From owner-freebsd-hackers Tue Jan 2 08:10:58 1996 Return-Path: owner-hackers Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id IAA25480 for hackers-outgoing; Tue, 2 Jan 1996 08:10:58 -0800 (PST) Received: from doolittle.ycc.Kodak.COM ([129.126.74.2]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id IAA25470 for ; Tue, 2 Jan 1996 08:10:50 -0800 (PST) Received: from crestone.ycc.Kodak.COM by doolittle.ycc.Kodak.COM with SMTP id AA14157 (5.67b/IDA-1.5 for freebsd-hackers@FreeBSD.ORG); Tue, 2 Jan 1996 11:09:21 -0500 Received: from golly.PCD1 (golly.ycc.Kodak.COM) by crestone.ycc.kodak.com with SMTP id AA08104 (5.65c/IDA-1.5 for ); Tue, 2 Jan 1996 11:09:18 -0500 Received: by golly.PCD1 (5.0/SMI-SVR4) id AA28255; Tue, 2 Jan 1996 11:09:17 +0500 Date: Tue, 2 Jan 1996 11:09:17 +0500 From: lyman@ycc.Kodak.COM (Ralph Lyman) Message-Id: <9601021609.AA28255@golly.PCD1> To: freebsd-hackers@FreeBSD.ORG Subject: Problems running a ahc 2940 with a 100Mhz pentium? X-Sun-Charset: US-ASCII Sender: owner-hackers@FreeBSD.ORG Precedence: bulk Hello, I have just upgraded my processor from a 75Mhz pentium to a 100Mhz pentium (it was given to me). That is the only change in the system. The rest of the system consists of... ASUS P54SP4 motherboard W/ latest bios (1.06 I think) 32 Meg of 70ns ram. ahc 2940 controller Number 9 GXE pro video card W/2 Meg one Seagate ST31230 1 gig drive (Dos & Win NT) one Seagate ST42400 2 gig drive. (FreeBSD is on this one) Toshiba 3501b CD-rom drive. The slices are. 25 Meg / 30 Meg /var 64Meg swap ~1 gig /usr Running at 75 MHZ it ran fine running 2.1-Release and 2.2-current. When I put the 100MHZ part in and change the jumpers on the motherboard I get sd1(ahc0:1:0):ILLEGAL REQUEST info?:2028cf asc:20,0 Invalid command operation code sks:c0,0 Sometimes the info?: is 1f60cf. This is in both 2.1-release and 2.2-current. This happens only once in a while and load does not seem to matter. If I let the system just sit there it happens. I have built the kernel with the debugger in it and break in the sd_sense_handler() function so I can trace the error it happens after an interrupt. I don't have the traces with me now so I can't show them but they seems to come from different places (idle, vmsync, ...). I have also been able to dump the scsi command that seems to be the problem and it is always an extended write (WRITE_BIG). Can anyone point me in a direction to look for this? Like I said it does not seem to affect the system but I would like to fix this. It has been a long time since I have looked at scsi code so I am having fun. Ralph Lyman. ----------------------------------------|---------------------- | Ralph D. Lyman | (W) (716) 726-6210 | | Eastman Kodak Co. | | | 901 Elmgrove Road | | | 2/5/EP MC: 35400 | | | Rochester, N.Y. 14653-5400 | lyman@ycc.kodak.com | |---------------------------------------|---------------------|