Re: growisofs 7.1: problem past first layer on BD-R DL

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

Hi,

> Well, it'll enable defect management, whatever that really does.

Normally nothing good, except if you really have a small local
damage. To my experience, Defect Management lets media fail
earlier than without. And it makes them slow.

To get full speed with formatted BD-R and with BD-RE use
cdrskin option
  stream_recording=on
which will ask the drive to avoid Defect Management.


> The drive can do both with and without defect management;

My BDR-205 and the Optiarc can do too. But they fail nevertheless
on reasons of indivdual damage or media compatibility problems.
They should work, they think they can, but then ...
... it makes some difference whether the medium is formatted or
not formatted.
But in those situations, drive and media are balancing on the
edge of failure. Sometimes they fall, sometimes they stay up.


> my idea of removing the option and using defect management is based
> on the suspicion that the failing block is indeed problematic on
> the media, and can thus be worked around.

Rather not. The MMC reply code
  5 21 02 INVALID ADDRESS FOR WRITE
clearly says that there is something wrong with the block address of
the SCSI WRITE command.
But growisofs, as any burn program, takes care to increment the
block address in sync with the written amount.
So it would be some firmware error on the first hand.

But well, a mad drive can be deceiving to any degree.

I assume growisofs issued a RESERVE TRACK command before it
began writing. Although i trust it to send the correct parameters,
it might be worth a try to avoid it.
With cdrskin, this can be done by option
  -tao

(-tao is inherited from CD write type TAO. On BD-R and DVD+R it
 just has the effect to omit RESERVE TRACK and rather to write
 with open end.)


> I've never yet seen defect management remap
> anything, that's why I switch it off normally.

I got some BD-RE with replaced blocks. But normally
  cdrskin dev=/dev/sr3 --list_formats
says something like
  BD Spare Area: 0 blocks consumed, 393216 blocks available

----

I am reading over the BD-R description in MMC-5. 
It seems that dual-layer peculiarities are supposed to be
completely hidden from the "host" (i.e. the burn program
and its computer).
This differs much from the description of DVD+R DL, where
the opportunity to define the layer break address is explicitely
mentioned.

The command BFh SEND DISC STRUCTURE, which is used for DVD+R DL
layer break definition, is not associated to any of the features
promised by profile 41h BD-R SRM. (Well one could try to apply
growisofs_mmc.cpp function  plus_r_dl_split() to BD-R. It should
do not much harm if it fails. One would have to skip the DVD+R
specific inquiry, though.)


Have a nice day :)

Thomas