Patchwork [Quantal] pull-request: support for SATA device sleep

login
register
mail settings
Submitter Leann Ogasawara
Date Sept. 17, 2012, 8:22 p.m.
Message ID <50578671.3080704@canonical.com>
Download mbox
Permalink /patch/184532/
State New
Headers show

Pull-request

git://kernel.ubuntu.com/ogasawara/ubuntu-quantal.git lp1032228

Comments

Leann Ogasawara - Sept. 17, 2012, 8:22 p.m.
Hi All,

It's been requested we pick up the following patches to provide support
for SATA Device Sleep.  I've cherry-picked the first patch, b1354cbb,
from upstream which then allowed me to cleanly cherry-pick the next 3
patches from the maintainers tree:

git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev.git NEXT

I've pushed the entire patch set to the following branch for review. 
I've requested testing from Tim on AMD hardware.  I've also tested
locally on Intel HW and confirm there are no regressions.  For anyone
else interested in testing, I've posted my test kernel to the following
location:

http://people.canonical.com/~ogasawara/lp1032228
<http://people.canonical.com/%7Eogasawara/lp1032228>

Thanks,
Leann

The following changes since commit 5e08f8c334bb3df6e18b2d1f970b0749288916e9:

  drm/i915: add more Haswell PCI IDs (2012-09-17 13:08:11 -0600)

are available in the git repository at:

  git://kernel.ubuntu.com/ogasawara/ubuntu-quantal.git lp1032228

for you to fetch changes up to 6455bd3beb219fa10f8bac905a9aea28db44b18c:

  UBUNTU: SAUCE: [libata] export ata_dev_set_feature() (2012-09-17
12:42:08 -0700)

----------------------------------------------------------------
Jeff Garzik (1):
      UBUNTU: SAUCE: [libata] export ata_dev_set_feature()

Lin Ming (1):
      libata: detect Device Attention support

Shane Huang (2):
      UBUNTU: SAUCE: ata: define enum constants for IDENTIFY DEVICE
      UBUNTU: SAUCE: ahci: implement aggressive SATA device sleep support

 drivers/ata/ahci.h        |   14 +++++++
 drivers/ata/libahci.c     |   97
++++++++++++++++++++++++++++++++++++++++++++-
 drivers/ata/libata-core.c |   33 ++++++++++++---
 drivers/ata/libata-eh.c   |   12 +++---
 drivers/ata/libata.h      |    2 +
 include/linux/ata.h       |   29 ++++++++++----
 include/linux/libata.h    |    6 +++
 7 files changed, 175 insertions(+), 18 deletions(-)
Tim Gardner - Sept. 17, 2012, 8:39 p.m.
On 09/17/2012 02:22 PM, Leann Ogasawara wrote:
> Hi All,
> 
> It's been requested we pick up the following patches to provide support
> for SATA Device Sleep.  I've cherry-picked the first patch, b1354cbb,
> from upstream which then allowed me to cleanly cherry-pick the next 3
> patches from the maintainers tree:
> 
> git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev.git NEXT
> 
> I've pushed the entire patch set to the following branch for review. 
> I've requested testing from Tim on AMD hardware.  I've also tested
> locally on Intel HW and confirm there are no regressions.  For anyone
> else interested in testing, I've posted my test kernel to the following
> location:
> 
> http://people.canonical.com/~ogasawara/lp1032228
> <http://people.canonical.com/%7Eogasawara/lp1032228>
> 
> Thanks,
> Leann
> 

I think we should delay application of patches in bug #1032228 until we
can get some hard data. One of those patches is pretty intrusive. I'm
not sure its worth the power savings given the maintenance hassles this
might create.

