig Thu May 22 11:34:30 1997 >--- ./share/zoneinfo/Makefile Wed Jul 30 01:03:24 1997 >*************** >*** 16,22 **** > TZFILES+= backward > .endif > >! .if exists(${.OBJDIR}/yearistype) > YEARISTYPE= ${.OBJDIR}/yearistype > .else > YEARISTYPE= ${.CURDIR}/yearistype >--- 16,22 ---- > TZFILES+= backward > .endif > >! .if exists(${.OBJDIR}) > YEARISTYPE= ${.OBJDIR}/yearistype > .else > YEARISTYPE= ${.CURDIR}/yearistype > >Something tells me this hasn't been used for a while ;): The old version worked, except when yearistype hasn't already been created and make is invoked as with args `all install' to create and install in the same invocation. There are more bogons here. The whole ifdef is a verbose and confusing way write `YEARISTYPE= ${.OBJDIR}/yearistype. yearistype is always created in the object directory and ${.OBJDIR} always exists (although it's not clear where it is when make is invoked with args `obj all ...'). >--- ./bin/Makefile.inc Wed Jul 30 01:51:06 1997 >... >--- ./bin/ed/Makefile Wed Jul 30 01:11:58 1997 > [LIBCYPHER changes] >Why doesn't libcipher have an entry in bsd.libnames.mk? and shouldn't there Because I don't use it :-). bsd.libnames.mk is no longer used in -current, so this problem will go away automatically. >be some ${DESTDIR}'s near those ${BINDIR}'s (Also in ./bin/test/Makefile)? >That's what ./share/mk/bsd.README says... There's lots of those around the >tree. Not sure. See similar fixes for moving defines for init from sbin/Makefile.inc to sbin/init/Makefile. >Oops, one last thing: should ./usr.bin/lex/lib be built twice in the >bootstrapping of 'make world' (as part of lib-tools and libraries) and It doesn't hurt. >should the whole of ./lib be rebuilt, including stuff that's been done a >minute earlier? only libc and libm are needed, as far as I can see, to make >the build-tools targets before the libraries are built anyway as part of the >'make all'. The idea is that the library tools, especially the compiler, might have changed, so everything needs to rebuilt. Too many things are probably built before various toolsets are built, but this isn't easy to fix without modifying some of the tools to have fewer prerequisites. Satoshi has a version of src/Makefile that is more careful about building and using all prerequisites. It builds even more unnecessary things. Bruce From owner-freebsd-current Thu Jul 31 04:02:52 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA28293 for current-outgoing; Thu, 31 Jul 1997 04:02:52 -0700 (PDT) Received: from critter.dk.tfs.com (critter.phk.freebsd.dk [195.8.129.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA28281 for ; Thu, 31 Jul 1997 04:02:35 -0700 (PDT) Received: from critter.dk.tfs.com (localhost [127.0.0.1]) by critter.dk.tfs.com (8.8.6/8.8.5) with ESMTP id NAA01274; Thu, 31 Jul 1997 13:01:29 +0200 (CEST) To: "Jordan K. Hubbard" cc: Chuck Robey , FreeBSD current From: Poul-Henning Kamp Subject: Re: core group topics In-reply-to: Your message of "Thu, 31 Jul 1997 00:34:53 PDT." <20284.870334493@time.cdrom.com> Date: Thu, 31 Jul 1997 13:01:28 +0200 Message-ID: <1272.870346888@critter.dk.tfs.com> Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In message <20284.870334493@time.cdrom.com>, "Jordan K. Hubbard" writes: >> I agree. How long have we had to discuss it so far? I think you'll have >> to agree, it's been more than long enough. It feels more like a case of >> "if we ignore it long enough, maybe it'll go away." There are ways to > >Well, that's one way to look at it. Another way to look at it is that >a significant percentage of people looking at the whole idea have so >far gone "No, I'm outright OPPOSED to this idea. I do not want ELF." >It's not as passive a situation as you might think. Are we getting the kernel & userland mixed up here ? I belive there is strong agreement for ELF for the kernel... -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. Power and ignorance is a disgusting cocktail. From owner-freebsd-current Thu Jul 31 04:34:55 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA00676 for current-outgoing; Thu, 31 Jul 1997 04:34:55 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA00630; Thu, 31 Jul 1997 04:34:42 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.6/8.6.9) id EAA16307; Thu, 31 Jul 1997 04:34:38 -0700 (PDT) Date: Thu, 31 Jul 1997 04:34:38 -0700 (PDT) Message-Id: <199707311134.EAA16307@silvia.HIP.Berkeley.EDU> To: committers@freebsd.org CC: current@freebsd.org, stable@freebsd.org Reply-to: committers@freebsd.org Subject: A whole new world From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Here is a patch for 2.2-stable. As I said elsewhere, I don't have a 3.0-machine anymore, so that has to wait until thud comes back online. I believe -current people will have a reasonably good success by checking out a 2.2 /usr/src/Makefile, applying the patch, and just using that Makefile (IIRC, the previous patch made -current and -stable src/Makefile's look almost the same) in conjunction with the following patches to src/share/mk: %%%%%%%%%%%%%%%%%%% Index: share/mk/bsd.lib.mk =================================================================== RCS file: /usr/cvs/src/share/mk/bsd.lib.mk,v retrieving revision 1.59 diff -u -r1.59 bsd.lib.mk --- bsd.lib.mk 1997/06/21 15:40:32 1.59 +++ bsd.lib.mk 1997/07/01 21:15:03 @@ -186,10 +186,12 @@ > $$TMP; \ mv $$TMP ${DEPENDFILE} .endif +.if !defined(NOEXTRADEPEND) _EXTRADEPEND:: echo lib${LIB}.so.${SHLIB_MAJOR}.${SHLIB_MINOR}: \ `${LDDESTDIRENV} ${CC} -shared -Wl,-f ${LDDESTDIR} ${LDADD}` \ >> ${DEPENDFILE} +.endif .if !target(install) .if !target(beforeinstall) Index: share/mk/bsd.prog.mk =================================================================== RCS file: /usr/cvs/src/share/mk/bsd.prog.mk,v retrieving revision 1.52 diff -u -r1.52 bsd.prog.mk --- bsd.prog.mk 1997/06/28 08:14:10 1.52 +++ bsd.prog.mk 1997/07/01 21:15:03 @@ -68,7 +68,7 @@ .endif .endif -.if defined(PROG) +.if defined(PROG) && !defined(NOEXTRADEPEND) _EXTRADEPEND: echo ${PROG}: `${CC} -Wl,-f ${CFLAGS} ${LDFLAGS} ${LDDESTDIR} \ ${LDADD:S/^/-Wl,/}` >> ${DEPENDFILE} %%%%%%%%%%%%%%%%%%% This isn't really necessary unless you are building -current on a very old system. If your source tree is -stable, you don't need the above patch (it won't apply anyway ;). Now, here is a patch to Makefile. %%%%%%%%%%%%%%%%%%% Index: Makefile =================================================================== RCS file: /usr/cvs/src/Makefile,v retrieving revision 1.109.2.8 diff -u -r1.109.2.8 Makefile --- Makefile 1997/07/31 08:32:42 1.109.2.8 +++ Makefile 1997/07/31 08:41:42 @@ -31,6 +31,9 @@ # /usr/share/mk. These include: # obj depend all install clean cleandepend cleanobj +.if (!make(world)) && (!make(buildworld)) && (!make(installworld)) +.MAKEFLAGS:= ${.MAKEFLAGS} -m ${.CURDIR}/share/mk +.endif # Put initial settings here. SUBDIR= @@ -113,8 +116,8 @@ .endif .endif -SUP?= sup -SUPFLAGS?= -v +SUP?= sup +SUPFLAGS?= -v # # While building tools for bootstrapping, we dont need to waste time on @@ -135,98 +138,151 @@ @echo "make world started on `LC_TIME=C date`" @echo "--------------------------------------------------------------" .if target(pre-world) + @echo @echo "--------------------------------------------------------------" @echo " Making 'pre-world' target" @echo "--------------------------------------------------------------" cd ${.CURDIR} && ${MAKE} pre-world - @echo .endif - @echo "--------------------------------------------------------------" - @echo " Making hierarchy" - @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} hierarchy + cd ${.CURDIR} && ${MAKE} buildworld + cd ${.CURDIR} && ${MAKE} installworld +.if target(post-world) @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding /usr/share/mk" + @echo " Making 'post-world' target" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} mk + cd ${.CURDIR} && ${MAKE} post-world +.endif @echo @echo "--------------------------------------------------------------" - @echo " Cleaning up the source tree" + @echo "make world completed on `LC_TIME=C date`" @echo "--------------------------------------------------------------" -.if defined(NOCLEAN) - @echo "Not cleaning anything! I sure hope you know what you are doing!" + +.if defined(MAKEOBJDIRPREFIX) +WORLDTMP= ${MAKEOBJDIRPREFIX}${.CURDIR}/tmp .else - cd ${.CURDIR} && ${MAKE} ${CLEANDIR} +WORLDTMP= /usr/obj${.CURDIR}/tmp .endif +STRICTTMPPATH= ${WORLDTMP}/sbin:${WORLDTMP}/usr/sbin:${WORLDTMP}/bin:${WORLDTMP}/usr/bin +TMPPATH= ${STRICTTMPPATH}:${PATH} + +# XXX COMPILER_PATH is needed for finding cc1, ld and as +# XXX GCC_EXEC_PREFIX is for *crt.o. It is probably unnecssary now +# tbat LIBRARY_PATH is set. We still can't use -nostdlib, since gcc +# wouldn't link *crt.o or libgcc if it were used. +# XXX LD_LIBRARY_PATH is for ld.so. It is also used by ld, although we don't +# want that - all compile-time library paths should be resolved by gcc. +# It fails for set[ug]id executables (are any used?). +COMPILER_ENV= BISON_SIMPLE=${WORLDTMP}/usr/share/misc/bison.simple \ + COMPILER_PATH=${WORLDTMP}/usr/libexec:${WORLDTMP}/usr/bin \ + GCC_EXEC_PREFIX=${WORLDTMP}/usr/lib/ \ + LD_LIBRARY_PATH=${WORLDTMP}${SHLIBDIR} \ + LIBRARY_PATH=${WORLDTMP}${SHLIBDIR}:${WORLDTMP}/usr/lib + +BMAKEENV= PATH=${TMPPATH} ${COMPILER_ENV} NOEXTRADEPEND=t +XMAKEENV= PATH=${STRICTTMPPATH} ${COMPILER_ENV} \ + CC='cc -nostdinc' # XXX -nostdlib + +# used to compile and install 'make' in temporary build tree +IBMAKE= ${BMAKEENV} ${MAKE} DESTDIR=${WORLDTMP} +# bootstrap make +BMAKE= ${BMAKEENV} ${WORLDTMP}/usr/bin/${MAKE} DESTDIR=${WORLDTMP} +# cross make used for compilation +XMAKE= ${XMAKEENV} ${WORLDTMP}/usr/bin/${MAKE} DESTDIR=${WORLDTMP} +# cross make used for final installation +IXMAKE= ${XMAKEENV} ${WORLDTMP}/usr/bin/${MAKE} + +# +# buildworld +# +# Attempt to rebuild the entire system, with reasonable chance of +# success, regardless of how old your existing system is. +# +buildworld: +.if !defined(NOCLEAN) @echo -.if !defined(NOOBJDIR) @echo "--------------------------------------------------------------" - @echo " Rebuilding the obj tree" + @echo " Cleaning up the temporary build tree" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} obj - @echo + mkdir -p ${WORLDTMP} + chflags -R noschg ${WORLDTMP}/ + rm -rf ${WORLDTMP} .endif + @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding bootstrap tools" + @echo " Making make" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} bootstrap + mkdir -p ${WORLDTMP}/usr/bin + cd ${.CURDIR}/usr.bin/make && \ + ${IBMAKE} -I${.CURDIR}/share/mk ${OBJDIR} clean cleandepend depend && \ + ${IBMAKE} -I${.CURDIR}/share/mk ${MK_FLAGS} all install clean cleandepend @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding tools necessary to build the include files" + @echo " Making hierarchy" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} include-tools + cd ${.CURDIR} && ${BMAKE} hierarchy +.if !defined(NOCLEAN) @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding /usr/include" + @echo " Cleaning up the obj tree" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} includes + cd ${.CURDIR} && ${BMAKE} ${CLEANDIR} +.endif +.if !defined(NOOBJDIR) @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding tools needed to build the libraries" + @echo " Rebuilding the obj tree" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} lib-tools + cd ${.CURDIR} && ${BMAKE} obj +.endif @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding /usr/lib" + @echo " Rebuilding bootstrap tools" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} libraries + cd ${.CURDIR} && ${BMAKE} bootstrap @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding symorder, groff and zic(8)" + @echo " Rebuilding tools necessary to build the include files" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} build-tools + cd ${.CURDIR} && ${BMAKE} include-tools @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding dependencies" + @echo " Rebuilding /usr/include" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} depend + cd ${.CURDIR} && ${BMAKE} includes @echo @echo "--------------------------------------------------------------" - @echo " Building everything.." + @echo " Rebuilding tools needed to build the libraries" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} all + cd ${.CURDIR} && ${BMAKE} lib-tools @echo @echo "--------------------------------------------------------------" - @echo " Installing everything.." + @echo " Rebuilding /usr/lib" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} install + cd ${.CURDIR} && ${BMAKE} libraries @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding man page indexes" + @echo " Rebuilding all other tools needed to build the world" @echo "--------------------------------------------------------------" - cd ${.CURDIR}/share/man && ${MAKE} makedb -.if target(post-world) + cd ${.CURDIR} && ${BMAKE} build-tools @echo @echo "--------------------------------------------------------------" - @echo " Making 'post-world' target" + @echo " Rebuilding dependencies" @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} post-world -.endif + cd ${.CURDIR} && ${XMAKE} depend @echo @echo "--------------------------------------------------------------" - @echo "make world completed on `LC_TIME=C date`" + @echo " Building everything.." @echo "--------------------------------------------------------------" + cd ${.CURDIR} && ${XMAKE} all + +# +# installworld +# +# Installs everything compiled by a 'buildworld'. +# +installworld: + cd ${.CURDIR} && ${IXMAKE} reinstall # # reinstall @@ -242,11 +298,6 @@ cd ${.CURDIR} && ${MAKE} hierarchy @echo @echo "--------------------------------------------------------------" - @echo " Rebuilding /usr/share/mk" - @echo "--------------------------------------------------------------" - cd ${.CURDIR} && ${MAKE} mk - @echo - @echo "--------------------------------------------------------------" @echo " Installing everything.." @echo "--------------------------------------------------------------" cd ${.CURDIR} && ${MAKE} install @@ -256,7 +307,6 @@ @echo "--------------------------------------------------------------" cd ${.CURDIR}/share/man && ${MAKE} makedb - # # update # @@ -283,7 +333,6 @@ cd ${.CURDIR} && cvs -q update -P -d .endif - # # most # @@ -351,17 +400,17 @@ cd ${.CURDIR}/etc && ${MAKE} distrib-dirs # -# mk - update the /usr/share/mk makefiles. -# -mk: - cd ${.CURDIR}/share/mk && ${MAKE} install - -# # bootstrap - [re]build tools needed to run the actual build, this includes # tools needed by 'make depend', as some tools are needed to generate source # for the dependency information to be gathered from. # bootstrap: +.if defined(DESTDIR) + rm -f ${DESTDIR}/usr/src/sys + ln -s ${.CURDIR}/sys ${DESTDIR}/usr/src + cd ${.CURDIR}/include && find -dx . | cpio -dump ${DESTDIR}/usr/include + cd ${.CURDIR}/include && make symlinks +.endif cd ${.CURDIR}/usr.bin/make && ${MAKE} depend && \ ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} cd ${.CURDIR}/usr.bin/xinstall && ${MAKE} depend && \ @@ -373,11 +422,13 @@ # include-tools - generally the same as 'bootstrap', except that it's for # things that are specifically needed to generate include files. # -# XXX should be merged with bootstrap, it's not worth keeeping them seperate +# XXX should be merged with bootstrap, it's not worth keeeping them separate. +# Well, maybe it is now. We force 'cleandepend' here to avoid dependencies +# on cleaned away headers in ${WORLDTMP}. # include-tools: - cd ${.CURDIR}/usr.bin/rpcgen && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} depend all install ${CLEANDIR} ${OBJDIR} + cd ${.CURDIR}/usr.bin/rpcgen && ${MAKE} cleandepend depend && \ + ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} # # includes - possibly generate and install the include files. @@ -388,8 +439,7 @@ mtree -deU -f ${.CURDIR}/etc/mtree/BSD.include.dist \ -p ${DESTDIR}/usr/include .endif - cd ${.CURDIR}/include/rpcsvc && ${MAKE} all - cd ${.CURDIR}/include && ${MAKE} install + cd ${.CURDIR}/include && ${MAKE} all install cd ${.CURDIR}/gnu/include && ${MAKE} install cd ${.CURDIR}/gnu/lib/libreadline && ${MAKE} beforeinstall cd ${.CURDIR}/gnu/lib/libregex && ${MAKE} beforeinstall @@ -422,36 +472,31 @@ cd ${.CURDIR}/lib/libss && ${MAKE} beforeinstall cd ${.CURDIR}/lib/libscsi && ${MAKE} beforeinstall cd ${.CURDIR}/lib/libutil && ${MAKE} beforeinstall + cd ${.CURDIR}/lib/libz && ${MAKE} beforeinstall # # lib-tools - build tools to compile and install the libraries. # +# XXX gperf is required for cc +# XXX a new ld and tsort is required for cc lib-tools: - cd ${.CURDIR}/usr.bin/tsort && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/gnu/usr.bin/ld && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/gnu/usr.bin/as && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/usr.bin/ar && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/usr.bin/ranlib && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/usr.bin/nm && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/usr.bin/lex/lib && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/usr.bin/compile_et && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} && \ - rm -f /usr/sbin/compile_et - cd ${.CURDIR}/usr.bin/mk_cmds && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/gnu/usr.bin/bison && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/gnu/usr.bin/gperf && ${MAKE} depend && \ - ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} - cd ${.CURDIR}/gnu/usr.bin/cc && ${MAKE} depend && \ +.for d in \ + gnu/usr.bin/gperf \ + gnu/usr.bin/ld \ + usr.bin/tsort \ + gnu/usr.bin/as \ + gnu/usr.bin/bison \ + gnu/usr.bin/cc \ + usr.bin/ar \ + usr.bin/compile_et \ + usr.bin/lex/lib \ + usr.bin/mk_cmds \ + usr.bin/nm \ + usr.bin/ranlib \ + usr.bin/uudecode + cd ${.CURDIR}/$d && ${MAKE} depend && \ ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} +.endfor # # libraries - build and install the libraries @@ -501,15 +546,70 @@ # # build-tools - build and install any other tools needed to complete the # compile and install. +# ifdef stale +# bc and cpp are required to build groff. Otherwise, the order here is +# mostly historical, i.e., bogus. +# chmod is used to build gcc's tmpmultilib[2] at obscure times. +# endif stale +# XXX uname is a bug - the target should not depend on the host. # build-tools: .for d in \ + bin/cat \ + bin/chmod \ + bin/cp \ + bin/date \ + bin/dd \ + bin/echo \ + bin/expr \ + bin/hostname \ + bin/ln \ + bin/ls \ + bin/mkdir \ + bin/mv \ + bin/rm \ + bin/sh \ + bin/test \ + gnu/usr.bin/awk \ + gnu/usr.bin/bc \ + gnu/usr.bin/grep \ + gnu/usr.bin/groff \ + gnu/usr.bin/gzip \ + gnu/usr.bin/man/makewhatis \ + gnu/usr.bin/sort \ + gnu/usr.bin/texinfo \ share/info \ - gnu/usr.bin/texinfo \ + usr.bin/basename \ + usr.bin/cap_mkdb \ + usr.bin/chflags \ + usr.bin/cmp \ + usr.bin/col \ + usr.bin/cpp \ + usr.bin/expand \ + usr.bin/file2c \ + usr.bin/find \ + usr.bin/gencat \ + usr.bin/lorder \ + usr.bin/m4 \ + usr.bin/mkdep \ + usr.bin/paste \ + usr.bin/sed \ + usr.bin/size \ + usr.bin/soelim \ + usr.bin/strip \ usr.bin/symorder \ - usr.sbin/zic \ - gnu/usr.bin/awk \ - gnu/usr.bin/groff + usr.bin/touch \ + usr.bin/tr \ + usr.bin/true \ + usr.bin/uname \ + usr.bin/uuencode \ + usr.bin/vgrind \ + usr.bin/vi \ + usr.bin/wc \ + usr.bin/yacc \ + usr.sbin/chown \ + usr.sbin/mtree \ + usr.sbin/zic cd ${.CURDIR}/$d && ${MAKE} depend && \ ${MAKE} ${MK_FLAGS} all install ${CLEANDIR} ${OBJDIR} .endfor %%%%%%%%%%%%%%%%%%% Basically what it does is to split "world" into two parts, "buildworld" and "installworld", which can be run together or separately (or even on different machines). The "buildworld" target does not install anything outside /usr/obj; you still need to be root to run it (the default install rules want to set the ownership, etc.), but it's now possible to upgrade a bunch of -stable machines by running "make buildworld" on a 2.1.5 NFS server and then running "make installworld" on each of the clients, while not compromising the stability of the server at all. (Trust me, I've done exactly that last week. :) One thing this patch does not address (i.e., I tried, but gave up) is that your source tree still has to be called "/usr/src" -- there are far too many brokenness in the tree for me to fix by myself (and those can be hunted down/fixed independently of this project anyway). If that can be fixed, the above NFS server in the example above can even build binaries for -current and -stable machines simultaneously. If someone can take a stab at this, it is great. I would like people, especially those with very old systems, to try to upgrade your machine by using this Makefile. Everything (except for upgrading /etc) should be done with a single "make world", followed by a kernel rebuild and reboot; if it doesn't, then there is something wrong, and I want to know about it. Thanks! Satoshi From owner-freebsd-current Thu Jul 31 04:41:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA01328 for current-outgoing; Thu, 31 Jul 1997 04:41:42 -0700 (PDT) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA01323 for ; Thu, 31 Jul 1997 04:41:40 -0700 (PDT) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id EAA28471; Thu, 31 Jul 1997 04:42:32 -0700 (PDT) Message-Id: <199707311142.EAA28471@implode.root.com> To: Poul-Henning Kamp cc: "Jordan K. Hubbard" , Chuck Robey , FreeBSD current Subject: Re: core group topics In-reply-to: Your message of "Thu, 31 Jul 1997 13:01:28 +0200." <1272.870346888@critter.dk.tfs.com> From: David Greenman Reply-To: dg@root.com Date: Thu, 31 Jul 1997 04:42:32 -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >In message <20284.870334493@time.cdrom.com>, "Jordan K. Hubbard" writes: >>> I agree. How long have we had to discuss it so far? I think you'll have >>> to agree, it's been more than long enough. It feels more like a case of >>> "if we ignore it long enough, maybe it'll go away." There are ways to >> >>Well, that's one way to look at it. Another way to look at it is that >>a significant percentage of people looking at the whole idea have so >>far gone "No, I'm outright OPPOSED to this idea. I do not want ELF." >>It's not as passive a situation as you might think. > >Are we getting the kernel & userland mixed up here ? I belive there >is strong agreement for ELF for the kernel... There is? I don't recall discussing that at all, and if we do eventually go the ELF way, I think the kernel would be one of the last areas we'd want to convert since it further complicates our cramped bootblocks. The only issue I have against ELF is that I'm concerned that the overhead for processing the much more sophisticated header at exec time might have a serious impact on exec performance (something I'm particularly sensitive to since I wrote the a.out exec code for FreeBSD). -DG David Greenman Core-team/Principal Architect, The FreeBSD Project From owner-freebsd-current Thu Jul 31 05:11:38 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA02984 for current-outgoing; Thu, 31 Jul 1997 05:11:38 -0700 (PDT) Received: from critter.dk.tfs.com (critter.phk.freebsd.dk [195.8.129.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA02978 for ; Thu, 31 Jul 1997 05:11:31 -0700 (PDT) Received: from critter.dk.tfs.com (localhost [127.0.0.1]) by critter.dk.tfs.com (8.8.6/8.8.5) with ESMTP id OAA00246; Thu, 31 Jul 1997 14:10:27 +0200 (CEST) To: dg@root.com cc: Poul-Henning Kamp , "Jordan K. Hubbard" , Chuck Robey , FreeBSD current From: Poul-Henning Kamp Subject: Re: core group topics In-reply-to: Your message of "Thu, 31 Jul 1997 04:42:32 PDT." <199707311142.EAA28471@implode.root.com> Date: Thu, 31 Jul 1997 14:10:27 +0200 Message-ID: <244.870351027@critter.dk.tfs.com> Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In message <199707311142.EAA28471@implode.root.com>, David Greenman writes: >>Are we getting the kernel & userland mixed up here ? I belive there >>is strong agreement for ELF for the kernel... > > There is? I don't recall discussing that at all, and if we do eventually go >the ELF way, I think the kernel would be one of the last areas we'd want to >convert since it further complicates our cramped bootblocks. Have you looked at multiboot/GRUB from Eric Boleyn (sp?) yet ? > The only issue I have against ELF is that I'm concerned that the overhead >for processing the much more sophisticated header at exec time might have a >serious impact on exec performance (something I'm particularly sensitive to >since I wrote the a.out exec code for FreeBSD). Well, I belive a.out will still be an option or how ? -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. Power and ignorance is a disgusting cocktail. From owner-freebsd-current Thu Jul 31 05:17:06 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA03245 for current-outgoing; Thu, 31 Jul 1997 05:17:06 -0700 (PDT) Received: from implode.root.com (implode.root.com [198.145.90.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA03240 for ; Thu, 31 Jul 1997 05:17:03 -0700 (PDT) Received: from implode.root.com (localhost [127.0.0.1]) by implode.root.com (8.8.5/8.8.5) with ESMTP id FAA28904; Thu, 31 Jul 1997 05:17:56 -0700 (PDT) Message-Id: <199707311217.FAA28904@implode.root.com> To: Poul-Henning Kamp cc: "Jordan K. Hubbard" , Chuck Robey , FreeBSD current Subject: Re: core group topics In-reply-to: Your message of "Thu, 31 Jul 1997 14:10:27 +0200." <244.870351027@critter.dk.tfs.com> From: David Greenman Reply-To: dg@root.com Date: Thu, 31 Jul 1997 05:17:56 -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >In message <199707311142.EAA28471@implode.root.com>, David Greenman writes: > >>>Are we getting the kernel & userland mixed up here ? I belive there >>>is strong agreement for ELF for the kernel... >> >> There is? I don't recall discussing that at all, and if we do eventually go >>the ELF way, I think the kernel would be one of the last areas we'd want to >>convert since it further complicates our cramped bootblocks. > >Have you looked at multiboot/GRUB from Eric Boleyn (sp?) yet ? No, but I am somewhat familiar with it. If we want to use it without GPL restrictions, we have to remove the ELF support (since it is GPL'd) - a.out support isn't encumbered (or at least Erich was willing to change the copyright terms so that it is not). -DG David Greenman Core-team/Principal Architect, The FreeBSD Project From owner-freebsd-current Thu Jul 31 05:29:57 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA04266 for current-outgoing; Thu, 31 Jul 1997 05:29:57 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA04253 for ; Thu, 31 Jul 1997 05:29:54 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id VAA28215; Thu, 31 Jul 1997 21:59:31 +0930 (CST) From: Michael Smith Message-Id: <199707311229.VAA28215@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: <199707311142.EAA28471@implode.root.com> from David Greenman at "Jul 31, 97 04:42:32 am" To: dg@root.com Date: Thu, 31 Jul 1997 21:59:31 +0930 (CST) Cc: phk@dk.tfs.com, jkh@time.cdrom.com, chuckr@Glue.umd.edu, freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk David Greenman stands accused of saying: > There is? I don't recall discussing that at all, and if we do > eventually go the ELF way, I think the kernel would be one of the > last areas we'd want to convert since it further complicates our > cramped bootblocks. I've already volunteered to do much of the grunt work dealing with that if someone will help me over my x86 cluelessness in getting a third-stage bootstrap running. > The only issue I have against ELF is that I'm concerned that the overhead > for processing the much more sophisticated header at exec time might have a > serious impact on exec performance (something I'm particularly sensitive to > since I wrote the a.out exec code for FreeBSD). You raised this one last time, and forgive me if I asked the same question then too - could the exec parser be common-cased to deal with "typical" executables? For that matter, have you done any actual comparisons using the ELF parser currently in place? -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Thu Jul 31 05:36:50 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id FAA04868 for current-outgoing; Thu, 31 Jul 1997 05:36:50 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id FAA04863 for ; Thu, 31 Jul 1997 05:36:46 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id WAA28265; Thu, 31 Jul 1997 22:06:25 +0930 (CST) From: Michael Smith Message-Id: <199707311236.WAA28265@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: <244.870351027@critter.dk.tfs.com> from Poul-Henning Kamp at "Jul 31, 97 02:10:27 pm" To: phk@dk.tfs.com (Poul-Henning Kamp) Date: Thu, 31 Jul 1997 22:06:25 +0930 (CST) Cc: dg@root.com, phk@dk.tfs.com, jkh@time.cdrom.com, chuckr@Glue.umd.edu, freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Poul-Henning Kamp stands accused of saying: > >the ELF way, I think the kernel would be one of the last areas we'd want to > >convert since it further complicates our cramped bootblocks. > > Have you looked at multiboot/GRUB from Eric Boleyn (sp?) yet ? GRUB is very interesting in some ways, but a) I couldn't actually build it (Linux-centric build process?), and b) is very much the hackers power-toy. I could see it being quite a daunting experience for a "new" user. (It also uses hardcoded block numbers for the 1.5 stage UFS bootstrap, something I am allergic to as a rule.) -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Thu Jul 31 08:02:29 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id IAA13835 for current-outgoing; Thu, 31 Jul 1997 08:02:29 -0700 (PDT) Received: from ox.ismi.net (root@ox.ismi.net [206.31.56.6]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id IAA13829 for ; Thu, 31 Jul 1997 08:02:24 -0700 (PDT) Received: from aerosmith.dyn.ml.org (mrr@pm5-23.ismi.net [207.51.208.194]) by ox.ismi.net (8.6.9/8.6.9) with SMTP id LAA05453; Thu, 31 Jul 1997 11:00:31 -0400 Date: Thu, 31 Jul 1997 10:54:17 -0400 (EDT) From: "Michael R. Rudel" To: Tim Liddelow cc: "Jordan K. Hubbard" , Chuck Robey , FreeBSD current Subject: Re: core group topics In-Reply-To: <33E04610.81C2C2EC@cybec.com.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Thu, 31 Jul 1997, Tim Liddelow wrote: > Jordan K. Hubbard wrote: > > > > Erm, "none excepting inertia and fear of chaos, confusion, shock and > > increased entropy all around." :) Inflicting another ABI on our user > > base, one which will result in a new class of packages which all older > > FreeBSD releases cannot run, is NOT something to be considered > > lightly. > > > > Jordan > > It doesn't have to be _that_ hard. The Linux camp did the dual ABI > thing for quite a while. Sure - the transition wasn't all beer and > skittles but they did it. And they're reaping the rewards (see > previous postings) for it. > > Another thing - so what if OLDER FreeBSD releases can't run the ELF > packages ? You have symlinks for packages for a certain release, > right ? There's no way if I am running FreeBSD 4.0 (ELF) that I would > try to add an a.out package. And vise versa. OLDER FreeBSD releases > will continue to use the a.out packages. New packages would have > to be made for both 2.1.7.x release(s) and the newer ELF release. Well, new FreeBSD realeses will be able to run both ELF and a.out, correct? So, really, as long as people running an old release don't try to grab packages for a new release, this should work fine. We could really stress the fact that, when the ELF release comes up, people -NEED- to upgrade, as this will (and, IMO, 3.0 always has been) a major milestone. > > Noone is suggesting there aren't going to be hurdles. What has > to be considered is the cost in remaining a.out, and the advantages > of ELF in the longer term. > > We rely on third party (GNU) software that supports ELF "out of the > box". If we continue down the a.out path for too long, we risk > obscurity in terms of language support, etc. When gcc 2.8 comes > out (fingers crossed) will it support a.out correctly ? C++ > support on FreeBSD is something that really interests/concerns me. > FreeBSD plugs itself as a great networking/internet server. What > about a development box ? ELF is great for development - e.g. Java, > C++, etc. > > Just my $0.02. > > Tim. > -- Michael R. Rudel -=- FreeBSD: There are no limits -=- mrr@aerosmith.dyn.ml.org FreeBSD aerosmith.dyn.ml.org 3.0-CURRENT PGP Key Block: finger mrrpgp@aerosmith.dyn.ml.org When you are born your afraid of the darkness ... Then your afraid of the light ... I'm not afraid when I dance with my shadows ... This time I'm gonna get it right ... -- Aerosmith: Taste of India From owner-freebsd-current Thu Jul 31 08:24:24 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id IAA15039 for current-outgoing; Thu, 31 Jul 1997 08:24:24 -0700 (PDT) Received: from friley01.res.iastate.edu (friley01.res.iastate.edu [129.186.189.