From owner-freebsd-smp Thu Mar 23 20:50:13 2000 Delivered-To: freebsd-smp@freebsd.org Received: from skippyii.compar.com (mail.compar.com [216.208.38.130]) by hub.freebsd.org (Postfix) with ESMTP id 4D9E237B5CD; Thu, 23 Mar 2000 20:49:54 -0800 (PST) (envelope-from matt@gsicomp.on.ca) Received: from matt (HSE-Kitchener-ppp84482.sympatico.ca [216.209.96.51]) by skippyii.compar.com (8.9.3/8.9.1) with SMTP id XAA13460; Thu, 23 Mar 2000 23:54:19 -0500 (EST) (envelope-from matt@gsicomp.on.ca) Message-ID: <001401bf954c$5fd5bf40$1200a8c0@gsicomp.on.ca> From: "Matthew Emmerton" To: , Subject: Problems with 3.4-RELEASE SMP + AIC7870 Date: Thu, 23 Mar 2000 23:49:13 -0500 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.6600 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600 Sender: owner-freebsd-smp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org I'm running a 3.4-RELEASE SMP system, equipped with 2xP133 and an onboard AIC7870 PCI SCSI controller, and recently have been plagued by a whack of SCB errors. dmesg (relevant parts only) --------------------------- FreeBSD 3.4-RELEASE #0: Sun Jan 16 18:14:14 EST 2000 root@gabby.gsicomp.on.ca:/usr/src-3.4/sys/compile/GABBY.20000116.01 Timecounter "i8254" frequency 1193182 Hz CPU: Pentium/P54C (586-class CPU) Origin = "GenuineIntel" Id = 0x52c Stepping = 12 Features=0x3bf real memory = 50331648 (49152K 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: 0x000f0011, at 0xfec00000 Preloaded elf kernel "kernel" at 0xc02d4000. Preloaded userconfig_script "/boot/kernel.conf" at 0xc02d409c. ahc0: rev 0x03 int a irq 11 on pci0.11.0 ahc0: aic7870 Wide Channel A, SCSI Id=7, 16/255 SCBs Intel Pentium detected, installing workaround for F00F bug APIC_IO: Testing 8254 interrupt delivery APIC_IO: routing 8254 via pin 2 IP packet filtering initialized, divert enabled, rule-based forwarding disabled, logging disabled Waiting 15 seconds for SCSI devices to settle da0 at ahc0 bus 0 target 0 lun 0 da0: Fixed Direct Access SCSI-2 device da0: 10.000MB/s transfers (10.000MHz, offset 15), Tagged Queueing Enabled da0: 2049MB (4197405 512 byte sectors: 64H 32S/T 2049C) cd0 at ahc0 bus 0 target 4 lun 0 cd0: Removable CD-ROM SCSI-2 device cd0: 10.000MB/s transfers (10.000MHz, offset 15) cd0: Attempt to query device size failed: NOT READY, Medium not present This configuration was rock-solid for it's "normal" workload, but whenever I attempted to do something that bashed the disk (such as building gcc), I would get errors like these: /kernel: swap_pager: indefinate wait buffer: device 0x30411, blkno: 8688, size 8192 /kernel: (probe0:ahc0:0:0:0): SCB 0xa - timed out in message out phase, SEQADDR == 0x151 The system would then freeze. Then things got worse. To supplement my existing 2 GB Seagate SCSI-2 drive, I dropped in a 1 GB Fujitsu SCSI-2 drive. (It's running in SCSI-1 mode for debugging purposes.) da1 at ahc0 bus 0 target 2 lun 0 da1: Fixed Direct Access SCSI-CCS device da1: 3.300MB/s transfers da1: 1033MB (2117025 512 byte sectors: 64H 32S/T 1033C) Now, I consistently get (on an almost daily basis) messages like the following: /kernel: Timedout SCB handled by another timeout and the system locks for 10-15 seconds, and then goes along it's merry way; in addition, the errors reported above now occur with greater frequency. Scanning the archives, I came upon these threads: http://www.freebsd.org/cgi/getmsg.gsi?fetch=66633+69631+/usr/local/www/db/te xt/1998/freebsd-scsi/19980308.freebsd-scsi http://www.freebsd.org/cgi/getmsg.gsi?fetch=123866+125897+/usr/local/www/db/ text/1998/freebsd-scsi/19980531.freebsd-scsi which suggested UP mode, which seemed to work just fine. However, my thought was the "problem" that was seen 1.5 years ago, presumably using some hybrid 2.2-stable + CAM codebase, would have been fixed by now. (And yes, I've checked my cables and termination.) If something thinks that there definitely is a timing problem or something with the SMP + AIC code, I'd be willing to hack away at it - mind you, "remote gdb" is a totally foreign concept to me. Thanks, -- Matthew Emmerton GSI Computer Services To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message