libATA SATA errors on DVD bad sectors...

["Maciej Żenczykowski" <zenczykowski@gmail.com>]

I'm getting the following on bad sectors on a pretty badly scratched DVD:

Sep 21 20:23:20 zeus kernel: ata12.00: exception Emask 0x0 SAct 0x0
SErr 0x0 action 0x6
Sep 21 20:23:20 zeus kernel: ata12.00: port_status 0x20280000
Sep 21 20:23:20 zeus kernel: ata12.00: cmd
a0/01:00:00:00:08/00:00:00:00:00/a0 tag 0 dma 2048 in
Sep 21 20:23:20 zeus kernel:         cdb a8 00 00 16 02 6c 00 00  00
01 00 00 00 00 00 00
Sep 21 20:23:20 zeus kernel:         res
51/30:03:00:00:00/00:00:00:00:00/e0 Emask 0x2 (HSM violation)
Sep 21 20:23:20 zeus kernel: ata12.00: status: { DRDY ERR }
Sep 21 20:23:20 zeus kernel: ata12: hard resetting link
Sep 21 20:23:21 zeus kernel: ata12: SATA link up 1.5 Gbps (SStatus 113
SControl 300)
Sep 21 20:23:21 zeus kernel: ata12.00: configured for UDMA/100
Sep 21 20:23:21 zeus kernel: ata12: EH complete

Sep 21 20:23:24 zeus kernel: ata12.00: exception Emask 0x0 SAct 0x0
SErr 0x0 action 0x6
Sep 21 20:23:24 zeus kernel: ata12.00: port_status 0x20280000
Sep 21 20:23:24 zeus kernel: ata12.00: cmd
a0/01:00:00:00:08/00:00:00:00:00/a0 tag 0 dma 2048 in
Sep 21 20:23:24 zeus kernel:         cdb a8 00 00 16 02 6d 00 00  00
01 00 00 00 00 00 00
Sep 21 20:23:24 zeus kernel:         res
51/30:03:00:00:00/00:00:00:00:00/e0 Emask 0x2 (HSM violation)
Sep 21 20:23:24 zeus kernel: ata12.00: status: { DRDY ERR }
Sep 21 20:23:24 zeus kernel: ata12: hard resetting link
Sep 21 20:23:24 zeus kernel: ata12: SATA link up 1.5 Gbps (SStatus 113
SControl 300)
Sep 21 20:23:25 zeus kernel: ata12.00: configured for UDMA/100
Sep 21 20:23:25 zeus kernel: ata12: EH complete

The command being run is:
sg_dd blk_sgio=1 bpt=1 bs=2048 cdbsz=12 coe=0 coe_limit=0 if=/dev/srX
odir of=badsector.bin count=1 skip=$i

on a Fedora 11 box:
Linux zeus 2.6.30.5-43.fc11.x86_64 #1 SMP Thu Aug 27 21:39:52 EDT 2009
x86_64 x86_64 x86_64 GNU/Linux

These only seem to show up when the DVD is inserted into /dev/sr1.
/dev/sr0 doesn't seem to spew this crap (although both drives fail to
read the sector).

/sys/block/sr0 ->
../devices/pci0000:00/0000:00:1f.2/host0/target0:0:0/0:0:0:0/block/sr0
/sys/block/sr1 ->
../devices/pci0000:00/0000:00:1e.0/0000:03:09.0/host11/target11:0:0/11:0:0:0/block/sr1

ata1.00: HL-DT-ST BD-RE GGW-H20L YL05 UDMA/133
ata12.00: LITE-ON DVDRW LH-20A1L BL06 UDMA/100

# lspci [-n] | egrep '1e.0|1f.2|3:09'
00:1f.2 0106: 8086:2922 SATA controller: Intel Corporation
82801IR/IO/IH (ICH9R/DO/DH) 6 port SATA AHCI Controller (rev 02)
00:1e.0 0604: 8086:244e PCI bridge: Intel Corporation 82801 PCI Bridge (rev 92)
03:09.0 0180: 105a:3d17 Mass storage controller: Promise Technology,
Inc. PDC40718 (SATA 300 TX4) (rev 02)

ie. the Lite-On drive on the promise TX4 controller has issues, while
the LG on the intel AHCI controller seems fine...

Would it make sense to add this somewhere into the code as a 'bad
sector' error condition and not perform a hard reset?

It may be worth pointing out that the LiteOn drive is significantly
faster in dealing with the bad sectors, and actually successfully
reads a far larger number of them (about 15% of the sectors that the
LG can't read, the Lite-On can, and in about 2/3 of the time).  I
would guess that if it didn't have to hard reset the bus, it'd be even
faster...

- Maciej