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA15025 for ; Thu, 31 Jul 1997 08:24:03 -0700 (PDT) Received: from friley01.res.iastate.edu (loopback [127.0.0.1]) by friley01.res.iastate.edu (8.8.6/8.8.5) with ESMTP id KAA14268 for ; Thu, 31 Jul 1997 10:27:49 -0500 (CDT) Message-Id: <199707311527.KAA14268@friley01.res.iastate.edu> To: freebsd-current@freebsd.org Subject: Re: core group topics In-reply-to: Your message of Wed, 30 Jul 1997 18:49:25 -0400. Date: Thu, 31 Jul 1997 10:27:49 -0500 From: Chris Csanady Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >2) John Polstra's ELF work. He's done all the work that anyone could >possibly want, to move FreeBSD to ELF. There are several good reasons to >move, and none excepting inertia not to. I know that moving FreeBSD to >ELF isn't a high priority project, but all the work's been done, the >benefits are real, and it's high time to consider it. As an aside.. from the past discussions about Doug Rabson's work on a new module system--It sounded as if things would be considerably simpler if we were to use an ELF kernel from the start. By the way, can anyone comment how things are going with respect to this? Chris From owner-freebsd-current Thu Jul 31 08:36:30 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id IAA15927 for current-outgoing; Thu, 31 Jul 1997 08:36:30 -0700 (PDT) Received: from pandora.hh.kew.com (ahd@kendra.ne.mediaone.net [24.128.53.73]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id IAA15918 for ; Thu, 31 Jul 1997 08:36:24 -0700 (PDT) Received: (from ahd@localhost) by pandora.hh.kew.com (8.8.5/8.8.5) id LAA00915 for freebsd-current@FreeBSD.ORG; Thu, 31 Jul 1997 11:36:22 -0400 (EDT) Date: Thu, 31 Jul 1997 11:36:22 -0400 (EDT) From: Drew Derbyshire Message-Id: <199707311536.LAA00915@pandora.hh.kew.com> To: freebsd-current@FreeBSD.ORG Subject: Tuning ISA IDE drives Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Okay, I read LINT for tuning the IDE controller ... what happens if I just crank the puppy for max sectors and check for 32 bit I/O? Under what conditions DOESN'T this work properly? (System is an ISA/VLB 486/66, an IBM PS/1.) -- Drew Derbyshire Internet: ahd@kew.com Kendra Electronic Wonderworks Telephone: 617-279-9812 "Who are these men of lust, greed, and glory?" - Supertramp From owner-freebsd-current Thu Jul 31 10:12:53 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA21262 for current-outgoing; Thu, 31 Jul 1997 10:12:53 -0700 (PDT) Received: from rover.village.org (rover.village.org [204.144.255.49]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id KAA21253 for ; Thu, 31 Jul 1997 10:12:48 -0700 (PDT) Received: from rover.village.org [127.0.0.1] by rover.village.org with esmtp (Exim 1.60 #1) id 0wtymC-0002jm-00; Thu, 31 Jul 1997 11:12:36 -0600 To: Chuck Robey Subject: Re: core group topics Cc: FreeBSD current In-reply-to: Your message of "Wed, 30 Jul 1997 18:49:25 EDT." References: Date: Thu, 31 Jul 1997 11:12:36 -0600 From: Warner Losh Message-Id: Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In message Chuck Robey writes: : 1) Satoshi's new Makefile. His makefile has been around plenty long : enough for testing, and while several people have reported good results, : I've seen no negatives reported at all. Will it get ignored or committed? Heck, you have commit privs, so you could do it yourself :-). That's what I've been doing with the OpenBSD security fixes, just doing it when I have time, and backing things out that escape my testing procedures. Granted this is a bigger deal thing than my stuff. : ELF If we do this, we should do it in 3.0. That's the perfect time because all the libraries are changing major numbers (or most) and people would expect much disruption. This would be a major diff between 2.2.x and 3.x. Heck, at least integrate it into the tree and make it possible to build the new stuff with a DOELF=yes in /etc/make.conf. That is the next logical step, no? Warner From owner-freebsd-current Thu Jul 31 10:19:37 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA21657 for current-outgoing; Thu, 31 Jul 1997 10:19:37 -0700 (PDT) Received: from rover.village.org (rover.village.org [204.144.255.49]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id KAA21648 for ; Thu, 31 Jul 1997 10:19:31 -0700 (PDT) Received: from rover.village.org [127.0.0.1] by rover.village.org with esmtp (Exim 1.60 #1) id 0wtysb-0002kb-00; Thu, 31 Jul 1997 11:19:13 -0600 To: Michael Smith Subject: Re: core group topics Cc: chuckr@glue.umd.edu (Chuck Robey), jkh@time.cdrom.com, freebsd-current@freefall.freebsd.org In-reply-to: Your message of "Thu, 31 Jul 1997 17:12:10 +0930." <199707310742.RAA27102@genesis.atrad.adelaide.edu.au> References: <199707310742.RAA27102@genesis.atrad.adelaide.edu.au> Date: Thu, 31 Jul 1997 11:19:13 -0600 From: Warner Losh Message-Id: Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In message <199707310742.RAA27102@genesis.atrad.adelaide.edu.au> Michael Smith writes: : > compatibly. How about we move the discussion from "ignore" stage to an : > enumeration of possible transition problems, so that last problems can be : > solved? I don't honestly think they're all that bad, myself. : : Er, we've been doing this on -hackers the last couple of days. 8) Yes. I'd like to reiterate that it might be time to integrate the ELF kit into the tree, but require (for now) a DOELF=yes in /etc/make.conf. This would allow for even easier experimentation and might make it easier to identify the transition issues involved. ELF kit could also be back ported to 2.2. It might be sensible to have DOELF=no in 2.2 and yes in 3.0. What are the issues with old systems anyway? I thought we had ELF support in the kernel already... Would we need a compatelf package for 2.2 much like we have compat10, et al? Warner From owner-freebsd-current Thu Jul 31 10:23:41 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA21913 for current-outgoing; Thu, 31 Jul 1997 10:23:41 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [206.246.122.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id KAA21908 for ; Thu, 31 Jul 1997 10:23:36 -0700 (PDT) Received: from Journey2.mat.net (journey2.mat.net [206.246.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id NAA13648; Thu, 31 Jul 1997 13:23:25 -0400 Date: Thu, 31 Jul 1997 13:23:22 -0400 (EDT) From: Chuck Robey X-Sender: chuckr@Journey2.mat.net To: Warner Losh cc: FreeBSD current Subject: Re: core group topics In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Thu, 31 Jul 1997, Warner Losh wrote: > In message Chuck Robey writes: > : 1) Satoshi's new Makefile. His makefile has been around plenty long > : enough for testing, and while several people have reported good results, > : I've seen no negatives reported at all. Will it get ignored or committed? > > Heck, you have commit privs, so you could do it yourself :-). That's > what I've been doing with the OpenBSD security fixes, just doing it > when I have time, and backing things out that escape my testing > procedures. > > Granted this is a bigger deal thing than my stuff. I felt that it's importance was such that only core should do it. Certainly I could do it. Let me ask Satoshi, he's core himself. > > : ELF > > If we do this, we should do it in 3.0. That's the perfect time > because all the libraries are changing major numbers (or most) and > people would expect much disruption. This would be a major diff > between 2.2.x and 3.x. > > Heck, at least integrate it into the tree and make it possible to > build the new stuff with a DOELF=yes in /etc/make.conf. That is the > next logical step, no? > > Warner > > > > ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 213 Lakeside Drive Apt T-1 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-current Thu Jul 31 10:52:44 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA23043 for current-outgoing; Thu, 31 Jul 1997 10:52:44 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA23037 for ; Thu, 31 Jul 1997 10:52:41 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id KAA06386; Thu, 31 Jul 1997 10:51:47 -0700 (PDT) Message-Id: <199707311751.KAA06386@austin.polstra.com> To: Terry Lambert cc: msmith@atrad.adelaide.edu.au, current@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( In-reply-to: Your message of "Wed, 30 Jul 1997 10:05:24 PDT." <199707301705.KAA05229@phaeton.artisoft.com> Date: Thu, 31 Jul 1997 10:51:47 -0700 From: John Polstra Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > > not. (The usual reply to this is "let's just make our ELF tools > > prepend the underscore too," although I bet you know better than > > that. You're not allowed to do it because the ELF spec says you > > can't. If we want ELF, then let's have ELF and not elf.out or > > a.elf.) > > Don't be so sure that I am not an "underscore advocate". There's > plenty of code that distinguishes the assembly from the C symbol > name space using underscores... the non-underscore functions are > a means of providing duplicate symbol hiding. Sorry, but for me this issue is non-negotiable. From the ELF spec: External C symbols have the same names in C, assembly code, and object files' symbol tables. If you add the underscore, it's not ELF. > In many cases, the removal of the C symbol underscore is BAD(tm), > in that it causes a collision with an assembly function of the same > name. In ELF, global assembly language symbols that are intended to be hidden from C generally look like ".foo". It's just a different convention, that's all. I like the ELF convention better, because it eliminates all confusion about whether you should pass "foo" or "_foo" to functions like dlopen() and nlist(). -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Thu Jul 31 11:43:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA25568 for current-outgoing; Thu, 31 Jul 1997 11:43:21 -0700 (PDT) Received: from friley01.res.iastate.edu (friley01.res.iastate.edu [129.186.189.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA25554 for ; Thu, 31 Jul 1997 11:43:05 -0700 (PDT) Received: from friley01.res.iastate.edu (loopback [127.0.0.1]) by friley01.res.iastate.edu (8.8.6/8.8.5) with ESMTP id NAA14531 for ; Thu, 31 Jul 1997 13:45:53 -0500 (CDT) Message-Id: <199707311845.NAA14531@friley01.res.iastate.edu> To: freebsd-current@freebsd.org Subject: Myrinet, etc.. (Re: code talks: announcing EIDE bus master patches) In-reply-to: Your message of Thu, 31 Jul 1997 12:47:07 +0930. <199707310317.MAA25355@genesis.atrad.adelaide.edu.au> Date: Thu, 31 Jul 1997 13:45:53 -0500 From: Chris Csanady Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >Stefan Esser stands accused of saying: >> > (*) PCI device == a Myrinet M2F-PCI32 card. This is a programmable >> > gigabit networking card. It has a 256k bank of SRAM on the card, and >> > is very good for doing things like measuring PCI b/w. The tests were >> > done from user space operating on mmap'ed device memory & a kernel >> > allocated chunk of RAM to do DMA xfers to/from. It also runs IP >> > traffic at better than 300Mb/sec. >> I'd also like to point out that the 300Mb's is acheived using an IP stack layered on top of some active messaging protocol. (which is implemented on the io processor on the nic.) As far as the TCP/IP stack under FreeBSD, all you can push through it is about 150Mb/s. This is somewhat unfortunate, although surprisingly linux doesn't seem to manage much better. I think it would be quite nice if we could correctly implement a zero copy architecture.. Chris Csanady From owner-freebsd-current Thu Jul 31 12:06:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA26927 for current-outgoing; Thu, 31 Jul 1997 12:06:26 -0700 (PDT) Received: from Ilsa.StevesCafe.com (Ilsa.StevesCafe.com [205.168.119.129]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA26921 for ; Thu, 31 Jul 1997 12:06:21 -0700 (PDT) Received: from Ilsa.StevesCafe.com (localhost [127.0.0.1]) by Ilsa.StevesCafe.com (8.8.6/8.8.5) with ESMTP id NAA18830; Thu, 31 Jul 1997 13:05:57 -0600 (MDT) Message-Id: <199707311905.NAA18830@Ilsa.StevesCafe.com> X-Mailer: exmh version 2.0gamma 1/27/96 From: Steve Passe To: John Polstra cc: current@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( In-reply-to: Your message of "Thu, 31 Jul 1997 10:51:47 PDT." <199707311751.KAA06386@austin.polstra.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 31 Jul 1997 13:05:57 -0600 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Hi, FWIW, my vote is to bite the bullet and go with ELF on 3.0, kernel and apps. As others have said, its the right point & opportunity to make the transition. If not now, when? -- Steve Passe | powered by smp@csn.net | Symmetric MultiProcessor FreeBSD From owner-freebsd-current Thu Jul 31 12:26:54 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA27777 for current-outgoing; Thu, 31 Jul 1997 12:26:54 -0700 (PDT) Received: from whistle.com (s205m131.whistle.com [207.76.205.131]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA27768 for ; Thu, 31 Jul 1997 12:26:51 -0700 (PDT) Received: (from smap@localhost) by whistle.com (8.7.5/8.6.12) id MAA23816; Thu, 31 Jul 1997 12:26:20 -0700 (PDT) Received: from bubba.whistle.com(207.76.205.7) by whistle.com via smap (V1.3) id sma023812; Thu Jul 31 12:25:52 1997 Received: (from archie@localhost) by bubba.whistle.com (8.8.5/8.6.12) id MAA27062; Thu, 31 Jul 1997 12:25:52 -0700 (PDT) From: Archie Cobbs Message-Id: <199707311925.MAA27062@bubba.whistle.com> Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707311905.NAA18830@Ilsa.StevesCafe.com> from Steve Passe at "Jul 31, 97 01:05:57 pm" To: smp@csn.net (Steve Passe) Date: Thu, 31 Jul 1997 12:25:52 -0700 (PDT) Cc: jdp@polstra.com, current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL31 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > FWIW, my vote is to bite the bullet and go with ELF on 3.0, kernel > and apps. As others have said, its the right point & opportunity to > make the transition. If not now, when? FWIW, I agree as well. Also, the JDK is important to me. -Archie ___________________________________________________________________________ Archie Cobbs * Whistle Communications, Inc. * http://www.whistle.com From owner-freebsd-current Thu Jul 31 13:26:29 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA01309 for current-outgoing; Thu, 31 Jul 1997 13:26:29 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id NAA01299 for ; Thu, 31 Jul 1997 13:26:26 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id NAA07684; Thu, 31 Jul 1997 13:22:21 -0700 From: Terry Lambert Message-Id: <199707312022.NAA07684@phaeton.artisoft.com> Subject: Re: core group topics To: dg@root.com Date: Thu, 31 Jul 1997 13:22:20 -0700 (MST) Cc: phk@dk.tfs.com, jkh@time.cdrom.com, chuckr@Glue.umd.edu, freebsd-current@freefall.FreeBSD.org In-Reply-To: <199707311142.EAA28471@implode.root.com> from "David Greenman" at Jul 31, 97 04:42:32 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > >Are we getting the kernel & userland mixed up here ? I belive there > >is strong agreement for ELF for the kernel... > > There is? I don't recall discussing that at all, and if we do eventually go > the ELF way, I think the kernel would be one of the last areas we'd want to > convert since it further complicates our cramped bootblocks. This is only true if you maintain support for booting non-ELF kernels; I view this as more of a transition issue than anything else. > The only issue I have against ELF is that I'm concerned that the overhead > for processing the much more sophisticated header at exec time might have a > serious impact on exec performance (something I'm particularly sensitive to > since I wrote the a.out exec code for FreeBSD). This is a bvery good point. But I think the benefits outweigh the costs. In particualr, the kernel could map the ld.so instead of making the crt0.o do it; this would buy back some of the overhead right away, if it stayed mapped in the cloned porcess address space from the fork to the exec. The same is actually true of the libc.so and other shared libraries, albiet a bit of an arbitrary mapping stretch (ie: pick a high address and map down). Specifically, I'd only place libc in this boat, seeing as how few programs can exist without system calls of one kind or another. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Thu Jul 31 13:27:55 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA01431 for current-outgoing; Thu, 31 Jul 1997 13:27:55 -0700 (PDT) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.50]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id NAA01421 for ; Thu, 31 Jul 1997 13:27:50 -0700 (PDT) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id NAA07693; Thu, 31 Jul 1997 13:23:44 -0700 From: Terry Lambert Message-Id: <199707312023.NAA07693@phaeton.artisoft.com> Subject: Re: core group topics To: phk@dk.tfs.com (Poul-Henning Kamp) Date: Thu, 31 Jul 1997 13:23:44 -0700 (MST) Cc: dg@root.com, phk@dk.tfs.com, jkh@time.cdrom.com, chuckr@Glue.umd.edu, freebsd-current@freefall.FreeBSD.org In-Reply-To: <244.870351027@critter.dk.tfs.com> from "Poul-Henning Kamp" at Jul 31, 97 02:10:27 pm X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > >the ELF way, I think the kernel would be one of the last areas we'd want to > >convert since it further complicates our cramped bootblocks. > > Have you looked at multiboot/GRUB from Eric Boleyn (sp?) yet ? Far be it for me to come off sounding anti-ELF, when I've been so violently in favor of it for so long, but... I had a number of large problems getting the GRUB code to actually work. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. From owner-freebsd-current Thu Jul 31 14:07:39 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA03763 for current-outgoing; Thu, 31 Jul 1997 14:07:39 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [206.246.122.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id OAA03755 for ; Thu, 31 Jul 1997 14:07:35 -0700 (PDT) Received: from Journey2.mat.net (journey2.mat.net [206.246.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id RAA22904; Thu, 31 Jul 1997 17:05:44 -0400 Date: Thu, 31 Jul 1997 17:05:57 -0400 (EDT) From: Chuck Robey X-Sender: chuckr@Journey2.mat.net To: Terry Lambert cc: Poul-Henning Kamp , dg@root.com, jkh@time.cdrom.com, freebsd-current@freefall.FreeBSD.org Subject: Re: core group topics In-Reply-To: <199707312023.NAA07693@phaeton.artisoft.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Thu, 31 Jul 1997, Terry Lambert wrote: > > >the ELF way, I think the kernel would be one of the last areas we'd want to > > >convert since it further complicates our cramped bootblocks. > > > > Have you looked at multiboot/GRUB from Eric Boleyn (sp?) yet ? > > Far be it for me to come off sounding anti-ELF, when I've been > so violently in favor of it for so long, but... I had a number > of large problems getting the GRUB code to actually work. I know very little about boot loaders, but I was wondering if anyone had seen a tool called solo, which seems to be a bootloder by Frank Barrus? The important part from the README has: Complete, unaltered, original distributions of this and related code may be obtained from: http://www.csh.rit.edu/~shaggy/software.html or: ftp://ftp.csh.rit.edu/pub/csh/shaggy > > > Terry Lambert > terry@lambert.org > --- > Any opinions in this posting are my own and not those of my present > or previous employers. > > ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 213 Lakeside Drive Apt T-1 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-current Thu Jul 31 14:10:28 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA04063 for current-outgoing; Thu, 31 Jul 1997 14:10:28 -0700 (PDT) Received: from rah.star-gate.com (rah.star-gate.com [204.188.121.18]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA04058 for ; Thu, 31 Jul 1997 14:10:21 -0700 (PDT) Received: from rah.star-gate.com (localhost.star-gate.com [127.0.0.1]) by rah.star-gate.com (8.8.5/8.8.5) with ESMTP id OAA08041; Thu, 31 Jul 1997 14:09:52 -0700 (PDT) Message-Id: <199707312109.OAA08041@rah.star-gate.com> To: Archie Cobbs cc: smp@csn.net (Steve Passe), jdp@polstra.com, current@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( In-reply-to: Your message of "Thu, 31 Jul 1997 12:25:52 PDT." <199707311925.MAA27062@bubba.whistle.com> Date: Thu, 31 Jul 1997 14:09:52 -0700 From: Amancio Hasty Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Me too 8) an elf is important for C++. Amancio >From The Desk Of Archie Cobbs : > > > FWIW, my vote is to bite the bullet and go with ELF on 3.0, kernel > > and apps. As others have said, its the right point & opportunity to > > make the transition. If not now, when? > > FWIW, I agree as well. Also, the JDK is important to me. > > -Archie > > ___________________________________________________________________________ > Archie Cobbs * Whistle Communications, Inc. * http://www.whistle.com From owner-freebsd-current Thu Jul 31 14:44:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA05587 for current-outgoing; Thu, 31 Jul 1997 14:44:36 -0700 (PDT) Received: from smtp2.xs4all.nl (smtp2.xs4all.nl [194.109.6.52]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA05580 for ; Thu, 31 Jul 1997 14:44:31 -0700 (PDT) Received: from asterix.xs4all.nl (root@asterix.xs4all.nl [194.109.6.11]) by smtp2.xs4all.nl (8.8.6/XS4ALL) with ESMTP id XAA23925 for ; Thu, 31 Jul 1997 23:44:26 +0200 (CEST) Received: from plm.xs4all.nl (uucp@localhost) by asterix.xs4all.nl (8.8.6/8.8.6) with UUCP id XAA11202 for freebsd-current@freebsd.org; Thu, 31 Jul 1997 23:42:55 +0200 (MET DST) Received: (from plm@localhost) by plm.xs4all.nl (8.8.5/8.7.3) id WAA05489; Thu, 31 Jul 1997 22:54:37 +0200 (MET DST) To: freebsd-current@freebsd.org Subject: Re: core group topics References: <87pvrzvto5.fsf@totally-fudged-out-message-id> From: Peter Mutsaers Date: 31 Jul 1997 22:54:37 +0200 In-Reply-To: TLiddelow@cybec.com.au's message of Thu, 31 Jul 1997 18:00:16 +1000 Message-ID: <87204fndrm.fsf@plm.xs4all.nl> Lines: 52 X-Mailer: Gnus v5.4.25/Emacs 19.34 Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >> On Thu, 31 Jul 1997 18:00:16 +1000, TLiddelow@cybec.com.au (Tim >> Liddelow) said: TL> Jordan K. Hubbard wrote: >> >> Erm, "none excepting inertia and fear of chaos, confusion, shock and >> increased entropy all around." :) Inflicting another ABI on our user >> base, one which will result in a new class of packages which all older >> FreeBSD releases cannot run, is NOT something to be considered >> lightly. >> >> Jordan TL> It doesn't have to be _that_ hard. The Linux camp did the dual ABI TL> thing for quite a while. Sure - the transition wasn't all beer and TL> skittles but they did it. And they're reaping the rewards (see TL> previous postings) for it. I agree completely. I was using Linux at the time that ELF first appeared there. I fetched the ELF packages and libraries and converted my system slowly myself. It was pretty painless, for anyone who knows a little bit about Unix and who has some common sense. There was some upset about this switch in the Linux camp, but only because: - there are many clueless Linux users (much more than freebsd users) AND - Linux doesn't come in one single consistent distrubution. This combination (clueless users trying to install software themselves without a clear ports mechanism etc.) was pretty tough for many; note that this switch occurred even before the somewhat better distributions such as Debian or Red Hat existed. Most users at that time were still assembling their Linux systems by themselves. FreeBSD users will have much less troubles with a transition to ELF as the Linux camp had because 1. Either FreeBSD users are more knowledgeble and/or 2. FreeBSD comes better packaged. Users don't go out and fetch packages from many sites and install them by themselves. Instead, they have the single and consistent base distribution, and add packages via the ports. From owner-freebsd-current Thu Jul 31 14:59:50 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA06519 for current-outgoing; Thu, 31 Jul 1997 14:59:50 -0700 (PDT) Received: from shale.csir.co.za (shale.csir.co.za [146.64.46.5]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA06510 for ; Thu, 31 Jul 1997 14:59:45 -0700 (PDT) Received: (from reg@localhost) by shale.csir.co.za (8.8.6/8.8.5) id AAA13118; Fri, 1 Aug 1997 00:00:19 +0200 (SAT) Message-ID: <19970801000015.50699@shale.csir.co.za> Date: Fri, 1 Aug 1997 00:00:15 +0200 From: Jeremy Lea To: "Jordan K. Hubbard" Cc: current@FreeBSD.ORG Subject: Re: More bogons in Makefiles... References: <19970730225235.58179@shale.csir.co.za> <19894.870331035@time.cdrom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.76 In-Reply-To: <19894.870331035@time.cdrom.com>; from Jordan K. Hubbard on Wed, Jul 30, 1997 at 11:37:15PM -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Wed, Jul 30, 1997 at 11:37:15PM -0700, Jordan K. Hubbard wrote: > Assuming anything about the contents of the final destination > directories for what may be a simple build is highly bogus and, while > this is still done in a number of places in /usr/src, the eventual > intent is to get rid of this and make /usr/src build in a properly > self-referential fashion. > > > src tree. I percieve the purpose of 'make world' to be to make sure that > > these libraries get installed before the build happens. Or are these > > It's the current kludge of `make world' to do this, but it's still not > its most ideal purpose. :) Ok, I think I'm getting the idea here, from what's been said here, in the "core group topics" thread and Satoshi's Makefile: The build process must be able to go from a src tree in '/a/foo' (defaulting to '/usr/src', which maybe read only), build with cross-compiling, a.out-> ELF or some such a build environment in another tree '/a/bar' (defaulting to '/usr/obj/usr/src/tmp' in Satoshi's patch or ${WORLDTMP}, a mix of '/usr/obj' or ${MAKEOBJDIRPREFIX} and '/' in current system) which can then by used to build the world in a temporary tree '/b/bas' (defaulting to '/usr/obj') which can then be used for multiple reinstalls via NFS into '/' or ${DESTDIR} for releases. *And* it must be able to do this without to much magic in the ording of steps in Makefile's and be 'make -j' safe. It's a wonder it works at all ;) I s'pose the second requirement is that it also work without having to require separate build environments etc, so that people can use it to stay -current while only building things that change, much like is done by most users. Anyway, where should I concentrate my efforts then ;)... I s'pose on making sure that everything gets built into ${WORLDTMP} and ${MAKEOBJDIRPREFIX} and ${DESTDIR}. What I've been doing at the moment is getting the tree to build and install in one pass, since I dont have enough space to store /usr/obj and I'm being stuborn about getting more... the changes I've fed through already and about 100 lines more (excluding gratitious diffs to /usr/src/Makefile) allow me to do the following in each directory (except for the next problem...): cd {.CURDIR} && make obj && make all install cleandir (I wish 'make obj all install cleandir' worked)... You'll notice this also fairly close to what most of the targets in the bootstrapping sections of 'make world' are doing. I'm worried by all the cross linking to libraries in /usr/obj because I'm not sure what I'm breaking or not picking up correctly. > > Is there a clean way of building and installing the lib in > > ${.OBJDIR}/.. as a first step in the Makefile, and cleaning it from the > > other subdirectory's Makefile if only that subdirectory was built? > > Not really, no. Not a clean way. :-) This is where most of the patching comes in... you have to make changes to a lot of makefiles to have them stop and make obj && make all && make install && make cleandir across their subtree. Would it be possible to add another .mk file which these could use rather than bsd.subdir.mk, with a before all traget which would build the shared directories and install them in ${.OBJDIR} at the top level and then an afterclean target to remove them. I'm not sure how to quite set up the dependancies so that the subdirs could perform a 'cd ${.CURDIR}../libfoo && make all install' if they were built seprately and have them clean up after themselves (in much the same fashion as ports do), but building the subdirs individually seems like a strange practice anyway, and they should probably just print a warning or cd .. and make. I'm working on diffs for this at the moment. You need about 15Mb free for /usr/obj a 'make world' like this... I wonder if you can build this in a mfs... should be fairly fast ;) > Yes, as you can see, things aren't always 100% consistent in there and > you've certainly identified some definite bogons. I or someone will > review and commit from this list fairly shortly. Thanks. No rush though, since they obviously aren't tripping to many people up. It's just good to remove possible problems I think. > You seem to be doing pretty well so far. :-) Thank you. -Jeremy -- | "In this world of temptation, I will stand for what is right. --+-- With a heart of salvation, I will hold up the light. | If I live or if I die, if I laugh or if I cry, | in this world of temptation, I will stand." -Pam Thum From owner-freebsd-current Thu Jul 31 15:29:31 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA08087 for current-outgoing; Thu, 31 Jul 1997 15:29:31 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA08080 for ; Thu, 31 Jul 1997 15:29:27 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id PAA20234; Thu, 31 Jul 1997 15:28:44 -0700 (PDT) To: Jeremy Lea cc: current@FreeBSD.ORG Subject: Re: More bogons in Makefiles... In-reply-to: Your message of "Fri, 01 Aug 1997 00:00:15 +0200." <19970801000015.50699@shale.csir.co.za> Date: Thu, 31 Jul 1997 15:28:44 -0700 Message-ID: <20230.870388124@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Overall, I think that the focus should be on clarity over disk space savings. Your stubbornness about obtaining more space aside, I think that the trend is towards disk space becoming cheaper than programmer brain cells, and I would not want to see the Makefiles cluttered with changes aimed at supporting partial builds - I think that the whole partial build problem is a fool's errand and not worth the time better spent in making dependencies work more consistently in a complete src tree. Jordan From owner-freebsd-current Thu Jul 31 15:50:10 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA09164 for current-outgoing; Thu, 31 Jul 1997 15:50:10 -0700 (PDT) Received: from shale.csir.co.za (shale.csir.co.za [146.64.46.5]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA09112 for ; Thu, 31 Jul 1997 15:49:57 -0700 (PDT) Received: (from reg@localhost) by shale.csir.co.za (8.8.6/8.8.5) id AAA14411; Fri, 1 Aug 1997 00:50:22 +0200 (SAT) Message-ID: <19970801005017.53925@shale.csir.co.za> Date: Fri, 1 Aug 1997 00:50:17 +0200 From: Jeremy Lea To: Bruce Evans Cc: current@FreeBSD.ORG Subject: Re: More bogons in Makefiles... References: <199707311018.UAA26613@godzilla.zeta.org.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.76 In-Reply-To: <199707311018.UAA26613@godzilla.zeta.org.au>; from Bruce Evans on Thu, Jul 31, 1997 at 08:18:54PM +1000 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Thu, Jul 31, 1997 at 08:18:54PM +1000, Bruce Evans wrote: > Paths relative to the source tree are preferred, although this causes > problems with partial source trees. Paths relative to the object tree > are used by many sloppy Makefiles Ok, I'll look out for these while I'm bumbling around. > with a per-source directory obj symlink)). Non sloppy makefiles would > use a more complicated search for the object directory that would > actually be used if `make' were run in the external source directory. An example being? ;) > The path here should be ${.CURDIR/../../secure}. I'll watch out for this too then... > There are more bogons here. The whole ifdef is a verbose and confusing > way write `YEARISTYPE= ${.OBJDIR}/yearistype. yearistype is always > created in the object directory and ${.OBJDIR} always exists (although > it's not clear where it is when make is invoked with args `obj all ...'). >From this I read ${.OBJDIR} == ${.CURDIR} iff no 'make obj'. If that is correct then there is a lot of funny business going on, because there are a lot of checks this form. > Because I don't use it :-). bsd.libnames.mk is no longer used in -current, > so this problem will go away automatically. Looks suspiciously like bsd.lib.mk includes this file... but now I'm lost ;) where do all the other defines for libraries come from? > >Oops, one last thing: should ./usr.bin/lex/lib be built twice in the > >bootstrapping of 'make world' (as part of lib-tools and libraries) and > > It doesn't hurt. It uses CPU cycles without any gain, therefore the net result is -ve. Thanks for the input. -Jeremy -- | "In this world of temptation, I will stand for what is right. --+-- With a heart of salvation, I will hold up the light. | If I live or if I die, if I laugh or if I cry, | in this world of temptation, I will stand." -Pam Thum From owner-freebsd-current Thu Jul 31 16:12:23 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA10042 for current-outgoing; Thu, 31 Jul 1997 16:12:23 -0700 (PDT) Received: from shale.csir.co.za (shale.csir.co.za [146.64.46.5]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA10035 for ; Thu, 31 Jul 1997 16:12:14 -0700 (PDT) Received: (from reg@localhost) by shale.csir.co.za (8.8.6/8.8.5) id BAA16153; Fri, 1 Aug 1997 01:12:01 +0200 (SAT) Message-ID: <19970801011154.07240@shale.csir.co.za> Date: Fri, 1 Aug 1997 01:11:54 +0200 From: Jeremy Lea To: "Jordan K. Hubbard" Cc: current@FreeBSD.ORG Subject: Re: More bogons in Makefiles... References: <19970801000015.50699@shale.csir.co.za> <20230.870388124@time.cdrom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.76 In-Reply-To: <20230.870388124@time.cdrom.com>; from Jordan K. Hubbard on Thu, Jul 31, 1997 at 03:28:44PM -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Hi, On Thu, Jul 31, 1997 at 03:28:44PM -0700, Jordan K. Hubbard wrote: > Overall, I think that the focus should be on clarity over disk space > savings. Your stubbornness about obtaining more space aside, I think > that the trend is towards disk space becoming cheaper than programmer > brain cells, and I would not want to see the Makefiles cluttered with > changes aimed at supporting partial builds - I think that the whole > partial build problem is a fool's errand and not worth the time better > spent in making dependencies work more consistently in a complete > src tree. > > Jordan Oh, I agree... why do you think the other 100 lines of diffs weren't in the message ;) But there is nothing which wastes brain cells more than spaghetti, and some of the Makefiles make even the worst assembly look good. The more the src tree gets interwoven the more difficult it becomes to upgrade components, remove components, etc. I've taken the view that each src directory contains a component, and that they should be able to know which other components they depend on, much like the ports do, but without the problems of fetching and patching. Basically a component must be able to say 'build x, y, and z (if it's around) before me'. I dont think that is any different to what you are saying, or have been saying for as long as I've been reading these lists. I'm only stubborn about disk space because it forces me to learn... lots of processing power and diskspace will only get you Free95 ;) -Jeremy -- | "In this world of temptation, I will stand for what is right. --+-- With a heart of salvation, I will hold up the light. | If I live or if I die, if I laugh or if I cry, | in this world of temptation, I will stand." -Pam Thum From owner-freebsd-current Thu Jul 31 16:34:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA10990 for current-outgoing; Thu, 31 Jul 1997 16:34:18 -0700 (PDT) Received: from rah.star-gate.com (rah.star-gate.com [204.188.121.18]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA10981 for ; Thu, 31 Jul 1997 16:34:14 -0700 (PDT) Received: (from hasty@localhost) by rah.star-gate.com (8.8.5/8.8.5) id QAA08703 for current@freebsd.org; Thu, 31 Jul 1997 16:34:01 -0700 (PDT) Date: Thu, 31 Jul 1997 16:34:01 -0700 (PDT) From: Amancio Hasty Message-Id: <199707312334.QAA08703@rah.star-gate.com> To: current@freebsd.org Subject: elf patches for current? Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I have a couple or large comm packages (ilu and ace ) which I will like to try out with elf. Tnks, Amancio From owner-freebsd-current Thu Jul 31 16:54:38 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA11765 for current-outgoing; Thu, 31 Jul 1997 16:54:38 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA11760 for ; Thu, 31 Jul 1997 16:54:28 -0700 (PDT) Received: from hurricane.cs.duke.edu (hurricane.cs.duke.edu [152.3.145.1]) by duke.cs.duke.edu (8.8.5/8.8.5) with ESMTP id TAA09826; Thu, 31 Jul 1997 19:54:22 -0400 (EDT) Received: (from gallatin@localhost) by hurricane.cs.duke.edu (8.8.4/8.7.3) id TAA29288; Thu, 31 Jul 1997 19:54:22 -0400 (EDT) Date: Thu, 31 Jul 1997 19:54:22 -0400 (EDT) Message-Id: <199707312354.TAA29288@hurricane.cs.duke.edu> From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Chris Csanady Cc: freebsd-current@FreeBSD.ORG Subject: Re: Myrinet, etc.. (Re: code talks: announcing EIDE bus master patches) In-Reply-To: <199707311845.NAA14531@friley01.res.iastate.edu> References: <199707310317.MAA25355@genesis.atrad.adelaide.edu.au> <199707311845.NAA14531@friley01.res.iastate.edu> X-Mailer: VM 6.32 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Chris Csanady writes: > > >Stefan Esser stands accused of saying: > >> > (*) PCI device == a Myrinet M2F-PCI32 card. This is a programmable > >> > gigabit networking card. It has a 256k bank of SRAM on the card, and > >> > is very good for doing things like measuring PCI b/w. The tests were > >> > done from user space operating on mmap'ed device memory & a kernel > >> > allocated chunk of RAM to do DMA xfers to/from. It also runs IP > >> > traffic at better than 300Mb/sec. > >> > > I'd also like to point out that the 300Mb's is acheived using an IP > stack layered on top of some active messaging protocol. (which is > implemented on the io processor on the nic.) > > As far as the TCP/IP stack under FreeBSD, all you can push through it > is about 150Mb/s. This is somewhat unfortunate, although surprisingly > linux doesn't seem to manage much better. I think it would be quite > nice if we could correctly implement a zero copy architecture.. > > Chris Csanady Actually, that's not true. I'm using an in-kernel IP driver layered on top of Myricom's MyriApi general-purpose messaging software (which is kernel resident). It hooks into the network stack the way any ethernet driver would (and gets hit with a copy for each xmit from or receive into user space like anything else does). There is software that's downloaded onto the interface card, but it has no protocol specific knowledge. When its sitting under the IP driver, it is used the same way the firwmware on any ethernet card would be used -- the driver tells it a MAC address that it wants something sent to, and it sends it, etc. The fact that we're running over Myrinet & not ethernet is transparent to any application. In this configuration, with the hardware I currently have available (memory b/w challenged Pentium Pros) I can receive UDP traffic at about 356 Mb/sec and send UDP traffic at a rate of a about 280Mb/sec. TCP streams are around 275Mb/sec. I think my transmission speed is limited both by the memory b/w limitations of Pentium Pro's (from the uiomove copy to/from user space) & the fact that the myri MTU is more than twice the 4k hw pagesize of the Intel platform, so I'm incurring a lot of dma setup overhead on transmits. (I manage my own set of MTU-sized receive buffers, so there's only 1 dma required for each receive). I've just noticed that -current appears to support having MCLBYTES > page size; this should improve transmit speed some. Please don't get me wrong: I agree that it really would be nice to have a zero-copy IP stack. Drew PS: For more performance numbers check out the netperf pages (http://www.cup.hp.com/netperf/numbers/NetperfBrowse.html)..I'm happy to report that FreeBSD w/Myrinet is #2 (behind a pair of 10 & 16 processor SGI Power Challenges connected via HIPPI), and/or our group's pages (http://www.cs.duke.edu/ari/manic/ip). ------------------------------------------------------------------------------ Andrew Gallatin http://www.cs.duke.edu/~gallatin Duke University Email: gallatin@cs.duke.edu Department of Computer Science Phone: (919) 660-6590 From owner-freebsd-current Thu Jul 31 17:07:44 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA12227 for current-outgoing; Thu, 31 Jul 1997 17:07:44 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA12219 for ; Thu, 31 Jul 1997 17:07:38 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id JAA00566; Fri, 1 Aug 1997 09:36:15 +0930 (CST) From: Michael Smith Message-Id: <199708010006.JAA00566@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: from Chuck Robey at "Jul 31, 97 05:05:57 pm" To: chuckr@glue.umd.edu (Chuck Robey) Date: Fri, 1 Aug 1997 09:36:14 +0930 (CST) Cc: terry@lambert.org, phk@dk.tfs.com, dg@root.com, jkh@time.cdrom.com, freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Chuck Robey stands accused of saying: > > > > Far be it for me to come off sounding anti-ELF, when I've been > > so violently in favor of it for so long, but... I had a number > > of large problems getting the GRUB code to actually work. > > I know very little about boot loaders, but I was wondering if anyone had > seen a tool called solo, which seems to be a bootloder by Frank Barrus? Yup. It's another GPL'd unbuildable monstrosity 8) -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Thu Jul 31 17:08:53 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA12304 for current-outgoing; Thu, 31 Jul 1997 17:08:53 -0700 (PDT) Received: from mail.cdsnet.net (mail.cdsnet.net [204.118.244.5]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA12297 for ; Thu, 31 Jul 1997 17:08:50 -0700 (PDT) Received: from chop.cdsnet.net (chop.cdsnet.net [204.118.244.3]) by mail.cdsnet.net (8.8.6/8.8.6) with ESMTP id RAA15967; Thu, 31 Jul 1997 17:08:49 -0700 (PDT) Received: from nowin (1Cust116.max3.boston.ma.ms.uu.net [153.35.70.244]) by chop.cdsnet.net (8.8.5/8.8.5) with SMTP id RAA08430; Thu, 31 Jul 1997 17:08:45 -0700 (PDT) Message-Id: <3.0.3.32.19970731194947.030b5044@ranier.altavista-software.com> X-Sender: 3ampop@ranier.altavista-software.com X-Mailer: QUALCOMM Windows Eudora Pro Version 3.0.3 (32) Date: Thu, 31 Jul 1997 19:49:47 -0400 To: Terry Lambert From: Matt Thomas Subject: Re: core group topics Cc: freebsd-current@freefall.FreeBSD.org In-Reply-To: <199707312022.NAA07684@phaeton.artisoft.com> References: <199707311142.EAA28471@implode.root.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk At 01:22 PM 7/31/97 -0700, Terry Lambert wrote: >> >Are we getting the kernel & userland mixed up here ? I belive there >> >is strong agreement for ELF for the kernel... >> >> There is? I don't recall discussing that at all, and if we do eventually go >> the ELF way, I think the kernel would be one of the last areas we'd want to >> convert since it further complicates our cramped bootblocks. > >This is only true if you maintain support for booting non-ELF kernels; >I view this as more of a transition issue than anything else. > > >> The only issue I have against ELF is that I'm concerned that the overhead >> for processing the much more sophisticated header at exec time might have a >> serious impact on exec performance (something I'm particularly sensitive to >> since I wrote the a.out exec code for FreeBSD). > >This is a bvery good point. But I think the benefits outweigh the >costs. In particualr, the kernel could map the ld.so instead of >making the crt0.o do it; this would buy back some of the overhead >right away, if it stayed mapped in the cloned porcess address space >from the fork to the exec. The same is actually true of the libc.so >and other shared libraries, albiet a bit of an arbitrary mapping >stretch (ie: pick a high address and map down). Specifically, I'd >only place libc in this boat, seeing as how few programs can exist >without system calls of one kind or another. That's how NetBSD does it. It loads the executable and the interpt. (ld-elf.so) and transfers controls to the interpt instead of the executable. -- Matt Thomas Internet: matt@3am-software.com 3am Software Foundry WWW URL: http://www.3am-software.com/bio/matt.html Nashua, NH Disclaimer: I disavow all knowledge of this message From owner-freebsd-current Thu Jul 31 17:11:37 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA12430 for current-outgoing; Thu, 31 Jul 1997 17:11:37 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA12412; Thu, 31 Jul 1997 17:11:28 -0700 (PDT) Received: from hurricane.cs.duke.edu (hurricane.cs.duke.edu [152.3.145.1]) by duke.cs.duke.edu (8.8.5/8.8.5) with ESMTP id UAA09952; Thu, 31 Jul 1997 20:10:57 -0400 (EDT) Received: (from gallatin@localhost) by hurricane.cs.duke.edu (8.8.4/8.7.3) id UAA24020; Thu, 31 Jul 1997 20:10:56 -0400 (EDT) Date: Thu, 31 Jul 1997 20:10:56 -0400 (EDT) Message-Id: <199708010010.UAA24020@hurricane.cs.duke.edu> From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Michael Smith Cc: se@FreeBSD.ORG, freebsd-current@FreeBSD.ORG Subject: Re: code talks: announcing EIDE bus master patches In-Reply-To: <199707310311.MAA25330@genesis.atrad.adelaide.edu.au> References: <199707301924.PAA08936@hurricane.cs.duke.edu> <199707310311.MAA25330@genesis.atrad.adelaide.edu.au> X-Mailer: VM 6.32 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Michael Smith writes: > Andrew Gallatin stands accused of saying: > > > > I suppose this is as good a time to ask as any: I have a FreeBSD > > driver 95% done for Myrinet cards such as this. Who can I talk to > > about the possibility of getting it committed to -current? There are > > some sticky issues that may prevent its inclusion (specifically > > GPL'ed libraries that are required to communicate w/the vendor > > supported "firmware" that runs on the card & which I cannot think of a > > clean way to build). > > The GPL'd libraries won't prevent it being included; there is a > precedent for handling useful GPL kernel components. > > If nobody else sticks their hand up, I'd be happy to do the actual > grunt work involved in committing it. Thanks for the offer! Let's take this dialogue offline for now.. There are some details I'd like to talk about w/you that I don't want to bore everybody with. Drew ------------------------------------------------------------------------------ Andrew Gallatin http://www.cs.duke.edu/~gallatin Duke University Email: gallatin@cs.duke.edu Department of Computer Science Phone: (919) 660-6590 From owner-freebsd-current Thu Jul 31 19:02:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA17261 for current-outgoing; Thu, 31 Jul 1997 19:02:42 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA17251; Thu, 31 Jul 1997 19:02:38 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id TAA03183; Thu, 31 Jul 1997 19:02:25 -0700 (PDT) Message-Id: <199708010202.TAA03183@austin.polstra.com> To: hsu@FreeBSD.ORG Subject: Re: pccard and -current; a long way to go. :-( In-Reply-To: <199707300710.AAA14949@hub.freebsd.org> Organization: Polstra & Co., Seattle, WA Cc: current@FreeBSD.ORG Date: Thu, 31 Jul 1997 19:02:25 -0700 From: John Polstra Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In article <199707300710.AAA14949@hub.freebsd.org>, Jeffrey Hsu wrote: > I see this problem too in conjunction with language tools. My next > JDK port which I'm working on now (not sure if I should do a stable > 1.1.4 port or just skip ahead to 1.2, neither of which the Linux > people have yet) would happily use ELF if it solves the dlsym(RTLD_NEXT) > problem (namely, we don't have it) That's not really related to ELF. It's caused by my own neglect. :-) Being reminded of it made me feel so guilty that I implemented it this afternoon. I'll send you some patches to test if you're interested. > and it also eliminates a bunch > of kludges in the Java interpreter code that I have to make because > we prepend underscores and the rest of the world doesn't. You're talking about calls to dlsym(), I assume. Personally, I think that the argument to dlsym() ought to be the symbol name *as it is written in C*. For a.out, dlsym() itself should add the leading underscore if necessary. I.e., you shouldn't have to call the function differently depending on what your object format is. Unfortunately, it wasn't done that way originally in FreeBSD. It would break some things to change it now. Of course, it would break most of the same things if we switched to ELF ... John -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Thu Jul 31 19:48:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA19711 for current-outgoing; Thu, 31 Jul 1997 19:48:34 -0700 (PDT) Received: from friley01.res.iastate.edu (friley01.res.iastate.edu [129.186.189.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA19703 for ; Thu, 31 Jul 1997 19:48:29 -0700 (PDT) Received: from friley01.res.iastate.edu (loopback [127.0.0.1]) by friley01.res.iastate.edu (8.8.6/8.8.5) with ESMTP id VAA15322; Thu, 31 Jul 1997 21:52:03 -0500 (CDT) Message-Id: <199708010252.VAA15322@friley01.res.iastate.edu> To: Andrew Gallatin cc: freebsd-current@freebsd.org Subject: Re: Myrinet, etc.. (Re: code talks: announcing EIDE bus master patches) In-reply-to: Your message of Thu, 31 Jul 1997 19:54:22 -0400. <199707312354.TAA29288@hurricane.cs.duke.edu> Date: Thu, 31 Jul 1997 21:52:03 -0500 From: Chris Csanady Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > > I'd also like to point out that the 300Mb's is acheived using an IP > > stack layered on top of some active messaging protocol. (which is > > implemented on the io processor on the nic.) > > > > As far as the TCP/IP stack under FreeBSD, all you can push through it > > is about 150Mb/s. This is somewhat unfortunate, although surprisingly > > linux doesn't seem to manage much better. I think it would be quite > > nice if we could correctly implement a zero copy architecture.. > > > > Chris Csanady > >Actually, that's not true. I'm using an in-kernel IP driver layered >on top of Myricom's MyriApi general-purpose messaging software (which >is kernel resident). It hooks into the network stack the way any >ethernet driver would (and gets hit with a copy for each xmit from or >receive into user space like anything else does). There is software >that's downloaded onto the interface card, but it has no protocol >specific knowledge. When its sitting under the IP driver, it is used >the same way the firwmware on any ethernet card would be used -- the >driver tells it a MAC address that it wants something sent to, and it >sends it, etc. Sorry, I guess I assumed wrong. However, the numbers you are reporting are vastly different from he ones mentioned on myrinet's web page for 2.2.1. >The fact that we're running over Myrinet & not ethernet is transparent >to any application. In this configuration, with the hardware I >currently have available (memory b/w challenged Pentium Pros) I can >receive UDP traffic at about 356 Mb/sec and send UDP traffic at a rate >of a about 280Mb/sec. TCP streams are around 275Mb/sec. That really is quite impressive--is your driver available? I would like to play with it when I get a chance. :) I'm also quite interested in how your interrupt/buffer architecture looks. As for memory bandwith limitations, I can't wait for the alpha port. ;) Chris Csanady From owner-freebsd-current Thu Jul 31 20:51:20 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA23059 for current-outgoing; Thu, 31 Jul 1997 20:51:20 -0700 (PDT) Received: from emout10.mail.aol.com (emout10.mx.aol.com [198.81.11.25]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA23054 for ; Thu, 31 Jul 1997 20:51:18 -0700 (PDT) From: StevenR362@aol.com Received: (from root@localhost) by emout10.mail.aol.com (8.7.6/8.7.3/AOL-2.0.0) id XAA00604; Thu, 31 Jul 1997 23:50:43 -0400 (EDT) Date: Thu, 31 Jul 1997 23:50:43 -0400 (EDT) Message-ID: <970731235042_1593313789@emout10.mail.aol.com> To: bde@zeta.org.au, cgull@smoke.marlboro.vt.us, freebsd-current@freebsd.org Subject: Re: problems with IDE DMA Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In a message dated 97-07-31 06:02:42 EDT, bde@zeta.org.au (Bruce Evans) writes: > After successfully testing IDE DMA for a little while (reading rwd0, > fsck -n and bonnie -s 64 worked fine) I got the following messages > after the write stages for bonnie -s 64: > > wd0: interrupt timeout: > wd0: status 50 error 0 > wd0: interrupt timeout: > wd0: status 58 error 1 > wd0: interrupt timeout: > wd0: status 58 error 1 > wd0: interrupt timeout: > wd0: status 58 error 1 > wd0: Last time I say: interrupt timeout. Probably a portable PC.: > wd0: status 58 error 1 > > The system then drifted South: > - no more interrupt timeouts, but wd seemed to get stuck in single > sector mode. ls on the bonnie output file was very slow, and > systat reported about 20 interrupts/sec, 10 "blks", and often > "***** msps" (division by 0?). > - long before the 13107 seconds required to read 128MB at 10K/sec, > the system hung (the keyboard LEDs didn't work). > - the test drive was dead after hard reset. > - the test drive was OK after cycling the power. > I very much doubt that this is a problem with the new DMA code. As I have been seeing this happen off and on since 386BSD with patchkit 2.4 days. It seems to happen when you have multiple processes hitting the drive hard and perhaps an interrupt gets lost. I sometimes see it during make worlds or when doing a cvs update in /usr/src and /usr/ports simultaneously. These errors appear on the console and the hard drive LED goes solid on with the whole system locked up inside of 30 seconds. I believe it is a timing related race condition in the wd.c driver. As various FreeBSD editions exhibit it where others don't. Current, checked out a couple of hours before the Lite2 merge at the beginning of the year was very bad in this regard. Whereas current from the past few months has been very solid. Steve From owner-freebsd-current Thu Jul 31 20:51:30 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA23079 for current-outgoing; Thu, 31 Jul 1997 20:51:30 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA23074 for ; Thu, 31 Jul 1997 20:51:27 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.6/8.6.9) id UAA17854; Thu, 31 Jul 1997 20:51:05 -0700 (PDT) Date: Thu, 31 Jul 1997 20:51:05 -0700 (PDT) Message-Id: <199708010351.UAA17854@silvia.HIP.Berkeley.EDU> To: imp@village.org CC: chuckr@glue.umd.edu, freebsd-current@freefall.freebsd.org In-reply-to: (message from Warner Losh on Thu, 31 Jul 1997 11:12:36 -0600) Subject: Re: core group topics From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk * Heck, you have commit privs, so you could do it yourself :-). That's * what I've been doing with the OpenBSD security fixes, just doing it * when I have time, and backing things out that escape my testing * procedures. * * Granted this is a bigger deal thing than my stuff. Um, it's generally not considered a good manner to commit someone else's work if that someone else also has a commit priviledge and that someone else didn't ask you to do so. :) You guys just got the latest version (with Bruce's fixes) this morning, if there aren't any objections in a few days, I'll commit it myself. Satoshi From owner-freebsd-current Thu Jul 31 21:10:17 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id VAA24223 for current-outgoing; Thu, 31 Jul 1997 21:10:17 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA24218 for ; Thu, 31 Jul 1997 21:10:13 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id VAA03660 for ; Thu, 31 Jul 1997 21:10:12 -0700 (PDT) Message-Id: <199708010410.VAA03660@austin.polstra.com> To: current@freebsd.org Subject: Re: core group topics In-Reply-To: <199707311142.EAA28471@implode.root.com> References: <199707311142.EAA28471@implode.root.com> Organization: Polstra & Co., Seattle, WA Date: Thu, 31 Jul 1997 21:10:12 -0700 From: John Polstra Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In article <199707311142.EAA28471@implode.root.com>, David Greenman wrote: > The only issue I have against ELF is that I'm concerned that the overhead > for processing the much more sophisticated header at exec time might have a > serious impact on exec performance (something I'm particularly sensitive to > since I wrote the a.out exec code for FreeBSD). This is a common concern, but I'm convinced it's not a problem. I think ELF can be loaded as fast as a.out, within the limits of measurability. Well, almost as fast, anyway. People look at the ELF spec and they think, "Gads, all those different kinds of sections! It must be hard to load one of these things." But to load an ELF program or shared library, you don't even look at the section table. Instead, you use the Program Header, which is specially constructed just for this purpose. The Program Header describes text, data, and bss segments, just like a.out does. It's not much different from a.out in complexity. To illustrate, the loop that does the loading in the ELF bootloader looks like this: printf("segments:"); for (i = 0; i < head.e_phnum; i++) { ph = (Elf32_Phdr*)(phbuf + head.e_phentsize * i); if (ph->p_type == PT_LOAD) { ph->p_vaddr &= ADDR_MASK; printf(" 0x%x-0x%x", ph->p_vaddr, ph->p_vaddr + ph->p_memsz); if (ph->p_filesz > 0) { poff = ph->p_offset; xread((void *)ph->p_vaddr, ph->p_filesz); } if (ph->p_filesz < ph->p_memsz) { pbzero((void *)(ph->p_vaddr + ph->p_filesz), ph->p_memsz - ph->p_filesz); } } } printf(" \n"); The loop typically iterates over 6-8 items. Of those, only 2 satisfy the "(ph->p_type == PT_LOAD)" condition -- the rest do nothing. I think it's virtually as fast as loading an a.out file, and nobody has tried to optimize it yet. The kernel exec code in "imgact_elf.c" looks more complicated. But that's only because it contains a whole bunch of debugging cruft. There's a lot of room for optimization in it. John -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Thu Jul 31 21:34:30 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id VAA25620 for current-outgoing; Thu, 31 Jul 1997 21:34:30 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [206.246.122.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id VAA25583 for ; Thu, 31 Jul 1997 21:34:25 -0700 (PDT) Received: from Journey2.mat.net (journey2.mat.net [206.246.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id AAA11109; Fri, 1 Aug 1997 00:26:37 -0400 Date: Fri, 1 Aug 1997 00:26:52 -0400 (EDT) From: Chuck Robey X-Sender: chuckr@Journey2.mat.net To: Michael Smith cc: terry@lambert.org, phk@dk.tfs.com, dg@root.com, jkh@time.cdrom.com, freebsd-current@freefall.FreeBSD.org Subject: Re: core group topics In-Reply-To: <199708010006.JAA00566@genesis.atrad.adelaide.edu.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Fri, 1 Aug 1997, Michael Smith wrote: > Chuck Robey stands accused of saying: > > > > > > Far be it for me to come off sounding anti-ELF, when I've been > > > so violently in favor of it for so long, but... I had a number > > > of large problems getting the GRUB code to actually work. > > > > I know very little about boot loaders, but I was wondering if anyone had > > seen a tool called solo, which seems to be a bootloder by Frank Barrus? > > Yup. It's another GPL'd unbuildable monstrosity 8) I spoke with him after his announcement of maybe 3 months ago, asking why he had to limit it to GPL, so that FreeBSDers couldn't consider he. I got told he'd consider it, and it looks like he did, from the most recent release licensing terms. The non-profit part was removed, just says you have to give credit in the documentation, and the source copyright has to remain. I'll try to compile it. Maybe you're right, and it's still unbuildable. > > -- > ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ > ]] Genesis Software genesis@gsoft.com.au [[ > ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ > ]] realtime instrument control. (ph) +61-8-8267-3493 [[ > ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ > > ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 213 Lakeside Drive Apt T-1 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-current Thu Jul 31 22:27:06 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA29243 for current-outgoing; Thu, 31 Jul 1997 22:27:06 -0700 (PDT) Received: from news1.gtn.com (news1.gtn.com [192.109.159.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA29234 for ; Thu, 31 Jul 1997 22:26:59 -0700 (PDT) Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) with UUCP id HAA01501 for current@freebsd.org; Fri, 1 Aug 1997 07:00:21 +0200 (MET DST) Received: (from andreas@localhost) by klemm.gtn.com (8.8.6/8.8.6) id GAA12069; Fri, 1 Aug 1997 06:56:33 +0200 (CEST) Message-ID: <19970801065632.28286@gtn.com> Date: Fri, 1 Aug 1997 06:56:32 +0200 From: Andreas Klemm To: current@freebsd.org Subject: httpd in free(): warning: modified (page-) pointer. Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Am using apache-stable on FreeBSD-2.2-STABLE and 3.0-CURRENT compiled in newest php version. On FreeBSD-2.2 everything is fine calling pgp html pages. On FreeBSD-current I get the following warning as root logged in: httpd in free(): warning: modified (page-) pointer. httpd in free(): warning: modified (page-) pointer. httpd in free(): warning: modified (page-) pointer. httpd in free(): warning: modified (page-) pointer. This message only appears, if I load a .phtml page, so I think it indicates some trouble in the php perl module I compiled in. But why only on -current ??? What to do next ??? Andreas /// -- Andreas Klemm | klemm.gtn.com - powered by Symmetric MultiProcessor FreeBSD http://www.freebsd.org/~fsmp/SMP/SMP.html http://www.freebsd.org/~fsmp/SMP/benches.html From owner-freebsd-current Thu Jul 31 23:01:41 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA01365 for current-outgoing; Thu, 31 Jul 1997 23:01:41 -0700 (PDT) Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA01339; Thu, 31 Jul 1997 23:01:30 -0700 (PDT) Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.8.6/8.6.9) id XAA18505; Thu, 31 Jul 1997 23:01:28 -0700 (PDT) Date: Thu, 31 Jul 1997 23:01:28 -0700 (PDT) Message-Id: <199708010601.XAA18505@silvia.HIP.Berkeley.EDU> To: current@freebsd.org, stable@freebsd.org CC: ports@freebsd.org Reply-to: ports@freebsd.org Subject: ports-current/packages-current discontinued From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk -----BEGIN PGP SIGNED MESSAGE----- Hi all, As I have mentioned in several places, I am now going to officially discontinue support of FreeBSD-current by the ports tree. The ports tree will now support FreeBSD-stable (the "2.2 branch"). As you know, the ports tree has always supported -current. Why change it now? Note that the "ports support -current" used to mean "ports support the next release" or "ports support the tree that is to become the next release, so we will have packages all built and ready when the release is made". It was true three years ago, when there was only one development tree (-current) that would become the next release. However, since then the tree was branched, the ports team has always been sandwiched between the need to support both -current and -stable. When the tree was initially branched, the promise was that the 2.1-stable branch was going to be short-lived and will only have one release, so we will just ship ports/packages from the previous CDROM. They will be only a few months old, and we'll move onto 2.2. Unfortunately, that didn't happen, the 2.1-stable tree lingered for almost two years until 2.2 was finally released (or 15 months since the 2.1 release), without getting any new support for ports. And we heard no end of it from the users, who upgraded from 2.1 to 2.1.5 and then 2.1.7 and still found the old moldy ports/packages in there. Or those who cvsup'd the ports tree and found that many things don't build on their system. Learning from that experience, we have tried to keep the ports tree compile both on 3.0-current and 2.2-stable when those two were branched. Many man-hours have gone into the tree for this effort, and we have succeeded to get it to work for more than half a year, despite some extensive changes in 3.0-current (like Garrett's network header updates). I have also provided a package containing the minimal set of utilities (the "2.2-stable upgrade kit") for 2.2.1R and 2.2.2R users so they can use the latest ports without even tracking 2.2-stable. Another thing I have done was to merge necessary changes from 3.0-current to 2.2-stable (e.g., install-info) to make the two systems reasonably close to each other. However, the recent rash of commits to the 3.0 branch without regards of consistency and compatibility with 2.2-stable made it necessary to make a decision. Also, it is often the case that -current doesn't even compile, which made it difficult for me to maintain a reasonable build machine. As we will have (at least) one more release from the 2.2-stable branch, it doesn't really make sense for us to follow the circus going on in 3.0-current now. We will make sure everything compiles and works in 2.2-stable, so when the next release comes out, we'll have a functional set of packgaes. (Also, people following -current are generally more capable of building stuff themselves. :) Note that the above doesn't mean we are not going to discontinue "#if __FreeBSD_version > ..." style of patches -- compatible porting in that manner has always been encouraged, and will continue to be so. (That will make it easier when we move over to that branch.) It just means that when it becomes impossible to support both branches without severe loads to our already scarce manpower and/or disgusting hacks, we will choose to support -stable. As many you have undoubtedly noticed, I am in the process of building packages for 2.2-stable and copying them to ftp.freebsd.org's new packages-stable directory. I will also move the ports tree on the ftp site from under "FreeBSD-current" to somewhere else to avoid confusion, and remove the "ports-current" symbolic link. The packages-current directory will eventually be deleted too (unless some kind soul steps forward to build them for me). This is truly a sad day. The irony of this all is that it really didn't have to happen, and it would have been possible for us to support both 3.0-current and 2.2-stable if more consideration has been made before some commits into -current. For instance, there is no reason why tcl-8.0beta2 had to go into the tree now, whether the tree is called -current or not. It is dubbed BETA of all things on the master site, and we haven't even had it in our ports area to test it. As some people have noted, there is no matching tk version in our ports tree either. The latest incident just reinforces my belief that it was a mistake in the first place to include something like tcl in the base distribution (note that it was also done without consensus back then), and one we may regret for a long time to come. Yours truly, Satoshi and the awesome Ports Team -----BEGIN PGP SIGNATURE----- Version: 2.6.2 iQCVAwUBM+F7mINA0SoeCNiJAQFxCgP/ZeTbsaIXqCgPOvBImJa3FOQdUjouo6IM jmbsoYqxgJc6bYZOPDxvyRWK2ne845FjHywSyssCNfQU82qgg2Vdot0JwXoKTGm/ 5x5ZG6nHCgRjXVEsEC6SqI3Aw9iwA0vgD+dRB5LLVhCMs/BbA5ejkSyxgq+RITKG LzQ6koCTHRg= =PHTA -----END PGP SIGNATURE----- From owner-freebsd-current Thu Jul 31 23:24:09 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA02967 for current-outgoing; Thu, 31 Jul 1997 23:24:09 -0700 (PDT) Received: from skynet.ctr.columbia.edu (skynet.ctr.columbia.edu [128.59.64.70]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id XAA02961 for ; Thu, 31 Jul 1997 23:24:03 -0700 (PDT) Received: (from wpaul@localhost) by skynet.ctr.columbia.edu (8.6.12/8.6.9) id CAA12780; Fri, 1 Aug 1997 02:24:09 -0400 From: Bill Paul Message-Id: <199708010624.CAA12780@skynet.ctr.columbia.edu> Subject: Re: httpd in free(): warning: modified (page-) pointer. To: andreas@klemm.gtn.com (Andreas Klemm) Date: Fri, 1 Aug 1997 02:24:07 -0400 (EDT) Cc: current@freebsd.org In-Reply-To: <19970801065632.28286@gtn.com> from "Andreas Klemm" at Aug 1, 97 06:56:32 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Of all the gin joints in all the towns in all the world, Andreas Klemm had to walk into mine and say: > Am using apache-stable on FreeBSD-2.2-STABLE and 3.0-CURRENT > > compiled in newest php version. > > On FreeBSD-2.2 everything is fine calling pgp html pages. > On FreeBSD-current I get the following warning as root > logged in: > > httpd in free(): warning: modified (page-) pointer. > httpd in free(): warning: modified (page-) pointer. > httpd in free(): warning: modified (page-) pointer. > httpd in free(): warning: modified (page-) pointer. > > This message only appears, if I load a .phtml page, > so I think it indicates some trouble in the php perl > module I compiled in. Between apache and perl, I think it would be an incredible coincidence if there wasn't a bug in the code somewhere. > But why only on -current ??? What to do next ??? The fact that malloc()/free() only complains on FreeBSD-current does not mean that the bug isn't also present on FreeBSD-2.2.x. It's there too: you just don't see it. It may be that it only shows up on current due to some change in the malloc code. However, there is most certainly a bug in the pgp module you compiled into apache (or there's a bug in apache which is triggered by the extra module). If the code is supported by someone, write them a note and tell them it's buggy. There are a couple of things you can do to try tracking down the problem: - Grab the malloc.c module from the libc source and compile it directly into apache. Build the whole binary with -g. (This will override the malloc in libc, which is not compiled with debug symbols.) Run apache with gdb and set a breakpoint in the malloc code where this error message occurs. Then reproduce the condition that causes the error; gdb will stop the server at the breakpoint and you can get a function call trace which will hopefully show you the error. - Compile apache with the Electric Fence 2.0.5 debugging malloc library. Compile with -g, as before, link with -lefence and run the server normally. The odds are that the problem here is heap corruption; with Electric Fence, the instant you stray out of bounds of a buffer allocated with malloc(), the program will get smacked with a SIGBUS and die, leaving a core dump behind. The core can be used with gdb to locate the offending code. Caveats: o With our gdb, it is sometimes difficult to use a core dump to properly debug things: you will be able to get a stack trace and locate the exact line in the source where the program died, however not all of the environment is preserved: variables are not properly reloaded with the values they posessed when the crash took place. If the program crashes while being run in gdb, then you can examine variables and memory contents with no problems, but trying to do it with just the crash dump and the executable doesn't seem to work. (It seems to work with kernel crash dumps though.) I wish it did; it would make my life a lot easier. o Electric Fence works by using mmap() to allocate memory instead of brk()/sbrk(). To trap out of bounds references, it allocates an extra page of memory immediately after the memory requested with malloc()/calloc()/etc and uses mprotect() to disallow read and write access to it. In this way, you use the MMU itself to signal bad accesses: the instant an instruction is executed that tries to reference the protected page, a trap is generated and your program dies with a SIGBUS. If the executable and libefence.a are both compiled with -g, you can then use gdb to locate the exact line of source where the invalid reference took place. The downside to this is that the malloc() in libefence.a takes longer to work and uses more memory than normal. If you can trip the bug right away, then this probably won't pose any problems, but it can get a bit clumsy with programs that run for a long time or allocate lots of memory. Before you ask, Electric Fence is available at ftp.pixar.com:/pub/bruce. You need to fiddle with it slightly to compile it on FreeBSD, but in general I find it works quite well. I've been using it while working on my NIS+ code and once or twice on ypserv, and it's helped a lot. More people should make use of it. -Bill -- ============================================================================= -Bill Paul (212) 854-6020 | System Manager, Master of Unix-Fu Work: wpaul@ctr.columbia.edu | Center for Telecommunications Research Home: wpaul@skynet.ctr.columbia.edu | Columbia University, New York City ============================================================================= "It is not I who am crazy; it is I who am mad!" - Ren Hoek, "Space Madness" ============================================================================= From owner-freebsd-current Fri Aug 1 03:06:45 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA12609 for current-outgoing; Fri, 1 Aug 1997 03:06:45 -0700 (PDT) Received: from rah.star-gate.com (rah.star-gate.com [204.188.121.18]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA12601 for ; Fri, 1 Aug 1997 03:06:41 -0700 (PDT) Received: from rah.star-gate.com (localhost.star-gate.com [127.0.0.1]) by rah.star-gate.com (8.8.5/8.8.5) with ESMTP id DAA03932 for ; Fri, 1 Aug 1997 03:06:40 -0700 (PDT) Message-Id: <199708011006.DAA03932@rah.star-gate.com> X-Mailer: exmh version 2.0gamma 1/27/96 To: current@FreeBSD.ORG Subject: second post: elf patches for -current Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 01 Aug 1997 03:06:40 -0700 From: Amancio Hasty Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk I downloaded the elfkit from John's ftp site and noticed that the README file was dated from 1996/04/16 and that it requires the include files from FreeBSD 2.1 . Doesn't make sense that we should patches for freebsd -current that does not require the include files from 2.1? Tnks, Amancio From owner-freebsd-current Fri Aug 1 04:46:32 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA15506 for current-outgoing; Fri, 1 Aug 1997 04:46:32 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA15496 for ; Fri, 1 Aug 1997 04:46:25 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id VAA07673; Fri, 1 Aug 1997 21:39:41 +1000 Date: Fri, 1 Aug 1997 21:39:41 +1000 From: Bruce Evans Message-Id: <199708011139.VAA07673@godzilla.zeta.org.au> To: bde@zeta.org.au, reg@shale.csir.co.za Subject: Re: More bogons in Makefiles... Cc: current@freebsd.org Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >> with a per-source directory obj symlink)). Non sloppy makefiles would >> use a more complicated search for the object directory that would >> actually be used if `make' were run in the external source directory. > >An example being? ;) I don't know of any non-sloppy ones. `make''s object directory search path is too complicated to duplicate in dozens of Makefiles. `make -V .OBJDIR' was supposed to be useful for getting `make' to do the search, but it has various problems including endless recursion and/or unacceptable speed when used in complicated and/or large Makefiles. >> There are more bogons here. The whole ifdef is a verbose and confusing >> way write `YEARISTYPE= ${.OBJDIR}/yearistype. yearistype is always >> created in the object directory and ${.OBJDIR} always exists (although >> it's not clear where it is when make is invoked with args `obj all ...'). > >>From this I read ${.OBJDIR} == ${.CURDIR} iff no 'make obj'. If that is No, ${.OBJDIR} is the current directory for building sources. yearistype is always built there. ${.CURDIR} is the current directory when `make' is started. yearistype is only built in ${.CURDIR} if ${.CURDIR} happens to be the same directory as ${.OBJDIR} (I think it can be a different path in some cases involving symlinks). >correct then there is a lot of funny business going on, because there are a >lot of checks this form. Maybe. >> Because I don't use it :-). bsd.libnames.mk is no longer used in -current, >> so this problem will go away automatically. > >Looks suspiciously like bsd.lib.mk includes this file... but now I'm lost ;) >where do all the other defines for libraries come from? Nowhere. Dependencies are generated by the _EXTRADEPEND target, and ${DPADD} is not used (except in bsd.kmod.mk), so none of the defines for libraries are used. Bruce From owner-freebsd-current Fri Aug 1 07:45:26 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA23255 for current-outgoing; Fri, 1 Aug 1997 07:45:26 -0700 (PDT) Received: from mail11.digital.com (mail11.digital.com [192.208.46.