Re: [git patches] libata updates for 2.6.34

[Zeno Davatz <zdavatz@gmail.com>]

On Mon, Mar 15, 2010 at 2:32 PM, Jeff Garzik <jeff@garzik.org> wrote:
> On 03/15/2010 09:21 AM, Zeno Davatz wrote:
>>
>> On Mon, Mar 15, 2010 at 2:06 PM, Jeff Garzik<jeff@garzik.org>  wrote:
>>>
>>> On 03/15/2010 03:33 AM, Zeno Davatz wrote:
>>>>
>>>> On Mon, Mar 15, 2010 at 3:55 AM, Jeff Garzik<jeff@garzik.org>    wrote:
>>>>>
>>>>> On 03/09/2010 11:26 PM, Tejun Heo wrote:
>>>>>>
>>>>>> Hello, Linus, Jeff.
>>>>>>
>>>>>> On 03/10/2010 07:12 AM, Jeff Garzik wrote:
>>>>>>>
>>>>>>> Coincedentally, it looks like someone else just reported the same
>>>>>>> problem, with 2.6.34-rc1.
>>>>>>>
>>>>>>> It definitely sounds like a race.  READ DMA is a DMA command as the
>>>>>>> name
>>>>>>> implies, so that eliminates the possibility of polling-related paths
>>>>>>> in
>>>>>>> ata_sff_interrupt (libata-sff.c).
>>>>>>>
>>>>>>> I'll flip some of my machines to the icky slow boring piix mode,
>>>>>>> rather
>>>>>>> than sexy AHCI mode :) to see if I can reproduce.  I have had a
>>>>>>> feeling
>>>>>>> that we needed a more sophisticated IRQ handling setup, this may be
>>>>>>> what
>>>>>>> was needed.  Lost interrupt recovery should occur faster than 30
>>>>>>> seconds
>>>>>>> in any case, and should not require a hard reset if the hardware
>>>>>>> functions just fine outside of the lost-interrupt / race that just
>>>>>>> occurred.
>>>>>>
>>>>>> Yeap, there is a race condition with clearing which I don't think we
>>>>>> can solve completely but with some modification I think we can at
>>>>>> least cover known failure cases.
>>>>>>
>>>>>> For longer term, I don't think we can solve this by diddling with the
>>>>>> SFF registers.  The interface is just way too ancient and horrid to
>>>>>> build anything reliable on top of.  I'm planning on implementing
>>>>>> smarter IRQ storm handling and stepped timeouts for ATA commands.
>>>>>
>>>>> A tester on this bug
>>>>>        http://bugzilla.kernel.org/show_bug.cgi?id=15537
>>>>> seemed to find success with the patch.
>>>>
>>>> Thanks for the Update!
>>>>
>>>> I will wait some more and then test rc-2.
>>>
>>> Can you test the patch, please?
>>
>> Sure. I done:
>>
>> /usr/src/linux>  sudo patch -p1<  teo
>> patching file drivers/ata/libata-sff.c
>> Hunk #1 FAILED at 1667.
>> Hunk #2 FAILED at 1700.
>> Hunk #3 FAILED at 1718.
>> Hunk #4 FAILED at 1770.
>> Hunk #5 FAILED at 1792.
>> Hunk #6 FAILED at 1801.
>> Hunk #7 FAILED at 1818.
>> 7 out of 7 hunks FAILED -- saving rejects to file
>> drivers/ata/libata-sff.c.rej
>>
>> Is that the expected outcome? Can I ignore the "failed" output?
>
> With 7 out of 7 hunks failing, nothing got modified.
>
> Is your source tree an unmodified, vanilla 2.6.34-rc1 tree?

I normally only run "sudo git pull" on the git directory of Linus
every time a new RC is out. Then I do a "make silentoldconfig" a
"make", cp the bzImage, lilo -v and reboot.

I have not applied any patches on my kernel directory for ages.

Best
Zeno