[Intrepid, Jaunty] SRU LP#290506

[Tim Gardner]

Tim Gardner wrote:
> Stefan Bader wrote:
>> https://bugs.launchpad.net/bugs/290506
>>
>> Jaunty would only be missing the last patch.
>>
>> SRU justification:
>>
>> Impact: Webcams using the uvcvideo driver suffered several problems
>> starting with Intrepid (starting with the move to v4l2 and the use of
>> the v4l1 library).
>>
>> Fix: This patches update several issues with the driver (also adding
>> support for a few models) and has been verified to help issues with
>> those webcams. All but the last patch are taken from upstream. The last
>> one is taken from the uvcvideo repository and should also hit upstream
>> sooner or later. All patches only change the uvcvideo driver. Impact on
>> the overall system should be minimal.
>>
>> Testcase: Webcams not working properly in cheese without.
>>
>>
>>
> 
> I would really like to see these split into multiple SRUs, each with
> their own LP report. Only 3 of these patches appear to have direct
> applicability to the v4l2 issues, the rest are either device enablement
> or just a damn good idea.
> 
> v4l2 (currently #290506)
> 0007-V4L-DVB-9031-uvcvideo-Fix-incomplete-frame-drop.patch
> 0009-V4L-DVB-9035-uvcvideo-Declare-missing-camera-and.patch
> 0012-UBUNTU-uvcvideo-Commit-streaming-parameters-when-e.patch
> 
> General Stability:
> 0001-V4L-DVB-8729-Use-DIV_ROUND_UP.patch
> 0004-V4L-DVB-8846-uvcvideo-Supress-spurious-EOF-in-e.patch
> 0008-V4L-DVB-usbvideo-don-t-use-part-of-buffer-for-USB-tr.patch
> 0010-V4L-DVB-9036-uvcvideo-Fix-control-cache-access-w.patch
> 
> Device Enablement:
> 0005-V4L-DVB-8847-uvcvideo-Add-support-for-a-Bison-El.patch
> 0006-V4L-DVB-9030-uvcvideo-Add-support-for-Advent-42.patch
> 0011-V4L-DVB-9169-uvcvideo-Support-two-new-Bison-Elec.patch
> 
> Suspend/Resume:
> 0002-V4L-DVB-8754-uvcvideo-Implement-the-USB-power-ma.patch
> 

Let me further clarify. I'm fine with collapsing all of the patches
within a category into one SRU. For example, all of the patches in
'General Stability' can go into one SRU, etc.

I'm just thinking about regression analysis issues. Martin will be very
unhappy if we drop this huge number of changes on him with no way to
distinguish root cause of possible regressions.

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