Hello All
I am experiencing extremely dangerous problems with AM53C974 driver
in 2.2.9 and 2.2.10 kernels (Mandrake 6.0).
Kernel panics when trying to transfer large files or intensively
working with SCSI drive from within X. I tried to recompile the kernel
without using the driver as a module, but this kernel (I didn't forget
to update initrd :) but still don't know whether it is needed; however,
when removing it from lilo.conf got same weirdness) crashes right after
starting to exec init scripts...
The symptoms (when booting recompiled kernel, other cases look similar):
Unmounting old root ... okay
Freeing unused kernel memory: 48k freed
scsi: aborting command due to timeout : pid 44, scsi 0, channel 0, id 0,
lun 0 Read (6) 03 c2 6b 68 00
SCSI host0 abort (pid 44) timed out - resetting
SCSI bus is being reset for host 0 channel 0.
AM53C974_reset called
AM53C974 register dump:
IO base: 0xff00 CTCREG: 0x0000; CMDREG: 0x90; STATREG: 0x11; ISREG: 0xd4
CFIREG: 0x80; CNTLREG1-4: 0x57; 0x40; 0x18; 0x44
DMACMD: 0xc3; DMASTC: 0x0400; DMASPA: 0xe2b800
DMAWBC: 0x0000; DMAWAC: 0xe2bc00; DMASTATUS: 0x08
==============
Sometimes it manages to issue smth like "kernel_panic : scsi_free" and
then dies. However, Magic SysRq still works. Sometimes I was able to
reboot the
system with C+A+D.
The HD indicator light is on during, sometimes turns off after some long
time.
The setup: HP Vectra XU 5/90C, Dual P90, NX chipset, video doesn't
matter
(tried Matrox Impression+ and Expert S3 Virge), 48M memory, no IDE
drives (CMD640 :) ).
Some old version of Debian works rock solid on the same setup.
Tried Uniprocessor kernel as well as SMP - the same.
Does anybody have any pointers to patches/workarounds?
Thanks a lot for your time and patience,
Kirill
Kirill Bolshakov
DCN Dept.
[EMAIL PROTECTED]
