On 04/05/2011 02:50 PM, Tim Gardner wrote:
> On 04/05/2011 01:27 AM, Stefan Bader wrote:
>> On 04/04/2011 09:24 PM, Leann Ogasawara wrote:
>>> Hi Stefan,
>>> When opening the Oneiric repo, the following patch produces a build
>>> failure due to the fact that it's still a consumer of the BKL:
>>> Author: Stefan Bader<email@example.com>
>>> Date: Tue Mar 4 22:10:36 2008 +0000
>>> UBUNTU: SAUCE: (no-up) Restore VT fonts on switch
>>> Not all X drivers save and restore fonts on text VTs. Add code to the
>>> kernel to explicitly save and restore them on VT switches.
>>> Signed-off-by: Matthew Garrett<firstname.lastname@example.org>
>>> Signed-off-by: Stefan Bader<email@example.com>
>>> Signed-off-by: Ben Collins<firstname.lastname@example.org>
>>> I'm inclined to just drop this patch completely from Oneiric given that
>>> the patch is 3yrs old, marked as SAUCE: (no-up), and the state of X
>>> drivers has improved since then. We can always re-apply it should
>>> someone scream. I wanted to get your thoughts first?
>> I guess it depends a bit what fallbacks we got. If memory serves this was only
>> for vga framebuffer and only helping if the X driver did not handle the fonts.
>> So basically helping for the special last resort graphical mode.
>> Our reasoning before was going along the lines we did for the mystic rfkill
>> special drivers we hump along. Which is, as long as nobody complains about it
>> being there and it continues to simply apply, why not keep it just in case.
>> Given that most users end up with i915, radeon or nouveau frambuffer devices (or
>> (u)efi on o) it probably does not matter that much. More the old stable
>> maintainers thinking: As long as there ain't complaints, dun't tuch it.
> I'm not really sure what Stefan just said, but I think we should be able to drop
> this patch.
*sigh* Sometimes I wonder why I bother... :-P What I said was that this patch
could still be useful, even if the number of people affected likely is going down.
As long as carrying the patch does not cause problems, why should we remove it?
Just because upstream did not care that much for that detail?
kernel-team mailing list