[RESEND][Quantal] [Precise SRU] drm/radeon: fix unintended display toggles

[Tim Gardner]

On 02/25/2013 04:00 AM, Alex Hung wrote:
> SRU Justification:
> 
> Impact:
>   An additional KEY_SWITCHVIDEOMODE keycode is sent with when
>   brightness up/down hotkey are pressed on some systems with AMD's
>   VGA chips. It toggles displays unintendedly when multiple monitors
>   are connected or causes LCD panel flickers with no external monitors.
> 
> Fix:
>   The standard video events may be overloaded for device specific
>   purposes. For example AMD ACPI interface overloads
>   ACPI_VIDEO_NOTIFY_PROBE (0x81) to signal AMD-specific events.
>   This patch gives the handlers the chance to examine the event and
>   block the keypress if the event is device specific. 
> 
>   The patches are included in upstream and are cherry-picked from Linus's repo.
> 
> Test:
>   I tested on ASUS 1015U, and it removed the additional keycode.
> 
> Note:
> 
> The following changes since commit 965b38cd16f964011172991c2e67b76562828b74:
> 
>   UBUNTU: Ubuntu-lts-3.5.0-25.38
> 
> are available in the git repository at:
>   git://kernel.ubuntu.com/alexhung/ubuntu-precise.git quantal-ati
> 
> for you to fetch upto bf241ddc8f9ee72fda14fe053e800b10e6c24b3c including the following patches:
> 
>   drm/radeon: block the keypress on ATIF events
>   ACPI video: allow events handlers to veto the keypress
>   drm/radeon: re-organize the acpi notifier callback
>   drm/radeon: implement handler for ACPI event
>   drm/radeon: implement wrapper for GET_SYSTEM_PARAMS
>   drm/radeon: implement radeon_atif_verify_interface
>   drm/radeon: refactor radeon_atif_call
>   drm/radeon: add backlight control for atom devices (v2)
>   drm/radeon: rework legacy backlight control
>   drm/radeon: track whether the GPU controls the backlight (v2)
>   drm/radeon: add a license header to radeon_apci.c
>   drm/radeon: add new AMD ACPI header and update relevant code (v2)
> 
> Please note drm/radeon: implement radeon_atif_verify_interface (5213fc0f09) has resolved conflicts.
> 
> BugLink: http://bugs.launchpad.net/bugs/1110219
> 

Aside from the one Radeon based laptop that you've tried, what other
regression testing has been done ? This is a giant patch set for what
should be a relatively stable kernel. Is this truly the smallest patch
you could come up with ?
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1110219/comments/4
implies that perhaps you might be able to get by with a smaller backport.

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