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA23250 for ; Fri, 1 Aug 1997 07:45:22 -0700 (PDT) From: garyj@frt.dec.com Received: from cssmuc.frt.dec.com (cssmuc.frt.dec.com [16.186.96.161]) by mail11.digital.com (8.7.5/UNX 1.5/1.0/WV) with SMTP id KAA13399 for ; Fri, 1 Aug 1997 10:39:40 -0400 (EDT) Received: from localhost by cssmuc.frt.dec.com; (5.65v3.2/1.1.8.2/14Nov95-0232PM) id AA13051; Fri, 1 Aug 1997 16:39:30 +0200 Message-Id: <9708011439.AA13051@cssmuc.frt.dec.com> X-Mailer: exmh version 1.6.4 10/10/95 To: current@freebsd.org In-Reply-To: Message from Bill Paul of Fri, 01 Aug 97 02:24:07 EDT. Reply-To: gjennejohn@frt.dec.com Subject: Re: httpd in free(): warning: modified (page-) pointer. Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 01 Aug 97 16:39:30 +0200 X-Mts: smtp Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk wpaul@skynet.ctr.columbia.edu writes: > Caveats: > > o With our gdb, it is sometimes difficult to use a core dump > to properly debug things: you will be able to get a stack > trace and locate the exact line in the source where the program > died, however not all of the environment is preserved: variables > are not properly reloaded with the values they posessed when > the crash took place. If the program crashes while being run > in gdb, then you can examine variables and memory contents > with no problems, but trying to do it with just the crash dump > and the executable doesn't seem to work. (It seems to work with > kernel crash dumps though.) I wish it did; it would make my life > a lot easier. > has it always had this problem or did it start after the changes in -current to eliminate the user structure ? --- Gary Jennejohn (work) gjennejohn@frt.dec.com (home) Gary.Jennejohn@munich.netsurf.de (play) gj@freebsd.org From owner-freebsd-current Fri Aug 1 09:16:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id JAA27277 for current-outgoing; Fri, 1 Aug 1997 09:16:43 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id JAA27270 for ; Fri, 1 Aug 1997 09:16:35 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id CAA16320; Sat, 2 Aug 1997 02:13:34 +1000 Date: Sat, 2 Aug 1997 02:13:34 +1000 From: Bruce Evans Message-Id: <199708011613.CAA16320@godzilla.zeta.org.au> To: current@FreeBSD.ORG, garyj@frt.dec.com Subject: Re: httpd in free(): warning: modified (page-) pointer. Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >wpaul@skynet.ctr.columbia.edu writes: >> Caveats: >> >> o With our gdb, it is sometimes difficult to use a core dump >> to properly debug things: you will be able to get a stack >> trace and locate the exact line in the source where the program >> died, however not all of the environment is preserved: variables >> are not properly reloaded with the values they posessed when >> the crash took place. If the program crashes while being run >... >has it always had this problem or did it start after the changes >in -current to eliminate the user structure ? This is probably the bug that started in FreeBSD-2.0 and was fixed a while ago (4.4BSDLite doesn't write vm_daddr to cores, and gdb was configured to use vm_daddr). Bruce From owner-freebsd-current Fri Aug 1 10:23:56 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA00744 for current-outgoing; Fri, 1 Aug 1997 10:23:56 -0700 (PDT) Received: from news1.gtn.com (news1.gtn.com [192.109.159.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA00739 for ; Fri, 1 Aug 1997 10:23:52 -0700 (PDT) Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) with UUCP id RAA27945; Fri, 1 Aug 1997 17:00:19 +0200 (MET DST) Received: (from andreas@localhost) by klemm.gtn.com (8.8.6/8.8.6) id QAA18420; Fri, 1 Aug 1997 16:48:31 +0200 (CEST) Message-ID: <19970801164831.09171@klemm.gtn.com> Date: Fri, 1 Aug 1997 16:48:31 +0200 From: Andreas Klemm To: Bill Paul Cc: current@freebsd.org Subject: Re: httpd in free(): warning: modified (page-) pointer. References: <19970801065632.28286@gtn.com> <199708010624.CAA12780@skynet.ctr.columbia.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 In-Reply-To: <199708010624.CAA12780@skynet.ctr.columbia.edu>; from Bill Paul on Fri, Aug 01, 1997 at 02:24:07AM -0400 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Ok, thanks [...] > If the code is supported by someone, write them a note and tell them > it's buggy. There are a couple of things you can do to try tracking > down the problem: [...] -- Andreas Klemm | klemm.gtn.com - powered by Symmetric MultiProcessor FreeBSD http://www.freebsd.org/~fsmp/SMP/SMP.html http://www.freebsd.org/~fsmp/SMP/benches.html From owner-freebsd-current Fri Aug 1 10:35:38 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA01355 for current-outgoing; Fri, 1 Aug 1997 10:35:38 -0700 (PDT) Received: from news1.gtn.com (news1.gtn.com [194.77.0.15]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA01336; Fri, 1 Aug 1997 10:35:18 -0700 (PDT) Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) with UUCP id SAA19408; Fri, 1 Aug 1997 18:30:21 +0200 (MET DST) Received: (from andreas@localhost) by klemm.gtn.com (8.8.6/8.8.6) id SAA07707; Fri, 1 Aug 1997 18:24:49 +0200 (CEST) Message-ID: <19970801182448.26268@klemm.gtn.com> Date: Fri, 1 Aug 1997 18:24:48 +0200 From: Andreas Klemm To: ports@freebsd.org Cc: current@freebsd.org, stable@freebsd.org Subject: Re: ports-current/packages-current discontinued References: <199708010601.XAA18505@silvia.HIP.Berkeley.EDU> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Mutt 0.79 In-Reply-To: <199708010601.XAA18505@silvia.HIP.Berkeley.EDU>; from Satoshi Asami on Thu, Jul 31, 1997 at 11:01:28PM -0700 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk On Thu, Jul 31, 1997 at 11:01:28PM -0700, Satoshi Asami wrote: > We will make sure everything compiles and > works in 2.2-stable, so when the next release comes out, we'll have a > functional set of packgaes. (Also, people following -current are > generally more capable of building stuff themselves. :) But unable to build really working 2.2 packages, because they only have -current :-/ > This is truly a sad day. The irony of this all is that it really > didn't have to happen, and it would have been possible for us to > support both 3.0-current and 2.2-stable if more consideration has been > made before some commits into -current. Well, wouldn╢t it make more sense to discuss, to back out the changes that introduces these problems ? Very very sad is, that in my eyes tcl 8.x brings us really _nothing_ from the Operating systems view, but finally it breaks the whole concept, that -current ports can be used in 99% of cases in 2.2 and -current. We want to improve the Operating System -> that╢s FreeBSD, not compete with Linux to pickup the last bits and bytes, that break something. Why was TCL not brought into the ports collection ??? We already have several TCL versions there for backward compatibility ... why not put the newest stuff into there with respect to the problems. that now arise ? Generally we had the policy, to use the ports collection for new stuff, that belongs to FreeBSD. I think it╢s time to re-think, what really belongs into the operating system and what not ... Wouln╢t it be better to come to a decision, that such things like perl and tcl, _if_ they are needed in the base system, should be of the same version in -current and -stable ?! > For instance, there is no reason why tcl-8.0beta2 had to go > into the tree now, whether the tree is called -current or not. I fully agree, if I read now the big _disadvantages_ > It is dubbed BETA of all things on the master site, and we haven't > even had it in our ports area to test it. Then I╢d vote "back it out!" ;-) > As some people have noted, there is no matching tk version in our > ports tree either. The latest incident just reinforces my belief that > it was a mistake in the first place to include something like tcl in > the base distribution (note that it was also done without consensus > back then), and one we may regret for a long time to come. I think it╢s time to "purify" our sources. I think it╢s a bad idea, to have only a ports collection based on -STABLE. Ports should at least run with -STABLE and -CURRENT. I only run -current, create ports myself and I╢m unable to build up a -STABLE system to test my ports for 2.2-STABLE systems. And I don╢t want to go back to -STABLE, because I have a SMP system and like the bleeding edge more ... I╢d like to say "peope, let╢s try to find a way out of this mess" .... Andreas /// -- Andreas Klemm | klemm.gtn.com - powered by Symmetric MultiProcessor FreeBSD http://www.freebsd.org/~fsmp/SMP/SMP.html http://www.freebsd.org/~fsmp/SMP/benches.html From owner-freebsd-current Fri Aug 1 11:27:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA04351 for current-outgoing; Fri, 1 Aug 1997 11:27:22 -0700 (PDT) Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA04346 for ; Fri, 1 Aug 1997 11:27:17 -0700 (PDT) Received: from hurricane.cs.duke.edu (hurricane.cs.duke.edu [152.3.145.1]) by duke.cs.duke.edu (8.8.5/8.8.5) with ESMTP id OAA24025; Fri, 1 Aug 1997 14:27:14 -0400 (EDT) Received: (from gallatin@localhost) by hurricane.cs.duke.edu (8.8.4/8.7.3) id OAA27698; Fri, 1 Aug 1997 14:27:14 -0400 (EDT) Date: Fri, 1 Aug 1997 14:27:14 -0400 (EDT) Message-Id: <199708011827.OAA27698@hurricane.cs.duke.edu> From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit To: Chris Csanady Cc: freebsd-current@freebsd.org Subject: Re: Myrinet, etc.. (Re: code talks: announcing EIDE bus master patches) In-Reply-To: <199708010252.VAA15322@friley01.res.iastate.edu> References: <199707312354.TAA29288@hurricane.cs.duke.edu> <199708010252.VAA15322@friley01.res.iastate.edu> X-Mailer: VM 6.32 under 19.15 XEmacs Lucid Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Chris Csanady writes: > > > > I'd also like to point out that the 300Mb's is acheived using an IP > > > stack layered on top of some active messaging protocol. (which is > > > implemented on the io processor on the nic.) > > > > > > As far as the TCP/IP stack under FreeBSD, all you can push through it > > > is about 150Mb/s. This is somewhat unfortunate, although surprisingly > > > linux doesn't seem to manage much better. I think it would be quite > > > nice if we could correctly implement a zero copy architecture.. > > > > > > Chris Csanady > > > >Actually, that's not true. I'm using an in-kernel IP driver layered > >on top of Myricom's MyriApi general-purpose messaging software (which > >is kernel resident). It hooks into the network stack the way any > >ethernet driver would (and gets hit with a copy for each xmit from or > >receive into user space like anything else does). There is software > >that's downloaded onto the interface card, but it has no protocol > >specific knowledge. When its sitting under the IP driver, it is used > >the same way the firwmware on any ethernet card would be used -- the > >driver tells it a MAC address that it wants something sent to, and it > >sends it, etc. > > Sorry, I guess I assumed wrong. Me too.. I hope I didn't over-react. > However, the numbers you are > reporting are vastly different from he ones mentioned on myrinet's > web page for 2.2.1. Yeah.. the UDP numbers on Myri's page are a little out of date, plus the TCP numbers are from a benchmark run w/TCP send & receive space at their 16k defaults. The b/w nearly doubles if you crank them up to 64k. Also, my 356Mb/sec is taken when receiving from an alpha. I cannot transmit that fast with an Intel box yet. > >The fact that we're running over Myrinet & not ethernet is transparent > >to any application. In this configuration, with the hardware I > >currently have available (memory b/w challenged Pentium Pros) I can > >receive UDP traffic at about 356 Mb/sec and send UDP traffic at a rate > >of a about 280Mb/sec. TCP streams are around 275Mb/sec. > > That really is quite impressive--is your driver available? I would > like to play with it when I get a chance. :) I'm also quite > interested in how your interrupt/buffer architecture looks. Sure. There are pointers to both the distribution tarball & a CVS-web browsable interface at http://www.cs.duke.edu/ari/manic/ip/ > As for memory bandwith limitations, I can't wait for the alpha > port. ;) > > Chris Csanady In terms of memory b/w, the alphas are ... "interesting". The older 21171 core-logic chipset alphas (AlphaStation 500, AlphaStation 600, EB164, PC164) leave a bit to be desired, especially when both the CPU and a gigabit Myrinet interface are competing for the memory system. The newer 21174 core-logic chipset alphas (Personal Workstation series, PC164lx) are much better. See http://www.cs.duke.edu/ari/manic/ip/dma-perf/dma-bcopy.html for some graphs that illustrate this point. Drew ------------------------------------------------------------------------------ Andrew Gallatin http://www.cs.duke.edu/~gallatin Duke University Email: gallatin@cs.duke.edu Department of Computer Science Phone: (919) 660-6590 From owner-freebsd-current Fri Aug 1 11:35:28 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA04714 for current-outgoing; Fri, 1 Aug 1997 11:35:28 -0700 (PDT) Received: from mailhost.anasazi.com (mailhost.anasazi.com [138.113.128.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id LAA04679; Fri, 1 Aug 1997 11:34:56 -0700 (PDT) Received: from chad.anasazi.com by mailhost.anasazi.com (5.65/3.7b) id AA26033; Fri, 1 Aug 97 11:34:06 -0700 Received: by chad.anasazi.com (5.65/3.7) id AA16893; Fri, 1 Aug 97 11:34:03 -0700 From: chad@anasazi.com (Chad R. Larson) Message-Id: <9708011834.AA16893@chad.anasazi.com> Subject: Re: ports-current/packages-current discontinued To: andreas@klemm.gtn.com (Andreas Klemm) Date: Fri, 1 Aug 1997 11:34:02 -0700 (MST) Cc: ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG In-Reply-To: <19970801182448.26268@klemm.gtn.com> from "Andreas Klemm" at Aug 1, 97 06:24:48 pm Reply-To: chad@anasazi.com X-Mailer: ELM [version 2.4 PL23] Content-Type: text Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > On Thu, Jul 31, 1997 at 11:01:28PM -0700, Satoshi Asami wrote: > > We will make sure everything compiles and > > works in 2.2-stable, so when the next release comes out, we'll have a > > functional set of packgaes. (Also, people following -current are > > generally more capable of building stuff themselves. :) > > > > [reference to TCL/TK deleted] > > It is dubbed BETA of all things on the master site, and we haven't > > even had it in our ports area to test it. > > Then I=B4d vote "back it out!" ;-) I'd second that. I'd like to see the "Release" of FreeBSD to contain the core system. User contributed stuff should all be integrated through the ports process. -crl -- Chad R. Larson (CRL22) Brother, can you paradigm? 602-870-3330 chad@anasazi.com chad@anasaz.UUCP chad@dcfinc.com Anasazi, Inc. - 7500 North Dreamy Draw Drive, Suite 120, Phoenix, Az 85020 From owner-freebsd-current Fri Aug 1 15:25:20 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA15710 for current-outgoing; Fri, 1 Aug 1997 15:25:20 -0700 (PDT) Received: from skynet.ctr.columbia.edu (skynet.ctr.columbia.edu [128.59.64.70]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id PAA15698 for ; Fri, 1 Aug 1997 15:25:14 -0700 (PDT) Received: (from wpaul@localhost) by skynet.ctr.columbia.edu (8.6.12/8.6.9) id SAA13792; Fri, 1 Aug 1997 18:25:19 -0400 From: Bill Paul Message-Id: <199708012225.SAA13792@skynet.ctr.columbia.edu> Subject: Re: httpd in free(): warning: modified (page-) pointer. To: bob@luke.pmr.com Date: Fri, 1 Aug 1997 18:25:17 -0400 (EDT) Cc: current@freebsd.org In-Reply-To: <19970801082351.40895@pmr.com> from "Bob Willcox" at Aug 1, 97 08:23:51 am X-Mailer: ELM [version 2.4 PL24] Content-Type: text Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Of all the gin joints in all the towns in all the world, Bob Willcox had to walk into mine and say: > On Fri, Aug 01, 1997 at 02:24:07AM -0400, Bill Paul wrote: > > > > Before you ask, Electric Fence is available at ftp.pixar.com:/pub/bruce. > > Do you know of somewhere else that it might reside? From the looks > of it, /pub/bruce no longer exists on ftp.pixar.com. :-( Whups. Hadn't checked there in a while. There are copies on all the Linux sites. It's also mirrored on ftp.cdrom.com (everyone's favorite FTP server) under /pub/linux/sunsite/devel/lang/c. -Bill -- ============================================================================= -Bill Paul (212) 854-6020 | System Manager, Master of Unix-Fu Work: wpaul@ctr.columbia.edu | Center for Telecommunications Research Home: wpaul@skynet.ctr.columbia.edu | Columbia University, New York City ============================================================================= "It is not I who am crazy; it is I who am mad!" - Ren Hoek, "Space Madness" ============================================================================= From owner-freebsd-current Fri Aug 1 16:04:13 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA17291 for current-outgoing; Fri, 1 Aug 1997 16:04:13 -0700 (PDT) Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id QAA17230; Fri, 1 Aug 1997 16:04:00 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr3-8.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA04688 (5.67b/IDA-1.5); Sat, 2 Aug 1997 01:03:56 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.6/8.6.9) id AAA00557; Sat, 2 Aug 1997 00:29:09 +0200 (CEST) X-Face: " Date: Sat, 2 Aug 1997 00:27:48 +0200 From: Stefan Esser To: Andrew Gallatin Cc: freebsd-current@FreeBSD.ORG, Stefan Esser Subject: Re: code talks: announcing EIDE bus master patches References: <19970729210723.18104@mi.uni-koeln.de> <199707291949.VAA00271@sos.freebsd.dk> <19970729221036.52544@mi.uni-koeln.de> <199707301924.PAA08936@hurricane.cs.duke.edu> <19970730220038.02422@mi.uni-koeln.de> <199707302036.QAA23330@hurricane.cs.duke.edu> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.74 In-Reply-To: <199707302036.QAA23330@hurricane.cs.duke.edu>; from Andrew Gallatin on Wed, Jul 30, 1997 at 04:36:30PM -0400 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Jul 30, Andrew Gallatin wrote: > > But in any case: In order to use PCI burst transfers, you > > have to write to sequential memory addresses, e.g. to some > > double-mapped SRAM buffer. > > Double-mapped? What does that mean? Sorry, it was too late in then night, and I just could not find the correct word. What I meant was "dual-ported", but I think you already guessed that ... > > That's *very* different from an I/O port, and you should be able > > to copy data to that SRAM at a much higher rate than 8 or 13MB/s! > > Right. The Natoma is using burst transfers for PIO writes and getting > nearly 30MB/sec. Its apparently not bursting the reads, and is > getting only 8MB/sec. How do you get an Intel to burst reads? I > didn't think they could support multiple outstanding reads. I know > that alphas do, and there are apparently tricks you can use to make an > alpha burst reads & get 50MB/sec bandwidth, but I wasn't aware of > Intels that could get better than 15MB/sec for PIO reads. Yes, AFAIK there is no Intel (or other P5 or PPro compatible) chip set, that supports burst reads. Burst writes are easy (the FIFO in the host to PCI bridge checks that sequential addresses are written to, and can use the byte enables if not thea full DWORD is to be stored), but the host to PCI bridge had to perform a read-ahead for read bursts ... Most PCI bus-master controllers do not only know how to initiate burst transfers in either direction, they will also use read-line and read-multiple (and write-and-invalidate) where appropriate to reduce latencies (when crossing PCI bridges) and impact on the CPU. > I'm getting most of my information from emperical results obtained > with our P6's and Alphas using Myrinet cards, plus the DEC PCI Pamette > paper ("Systems Performance Measurement on PCI Pamette", Laurent Moll, > Mark Shand, Fccm97, ftp://pam.devinci.fr/pub/pamette/fccm97.ps.gz). Regards, STefan From owner-freebsd-current Fri Aug 1 16:04:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA17301 for current-outgoing; Fri, 1 Aug 1997 16:04:16 -0700 (PDT) Received: from Octopussy.MI.Uni-Koeln.DE (Octopussy.MI.Uni-Koeln.DE [134.95.166.20]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id QAA17289; Fri, 1 Aug 1997 16:04:09 -0700 (PDT) Received: from x14.mi.uni-koeln.de (annexr3-8.slip.Uni-Koeln.DE) by Octopussy.MI.Uni-Koeln.DE with SMTP id AA04693 (5.67b/IDA-1.5); Sat, 2 Aug 1997 01:04:01 +0200 Received: (from se@localhost) by x14.mi.uni-koeln.de (8.8.6/8.6.9) id AAA00617; Sat, 2 Aug 1997 00:44:18 +0200 (CEST) X-Face: " Date: Sat, 2 Aug 1997 00:42:56 +0200 From: Stefan Esser To: Michael Smith Cc: gallatin@cs.duke.edu, freebsd-current@FreeBSD.ORG, Stefan Esser Subject: Re: code talks: announcing EIDE bus master patches References: <19970730220038.02422@mi.uni-koeln.de> <199707310317.MAA25355@genesis.atrad.adelaide.edu.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.74 In-Reply-To: <199707310317.MAA25355@genesis.atrad.adelaide.edu.au>; from Michael Smith on Thu, Jul 31, 1997 at 12:47:07PM +0930 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Jul 31, Michael Smith wrote: > If I read the AMCC PCI matchmaker databook correctly, PIO to a > memory-mapped buffer is likely to be severely affected by whether the > PCI bridge in question supports cache line read/write operations. CPU writes to a PCI device are (at least in the case of x86 CPUs) always issued independently, and only the host to PCI bridge chip will see requests to sequentially increasing memory addresses and in that case just keep on sending data with no intervening PCI address cycles. This feature can be enabled in all host to PCI bridges since at least the Intel Saturn (introduced some 4 years ago), though it did not yet work reliably in all isituations, then. Today it is generally enabled and working. > Is this a potential issue here? No. No chipset (that I know of) performs read-ahead on CPU reads, even to PCI bus address ranges that have been marked prefetchable. But you can use a bus-master PCI chip (e.g. an NCR PCI SCSI chips) as a DMA controller, and will then be able issue burst transfers in any direction. A 53c810A does make for a cheap and fast DMA controller, too bad that it occupies another one of those scarce PCI slots :) Regards, STefan From owner-freebsd-current Fri Aug 1 18:15:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA23570 for current-outgoing; Fri, 1 Aug 1997 18:15:18 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA23565; Fri, 1 Aug 1997 18:15:12 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id KAA15728; Sat, 2 Aug 1997 10:33:30 +1000 (EST) Message-Id: <199708020033.KAA15728@unique.usn.blaze.net.au> To: committers@freebsd.org Cc: freebsd-current@freebsd.org Subject: rc.shutdown X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sat, 02 Aug 1997 10:33:30 +1000 From: David Nugent Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Just a heads-up on the just-added /etc/rc.shutdown support. To avoid a complaint from init(8) when shutting down, you'll need to manually copy src/etc/rc.shutdown to /etc after building world. Regards, David From owner-freebsd-current Fri Aug 1 18:46:20 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA24792 for current-outgoing; Fri, 1 Aug 1997 18:46:20 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA24787 for ; Fri, 1 Aug 1997 18:46:17 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id SAA09667; Fri, 1 Aug 1997 18:46:11 -0700 (PDT) Message-Id: <199708020146.SAA09667@austin.polstra.com> To: hasty@rah.star-gate.com Subject: Re: second post: elf patches for -current In-Reply-To: <199708011006.DAA03932@rah.star-gate.com> References: <199708011006.DAA03932@rah.star-gate.com> Organization: Polstra & Co., Seattle, WA Cc: current@freebsd.org Date: Fri, 01 Aug 1997 18:46:10 -0700 From: John Polstra Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In article <199708011006.DAA03932@rah.star-gate.com>, Amancio Hasty wrote: > > I downloaded the elfkit from John's ftp site and noticed that the > README file was dated from 1996/04/16 and that it requires the > include files from FreeBSD 2.1 . > > Doesn't make sense that we should patches for freebsd -current that > does not require the include files from 2.1? ElfKit is almost obsolete, because just about everything you need is now in -current. You can build the FreeBSD compiler for ELF by specifying "BINFORMAT=elf". The standard libc in -current (and I think in -2.2 also) can be built for ELF, and it works just fine. You don't have to do anything special; just compile it with the ELF tools. The current version of GNU binutils has all the FreeBSD-ELF support in it. Just configure it like this and then build it: ./configure --prefix=/usr/local/elf --host=i386-unknown-freebsd \ --target=i386-unknown-freebsdelf John -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Fri Aug 1 19:21:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA26230 for current-outgoing; Fri, 1 Aug 1997 19:21:03 -0700 (PDT) Received: from rah.star-gate.com (rah.star-gate.com [204.188.121.18]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA26225 for ; Fri, 1 Aug 1997 19:21:00 -0700 (PDT) Received: from rah.star-gate.com (localhost.star-gate.com [127.0.0.1]) by rah.star-gate.com (8.8.5/8.8.5) with ESMTP id TAA07109; Fri, 1 Aug 1997 19:20:54 -0700 (PDT) Message-Id: <199708020220.TAA07109@rah.star-gate.com> X-Mailer: exmh version 2.0gamma 1/27/96 To: John Polstra cc: current@freebsd.org Subject: Re: second post: elf patches for -current In-reply-to: Your message of "Fri, 01 Aug 1997 18:46:10 PDT." <199708020146.SAA09667@austin.polstra.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Fri, 01 Aug 1997 19:20:54 -0700 From: Amancio Hasty Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Tnks I will give it a try. Really, interested on this stuff for large C++ packages such as ACE. I suggest that you add a little README.current to your elfkit directory mostly because I think thats where people will go for information about elf on FreeBSD at any rate thats what I did 8) Tnks again, Amancio >From The Desk Of John Polstra : > In article <199708011006.DAA03932@rah.star-gate.com>, > Amancio Hasty wrote: > > > > I downloaded the elfkit from John's ftp site and noticed that the > > README file was dated from 1996/04/16 and that it requires the > > include files from FreeBSD 2.1 . > > > > Doesn't make sense that we should patches for freebsd -current that > > does not require the include files from 2.1? > > ElfKit is almost obsolete, because just about everything you need > is now in -current. You can build the FreeBSD compiler for ELF by > specifying "BINFORMAT=elf". The standard libc in -current (and I > think in -2.2 also) can be built for ELF, and it works just fine. > You don't have to do anything special; just compile it with the > ELF tools. > > The current version of GNU binutils has all the FreeBSD-ELF support > in it. Just configure it like this and then build it: > > ./configure --prefix=/usr/local/elf --host=i386-unknown-freebsd \ > --target=i386-unknown-freebsdelf > > John > -- > John Polstra jdp@polstra.com > John D. Polstra & Co., Inc. Seattle, Washington USA > "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Fri Aug 1 22:29:20 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA03055 for current-outgoing; Fri, 1 Aug 1997 22:29:20 -0700 (PDT) Received: from mail.webspan.net (root@mail.webspan.net [206.154.70.7]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA03049 for ; Fri, 1 Aug 1997 22:29:17 -0700 (PDT) Received: from orion.webspan.net (orion.webspan.net [206.154.70.5]) by mail.webspan.net (WEBSPAN/970608) with ESMTP id BAA01848; Sat, 2 Aug 1997 01:29:16 -0400 (EDT) Received: from orion.webspan.net (localhost [127.0.0.1]) by orion.webspan.net (WEBSPAN/970608) with ESMTP id BAA26504; Sat, 2 Aug 1997 01:29:10 -0400 (EDT) To: dg@root.com cc: Poul-Henning Kamp , "Jordan K. Hubbard" , Chuck Robey , FreeBSD current From: "Gary Palmer" Subject: Re: core group topics In-reply-to: Your message of "Thu, 31 Jul 1997 04:42:32 PDT." <199707311142.EAA28471@implode.root.com> Date: Sat, 02 Aug 1997 01:29:10 -0400 Message-ID: <26502.870499750@orion.webspan.net> Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk David Greenman wrote in message ID <199707311142.EAA28471@implode.root.com>: > There is? I don't recall discussing that at all, and if we do > eventually go the ELF way, I think the kernel would be one of the > last areas we'd want to convert since it further complicates our > cramped bootblocks. All the more reason to go to 3 stage booting. It seems to have been one of these `nice to have, maybe someday' things like a fully working DEVFS. Gary -- Gary Palmer FreeBSD Core Team Member FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info From owner-freebsd-current Fri Aug 1 23:14:35 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA04655 for current-outgoing; Fri, 1 Aug 1997 23:14:35 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA04648; Fri, 1 Aug 1997 23:14:30 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id PAA09337; Sat, 2 Aug 1997 15:44:20 +0930 (CST) From: Michael Smith Message-Id: <199708020614.PAA09337@genesis.atrad.adelaide.edu.au> Subject: Re: core group topics In-Reply-To: <26502.870499750@orion.webspan.net> from Gary Palmer at "Aug 2, 97 01:29:10 am" To: gpalmer@FreeBSD.ORG (Gary Palmer) Date: Sat, 2 Aug 1997 15:44:19 +0930 (CST) Cc: dg@root.com, phk@dk.tfs.com, jkh@time.cdrom.com, chuckr@Glue.umd.edu, freebsd-current@freefall.FreeBSD.org X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Gary Palmer stands accused of saying: > > There is? I don't recall discussing that at all, and if we do > > eventually go the ELF way, I think the kernel would be one of the > > last areas we'd want to convert since it further complicates our > > cramped bootblocks. > > All the more reason to go to 3 stage booting. It seems to have been > one of these `nice to have, maybe someday' things like a fully working > DEVFS. What can I say? Grab the standalone library stuff from NetBSD, the boot3 code that Bill Paul released last year (I can send you a copy if you like), and make it happen. All it will take is some patience and a little i386 knowledge. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Sat Aug 2 02:28:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id CAA13854 for current-outgoing; Sat, 2 Aug 1997 02:28:16 -0700 (PDT) Received: from dfw-ix15.ix.netcom.com (dfw-ix15.ix.netcom.com [206.214.98.15]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id CAA13814; Sat, 2 Aug 1997 02:28:04 -0700 (PDT) Received: (from smap@localhost) by dfw-ix15.ix.netcom.com (8.8.4/8.8.4) id EAA10704; Sat, 2 Aug 1997 04:27:32 -0500 (CDT) Received: from sjx-ca30-16.ix.netcom.com(204.31.235.176) by dfw-ix15.ix.netcom.com via smap (V1.3) id sma010692; Sat Aug 2 04:27:28 1997 Received: (from asami@localhost) by blimp.mimi.com (8.8.6/8.6.9) id CAA11125; Sat, 2 Aug 1997 02:27:25 -0700 (PDT) Date: Sat, 2 Aug 1997 02:27:25 -0700 (PDT) Message-Id: <199708020927.CAA11125@blimp.mimi.com> To: andreas@klemm.gtn.com CC: ports@freebsd.org, current@freebsd.org, stable@freebsd.org In-reply-to: <19970801182448.26268@klemm.gtn.com> (message from Andreas Klemm on Fri, 1 Aug 1997 18:24:48 +0200) Subject: Re: ports-current/packages-current discontinued From: asami@cs.berkeley.edu (Satoshi Asami) Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk * But unable to build really working 2.2 packages, because they * only have -current :-/ Well, I'll be doing thot from now on. Watch packages-stable on your nearest mirror site -- other than building packages for new/upgraded ports, I've been recompiling the entire tree every month or so, and that will still continue after switching to -stable. * Well, wouldn't it make more sense to discuss, to back out the * changes that introduces these problems ? You are absolutely right. * Why was TCL not brought into the ports collection ??? * We already have several TCL versions there for backward * compatibility ... why not put the newest stuff into there * with respect to the problems. that now arise ? Beats me. ;) * Wouln't it be better to come to a decision, that such * things like perl and tcl, _if_ they are needed in the * base system, should be of the same version in -current * and -stable ?! That's a big "if" there. I think the latest incident are making many people wonder "is it really worth having tcl in the base tree?". The problem with tcl is that it changes too much and often new versions are not backwards compatible with the old ones. It's not much of a problem if it isn't such an essential part of the FreeBSD environment -- but it is, and there are so many ports that depend on a particular version of tcl, it is simply too much of a pain to have "the" version in the base system yank the ports around every time it is upgraded. (The last non-continuation occurred between tcl73/tk36 and tcl74/tk40 -- but at that time, tcl was not part of the base tree so we could deal with it entirely in the ports side.) There are only two ways out of this, as far as I can tell; remove tcl from the base system (pst has done most of the work for this), or completely ignore the one in the base system and always use tcl from the ports collection (I'm not sure how hard this is -- we may need an enhanced version of LIB_DEPENDS or something). * Ports should at least run with -STABLE and -CURRENT. I only run Absolutely. I (and many others) have spend much time and effort on this; I would like to continue doing so. Satoshi P.S. By the way, every single one of the packages (that worked for -current) still compiled cleanly after I switched to -stable; the only ones that broke since my last -current build (about two weeks ago) was tk41 (broken by half-baked upgrade/switch of dependency, since fixed) and tclX75 (broken by tk41, haven't figured this one out yet). From owner-freebsd-current Sat Aug 2 03:13:40 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id DAA16018 for current-outgoing; Sat, 2 Aug 1997 03:13:40 -0700 (PDT) Received: from genesis.atrad.adelaide.edu.au (genesis.atrad.adelaide.edu.au [129.127.96.120]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id DAA15998; Sat, 2 Aug 1997 03:13:30 -0700 (PDT) Received: (from msmith@localhost) by genesis.atrad.adelaide.edu.au (8.8.5/8.7.3) id TAA09852; Sat, 2 Aug 1997 19:43:14 +0930 (CST) From: Michael Smith Message-Id: <199708021013.TAA09852@genesis.atrad.adelaide.edu.au> Subject: Re: ports-current/packages-current discontinued In-Reply-To: <199708020927.CAA11125@blimp.mimi.com> from Satoshi Asami at "Aug 2, 97 02:27:25 am" To: asami@cs.berkeley.edu (Satoshi Asami) Date: Sat, 2 Aug 1997 19:43:14 +0930 (CST) Cc: andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk I've stayed out of this thread long enough. 