Re: [PATCH 3/3] drm/xe: Use the encoder suspend helper also used by the i915 driver
On Mon, Jun 17, 2024 at 08:54:13PM +0300, Imre Deak wrote:
> On Mon, Jun 17, 2024 at 01:44:06PM -0400, Rodrigo Vivi wrote:
> > On Mon, Jun 17, 2024 at 08:03:56PM +0300, Imre Deak wrote:
> > > Use the suspend encoders helper which is also used by the i915 driver.
> > > This fixes an issue in the xe driver where the encoder
> > > suspend_complete() hook is not called and was an overlook when this hook
> > > was added in
> >
> > I believe it is okay to stay in the same line and then break the commit
> > subject in the middle for the next line...
>
> Ok, looks like done elsewhere as well, will reformat.
nah, not needed. this is clear enough if checkpatch is not complainin no need
to change.
>
> > > Commit b61fad5f7e5d ("drm/i915/tc: Call TypeC port flush_work/cleanup
> > > without modeset locks held")
> > >
> >
> > I wonder if we should have a fixes, but probably okay to skip it.
>
> I wasn't sure, since strictly the tree itself at the above commit didn't
> have the issue. I suppose if needed this would need to be sent to 6.9
> stable manually.
good point!
thanks a lot,
Rodrigo.
>
> > Reviewed-by: Rodrigo Vivi
>
> Thanks.
>
> >
> > > for i915, but not added to the xe driver (which was still in a separate
> > > source tree at the above point).
> > >
> > > Signed-off-by: Imre Deak
> > > ---
> > > drivers/gpu/drm/xe/display/xe_display.c | 18 ++
> > > 1 file changed, 2 insertions(+), 16 deletions(-)
> > >
> > > diff --git a/drivers/gpu/drm/xe/display/xe_display.c
> > > b/drivers/gpu/drm/xe/display/xe_display.c
> > > index 78cccbe289475..8b83dcff72e17 100644
> > > --- a/drivers/gpu/drm/xe/display/xe_display.c
> > > +++ b/drivers/gpu/drm/xe/display/xe_display.c
> > > @@ -23,6 +23,7 @@
> > > #include "intel_display_types.h"
> > > #include "intel_dmc.h"
> > > #include "intel_dp.h"
> > > +#include "intel_encoder.h"
> > > #include "intel_fbdev.h"
> > > #include "intel_hdcp.h"
> > > #include "intel_hotplug.h"
> > > @@ -270,21 +271,6 @@ void xe_display_irq_postinstall(struct xe_device
> > > *xe, struct xe_gt *gt)
> > > gen11_de_irq_postinstall(xe);
> > > }
> > >
> > > -static void intel_suspend_encoders(struct xe_device *xe)
> > > -{
> > > - struct drm_device *dev = >drm;
> > > - struct intel_encoder *encoder;
> > > -
> > > - if (has_display(xe))
> > > - return;
> > > -
> > > - drm_modeset_lock_all(dev);
> > > - for_each_intel_encoder(dev, encoder)
> > > - if (encoder->suspend)
> > > - encoder->suspend(encoder);
> > > - drm_modeset_unlock_all(dev);
> > > -}
> > > -
> > > static bool suspend_to_idle(void)
> > > {
> > > #if IS_ENABLED(CONFIG_ACPI_SLEEP)
> > > @@ -315,7 +301,7 @@ void xe_display_pm_suspend(struct xe_device *xe, bool
> > > runtime)
> > >
> > > intel_hpd_cancel_work(xe);
> > >
> > > - intel_suspend_encoders(xe);
> > > + intel_encoder_suspend_all(>display);
> > >
> > > intel_opregion_suspend(xe, s2idle ? PCI_D1 : PCI_D3cold);
> > >
> > > --
> > > 2.43.3
> > >
Re: [PATCH 3/3] drm/xe: Use the encoder suspend helper also used by the i915 driver
On Mon, Jun 17, 2024 at 01:44:06PM -0400, Rodrigo Vivi wrote:
> On Mon, Jun 17, 2024 at 08:03:56PM +0300, Imre Deak wrote:
> > Use the suspend encoders helper which is also used by the i915 driver.
