I'm having an odd problem with my SATA controller and/or drives. Every
so often the kernel appears to shutdown my controller (or something
similar) and access to the hard drives is lost. I get the following is
syslog:

Mar 28 05:19:45 galaxia kernel: [832818.100722] irq 201: nobody cared
(try booting with the "irqpoll" option)
Mar 28 05:21:15 galaxia kernel: [832818.114955]
Mar 28 05:23:15 galaxia kernel: [832818.114956] Call Trace:
{__report_bad_irq+53}
{note_interrupt+538}
Mar 28 05:23:15 galaxia kernel: [832818.114984]
{__do_IRQ+215} {do_IRQ+47}
Mar 28 05:23:15 galaxia kernel: [832818.115004]
{ret_from_intr+0}
{preempt_schedule_irq+93}
Mar 28 05:23:15 galaxia kernel: [832818.115015]
{retint_kernel+38}
{:vmmon:Task_Switch+2407}
Mar 28 05:23:15 galaxia kernel: [832818.115037]
{:vmmon:Vmx86_RunVM+82}
{:vmmon:LinuxDriver_Ioctl+361}
Mar 28 05:23:15 galaxia kernel: [832818.115066]
{:vmmon:LinuxDriver_Ioctl+386}
{lock_timer_base+33}
Mar 28 05:23:15 galaxia kernel: [832818.115086]
{:vmnet:VNetUserIfPoll+0}
{:vmmon:LinuxDriver_CompatIoctl+48}
Mar 28 05:23:15 galaxia kernel: [832818.115116]
{compat_sys_ioctl+241}
{do_gettimeofday+94}
Mar 28 05:23:15 galaxia kernel: [832818.115127]
{cstar_do_call+27}
Mar 28 05:23:15 galaxia kernel: [832818.115149] handlers:
Mar 28 05:23:15 galaxia kernel: [832818.122398] [_end
+130070880/2132185088] (pdc_interrupt+0x0/0x200 [sata_promise])
Mar 28 05:29:15 galaxia kernel: [832818.129778] Disabling IRQ #201
Mar 28 05:29:15 galaxia kernel: [832818.137046] rtc: lost some
interrupts at 1024Hz.
Mar 28 05:29:15 galaxia kernel: [832848.085956] ata4: command timeout
Mar 28 05:29:15 galaxia kernel: [832848.093330] ata4: status=0x50
{ DriveReady SeekComplete }
Mar 28 05:29:15 galaxia kernel: [832848.093336] sdb: Current: sense
key: No Sense
Mar 28 05:29:15 galaxia kernel: [832848.093339] Additional sense:
No additional sense information
Mar 28 05:36:15 galaxia kernel: [832848.832518] ata5: command timeout
Mar 28 05:36:15 galaxia kernel: [832848.839765] ata5: status=0x50
{ DriveReady SeekComplete }
Mar 28 05:36:15 galaxia kernel: [832848.839771] sdc: Current: sense
key: No Sense
Mar 28 05:36:15 galaxia kernel: [832848.839774] Additional sense:
No additional sense information
Mar 28 05:36:15 galaxia kernel: [832849.535177] ata3: command timeout
Mar 28 05:36:15 galaxia kernel: [832849.542383] ata3: status=0x50
{ DriveReady SeekComplete }
Mar 28 05:36:15 galaxia /USR/SBIN/CRON[17836]: (root) CMD (if [ -x /
usr/bin/mrtg ] && [ -r /etc/mrtg.cfg ]; then env LANG=C /usr/bin/mrtg
--lock-file /var/lib/mrtg/mrtg /etc/mrtg.cfg >> /var/log/mrtg/mrtg.log
2>&1; fi)
Mar 28 06:22:46 galaxia kernel: [832849.542389] sda: Current: sense
key: No Sense
Mar 28 06:22:46 galaxia kernel: [832849.542392] Additional sense:
No additional sense information
Mar 28 06:22:46 galaxia kernel: [832878.036506] ata6: command timeout
Mar 28 06:22:46 galaxia kernel: [832878.043594] ata6: status=0x50
{ DriveReady SeekComplete }
Mar 28 06:22:46 galaxia /USR/SBIN/CRON[17840]: (root) CMD (if [ -x /
usr/bin/mrtg ] && [ -r /etc/mrtg.cfg ]; then env LANG=C /usr/bin/mrtg
--lock-file /var/lib/mrtg/mrtg /etc/mrtg.cfg >> /var/log/mrtg/mrtg.log
2>&1; fi)
Mar 28 06:22:46 galaxia kernel: [832878.043599] sdd: Current: sense
key: No Sense
Mar 28 06:22:46 galaxia kernel: [832878.043601] Additional sense:
No additional sense information
Mar 28 06:22:46 galaxia kernel: [832878.043673] ata4: command timeout
Mar 28 06:22:46 galaxia kernel: [832878.050375] ata4: status=0x50
{ DriveReady SeekComplete }
Mar 28 06:22:46 galaxia kernel: [832878.050378] sdb: Current: sense
key: No Sense
Mar 28 06:22:46 galaxia kernel: [832878.050380] Additional sense:
No additional sense information
Mar 28 06:22:46 galaxia kernel: [832878.779077] ata5: command timeout

The end result is that I have to reboot.

I have 4 hard drives in my machine:
[ 20.148076] SCSI device sda: 488397168 512-byte hdwr sectors
(250059 MB)
[ 20.170046] SCSI device sdb: 625142448 512-byte hdwr sectors
(320073 MB)
[ 20.229639] SCSI device sdc: 398297088 512-byte hdwr sectors
(203928 MB)
[ 20.250513] SCSI device sdd: 625142448 512-byte hdwr sectors
(320073 MB)

They are setup in a RAID configuration using linux software RAID (md).
The SATA controller I'm using is a Promise TX4:
05:0b.0 Mass storage controller: Promise Technology, Inc. PDC20718
(SATA 300 TX4) (rev 02)
Subsystem: Promise Technology, Inc. PDC20718 (SATA 300 TX4)

Now I seem to have roughly pinpointed it to when the machine gets
"knocked". It sits on my desk and if the chair hits the table, it is
possible for this to occur. As I said this is an "educated guess"
based on circumstances. It could very well be the cables but I've
swapped a couple of them. It's still on my "to do" list to go out and
purchase all new ones. I thought I would post here and see if anyone
else has encountered this problem.

thanks,

Aaron