Hi,

Le vendredi 30 octobre 2026 à 20:05 +0200, Igor Velkov a écrit :
> On Fri, 30 Oct 2026 08:49:03 -0400, Nicolas Dufresne wrote:
> > So if CONFIG_PM is not set, the initial resume will never occur, and first mmio
> > access will likely hang the CPU. We need a clean pattern to handle this, and
> > perhaps review other similar drivers while at it.
> 
> PM=n only happens with COMPILE_TEST: ARCH_ROCKCHIP selects PM on arm64 and
> arm. rkvdec had the same gap until 6a846f7d72c7; rga and rkisp1 have it today.
> v2 keeps the clocks enabled from probe when PM is off, as etnaviv and v3d do:
> https://lore.kernel.org/r/20261030-rkvdec-runtime-pm-clocks-v2-1-8ff510a231de@iav.lv
> rga and rkisp1 as separate patches?

Thanks for v2, and yes, best to send0 independent patches per driver assuming
you have time to do so.

regards,
Nicolas

> 
> _______________________________________________
> Linux-rockchip mailing list
> Linux-rockchip@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-rockchip