[PATCH] drm/amdgpu: During s0ix don't wait to signal GFXOFF
Limonciello, Mario
Mario.Limonciello at amd.com
Fri Oct 1 13:40:53 UTC 2021
[Public]
> -----Original Message-----
> From: Alex Deucher <alexdeucher at gmail.com>
> Sent: Friday, October 1, 2021 08:26
> To: Lazar, Lijo <Lijo.Lazar at amd.com>
> Cc: amd-gfx list <amd-gfx at lists.freedesktop.org>; Deucher, Alexander
> <Alexander.Deucher at amd.com>; Limonciello, Mario
> <Mario.Limonciello at amd.com>; Zhang, Hawking <Hawking.Zhang at amd.com>;
> Wang, Yang(Kevin) <KevinYang.Wang at amd.com>; Quan, Evan
> <Evan.Quan at amd.com>
> Subject: Re: [PATCH] drm/amdgpu: During s0ix don't wait to signal GFXOFF
>
> On Fri, Oct 1, 2021 at 6:16 AM Lijo Lazar <lijo.lazar at amd.com> wrote:
> >
> > In the rare event when GFX IP suspend coincides with a s0ix entry, don't
> > schedule a delayed work, instead signal PMFW immediately to allow GFXOFF
> > entry. GFXOFF is a prerequisite for s0ix entry. PMFW needs to be
> > signaled about GFXOFF status before amd-pmc module passes OS HINT
> > to PMFW telling that everything is ready for a safe s0ix entry.
> >
> > Bug:
> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgitlab.fr
> eedesktop.org%2Fdrm%2Famd%2F-
> %2Fissues%2F1712&data=04%7C01%7CMario.Limonciello%40amd.com%7
> C0ff4fe8eaf34471369ff08d984df1a33%7C3dd8961fe4884e608e11a82d994e183
> d%7C0%7C0%7C637686916025223001%7CUnknown%7CTWFpbGZsb3d8eyJWIj
> oiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C100
> 0&sdata=x4FS7%2B8uSPiNwYhQdDLekjBabrQwvkBfb%2BjlVbxJWB0%3D&a
> mp;reserved=0
> >
> > Signed-off-by: Lijo Lazar <lijo.lazar at amd.com>
>
> Reviewed-by: Alex Deucher <alexander.deucher at amd.com>
Reviewed-by: Mario Limonciello <mario.limonciell at amd.com>
>
> > ---
> > drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c | 14 ++++++++++++--
> > 1 file changed, 12 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> > index e7f06bd0f0cd..1916ec84dd71 100644
> > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c
> > @@ -31,6 +31,8 @@
> > /* delay 0.1 second to enable gfx off feature */
> > #define GFX_OFF_DELAY_ENABLE msecs_to_jiffies(100)
> >
> > +#define GFX_OFF_NO_DELAY 0
> > +
> > /*
> > * GPU GFX IP block helpers function.
> > */
> > @@ -558,6 +560,8 @@ int amdgpu_gfx_enable_kcq(struct amdgpu_device
> *adev)
> >
> > void amdgpu_gfx_off_ctrl(struct amdgpu_device *adev, bool enable)
> > {
> > + unsigned long delay = GFX_OFF_DELAY_ENABLE;
> > +
> > if (!(adev->pm.pp_feature & PP_GFXOFF_MASK))
> > return;
> >
> > @@ -573,8 +577,14 @@ void amdgpu_gfx_off_ctrl(struct amdgpu_device
> *adev, bool enable)
> >
> > adev->gfx.gfx_off_req_count--;
> >
> > - if (adev->gfx.gfx_off_req_count == 0 && !adev->gfx.gfx_off_state)
> > - schedule_delayed_work(&adev->gfx.gfx_off_delay_work,
> GFX_OFF_DELAY_ENABLE);
> > + if (adev->gfx.gfx_off_req_count == 0 &&
> > + !adev->gfx.gfx_off_state) {
> > + /* If going to s2idle, no need to wait */
> > + if (adev->in_s0ix)
> > + delay = GFX_OFF_NO_DELAY;
> > + schedule_delayed_work(&adev->gfx.gfx_off_delay_work,
> > + delay);
> > + }
> > } else {
> > if (adev->gfx.gfx_off_req_count == 0) {
> > cancel_delayed_work_sync(&adev->gfx.gfx_off_delay_work);
> > --
> > 2.17.1
> >
More information about the amd-gfx
mailing list