> Hello,
>
> I got under kernel 2.2.9 (redhat 5.2) the following errors I have
> never seen before:
>
> (...)
> May 16 11:01:05 jl2 kernel: attempt to access beyond end of device
> May 16 11:01:05 jl2 kernel: 16:03: rw=0, want=2020083719, limit=1068322
> May 16 11:01:05 jl2 kernel: attempt to access beyond end of device
> May 16 11:01:05 jl2 kernel: 16:03: rw=0, want=1761607682, limit=1068322
> May 16 11:01:05 jl2 kernel: attempt to access beyond end of device
> May 16 11:01:05 jl2 kernel: 16:03: rw=0, want=2147482609, limit=1068322
> (...)
> May 16 11:01:05 jl2 kernel: attempt to access beyond end of device
> May 16 11:01:05 jl2 kernel: 16:03: rw=0, want=134539165, limit=1068322
> May 16 11:01:05 jl2 kernel: hdc: dma_intr: status=0x80 { Busy }
> May 16 11:01:05 jl2 kernel: hdc: DMA disabled
> May 16 11:01:05 jl2 kernel: hdd: DMA disabled
> May 16 11:01:05 jl2 kernel: attempt to access beyond end of device
> May 16 11:01:05 jl2 kernel: 16:03: rw=0, want=1746433536, limit=1068322
> (...)
> May 16 11:01:05 jl2 kernel: 16:03: rw=0, want=267814149, limit=1068322
> May 16 11:01:12 jl2 kernel: ide1: reset: success
>
> Is this something I should worry about?
>
> I found some people reporting similar messages, thinking they
> have something to do with the sym/ncr53c875-scsi-driver resp.
> with broken scsi disks. It is true that I have a scsi controller,
> but only for my tape drive; all of my disks are ide ones and
> according to e2fsck and scandisk I don't recognize any damage
> to my disks. Maybe some of these people are interested in the
> fact, that in my case it doesn't seem to be a disk problem.
>
> But - can I do anything to avoid those error messages?
This problem may be due to bad DMA (hardware) or to software bug.
The following software change has been reported to fix the problem
in some kernel, a couple of weeks ago.
--- linux/drivers/scsi/sd.c.ORIG Tue May 18 21:45:05 1999
+++ linux/drivers/scsi/sd.c Tue May 18 21:45:56 1999
@@ -1050,6 +1050,7 @@
SCpnt->transfersize = rscsi_disks[dev].sector_size;
SCpnt->underflow = this_count << 9;
+ SCpnt->cmd_len = 0;
scsi_do_cmd (SCpnt, (void *) cmd, buff,
this_count * rscsi_disks[dev].sector_size,
rw_intr,
I have posted it to a linux list with some explanations and it seems that
this has had the same effect as if I just had pissed into a violin :), by
the way.
Gérard.
-
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/