[Debburn-devel] LG CH08LS10 not burning CDs

[Thomas Schmitt]

Hi,

> Found this and killed it
> 
>  hald-addon-storage: polling /dev/sr0 (every 2 sec)

That is a very confession-happy criminal.
Pity that it seems not to be the cause of
the problem.


> I have an age-old CD-RW I could play with (if it still
> works) and a single DVD-RW which is just as old.

DVD-RW are not a good benchmark in case of
trouble. They often bring their own troubles.

If you want to stay with wodim, then try to
get a DVD+RW.
(I see below that k3B already tested growisofs
 for you. I think the current hardware
 configuration is in some way unhealthy.)


> At the moment, I am burning an image onto the
> DVD-RW at the blazing speed of 400-500 kB/s aka
> 0.30x speed.

Well, if the drive takes the data, then it is not
molested by other processes.
This statement depends on the DVD-RW formatting
state.
Check it by this info command:

  wodim -v dev=/dev/sr0 -v -atip 2>&1 | grep Current:

The possible outcome is

  Current: 0x0014 (DVD-RW sequential recording)

which should be vulnerable to disturbance, or

  Current: 0x0013 (DVD-RW restricted overwrite)

which should not be vulnerable.

If you have the latter, then we still cannot
outrule disturbance by foreign processes.
In that case the wodim shell commands below
might show the same bad effect as with CD.


> > If not, try some different media brands.
> I tried some cheap Maxell ones (cheap enough to waste
> a dozen or two with various methods, now) and Verbatim
> Azo.

That should suffice as test.
As said it is quite unusual that CD-R are bad.


> It is probably important to emphasize that the CD blanks
> are _not_ usable after a test run, any more. I.e.,
> _something_ happens.

This would match so well the known hald problems.
Damn.


> It wants to burn with 40x, but I learned the hard way to
> always choose the slowest speed offered by any drive.

One can hardly talk a drive into burning faster
than it is willing to announce.


> Also of note is that while the drive was on the 'Raid'
> SATA port, I could not boot from CD. When I switched
> over to the 'normal' SATA ports (with a new cable) I
> could boot from CD.

That should be a matter of your BIOS.
It does not indicate any inability of the drive.


The low speed with DVD-RW could indicate bus
problems. With IDE i would say that for some
reason DMA is disabled. But for SATA this
slowliness is quite unusual.

The burns in 
  grml_iso_defaults_24_speed_no_hald-addon-storage
  grml_iso_defaults_auto_speed_no_hald-addon-storage
failed just like with hald-addon-storage.
(Is it sure that it is really not running any
 more ? Not restarted automatically ?)


The file 
  dvd-asus-ion_dvd-rw_defaults
reveils that k3B already made a growisofs test
for you:
> /usr/bin/growisofs -Z /dev/sr0=/dev/fd/0 ...
...
> 155451392/3246325760 ( 4.8%) @0.3x, remaining 109:21 RBU 100.0% UBU   8.6%
> 156925952/3246325760 ( 4.8%) @0.3x, remaining 109:15 RBU 100.0% UBU 100.0%
> 158531584/3246325760 ( 4.9%) @0.3x, remaining 109:04 RBU 100.0% UBU  14.8%
> 160464896/3246325760 ( 4.9%) @0.4x, remaining 108:58 RBU 100.0% UBU   0.0%
> 162070528/3246325760 ( 5.0%) @0.3x, remaining 108:47 RBU 100.0% UBU  17.3%

The drive buffer (UBU) is running empty, whereas
the program fifo (RBU) is full.
So the transport of data between computer and
drive is hampered. Reason might be the drive
controller or the bus controller of the computer.

Is there possibly a USB cable and in the
computer an USB 1.0 bus involved ?
(Would explain slow speed but not wodim failure.)

How fast can you read from the drive ?
E.g. 100 MB from a CD or DVD
  time dd if=/dev/sr0 bs=1M count=100


For a DVD-RW test with wodim you would have to
use it as shell command. Like:

  wodim -v dev=/dev/sr0 blank=fast
  wodim -v dev=/dev/sr0 -dao /my/test/file/of/a/few/MB

Since we are not interested in reading the result
you may use any larger file as input.
If you want to eventually mount and read it,
then take some ISO 9660 image as input. E.g.
copied from some readable CD by:
  dd if=/dev/sr0 of=/tmp/cd_image.iso
(This will end in an i/o error when the CD end
 is reached. Harmless.)


Aha. More experiment results:

- dvd-asus-ion_dvd-r_defaults
>  :-[ WRITE at LBA=100h failed with SK=5h/INVALID ADDRESS FOR WRITE]: Invalid argument
  Looks again much like a hald interference.
  Address 100 hex is 256 decimal = 500 kB.

- grml_iso_-cd-rw_defaults_auto_speed
  The already known SAO failure.


------------------------------------------------

Ok. For now we have:
- wodim fails to write CD-R
- growisofs is terribly slow

Both is not very suspicious to be caused by the
burn porgrams.
The possibility of disturbance by another process
is not yet outruled. We have no clear indication
for it, though.


If you cannot spot any obstacle in the operating
system any more then you will have to test the
drive with a different computer.
A failed attempt with MS-Windows (spit spit) and
a widely used burn program like Nero should
convince the dealer.


BTW: i have a LG BDDVDRW GGC-H20 combo at USB 2.0
and it works fine with CD and DVD. My USB does
not reach full 16xDVD throughput, though. 
Nevertheless 10x to 11x is no problem for
growisofs or untuned cdrskin. (I can tune it up
to nearly 15x by sending larger data chunks.)

So it is not a matter of LG or having a DVD/BD
combo drive. (I wonder how long the combo era
will last this time. The CD/DVD combos were
around for about 2 years.)


Have a nice day :)

Thomas