> I have bizarre problems with reading burned CDRs under 2.2.0-prex
> affecting as far at least 2.2.0-pre[5-7]. The last part of the data
> on the disk is not read due to io errors. Probably I've done something
> really dumb, but anyway.....
> btw I'm not subscribed so please cc me with any replies.
Oh, how nice to see that somebody else is also seeing the problem :)
I actually exchanged some mails with Joerg Schilling regarding this
problem; this was not entirely constructive as he insisted on this
being a bug in the Linux kernel -- he called it the "read-ahead bug"
and told me to upgrade to the latest dev kernel (I was already running
it) or switching to Solaris (oh yes).
The other fact that could be related, so he told me, is that cdrecord
burns CD's in track-at-once mode, not in disk-at-once mode. Not sure
if this is the root of the problem.
I gave up trying to find out the cause and learned to live with this
"misfeature": After creating an ISO image, I simply append an MB of
zero's and burn the extended file: dd if=/dev/zero >>cd.iso bs=1k
count=1024.
> PPro
> Advansys Ultra/narrow SCSI
> TEAC CDR55S CD-recorder
Mine is a K5, ncr53c8xx, but I have the same recording device.
> cp /mnt/cdrom2/* ./junk terminates with io errors before the end of
> the last file on disk. The log file shows messages like:
Even more disturbing is that you can't read out the disk using "cp
/dev/cdrom cd.iso"...
> The disks read fine (or at least without reported errors) either
> from /dev/hdc (Mitsumi FX120T ATAPI drive) under the same kernel
> versions, and read fine using the stock Red Hat 2.0.36 kernels
> using either /dev/sr0 or /dev/hdc. Given that it reads fine
> on other systems it seems like either the advansys or sr
> driver is the cause.
Interesting. My CD's show this error even when read with an IDE drive,
under all kernel versions.
My guess is it's either the CDR55S or the TEAC driver in cdrecord.
-- michael krause [aka raw style / lego] - www.tu-harburg.de/~semk2104/
- 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/