8) Satoshi has, IMHO, done the right thing, but for all the wrong reasons. Ports should have less (or no) "official" support on -current. If you are running -current, you should be able to take care of yourself. Naturally, merging fixes to help them run on -current is desirable, as it will make the cutover at the next major upgrade easier. > * Well, wouldn't it make more sense to discuss, to back out the > * changes that introduces these problems ? > > You are absolutely right. We've done this. Countless times. Nobody appears to have actually done very much about the results of these discussions. It is, however, apparent that moving in any _other_ direction that that eventually agreed on will only make things worse. > * Wouln't it be better to come to a decision, that such > * things like perl and tcl, _if_ they are needed in the > * base system, should be of the same version in -current > * and -stable ?! This indicates that you haven't even watched, let alone studied, the issues that have been raised in these discussions before. > There are only two ways out of this, as far as I can tell; remove tcl > from the base system (pst has done most of the work for this), or > completely ignore the one in the base system and always use tcl from > the ports collection (I'm not sure how hard this is -- we may need an > enhanced version of LIB_DEPENDS or something). The correct answer to this is, of course, that any port that requires a specific Tcl version or range of versions should require one of those versions out of the ports collection. A port failing to operate regardless of the Tcl version in the base distribution is _fundamentally_broken_, and should be fixed. -- ]] Mike Smith, Software Engineer msmith@gsoft.com.au [[ ]] Genesis Software genesis@gsoft.com.au [[ ]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[ ]] realtime instrument control. (ph) +61-8-8267-3493 [[ ]] Unix hardware collector. "Where are your PEZ?" The Tick [[ From owner-freebsd-current Sat Aug 2 04:10:11 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA19569 for current-outgoing; Sat, 2 Aug 1997 04:10:11 -0700 (PDT) Received: from news1.gtn.com (news1.gtn.com [192.109.159.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA19564 for ; Sat, 2 Aug 1997 04:10:03 -0700 (PDT) Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) with UUCP id LAA01589; Sat, 2 Aug 1997 11:45:15 +0200 (MET DST) Received: (from andreas@localhost) by klemm.gtn.com (8.8.6/8.8.6) id LAA03986; Sat, 2 Aug 1997 11:14:03 +0200 (CEST) Message-ID: <19970802111403.26993@klemm.gtn.com> Date: Sat, 2 Aug 1997 11:14:03 +0200 From: Andreas Klemm To: John Polstra Cc: hasty@rah.star-gate.com, current@FreeBSD.ORG Subject: Re: second post: elf patches for -current References: <199708011006.DAA03932@rah.star-gate.com> <199708020146.SAA09667@austin.polstra.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Mutt 0.79 In-Reply-To: <199708020146.SAA09667@austin.polstra.com>; from John Polstra on Fri, Aug 01, 1997 at 06:46:10PM -0700 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Fri, Aug 01, 1997 at 06:46:10PM -0700, John Polstra wrote: > > ElfKit is almost obsolete, because just about everything you need > is now in -current. You can build the FreeBSD compiler for ELF by > specifying "BINFORMAT=elf". The standard libc in -current (and I > think in -2.2 also) can be built for ELF, and it works just fine. > You don't have to do anything special; just compile it with the > ELF tools. If I define BINFORMAT=elf in /etc/make.conf ... can I do a make world then ? Or do I have to be very careful about that ? Are there any restrictions if I╢d decide, to give ELF a try ? > The current version of GNU binutils has all the FreeBSD-ELF support > in it. Just configure it like this and then build it: > > ./configure --prefix=/usr/local/elf --host=i386-unknown-freebsd \ > --target=i386-unknown-freebsdelf Do I need the binutils if I do a complete make world ? Only want to make sure, not to screw up my system within a few minutes ;-) -- Andreas Klemm | klemm.gtn.com - powered by Symmetric MultiProcessor FreeBSD http://www.freebsd.org/~fsmp/SMP/SMP.html http://www.freebsd.org/~fsmp/SMP/benches.html From owner-freebsd-current Sat Aug 2 04:14:35 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id EAA19684 for current-outgoing; Sat, 2 Aug 1997 04:14:35 -0700 (PDT) Received: from news1.gtn.com (news1.gtn.com [192.109.159.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id EAA19665; Sat, 2 Aug 1997 04:14:24 -0700 (PDT) Received: (from uucp@localhost) by news1.gtn.com (8.7.2/8.7.2) with UUCP id MAA07164; Sat, 2 Aug 1997 12:30:14 +0200 (MET DST) Received: (from andreas@localhost) by klemm.gtn.com (8.8.6/8.8.6) id MAA02251; Sat, 2 Aug 1997 12:23:32 +0200 (CEST) Message-ID: <19970802122332.51323@klemm.gtn.com> Date: Sat, 2 Aug 1997 12:23:32 +0200 From: Andreas Klemm To: Satoshi Asami Cc: ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued References: <19970801182448.26268@klemm.gtn.com> <199708020927.CAA11125@blimp.mimi.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Mutt 0.79 In-Reply-To: <199708020927.CAA11125@blimp.mimi.com>; from Satoshi Asami on Sat, Aug 02, 1997 at 02:27:25AM -0700 X-Disclaimer: A free society is one where it is safe to be unpopular X-Operating-System: FreeBSD 3.0-CURRENT SMP Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, Aug 02, 1997 at 02:27:25AM -0700, Satoshi Asami wrote: > * But unable to build really working 2.2 packages, because they > * only have -current :-/ > > Well, I'll be doing thot from now on. Watch packages-stable on your > nearest mirror site -- other than building packages for new/upgraded > ports, I've been recompiling the entire tree every month or so, and > that will still continue after switching to -stable. Hmmm, words like "if it ain╢t broken, don╢t fix it". The whole thing intorduces new words an phrases as "ports-stable" removes well known things like ports-current. Wouldn╢t it be simpler to keep things as they are and to consolidate on a basis, that makes it again possible for you so as if nothing bad has happened ? ;-) > * Well, wouldn't it make more sense to discuss, to back out the > * changes that introduces these problems ? > > You are absolutely right. Shouldn╢t we then wait for an agreement without changing things or do you want to push the discussion forward by introducing ports-stable ? > The problem with tcl is that it changes too much and often new > versions are not backwards compatible with the old ones. It's not > much of a problem if it isn't such an essential part of the FreeBSD > environment -- but it is, and there are so many ports that depend on a > particular version of tcl, it is simply too much of a pain to have > "the" version in the base system yank the ports around every time it > is upgraded. (The last non-continuation occurred between tcl73/tk36 > and tcl74/tk40 -- but at that time, tcl was not part of the base > tree so we could deal with it entirely in the ports side.) I think FreeBSD doesn╢t have to use both scripting languages. Perl is _the_ straight forward enhancement of sh & Co., so that it should be ok to write utilities based on this. Desktop Managament systems or administration tool could be as well packaged as port with a proper dependency to the correct TCL/TK version. During system installation you could ask if the user wants to install the foo/bar administration tool package and FreeBSD would be purified from the bloat of huge "work in progress" packages ... > There are only two ways out of this, as far as I can tell; remove tcl > from the base system (pst has done most of the work for this), or > completely ignore the one in the base system and always use tcl from > the ports collection (I'm not sure how hard this is -- we may need an > enhanced version of LIB_DEPENDS or something). See above, I╢d vote now to remove TCP completely after seeing that results. > Absolutely. I (and many others) have spend much time and effort on > this; I would like to continue doing so. Thanks for your and all the others most valueable work. My big wish would be, not to debate about this too much, we all know our wonderful working ports mechanism, so then please use it when it is recommended. And I think it would be here. Would you all agree, to make ports of FreeBSD system utilities that rely on TCL ? Is it possible to remove TCL from the base system and keep the ports mechism as it is without introducing new declarations ? -- Andreas Klemm | klemm.gtn.com - powered by Symmetric MultiProcessor FreeBSD http://www.freebsd.org/~fsmp/SMP/SMP.html http://www.freebsd.org/~fsmp/SMP/benches.html From owner-freebsd-current Sat Aug 2 07:07:32 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id HAA24763 for current-outgoing; Sat, 2 Aug 1997 07:07:32 -0700 (PDT) Received: from critter.dk.tfs.com (critter.phk.freebsd.dk [195.8.133.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id HAA24747 for ; Sat, 2 Aug 1997 07:07:24 -0700 (PDT) Received: from critter.dk.tfs.com (localhost [127.0.0.1]) by critter.dk.tfs.com (8.8.6/8.8.5) with ESMTP id QAA01521 for ; Sat, 2 Aug 1997 16:06:24 +0200 (CEST) To: current@freebsd.org Subject: DMI anyone ? From: Poul-Henning Kamp Date: Sat, 02 Aug 1997 16:06:24 +0200 Message-ID: <1519.870530784@critter.dk.tfs.com> Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Does anybody have in insight or interest in the DMI stuff ? (http://www.dmtf.org) -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. Future will arrive by its own means, progress not so. From owner-freebsd-current Sat Aug 2 09:03:23 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id JAA29036 for current-outgoing; Sat, 2 Aug 1997 09:03:23 -0700 (PDT) Received: from austin.polstra.com (austin.polstra.com [206.213.73.10]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id JAA29031 for ; Sat, 2 Aug 1997 09:03:19 -0700 (PDT) Received: from austin.polstra.com (jdp@localhost) by austin.polstra.com (8.8.6/8.8.5) with ESMTP id JAA16383; Sat, 2 Aug 1997 09:03:08 -0700 (PDT) Message-Id: <199708021603.JAA16383@austin.polstra.com> To: Andreas Klemm cc: hasty@rah.star-gate.com, current@FreeBSD.ORG Subject: Re: second post: elf patches for -current In-reply-to: Your message of "Sat, 02 Aug 1997 11:14:03 +0200." <19970802111403.26993@klemm.gtn.com> Date: Sat, 02 Aug 1997 09:03:08 -0700 From: John Polstra Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > If I define BINFORMAT=elf in /etc/make.conf ... can I do a make > world then ? No. You need the binutils assembler and linker, at least. Also, you need the ElfKit dynamic linker. Also, there was a little patch to binutils "nm" in ElfKit which GNU did not accept for compatibility reasons. Without the patch, "lorder" doesn't work right. I think I could fix lorder so it works in either case, though. > Or do I have to be very careful about that ? > [...] > Only want to make sure, not to screw up my system within a few > minutes ;-) Let's put it this way: I've never been brave enough to try it. :-O But I don't have a system that I'm willing to reinstall if necessary. Looking forward to your report ... John -- John Polstra jdp@polstra.com John D. Polstra & Co., Inc. Seattle, Washington USA "Self-knowledge is always bad news." -- John Barth From owner-freebsd-current Sat Aug 2 09:25:32 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id JAA29912 for current-outgoing; Sat, 2 Aug 1997 09:25:32 -0700 (PDT) Received: from ady.warp.starnets.ro (ady.warp.starnets.ro [193.226.124.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id JAA29887; Sat, 2 Aug 1997 09:25:03 -0700 (PDT) Received: from localhost (ady@localhost) by ady.warp.starnets.ro (8.8.5/8.8.5) with SMTP id TAA05353; Sat, 2 Aug 1997 19:22:26 +0300 (EEST) Date: Sat, 2 Aug 1997 19:22:26 +0300 (EEST) From: Penisoara Adrian To: Michael Smith cc: Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <199708021013.TAA09852@genesis.atrad.adelaide.edu.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Hi, On Sat, 2 Aug 1997, Michael Smith wrote: > Ports should have less (or no) "official" support on -current. If you > are running -current, you should be able to take care of yourself. Well, what about folks running 3.0-current for SMP reasons and that have not so much ideea about UN*X/FreeBSD hacking ? By chance it happens that myself too I am in the same position... > Naturally, merging fixes to help them run on -current is desirable, as > it will make the cutover at the next major upgrade easier. Not only that, the ports should be buildable on every version of FreeBSD, if possible; it comes to my mind something like GNU's autoconfig scripts... I'd really like to see *one* single ports tree, buildable on every FreeBSD version; this might be hard, I know, but I think it well worths working on it. Ady (@warp.starnets.ro) From owner-freebsd-current Sat Aug 2 10:03:23 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA01642 for current-outgoing; Sat, 2 Aug 1997 10:03:23 -0700 (PDT) Received: from pcnet1.pcnet.com (pcnet1.pcnet.com [204.213.232.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA01636 for ; Sat, 2 Aug 1997 10:03:15 -0700 (PDT) Received: (from eischen@localhost) by pcnet1.pcnet.com (8.8.6/8.8.6) id NAA07648; Sat, 2 Aug 1997 13:00:05 -0400 (EDT) Date: Sat, 2 Aug 1997 13:00:05 -0400 (EDT) From: Daniel Eischen Message-Id: <199708021700.NAA07648@pcnet1.pcnet.com> To: andreas@klemm.gtn.com, jdp@polstra.com Subject: Re: second post: elf patches for -current Cc: current@FreeBSD.ORG, hasty@rah.star-gate.com Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > > If I define BINFORMAT=elf in /etc/make.conf ... can I do a make > > world then ? > > No. You need the binutils assembler and linker, at least. > > Also, you need the ElfKit dynamic linker. > > Also, there was a little patch to binutils "nm" in ElfKit which GNU > did not accept for compatibility reasons. Without the patch, > "lorder" doesn't work right. I think I could fix lorder so it works > in either case, though. Don't you also need the elf-branding patch to bfd/elf.c? Unless you want to brand your own elfs with brandelf. I didn't find this patch incorporated into binutils-2.7. Perhaps binutils-2.8 has it? Dan Eischen eischen@pcnet.com From owner-freebsd-current Sat Aug 2 10:05:42 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA01743 for current-outgoing; Sat, 2 Aug 1997 10:05:42 -0700 (PDT) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.2.228.19]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA01735 for ; Sat, 2 Aug 1997 10:05:39 -0700 (PDT) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.5/8.6.9) id DAA21674 for current@freebsd.org; Sun, 3 Aug 1997 03:03:01 +1000 Date: Sun, 3 Aug 1997 03:03:01 +1000 From: Bruce Evans Message-Id: <199708021703.DAA21674@godzilla.zeta.org.au> To: current@freebsd.org Subject: cvs vs default 16MB data limit Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk `cvs co -r CSRG dict' fails attempting to allocate a measly 2.4MB for dict/web2. This is with the default data limit of 16MB. Bruce From owner-freebsd-current Sat Aug 2 10:21:57 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA02227 for current-outgoing; Sat, 2 Aug 1997 10:21:57 -0700 (PDT) Received: from labs.usn.blaze.net.au (root@labs.usn.blaze.net.au [203.17.53.30]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id KAA02207; Sat, 2 Aug 1997 10:21:45 -0700 (PDT) Received: from labs.usn.blaze.net.au (davidn@local [127.0.0.1]) by labs.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id DAA00921; Sun, 3 Aug 1997 03:20:13 +1000 (EST) Message-Id: <199708021720.DAA00921@labs.usn.blaze.net.au> X-Mailer: exmh version 2.0zeta 7/24/97 To: Michael Smith cc: asami@cs.berkeley.edu (Satoshi Asami), andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 19:43:14 +0930." <199708021013.TAA09852@genesis.atrad.adelaide.edu.au> X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sun, 03 Aug 1997 03:20:12 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > > There are only two ways out of this, as far as I can tell; remove tcl > > from the base system (pst has done most of the work for this), or > > completely ignore the one in the base system and always use tcl from > > the ports collection (I'm not sure how hard this is -- we may need an > > enhanced version of LIB_DEPENDS or something). > > The correct answer to this is, of course, that any port that requires > a specific Tcl version or range of versions should require one of > those versions out of the ports collection. A port failing to operate > regardless of the Tcl version in the base distribution is > _fundamentally_broken_, and should be fixed. This is a gross over-simplification. The ports collection cannot ignore the rest of the system on which it is installed. That's the whole point of this discussion. Well, perhaps it isn't that much of an over-simplification, if you consider that all of these problems are fixed if you remove tcl from the base dist, in -current and in 2.2-stable as well (yes, this would be a lot of work, and yes, I'll volunteer to help). Just out of curiosity, why exactly was tcl8.0 *beta* added to the base distribution? I saw a message from Jordan which said "this will make my life easier" giving rather obscure reasons, but I can't express the amazement I experienced when I first saw the commit. This upgrade has caused me no end of problems, since it broke many tcl and tk apps I run on several systems. Yes, sure, this is -current, and we have to live with these inconveniences, but I'd be far more comfortable living with this particular inconvenience if the end result was the removal of tcl from the base distribution altogether, and moving it to ports. This is not an anti-tcl stance, but the opposite; I use tcl/tk widely myself, but I'd far prefer to track versions through the ports system than have to rely on the base distribution staying up-to-date, purely because many apps require different versions. The same with perl. The base distribution is simply not the place for these things; surely that much is obvious by now? Upgrading the base system's version to an untested beta with some very significant internal changes, even in -current, is the most stupid move I have seen in the FreeBSD project since my involvement. I can count the number of such incidents I've witnessed in the last couple of years on one hand, so it's not like the project is infested with stupidity. It was very ill-considered, and Satoshi's position here is critical. That he apparently got no say in this is incredible, to say the least. It seems obvious to an outsider that there are some very fundamental communication problems within the core team. Seeing this fixed is even more critical than where tcl or perl happen to reside. Regards, David -- David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 10:54:35 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id KAA03704 for current-outgoing; Sat, 2 Aug 1997 10:54:35 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [206.246.122.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id KAA03663; Sat, 2 Aug 1997 10:54:25 -0700 (PDT) Received: from Journey2.mat.net (journey2.mat.net [206.246.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id NAA17995; Sat, 2 Aug 1997 13:54:23 -0400 Date: Sat, 2 Aug 1997 13:54:35 -0400 (EDT) From: Chuck Robey X-Sender: chuckr@Journey2.mat.net To: current@FreeBSD.ORG cc: ports@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <199708021720.DAA00921@labs.usn.blaze.net.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 3 Aug 1997, David Nugent wrote: > Upgrading the base system's version to an untested beta with some very > significant internal changes, even in -current, is the most stupid move > I have seen in the FreeBSD project since my involvement. I can count the > number of such incidents I've witnessed in the last couple of years on one > hand, so it's not like the project is infested with stupidity. It was > very ill-considered, and Satoshi's position here is critical. That he > apparently got no say in this is incredible, to say the least. It seems > obvious to an outsider that there are some very fundamental communication > problems within the core team. Seeing this fixed is even more critical > than where tcl or perl happen to reside. [I trimmed out all the single names, you guys are all subscribed] Another thing to consider is the basic cleanliness of FreeBSD to tcl applications that don't arrive via the ports mechanism. Our present environment is such that even a well written tcl app hasn't the least chance of configuring itself correctly. In our respect for tcl, we've constructed a tcl-hostile system. ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 213 Lakeside Drive Apt T-1 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-current Sat Aug 2 11:08:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA04351 for current-outgoing; Sat, 2 Aug 1997 11:08:18 -0700 (PDT) Received: from critter.dk.tfs.com (critter.phk.freebsd.dk [195.8.133.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA04311; Sat, 2 Aug 1997 11:08:11 -0700 (PDT) Received: from critter.dk.tfs.com (localhost [127.0.0.1]) by critter.dk.tfs.com (8.8.6/8.8.5) with ESMTP id UAA01775; Sat, 2 Aug 1997 20:05:18 +0200 (CEST) To: David Nugent cc: Michael Smith , asami@cs.berkeley.edu (Satoshi Asami), andreas@klemm.gtn.com, ports@freebsd.org, current@freebsd.org, stable@freebsd.org From: Poul-Henning Kamp Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sun, 03 Aug 1997 03:20:12 +1000." <199708021720.DAA00921@labs.usn.blaze.net.au> Date: Sat, 02 Aug 1997 20:05:18 +0200 Message-ID: <1773.870545118@critter.dk.tfs.com> Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In message <199708021720.DAA00921@labs.usn.blaze.net.au>, David Nugent writes: >Just out of curiosity, why exactly was tcl8.0 *beta* added to the base >distribution? Because -current is "not yet release code" and tcl8.0b2 is "almost but not quite release code". That's why. It sounds to me like a lot of people should seriously reconsider if they ought to run -current :-( -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. Power and ignorance is a disgusting cocktail. From owner-freebsd-current Sat Aug 2 11:29:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA05043 for current-outgoing; Sat, 2 Aug 1997 11:29:18 -0700 (PDT) Received: from helios.dnttm.ru (root@dnttm.wave.ras.ru [194.85.104.197]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA05037; Sat, 2 Aug 1997 11:29:09 -0700 (PDT) Received: (from uucp@localhost) by helios.dnttm.ru (8.8.5/8.8.5/IP-3) with UUCP id WAA07806; Sat, 2 Aug 1997 22:22:56 +0400 Received: from tejblum.dnttm.rssi.ru (localhost [127.0.0.1]) by tejblum.dnttm.rssi.ru (8.8.6/8.8.5) with ESMTP id WAA00744; Sat, 2 Aug 1997 22:24:32 +0400 (MSD) Message-Id: <199708021824.WAA00744@tejblum.dnttm.rssi.ru> X-Mailer: exmh version 2.0gamma 1/27/96 To: Michael Smith cc: asami@cs.berkeley.edu (Satoshi Asami), ports@FreeBSD.ORG, current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 19:43:14 +0930." <199708021013.TAA09852@genesis.atrad.adelaide.edu.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 02 Aug 1997 22:24:30 +0400 From: Dmitrij Tejblum Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Ports should have less (or no) "official" support on -current. No. Ports should have more (full) "official" support for 2.2. If Satoshi can not build packages for both 2.2 and -current, I prefer packages for 2.2. [...] > A port failing to operate > regardless of the Tcl version in the base distribution is > _fundamentally_broken_, and should be fixed. It is slightly other issue, but see: I still can run binaries linked with libc.so.2.2 on my -current machine. But after I did 'make world' today I am unable to run binaries linked with libtcl75.so.1.1 (wish4.1), because: application-specific initialization failed: Can't find a usable init.tcl in the following directories: /usr/libdata/tcl /usr/local/lib/tcl7.5 /usr/tcl7.5/library /usr/local/library This probably means that Tcl wasn't installed properly. Is the correct answer "Install Tcl 7.5 from ports"? IMHO, something is broken here, and should be fixed. Dima From owner-freebsd-current Sat Aug 2 11:34:40 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA05408 for current-outgoing; Sat, 2 Aug 1997 11:34:40 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA05385; Sat, 2 Aug 1997 11:34:32 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id LAA15390; Sat, 2 Aug 1997 11:32:38 -0700 (PDT) To: David Nugent cc: Michael Smith , asami@cs.berkeley.edu (Satoshi Asami), andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sun, 03 Aug 1997 03:20:12 +1000." <199708021720.DAA00921@labs.usn.blaze.net.au> Date: Sat, 02 Aug 1997 11:32:38 -0700 Message-ID: <15386.870546758@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > I have seen in the FreeBSD project since my involvement. I can count the > number of such incidents I've witnessed in the last couple of years on one > hand, so it's not like the project is infested with stupidity. It was > very ill-considered, and Satoshi's position here is critical. That he > apparently got no say in this is incredible, to say the least. It seems > obvious to an outsider that there are some very fundamental communication > problems within the core team. Seeing this fixed is even more critical > than where tcl or perl happen to reside. What most people here don't understand is that there was very fierce debate about this behind the scenes for some time before it was done, and what started as a TCL debate blew up into a whole "should FreeBSD have everything unbundled, from the compilers to perl, or should it bundle all the high level tools so that other system tools can be written which depend on them?" sort of fracas. One man's useful tool is another man's wasteful, unnecessary bloat, it seems, and rather than see an anti-bloatist campaign which would have led to the removal of tcl, perl, xntpd, tn3270 and a host of other utilities which are currently not deemed "essential" by the anti-bloatists, I think it was sort of deemed the lesser of two evils to just let the bloatists win the TCL debate. At least that's how I see it from my perspective - frankly, after the debate in question was over (which, again, most people here were spared), I decided I didn't even want to think about the issue for awhile and that's why I've been silent in the face of Satoshi's impassioned pleas - I don't want to go back to the bargaining table and have to decide which utilities will get the axe. Once you start with TCL, it will *not* stop there - I can only assure you of that. Perl will follow immediately behind, as will much other stuff (yes perl fans, there are many out there who consider your favorite utility language an evil, bloated monster which should not be bundled with FreeBSD at all). What we have now is a rough state of equilibrium between the two sides (who are fundamentally at odds as to what constitutes a reasonable bundling policy) and while TCL may be causing some grief, I think the bloatists are content with that state of affairs and nothing else is on the bundling horizon that I can see. Nuke TCL and you will swing the balance in the other direction, with a lot more than just TCL biting the dust as a result. Maybe that's not such a bad thing, but just so you understand how much of a "linchpin" issue this one is. Jordan From owner-freebsd-current Sat Aug 2 11:39:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id LAA05707 for current-outgoing; Sat, 2 Aug 1997 11:39:14 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id LAA05688; Sat, 2 Aug 1997 11:39:09 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.171]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA126; Sat, 2 Aug 1997 14:30:12 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id OAA01081; Sat, 2 Aug 1997 14:38:46 -0400 (EDT) Message-ID: <19970802143845.26672@scsn.net> Date: Sat, 2 Aug 1997 14:38:45 -0400 From: "Donald J. Maddox" To: Poul-Henning Kamp Cc: David Nugent , Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued Reply-To: dmaddox@scsn.net References: <199708021720.DAA00921@labs.usn.blaze.net.au> <1773.870545118@critter.dk.tfs.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 In-Reply-To: <1773.870545118@critter.dk.tfs.com>; from Poul-Henning Kamp on Sat, Aug 02, 1997 at 08:05:18PM +0200 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, Aug 02, 1997 at 08:05:18PM +0200, Poul-Henning Kamp wrote: > In message <199708021720.DAA00921@labs.usn.blaze.net.au>, David Nugent writes: > > >Just out of curiosity, why exactly was tcl8.0 *beta* added to the base > >distribution? > > Because -current is "not yet release code" and tcl8.0b2 is "almost but > not quite release code". > > That's why. > > It sounds to me like a lot of people should seriously reconsider if > they ought to run -current :-( > > -- > Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. > http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. > whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. > Power and ignorance is a disgusting cocktail. This seems to me to be a very silly attitude. Today's -current is tomorrow's release... If this foolishness of continuing to maintain tcl in the base tree is not stopped _now_, we will have to live with it for the next N releases. _Please_ end it *now*. From owner-freebsd-current Sat Aug 2 12:09:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA07061 for current-outgoing; Sat, 2 Aug 1997 12:09:14 -0700 (PDT) Received: from dyson.iquest.net (dyson.iquest.net [198.70.144.127]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA07026 for ; Sat, 2 Aug 1997 12:08:40 -0700 (PDT) Received: (from root@localhost) by dyson.iquest.net (8.8.6/8.8.5) id OAA13852; Sat, 2 Aug 1997 14:08:32 -0500 (EST) From: "John S. Dyson" Message-Id: <199708021908.OAA13852@dyson.iquest.net> Subject: Re: cvs vs default 16MB data limit In-Reply-To: <199708021703.DAA21674@godzilla.zeta.org.au> from Bruce Evans at "Aug 3, 97 03:03:01 am" To: bde@zeta.org.au (Bruce Evans) Date: Sat, 2 Aug 1997 14:08:32 -0500 (EST) Cc: current@FreeBSD.ORG Reply-To: dyson@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL31 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > `cvs co -r CSRG dict' fails attempting to allocate a measly 2.4MB > for dict/web2. This is with the default data limit of 16MB. > What is the reason for the default data limit of 16MB anyway? Is there a reason to keep it so low? (I know that it might be side-stepping an issue, but I have had to modify the NCI distribution for the "default.") It is okay, but why so small? John From owner-freebsd-current Sat Aug 2 12:09:51 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA07083 for current-outgoing; Sat, 2 Aug 1997 12:09:51 -0700 (PDT) Received: from precipice.shockwave.com (ppp-206-170-5-32.rdcy01.pacbell.net [206.170.5.32]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA07058; Sat, 2 Aug 1997 12:09:11 -0700 (PDT) Received: from shockwave.com (localhost [127.0.0.1]) by precipice.shockwave.com (8.8.6/8.7.3) with ESMTP id MAA06706; Sat, 2 Aug 1997 12:07:09 -0700 (PDT) Message-Id: <199708021907.MAA06706@precipice.shockwave.com> To: David Nugent cc: committers@FreeBSD.ORG, freebsd-current@FreeBSD.ORG Subject: Re: rc.shutdown In-reply-to: Your message of "Sat, 02 Aug 1997 10:33:30 +1000." <199708020033.KAA15728@unique.usn.blaze.net.au> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 02 Aug 1997 12:07:08 -0700 From: Paul Traina Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk This warning is bogus, rc.shutdown is not required for proper operation, so why complain? From owner-freebsd-current Sat Aug 2 12:12:59 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA07282 for current-outgoing; Sat, 2 Aug 1997 12:12:59 -0700 (PDT) Received: from sos.freebsd.dk (sos.freebsd.dk [195.8.129.33]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA07261; Sat, 2 Aug 1997 12:12:33 -0700 (PDT) Received: (from sos@localhost) by sos.freebsd.dk (8.8.6/8.7.3) id VAA00512; Sat, 2 Aug 1997 21:11:15 +0200 (MEST) From: SЬren Schmidt Message-Id: <199708021911.VAA00512@sos.freebsd.dk> Subject: Re: ports-current/packages-current discontinued In-Reply-To: <199708021824.WAA00744@tejblum.dnttm.rssi.ru> from Dmitrij Tejblum at "Aug 2, 97 10:24:30 pm" To: dima@tejblum.dnttm.rssi.ru (Dmitrij Tejblum) Date: Sat, 2 Aug 1997 21:11:15 +0200 (MEST) Cc: msmith@atrad.adelaide.edu.au, asami@cs.berkeley.edu, ports@FreeBSD.ORG, current@FreeBSD.ORG X-Mailer: ELM [version 2.4ME+ PL30 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk In reply to Dmitrij Tejblum who wrote: > > It is slightly other issue, but see: I still can run binaries linked with > libc.so.2.2 on my -current machine. But after I did 'make world' today I am > unable to run binaries linked with libtcl75.so.1.1 (wish4.1), because: > > application-specific initialization failed: Can't find a usable init.tcl in > the following directories: > /usr/libdata/tcl /usr/local/lib/tcl7.5 /usr/tcl7.5/library /usr/local/library > This probably means that Tcl wasn't installed properly. > > Is the correct answer "Install Tcl 7.5 from ports"? IMHO, something is broken > here, and should be fixed. Yeah, like removing tcl from the base system.... -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- SЬren Schmidt (sos@FreeBSD.org) FreeBSD Core Team Even more code to hack -- will it ever end .. From owner-freebsd-current Sat Aug 2 12:26:05 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA07852 for current-outgoing; Sat, 2 Aug 1997 12:26:05 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA07811; Sat, 2 Aug 1997 12:25:42 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id MAA15696; Sat, 2 Aug 1997 12:23:22 -0700 (PDT) To: dmaddox@scsn.net cc: David Nugent , Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 15:11:35 EDT." <19970802151135.60481@scsn.net> Date: Sat, 02 Aug 1997 12:23:21 -0700 Message-ID: <15692.870549801@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Most of this argument is just silly. Even the most vehement anti- > bloatists don't consider perl to be "an evil, bloated monster"; they just > consider _it's inclusion in the base distribution_ to be AEBM. While you I don't see the difference from the POV of this discussion so this paragraph of yours doesn't really parse for me. > In any case, I see none of this bloatist v. antibloatist propaganda > as cogent here. Tcl should not be part of the base system because it It's imminently cogent - this is NOT just a technical issue, it's an emotional one, and if you think that all software decisions are made on purely technical merits then I have a certain tower in Paris which I could make you a _great_ deal on. ;-) Jordan From owner-freebsd-current Sat Aug 2 12:35:53 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA08345 for current-outgoing; Sat, 2 Aug 1997 12:35:53 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA08326; Sat, 2 Aug 1997 12:35:46 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.148]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA190; Sat, 2 Aug 1997 15:25:42 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id PAA01228; Sat, 2 Aug 1997 15:32:16 -0400 (EDT) Message-ID: <19970802153150.47714@scsn.net> Date: Sat, 2 Aug 1997 15:31:50 -0400 From: "Donald J. Maddox" To: "Jordan K. Hubbard" Cc: dmaddox@scsn.net, David Nugent , Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued Reply-To: dmaddox@scsn.net References: <19970802151135.60481@scsn.net> <15692.870549801@time.cdrom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 In-Reply-To: <15692.870549801@time.cdrom.com>; from Jordan K. Hubbard on Sat, Aug 02, 1997 at 12:23:21PM -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, Aug 02, 1997 at 12:23:21PM -0700, Jordan K. Hubbard wrote: > > Most of this argument is just silly. Even the most vehement anti- > > bloatists don't consider perl to be "an evil, bloated monster"; they just > > consider _it's inclusion in the base distribution_ to be AEBM. While you > > I don't see the difference from the POV of this discussion so this > paragraph of yours doesn't really parse for me. Huh? > > > In any case, I see none of this bloatist v. antibloatist propaganda > > as cogent here. Tcl should not be part of the base system because it > > It's imminently cogent - this is NOT just a technical issue, it's > an emotional one, and if you think that all software decisions are > made on purely technical merits then I have a certain tower in Paris > which I could make you a _great_ deal on. ;-) No, in light of this discussion, it's clear that they are not. I humbly submit that they _should be_. Maybe you gentlemen of core should take a step back from the situation, take a few deep breaths, and reconsider this issue. From owner-freebsd-current Sat Aug 2 12:36:07 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA08445 for current-outgoing; Sat, 2 Aug 1997 12:36:07 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA08356; Sat, 2 Aug 1997 12:35:55 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.148]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA223; Sat, 2 Aug 1997 15:25:59 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id PAA01242; Sat, 2 Aug 1997 15:34:45 -0400 (EDT) Message-ID: <19970802153427.60153@scsn.net> Date: Sat, 2 Aug 1997 15:34:28 -0400 From: "Donald J. Maddox" To: =?iso-8859-1?Q?S=F8ren_Schmidt?