From owner-freebsd-current Sun Aug 5 19:33:11 2001 Delivered-To: freebsd-current@freebsd.org Received: from smtp015.mail.yahoo.com (smtp015.mail.yahoo.com [216.136.173.59]) by hub.freebsd.org (Postfix) with SMTP id 766A037B401 for ; Sun, 5 Aug 2001 19:33:03 -0700 (PDT) (envelope-from kc5vdj@yahoo.com) Received: from mkc-65-28-47-209.kc.rr.com (HELO yahoo.com) (65.28.47.209) by smtp.mail.vip.sc5.yahoo.com with SMTP; 6 Aug 2001 02:33:03 -0000 X-Apparently-From: Message-ID: <3B6E01DE.A78FB3FE@yahoo.com> Date: Sun, 05 Aug 2001 21:33:02 -0500 From: Jim Bryant Reply-To: kc5vdj@yahoo.com X-Mailer: Mozilla 4.77 [en] (X11; U; Linux 2.2.12 i386) X-Accept-Language: en MIME-Version: 1.0 To: Gordon Tetlow Cc: current@FreeBSD.ORG Subject: Re: ahc fails to attach in -current (was: snapshot installation woes) References: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG Gordon Tetlow wrote: > > On Sat, 4 Aug 2001, Gordon Tetlow wrote: > > > Sure enough, that fixed the kernel panic, but here's the next odd piece, > > my hard drive wasn't showing up! I have a rather standard Adaptec AHA-2940 > > dmesg reports that ahc0 is there. The lines from the dmesg are (hand > > typed): > > > > ahc0: port 0x5000-0x50ff mem 0x50000000-0x50000fff irq 15 at device 15.0 on pci0 > > device_probe_and_attach: ahc0 attach returned 12 > > > > errno.h says ENOMEM is 12, so it seems that something in the ahc driver is > > unable to allocate memory. Dunno where or why, but something is fouling > > it up. By contrast 4.3-RELEASE doesn't have any issues (I'll try a recent > > snapshot if that would help). Sorry I can't help out any more, but the > > debugging tools in the installation disks seem to lacking > > (understandably). > > Okay, I tried with a 4.4-PRERELEASE bootdisk that was available on > current.jp.freebsd.org and dmesg came up with the following: > > ahc0: port 0x5000-0x50ff mem 0x50000000-0x50000fff irq 15 at device 15.0 on pci0 > aic7880: Wide Channel A, SCSI Id=7, 16/255 SCBs > > Since -stable and -current are using the exact same driver, there is > something more subtle (and sinister?) going on that I can't figure out. At > this point, I'm throwing my hands up in the air unless someone can give me > a better idea as to what the possible problem could be. I'd really like to > try -current on this box as it's a dual proc PPro 200. I don't know if -current and -stable are using the same driver. Something *IS* different. Please see my post from about a month or so back in the aic7xxx group. Nobody responded to this. Where it would boot, with both controllers [3940UW on the MB] under -stable, it came up with the same message on ahc0 then after probing ahc1 and finding something there, reassigning ahc1 to ahc0. I ended up solving my own problem on this, although it did cause yet other problems. I had, because of IRQ issues, disabled the secondary IDE controller in the BIOS, to allow for both SCSI controller sides to have independant IRQs. This was causing the exact same message you note [see the dmesg output in the aic7xxx list]. Interesting to note that even with the secondary IDE controller disabled on the MB, -current failed to recognize this, and then in turn enabled it. Once I turned on the second IDE controller, this whole thing cleared up, albeit, with side-effects [now my Hauppauge WinTV/Theatre card is in conflict, and will cause a spontaneous reboot in both winblowz and fbsd when the TV is activated, although radio works fine]. Are you disabling motherboard IRQs in a similar fashion? Play with your IRQs, and have "Plug and Play" set in your BIOS. jim -- ET has one helluva sense of humor! He's always anal-probing right-wing schizos! _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message