On Mon, Apr 09, 2018 at 09:39:06PM +0200, Takashi Iwai wrote: > It must be sleepable context as it's a standard platform driver probe > callback.
> And now looking at grep output, only sound/soc/intel contains so many > calls with GFP_ATOMIC. I bet that almost all can be done with > GFP_KERNEL, maybe only one or two in atom/sst would be conditionally > with GFP_ATOMIC. Yeah, it's not like it'd be the first code quality problem found in the Intel drivers.
Description: PGP signature