Kurt Garloff schrieb am Donnerstag, den 23. November 2000:
> On Wed, Nov 22, 2000 at 02:06:18AM +0100, Matthias Andree wrote:
> > I ran that script several times since it did not collect all devices,
>
> Strange.
>
> While 2.0e3 contains a bug that can cause an OOps inside the driver (just
> use the echo "INQUIRY 0" >/proc/scsi/tmscsim/?), the normal bus rescanning
> should not be able to trigger it. The above looks like the bug is occuring
> somewhere else.
> Having said this, I'd like to ask you to try 2.0e6 of the tmscsim driver and
> check whether you are able to reproduce the bug.
Good thing first: I can't trigger the oopses with 2.0e6. I tried hard, didn't
manage to.
However, trying to trigger a race condition, I found that your script suffers
from too small a kernel API in this place, since it uses the
remove-single-device and all-single-devices in a non-locked manner, so if you
run several instances of your script in parallel (simple rescan-scsi-bus.sh -r
& will do) it will give a mess and may leave you with missing devices.
I don't currently recall if your web site states that your script may only be
run once at the same time; lock files would be nice :-]
To recall my devices:
#2 Plextor PX-32TS 1.02
#3 Yamaha CRW4416S 1.0j
#4 HP 35480A 12.09
#5 Plextor PX-20TS 1.00
#6 Tandberg TDC4222
The first run of your rescan-scsi-script showed that the "scsi
singledevice 1 0 2 0" command was aborted due to timeout, and the
Plextor PX-32TS is known to require at least 5 s settling time after a
bus reset in my Tekram DC-390U (SYM53C875) BIOS, so it may actually be the
drive itself that gives us headaches.
Result of first rescan-scsi-bus.sh -r:
all 5 devices removed, and only #3 ... #6 added (#2, Plextor PX-32
missing)
During scanning, I got these (the first two aborts and the SCSI bus reset look
right since I hotplugged the entire bus to the other adaptor)
scsi singledevice 1 0 2 0
scsi : aborting command due to timeout : pid 46, scsi1, channel 0, id 2, lun 0 Inquiry 00 00 00 ff 00
DC390: Abort command (pid 46, Device 02-00)
DC390: SRB: Xferred 00000000, Remain 00000100, State 00000100, Phase 01
DC390: AdpaterStatus: 00, SRB Status 00
DC390: Status of last IRQ (DMA/SC/Int/IRQ): 00818418
DC390: Register dump: SCSI block:
DC390: XferCnt Cmd Stat IntS IRQS FFIS Ctl1 Ctl2 Ctl3 Ctl4
DC390: 0000ff 90 01 c4 00 81 17 48 08 84
DC390: FIFO: 20
DC390: Register dump: DMA engine:
DC390: Cmd STrCnt SBusA WrkBC WrkAC Stat SBusCtrl
DC390: 83 00000100 071c5da4 00000100 071c5da4 00 03184200
DC390: Register dump: PCI Status: c200
DC390: In case of driver trouble read linux/drivers/scsi/README.tmscsim
DC390: Abort current command (pid 46, SRB c7e3c13c)
DC390: Aborted pid 46 with status 3
scsi : aborting command due to timeout : pid 46, scsi1, channel 0, id 2, lun 0 Inquiry 00 00 00 ff 00
DC390: Abort command (pid 46, Device 02-00)
DC390: SRB: Xferred 00000000, Remain 00000100, State 00000100, Phase 01
DC390: AdpaterStatus: 00, SRB Status 00
DC390: Status of last IRQ (DMA/SC/Int/IRQ): 00818418
DC390: Register dump: SCSI block:
DC390: XferCnt Cmd Stat IntS IRQS FFIS Ctl1 Ctl2 Ctl3 Ctl4
DC390: 0000ff 90 01 c4 00 80 17 48 08 84
DC390: Register dump: DMA engine:
DC390: Cmd STrCnt SBusA WrkBC WrkAC Stat SBusCtrl
DC390: 83 00000100 071c5da4 00000100 071c5da4 00 03184200
DC390: Register dump: PCI Status: c200
DC390: In case of driver trouble read linux/drivers/scsi/README.tmscsim
DC390: Abort current command (pid 46, SRB c7e3c13c)
DC390: Aborted pid 46 with status 3
SCSI host 1 abort (pid 46) timed out - resetting
SCSI bus is being reset for host 1 channel 0.
DC390: RESET ... done
scsi singledevice 1 0 3 0
DC390: Illegal Operation detected (00c38418)!
DC390: SRB: Xferred 00000000, Remain 00000000, State 00000040, Phase 05
DC390: AdpaterStatus: 00, SRB Status 00
DC390: Status of last IRQ (DMA/SC/Int/IRQ): 00c38418
DC390: Register dump: SCSI block:
DC390: XferCnt Cmd Stat IntS IRQS FFIS Ctl1 Ctl2 Ctl3 Ctl4
DC390: 0000ff 42 03 84 00 80 17 48 08 84
DC390: Register dump: DMA engine:
DC390: Cmd STrCnt SBusA WrkBC WrkAC Stat SBusCtrl
DC390: 00 00000100 071c5da4 00000100 071c5da4 00 031a4602
DC390: Register dump: PCI Status: c200
DC390: In case of driver trouble read linux/drivers/scsi/README.tmscsim
A subsequent rescan-scsi-bus.sh will find the PX-32TS and add it, this time,
without bus resets and aborts.
Should the reset->inquiry delay be applied after ANY reset? Is it actually
applied but too short for my Plextor?
This archive was generated by hypermail 2b29 : Thu Nov 30 2000 - 21:00:17 EST