[Precise][Patch 0/1] Revert "samsung-laptop: make the dmi check less strict"

[Joseph Salisbury]

On 08/28/2012 04:06 PM, Herton Ronaldo Krzesinski wrote:
> On Tue, Aug 28, 2012 at 03:34:50PM -0400, joseph.salisbury at canonical.com wrote:
>> From: Joseph Salisbury<joseph.salisbury at canonical.com>
>>
>> BugLink: http://bugs.launchpad.net/bugs/1028151
>>
>> == Precise SRU Justification ==
>>
>> Reverting this patch means that samsung specific functionality(ie. rfkill, leds and so on) will not work on any systems not specifically white-listed in the samung-laptop module.  There has been only two people affected by the original bug, lp1012284.
>>
>> Not reverting this patch would would mean an 'unknown' number of systems may run into issues when loading the samsung-laptop module.  According to bug lp1028151, there are at least 27 "Affects me too".  There is a work-around in this case, which is to blacklist the module.
> Ack on the revert. BTW, a request to revert on upstream stable (3.2)
> should also be made.
Requesting revert on upstream stable today.

>
> The real issue here, is that with commit to fix lp1012284, samsung-laptop
> started to load on far more machines, including the ones where brightness
> were already being handled by ACPI video module. samsung-laptop and
> ACPI video doing the brightness handling don't goes well.
>
> This conclusion is drawn since commit f34cd9ca ("samsung-laptop: don't
> handle backlight if handled by acpi/video") would fix the regression too
> instead of doing the revert. One of Joseph's testing kernels in the
> bug applied f34cd9ca and was confirmed to also fix the problem.
>
> But to pick commit f34cd9ca, it would need to be manually backported, or
> including these two commits before to apply successfuly, which are may
> be too many changes for anything stable:
> a6df48943a408b493d1aa141791d614a529d484e
> 5dea7a2094d5e60fe8f8ec4277d22d7ad6fa8c26
>
> The backport is not hard, but the code changes ends up different. Not a
> problem, but then looking at samsung-laptop changes, commit f34cd9ca
> seems to not be enough to address the regression, or would introduce
> another regression for some users, fixed by these commits:
> a979e2e2af7d5b4bb3b20f6a716c627bb23a6753
> samsung-laptop: unregister ACPI video module for some well known laptops
> 09d5677cf18f1b9f0e092eb83e9ba6a771c15c5c
> samsung-laptop: X360 ACPI backlight device is broken
>
> So it would be a lot of changes for an stable, and the quirk problem
> doesn't look to be really solved, perhaps only less quirk entries are
> needed now (for machines which advertises acpi/video interface, but it
> doesn't work, and samsung-laptop is really needed). It seems reasonable
> to just revert the change right now on stables.
>
>>
>> == Fix ==
>> Revert the following commit:
>>
>> commit 9a2e71ce177a1602a0502fbcbb3f11e4bae584e4
>> Author: Corentin Chary<corentincj at iksaif.net>
>> Date:   Sat Nov 26 11:00:10 2011 +0100
>>
>>      samsung-laptop: make the dmi check less strict
>>
>>
>>
>> Joseph Salisbury (1):
>>    Revert "samsung-laptop: make the dmi check less strict"
>>
>>   drivers/platform/x86/samsung-laptop.c |  235 +++++++++++++++++++++++++++++++--
>>   1 file changed, 227 insertions(+), 8 deletions(-)
>>
>> -- 
>> 1.7.9.5
>>
>>
>> -- 
>> kernel-team mailing list
>> kernel-team at lists.ubuntu.com
>> https://lists.ubuntu.com/mailman/listinfo/kernel-team
>>