[Pm-utils] [PATCH 1/8] pm-utils 1.2.3 proposed patches

[Michael Biebl]

2008/11/29 Michael Biebl <mbiebl at gmail.com>:
> 2008/11/29 Matthew Garrett <mjg59 at srcf.ucam.org>:
>> On Sat, Nov 29, 2008 at 06:17:20PM +0100, Michael Biebl wrote:
>>> 2008/11/29 Matthew Garrett <mjg59 at srcf.ucam.org>:
>>> > The acpi_sleep parameter is a documented kernel interface and many
>>> > people currently rely upon it. It's not reasonable to override it.
>>> >
>>>
>>> Have you followed the discussion, why we added this?
>>> If we don't clear the acpi flags, we get all sort of weird and
>>> unexpected results.
>>
>> So make it conditional on whether the user passed acpi_sleep to the
>> kernel. You can check /proc/cmdline.
>
> Would still be confusing as hell, if e.g. a user wants to test quirks and runs
> one of the vbe-* quirks, but what he actually gets is a strange mix of
> acpi_flags and vbe-* quirks.
>
> Past experience has shown, that users were confused by the old behaviour.
>
> It also bears the risk, that we get wrong suspend quirks from users
> and add them to hal-info (unless we don't forget to ask each time if
> the users has passed acpi flags via the kernel cmd line).
>

Oh, and we obviously also want, that users send us their fdi files
with the (correct) quirks for their plattform, so everyone benefits.

Supporting flags via /proc/cmdline  in pm-utils benefits noone, imho,
and we shouldn't encourage the usage of this interface.

Michael

-- 
Why is it that all of the instruments seeking intelligent life in the
universe are pointed away from Earth?