APPLIED most: [PULL REQ][Trusty SRU] Updates for X-Gene platforms

[Tim Gardner]

On 06/30/2014 08:52 AM, Dann Frazier wrote:
> On Sun, Jun 29, 2014 at 8:08 PM, Tim Gardner <tim.gardner at canonical.com> wrote:
>> On 06/29/2014 11:35 AM, Dann Frazier wrote:
>>>
>>> This branch includes a backported gpio driver and a couple of SATA
>>> fixes for the X-Gene platform.
>>> One of the SATA fixes is a revert of a patch that came in via upstream
>>> stable that causes a regression for xgene-ahci. See commit logs and
>>> referenced BugLinks for full details.
>>>
>>> git://kernel.ubuntu.com/dannf/trusty-xgene.git for-ubuntu-20140629
>>>
>>> Change Summary
>>> =============
>>> LP: #1334823:
>>> 70a09a36 UBUNTU: [Config] CONFIG_GPIO_DWAPB=m
>>> d1dfa16 UBUNTU: SAUCE: (no-up) arm64: dts: Add Designware GPIO dts
>>> binding to APM X-Gene platform
>>> 95ab481 gpio: dwapb: use a second irq chip
>>> 2e55022 gpio: dwapb: drop irq_setup_generic_chip()
>>> b0f208a gpio: add a driver for the Synopsys DesignWare APB GPIO block
>>>
>>> LP: #1335636:
>>> 950863d UBUNTU: SAUCE: (no-up) phy-xgene: Use correct tuning for Mustang
>>>
>>> LP: #1335645:
>>> e411219 Revert "libata/ahci: accommodate tag ordered controllers"
>>>
>>
>> I've applied all but 'Revert "libata/ahci: accommodate tag ordered
>> controllers"'. I'm pretty leery of reverting a stable upstream patch in
>> order to accommodate early rev x-gene silicon. How about the approach
>> suggested in (marc.info/?l=linux-kernel&m=140202343830821&w=2) ? That has
>> the advantage of not affecting anything other then x-gene.
> 
> There's that approach, and there's another proposed by APM:
>   http://www.spinics.net/lists/linux-scsi/msg75406.html
> 
> I've been carrying APM's proposed fix in my "hyperscale" tree to fix
> builds for 3.16, and it seems fine. I'm in the process of testing
> those patches on a trusty base, which I've gone ahead and pushed to:
>   git://kernel.ubuntu.com/dannf/trusty-xgene.git for-ubuntu-20140630
> 

Those 2 patches look fine. Is that what you want to do ?

> I've also pinged the maintainer to please respond to Tejun's question here:
>   http://www.spinics.net/lists/linux-scsi/msg75482.html
> 
>> Is 'gpio: dwapb: drop irq_setup_generic_chip()' a candidate for 3.15 stable
>> ?
> 
> Sure looks like it.
> 

hint, hint.

rtg
-- 
Tim Gardner tim.gardner at canonical.com