[indiana-discuss] changes in the ata driver between builds 111b and 117?

[Michael Schuster]

I hope this is the right place to ask this:

on my private no-name PC (Asus P5N73AM MB, Nvidia Graphics, internal sata 
disk, external USB disk), I've been seeing hangs and panics during boot 
since I upgraded from 111b to 117, this behaviour persists on 118.
I saw a few panic messages which looked like I/O to me, but I have not 
managed to capture a dump; I managed to get a screen shot of a hang during 
boot of 118 just a few minutes ago (see attachment).

(I see complaints about two disks:
" /pci at 0..../ide at 1 (ata3):
timeout:reset bus ..."
and
"/pci at 0....ide@1/cmdk at 0.0 (Disk0):
Error for command 'read sector' Error Level: Informational
Sense Key: aborted command")

when I reboot to 111b and look at fmdump -e, I see a list like this:

Jul 28 19:56:51.4645 ereport.io.scsi.cmd.disk.dev.uderr
Jul 28 21:47:41.8053 ereport.io.scsi.cmd.disk.dev.uderr
Jul 28 21:49:05.4363 ereport.io.scsi.cmd.disk.dev.uderr
Jul 28 22:00:34.6571 ereport.io.scsi.cmd.disk.dev.uderr
Jul 28 22:01:57.8560 ereport.io.scsi.cmd.disk.dev.uderr

and an example of the output of fmdump -evV looks like this:

Jul 28 2009 19:56:51.464542428 ereport.io.scsi.cmd.disk.dev.uderr
nvlist version: 0
         class = ereport.io.scsi.cmd.disk.dev.uderr
         ena = 0x1f4b22bbb200001
         detector = (embedded nvlist)
         nvlist version: 0
                 version = 0x0
                 scheme = dev
                 device-path = /pci at 0,0/pci1043,82ae at 4,1/storage at 4/disk at 0,0
         (end detector)

         driver-assessment = fail
         op-code = 0x1a
         cdb = 0x1a 0x0 0x8 0x0 0x18 0x0
         pkt-reason = 0x0
         pkt-state = 0x1f
         pkt-stats = 0x0
         stat-code = 0x0
         un-decode-info = sd_get_write_cache_enabled: Mode Sense caching 
page code mismatch 48

         un-decode-value =
         __ttl = 0x1
         __tod = 0x4a6fba73 0x1bb05adc

In order to see whether the disk (or more precisely, the partition I've 
assigned to opensolaris) has any difficulties, I tried filling it up with 
"cat bigfile > anotherbigfile" in a loop until the pool was almost 
completely full, and got no errors (all this running 111b).

Does anyone have any idea what might be going on here?

tia
Michael
-- 
Michael Schuster        http://blogs.sun.com/recursion
Recursion, n.: see 'Recursion'