>> I kept having troubles with the aic7xxx driver in stock 2.0.30.
Hi,
sorry for the long mail; I don't know what is really important.
I didn't have any (noticeable) problems with stock 2.0.30...
I ran 2.0.30+pre31-2+werner-22Jul+aic7xxx-Jul22.tgz on our main server
and had great troubles with scsi timeouts/aborts/resets.
Yesterday I put a third 4GB disk to this machine and tried
2.0.30+pre31-2+bill-readahead+werner-01Aug+aic7xxx-2.1.42-July27.tgz.
(I don't know if I can use aic7xxx-2.1.42 with the 2.0.30+ kernel...)
I got many of the following messages (with only two 4GB Quantum XP34300W):
Aug 4 00:20:09 : aic7xxx: Command complete near Qfull count, qoutcnt = 16.
Aug 4 00:20:23 : scsi : aborting command due to timeout : pid 5864444, scsi0, channel 0, id 1, lun 0 0x2a 00 00 6e df 62 00 00 06 00
Aug 4 00:20:23 : scsi : aborting command due to timeout : pid 5864445, scsi0, channel 0, id 1, lun 0 0x2a 00 00 6e df 88 00 00 02 00
Aug 4 00:20:25 : aic7xxx: Command complete near Qfull count, qoutcnt = 16.
Aug 4 00:20:38 : aic7xxx: Command complete near Qfull count, qoutcnt = 15.
Aug 4 00:20:40 : scsi : aborting command due to timeout : pid 5864444, scsi0, channel 0, id 1, lun 0 0x2a 00 00 6e df 62 00 00 06 00
Aug 4 00:20:40 : SCSI host 0 abort (pid 5864444) timed out - resetting
Aug 4 00:20:40 : SCSI bus is being reset for host 0 channel 0.
Aug 4 00:20:40 : (scsi0:1:0) Abort_reset, scb flags 0x1, while idle, LASTPHASE = 0x1, SCSISIGI 0x0, SEQADDR 0x5, SSTAT0 0x5, SSTAT1 0xa
Aug 4 00:20:40 : (scsi0:1:0) Queueing an Abort SCB.
Aug 4 00:20:40 : scsi : aborting command due to timeout : pid 5864445, scsi0, channel 0, id 1, lun 0 <4>(scsi0:1:0): Abort message sent.
Aug 4 00:20:40 : 0x2a 00 00 6e df 88 00 00 02 00
Aug 4 00:20:40 : SCSI host 0 abort (pid 5864445) timed out - resetting
Aug 4 00:20:40 : SCSI bus is being reset for host 0 channel 0.
Aug 4 00:20:40 : (scsi0:1:0) SCB 3 abort Tag completed.
Aug 4 00:20:40 : (scsi0:1:0) Reset device, active_scb 6
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 0/chan A)
Aug 4 00:20:40 : message repeated 11 times
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 1/chan A)
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 0/chan A)
Aug 4 00:20:40 : message repeated 6 times
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 1/chan A)
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 0/chan A)
Aug 4 00:20:40 : message repeated 6 times
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 1/chan A)
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 0/chan A)
Aug 4 00:20:40 : scsi0: (targ 1/chan A) matching scb to (targ 1/chan A)
Aug 4 00:20:40 : (scsi0:1:0) Aborting scb 3
Aug 4 00:20:57 : scsi : aborting command due to timeout : pid 5869746, scsi0, channel 0, id 1, lun 0 0x08 13 02 76 04 00
Aug 4 00:20:57 : SCSI host 0 channel 0 reset (pid 5864445) timed out - trying harder
Aug 4 00:20:57 : SCSI bus is being reset for host 0 channel 0.
Aug 4 00:20:57 : (scsi0:-1:0) Reset device, active_scb 15
Aug 4 00:20:57 : scsi0: (targ -1/chan A) matching scb to (targ 0/chan A)
Aug 4 00:20:57 : message repeated 30 times
Aug 4 00:20:57 : scsi0: (targ -1/chan A) matching scb to (targ 1/chan A)
Aug 4 00:20:57 : scsi0: (targ -1/chan A) matching scb to (targ 0/chan A)
Aug 4 00:20:57 : scsi0: (targ -1/chan A) matching scb to (targ 1/chan A)
Aug 4 00:20:57 : scsi0: Resetting current channel A
Aug 4 00:20:57 : scsi0: Channel reset, sequencer restarted
Aug 4 00:20:57 : (scsi0:0:0) Aborting scb 2
[other scb values within this range]
Aug 4 00:20:57 : (scsi0:1:0) Aborting scb 32
I tried the new one aic7xxx-2.1.42-July27 with
CONFIG_SCSI_AIC7XXX=y
CONFIG_AIC7XXX_TAGGED_QUEUEING=y
CONFIG_AIC7XXX_CMDS_PER_LUN=16
CONFIG_AIC7XXX_PAGE_ENABLE=y
# CONFIG_AIC7XXX_USE_EXT_SCBRAM is not set
# CONFIG_AIC7XXX_PROC_STATS is not set
CONFIG_AIC7XXX_RESET_DELAY=15
and ran one iozone 200MB on each disk and a make clean in the kernel source
tree; then the scsi bus locked up (led on the adaptec was on; no flashing).
