Scary syslog messages from Adaptac SCSI controller
Jeffrey J. Kosowsky
I am running my RH7.3 system off of several partions of a 20 GB EIDE drive.
However, I also have an old 2 GB SCSI drive hanging off an Adaptac
2940 controller.
When fscking some old ext2 partitions on the SCSI drive, fsck froze
for a couple of minutes and then syslog spit out the following (scary)
messages:
Sep 1 20:19:11 surgeon kernel: scsi0:0:0:0: Attempting to queue an ABORT message
Sep 1 20:19:11 surgeon kernel: scsi0: Dumping Card State in Data-in phase, at SEQADDR 0x7a
Sep 1 20:19:11 surgeon kernel: ACCUM = 0x0, SINDEX = 0x8, DINDEX = 0x8f, ARG_2 = 0xff
Sep 1 20:19:11 surgeon kernel: HCNT = 0xf0 SCBPTR = 0x1
Sep 1 20:19:11 surgeon kernel: SCSISEQ = 0x12, SBLKCTL = 0x0
Sep 1 20:19:11 surgeon kernel: DFCNTRL = 0x78, DFSTATUS = 0x40
Sep 1 20:19:11 surgeon kernel: LASTPHASE = 0x40, SCSISIGI = 0x44, SXFRCTL0 = 0x80
Sep 1 20:19:11 surgeon kernel: SSTAT0 = 0x0, SSTAT1 = 0x2
Sep 1 20:19:11 surgeon kernel: STACK == 0x96, 0x186, 0x156, 0x0
Sep 1 20:19:11 surgeon kernel: SCB count = 4
Sep 1 20:19:11 surgeon kernel: Kernel NEXTQSCB = 0
Sep 1 20:19:11 surgeon kernel: Card NEXTQSCB = 0
Sep 1 20:19:11 surgeon kernel: QINFIFO entries:
Sep 1 20:19:11 surgeon kernel: Waiting Queue entries:
Sep 1 20:19:11 surgeon kernel: Disconnected Queue entries:
Sep 1 20:19:11 surgeon kernel: QOUTFIFO entries:
Sep 1 20:19:11 surgeon kernel: Sequencer Free SCB List: 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sep 1 20:19:11 surgeon kernel: Sequencer SCB Info: 0(c 0x60, s 0x7, l 0, t 0xff) 1(c 0x60, s 0x7, l 0, t 0x1) 2(c 0x0, s 0x0, l 0, t 0xff) 3(c 0x0, s 0x0, l 0, t 0xff) 4(c 0x0, s 0x0, l 0, t 0xff) 5(c 0x0, s 0x0, l 0, t 0xff) 6(c 0x0, s 0x0, l 0, t 0xff) 7(c 0x0, s 0x0, l 0, t 0xff) 8(c 0x0, s 0x0, l 0, t 0xff) 9(c 0x0, s 0x0, l 0, t 0xff) 10(c 0x0, s 0x0, l 0, t 0xff) 11(c 0x0, s 0x0, l 0, t 0xff) 12(c 0x0, s 0x0, l 0, t 0xff) 13(c 0x0, s 0x0, l 0, t 0xff) 14(c 0x0, s 0x0, l 0, t 0xff) 15(c 0x0, s 0x0, l 0, t 0xff)
Sep 1 20:19:11 surgeon kernel: Pending list: 1(c 0x60, s 0x7, l 0)
Sep 1 20:19:11 surgeon kernel: Kernel Free SCB list: 2 3
Sep 1 20:19:11 surgeon kernel: DevQ(0:0:0): 0 waiting
Sep 1 20:19:11 surgeon kernel: DevQ(0:5:0): 0 waiting
Sep 1 20:19:11 surgeon kernel: DevQ(0:6:0): 0 waiting
Sep 1 20:19:11 surgeon kernel: scsi0:0:0:0: Device is active, asserting ATN
Sep 1 20:19:11 surgeon kernel: Recovery code sleeping
Sep 1 20:19:11 surgeon kernel: Recovery code awake
Sep 1 20:19:11 surgeon kernel: aic7xxx_abort returns 0x2002
Sep 1 20:19:13 surgeon modprobe: modprobe: Can't locate module scsi_host
Sep 1 20:19:21 surgeon kernel: scsi0:0:0:0: Attempting to queue an ABORT message
Sep 1 20:19:21 surgeon kernel: scsi0: Dumping Card State in Data-in phase, at SEQADDR 0x79
Sep 1 20:19:21 surgeon kernel: ACCUM = 0x0, SINDEX = 0x8, DINDEX = 0x8f, ARG_2 = 0xff
Sep 1 20:19:21 surgeon kernel: HCNT = 0xf0 SCBPTR = 0x1
Sep 1 20:19:21 surgeon kernel: SCSISEQ = 0x12, SBLKCTL = 0x0
Sep 1 20:19:21 surgeon kernel: DFCNTRL = 0x78, DFSTATUS = 0x40
Sep 1 20:19:21 surgeon kernel: LASTPHASE = 0x40, SCSISIGI = 0x54, SXFRCTL0 = 0x80
Sep 1 20:19:21 surgeon kernel: SSTAT0 = 0x0, SSTAT1 = 0x2
Sep 1 20:19:21 surgeon kernel: STACK == 0x96, 0x186, 0x156, 0x0
Sep 1 20:19:21 surgeon kernel: SCB count = 4
Sep 1 20:19:21 surgeon kernel: Kernel NEXTQSCB = 2
Sep 1 20:19:21 surgeon kernel: Card NEXTQSCB = 0
Sep 1 20:19:21 surgeon kernel: QINFIFO entries: 0
Sep 1 20:19:21 surgeon kernel: Waiting Queue entries:
Sep 1 20:19:21 surgeon kernel: Disconnected Queue entries:
Sep 1 20:19:21 surgeon kernel: QOUTFIFO entries:
-----------
What is going on here? Is there something wrong with the disk drive?
the controller? the Linux drivers? the filesystem (or perhaps an
incompatibility between latest fsck and an old kernel 2.0 verysion of
ext2 filesystem)?
Note: that the partitions on the old SCSI drive seem to be OK, since I can boot
my old Linux 2.0.28 system off of them and they all fsck ok.
Also, if this is relevant, on boot-up I always get the "mysterious"
but supposedly harmless message:
kmod: failed to exec /sbin/modprobe -s -k scsi_hostadapter, errno = 2
Any suggestions here?
Thanks,
Jeff
E1
It appears that direct support for the Adaptec 2940 controller died off
with version 7.2.
Just after the middle of your syslog output, you'll find this line:
"Sep 1 20:19:13 surgeon modprobe: modprobe: Can't locate module"
Apparantly, there's no kernel module available for your hardware, which
would also explain your "kmod: failed to exec /sbin/modprobe -s -k
scsi_hostadapter, errno = 2" message at boot.
It appears that the host adapter is active (according to the syslog),
but the lack of a supporting system module is causing you problems with
running fsck.
Cause? No kernel module (driver) for your host adapter can be found in
your directory structure (at least not in the usual locations).
You might want to consult the man pages for modprobe to get a decent
idea of your options for your particular situation.
At the very worst, you might have to yank the Adaptec entirely and opt
for another storage solution. (If you upgraded from an earlier version
of RH, you should still have the applicable modules available to the
kernel, but I'm thinking that you did a clean install of 7.3, or did an
upgrade to a 7.2 installation.
Jeffrey J. Kosowsky posed the following question: