Alexander Clausen writes:
> Mikael Pettersson wrote:
> > >
> > > Based on several other libata sata drivers I think handling suspend in
> > > sata_promise may be as simple as hooking ata_pci_device_{suspend,resume},
> > > which is what the patch below does. If this doesn't work then the most
> > > likely solution is to augment the ->resume op with a Promise-specific
> > > controller reset (again similar to several other sata drivers).
> > >
> > > Care to give it a try? The patch is against 2.6.28-rc5 but should also
> > > work in 2.6.27.
> >
> > Alexander, have you had a chance to test this? Preferably together
> > with Tejun's hardreset improvement patch posted recently.
> > <http://marc.info/?l=linux-ide&m=122724338905689&w=2>
> >
> > With both this and Tejun's hardreset patch applied my Promise test
> > machine survives several suspend/resume cycles without problems.
>
> Did not work here. I applied both patches to 2.6.27.6, seems to behave exactly as before.
> dmesg attached.
Are you absolutely certain both patches were applied and the driver rebuilt?
I'm asking because I would have expected to see
[...] sata_promise 0000:00:0d.0: restoring config space at offset 0x1 (...)
but your kernel log does not contain anything like that. Both machines
I tested on today consistently included this message after the patch was
applied, but did not when the patch wasn't applied.
I did note some temporary errors (that sata_promise recovered from) on
one machine after resuming from suspend to ram that did not occur when
using standby. I'll update the patch to include a controller reinit in
the ->resume op, which should hopefully fix that.
On some possibly unrelated notes, I see that:
a) all your interrupts seem to be legacy IRQs < 16, aren't you
enabling the I/O-APIC?
b) your kernel includes the nvidia binary-only driver, which puts
a big unknown into the picture
/Mikael
--
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
