Hi,
I just upgraded to 2.6.8 because I've had this same problem for while already. My machine suddenly starts lagging and /dev/hde stops working after that. I get something like this with dmesg:

Disabling IRQ #11
EXT3-fs error (device hde1): ext3_get_inode_loc: unable to read inode block - inode=7897448, block=15794189
end_request: I/O error, dev hde, sector 126353559
irq 11: nobody cared!
Stack pointer is garbage, not printing trace
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11
EXT3-fs error (device hde1): ext3_get_inode_loc: unable to read inode block - inode=7897407, block=15794187
Debug: sleeping function called from invalid context at include/asm/semaphore.h:119
in_atomic():1, irqs_disabled():0
[] __might_sleep+0xb2/0xd3
[] drm_free+0xb7/0x150 [fglrx]
[] __ke_down_struct_sem+0x2a/0x44 [fglrx]
[] firegl_remove_all_drawables+0x3f/0x100 [fglrx]
[] firegl_release_helper+0x546/0x750 [fglrx]
[] firegl_takedown+0x32/0xaa0 [fglrx]
[] firegl_release+0x12f/0x190 [fglrx]
[] sys_fstat64+0x37/0x39
[] __fput+0x112/0x124
[] filp_close+0x59/0x86
[] sys_close+0x63/0x96
[] syscall_call+0x7/0xb
end_request: I/O error, dev hde, sector 4287
Buffer I/O error on device hde1, logical block 528
lost page write due to I/O error on hde1
irq 11: nobody cared!
Stack pointer is garbage, not printing trace
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11
irq 11: nobody cared!
Stack pointer is garbage, not printing trace
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11


On boot I get something like this

hde: sata_error = 0x00090000, watchdog = 1, siimage_mmio_ide_dma_test_irq
ieee1394: Host added: ID:BUS[0-00:1023] GUID[0004610000052c50]
hde: dma_timer_expiry: dma status == 0x61
hde: DMA timeout error
hde: dma timeout error: status=0xd0 { Busy }

hde: DMA disabled
ide2: reset phy, status=0x00000113, siimage_reset
ide2: reset timed-out, status=0xd0
hde: status timeout: status=0xd0 { Busy }

ide2: reset phy, status=0x00000113, siimage_reset
hde: drive not ready for command
irq 11: nobody cared!
[] __report_bad_irq+0x2a/0x8b
[] note_interrupt+0x6f/0x9f
[] do_IRQ+0x17c/0x1a8
[] common_interrupt+0x18/0x20
[] __do_softirq+0x2f/0x80
[] do_softirq+0x43/0x52
=======================
[] do_IRQ+0x155/0x1a8
[] common_interrupt+0x18/0x20
[] default_idle+0x23/0x26
[] cpu_idle+0x2c/0x35
[] start_kernel+0x165/0x17f
[] unknown_bootoption+0x0/0x149
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11
ide2: reset phy dead, status=0x00000000
ide2: host reset_poll failure for hde.
hde: status timeout: status=0xd0 { Busy }

ide2: reset phy, status=0x00000000, siimage_reset
ide2: reset phy dead, status=0x00000000
hde: drive not ready for command
irq 11: nobody cared!
[] __report_bad_irq+0x2a/0x8b
[] note_interrupt+0x6f/0x9f
[] do_IRQ+0x17c/0x1a8
[] reset_pollfunc+0x0/0x1b9
[] common_interrupt+0x18/0x20
[] reset_pollfunc+0x0/0x1b9
[] ide_timer_expiry+0x118/0x220
[] ide_timer_expiry+0x0/0x220
[] run_timer_softirq+0xce/0x1ae
[] __do_softirq+0x7e/0x80
[] do_softirq+0x43/0x52
=======================
[] do_IRQ+0x155/0x1a8
[] common_interrupt+0x18/0x20
[] default_idle+0x23/0x26
[] cpu_idle+0x2c/0x35
[] start_kernel+0x165/0x17f
[] unknown_bootoption+0x0/0x149
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11
ide2: reset phy dead, status=0x00000000
ide2: host reset_poll failure for hde.
end_request: I/O error, dev hde, sector 65
irq 11: nobody cared!
[] __report_bad_irq+0x2a/0x8b
[] note_interrupt+0x6f/0x9f
[] do_IRQ+0x17c/0x1a8
[] reset_pollfunc+0x0/0x1b9
[] common_interrupt+0x18/0x20
[] reset_pollfunc+0x0/0x1b9
[] ide_timer_expiry+0x118/0x220
[] ide_timer_expiry+0x0/0x220
[] run_timer_softirq+0xce/0x1ae
[] __do_softirq+0x7e/0x80
[] do_softirq+0x43/0x52
=======================
[] do_IRQ+0x155/0x1a8
[] common_interrupt+0x18/0x20
[] default_idle+0x23/0x26
[] cpu_idle+0x2c/0x35
[] start_kernel+0x165/0x17f
[] unknown_bootoption+0x0/0x149
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11
EXT3-fs: unable to read superblock
irq 11: nobody cared!
Stack pointer is garbage, not printing trace
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11
irq 11: nobody cared!
Stack pointer is garbage, not printing trace
handlers:
[] (ide_intr+0x0/0x18e)
[] (ohci_irq_handler+0x0/0x80e)
[] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #11


Sorry for long post, I would have put dmesg to my homepage but all my passwords are in /dev/hde and I can't access them now. My motherboard is EPOX 8rda3+

Looks like a hardware problem. Most probably your HD is broken.

disable dma for that drive. seems like a bug in the kernel for me. what does acpi say about irq's?

sl mritch.

I have acpi disabled. Here's info about interrupts:

And btw I didn't mention. This problem doesn't occur on every boot. I can sometimes use computer for days before I get this error again, and then it comes several times in short time. I've tried disabling usb and acpi but it didn't help.

hde1 seems to work fine with 2.4 kernel (well... I've tested only 2 hours so I can't be sure), so maybe it's not hardware problem...
I'll try disabling dma

dma maybe isn't the issue here. you've done right deaktivating usb, that's the thing i'd recommend. the "nobody cared" is iirc about something not handled by the bios = irq not assigned.

i don't know how stable sata is in 2.6 for your device so you may be at a better address posting this to http://www.tux.org/lkml/ .

another thing. do you have the option to unmask interrupts with that drive? (hdparm -u /dev/hde - read the hdparm manpage before!)

i'd say this is some prob with the interrupt assignment. so changing the used interrupt of the sata controller may fix that. (try another pci slot?)

sl mritch.