[Debburn-devel] LG CH08LS10 not burning CDs

[Thomas Schmitt]

Hi,

> http://paste.debian.net/58276/
> http://paste.debian.net/58278/

Same old problem.

> I will in single user mode try next night; maybe we should
> prepare something to send to LKML if these problems persist?

LKML is a harsh environment.
Better start with the distro community and get
a kernel expert as companion for an appearance
on LKML. Coordinate with surfed, so you appear
as one problem report.

Important will be that you can tell them a
simple and cheap way to reproduce it.
I'll try to help with that.


> If I should try any BIOS/grub settings, please tell me.

We have the suspicion that ACPI has to do with
it.
We have indication that AHCI aggravates or even
causes the problem.
We have indication that some software triggered
by udev might be polling fast and endlessly.

But with those kernel aspects i cannot be much
of help.


> I resigned to
> wasting the whole 50 CD-Rs, already.

You can burn them in -dummy mode. That keeps the
media blank. Are they really wasted after the
failures, at all ?
Test by
  cdrskin -v dev=/dev/sr0 -atip 2>&1 | \
  fgrep 'cdrskin: status'
Hope for
  cdrskin: status 1 burn_disc_blank "The drive holds a blank disc"
rather than
  cdrskin: status 4 burn_disc_full "There is a disc with data on it in the drive"
The result might be different for CDs which were
burned in SAO mode in contrast to TAO mode.


K3B is in no way needed to reproduce the probem.
This shell command should suffice for testing and
should not waste CD-R media:

  dd if=/dev/zero bs=1M count=100 | \
  wodim -v dev=/dev/sr0 -tao -dummy -

You might get the resulting media reported as
written as long as you did not unload and reload
it. After reload it should appear as blank.


Even more convenient is to watch for dmesg errors
to come when you insert a blank CD and do _not_
start a burn program.
As long as complaints about a bad CD appear, that
long you have a program molesting the drive in
a way that often prevents writing. 

E.g. this quote from
  http://paste.debian.net/58278/
does not stem from the failed burn run:
  [  363.644980] sr 1:0:0:0: [sr0] Add. Sense: Logical block address out of range
  [  363.644984] end_request: I/O error, dev sr0, sector 0
  [  363.644988] Buffer I/O error on device sr0, logical block 0

The grade of vulnerability differs between drives
and firmware revisions. But in general it is
fatal for CD burning if you cannot keep other
entities from accessing the drive.
(Again: Only CD and DVD- media are that picky.)


> Also, if anyone thinks they could use any of those CDs I burned,
> please tell me and I can send them by mail.

No use for really spoiled ones.
Check them with cdrskin, xorriso, or cdrecord
option -minfo, whether they are declared written.
In that case you can trash them.

I would be interested in testing such a drive.
(Best in an external box with eSATA. But i could
also try to operate a screwdriver as last resort.
My location is germany.)


Have a nice day :)

Thomas