Przemysław Kopa reports that since commit b516ea586d71 ("PCI: Enable NVIDIA HDA controllers"), the discrete GPU Nvidia GeForce GT 540M on his 2011 Samsung laptop refuses to runtime suspend, resulting in a power regression and excessive heat.
Rivera Valdez witnesses the same issue with a GeForce GT 525M (GF108M) of the same era, as does another Arch Linux user named "R0AR" with a more recent GeForce GTX 1050 Ti (GP107M).
The commit exposes the discrete GPU's HDA controller and all four codecs on the controller do not set the CLKSTOP and EPSS bits in the Supported Power States Response. They also do not set the PS-ClkStopOk bit in the Get Power State Response. hda_codec_runtime_suspend() therefore does not call snd_hdac_codec_link_down(), which prevents each codec and the PCI device from runtime suspending.
The same issue is present on some AMD discrete GPUs and we addressed it by forcing runtime PM despite the bits not being set, see commit 57cb54e53bdd ("ALSA: hda - Force to link down at runtime suspend on ATI/AMD HDMI").
Do the same for Nvidia HDMI codecs.
Fixes: b516ea586d71 ("PCI: Enable NVIDIA HDA controllers") Link: https://bbs.archlinux.org/viewtopic.php?pid=1865512 Link: https://bugs.freedesktop.org/show_bug.cgi?id=75985#c81 Reported-by: Przemysław Kopa prymoo@gmail.com Reported-by: Rivera Valdez riveravaldez@ysinembargo.com Signed-off-by: Lukas Wunner lukas@wunner.de Cc: Daniel Drake dan@reactivated.net Cc: stable@vger.kernel.org # v5.3+ --- sound/pci/hda/patch_hdmi.c | 2 ++ 1 file changed, 2 insertions(+)
diff --git a/sound/pci/hda/patch_hdmi.c b/sound/pci/hda/patch_hdmi.c index bca5de7..795cbda 100644 --- a/sound/pci/hda/patch_hdmi.c +++ b/sound/pci/hda/patch_hdmi.c @@ -3474,6 +3474,8 @@ static int patch_nvhdmi(struct hda_codec *codec) nvhdmi_chmap_cea_alloc_validate_get_type; spec->chmap.ops.chmap_validate = nvhdmi_chmap_validate;
+ codec->link_down_at_suspend = 1; + generic_acomp_init(codec, &nvhdmi_audio_ops, nvhdmi_port2pin);
return 0;