Re: 4.7GB DVD-RAM geometry wrong?

[Hisaaki Shibata <shibata@luky.org>]

Thanks Dir Axboe,

> I don't have the -ac[5] tree handy, in the 2.4.9-pre4 there's no BUG in
> get_empty_inode(). For DVD-RAM + UDF, I would suggest using the version
> from CVS. I know it quite recently got some fixes that removed a BUG or
> two when used with the -ac kernels.
> 
> So could you reproduce with latest -ac or Linus kernel + CVS UDF?

At first, I tried it with 2.4.9-pre4 and udf-0.9.4.1.
I did not get segv nor Oops with them.

But it seems bad yet.

My operations are here.
# mkudf /dev/hdb
# mount -t udf /dev/hdb /mnt/floppy/
# cp -auv ~/dir /mnt/floppy/ 1> /tmp/cp.log 2> /tmp/cp.log &

"cp" seems good until following "df" point.
/dev/hdb               4470016   1639798   2830218  37% /mnt/floppy

After that point, "tail -f /tmp/cp.log" shows "cp going :-)" and
my DVD-RAM driveseeks like working good.
But result of "df" command is not change. 
(it shows 37% forever...)

"df -i" shows
/dev/hdb             1440735   25626 1415109    2% /mnt/floppy

tails of dmesg are;
UDF-fs INFO UDF 0.9.4.1-rw (2001/06/13) Mounting volume '', timestamp 2001/08/16 21:54 (121c)
hdb: irq timeout: status=0xd0 { Busy }
hdb: ATAPI reset complete
hdb: irq timeout: status=0xd0 { Busy }
hdb: ATAPI reset complete
hdb: irq timeout: status=0xd0 { Busy }
hdb: ATAPI reset complete
hdb: irq timeout: status=0xd0 { Busy }
hdb: ATAPI reset complete
hdb: irq timeout: status=0xd0 { Busy }
hdb: ATAPI reset complete

I will try with CVS udf patch and will report it.

-- 
 WWWWW  shibata@luky.org
 |O-O|  Hisaaki Shibata @ Fukuoka, JAPAN
0(mmm)0 IRC: #luky
   ~    http://his.luky.org/ http://hoop.euqset.org/