sil680 MMIO changes moved to branch

sil680 MMIO changes moved to branch

[Jeff Garzik]

I moved the sil680 MMIO changes from 'upstream' (queued for 2.6.23) to a 
side branch 'sil680-mmio'.

To be clear, I am still quite interested in the patch (w/ the reset 
fix), but I did not want to delay other unrelated stuff that was about 
to be queued into #upstream.

The patch will still appear in -mm, and it is trivial to pull it from 
#sil680-mmio into #upstream once people are happy.

	Jeff

Re: sil680 MMIO changes moved to branch

[Benjamin Herrenschmidt]

On Thu, 2007-05-24 at 23:33 -0400, Jeff Garzik wrote:
> I moved the sil680 MMIO changes from 'upstream' (queued for 2.6.23) to
> a 
> side branch 'sil680-mmio'.
> 
> To be clear, I am still quite interested in the patch (w/ the reset 
> fix), but I did not want to delay other unrelated stuff that was
> about 
> to be queued into #upstream.
> 
> The patch will still appear in -mm, and it is trivial to pull it from 
> #sil680-mmio into #upstream once people are happy. 

Fine with me. I've been busy with other things and waiting for you and
Alan to decide what should be done for the reset stuff anway :-)

So Alan, do you think we should add something or do you think that just
reading altstat like ata_pause does is good enough ?

Cheers,
Ben.

Re: sil680 MMIO changes moved to branch

[Jeff Garzik]

Benjamin Herrenschmidt wrote:
> On Thu, 2007-05-24 at 23:33 -0400, Jeff Garzik wrote:
>> I moved the sil680 MMIO changes from 'upstream' (queued for 2.6.23) to
>> a 
>> side branch 'sil680-mmio'.
>>
>> To be clear, I am still quite interested in the patch (w/ the reset 
>> fix), but I did not want to delay other unrelated stuff that was
>> about 
>> to be queued into #upstream.
>>
>> The patch will still appear in -mm, and it is trivial to pull it from 
>> #sil680-mmio into #upstream once people are happy. 
> 
> Fine with me. I've been busy with other things and waiting for you and
> Alan to decide what should be done for the reset stuff anway :-)

Just replace the ata_std_softreset() call with code that does MMIO 
flushes on a safe register?


> So Alan, do you think we should add something or do you think that just
> reading altstat like ata_pause does is good enough ?

Reading AltStatus is fine for this.

	Jeff

Re: sil680 MMIO changes moved to branch

[Benjamin Herrenschmidt]

On Fri, 2007-05-25 at 00:26 -0400, Jeff Garzik wrote:
> Just replace the ata_std_softreset() call with code that does MMIO 
> flushes on a safe register?

Yeah well... if just reading altstatus is enough, I don't see the point
of doing anything MMIO-dependant or whatever. I can just add reading of
it after all 3 SRST bit toggles unconditionally.

Ben.

Re: sil680 MMIO changes moved to branch

[Jeff Garzik]

Benjamin Herrenschmidt wrote:
> On Fri, 2007-05-25 at 00:26 -0400, Jeff Garzik wrote:
>> Just replace the ata_std_softreset() call with code that does MMIO 
>> flushes on a safe register?
> 
> Yeah well... if just reading altstatus is enough, I don't see the point
> of doing anything MMIO-dependant or whatever. I can just add reading of
> it after all 3 SRST bit toggles unconditionally.

Sorry, to be clear I would not want to change the core code to bang 
AltStatus in the middle of the SRST twiddles...

The core code only fails (in theory) for PATA+MMIO cases, and the 
easiest fix should be to read some innocuous vendor-specific register. 
That allows the core code to stay unchanged, requiring only two drivers 
(pdc and sil680) to be updated with this flushing behavior.

All other discussion of ata_pause() seemed like an unrelated side 
discussion, unless I missed something?

	Jeff

Re: sil680 MMIO changes moved to branch

[Benjamin Herrenschmidt]

On Fri, 2007-05-25 at 00:41 -0400, Jeff Garzik wrote:
> Sorry, to be clear I would not want to change the core code to bang 
> AltStatus in the middle of the SRST twiddles...
> 
> The core code only fails (in theory) for PATA+MMIO cases, and the 
> easiest fix should be to read some innocuous vendor-specific
> register. 
> That allows the core code to stay unchanged, requiring only two
> drivers 
> (pdc and sil680) to be updated with this flushing behavior.

Ok. Will do that then.

> All other discussion of ata_pause() seemed like an unrelated side 
> discussion, unless I missed something?

Yes, I think we drifted because ata_pause is the other known case where
MMIO needs flushing.

Re: sil680 MMIO changes moved to branch

[Alan Cox]

> Fine with me. I've been busy with other things and waiting for you and
> Alan to decide what should be done for the reset stuff anway :-)
> 
> So Alan, do you think we should add something or do you think that just
> reading altstat like ata_pause does is good enough ?

Reading altstat is unneccessary for PIO so we do want to get these
changes in so we can sort out performance of command sending, pause and
anything else where it would help.

The PDC driver may well want to use altstatus reads for this but thats
driver specific.