= Cc: Dmitrij Tejblum , msmith@atrad.adelaide.edu.au, asami@cs.berkeley.edu, ports@FreeBSD.ORG, current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued Reply-To: dmaddox@scsn.net References: <199708021824.WAA00744@tejblum.dnttm.rssi.ru> <199708021911.VAA00512@sos.freebsd.dk> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Mutt 0.79 In-Reply-To: =?iso-8859-1?Q?=3C199708021911=2EVAA00512=40sos=2Efreebsd=2Edk=3E=3B_fro?= =?iso-8859-1?Q?m_S=F8ren_Schmidt_on_Sat=2C_Aug_02=2C_1997_at_09=3A11=3A1?= =?iso-8859-1?Q?5PM_+0200?= Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, Aug 02, 1997 at 09:11:15PM +0200, SЬren Schmidt wrote: > In reply to Dmitrij Tejblum who wrote: > > > > It is slightly other issue, but see: I still can run binaries linked with > > libc.so.2.2 on my -current machine. But after I did 'make world' today I am > > unable to run binaries linked with libtcl75.so.1.1 (wish4.1), because: > > > > application-specific initialization failed: Can't find a usable init.tcl in > > the following directories: > > /usr/libdata/tcl /usr/local/lib/tcl7.5 /usr/tcl7.5/library /usr/local/library > > This probably means that Tcl wasn't installed properly. > > > > Is the correct answer "Install Tcl 7.5 from ports"? IMHO, something is broken > > here, and should be fixed. > > Yeah, like removing tcl from the base system.... > Hear, hear! From owner-freebsd-current Sat Aug 2 12:36:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA08513 for current-outgoing; Sat, 2 Aug 1997 12:36:21 -0700 (PDT) Received: from genghis.eng.demon.net (genghis.eng.demon.net [193.195.45.10]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id MAA08458; Sat, 2 Aug 1997 12:36:09 -0700 (PDT) Received: from genghis.eng.demon.net [193.195.45.10] by genghis.eng.demon.net with esmtp (Exim 1.62 #1) id 0wujxY-00006R-00; Sat, 2 Aug 1997 20:35:28 +0100 To: Poul-Henning Kamp cc: ports@freebsd.org, current@freebsd.org, stable@freebsd.org Subject: Re: ports-current/packages-current discontinued Organization: Demon Internet Ltd. Reply-To: ade@demon.net In-reply-to: Your message of "Sat, 02 Aug 1997 21:15:33 +0200." <1987.870549333@critter.dk.tfs.com> Date: Sat, 02 Aug 1997 20:35:28 +0100 From: Ade Lovett Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Poul-Henning Kamp writes: > >In message , Ade Lovett writes: >>This seems to be something of an oversimplification. Whilst there >>are undoubtedly people who are much more likely to be better off >>running release (or perhaps -stable) code, there are quite a number >>of people who need features (SMP, for example) that are only present >>in -current. > >You know, I actualle don't see that as an excuse for running -current, >but if you insist, at least don't try to use that kind of argument >for trying to turn -current into -stable, OK ? I'm suggesting no such thing. What I am suggesting is that the 'base' FreeBSD system is now too big, and too complicated, for its own good. Perl and TCL shouldn't really be in the base system at all. Especially perl, given that it's version 4. Should I wish to install perl5, either direct from CPAN, or through ports, there's even no easy way to ensure that all traces of perl4 have been removed from my machine, which may (or may not) cause subtle problems in the future. Indeed, perl and tcl bring up an interesting discrepancy. If -current is containing newer versions of tcl than in -release or -stable, why isn't the same being done for perl? After all, now that perl 5 has been around for quite a while, and if perl is to remain part of the base system, shouldn't it be being incorporated into -current, with perl 4 being dropped? Having duplicate code, in the base system, and in ports, is just causing anguish all around. That much should be obvious :( Either perl belongs in the base system, or it doesn't. If it does, then the ports version should be killed off, if it doesn't, it needs to be removed from /usr/src. Similarly for tcl. Similarly for a whole host of other things. You (plural, not aimed at anyone in particular) can't have it both ways -- it just causes far too much grief in the long run. >>The problems come about when the base operating system contains >>components that really shouldn't be there at all. TCL, Perl certainly >>fall into this category, and there's probably quite a bit else which >>would be better off in either ports, or an 'additions' package. > >I consider "options" equally bad in protocols and operating systems. We already have "options". At install time I can choose to add a bunch of things to my machine should my heart so desire. games, proflibs, catpages, manpages etc.. Splitting up "bin" into two or more components, just in the same way that selecting "src" will offer sub-choices, depending on whether the binaries are absolutely definitely critical for system to run, or whether they are useful 'core' bolt-ons, doesn't seem to be too much of a change in direction. -aDe -- Ade Lovett, Demon Internet Ltd. From owner-freebsd-current Sat Aug 2 12:43:30 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA08981 for current-outgoing; Sat, 2 Aug 1997 12:43:30 -0700 (PDT) Received: from critter.dk.tfs.com (critter.phk.freebsd.dk [195.8.133.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id MAA08955; Sat, 2 Aug 1997 12:43:17 -0700 (PDT) Received: from critter.dk.tfs.com (localhost [127.0.0.1]) by critter.dk.tfs.com (8.8.6/8.8.5) with ESMTP id VAA01989; Sat, 2 Aug 1997 21:15:33 +0200 (CEST) To: ade@demon.net cc: Poul-Henning Kamp , ports@freebsd.org, current@freebsd.org, stable@freebsd.org From: Poul-Henning Kamp Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 20:03:07 BST." Date: Sat, 02 Aug 1997 21:15:33 +0200 Message-ID: <1987.870549333@critter.dk.tfs.com> Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In message , Ade Lovett writes: >>It sounds to me like a lot of people should seriously reconsider if >>they ought to run -current :-( > > >This seems to be something of an oversimplification. Whilst there >are undoubtedly people who are much more likely to be better off >running release (or perhaps -stable) code, there are quite a number >of people who need features (SMP, for example) that are only present >in -current. You know, I actualle don't see that as an excuse for running -current, but if you insist, at least don't try to use that kind of argument for trying to turn -current into -stable, OK ? >The problems come about when the base operating system contains >components that really shouldn't be there at all. TCL, Perl certainly >fall into this category, and there's probably quite a bit else which >would be better off in either ports, or an 'additions' package. I consider "options" equally bad in protocols and operating systems. -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. Power and ignorance is a disgusting cocktail. From owner-freebsd-current Sat Aug 2 12:55:38 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id MAA10028 for current-outgoing; Sat, 2 Aug 1997 12:55:38 -0700 (PDT) Received: from genghis.eng.demon.net (genghis.eng.demon.net [193.195.45.10]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id MAA09994; Sat, 2 Aug 1997 12:55:17 -0700 (PDT) Received: from genghis.eng.demon.net [193.195.45.10] by genghis.eng.demon.net with esmtp (Exim 1.62 #1) id 0wukGW-00007A-00; Sat, 2 Aug 1997 20:55:04 +0100 To: "Jordan K. Hubbard" cc: ports@freebsd.org, current@freebsd.org, stable@freebsd.org Subject: Re: ports-current/packages-current discontinued Organization: Demon Internet Ltd. Reply-To: ade@demon.net In-reply-to: Your message of "Sat, 02 Aug 1997 11:32:38 PDT." <15386.870546758@time.cdrom.com> Date: Sat, 02 Aug 1997 20:55:03 +0100 From: Ade Lovett Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk "Jordan K. Hubbard" writes: > >Nuke TCL and you will swing the balance in the other direction, with a >lot more than just TCL biting the dust as a result. Maybe that's not >such a bad thing, but just so you understand how much of a "linchpin" >issue this one is. Surely this would depend on how far things "bite the dust". Personally, I don't think it would be unreasonable to have a situation where the base system is made up of two distinct, but heavily inter-related parts, namely that chunk of the current core system that is definitely needed to run any kind of system at all, and other parts which aren't necessarily required, but which are stamped with a kind of "seal of approval" for use with FreeBSD. We then have a situation where both sides of the "bloat" are happy, those that want a minimalistic approach can install the required components and leave out the approved integrated packages, others that prefer a full-featured system, without having to compile/install anything else, can install both. In terms of what FreeBSD (the entity) is, however, both the required and extra package components should be considered as equally important. Possibly, a somewhat simplistic attitude, but certainly something to consider, no? -aDe -- Ade Lovett, Demon Internet Ltd. From owner-freebsd-current Sat Aug 2 13:39:04 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA12769 for current-outgoing; Sat, 2 Aug 1997 13:39:04 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA12764 for ; Sat, 2 Aug 1997 13:39:02 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id NAA16096; Sat, 2 Aug 1997 13:38:09 -0700 (PDT) To: ade@demon.net cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 20:55:03 BST." Date: Sat, 02 Aug 1997 13:38:09 -0700 Message-ID: <16092.870554289@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk [Directed only to -current; 3 mailing lists is too many] > Surely this would depend on how far things "bite the dust". > Personally, I don't think it would be unreasonable to have a > situation where the base system is made up of two distinct, but > heavily inter-related parts, namely that chunk of the current core > system that is definitely needed to run any kind of system at all, > and other parts which aren't necessarily required, but which are > stamped with a kind of "seal of approval" for use with FreeBSD. That option's already been discussed to death. ;-) The question is how to do this effectively, not in the concept. The concept is the easy part. ;) Jordan From owner-freebsd-current Sat Aug 2 13:45:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA13059 for current-outgoing; Sat, 2 Aug 1997 13:45:21 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA13054 for ; Sat, 2 Aug 1997 13:45:16 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id NAA16130; Sat, 2 Aug 1997 13:44:22 -0700 (PDT) To: dmaddox@scsn.net cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 14:38:45 EDT." <19970802143845.26672@scsn.net> Date: Sat, 02 Aug 1997 13:44:22 -0700 Message-ID: <16126.870554662@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk [Please trim your cc's - this is getting out of control] > This seems to me to be a very silly attitude. Today's -current is > tomorrow's release... If this foolishness of continuing to maintain > tcl in the base tre e is not stopped _now_, we will have to live > with it for the next N releases. _Please_ end it *now*. This argument is hardly likely to win many points. If this matter is to be concluded reasonably then it's going to be because the principals involved agreed to sit down and discuss it *rationally*. Demanding that people do things "now" is just fanning the flames and bringing yet more onwonted emotion to an already overly-emotional issue. In other words, Donald, you're not being part of the solution, you're being part of the problem. Either change that aspect of your debating style or kindly stay out of the discussion since you're not helping. Jordan From owner-freebsd-current Sat Aug 2 13:51:33 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA13445 for current-outgoing; Sat, 2 Aug 1997 13:51:33 -0700 (PDT) Received: from critter.dk.tfs.com (critter.phk.freebsd.dk [195.8.133.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA13421; Sat, 2 Aug 1997 13:51:25 -0700 (PDT) Received: from critter.dk.tfs.com (localhost [127.0.0.1]) by critter.dk.tfs.com (8.8.6/8.8.5) with ESMTP id WAA02299; Sat, 2 Aug 1997 22:49:03 +0200 (CEST) To: ade@demon.net cc: "Jordan K. Hubbard" , ports@freebsd.org, current@freebsd.org, stable@freebsd.org From: Poul-Henning Kamp Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 20:55:03 BST." Date: Sat, 02 Aug 1997 22:49:03 +0200 Message-ID: <2297.870554943@critter.dk.tfs.com> Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >Possibly, a somewhat simplistic attitude, but certainly something >to consider, no? >-- >Ade Lovett, Demon Internet Ltd. And it has been considered, and as far as I know fell because what nobody saw any chance of "minimal system" being definable so that any significant majority would agree to it. -- Poul-Henning Kamp | phk@FreeBSD.ORG FreeBSD Core-team. http://www.freebsd.org/~phk | phk@login.dknet.dk Private mailbox. whois: [PHK] | phk@tfs.com TRW Financial Systems, Inc. Power and ignorance is a disgusting cocktail. From owner-freebsd-current Sat Aug 2 13:53:58 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA13622 for current-outgoing; Sat, 2 Aug 1997 13:53:58 -0700 (PDT) Received: from genghis.eng.demon.net (genghis.eng.demon.net [193.195.45.10]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id NAA13612 for ; Sat, 2 Aug 1997 13:53:51 -0700 (PDT) Received: from genghis.eng.demon.net [193.195.45.10] by genghis.eng.demon.net with esmtp (Exim 1.62 #1) id 0wulBA-00008k-00; Sat, 2 Aug 1997 21:53:36 +0100 To: "Jordan K. Hubbard" cc: current@freebsd.org Subject: Re: ports-current/packages-current discontinued Organization: Demon Internet Ltd. Reply-To: ade@demon.net In-reply-to: Your message of "Sat, 02 Aug 1997 13:38:09 PDT." <16092.870554289@time.cdrom.com> Date: Sat, 02 Aug 1997 21:53:35 +0100 From: Ade Lovett Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk "Jordan K. Hubbard" writes: > >That option's already been discussed to death. ;-) *smile* I'm sure it has.. dontcha just love the cyclic nature of project management concepts? :) >The question is how to do this effectively, not in the concept. >The concept is the easy part. ;) Well, given the major amount of work this is going to require in order to do things effectively, surely the first step is a mandate from the core team that a particular 'way forward' is the right way to go. IMHO, 'leaving things as they are' isn't a valid option in this case, since we can see that things are starting to break down, as the system gets more complex. Once the decision is made, then things can progress to discussion about how best to implement the decision, followed by resourcing the whole thing. Until the decision is made about the way forward, however, then much of the talk here is moot. -aDe -- Ade Lovett, Demon Internet Ltd. From owner-freebsd-current Sat Aug 2 13:59:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id NAA13810 for current-outgoing; Sat, 2 Aug 1997 13:59:16 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id NAA13804 for ; Sat, 2 Aug 1997 13:59:13 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id NAA16221; Sat, 2 Aug 1997 13:58:14 -0700 (PDT) To: ade@demon.net cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 21:53:35 BST." Date: Sat, 02 Aug 1997 13:58:14 -0700 Message-ID: <16217.870555494@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Well, given the major amount of work this is going to require in > order to do things effectively, surely the first step is a > mandate from the core team that a particular 'way forward' is > the right way to go. IMHO, 'leaving things as they are' isn't > a valid option in this case, since we can see that things are > starting to break down, as the system gets more complex. > > Once the decision is made, then things can progress to discussion > about how best to implement the decision, followed by resourcing > the whole thing. > > Until the decision is made about the way forward, however, then > much of the talk here is moot. I think you're suffering from a fundamental misunderstanding of how these things work. :-) The core team can't really "mandate" anything without a much stronger fait-accompli (e.g. the work essentially has to be ready to fold in and the new mechanism approved in implementation as well as concept). We've tried that before in the past, and all that's ever come of it is an empty mandate - it doesn't serve as any kind of tie-breaker or incentive for forward progress (you're probably thinking of a more corporate model where once you have your "vision" you can go essentially *order* your minions to go carry it out for you). So it's really the other way around - until a truly defensible system of package layering and installation is both proposed and proven through some set of Makefile diffs which demonstrate the viability of the concept, it's moot. Jordan From owner-freebsd-current Sat Aug 2 14:12:12 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA14313 for current-outgoing; Sat, 2 Aug 1997 14:12:12 -0700 (PDT) Received: from fallout.campusview.indiana.edu (fallout.campusview.indiana.edu [149.159.1.1]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA14256; Sat, 2 Aug 1997 14:11:48 -0700 (PDT) Received: from localhost (jfieber@localhost) by fallout.campusview.indiana.edu (8.8.5/8.8.5) with SMTP id QAA18564; Sat, 2 Aug 1997 16:10:07 -0500 (EST) Date: Sat, 2 Aug 1997 16:10:07 -0500 (EST) From: John Fieber To: "Jordan K. Hubbard" cc: dmaddox@scsn.net, David Nugent , Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <15692.870549801@time.cdrom.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 2 Aug 1997, Jordan K. Hubbard wrote: > > In any case, I see none of this bloatist v. antibloatist propaganda > > as cogent here. Tcl should not be part of the base system because it > > It's imminently cogent - this is NOT just a technical issue, it's > an emotional one, and if you think that all software decisions are > made on purely technical merits then I have a certain tower in Paris > which I could make you a _great_ deal on. ;-) Ah, yes. There are many snakes in this pit. One I have not seen recently in the discussion in the inherent problem of incorporating into the base system a substantial component that is on a fundamentally different development schedule than the rest of the OS. For things with a relatively long update cycle, such as gcc, this isn't a huge problem, but for more rapidly developing items, like tcl, users stand a good chance of wanting an update between FreeBSD releases. Our only easy to use interim update mechanism is the ports collection. -john From owner-freebsd-current Sat Aug 2 14:14:24 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA14446 for current-outgoing; Sat, 2 Aug 1997 14:14:24 -0700 (PDT) Received: from genghis.eng.demon.net (genghis.eng.demon.net [193.195.45.10]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id OAA14435 for ; Sat, 2 Aug 1997 14:14:18 -0700 (PDT) Received: from genghis.eng.demon.net [193.195.45.10] by genghis.eng.demon.net with esmtp (Exim 1.62 #1) id 0wulUw-00009C-00; Sat, 2 Aug 1997 22:14:02 +0100 To: "Jordan K. Hubbard" cc: current@freebsd.org Subject: Re: ports-current/packages-current discontinued Organization: Demon Internet Ltd. Reply-To: ade@demon.net In-reply-to: Your message of "Sat, 02 Aug 1997 13:58:14 PDT." <16217.870555494@time.cdrom.com> Date: Sat, 02 Aug 1997 22:14:02 +0100 From: Ade Lovett Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk "Jordan K. Hubbard" writes: > >I think you're suffering from a fundamental misunderstanding of how >these things work. :-) Well, yes and no :) I freely admit to being more "corporately-aware" than how things are necessarily done with the core teams of the various *BSD's out in the world. However, at certain critical points in any products lifespan, there are certain key decisions that do have to be made which will have a major impact on the future. I believe that issues raised with regards to the tcl thing are now showing that a critical period has been reached and something "has to be done"[tm] about it. In an ideal world, everybodies point of view could be considered, the pro's and con's of each suggestion weighed up, and a decision reached. Unfortunately, design-by-large-committee very rarely (if at all) works effectively. For this reason, and this reason alone, I suggested that the core-team "mandate" an appropriate decision simply because they're an identifiable and small entity. Perhaps "mandate" was the wrong word to use, though I can't for the life of me think what the right word is :) >So it's really the other way around - until a truly defensible system >of package layering and installation is both proposed and proven >through some set of Makefile diffs which demonstrate the viability of >the concept, it's moot. I'm not entirely sure that it's going to be as simple as a set of Makefile diffs though :( For example, if one considers the approach of a FreeBSD-base + FreeBSD-approved-packages system, then substantial changes would have to be made to the entire source tree. Whether such changes would take the top-level form of a set of changes to allow "make world-base" and "make world-all", or two entirely separate source trees for base and approved-packages, or some other method, we're talking about an enormous amount of work for a proof-of-concept system, especially with the constant changing of the source trees themselves. We live in interesting times. -aDe -- Ade Lovett, Demon Internet Ltd. From owner-freebsd-current Sat Aug 2 14:18:17 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA14641 for current-outgoing; Sat, 2 Aug 1997 14:18:17 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA14633 for ; Sat, 2 Aug 1997 14:18:14 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.30]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA87; Sat, 2 Aug 1997 17:08:27 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id RAA01670; Sat, 2 Aug 1997 17:00:28 -0400 (EDT) Message-ID: <19970802165908.37359@scsn.net> Date: Sat, 2 Aug 1997 16:59:08 -0400 From: "Donald J. Maddox" To: "Jordan K. Hubbard" Cc: dmaddox@scsn.net, current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued Reply-To: dmaddox@scsn.net References: <19970802143845.26672@scsn.net> <16126.870554662@time.cdrom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 In-Reply-To: <16126.870554662@time.cdrom.com>; from Jordan K. Hubbard on Sat, Aug 02, 1997 at 01:44:22PM -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, Aug 02, 1997 at 01:44:22PM -0700, Jordan K. Hubbard wrote: > [Please trim your cc's - this is getting out of control] > > > This seems to me to be a very silly attitude. Today's -current is > > tomorrow's release... If this foolishness of continuing to maintain > > tcl in the base tre e is not stopped _now_, we will have to live > > with it for the next N releases. _Please_ end it *now*. > > This argument is hardly likely to win many points. > > If this matter is to be concluded reasonably then it's going to be > because the principals involved agreed to sit down and discuss it > *rationally*. Demanding that people do things "now" is just fanning Jordan, that seems to be in direct contradiction to your last email to me on this matter. Seems to me you were stating specifically that this issue was a _very emotional_ one, and it seemed that you put a lot of value on this emotional aspect. It was I who was arguing that this should be considered in a more rational fashion. > the flames and bringing yet more onwonted emotion to an already > overly-emotional issue. > > In other words, Donald, you're not being part of the solution, you're > being part of the problem. Either change that aspect of your debating > style or kindly stay out of the discussion since you're not helping. > > Jordan I'm sorry you see it that way, Jordan. I have simply endured too many of these ego-wars amongst core-members to remain silent through this one. I cannot see how any reasonable person cannot see at this point that having tcl in the base distribution was a rather MAJOR mistake from square one, and I, for one, am disgusted that people who I _used_ to have some measure of respect for could be so moved by ego alone to allow this situation to continue. From owner-freebsd-current Sat Aug 2 14:27:08 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA15090 for current-outgoing; Sat, 2 Aug 1997 14:27:08 -0700 (PDT) Received: from hwcn.org (main.hwcn.org [199.212.94.65]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA15068; Sat, 2 Aug 1997 14:27:02 -0700 (PDT) Received: from james.freenet.hamilton.on.ca (ac199@james.hwcn.org [199.212.94.66]) by hwcn.org (8.8.6/8.8.6) with ESMTP id RAA15744; Sat, 2 Aug 1997 17:27:34 -0400 (EDT) Received: from localhost (ac199@localhost) by james.freenet.hamilton.on.ca (8.8.6/8.8.6) with SMTP id RAA25135; Sat, 2 Aug 1997 17:27:51 -0400 (EDT) X-Authentication-Warning: james.freenet.hamilton.on.ca: ac199 owned process doing -bs Date: Sat, 2 Aug 1997 17:27:48 -0400 (EDT) From: Tim Vanderhoek X-Sender: ac199@james.freenet.hamilton.on.ca Reply-To: hoek@hwcn.org To: "Jordan K. Hubbard" cc: David Nugent , Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <15386.870546758@time.cdrom.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 2 Aug 1997, Jordan K. Hubbard wrote: > Once you start with TCL, it will *not* stop there - I can only > assure you of that. Perl will follow immediately behind, as > will much other stuff (yes perl fans, there are many out there > who consider your favorite utility language an evil, bloated > monster which should not be bundled with FreeBSD at all). What > we have now is a rough state of Meta-ports!! Their time has come! Not only could meta-packages answer bloatist concerns that "an installed FreeBSD system is useless unless one adds 1001 extra packages", but they could provide increased flexibility. I believe bsd.port.mk could support meta-packages right now, but I think some extra mods to it could make the whole thing work quite well indeed. [meta-ports, for those who have forgotten, are just normal ports, but whose only purpose is to build other ports.] -- Outnumbered? Maybe. Outspoken? Never! tIM...HOEk From owner-freebsd-current Sat Aug 2 14:39:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA15657 for current-outgoing; Sat, 2 Aug 1997 14:39:21 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA15631; Sat, 2 Aug 1997 14:38:57 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id HAA14498; Sun, 3 Aug 1997 07:33:40 +1000 (EST) Message-Id: <199708022133.HAA14498@unique.usn.blaze.net.au> To: "Jordan K. Hubbard" cc: Michael Smith , asami@cs.berkeley.edu (Satoshi Asami), andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 11:32:38 MST." <15386.870546758@time.cdrom.com> X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sun, 03 Aug 1997 07:33:39 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> I have seen in the FreeBSD project since my involvement. I can count the >> number of such incidents I've witnessed in the last couple of years on one >> hand, so it's not like the project is infested with stupidity. It was >> very ill-considered, and Satoshi's position here is critical. That he >> apparently got no say in this is incredible, to say the least. It seems >> obvious to an outsider that there are some very fundamental communication >> problems within the core team. Seeing this fixed is even more critical >> than where tcl or perl happen to reside. > >What most people here don't understand is that there was very fierce >debate about this behind the scenes for some time before it was done, >and what started as a TCL debate blew up into a whole "should FreeBSD >have everything unbundled, from the compilers to perl, or should it >bundle all the high level tools so that other system tools can be >written which depend on them?" sort of fracas. One man's useful tool >is another man's wasteful, unnecessary bloat, it seems, and rather [..] Yes, yes, yes. But all completely besides the point. I don't care one way or another about "bloat". It is irrelevent to the point I was making. "Bloat" is subjective, as you point out, but it has nothing to do with how tools are updated and integrated into the system. Tools which are imported into the base system are *necessarily* more static, because of the very real concerns that have caused problems here. They won't change as often, they can't change as often, else we should by rights be carefully following perl development IN THE SOURCE TREE rather than having not only an obsolete version of perl, one completely unsupported and with seriously security bugs, and one which hardly anyone uses anyway. Tcl, until this apparently unpopular import, was at least in a better position than that. I've said before that perl4 must die. I don't particularly care whether it is imported into the source tree or left as a port. It is stale, it is old, it is obsolete and completely unsupported. It has bugs, limitations and security holes. It must go. Period. My only concern about importing a current version into the source tree is the same as tcl - what guarantees are there that this will be kept up to date? The ports system, in spite of its faults and shortfalls, *works*, and works extremely well, mainly because of the efforts of porters and Satoshi's management. But I hardly need to point this out to you of all people. The ports system integrates third party software into the base system seemlessly, provides upgrade paths and has people actively supporting (the most popular packages, anyway) and updating their favourite apps. Sure, there are mistakes there, like the recent apache vs. apache current debate, but these are far more easily handled in the less expensive and more low-maintenance system that ports has been designed to handle. Again, this has nothing to do with any "bloatist" point of view, nor is it anti-perl or anti-tcl either - I use both daily and enhance systems constantly with both of them. BUT: How is tcl essential to a working FreeBSD environment? How is perl, for that matter, other than a few scattered scripts which can and do work quite happily with current versions of perl (and can easily be replaced with /bin/sh scripts anyway)? Why are we nailing ourselves to versions of third party software which are prone to become stale and unmaintained? Why can't the ports system be used for them? >Once you start with TCL, it will *not* stop there - I can only assure >you of that. Perl will follow immediately behind, as will much other >stuff (yes perl fans, there are many out there who consider your >favorite utility language an evil, bloated monster which should not be >bundled with FreeBSD at all). What we have now is a rough state of Surely, you're not suggesting that just because something is in not in the FreeBSD base system that it is "evil"? Oh come on, this is such a childish attitude I can't believe I'm hearing it. The question is not how large a package is, it is how useful it is in the base distribution - how dependant upon it a FreeBSD system is in installation and setup, and how integrated it is into the system. Neither perl nor tcl have any grip in the system where they could be considered so essential that FreeBSD will not run without them. It is more *appropriate* that they be maintained, integrated and - ESPECIALLY - kept current via the ports system. This is a major PLUS for users, not the opposite as you seem to be assuming. perl4 in the tree is living proof of exactly where we end up with the approach that is being taken. >equilibrium between the two sides (who are fundamentally at odds as to >what constitutes a reasonable bundling policy) and while TCL may be >causing some grief, I think the bloatists are content with that state >of affairs and nothing else is on the bundling horizon that I can see. >Nuke TCL and you will swing the balance in the other direction, with a >lot more than just TCL biting the dust as a result. Maybe that's not >such a bad thing, but just so you understand how much of a "linchpin" >issue this one is. No, I understand the issue well enough. But I think it is irrevelent. Having "current" versions of things in ports and "old" things in the tree is not only annoying, it is downright ridiculous. The opposite is almost as bad. I would quickly add that any generalisation here is dangerous - you have to consider the benefits of things being integrated into the source tree vs. the costs of possibly becoming stale. It is far easier to keep our cvs "hands" of a third party package and integrate it via the ports system. Diffs between a distribution and the "FreeBSD version" are absolutely plain without even having a source repository around. Things like bind, gcc, gdb etc. are a completely different ballgame. So, no, I don't see this as any sort of weighted argument. You can cry about bloat all day, and I simple don't care. That point of view misses the issue entirely. tcl is a major headache in terms of multiple version operability. Perl is likewise (do you do much perl debugging?). I'd live silently with these problems if the benefits outweighed the cost in terms of hassle, but to me, it just isn't worth it. Regards, David David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 14:41:50 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA15848 for current-outgoing; Sat, 2 Aug 1997 14:41:50 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA15842 for ; Sat, 2 Aug 1997 14:41:47 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id OAA16424; Sat, 2 Aug 1997 14:40:50 -0700 (PDT) To: ade@demon.net cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 22:14:02 BST." Date: Sat, 02 Aug 1997 14:40:50 -0700 Message-ID: <16420.870558050@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > I'm not entirely sure that it's going to be as simple as a set > of Makefile diffs though :( For example, if one considers the > approach of a FreeBSD-base + FreeBSD-approved-packages system, then > substantial changes would have to be made to the entire source tree. Exactly, which is part of what makes it such a difficult subject to wrestle with. :-( I'm all in favor of better mechanisms for dealing with this, don't get me wrong, I just think it's far far easier to say "we need mechanism x" than it is to define and implement mechanism x. :-) Jordan From owner-freebsd-current Sat Aug 2 14:44:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA15992 for current-outgoing; Sat, 2 Aug 1997 14:44:16 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA15926; Sat, 2 Aug 1997 14:43:36 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id HAA14589; Sun, 3 Aug 1997 07:40:47 +1000 (EST) Message-Id: <199708022140.HAA14589@unique.usn.blaze.net.au> To: Poul-Henning Kamp cc: Michael Smith , asami@cs.berkeley.edu (Satoshi Asami), andreas@klemm.gtn.com, ports@freebsd.org, current@freebsd.org, stable@freebsd.org Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 20:05:18 +0200." <1773.870545118@critter.dk.tfs.com> X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sun, 03 Aug 1997 07:40:47 +1000 From: David Nugent Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk >In message <199708021720.DAA00921@labs.usn.blaze.net.au>, David Nugent writes: >>Just out of curiosity, why exactly was tcl8.0 *beta* added to the base >>distribution? > >Because -current is "not yet release code" and tcl8.0b2 is "almost but >not quite release code". > >That's why. Well, sure, but This isn't why I asked. What features in tcl8.0 are required for a running FreeBSD system? Why is it essential that tcl be present in the base system AT ALL? What benefits are there in tcl being present here instead of being built by the ports/packages system? >It sounds to me like a lot of people should seriously reconsider if >they ought to run -current :-( Most of the systems I maintain are 2.2. However, I do software development on -current. I already mentioned that I'm quite prepared to live with any inconveniences (and there are plenty of those) absolutely without complaint. This has nothing to do with the issues I'm raising. This change appears to be entirely gratuitous. Please correct me if I am wrong. I'd seriously like to know what the benefits are of nailing ourselves to tcl8.0 in the FreeBSD base system will yield, or in fact what benefits having tcl7.5 is already giving 2.2 systems. Why is it worth the cost, hassle and upset, particularly with respect to the ports system. What compelling reasons are there or having it there as compared with the much more cheaply maintained ports system? Enquiring minds want to know. Regards, David David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 14:49:00 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id OAA16405 for current-outgoing; Sat, 2 Aug 1997 14:49:00 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id OAA16400 for ; Sat, 2 Aug 1997 14:48:58 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id OAA16482; Sat, 2 Aug 1997 14:48:04 -0700 (PDT) To: dmaddox@scsn.net cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 16:59:08 EDT." <19970802165908.37359@scsn.net> Date: Sat, 02 Aug 1997 14:48:04 -0700 Message-ID: <16478.870558484@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Jordan, that seems to be in direct contradiction to your last > email to me on this matter. Seems to me you were stating specifically > that this issue was a _very emotional_ one, and it seemed that you put > a lot of value on this emotional aspect. It was I who was arguing that > this should be considered in a more rational fashion. No, I was stating that this was a *problem* in this discussion and that debating it on purely technical terms would miss that crucial point, not that emoting had some intrinsic value in this debate. I was in no way trying to suggest that you should get all emotional in turn - stick to your technical points, please. :-) > I'm sorry you see it that way, Jordan. I have simply endured too many of > these ego-wars amongst core-members to remain silent through this one. I > cannot see how any reasonable person cannot see at this point that having > tcl in the base distribution was a rather MAJOR mistake from square one, Yes, and that's your opinion. Others have dissenting opinions or there would be no debate here. I'm just trying to make the point that pounding on the desk with your shoe is no way to bring the other side into agreement. Jordan From owner-freebsd-current Sat Aug 2 15:03:19 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA16993 for current-outgoing; Sat, 2 Aug 1997 15:03:19 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA16971; Sat, 2 Aug 1997 15:03:08 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id IAA14912; Sun, 3 Aug 1997 08:02:11 +1000 (EST) Message-Id: <199708022202.IAA14912@unique.usn.blaze.net.au> To: Paul Traina cc: committers@FreeBSD.ORG, freebsd-current@FreeBSD.ORG Subject: Re: rc.shutdown In-reply-to: Your message of "Sat, 02 Aug 1997 12:07:08 MST." <199708021907.MAA06706@precipice.shockwave.com> X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sun, 03 Aug 1997 08:02:11 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > This warning is bogus, rc.shutdown is not required for proper > operation, so why complain? The complaint is generic; it doesn't come from init itself, but from from /bin/sh. I think special-casing its non-existance is probably the way to go, in which case, the attempt to run it will simply be skipped. Any objections? This should also make it easy to integrate into 2.2. I can't see any reason not do do so, can you? It is a minor addition, a "new feature", sure, but there are no costs involved provided the warning is removed and the benefits for news servers in particular are there. The sysctl used is already in 2.2-stable. The only element missing is sysctlbyname(3) in 2.2's libc. [Peter Wemm - did you get my request to add it a couple of weeks ago? :-) If you have, I missed the commit message]. Regards, David David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 15:10:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA17453 for current-outgoing; Sat, 2 Aug 1997 15:10:21 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA17423; Sat, 2 Aug 1997 15:10:04 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id IAA15046; Sun, 3 Aug 1997 08:08:19 +1000 (EST) Message-Id: <199708022208.IAA15046@unique.usn.blaze.net.au> To: John Fieber cc: dmaddox@scsn.net, Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 16:10:07 EST." X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sun, 03 Aug 1997 08:08:18 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > Ah, yes. There are many snakes in this pit. One I have not seen > recently in the discussion in the inherent problem of > incorporating into the base system a substantial component that > is on a fundamentally different development schedule than the > rest of the OS. For things with a relatively long update cycle, > such as gcc, this isn't a huge problem, but for more rapidly > developing items, like tcl, users stand a good chance of wanting > an update between FreeBSD releases. Our only easy to use interim > update mechanism is the ports collection. Absolutely. This is much more succinct and clear than my own argument, and it is THE fundamental and most important issue at stake here. Well said. Regards, David David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 15:13:15 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA17726 for current-outgoing; Sat, 2 Aug 1997 15:13:15 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA17705; Sat, 2 Aug 1997 15:13:04 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id IAA15131; Sun, 3 Aug 1997 08:12:36 +1000 (EST) Message-Id: <199708022212.IAA15131@unique.usn.blaze.net.au> To: hoek@hwcn.org cc: Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 17:27:48 -0400." X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sun, 03 Aug 1997 08:12:36 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >Meta-ports!! Their time has come! >~ >[meta-ports, for those who have forgotten, are just normal >ports, but whose only purpose is to build other ports.] You lost me. :-) How is the structure any different from the current *_DEPENDS system? Regards, David David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 15:15:21 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA17913 for current-outgoing; Sat, 2 Aug 1997 15:15:21 -0700 (PDT) Received: from genghis.eng.demon.net (genghis.eng.demon.net [193.195.45.10]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id PAA17874 for ; Sat, 2 Aug 1997 15:15:01 -0700 (PDT) Received: from genghis.eng.demon.net [193.195.45.10] by genghis.eng.demon.net with esmtp (Exim 1.62 #1) id 0wumRi-0000B3-00; Sat, 2 Aug 1997 23:14:46 +0100 To: hoek@hwcn.org cc: current@freebsd.org Subject: Re: ports-current/packages-current discontinued Organization: Demon Internet Ltd. Reply-To: ade@demon.net In-reply-to: Your message of "Sat, 02 Aug 1997 17:27:48 EDT." Date: Sat, 02 Aug 1997 23:14:46 +0100 From: Ade Lovett Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk [Followups to -current only, attempting to keep the discussion in one place] Tim Vanderhoek writes: > >Meta-ports!! Their time has come! > >Not only could meta-packages answer bloatist concerns that "an >installed FreeBSD system is useless unless one adds 1001 extra >packages", but they could provide increased flexibility. Actually, I don't think we need to go as far as meta-ports insomuchas implementing a separate concept, rather we build in the concept of dependencies into ports (both in binary and source form). Example 1: I want to compile cvsup from source, which requires modula-3. The make process sees checks to see if I have a modula-3/binary package installed, if not, then it uses the default rule to create modula-3/binary, giving me the option of either ftp'ing (or whatever) the /binary version, or building from the /source version. Once all dependencies have been satisfied, it then goes on to register cvsup/source in my package tree (since I've ftp'd it from somewhere), builds cvsup/binary and registers and installs that. Example 2: I want to install a full FreeBSD binary system (lets say 2.2.2-RELEASE). The actual initial install process only installs a minimal system, but then also has an "extras" /binary package, which builds nothing itself, but has a whole host of dependencies for perl/binary, tcl/binary etc.. etc.. which it then proceeds to install from ftp, cdrom, floppy etc. Nothing particularly world-crushingly-new here, but the concept of full packaging, with proper interdependencies, version numbering etc does have a certain appeal to it. Of course, there are a whole load of examples of how packaging has failed miserably, or is unusable, etc.. but even a cursory look at such afflicted systems will show that it's the implementation of packaging that's at fault, rather than the idea. Packaging most certainly is not a global panacea, and there are a whole host of issues that would need to be addressed at a very early stage to ensure that we don't end up shooting ourselves in the foot, but the concept does have a certain charm to it. Is this kind of thing worth investigating in more detail? -aDe -- Ade Lovett, Demon Internet Ltd. From owner-freebsd-current Sat Aug 2 15:48:23 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA19473 for current-outgoing; Sat, 2 Aug 1997 15:48:23 -0700 (PDT) Received: from mail.scsn.net (scsn.net [206.25.246.12]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA19468 for ; Sat, 2 Aug 1997 15:48:16 -0700 (PDT) Received: from rhiannon.scsn.net ([208.133.153.36]) by mail.scsn.net (Post.Office MTA v3.1 release PO203a ID# 0-32322U5000L100S10000) with ESMTP id AAA135; Sat, 2 Aug 1997 18:38:30 -0400 Received: (from root@localhost) by rhiannon.scsn.net (8.8.6/8.8.5) id SAA02615; Sat, 2 Aug 1997 18:46:56 -0400 (EDT) Message-ID: <19970802184536.51442@scsn.net> Date: Sat, 2 Aug 1997 18:45:36 -0400 From: "Donald J. Maddox" To: "Jordan K. Hubbard" Cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued Reply-To: dmaddox@scsn.net References: <19970802165908.37359@scsn.net> <16478.870558484@time.cdrom.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.79 In-Reply-To: <16478.870558484@time.cdrom.com>; from Jordan K. Hubbard on Sat, Aug 02, 1997 at 02:48:04PM -0700 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, Aug 02, 1997 at 02:48:04PM -0700, Jordan K. Hubbard wrote: > No, I was stating that this was a *problem* in this discussion and > that debating it on purely technical terms would miss that crucial > point, not that emoting had some intrinsic value in this debate. I > was in no way trying to suggest that you should get all emotional in > turn - stick to your technical points, please. :-) How can one stick to the technical points while simultaneously somehow implicitly acknowledging that this is not strictly a technical issue? I don't know enough linguistic contortions to manage an email like that. > Yes, and that's your opinion. Others have dissenting opinions or > there would be no debate here. I'm just trying to make the point that > pounding on the desk with your shoe is no way to bring the other side > into agreement. Agreed, emphatically... But there _was_ no debate on this issue, at least none in a public forum. Tcl8.x just appeared out of nowhere. (Actually, just prior to this, I was laboring under the apparent misap- prehension that tcl was finally about to be removed from the base distibution.) From owner-freebsd-current Sat Aug 2 15:55:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA19836 for current-outgoing; Sat, 2 Aug 1997 15:55:03 -0700 (PDT) Received: from hwcn.org (main.hwcn.org [199.212.94.65]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA19810; Sat, 2 Aug 1997 15:54:53 -0700 (PDT) Received: from james.freenet.hamilton.on.ca (ac199@james.hwcn.org [199.212.94.66]) by hwcn.org (8.8.6/8.8.6) with ESMTP id SAA22918; Sat, 2 Aug 1997 18:55:26 -0400 (EDT) Received: from localhost (ac199@localhost) by james.freenet.hamilton.on.ca (8.8.6/8.8.6) with SMTP id SAA03479; Sat, 2 Aug 1997 18:55:41 -0400 (EDT) X-Authentication-Warning: james.freenet.hamilton.on.ca: ac199 owned process doing -bs Date: Sat, 2 Aug 1997 18:55:41 -0400 (EDT) From: Tim Vanderhoek X-Sender: ac199@james.freenet.hamilton.on.ca Reply-To: hoek@hwcn.org To: David Nugent cc: hoek@hwcn.org, Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <199708022212.IAA15131@unique.usn.blaze.net.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 3 Aug 1997, David Nugent wrote: > >Meta-ports!! Their time has come! > >~ > >[meta-ports, for those who have forgotten, are just normal > >ports, but whose only purpose is to build other ports.] > > You lost me. :-) > > How is the structure any different from the current *_DEPENDS > system? A normal port, eg. xonix-1.4, or tvp-0.9.7, will depend on other pieces of the port, but those depends are merely a side-effect ie. they are needed to compile xonix or tvp. However, a meta-port would consist of little more than a list of dependancies. Its purpose would not be to install a single piece of software (eg. tvp or xonix), but to install a whole list of software. To better support such meta-packages, bsd.port.mk could be expanded a little. A variable such as META, when turned on, could merge all the PLISTS into one big one, and force ``make package'' to include all the dependancy's binaries in the package (not currently done). Sysinstall could give these meta ports a special place in its world, making it obvious that one is expected to install them to get a "complete" system. This would have the large advantage that users of older releases could update their system by installing a meta-package or two. If the idea proves succesful, it could be possible to upgrade a whole system and kernel just by typing "pkg_add". :) Of course, this puts a lot of power in Satoshi's hands, and I think that may scare a certain few people. ;) (Also, there is precedent for installing parts of the "base system" through a port/package. Consider XFree86). -- Outnumbered? Maybe. Outspoken? Never! tIM...HOEk From owner-freebsd-current Sat Aug 2 15:58:22 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id PAA20058 for current-outgoing; Sat, 2 Aug 1997 15:58:22 -0700 (PDT) Received: from hwcn.org (main.hwcn.org [199.212.94.65]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id PAA20050 for ; Sat, 2 Aug 1997 15:58:19 -0700 (PDT) Received: from james.freenet.hamilton.on.ca (ac199@james.hwcn.org [199.212.94.66]) by hwcn.org (8.8.6/8.8.6) with ESMTP id SAA23315; Sat, 2 Aug 1997 18:58:53 -0400 (EDT) Received: from localhost (ac199@localhost) by james.freenet.hamilton.on.ca (8.8.6/8.8.6) with SMTP id SAA03971; Sat, 2 Aug 1997 18:59:08 -0400 (EDT) X-Authentication-Warning: james.freenet.hamilton.on.ca: ac199 owned process doing -bs Date: Sat, 2 Aug 1997 18:59:08 -0400 (EDT) From: Tim Vanderhoek X-Sender: ac199@james.freenet.hamilton.on.ca Reply-To: hoek@hwcn.org To: Ade Lovett cc: hoek@hwcn.org, current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 2 Aug 1997, Ade Lovett wrote: > Example 2: > I want to install a full FreeBSD binary system (lets say > 2.2.2-RELEASE). The actual initial install process only > installs a minimal system, but then also has an "extras" > /binary package, which builds nothing itself, but has a whole > host of dependencies for perl/binary, tcl/binary etc.. etc.. > which it then proceeds to install from ftp, cdrom, floppy etc. Basically, we have this already modulo a idealisms you threw in. :) Your example #2 is the meta-port concept that I mention. A couple new features in bsd.port.mk would be wanted, but the major groundwork is laid in the existing ports system. -- Outnumbered? Maybe. Outspoken? Never! tIM...HOEk From owner-freebsd-current Sat Aug 2 16:08:01 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA20558 for current-outgoing; Sat, 2 Aug 1997 16:08:01 -0700 (PDT) Received: from genghis.eng.demon.net (genghis.eng.demon.net [193.195.45.10]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id QAA20529 for ; Sat, 2 Aug 1997 16:07:57 -0700 (PDT) Received: from genghis.eng.demon.net [193.195.45.10] by genghis.eng.demon.net with esmtp (Exim 1.62 #1) id 0wunGs-0000CE-00; Sun, 3 Aug 1997 00:07:38 +0100 To: hoek@hwcn.org cc: current@freebsd.org Subject: Re: ports-current/packages-current discontinued Organization: Demon Internet Ltd. Reply-To: ade@demon.net In-reply-to: Your message of "Sat, 02 Aug 1997 18:59:08 EDT." Date: Sun, 03 Aug 1997 00:07:38 +0100 From: Ade Lovett Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Tim Vanderhoek writes: > >Basically, we have this already modulo a idealisms you threw in. >:) Your example #2 is the meta-port concept that I mention. A >couple new features in bsd.port.mk would be wanted, but the major >groundwork is laid in the existing ports system. Well, it was the "meta-port" label that concerned me, so I'm looking for a bit of clarification, that's all. Rather than implement a meta-port concept, which does nothing but bundle ports together, we add in (a) the idea that a port may not build anything itself, and (b) the idea of dependencies in both source *and* binary form. For example: port-A only has a list of dependencies (port-B) port-B builds something itself, and also depends on port-B1 and port-B2 port-B1 only available in /binary form (say it's a simple list of configuration files) port-B2 available in both /source and /binary form, no need to grab the /source form if the /binary is already present on the system -aDe -- Ade Lovett, Demon Internet Ltd. From owner-freebsd-current Sat Aug 2 16:51:18 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id QAA22473 for current-outgoing; Sat, 2 Aug 1997 16:51:18 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id QAA22468 for ; Sat, 2 Aug 1997 16:51:16 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id QAA16929; Sat, 2 Aug 1997 16:50:25 -0700 (PDT) To: dmaddox@scsn.net cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 18:45:36 EDT." <19970802184536.51442@scsn.net> Date: Sat, 02 Aug 1997 16:50:25 -0700 Message-ID: <16924.870565825@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > How can one stick to the technical points while simultaneously > somehow implicitly acknowledging that this is not strictly a technical > issue? I don't know enough linguistic contortions to manage an email > like that. I never said it would be easy. :-) > Agreed, emphatically... But there _was_ no debate on this issue, at > least none in a public forum. Tcl8.x just appeared out of nowhere. I think it's an acknowledged point that this could have been handled a lot better, and it's my hope that we can still come to some agreement as to how to handle this now, I'm just trying to keep everyone's back fur reasonably smoothed down in the meantime or we're only going to spend the next week shouting at one another. Take it from someone with far too much experience with the latter. Jordan From owner-freebsd-current Sat Aug 2 17:07:09 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA23294 for current-outgoing; Sat, 2 Aug 1997 17:07:09 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA23289 for ; Sat, 2 Aug 1997 17:07:05 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id RAA17044; Sat, 2 Aug 1997 17:06:07 -0700 (PDT) To: David Nugent cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sun, 03 Aug 1997 07:40:47 +1000." <199708022140.HAA14589@unique.usn.blaze.net.au> Date: Sat, 02 Aug 1997 17:06:07 -0700 Message-ID: <17040.870566767@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk [Trimmed to -current - PLEASE watch those cc lines, folks! We're all getting unforgivably sloppy here] > What features in tcl8.0 are required for a running FreeBSD system? > Why is it essential that tcl be present in the base system AT ALL? > What benefits are there in tcl being present here instead of being > built by the ports/packages system? OK, these are all good questions. I will attempt to answer some of them. 1. TCL needs, *at some point* (and note that the current move was rather premature, but let's not debate that here) to be part of the base system so that the installation tools can use it. We do intend on being heavy users of TCL, if not right this minute then in the future. 2. If it were in ports, we'd have a build problem since you wouldn't be able to build /usr/src/usr.sbin/setup (not existant yet, but it will be) without first building and installing a port. This would break the world target. Now the obvious "answer" is to somehow integrate ports with things that depend on it in /usr/src (I'm assuming that we'd also bite the bullet with perl and that things like adduser would also have this problem), but the question is how? Where do the distfiles live? How does the world target jump from src to ports in building a "complete" system with all the trimmings? How does this effect how we distribute ports and on which CD(s) the various files live? Most people only have one drive, and if you're going to support building srcs off CD (a definite goal) then what does that entail in the brave new world of merged ports and src? Those are the questions which *must* be answered, and answered well, before we can start truly pushing things out of /usr/src and exclusively into the ports collection. Elitist solutions which require dedicated network access or more than a jellybean 486 (of which we still have many in our userbase) need not apply. Jordan From owner-freebsd-current Sat Aug 2 17:30:36 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA24388 for current-outgoing; Sat, 2 Aug 1997 17:30:36 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA24380 for ; Sat, 2 Aug 1997 17:30:31 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id RAA17158; Sat, 2 Aug 1997 17:29:35 -0700 (PDT) To: David Nugent cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sun, 03 Aug 1997 07:33:39 +1000." <199708022133.HAA14498@unique.usn.blaze.net.au> Date: Sat, 02 Aug 1997 17:29:34 -0700 Message-ID: <17153.870568174@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk [ I'm getting REALLY TIRED of editing your cc lines - either get that straight or leave this discussion. Redirected to -current for what is hopefully the last time ] > The ports system, in spite of its faults and shortfalls, *works*, > and works extremely well, mainly because of the efforts of porters But it doesn't work in a stand-alone environment. I wish people would stop assuming that everyone in the world has a T1 to their desks or a CDROM full of convenient tarballs. It's just not the case and arguing that ports is a complete replacement for /usr/src is just naive in the extreme. I've said it before and I'll say it again - ports would make a fine mechanism for this IF IT WERE IMPROVED to deal with those sorts of "where are my sources?" issues, but that's not the case today. > >stuff (yes perl fans, there are many out there who consider your > >favorite utility language an evil, bloated monster which should not be > >bundled with FreeBSD at all). What we have now is a rough state of > > Surely, you're not suggesting that just because something is > in not in the FreeBSD base system that it is "evil"? Oh come on, > this is such a childish attitude I can't believe I'm hearing it. Then you need to get the wax out of your ears because what you're hearing is not what I said. :-) Perl is bloated and evil because it is a large, ugly piece of software and that's all I meant by this - whether it's part of src or ports bears no relation at all to its evilness, and I merely cited it as an example of something else that people are crying for to leave /usr/src. > The question is not how large a package is, it is how useful it > is in the base distribution - how dependant upon it a FreeBSD > system is in installation and setup, and how integrated it is Yes yes yes, we know all this. But by that same token, and what you perhaps fail to realize, is that "useful" is a highly subjective term and there's MUCH MORE besides TCL in /usr/src which fits the description of "not useful" and should be slated for the executioner's axe. What annoys me are the folks, of which you are apparently one of, who feel that we can all go back to sleep safely once the giant is slain, be that particular giant perl or TCL. In reality, they're simply symptoms of a greater maliase and I'd like to hear from those who truly have constructive solutions to offer in dealing with this ongoing problem. Kill TCL and you've solved the problem for a day. Kill our inflexible src mechanism and you've a more lasting solution, but nobody wants to tackle the real problem while there are so many more convenient peccadillos to flame about. Bah. Jordan From owner-freebsd-current Sat Aug 2 17:53:10 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id RAA25292 for current-outgoing; Sat, 2 Aug 1997 17:53:10 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id RAA25287 for ; Sat, 2 Aug 1997 17:53:04 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id KAA17855; Sun, 3 Aug 1997 10:52:41 +1000 (EST) Message-Id: <199708030052.KAA17855@unique.usn.blaze.net.au> To: "Jordan K. Hubbard" cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 17:06:07 MST." <17040.870566767@time.cdrom.com> X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sun, 03 Aug 1997 10:52:41 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk >> What features in tcl8.0 are required for a running FreeBSD system? >> Why is it essential that tcl be present in the base system AT ALL? >> What benefits are there in tcl being present here instead of being >> built by the ports/packages system? > >OK, these are all good questions. I will attempt to answer some >of them. > >1. TCL needs, *at some point* (and note that the current move was > rather premature, but let's not debate that here) to be part of > the base system so that the installation tools can use it. > We do intend on being heavy users of TCL, if not right this minute > then in the future. > >2. If it were in ports, we'd have a build problem since you wouldn't > be able to build /usr/src/usr.sbin/setup (not existant yet, but > it will be) without first building and installing a port. This > would break the world target. > >Now the obvious "answer" is to somehow integrate ports with things >that depend on it in /usr/src No, wrong answer, imho. This makes things way too complex. I already suggested a better solution to this a few weeks ago in private email, in which I cc'd you. The discussion, if you recall it (I never saw any reply from you, so I can only assume you read it) was originally about ncurses. It seems that ncurses is used by libdialog, and the only thing that seems to use either in our tree is sysinstall. Our ncurses is at version 1.8.4 dated October 1994, yet the current vendor version is 4.1. Again, we have this disparity of release schedules that John mentioned. And worse, because it isn't being actively maintained, it gets mouldy and ports depending on libncurse run into problems (ever tried installing a newer version of ncurses? I have, and it ain't pretty). I suggest that sysinstall and the release tools do not belong under src/ in the repository. They belong in a separate place, say "release/", and don't even need to track the same CVS tags. 95% of the commits you do on sysinstall are to all three branches, which I'm sure must be a headache for you. Certainly, the target for sysinstall needs to be modified based on release version, but you can handle this using makefile and preprocessor tags rather than CVS revisions (which is infinitely more complex to maintain, I'm sure). Now, if you want to move all of sysinstall's dependancies out to a separate tree, just like doc, you free the base system of quite a bit of cruft. sysinstall is built static, yet we have shared libs in our tree that nothing uses, unless they are built from ports. Shouldn't libncurses (shared lib) be therefore linked from /usr/local/lib? Is it just me, or does anyone else see the logic here? Why do we have things like libncurses and libdialog in our tree if (a) nothing uses it except sysinstall, and (b) it hasn't been updated in *years* (well, ncurses was updated recently, but this is incidental and I understand that only one vendor imported file was brought into HEAD anyway, so we're still well out of date regardless). If tcl is there for sysinstall or the Son Of Sysinstall, then clearly some restructuring is required. Now, you can have whatever you like as dependant bits under, say, release/, and you don't need to add all these bits to the world target which just gather dust after a few months. They can gather dust all they like under release/, since sysinstall no doubt depends on them and maybe these particular versions of them, but they aren't installed onto a new system, which is then free to use the ports/ packages system to install and maintain up-to-date rleases. "make release" is a separate phase from world anyway, and should logically be separated from the entire make world process. Or perhaps I'm missing something else in the larger picture. Regards, David David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 18:00:16 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA25642 for current-outgoing; Sat, 2 Aug 1997 18:00:16 -0700 (PDT) Received: from makai.lovett.com (root@makai.lovett.com [193.195.45.11]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id SAA25627 for ; Sat, 2 Aug 1997 18:00:07 -0700 (PDT) Received: from makai.lovett.com [193.195.45.11] (ade) by makai.lovett.com with esmtp (Exim 1.62 #1) id 0wup1O-0003bf-00; Sun, 3 Aug 1997 01:59:46 +0100 To: "Jordan K. Hubbard" cc: current@freebsd.org Subject: Re: ports-current/packages-current discontinued Organization: Demon Internet Ltd. Reply-To: ade@demon.net In-reply-to: Your message of "Sat, 02 Aug 1997 17:29:34 PDT." <17153.870568174@time.cdrom.com> Date: Sun, 03 Aug 1997 01:59:46 +0100 From: Ade Lovett Message-Id: Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk "Jordan K. Hubbard" writes: > >But it doesn't work in a stand-alone environment. I wish people would >stop assuming that everyone in the world has a T1 to their desks or a >CDROM full of convenient tarballs. It's just not the case and arguing >that ports is a complete replacement for /usr/src is just naive in the >extreme. Well, I for one am not arguing this to be the case. What we need is something along the ports line, but extended to deal with most (if not all) situations. But first, we have to consider the types of installation that we're supporting. It seems to me that we have the following types: 1. someone with a some kind of full net connection (maybe a T1, or even a dialup with lots of time :), that only downlaods a boot floppy. (net-poweruser) 2. someone who, by one means or another, has a complete source tree available to them, be it on cdrom, or they've downloaded the whole lot over the net, or whatever. (local-poweruser) 3. someone who has downloaded a basic system over the net (equivalent to grabbing the -release floppy sets from an ftp server), and then installs locally. (binary-enduser) The powerusers are easy to deal with, using the existing ports mechanism.. they have access to the full sources, and can just go (cd /usr/ports; make world) and go to sleep for a while. #3 is somewhat different. They don't have the time, nor the inclination, nor the space etc.. to build stuff from scratch. Yet they still want access to our packagised system (beyond the base FreeBSD code, which will continue to be installed in the usual manner). Surely the solution is to provide binary releases of our packagised software (indeed, we already have this with ports/ and packages/). Whatever we do, the enduser is still going to have to get their code from somewhere, be it local cdrom, or over the net. By reducing the core system to a minimum, we've saved them download time by them not having to pull down something which they're never going to use, so we provide options, either at download time (or install time) as to whether they want to download selected parts, or everything, at their own discretion. Why do I feel like I'm missing something fundamental here? :) -aDe -- Ade Lovett, Demon Internet Ltd. From owner-freebsd-current Sat Aug 2 