[19] in linux-scsi channel archive
aha152x driver hangs in 1.1.8x kernels
daemon@ATHENA.MIT.EDU (Alastair McKinstry)
Sat Jan 21 06:34:47 1995
To: linux-scsi@vger.rutgers.edu
Date: Sat, 21 Jan 1995 10:59:52 +0000
From: Alastair McKinstry <amck@maths.tcd.ie>
-----BEGIN PGP SIGNED MESSAGE-----
Is this a bug ?
I am running kernel 1.1.83 on a 486.
I have a Toshiba SCSI CD-ROM XM-5201, running on a Soundblaster SCSI =
card (aha-152x driver). Under certain conditions it can be made to hang.
Here are my hardware details, reported in /var/adm/messages:
Jan 18 17:13:51 magrathea syslogd: restart
Jan 18 17:13:52 magrathea kernel: Kernel logging (proc) started.
Jan 18 17:13:52 magrathea kernel: Console: colour EGA+ 80x25, 1 virtual co=
nsole (max 63)
Jan 18 17:13:52 magrathea kernel: Calibrating delay loop.. ok - 25.04 Bogo=
Mips
Jan 18 17:13:52 magrathea kernel: Serial driver version 4.00 with no seria=
l options enabled
Jan 18 17:13:52 magrathea kernel: tty00 at 0x03f8 (irq =3D 4) is a 16450
Jan 18 17:13:52 magrathea kernel: tty01 at 0x02f8 (irq =3D 3) is a 16550A
Jan 18 17:13:52 magrathea kernel: snd2 <SoundBlaster Pro 4.12> at 0x220 ir=
q 5 drq 1
Jan 18 17:13:52 magrathea kernel: snd6 <SoundBlaster 16 4.12> at 0x220 irq=
5 drq 1
Jan 18 17:13:52 magrathea kernel: snd7 <SoundBlaster MPU-401> at 0x330 irq=
5 drq 0
Jan 18 17:13:52 magrathea kernel: snd1 <Yamaha OPL-3 FM> at 0x388 irq 0 dr=
q 0
Jan 18 17:13:52 magrathea kernel: ftape: allocated 3 buffers aligned at: 0=
0388000
Jan 18 17:13:52 magrathea kernel: Floppy drive(s): fd0 is 1.44M
Jan 18 17:13:52 magrathea kernel: FDC 0 is a 8272A
Jan 18 17:13:52 magrathea kernel: aha152x: processing commandline: ok
Jan 18 17:13:52 magrathea kernel: detection complete
Jan 18 17:13:52 magrathea kernel: aha152x: vital data: PORTBASE=3D0x340, I=
RQ=3D11, SCSI ID=3D7, reconnect=3Ddisabled, parity=3Denabled
Jan 18 17:13:52 magrathea kernel: scsi0 : Adaptec 152x SCSI driver; $Revis=
ion: 1.6 $
Jan 18 17:13:52 magrathea kernel: scsi : 1 hosts.
Jan 18 17:13:52 magrathea kernel: Vendor: TOSHIBA Model: CD-ROM XM-520=
1TA Rev: 3014
Jan 18 17:13:52 magrathea kernel: Type: CD-ROM =
ANSI SCSI revision: 02
Jan 18 17:13:52 magrathea kernel: Detected scsi CD-ROM sr0 at scsi0, id 0,=
lun 0
Jan 18 17:13:52 magrathea kernel: scsi : detected 1 SCSI cdrom total.
Jan 18 17:13:52 magrathea kernel: Scd sectorsize =3D 2048 bytes.
Jan 18 17:13:52 magrathea kernel: Memory: 5316k/8448k available (676k kern=
el code, 384k reserved, 2072k data)
Jan 18 17:13:52 magrathea kernel: This processor honours the WP bit even w=
hen in supervisor mode. Good.
Jan 18 17:13:52 magrathea kernel: Swansea University Computer Society NET3=
.019
Jan 18 17:13:52 magrathea kernel: Swansea University Computer Society TCP/=
IP for NET3.019
Jan 18 17:13:52 magrathea kernel: IP Protocols: ICMP, UDP, TCP
Jan 18 17:13:52 magrathea kernel: Checking 386/387 coupling... Ok, fpu usi=
ng exception 16 error reporting.
