Re: [BUG] SCSI/aha152x lockup on 2.2.6 while reading faulty cd-rom

Simon J Mudd (sjmudd@bitmailer.net)
Sun, 18 Apr 1999 19:37:20 +0200 (CET)


On Sun, 18 Apr 1999, Simon J Mudd wrote:

> I can reproducibly hang 2.2.6 if reading a faulty cd-rom. I don't expect
> to be able to read the cd-rom, but I don't think the kernel should lock up
> completely in this situation. I know this is old hardware, but any help
> would be appreciated. If more info is needed I'll provide it on request.

Just tested the same sequence of events using redhat 5.2's "default
kernel" 2.0.36-0.7, and get the following output on the screen but NO
lockup.

scsi0: MEDIUM ERROR on channel 0, id 1, lun 0, CDB: Request Sense 00 00 00 10 00
Current error sr0b:00: sense key Medium Error
Additional sense indicates Unrecovered read error
CD-ROM I/O error: dev 0b:00, sector 885672
scsi : aborting command due to timeout : pid 1436, scsi0, channel 0, id 1,
lun 0 Read (6) 03 60 eb 03 00
aha152x: abort(), SCpnt=0x00fbd018, QUEUE STATUS:
issue_SC:
current_SC:
none
disconnected_SC:
0x00fbd018: target=1; lun=0; cmnd=(Read (6) 03 60 eb 03 00
); residual=1024; buffers=5; phase |disconnected|send_ident|; in
other(MESSAGE IN); phaseend; next=0x00000000

waiting: SCSISEQ (RESELI ); SCSISIG (DATA OUT); INTSTAT (lo); SSTAT
(BUSFREE PHASECHG ); SSTAT (); SXFRCTL0 (CH1 ); SIGNAL (); SELID (84),
SSTAT2 (SEMPTY ); SFCNT (0); SCSICNT (0), OFFCNT(0), SSTAT4 (); DMACNTRL0
(16BIT PIO WRITE INTEN ); DMASTAT (WORDRDY DFIFOEMP )
enabled interrupts (ENSELDI )
(ABORT) target selected, CD-ROM I/O error: dev 0b:00, sector 885676
CD-ROM I/O error: dev 0b:00, sector 885688
CD-ROM I/O error: dev 0b:00, sector 885672
cat: /mnt/cdrom/images/boot.img: Input/output error
[root@phoenix /root]#

I don't know if this information is sufficient to find the bug.

Let me know if I can provide more information.

Simon

--
Simon J Mudd, Madrid SPAIN  Tel: +34-91-408 4878  email: sjmudd@bitmailer.net
[short messages - from radio hams only]     ---->      ea4els@ea4els.ampr.org

- 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/