drm/etnaviv: keep MMU context across runtime suspend/resume
authorLucas Stach <l.stach@pengutronix.de>
Fri, 20 Aug 2021 20:18:26 +0000 (22:18 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 22 Sep 2021 10:27:55 +0000 (12:27 +0200)
commit6b1c223d8a53a9b08f06c65d932eeb3278022ee8
tree2ea2a8120eebfdb838b23f55b7782e0e9fa065bb
parentc63e6e0951ff4ab021947b834db37f591781ee28
drm/etnaviv: keep MMU context across runtime suspend/resume

commit 8f3eea9d01d7b0f95b0fe04187c0059019ada85b upstream.

The MMU state may be kept across a runtime suspend/resume cycle, as we
avoid a full hardware reset to keep the latency of the runtime PM small.

Don't pretend that the MMU state is lost in driver state. The MMU
context is pushed out when new HW jobs with a different context are
coming in. The only exception to this is when the GPU is unbound, in
which case we need to make sure to also free the last active context.

Cc: stable@vger.kernel.org # 5.4
Reported-by: Michael Walle <michael@walle.cc>
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Tested-by: Michael Walle <michael@walle.cc>
Tested-by: Marek Vasut <marex@denx.de>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/gpu/drm/etnaviv/etnaviv_gpu.c