Re: cdrecord with LSI53C1010 and Yamaha CRW2100S

[Gérard Roudier]

On 4 Sep 2001, C. Linus Hicks wrote:

> I have been unsuccessful trying to burn a CD-R since upgrading my
> computer and I'm not sure where to go with this. I decided to try here.
> I suspect a problem with the SYM53C8xx driver, and I think there may
> also be additional problems elsewhere. My old system worked. It was:
>
> RedHat 7.0 with 2.2.19 kernel
> ASUS P2B-DS (440BX chipset) with 2 600MHz Intel processors
> On-board Adaptec AIC-7890 with the new aic7xxx driver
> Add-on Adaptec 2940 PCI adapter (narrow devices attached here)
> Yamaha CRW6416S CD writer
>
> My new system:
>
> RedHat 7.1 with 2.4.8-ac11 kernel
> Tyan Thunder HEsl S2567 (ServerWorks chipset) with 2 1000MHz Intel
> processors
> On-board LSI 53C1010-66 (running at 33MHz) dual channel SCSI
> Add-on LSI21003 with 53C1010-33 (narrow devices here)
> SYM53C8xx driver
> Yamaha CRW2100S CD writer
>
> I looked for problem reports with the 2100S and didn't see any, rather
> several reports that it worked okay. I updated the firmware to the
> latest - 1.0N and it didn't help.
>
> I have tried several versions of cdrecord including the latest 1.11a05
> with associated tools to no avail.
>
> When I try burning a CD, sometimes I get errors and sometimes it appears
> to work. When I try to mount a CD-R that burned successfully in a
> Toshiba DVD-ROM drive, I get errors like this:
>
> [root@lh2 /root]# mount /mnt/cdrom
> mount: wrong fs type, bad option, bad superblock on /dev/cdrom,
>        or too many mounted file systems
> [root@lh2 /root]#
>
> And in /var/log/messages:
>
> Sep  4 16:31:18 LH2 kernel: sym53c1010-33-1:4: ERROR (0:18) (1-21-0)
> (10/30) @ (script 8e8:110007c1).
> Sep  4 16:31:18 LH2 kernel: sym53c1010-33-1: script cmd = 88080000
> Sep  4 16:31:18 LH2 kernel: sym53c1010-33-1: regdump: da 10 c0 30 47 10
> 04 0e 80 01 84 21 80 01 01 00 00 b0 d6 37 08 00 00 00.
> Sep  4 16:31:18 LH2 kernel: sym53c1010-33-1: ctest4/sist original
> 0x8/0x18  mod: 0x18/0x0
> Sep  4 16:31:18 LH2 kernel: sym53c1010-33-1: restart (scsi reset).
> Sep  4 16:31:18 LH2 kernel: sym53c1010-33-1: handling phase mismatch
> from SCRIPTS.
> Sep  4 16:31:18 LH2 kernel: sym53c1010-33-1: Downloading SCSI SCRIPTS.
> Sep  4 16:31:20 LH2 kernel: sym53c1010-33-1-<4,*>: FAST-20 SCSI 20.0
> MB/s (50.0 ns, offset 16)
> Sep  4 16:31:20 LH2 kernel:  I/O error: dev 0b:00, sector 68
> Sep  4 16:31:20 LH2 kernel: isofs_read_super: bread failed, dev=0b:00,
> iso_blknum=17, block=17
>
> I tried replacing the LSI21003 with a 2940 and burned a CD-R. The
> process went smoothly with no errors reported.

The ERROR (0:18) thing indicates a SCSI GROSS ERROR problem. In other
words, the controller is seing a severe transport error on the SCSI BUS.
This is often due to a SCSI BUS problem. Decreasing the data throughput
can make the problem less likely to happen, but this is obviously a poor
workaround.

Could you check the synchronous data parameters negotiated with the
Adaptec. If speed is lower than 20 MB/s and/or offset is lower than 16,
then we may well just be comparing oranges and apples.

> I mounted the burned CD
> in the Toshiba successfully and listed the contents. I did not think to
> verify the CD with the mastered file. I then switched back to the
> LSI21003 and tried to mount the burned CD-R. I got:

Hmmm... If you just hot-switch the device then the breakage seems normal.
Can I assume you didn't do so ?:)

> [root@lh2 /root]# mount /mnt/cdrom
> mount: Not a directory
> [root@lh2 /root]#
>
> And in /var/log/messages:
>
> Sep  4 15:15:47 LH2 kernel: scsi : aborting command due to timeout : pid
> 0, scsi1, channel 0, id 4, lun 0 Read (10) 00 00 00 00 1c 00 00 01 00
> Sep  4 15:15:47 LH2 kernel: sym53c8xx_abort: pid=0 serial_number=149685
> serial_number_at_timeout=149685
> Sep  4 15:15:48 LH2 kernel: SCSI host 1 abort (pid 0) timed out -
> resetting
> Sep  4 15:15:48 LH2 kernel: SCSI bus is being reset for host 1 channel
> 0.
> Sep  4 15:15:48 LH2 kernel: sym53c8xx_reset: pid=0 reset_flags=2
> serial_number=149685 serial_number_at_timeout=149685
> Sep  4 15:15:48 LH2 kernel: sym53c1010-33-1: restart (scsi reset).
> Sep  4 15:15:48 LH2 kernel: sym53c1010-33-1: handling phase mismatch
> from SCRIPTS.
> Sep  4 15:15:48 LH2 kernel: sym53c1010-33-1: Downloading SCSI SCRIPTS.
> Sep  4 15:15:48 LH2 kernel: sym53c1010-33-1-<4,*>: FAST-20 SCSI 20.0
> MB/s (50.0 ns, offset 16)
> Sep  4 15:15:48 LH2 kernel: Device 0b:00 not ready.
> Sep  4 15:15:48 LH2 kernel:  I/O error: dev 0b:00, sector 112
> Sep  4 15:15:49 LH2 kernel: ISOFS: unable to read i-node block
> Sep  4 15:15:49 LH2 kernel: Device 0b:00 not ready.
> Sep  4 15:15:49 LH2 kernel:  I/O error: dev 0b:00, sector 128
> Sep  4 15:15:49 LH2 kernel: ISOFS: unable to read i-node block
>
> Any help would be appreciated, and if anyone needs additional
> information I will be glad to do what I can.

I suggest you to define some lower data speed in the NVRAM and see if it
makes differences. OTOH, if the Adaptec is using a different (assumed
lower) speed you may give a try with the C1010-33 at the same speed.
You may let me know the results.

Regards,
  Gérard.