rtg
Tim Gardner - Sept. 18, 2012, 2:11 a.m.
On 09/17/2012 02:39 PM, Tim Gardner wrote:
> On 09/17/2012 02:22 PM, Leann Ogasawara wrote:
>> Hi All,
>>
>> It's been requested we pick up the following patches to provide support
>> for SATA Device Sleep.  I've cherry-picked the first patch, b1354cbb,
>> from upstream which then allowed me to cleanly cherry-pick the next 3
>> patches from the maintainers tree:
>>
>> git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev.git NEXT
>>
>> I've pushed the entire patch set to the following branch for review.
>> I've requested testing from Tim on AMD hardware.  I've also tested
>> locally on Intel HW and confirm there are no regressions.  For anyone
>> else interested in testing, I've posted my test kernel to the following
>> location:
>>
>> http://people.canonical.com/~ogasawara/lp1032228
>> <http://people.canonical.com/%7Eogasawara/lp1032228>
>>
>> Thanks,
>> Leann
>>
>
> I think we should delay application of patches in bug #1032228 until we
> can get some hard data. One of those patches is pretty intrusive. I'm
> not sure its worth the power savings given the maintenance hassles this
> might create.
>
> rtg
>

Colin - do you have any kit that we can use to measure this patch set ?
Leann Ogasawara - Sept. 18, 2012, 2:46 a.m.
On 09/17/2012 07:11 PM, Tim Gardner wrote:
> On 09/17/2012 02:39 PM, Tim Gardner wrote:
>> On 09/17/2012 02:22 PM, Leann Ogasawara wrote:
>>> Hi All,
>>>
>>> It's been requested we pick up the following patches to provide support
>>> for SATA Device Sleep.  I've cherry-picked the first patch, b1354cbb,
>>> from upstream which then allowed me to cleanly cherry-pick the next 3
>>> patches from the maintainers tree:
>>>
>>> git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev.git
>>> NEXT
>>>
>>> I've pushed the entire patch set to the following branch for review.
>>> I've requested testing from Tim on AMD hardware.  I've also tested
>>> locally on Intel HW and confirm there are no regressions.  For anyone
>>> else interested in testing, I've posted my test kernel to the following
>>> location:
>>>
>>> http://people.canonical.com/~ogasawara/lp1032228
>>> <http://people.canonical.com/%7Eogasawara/lp1032228>
>>>
>>> Thanks,
>>> Leann
>>>
>>
>> I think we should delay application of patches in bug #1032228 until we
>> can get some hard data. One of those patches is pretty intrusive. I'm
>> not sure its worth the power savings given the maintenance hassles this
>> might create.
>>
>> rtg
>>
>
> Colin - do you have any kit that we can use to measure this patch set ?
>

I've also asked the individuals who initiated this request for
additional metrics which they could share.  I'll pass any info along
once I have it.

Thanks,
Leann
Colin King - Sept. 18, 2012, 8:55 a.m.
On 18/09/12 03:11, Tim Gardner wrote:
> On 09/17/2012 02:39 PM, Tim Gardner wrote:
>> On 09/17/2012 02:22 PM, Leann Ogasawara wrote:
>>> Hi All,
>>>
>>> It's been requested we pick up the following patches to provide support
>>> for SATA Device Sleep.  I've cherry-picked the first patch, b1354cbb,
>>> from upstream which then allowed me to cleanly cherry-pick the next 3
>>> patches from the maintainers tree:
>>>
>>> git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev.git
>>> NEXT
>>>
>>> I've pushed the entire patch set to the following branch for review.
>>> I've requested testing from Tim on AMD hardware.  I've also tested
>>> locally on Intel HW and confirm there are no regressions.  For anyone
>>> else interested in testing, I've posted my test kernel to the following
>>> location:
>>>
>>> http://people.canonical.com/~ogasawara/lp1032228
>>> <http://people.canonical.com/%7Eogasawara/lp1032228>
>>>
>>> Thanks,
>>> Leann
>>>
>>
>> I think we should delay application of patches in bug #1032228 until we
>> can get some hard data. One of those patches is pretty intrusive. I'm
>> not sure its worth the power savings given the maintenance hassles this
>> might create.
>>
>> rtg
>>
>
> Colin - do you have any kit that we can use to measure this patch set ?
>
I don't believe I do, I've tested the newest kit that I have with this 
kernel and I was expecting to see one or more of the ahci flags: deso, 
sadm or sds, and I don't.  Either the my kit doesn't support this, or 
the detection isn't working (which I doubt).

Colin