Jan 18 17:13:52 magrathea kernel: Checking 'hlt' instruction... Ok.
Jan 18 17:13:52 magrathea kernel: Linux version 1.1.82 (amck@magrathea) (g=
cc version 2.5.8) #3 Wed Jan 18 03:14:51 GMT 1995
Jan 18 17:13:52 magrathea kernel: Partition check:
Jan 18 17:13:52 magrathea kernel: hda: WDC AC2420H, 405MB w/128KB Cache,=
CHS=3D989/15/56, MaxMult=3D16
Jan 18 17:13:52 magrathea kernel: hda: hda1 hda2 hda3 hda4
Jan 18 17:13:52 magrathea kernel: VFS: Mounted root (ext2 filesystem).
Jan 18 17:13:52 magrathea kernel: Adding Swap: 12596k swap-space
Jan 18 17:13:56 magrathea kernel: ftape v1.13b 5/7/94 (c) 1993, 1994 Bas L=
aarhoven (bas@vimec.nl)
Jan 18 17:13:56 magrathea kernel: QIC-117 driver for QIC-40 and QIC-80 ta=
pe drives
Jan 18 17:14:22 magrathea kernel: PCM device 1 not installed.
Jan 18 17:14:49 magrathea login: ROOT LOGIN ON tty1
I am using the cdplayer-2.0 program, and a small number of test programs
based on the IOCTL() calls seen in the source. I am not sure how to use
these properly; where can documentation be found ?
If I have ejected the tray using IOCTL(cd,CDROMEJECT), and then
play a CD using the CDROMPLAYTRKIND call, no error is reported, despite
the kernel knowing that the tray is ejected, as it gives the following mes=
sage
in /var/adm/messages: =
Jan 18 17:15:33 magrathea kernel: Device not ready. Make sure there is a =
disc in the drive.
Jan 18 17:15:33 magrathea kernel: VFS: Disk change detected on device 11/0
Jan 18 17:15:33 magrathea kernel: sr_photocd: ioctl error (TOSHIBA #2): 0x=
28000000
Jan 18 17:15:33 magrathea kernel: CDROM not ready. Make sure there is a d=
isc in the drive.
Is this a bug ? how is a program to know that the tray is currently
ejected ?
Note that the sr_photocd error message only occurs if root is running the
cdplayer; if an ordinary user runs it, just the other messages come up.
More seriously, if the tray is ejected using the manual eject button,
rather than by software, an attempt to play the CD rom causes the terminal
process to hang; as the process is caught in the ioctl() call, the process
is unkillable by kill -9 on another VC. However, the errors:
Jan 18 17:15:59 magrathea kernel: Device not ready. Make sure there is a =
disc in the drive.
Jan 18 17:15:59 magrathea kernel: VFS: Disk change detected on device 11/0
Jan 18 17:15:59 magrathea kernel: CDROM not ready. Make sure there is a d=
isc in the drive.
appear in the /var/adm/messages log, so the kernel knows the CD tray is =
not ready. Control is returned to the process and the program (mostly?)
continues as expected once the tray is re-inserted.
- --------
-^-
<amck@maths.tcd.ie> | Alastair McKinstry
Finger or mail me for PGP Public Key
PGP Key fingerprint =3D 19 72 38 40 D6 BF FD E0 21 17 96 05 B4 81 09 B1 =
When the going gets wierd, the wierd turn pro -- HST
-----BEGIN PGP SIGNATURE-----
Version: 2.6
iQCVAgUBLyDpJDzDKcCV/ZSpAQExtwP/S+sXa3XJdDdP9j1O9DtY1o7IB/vDxz+X
dMn6HhYzBE2cYMBXmW/S5fLbUXMrwoBLK4eat67VtybeBB2O7suXBQAegCrzhNNr
uG6RNvY91AQdDEiAKy6laRjwcN4fONp2CgGOyVMrceRAXErVqg3qd0664D1k4FWx
OT6w8//2wAc=3D
=3Dsom9
-----END PGP SIGNATURE-----
