Re: Pending July patches

[Tony Lindgren <tony@atomide.com>]

* andrzej zaborowski <balrog@zabor.org> [060810 11:07]:
> On 10/08/06, Tony Lindgren <tony@atomide.com> wrote:
> >* andrzej zaborowski <balrog@zabor.org> [060810 07:41]:
> >> >
> >> >Personally I didn't have problems with MMC/SD support on OMAP1 with
> >> >current git, except the 2GB and 4GB cards (but we have a workaround
> >> >for those). Instead I see strange things happen with the LCD -- It
> >> >takes about 1 second to initialise (in 2.6.16 you couldn't notice
> >>
> >> Oops, actually it was in the MMC/SD driver. Because the lines
> >>        if (!irqs_disabled())
> >>                return -EAGAIN;
> >> in mmc_omap_suspend were removed, LCD & backlight are now suspended
> >> before MMC and resumed first which results in a weird visual effect
> >> while cards are detected but no slowdown as far as I can tell.
> >
> >Do you have a patch for that?
> 
> Well, it's just aesthetics. If you put these two lines back in
> mmc_omap_suspend(), resuming will look better because LCD & backlight
> resumes last, but there's a trade-off, according to
> Documentation/power/devices.txt (which I only found out about today):
> when a driver returns -EAGAIN, it will be suspended as last but it can
> only be suspended with the whole system. Run-time suspending won't
> work (i.e. "echo 1 > /sys/devices/platform/mmci-omap.1/power/state"
> will fail) so it's probably not desirable.

OK, let's keep as it is then.

> I don't know how to force LCD to be resumed last.

Maybe the makefile order?

Tony