From owner-freebsd-current Mon Apr 6 20:07:37 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id UAA08649 for freebsd-current-outgoing; Mon, 6 Apr 1998 20:07:37 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from mantar.slip.netcom.com (mantar.slip.netcom.com [192.187.167.134]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id UAA08576 for ; Mon, 6 Apr 1998 20:07:32 -0700 (PDT) (envelope-from mantar@netcom.com) Received: from dual (DUAL [192.187.167.136]) by mantar.slip.netcom.com (8.8.8/8.8.8) with SMTP id UAA24404; Mon, 6 Apr 1998 20:06:17 -0700 (PDT) (envelope-from mantar@netcom.com) Message-Id: <199804070306.UAA24404@mantar.slip.netcom.com> X-Sender: null@mantar.slip.netcom.com X-Mailer: QUALCOMM Windows Eudora Pro Version 4.0 Date: Mon, 06 Apr 1998 20:06:16 -0700 To: Eric "Two Cats" Jones , obanta@unixgeeks.net (Oliver Banta) From: Manfred Antar Subject: Re: Current with XFree86 3.3.2 Cc: freebsd-current@FreeBSD.ORG In-Reply-To: <199804061243.HAA17043@psasolar.psa.pencom.com> References: <199804052149.QAA01036@aenima.unixgeeks.net> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG At 07:43 AM 4/6/98 -0500, Eric "Two Cats" Jones wrote: > >Oliver, > I had what I think is the same problem (although I was >building directly from the XF86 source base, not the port: the port >may avoid this problem). My problem was that the XF86 server >code implements their own malloc, for which they need the hardware >pagesize. The code checks whether to use sysconf(5) for this >info by doing #ifdef _SC_PAGESIZE, which until about 3/28 >was false. After 3/28, this value is defined, however the >backend code would cause sysconf(_SC_PAGESIZE) to return -1, >which XF86 didn't bother to check. It would later use this >pagesize to calculate what size chunk to allocate off the heap, >with disasterous consequences. There's a way to force the >code to use the getpagesize(3) call, but I can't remember >off the top of my head and I've deleted my distribution. Check >xc/programs/Xserver/os/xalloc.c and look for _SC_PAGESIZE... >you should be able to figure it out from there. > I suspect your log messages are coincidental. > >Eric Jones >Collective Technology > >Strange sunspot activity caused Oliver Banta to write: >> I recently installed 3.0-980311-SNAP on a Dual PPro 180. X was >> working fine until I did two things: 1) rebuilt world to -current as >> of April 3 and 2) rebuilt XFree86 3.3.2 from /usr/ports/x11/XFree86/. >> Now whenever I try to start any X server, it core dumps on me. >> >[...] >> >> To make sure it wasn't something dumb on my part, I ran XF86Setup >> which uses the XF86_VGA16 server, and it also core dumped on me. I >> ran this when I originally installed 3.0-SNAP and it worked fine. >> This is the error I get, in case it will help anyone figure out what's >> wrong: >> >> Apr 4 04:03:12 knuth /kernel: forward_hardclock: checkstate 0 >> Apr 4 04:03:13 knuth last message repeated 15 times >> Apr 4 04:03:15 knuth /kernel: pid 12296 (XF86_VGA16), uid 0: exited on signal 6 >> (core dumped) >> >> I am currently rebuilding world and will try again as soon as it >> finishes. I will also try it with SMP not enabled to see if that >> makes any difference. If anyone has any hints or suggestions, I'd >> love to hear them. >> >> Cheers, >> -- Oliver >> I changed line 682 in xc/programs/Xserver/os/xalloc.c to read pagesize = getpagesize(); and the XF86_SVGA server now works. I tried changing the permissions on the file but that didn't work. Changing the above line and rebuilding the server works so far up all day Manfred |=======================| | mantar@netcom.com | | Ph. (415) 681-6235 | |=======================| To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message