Re: growisofs : Wrong capacity when burn multisession BD-R

["Thomas Schmitt" <scdbackup@gmx.net>]

Hi,

Michael Herscheid wrote:
>  In short, growisofs reports the wrong capacity when burning 
> multisession bd-r media.
>  While readin source (growisofs_mmc.cpp), if found that the variable 
> bdr_plus_pow is set if the drive is captable to write pseudo overwrite (282)
> but bdr_plus_pow is used later (line 414) to indicate the media is used in
> pow mode.
>   So the correct fix would be to ask if the media is used as pow media 
> (0x51 READ DISC INFORMATION) before to decide if to use blocks already
> written or not.

In line 282 the drive seems to have stated that feature 0x38 POW is
current:
  (header[8+2]&1)==1
See MMC-5 5.2.2.4, Current Bit.

One could nevertheless inquire the number of Remaining POW Replacements
by READ DISC INFORMATION before assuming that the overwrite gesture
of growisofs will not fail after the new session was written.
(Afaik it overwrites per session one single 32 kB chunk at address 0.
 POW should hardly be used up after only 5 sessions.)


> READ TRACK INFORMATION[#6]:
>  Track State:           invisible incremental
>  Track Start Address:   13108288*2KB
>  Next Writable Address: 13108288*2KB
>  Free Blocks:           11199424*2KB

This seems not to be what growisofs expects as reply from a POW media.
The code suggests that it should look more like DVD+RW or BD-RE:
Only one track, starting at block 0 and reaching up to the formatted
media end. (Having a non-zero Next Writeable Adress, though.)


I would have done POW experiments myself if BD-R media weren't still
that expensive. So libburn uses them as sequential media for now,
formatted or unformatted. Experience with 50 GB media is sparse.
One user reported one success. I use 25 GB BD-R for daily backups
up to 300 sessions (more sessions seem to be bad on LG GGW H20).


Have a nice day :)

Thomas