[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: cdrecord and "get performance" scsi errors


me (about dvd+rw-mediainfo):
> > This stems from a CDB
> >   AC ?? 00 00 00 00 00 00 nn nn 03 00
> > MMC-5 says nothing about the Data Type (byte 1 of the CDB)
> > with Type 03h. I would expect that ?? is still 04 from the
> > previous command.
Joerg Schilling:
> ???
> Do you expect a bug in growisofs?

At least i cannot see that in byte 1 the value 04 from
the previous command is overwritten. So i expect this
04 to be sent again with the next command. In libburn
this command is sent with byte 1 == 00.
Well, it seems not to matter.

> [avoid] a DMA error by trying to
> reduce the 64 == 100 dec. to a lower number and check again

dvd+rw-tools usually first inquire the number of descriptors
by a "Maximum Number of Descriptors" (bytes 8+9) of 1.
I read from MMC-5 rather the idea to use 0 for this purpose.
Nevertheless in libburn i allow a reply of about 32 kB
without inquiring the reply size first.

In the case of Giulio's situation there are only two
descriptors returned. So it can hardly be a matter
of a fat data transfer.

If the drive or the system is taking offense from
a command that allows 100 descriptors (= 1600 byte)
then libburn would be affected too.

To Giulio:
One could find out by downloading and installing
and then running
  test/telltoc --drive /dev/scd1
watching for output like
  Speed descr. : 5540 kB/s, 4482.6 MiB, DVD+RW
  Speed descr. : 3324 kB/s, 4482.6 MiB, DVD+RW
If those don't appear, then libburn is in trouble.

Have a nice day :)


Reply to: