[PATCH 15/16] drm/mgag200: Reject modes when h-parameters are no multiple of 8

[Daniel Vetter]

On Tue, Jul 23, 2013 at 5:30 PM, Adam Jackson <ajax at redhat.com> wrote:
> On Tue, 2013-07-23 at 15:46 +1000, Dave Airlie wrote:
>> On Wed, Jul 17, 2013 at 11:07 PM, Egbert Eich <eich at suse.com> wrote:
>> > Matrox hardware only supports modes whose horizontal parameters are
>> > multiples of 8. This rules out a mode like 1366x768 for example.
>> >
>> > Signed-off-by: Egbert Eich <eich at suse.com>
>>
>> I'd like to get a second opinion from ajax for where this should be
>> dealt with, though this is probably okay.
>
> Patch seems like it's in the right place to me.

->mode_valid is only used when parsing the list of modes obtained from
the EDID. Userspace can still add new modes of its own and force the
kernel to use them. To plug that gap mgag200 also needs to reject any
modes in the crtc_helper->mode_fixup callback.

I've pondered whether we should unify these too callers, e.g. by using
the ->mode_fixup function from ->mode_valid with a fake adjusted_mode.
But I never got around to that ...
-Daniel
--
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch