spurious APIC interrupt + more

Vadim E. Kogan (vadim@vadim.gem.net)
Fri, 22 May 1998 22:59:10 -0700


I get this on .103 and pre-104:

spurious APIC interrupt, ayiee, should never happen.

Looks like it might be scsi related or disk related. I get alot when
doing something like make -j bzImage on scsi disk. It did NOT happen on
.99 (and I didn't try .100-.102). Can it just be duplicated interrupt?
Everything works Ok.

BTW, did any1 try to use MTRR write-combining for video + play some
sound ot the same time? I get dead computer immidiatly after mouse
leaves current window (i.e. WM changes something). That happens on ANY
kernel.

Another problem is w/ SCSI disk on NCR. After some time it just STOPs. I
can hear disk stopping. Controller doesn't even know about it, so I get
some

May 20 00:29:49 vadim kernel: scsidisk I/O error: dev 08:01, sector
1057826
May 20 00:29:49 vadim kernel: SCSI disk error : host 0 channel 0 id 6
lun 0 return code = 28000002
May 20 00:29:49 vadim kernel: extra data not valid Current error
sd08:01: sense key Hardware Error

and then just need to not to touch the disk (umount for example). BTW,
that happens not when disk is idle - that usually happens when Im doing
something. On reboot as soon as controller resets bus disks spins up
again and works Ok. This is:

ncr53c8xx: at PCI bus 0, device 18, function 0
ncr53c8xx: 53c825 detected
ncr53c825-0: rev=0x02, base=0xfebfff00, io_port=0xec00, irq=17
ncr53c825-0: ID 7, Fast-10, Parity Checking
ncr53c825-0: restart (scsi reset).
scsi0 : ncr53c8xx - revision 2.5f
scsi : 1 host.
ncr53c825-0-<6,0>: using tagged command queueing, up to 12 cmds/lun
Vendor: QUANTUM Model: FIREBALL_TM1280S Rev: 300X
Type: Direct-Access ANSI SCSI revision: 02
Detected scsi disk sda at scsi0, channel 0, id 6, lun 0
scsi : detected 1 SCSI disk total.
ncr53c825-0-<6,0>: FAST-10 SCSI 10.0 MB/s (100 ns, offset 8)
SCSI device sda: hdwr sector= 512 bytes. Sectors= 2503872 [1222 MB] [1.2
GB]

I guess that might be something my_comp_specific. I can either try
AHA2940 or reset the host adapter/bus/device. I prefer to play the
second game. The question is - what's the Right Way to reset it? Sould
it come down to scsi_try_host_reset? If so, what's the right way to get
to this function (just call it & hope it'll work?)

Config:
dual-ppro (SMP kernel)
IDE disk + SCSI disk (on NCR 53c825)
some S3 Trio64
Matrox Millenium PowerDOC
both under AcceleratedX hand :)
SB16 + ALSA (w/ spinlocks used even for DMA/etc - so no cli/sti left)

Need more info?

Vadim

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu