Module Autoloader or aic7xxx in 2.0.35 (fixed in pre 2.0.36)

Claudius Link (linkclau@ruf.uni-freiburg.de)
Tue, 13 Oct 1998 00:35:35 +0200 (MET DST)


Hello,

this is rather a notice that a bug in 2.0.35
seams to be fixed (in a future 2.0.36) than a bug report.

THIS BUG SEAMS TO BE FIXED in any of Alans pre 36 patches
(I tried it with 2.0.36-pre-patch-1.gz, 2.0.36-pre-patch-9.gz
and 2.0.36-pre-patch-12.gz )
AND it was not in 2.0.34

Description of the bug:
With kernel 2.0.36 and AHA-2940A Ultra SCSI host adapter,
"SCSI CD-ROM support" as an module and autoload of modules
enabled.

Any access on the CD-ROM resulted in a complete hang of the machine.
And at least sometime in this message:

Kernel panic aic7xxx: AWAITING_MSG for an SCB that does
not have a waiting message.
In swapper task - not syncing

The fix(es):
- Kernel 2.0.34 worked,
- 'insmod sr_mod' before the first access
- (SCSI CD-ROM support compiled into the kernel) or
- the patch 'Alan-Cox-2.0.36-pre9' solved the problem as well.

Even if I doubt that it helps, here is the output of
/proc/scsi/aic7xxx/0
just before successfully reproducing the panic
(or unsuccessfully mounting a cdrom :-( )

Adaptec AIC7xxx driver version: 5.0.19/3.2.4
Compile Options:
AIC7XXX_RESET_DELAY : 5
AIC7XXX_TAGGED_QUEUEING: Adapter Support Enabled
Check below to see which
devices use tagged queueing
AIC7XXX_PAGE_ENABLE : Enabled (This is no longer an option)
AIC7XXX_PROC_STATS : Enabled

Adapter Configuration:
SCSI Adapter: Adaptec AHA-2940A Ultra SCSI host adapter
Ultra Narrow Controller
Programmed I/O Base: b800
Adaptec SCSI BIOS: Enabled
IRQ: 10
SCBs: Active 0, Max Active 2,
Allocated 30, HW 3, Page 255
Interrupts: 4267
BIOS Control Word: 0x18b6
Adapter Control Word: 0x0059
Extended Translation: Enabled
Disconnect Enable Flags: 0x00ff
Ultra Enable Flags: 0x0000
Tag Queue Enable Flags: 0x0000
Ordered Queue Tag Flags: 0x0000
Default Tag Queue Depth: 8
Tagged Queue By Device array for aic7xxx host instance 0:
{255,255,255,255,255,255,255,255,255,255,255,255,255,255,255,255}
Actual queue depth per device for aic7xxx host instance 0:
{1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1}

Statistics:
(scsi0:0:0:0)
nxfers 3140 (2905 read;235 written)
blks(512) rd=27438; blks(512) wr=884
< 512 512-1K 1-2K 2-4K 4-8K 8-16K 16-32K 32-64K 64-128K >128K
Reads: 0 150 1102 125 660 817 38 13 0 0
Writes: 0 0 155 59 11 9 1 0 0 0

(scsi0:0:1:0)
nxfers 1080 (930 read;150 written)
blks(512) rd=2559; blks(512) wr=332
< 512 512-1K 1-2K 2-4K 4-8K 8-16K 16-32K 32-64K 64-128K >128K
Reads: 0 33 887 0 2 1 1 6 0 0
Writes: 0 0 136 14 0 0 0 0 0 0

I did not include any .config but if anyone thinks this or any other
information will be useful I will be happy to send it as well.

Claudius

-- 

Claudius Link linkclau@ruf.uni-freiburg.de

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/