[ANNOUNCEMENT] Wiki for discussing P35/IHC9(R)/SATA issues set up

[Sean C. Farley]

On Mon, 5 Nov 2007, Xin LI wrote:

> Aryeh M. Friedman wrote:
>> I have submitted the complete verbose dmesg, pciutil -l and
>> everything to several people and all of them have come to the
>> conclusion that there is no obvious smoking gone beyond ("it just
>> doesn't work")...  Thus the comment of ICH9 owners are on their
>> own... Next time I reboot I will send you a copy if you really want
>> it.
>
> Yes, please.  These information is very important for knowing how to
> change the situation, thanks.

I have two systems exhibiting problems:

1. Dell Optiplex 745:
ad4: 76293MB <WDC WD800JD-75MSA3 10.01E04> at ata2-master SATA150
acd0: CDRW <HL-DT-ST CD-RW/DVD-ROM GCC-H10N/C103> at ata3-master SATA150

...
ATA channel 2:
     Master:  ad4 <WDC WD800JD-75MSA3/10.01E04> Serial ATA II
     Slave:       no device present
ATA channel 3:
     Master: acd0 <HL-DT-ST CD-RW/DVD-ROM GCC-H10N/C103> Serial ATA v1.0
     Slave:       no device present
...

Problem 1:  ad4 recoginized as SATA/150 when it is a SATA/300.

Problem 2:  errors occur when reading a CD-RW (a BETA release)
Nov  5 15:57:55 newsys kernel: acd0: TIMEOUT - READ_BIG retrying (1 retry left)
Nov  5 15:58:27 newsys kernel: acd0: TIMEOUT - READ_BIG retrying (0 retries left)
Nov  5 15:59:00 newsys kernel: acd0: FAILURE - READ_BIG timed out
Nov  5 15:59:00 newsys kernel: g_vfs_done():acd0[READ(offset=387680256, length=14336)]error = 5

This showed up during install (used NFS to get around it), but I can
reproduce it by mounting the drive and running:
cd /cdrom/7-BETA && grep -Rl GENERIC

2. Asus P5K-E/WiFi:
ad10: 152627MB <Seagate ST3160827AS 3.42> at ata5-master SATA150
ad12: 152627MB <Seagate ST3160827AS 3.42> at ata6-master SATA150
acd0: <LITE-ON DVDRW LH-20A1S/9L08> DVDR drive at ata10 as master

...
ATA channel 5:
     Master: ad10 <ST3160827AS/3.42> Serial ATA v1.0
     Slave:       no device present
ATA channel 6:
     Master: ad12 <ST3160827AS/3.42> Serial ATA v1.0
     Slave:       no device present
...
ATA channel 10:
     Master: acd0 <LITE-ON DVDRW LH-20A1S/9L08> Serial ATA v1.0
     Slave:       no device present
...

Problem:  errors when reading an Unreal Tournament CD directly
(non-mounted)
Nov  4 17:45:00 thor kernel: ata10: reiniting channel ..
Nov  4 17:45:00 thor kernel: ata10: SATA connect time=0ms
Nov  4 17:45:00 thor kernel: ata10: SIGNATURE: eb140101
Nov  4 17:45:00 thor kernel: ata10: ahci_reset devices=0x4<ATAPI_MASTER>
Nov  4 17:45:00 thor kernel: ata10: reinit done ..
Nov  4 17:45:00 thor kernel: acd0: TIMEOUT - READ_BIG retrying (1 retry left)
Nov  4 17:45:31 thor kernel: ata10: reiniting channel ..
Nov  4 17:45:31 thor kernel: ata10: SATA connect time=0ms
Nov  4 17:45:31 thor kernel: ata10: SIGNATURE: eb140101
Nov  4 17:45:31 thor kernel: ata10: ahci_reset devices=0x4<ATAPI_MASTER>
Nov  4 17:45:31 thor kernel: ata10: reinit done ..
Nov  4 17:45:31 thor kernel: acd0: TIMEOUT - READ_BIG retrying (0 retries left)
Nov  4 17:46:01 thor kernel: ata10: reiniting channel ..
Nov  4 17:46:01 thor kernel: ata10: SATA connect time=0ms
Nov  4 17:46:01 thor kernel: ata10: SIGNATURE: eb140101
Nov  4 17:46:01 thor kernel: ata10: ahci_reset devices=0x4<ATAPI_MASTER>
Nov  4 17:46:01 thor kernel: ata10: reinit done ..
Nov  4 17:46:01 thor kernel: acd0: FAILURE - READ_BIG timed out

I also can receive:
Nov  4 16:08:47 thor kernel: acd0: FAILURE - MODE_SENSE_BIG timed out
Nov  4 17:40:55 thor kernel: acd0: FAILURE - READ_TRACK_INFO timed out

I put the dmesg.boot and pciconf logs along with a patch for ICH9R--is
it correct?--and a test program with pieces from burncd.c here[1].  For
me the test program, which does not burn, will cause timeout errors on
my Asus board but not the Dell (grep will as seen above).

Sean
   1. http://www.farley.org/freebsd/tmp/SATA/
-- 
scf at FreeBSD.org