18:46:06 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id SAA26989 for current-outgoing; Sat, 2 Aug 1997 18:46:06 -0700 (PDT) Received: from fly.HiWAAY.net (root@fly.HiWAAY.net [208.147.154.56]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA26958; Sat, 2 Aug 1997 18:45:52 -0700 (PDT) Received: from nexgen.hiwaay.net by fly.HiWAAY.net; (8.8.6/1.1.8.2/21Sep95-1003PM) id UAA12727; Sat, 2 Aug 1997 20:45:04 -0500 (CDT) Received: from nexgen (localhost [127.0.0.1]) by nexgen.hiwaay.net (8.8.6/8.8.4) with ESMTP id UAA04956; Sat, 2 Aug 1997 20:44:57 -0500 (CDT) Message-Id: <199708030144.UAA04956@nexgen.hiwaay.net> X-Mailer: exmh version 2.0zeta 7/24/97 To: ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG From: dkelly@hiwaay.net Subject: Re: ports-current/packages-current discontinued Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Sat, 02 Aug 1997 20:42:00 -0500 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk I trimmed the individual addresses from this reply but wonder if it really should be going to (3) FreeBSD lists? Maybe its time to consider a change in a bit different direction. Is anyone familiar with the inst system SGI uses? Its almsost unmanagable from an ASCII terminal but with the graphic tool its simple enough for my boss. Every component SGI provides is registered in their installation history. Some are not optional. Others are. And the inst tool knows what is required for what, and when a newer version is installed it knows what to delete. Try to install something without the prerequisites and it'll tell you exactly what is missing, and if its available you'll be offered the chance to install it too. Otherwise you may be swapping CDs until you find it. Needless to say, this would make it very easy for someone to delete all of TCL or perl or whatever from a base FreeBSD system if we had it. Or to install it. It would be handy for me right now as I've had some system freezes, possibly always in X (one time it rebooted in the middle of the night for no reason, one time in 2 years) and I'm about to make XFree86-3.3. Would be nice to be able to cleanly remove the prior version from underneath first. Maybe in the short term utilities such as TCL, perl, and x11 should be registered in /var/db/pkg even if they are part of the base installation so they could be removed later? It means other components which require these items would also have to be registered as a package. One hitch in the FreeBSD package system that SGI doesn't have is the concept of "upgrade". If I replace version 2.8 of something with version 2.9, to cleanly do it now I have to remove 2.8. But if other packages depend on it, then I can't delete it without deleting everything that depends on it. Of course there is always the problem of packages that work with 2.8 and not 2.9... Supposedly SGI tests for this before releasing the new one. -- David Kelly N4HHE, dkelly@hiwaay.net ===================================================================== The human mind ordinarily operates at only ten percent of its capacity -- the rest is overhead for the operating system. From owner-freebsd-current Sat Aug 2 19:17:00 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA28139 for current-outgoing; Sat, 2 Aug 1997 19:17:00 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA28132 for ; Sat, 2 Aug 1997 19:16:52 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id TAA17421; Sat, 2 Aug 1997 19:15:56 -0700 (PDT) To: dkelly@hiwaay.net cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 20:42:00 CDT." <199708030144.UAA04956@nexgen.hiwaay.net> Date: Sat, 02 Aug 1997 19:15:55 -0700 Message-ID: <17417.870574555@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > I trimmed the individual addresses from this reply but wonder if it really > should be going to (3) FreeBSD lists? Absolutely not (as stated in the mailing list charters in the handbook). Trimmed to -current only. Most of us don't need to see 3 or more copies of everything. :) > Every component SGI provides is registered in their installation history. > Some are not optional. Others are. And the inst tool knows what is required > for what, and when a newer version is installed it knows what to delete. That's basically the idea for Son Of Sysinstall. Jordan From owner-freebsd-current Sat Aug 2 19:50:17 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id TAA29726 for current-outgoing; Sat, 2 Aug 1997 19:50:17 -0700 (PDT) Received: from emout04.mail.aol.com (emout04.mx.aol.com [198.81.11.95]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id TAA29720 for ; Sat, 2 Aug 1997 19:50:14 -0700 (PDT) From: StevenR362@aol.com Received: (from root@localhost) by emout04.mail.aol.com (8.7.6/8.7.3/AOL-2.0.0) id WAA23563; Sat, 2 Aug 1997 22:49:43 -0400 (EDT) Date: Sat, 2 Aug 1997 22:49:43 -0400 (EDT) Message-ID: <970802224942_-356235525@emout04.mail.aol.com> To: jkh@time.cdrom.com, davidn@unique.usn.blaze.net.au cc: current@freebsd.org Subject: Re: ports-current/packages-current discontinued Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk In a message dated 97-08-02 20:24:04 EDT, jkh@time.cdrom.com (Jordan K. Hubbard) writes: > 1. TCL needs, *at some point* (and note that the current move was > rather premature, but let's not debate that here) to be part of > the base system so that the installation tools can use it. > We do intend on being heavy users of TCL, if not right this minute > then in the future. > > 2. If it were in ports, we'd have a build problem since you wouldn't > be able to build /usr/src/usr.sbin/setup (not existant yet, but > it will be) without first building and installing a port. This > would break the world target. Basically, the same thing that was done recently with the Handbook and it's tool dependencies needs to be done with the install tools. You move as much as is practical out of the /usr/src/release directories into a separate hierarchy. The new setup tool should be written in such a generic data driven manner that it can install all existing versions of FreeBSD or any future version by just updating some configuration files. These files can be put in /usr/share/examples just like cvsup does. In fact, now that I think of it cvsup would be a great model for a setup/install tool. Maybe you should twist John Polstra's arm to write the new setup tool in Modula 3 ;) and No it should not be added to the base system. Then we can rip both perl and tcl out of the base system. The true perl afficionados surely can't be using version 4.36 can they? Nor does anything currently depend on tcl in the base system. An install/setup tool truly does not need to be part of the base system either. It fits quite well in /stand and is very static for a specific release. Again, cvsup is a perfect example of a crucial install/setup program that is not part of the base system but is used to install multiple different FreeBSD versions. STeve From owner-freebsd-current Sat Aug 2 20:01:34 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA00298 for current-outgoing; Sat, 2 Aug 1997 20:01:34 -0700 (PDT) Received: from mail.webspan.net (root@mail.webspan.net [206.154.70.7]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA00293 for ; Sat, 2 Aug 1997 20:01:32 -0700 (PDT) Received: from orion.webspan.net (orion.webspan.net [206.154.70.5]) by mail.webspan.net (WEBSPAN/970608) with ESMTP id XAA26586; Sat, 2 Aug 1997 23:01:25 -0400 (EDT) Received: from orion.webspan.net (localhost [127.0.0.1]) by orion.webspan.net (WEBSPAN/970608) with ESMTP id XAA13470; Sat, 2 Aug 1997 23:01:25 -0400 (EDT) To: "Jordan K. Hubbard" cc: dkelly@hiwaay.net, current@FreeBSD.ORG From: "Gary Palmer" Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 19:15:55 PDT." <17417.870574555@time.cdrom.com> Date: Sat, 02 Aug 1997 23:01:24 -0400 Message-ID: <13465.870577284@orion.webspan.net> Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk "Jordan K. Hubbard" wrote in message ID <17417.870574555@time.cdrom.com>: > That's basically the idea for Son Of Sysinstall. What an apt name. Lets just hope it comes in time to answer the SOS of people reading this **** flame war. I'll admit I use perl. Why? Because at a past job it was used for a lot of stuff. Never used TCL (although I have a book on it at home which should get read sometime). If push came to shove, I'd prefer NOT having either of them installed by default. There is no clean way to ensure that we don't have recurrances of the current situation that we have with perl4 (and others) festering nicely in /usr/src. However, any move which will (later) allow us to rip such stuff out screaming is good. If SOS, down the road, can give us the necessary compartmentalism to do this, then a few months of pulled hair will be worth it. Gary -- Gary Palmer FreeBSD Core Team Member FreeBSD: Turning PC's into workstations. See http://www.FreeBSD.ORG/ for info From owner-freebsd-current Sat Aug 2 20:17:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA01148 for current-outgoing; Sat, 2 Aug 1997 20:17:43 -0700 (PDT) Received: from hwcn.org (main.hwcn.org [199.212.94.65]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA01141 for ; Sat, 2 Aug 1997 20:17:41 -0700 (PDT) Received: from james.freenet.hamilton.on.ca (ac199@james.hwcn.org [199.212.94.66]) by hwcn.org (8.8.6/8.8.6) with ESMTP id XAA15587; Sat, 2 Aug 1997 23:18:14 -0400 (EDT) Received: from localhost (ac199@localhost) by james.freenet.hamilton.on.ca (8.8.6/8.8.6) with SMTP id XAA05964; Sat, 2 Aug 1997 23:18:27 -0400 (EDT) X-Authentication-Warning: james.freenet.hamilton.on.ca: ac199 owned process doing -bs Date: Sat, 2 Aug 1997 23:18:26 -0400 (EDT) From: Tim Vanderhoek X-Sender: ac199@james.freenet.hamilton.on.ca Reply-To: hoek@hwcn.org To: Ade Lovett cc: hoek@hwcn.org, current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 3 Aug 1997, Ade Lovett wrote: > Rather than implement a meta-port concept, which does nothing but > bundle ports together, we add in (a) the idea that a port may not > build anything itself, and (b) the idea of dependencies in both > source *and* binary form. This is what I've proposed except without the binary dependencies. The reason for not adding automated package is that the pkg_* system has no way of knowing how to fetch a package from the 'net. Eg. I will sometimes download a package through the FreeNet, which means that I, 1. tip hwfn 2. J^H^H^H^H^H^H^H^H^H^H^H^H^H^Hftp://ftp.freebsd.org/pub/FreeBSD/packages-2.2.2/All 3. Choose a package 4. Download it with Zmodem. pkg_add has no way of knowing how to do this. Rather than adding package (or "binary") dependancies, my suggestion was that bsd.port.mk would, if META=yes, put all the dependancy binaries into the .tgz file resulting from ``make package''. > For example: > > port-A only has a list of dependencies (port-B) > > port-B builds something itself, and also depends > on port-B1 and port-B2 > > port-B1 only available in /binary form (say it's > a simple list of configuration files) > > port-B2 available in both /source and /binary > form, no need to grab the /source form > if the /binary is already present on the system I have a suspicion you lack some familiarity with the existing ports system. :) Regardless, in concept, the above is what I am suggesting. I want a new ports category, entitled "meta" (or whatever). Sysinstall would allow the user to choose which meta-packages they want installed. It would do this in such a way that these meta-packages seem like equal peer parts of the system. The fallings I see are A user may want the source to the "meta" packages. The current ports system doesn't provide a way to download both the binary and the source. Sysinstall would have to recognize meta-packages and do pkg_add followed by ``MASTER_SITE_FREEBSD=yes make extract''. This is entirely unintegrated with ``make world''. Someone running make world may or may not want all of their meta-packages recompiled. I think the answer here is that make world should rebuild everything, including all the ports that are installed w/ source. I suppose this would involve checking /var/db/pkg for installed ports, and then building those ports (provided they were installed as ports, not packages). This leaves no problems wrt some users having a T1, some users being occasional dial-up, and some users installing from a CD. Other problems Jordan listed. Why can't distfiles be kept where they are? As for CDs. Optimally, you'd want to get as many of the meta-packages onto the same CD that holds the rest base system, so they could be pkg_added without shuffling CDs. Building srcs off CDs is tricker. This is something that the ports system would have to solve. Not all ports need to be able to do it, but there would need to be a mechanism to mark which ones are capable of it. There would have to be some interesting changes to the ports system, but they would probably have the spin-off benefit of making the whole thing cleaner. I don't want to propose a way of doing this now, but perhaps someone else has some previously thought-out ideas. :-) -- Outnumbered? Maybe. Outspoken? Never! tIM...HOEk From owner-freebsd-current Sat Aug 2 20:33:49 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA01661 for current-outgoing; Sat, 2 Aug 1997 20:33:49 -0700 (PDT) Received: from hwcn.org (main.hwcn.org [199.212.94.65]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA01655 for ; Sat, 2 Aug 1997 20:33:46 -0700 (PDT) Received: from james.freenet.hamilton.on.ca (ac199@james.hwcn.org [199.212.94.66]) by hwcn.org (8.8.6/8.8.6) with ESMTP id XAA17284; Sat, 2 Aug 1997 23:34:19 -0400 (EDT) Received: from localhost (ac199@localhost) by james.freenet.hamilton.on.ca (8.8.6/8.8.6) with SMTP id XAA07887; Sat, 2 Aug 1997 23:34:33 -0400 (EDT) X-Authentication-Warning: james.freenet.hamilton.on.ca: ac199 owned process doing -bs Date: Sat, 2 Aug 1997 23:34:33 -0400 (EDT) From: Tim Vanderhoek X-Sender: ac199@james.freenet.hamilton.on.ca Reply-To: hoek@hwcn.org To: "Jordan K. Hubbard" cc: David Nugent , current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <17153.870568174@time.cdrom.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 2 Aug 1997, Jordan K. Hubbard wrote: > But it doesn't work in a stand-alone environment. I wish people would > stop assuming that everyone in the world has a T1 to their desks or a > CDROM full of convenient tarballs. It's just not the case and arguing > that ports is a complete replacement for /usr/src is just naive in the > extreme. I've said it before and I'll say it again - ports would make > a fine mechanism for this IF IT WERE IMPROVED to deal with those sorts > of "where are my sources?" issues, but that's not the case today. How does the ports system fail? Packages don't need sources. They are the moral equivalent to sysinstall adding only the binaries. When they do need sources, their's MASTER_SITE_OVERRIDE (which can be set to file://cdrom/distfiles). I've also always wanted an option to prevent bsd.port.mk from copying the fetched distfile insto ${DISTDIR}, so that could be added (necessary to prevent copying sources into /usr/ports/distfiles when MASTER_SITE_OVERRIDE=ftp://ftp.br.freebsd.org/pub/FreeBSD/). I'm keeping a mental list of extensions needed to bsd.port.mk and friends, and the idea involves more somewhat more coding than I initially realized (but I never suggested this as an "instant" solution, anyways), but I haven't seen anything absolutely unsolveable, yet. And really, the benefits go far beyond a solution to the "bloated-system" fight (which really doesn't interest me). Consider the person who was just complaining that there was no easy way to rip-out his old X installation. Had he used the port, it would be a simple pkg_delete. -- Outnumbered? Maybe. Outspoken? Never! tIM...HOEk From owner-freebsd-current Sat Aug 2 20:41:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA02075 for current-outgoing; Sat, 2 Aug 1997 20:41:03 -0700 (PDT) Received: from earth.mat.net (root@earth.mat.net [206.246.122.2]) by hub.freebsd.org (8.8.5/8.8.5) with SMTP id UAA02047; Sat, 2 Aug 1997 20:40:58 -0700 (PDT) Received: from Journey2.mat.net (journey2.mat.net [206.246.122.116]) by earth.mat.net (8.6.12/8.6.12) with SMTP id XAA09522; Sat, 2 Aug 1997 23:36:25 -0400 Date: Sat, 2 Aug 1997 23:36:41 -0400 (EDT) From: Chuck Robey X-Sender: chuckr@Journey2.mat.net To: "Jordan K. Hubbard" cc: dmaddox@scsn.net, David Nugent , Michael Smith , Satoshi Asami , andreas@klemm.gtn.com, ports@FreeBSD.ORG, current@FreeBSD.ORG, stable@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <15692.870549801@time.cdrom.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sat, 2 Aug 1997, Jordan K. Hubbard wrote: > > Most of this argument is just silly. Even the most vehement anti- > > bloatists don't consider perl to be "an evil, bloated monster"; they just > > consider _it's inclusion in the base distribution_ to be AEBM. While you > > I don't see the difference from the POV of this discussion so this > paragraph of yours doesn't really parse for me. > > > In any case, I see none of this bloatist v. antibloatist propaganda > > as cogent here. Tcl should not be part of the base system because it > > It's imminently cogent - this is NOT just a technical issue, it's > an emotional one, and if you think that all software decisions are > made on purely technical merits then I have a certain tower in Paris > which I could make you a _great_ deal on. ;-) Just want one clarification. Most of you who've talked with me know I am not one of the anti-bloatists (I like _big_ disks), so I want to make sure that having tcl out of the tree is not singularly associated with being an anti-bloatist. Everyone knows that tcl and tk are like siamese twins, bodily joined, and only separated at great risk. Don't tell me that we can't have tk in the tree, I know it requires X, and I'm not suggesting that. I'm saying that tcl requires tk, and if we can't have tk, we mustn't have tcl. All the other BSD environments are very friendly to tcl applications, so why does FreeBSD have to have such a tcl unfriendly environment? If we didn't have tcl ports, the complaint level would be even higher. It just seems like a scene from Catch-22. Why is FreeBSD a hard system to build tcl apps on? Because we like tcl ... does that make sense? This argument is not really centered on being bloatist, at least not totally. I would fight taking perl out of the kernel (I want perl5.004 brought in) but I'll be pleased to see tcl make an exit. Might a compromise be made, let tcl go away, in exchange for updating perl? ----------------------------+----------------------------------------------- Chuck Robey | Interests include any kind of voice or data chuckr@eng.umd.edu | communications topic, C programming, and Unix. 213 Lakeside Drive Apt T-1 | Greenbelt, MD 20770 | I run Journey2 and picnic, both FreeBSD (301) 220-2114 | version 3.0 current -- and great FUN! ----------------------------+----------------------------------------------- From owner-freebsd-current Sat Aug 2 20:56:56 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id UAA02668 for current-outgoing; Sat, 2 Aug 1997 20:56:56 -0700 (PDT) Received: from sendero.i-connect.net (sendero-ppp.i-Connect.Net [206.190.143.100]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id UAA02657; Sat, 2 Aug 1997 20:56:52 -0700 (PDT) Received: (from shimon@localhost) by sendero.i-connect.net (8.8.6/8.8.5) id UAA14882; Sat, 2 Aug 1997 20:57:07 -0700 (PDT) Message-ID: X-Mailer: XFMail 1.2-alpha [p0] on FreeBSD Content-Type: text/plain; charset=iso-8859-8 Content-Transfer-Encoding: 8bit MIME-Version: 1.0 Date: Sat, 02 Aug 1997 20:57:07 -0700 (PDT) Organization: Atlas Telecom From: Simon Shapiro To: FreeBSD-SCSI@FreeBSD.ORG, FreeBSD-Current@FreeBSD.ORG Subject: Kernel Build Fails - AHC Error Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk Hi Y'all, I know justin is on vacation, but did any of you encounter: cd ../../dev/aic7xxx; make obj; make BINDIR=/usr/src/3.0/src/sys/compile/SENDERO-smp all install yacc -d /usr/src/3.0/src/sys/dev/aic7xxx/gram.y /usr/obj/usr/src/3.0/src/sys/dev/aic7xxx created for /usr/src/3.0/src/sys/dev/aic7xxx yacc -d /usr/src/3.0/src/sys/dev/aic7xxx/gram.y mv y.tab.c gram.c mv y.tab.c gram.c cd ../../dev/aic7xxx; make obj; make BINDIR=/usr/src/3.0/src/sys/compile/SENDERO all install cc -O -I. -c gram.c mv: y.tab.c: No such file or directory *** Error code 1 This is in current as of today. Simon From owner-freebsd-current Sat Aug 2 21:15:28 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id VAA03142 for current-outgoing; Sat, 2 Aug 1997 21:15:28 -0700 (PDT) Received: from unique.usn.blaze.net.au (root@unique.usn.blaze.net.au [203.17.53.17]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id VAA03134 for ; Sat, 2 Aug 1997 21:15:14 -0700 (PDT) Received: from unique.usn.blaze.net.au (davidn@local [127.0.0.1]) by unique.usn.blaze.net.au (8.8.6/8.8.5) with ESMTP id OAA20765; Sun, 3 Aug 1997 14:14:33 +1000 (EST) Message-Id: <199708030414.OAA20765@unique.usn.blaze.net.au> To: hoek@hwcn.org cc: current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 23:34:33 -0400." X-Face: (W@z~5kg?"+5?!2kHP)+l369.~a@oTl^8l87|/s8"EH?Uk~P#N+Ec~Z&@;'LL!;3?y Date: Sun, 03 Aug 1997 14:14:32 +1000 From: David Nugent Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > How does the ports system fail? Yeah, I'd like to know too. I've never had a T1 on my desktop, but it has always worked fine for me, either from CDROM or over the internet, even at 14.4k as I personally was until recently. And I rarely even use pre-built packages. > Packages don't need sources. They are the moral equivalent to > sysinstall adding only the binaries. When they do need sources, > their's MASTER_SITE_OVERRIDE (which can be set to > file://cdrom/distfiles). I've also always wanted an option to > prevent bsd.port.mk from copying the fetched distfile insto > ${DISTDIR}, so that could be added (necessary to prevent copying > sources into /usr/ports/distfiles when > MASTER_SITE_OVERRIDE=ftp://ftp.br.freebsd.org/pub/FreeBSD/). Doesn't a symbolic link from /usr/ports/distfiles to the CDROM work? This is admittedly problematic if you keep your ports tree up to date, but it's a reasonable workaround I've used occasionally. :) Regards, David David Nugent - Unique Computing Pty Ltd - Melbourne, Australia davidn@freebsd.org davidn@blaze.net.au http://www.blaze.net.au/~davidn/ From owner-freebsd-current Sat Aug 2 22:35:03 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA05946 for current-outgoing; Sat, 2 Aug 1997 22:35:03 -0700 (PDT) Received: from linkdead.paranoia.com (linkdead.paranoia.com [204.145.225.245]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA05925 for ; Sat, 2 Aug 1997 22:34:57 -0700 (PDT) Received: from localhost (localhost [127.0.0.1]) by linkdead.paranoia.com (8.7.6/8.7.3) with SMTP id AAA24396; Sun, 3 Aug 1997 00:28:20 -0500 (CDT) Message-Id: <199708030528.AAA24396@linkdead.paranoia.com> X-Authentication-Warning: linkdead.paranoia.com: Host localhost [127.0.0.1] didn't use HELO protocol To: netbsd-users@netbsd.org, freebsd-current@freebsd.org, misc@openbsd.org, mail2news@news.news.demon.net Newsgroups: comp.unix.shell Subject: configuration file tricks of the masters Date: Sun, 03 Aug 1997 00:28:08 -0500 From: VaX#n8 Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I'm revising my dot-files this weekend and was thinking of compiling sample configuration files, dot files, rc files, and other customizations as a means of passing knowledge on to newbies, if it would not be duplicate effort. Focus would be on files which have a wide audience and portability. I am interested in: o clever hacks o generating multiple customized copies from one master o mixing automatically-generated and user-generated data o portable shell coding techniques Comments and submissions welcome. -- VaX#n8 http://www.paranoia.com/~vax League of Non-aligned Wizards From owner-freebsd-current Sat Aug 2 22:37:14 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA06086 for current-outgoing; Sat, 2 Aug 1997 22:37:14 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA06081 for ; Sat, 2 Aug 1997 22:37:11 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id WAA00540; Sat, 2 Aug 1997 22:36:46 -0700 (PDT) To: hoek@hwcn.org cc: David Nugent , current@freebsd.org Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 23:34:33 EDT." Date: Sat, 02 Aug 1997 22:36:46 -0700 Message-ID: <536.870586606@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > How does the ports system fail? > > Packages don't need sources. They are the moral equivalent to We're not talking about packages here - we're talking about sources. Jordan From owner-freebsd-current Sat Aug 2 22:38:57 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA06164 for current-outgoing; Sat, 2 Aug 1997 22:38:57 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA06159 for ; Sat, 2 Aug 1997 22:38:55 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id WAA00555; Sat, 2 Aug 1997 22:38:22 -0700 (PDT) To: Chuck Robey cc: dmaddox@scsn.net, David Nugent , current@freebsd.org Subject: Re: ports-current/packages-current discontinued In-reply-to: Your message of "Sat, 02 Aug 1997 23:36:41 EDT." Date: Sat, 02 Aug 1997 22:38:22 -0700 Message-ID: <552.870586702@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk [What's it going to take to get people to trim the headers, electric shocks? ;-) Trimmed yet again to include only -current] > anti-bloatist. Everyone knows that tcl and tk are like siamese twins, False. I've used TCL without a hint of Tk anywhere many times. It's more than generally useful as a scripting language, with or without a GUI. > that. I'm saying that tcl requires tk, and if we can't have tk, we > mustn't have tcl. All the other BSD environments are very friendly to tcl Sorry, but this argument is bogus. Jordan From owner-freebsd-current Sat Aug 2 22:50:47 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id WAA06590 for current-outgoing; Sat, 2 Aug 1997 22:50:47 -0700 (PDT) Received: from panix.com (panix.com [198.7.0.2]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id WAA06585 for ; Sat, 2 Aug 1997 22:50:44 -0700 (PDT) Received: from localhost (jbarrm@localhost) by panix.com (8.8.5/8.7/PanixU1.3) with SMTP id BAA23127 for ; Sun, 3 Aug 1997 01:50:42 -0400 (EDT) Date: Sun, 3 Aug 1997 01:50:41 -0400 (EDT) From: Barry Masterson To: FreeBSD-current Subject: XFree86-3.3, FBSD-2.2.2 & ViRGE/VX Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk I seem to remember reading on the -questions mailing list that XFree86-3.3 was not fully supported on the FreeBSD 2.2.2-R system. I'm running FreeBSD 2.2.2-R on a ASUS P55T2P4, with a Adaptec 2940AU. THe graphics card is a #9 9FX Reality 772 (ViRGE/VX). I built XFree86-3.3 (from the source code) because it offers better support for the ViRGE/VX chipset. My question is, Is there anything specific to FreeBSD 2.2.2-R that needs to be changed to best run XFree86-3.3; such as something from the FreeBSD-current source? Or, I should ask, is there something about 2.2.2-R that might cause the situation described below. The reason I'm asking, is that while 3.3 runs fine on this system, the XF86_SVGA server fails to restart after the first 'startx' session is closed on a freshly booted system. (running startx a second time results on a frozen keyboard and a blank screen). The author of XF86_S3V & XF86_SVGA is aware of this problem. They supplied me with the patches to fix it, but after applying them, and rebuilding, the 'no second restart' problem is still present. (The XF86_SVGA author can't reproduce the lockup on the patched 3.3.1 server) Luckily, the XF86_S3V server is ok. However, I've been told that the SVGA server is better, and that S3V will at some point be phased out. In an attempt to cover the bases, I thought I should ask this list if FreeBSD 2.2.2-R needs anything from the -current source tree to make the 3.3 or the 3.3.1 patched XF86_SVGA server happy. Please cc, not currently on -current. Thank you for listening. Barry Masterson jbarrm@panix.com >--->--->--->--->---> FreeBSD 2.2.2-R <---<---<---<---<---< From owner-freebsd-current Sat Aug 2 23:25:13 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA07594 for current-outgoing; Sat, 2 Aug 1997 23:25:13 -0700 (PDT) Received: from scanner.worldgate.com (scanner.worldgate.com [198.161.84.3]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA07589 for ; Sat, 2 Aug 1997 23:25:11 -0700 (PDT) Received: from znep.com (uucp@localhost) by scanner.worldgate.com (8.8.5/8.8.5) with UUCP id AAA18818; Sun, 3 Aug 1997 00:24:33 -0600 (MDT) Received: from localhost (marcs@localhost) by alive.znep.com (8.7.5/8.7.3) with SMTP id AAA16305; Sun, 3 Aug 1997 00:23:25 -0600 (MDT) Date: Sun, 3 Aug 1997 00:23:24 -0600 (MDT) From: Marc Slemko To: David Nugent cc: hoek@hwcn.org, current@FreeBSD.ORG Subject: Re: ports-current/packages-current discontinued In-Reply-To: <199708030414.OAA20765@unique.usn.blaze.net.au> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, 3 Aug 1997, David Nugent wrote: > > Packages don't need sources. They are the moral equivalent to > > sysinstall adding only the binaries. When they do need sources, > > their's MASTER_SITE_OVERRIDE (which can be set to > > file://cdrom/distfiles). I've also always wanted an option to > > prevent bsd.port.mk from copying the fetched distfile insto > > ${DISTDIR}, so that could be added (necessary to prevent copying > > sources into /usr/ports/distfiles when > > MASTER_SITE_OVERRIDE=ftp://ftp.br.freebsd.org/pub/FreeBSD/). > > Doesn't a symbolic link from /usr/ports/distfiles to the CDROM work? > This is admittedly problematic if you keep your ports tree up to > date, but it's a reasonable workaround I've used occasionally. :) A lndir should do the trick most of the time to keep /usr/ports/distfiles pointing to the cdrom but still allow things to be updated. From owner-freebsd-current Sat Aug 2 23:26:47 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA07641 for current-outgoing; Sat, 2 Aug 1997 23:26:47 -0700 (PDT) Received: from zibbi.mikom.csir.co.za (zibbi.mikom.csir.co.za [146.64.24.58]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA07636 for ; Sat, 2 Aug 1997 23:26:43 -0700 (PDT) Received: (from jhay@localhost) by zibbi.mikom.csir.co.za (8.8.6/8.8.5) id IAA06139 for current@freebsd.org; Sun, 3 Aug 1997 08:26:39 +0200 (SAT) From: John Hay Message-Id: <199708030626.IAA06139@zibbi.mikom.csir.co.za> Subject: clock_settime() error? To: current@freebsd.org Date: Sun, 3 Aug 1997 08:26:39 +0200 (SAT) X-Mailer: ELM [version 2.4ME+ PL31 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Hi, I have been playing with the latest xntpd and I got "Can't set time of day: Invalid argument" errors when it tried to step the time. I have traced it to the clock_settime() system call in kern_time.c. I think the test on line 171 is invalid: ------- if ((error = copyin(SCARG(uap, tp), &ats, sizeof(ats))) != 0) return (error); if (atv.tv_usec < 0 || ats.tv_nsec >= 1000000000) ^^^^^^^^^^^ return (EINVAL); TIMESPEC_TO_TIMEVAL(&atv, &ats); ------- At that time atv.tv_usec hasn't been assigned yet. It is only assigned a value two lines late in the TIMESPEC_TO_TIMEVAL() macro. Shouldn't it be ats.tv_sec? Or ats.tv_nsec maybe? Does anybody know what the check should be? John -- John Hay -- John.Hay@mikom.csir.co.za From owner-freebsd-current Sat Aug 2 23:27:00 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA07680 for current-outgoing; Sat, 2 Aug 1997 23:27:00 -0700 (PDT) Received: from rf900.physics.usyd.edu.au (rf900.physics.usyd.edu.au [129.78.129.109]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA07667 for ; Sat, 2 Aug 1997 23:26:54 -0700 (PDT) Received: (from dawes@localhost) by rf900.physics.usyd.edu.au (8.8.5/8.8.2) id QAA26840; Sun, 3 Aug 1997 16:26:30 +1000 (EST) Message-ID: <19970803162630.61774@rf900.physics.usyd.edu.au> Date: Sun, 3 Aug 1997 16:26:30 +1000 From: David Dawes To: Barry Masterson Cc: FreeBSD-current Subject: Re: XFree86-3.3, FBSD-2.2.2 & ViRGE/VX References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.69 In-Reply-To: ; from Barry Masterson on Sun, Aug 03, 1997 at 01:50:41AM -0400 Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk On Sun, Aug 03, 1997 at 01:50:41AM -0400, Barry Masterson wrote: >I seem to remember reading on the -questions mailing list that >XFree86-3.3 was not fully supported on the FreeBSD 2.2.2-R >system. XFree86 3.3 is fully supported on FreeBSD 2.2.2-R. I do 90% of my XFree86 development work on machines running 2.2.2 and 2.1.7, with the remaining 10% split between FreeBSD current, Linux and SVR4. >I'm running FreeBSD 2.2.2-R on a ASUS P55T2P4, with a Adaptec >2940AU. THe graphics card is a #9 9FX Reality 772 (ViRGE/VX). > >I built XFree86-3.3 (from the source code) because it offers >better support for the ViRGE/VX chipset. > >My question is, Is there anything specific to FreeBSD 2.2.2-R >that needs to be changed to best run XFree86-3.3; such as >something from the FreeBSD-current source? Or, I should ask, >is there something about 2.2.2-R that might cause the situation >described below. No. The situation you are describing almost certainly has nothing to do with the OS. >The reason I'm asking, is that while 3.3 runs fine on this system, >the XF86_SVGA server fails to restart after the first 'startx' >session is closed on a freshly booted system. (running startx >a second time results on a frozen keyboard and a blank screen). > >The author of XF86_S3V & XF86_SVGA is aware of this problem. They >supplied me with the patches to fix it, but after applying them, >and rebuilding, the 'no second restart' problem is still present. >(The XF86_SVGA author can't reproduce the lockup on the patched >3.3.1 server) The main XFree86 S3 ViRGE developers are running Linux and OS/2(!). The reason they can't reproduce the problem you are seeing is more likely that they are using a different ViRGE card to the one you are using. Are you absolutely certain that you've been testing the patched version? I hate to ask that, but it wouldn't be the first time someone has reported a fix not working because they were still running the older version without realising it. David From owner-freebsd-current Sat Aug 2 23:38:08 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.5/8.8.5) id XAA08145 for current-outgoing; Sat, 2 Aug 1997 23:38:08 -0700 (PDT) Received: from time.cdrom.com (root@time.cdrom.com [204.216.27.226]) by hub.freebsd.org (8.8.5/8.8.5) with ESMTP id XAA08139 for ; Sat, 2 Aug 1997 23:38:05 -0700 (PDT) Received: from time.cdrom.com (jkh@localhost.cdrom.com [127.0.0.1]) by time.cdrom.com (8.8.6/8.6.9) with ESMTP id XAA00692; Sat, 2 Aug 1997 23:37:42 -0700 (PDT) To: Barry Masterson cc: FreeBSD-current Subject: Re: XFree86-3.3, FBSD-2.2.2 & ViRGE/VX In-reply-to: Your message of "Sun, 03 Aug 1997 01:50:41 EDT." Date: Sat, 02 Aug 1997 23:37:42 -0700 Message-ID: <688.870590262@time.cdrom.com> From: "Jordan K. Hubbard" Sender: owner-freebsd-current@FreeBSD.ORG X-Loop: FreeBSD.org Precedence: bulk > I seem to remember reading on the -questions mailing list that > XFree86-3.3 was not fully supported on the FreeBSD 2.2.2-R > system. It's not the default XFree86 for it, but it's more than usable with 2.2.2. I don't know who told you it wasn't supported. > My question is, Is there anything specific to FreeBSD 2.2.2-R > that needs to be changed to best run XFree86-3.3; such as Nope. > The reason I'm asking, is that while 3.3 runs fine on this system, > the XF86_SVGA server fails to restart after the first 'startx' > session is closed on a freshly booted system. (running startx > a second time results on a frozen keyboard and a blank screen). That's weird - I can't say that I've seen this happen with my own XF86_SVGA server (running 2.2-stable). Jordan