Ok, after the problems we had with the 2.6.8 kernels, we are in good way
to fix this, thanks for that.

Now, i have a strange problem. When trying to burn DVD-RW media that
already had a track on it, i get :

$ growisofs -Z /dev/hdc=sarge-powerpc-1.iso
WARNING: /dev/hdc already carries isofs!
About to execute 'builtin_dd if=sarge-powerpc-1.iso of=/dev/hdc obs=32k
seek=0'
/dev/hdc: "Current Write Speed" is 2.0x1385KBps.
:-[ WRITE@LBA=0h failed with SK=5h/ASC=21h/ACQ=02h]: Invalid argument
:-( attempt to re-run with -dvd-compat -dvd-compat to engage DAO or apply full blanking procedure
:-( write failed: Invalid argument

(i didn't try the two -dvd-compat, as that resulted in media that is
unreadable on my target disk for dvd-r media).

Now, the funny thing here is that neither doing this as root or user
worked, but cdrecord with the dvd patch applied was successfull in
blanking the media, but not in burning afterward. Once
cdrecord+dvd-patch blanked the disk, i could without problem burn the
media with the above growisofs invocation.

Once i finish the burning, i can try running growisofs within strace -e
trace=ioctl or something, to see if it will sched some light on the
above problem.

I have a BTC dual format burner (8x dvd+r, 4x dvd-r, 4x dvd+rw, 2x dvd-rw)
whose /proc/ide/hdc/model says : DVD-RW IDE1008

Friendly,

Sven Luther

Arg, i spoke too quickly, the actual burning with growisofs failed with :

  4579885056/4592988160 (99.7%) @2.0x, remaining 0:04
  4589158400/4592988160 (99.9%) @2.0x, remaining 0:01
  :-[ WRITE@LBA=223860h failed with SK=5h/ASC=63h/ACQ=00h]: Input/output error
  builtin_dd: 2242672*2KB out @ average 2.0x1385KBps
  :-( write failed: Input/output error
  /dev/hdc: flushing cache
  /dev/hdc: updating RMA
  :-[ CLOSE TRACK failed with SK=2h/ASC=30h/ACQ=05h]: Wrong medium type
  /dev/hdc: closing session
  :-[ CLOSE SESSION failed with SK=2h/ASC=30h/ACQ=05h]: Wrong medium type

Wonder if the resulting image is usable though. This sounds suspiciously like
Bug #250436 though, but i will let it upto you to merge it if you consider it
to be the same problem.

I am all yours for any tests or other manipulation you want me to try.

Friendly,

Sven Luther

And altough i was able to read the media back with the burner (but maybe it
was still in cache or something such ?), on the target DVD reader (SAMSUNG
CDRW/DVD SM-352F) i could not see the media content.

Friendly,

Sven Luther

I have burned a dvd on my Pentium III to see if I was experimenting the same
thing, but it seems that it has burned ok, this is how the test was carried
out:

Machine is Pentium III based
Everything on it is IDE
DVD recorder is a _NEC DVD+/-RW ND-2510A ATAPI CD/DVD-ROM drive
Linux kernel 2.6.8 compiled by me for this machine from the sources
  published in kernel.org (the ones with the nfs bug) using native ATA,
  NO SCSI emulation.
Machine is running current sid, thus dvd+rw-tools 5.21.4.10.8-1
Media used for burning: already recorded DVD-RW
Command used to burn: growisofs -dvd-compat -Z /dev/dvd=sarge-i386-1.iso
Result: everything went smooth, dvd burned ok.

I'll try to test with a debian kernel and without -dvd-compat, but I don't
expect any changes.

Regards...

when running

/usr/bin/growisofs -Z /dev/hdb -dry-run
:-( unable to GET CONFIGURATION: Operation not permitted
:-( non-MMC unit?

cat /proc/ide/hdb/model

HL-DT-ST DVDRAM GSA-4040B

With  kernel-image-2.6.7-1-626 everything works. From my perspective
kernel kernel-image-2.6.8-1-626 from unstable breaks dvd+rw-tools in
testing so this bug should be filed against
kernel-image-2.6.8-1-626.

Hendrik