I patched kernel 2.1.107 to ncr-driver 0.5 and compiled stock 2.1.121.
I got M_REJECTs for both. I disconnected the uw disc. Still M_REJECT.
I disconnected DDS and MO-disk. No M_REJECT ...
Now i am running 2.1.107+0.5 with all devices, passive termination at
scsi 2 end, *TAGS=32,*SYNC=40, recompile kernel 121 (because of
ppp-errors), waiting for M_REJECT for
> uptime
> 4:31pm up 30 min, 3 users, load average: 1.07, 0.88, 0.58
> 5:03pm up 1:02, 3 users, load average: 0.00, 0.06, 0.35
U understand?
The only software dependency that might be true:
3.0g(107/117/121): ncr53c875-0-<0,0>: M_REJECT received (75:8)
ncr53c875-0-<0,0>: M_REJECT received (35:8)
0.[(4)5](107/117): ncr53c875-0-<0,0>: M_REJECT received (19:8)
with scsi II devices only:
0.5(107): ncr53c875-0-<0,0>: M_REJECT received (3e:8)
and of course:
2.5f(107): no M_REJECT
while
2.5f(117): ncr53c875-0-<0,0>: M_REJECT sent for 1-3-1-19-1e
ncr53c875-0-<1,0>: M_REJECT sent for 1-3-1-19-f
btw: smp=y, 2xi586/133, 128MB, gcc2.8.1;
Is there any chance to separate bus signal corruption from memory
corruption?
you wrote:
>It seems to occur during the device scan at start-up.
I dont believe this. Sometimes there was M_REJECT after i ran x11.
When i started 2.1.121+3.0g first, i got several block _like_
ncr53c875-0: restart (scsi reset).
ncr53c875-0: Downloading SCSI SCRIPTS.
ncr53c875-0-<1,*>: FAST-10 SCSI 10.0 MB/s (100 ns, offset 16)
ncr53c875-0-<0,0>: M_REJECT received (75:8).
mixed into the normal boot messages (start of init, init of isdn ...)
(The above four lines are from previous message, i didnīt save the new
ones. As they occured so regular, i wouldnīt expect to be unable to
reproduce them now. If i remember correctly, both discs appeared, disc
id 1 using 20MB.).
gerhard.
-
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/faq.html