I tried CTRL+ALT+DEL, but that doesn't help.
I'm now back on 2.0.30+pre31-2+bill-readahead+werner-01Aug. I did the same
test plus a make -j zImage modules after the successful make clean and all
went well (after some time of course ;-)
Aug 4 17:18:12 : aic7xxx: Enabling wide channel of AHA-2940 Ultra-Wide.
Aug 4 17:18:12 : AHA-2940 Ultra-WIDE (PCI-bus), I/O 0xd800, Mem 0xfa800000:
Aug 4 17:18:12 : irq 11
Aug 4 17:18:12 : bus release time 40 bclks
Aug 4 17:18:12 : data fifo threshold 100
Aug 4 17:18:12 : SCSI CHANNEL A:
Aug 4 17:18:12 : scsi id 7
Aug 4 17:18:12 : scsi selection timeout 256 ms
Aug 4 17:18:12 : scsi bus reset at power-on disabled
Aug 4 17:18:12 : scsi bus parity enabled
Aug 4 17:18:12 : scsi bus termination (low byte) disabled
Aug 4 17:18:12 : scsi bus termination (high byte) disabled
Aug 4 17:18:12 : aic7xxx: Resetting the SCSI bus...done.
Aug 4 17:18:12 : scsi0 : Adaptec AHA274x/284x/294x (EISA/VLB/PCI-Fast SCSI) 4.0/3.2/4.0
Aug 4 17:18:12 : scsi : 1 host.
Aug 4 17:18:12 : Started kswapd v 1.4.2.2
Aug 4 17:18:12 : scsi0: Target 0, channel A, now synchronous at 10.0MHz, offset 8.
Aug 4 17:18:12 : Vendor: Quantum Model: XP34300W Rev: 81HB
Aug 4 17:18:12 : Type: Direct-Access ANSI SCSI revision: 02
Aug 4 17:18:12 : Detected scsi disk sda at scsi0, channel 0, id 0, lun 0
Aug 4 17:18:12 : scsi0: Target 1, channel A, now synchronous at 10.0MHz, offset 8.
Aug 4 17:18:12 : Vendor: Quantum Model: XP34300W Rev: 81HB
Aug 4 17:18:12 : Type: Direct-Access ANSI SCSI revision: 02
Aug 4 17:18:12 : Detected scsi disk sdb at scsi0, channel 0, id 1, lun 0
Aug 4 17:18:12 : scsi0: Target 4, channel A, now synchronous at 10.0MHz, offset 8.
Aug 4 17:18:12 : Vendor: QUANTUM Model: XP34550W Rev: LXY4
Aug 4 17:18:12 : Type: Direct-Access ANSI SCSI revision: 02
Aug 4 17:18:12 : Detected scsi disk sdc at scsi0, channel 0, id 4, lun 0
Aug 4 17:18:12 : scsi : detected 3 SCSI disks total.
Aug 4 17:18:12 : SCSI device sda: hdwr sector= 512 bytes. Sectors= 8399520 [4101 MB] [4.1 GB]
Aug 4 17:18:12 : SCSI device sdb: hdwr sector= 512 bytes. Sectors= 8399520 [4101 MB] [4.1 GB]
Aug 4 17:18:12 : SCSI device sdc: hdwr sector= 512 bytes. Sectors= 8890760 [4341 MB] [4.3 GB]
Aug 4 17:18:12 : Partition check:
Aug 4 17:18:12 : sda: sda1 sda2 sda3 sda4
Aug 4 17:18:12 : sdb: sdb1 sdb2
Aug 4 17:18:12 : sdc: sdc1 sdc2
Hardware: Dual PPro (200Mhz, 256kB), 256MB, Miro 12SD, Adaptec 2940UW,
2x Quantum XP34300W, 1x Quantum XP34550W (1 XP34300W intern with
termination jumper; 1 XP34300W extern + 1 XP34550W extern with
termination jumper; Adaptec: low and high term disabled)
Bernhard
get my pgp public key by 'finger bheide@iicm.tu-graz.ac.at'
-----------------------------------------------------------------------------
Bernhard Heidegger bheide@hyperwave.com
Hyperwave Software Research & Development
Schloegelgasse 9/1, A-8010 Graz
Voice: ++43/316/820918-25 Fax: ++43/316/820918-99
-----------------------------------------------------------------------------