> > This fixes an issue in the xe driver where the encoder
> > suspend_complete() hook is not called and was an overlook when this hook
> > was added in
>
> I believe it is okay to stay in the same line and then break the commit
> subject in the middle for the next line...
Ok, looks like done elsewhere as well, will reformat.
> > Commit b61fad5f7e5d ("drm/i915/tc: Call TypeC port flush_work/cleanup
> > without modeset locks held")
> >
>
> I wonder if we should have a fixes, but probably okay to skip it.
I wasn't sure, since strictly the tree itself at the above commit didn't
have the issue. I suppose if needed this would need to be sent to 6.9
stable manually.
> Reviewed-by: Rodrigo Vivi
Thanks.
>
> > for i915, but not added to the xe driver (which was still in a separate
> > source tree at the above point).
> >
> > Signed-off-by: Imre Deak
> > ---
> > drivers/gpu/drm/xe/display/xe_display.c | 18 ++
> > 1 file changed, 2 insertions(+), 16 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/xe/display/xe_display.c
> > b/drivers/gpu/drm/xe/display/xe_display.c
> > index 78cccbe289475..8b83dcff72e17 100644
> > --- a/drivers/gpu/drm/xe/display/xe_display.c
> > +++ b/drivers/gpu/drm/xe/display/xe_display.c
> > @@ -23,6 +23,7 @@
> > #include "intel_display_types.h"
> > #include "intel_dmc.h"
> > #include "intel_dp.h"
> > +#include "intel_encoder.h"
> > #include "intel_fbdev.h"
> > #include "intel_hdcp.h"
> > #include "intel_hotplug.h"
> > @@ -270,21 +271,6 @@ void xe_display_irq_postinstall(struct xe_device *xe,
> > struct xe_gt *gt)
> > gen11_de_irq_postinstall(xe);
> > }
> >
> > -static void intel_suspend_encoders(struct xe_device *xe)
> > -{
> > - struct drm_device *dev = >drm;
> > - struct intel_encoder *encoder;
> > -
> > - if (has_display(xe))
> > - return;
> > -
> > - drm_modeset_lock_all(dev);
> > - for_each_intel_encoder(dev, encoder)
> > - if (encoder->suspend)
> > - encoder->suspend(encoder);
> > - drm_modeset_unlock_all(dev);
> > -}
> > -
> > static bool suspend_to_idle(void)
> > {
> > #if IS_ENABLED(CONFIG_ACPI_SLEEP)
> > @@ -315,7 +301,7 @@ void xe_display_pm_suspend(struct xe_device *xe, bool
> > runtime)
> >
> > intel_hpd_cancel_work(xe);
> >
> > - intel_suspend_encoders(xe);
> > + intel_encoder_suspend_all(>display);
> >
> > intel_opregion_suspend(xe, s2idle ? PCI_D1 : PCI_D3cold);
> >
> > --
> > 2.43.3
> >
Re: [PATCH 3/3] drm/xe: Use the encoder suspend helper also used by the i915 driver
On Mon, Jun 17, 2024 at 08:03:56PM +0300, Imre Deak wrote:
> Use the suspend encoders helper which is also used by the i915 driver.
> This fixes an issue in the xe driver where the encoder
> suspend_complete() hook is not called and was an overlook when this hook
> was added in
I believe it is okay to stay in the same line and then break the commit
subject in the middle for the next line...
>
> Commit b61fad5f7e5d ("drm/i915/tc: Call TypeC port flush_work/cleanup without
> modeset locks held")
>
I wonder if we should have a fixes, but probably okay to skip it.
Reviewed-by: Rodrigo Vivi
> for i915, but not added to the xe driver (which was still in a separate
> source tree at the above point).
>
> Signed-off-by: Imre Deak
> ---
> drivers/gpu/drm/xe/display/xe_display.c | 18 ++
> 1 file changed, 2 insertions(+), 16 deletions(-)
>
> diff --git a/drivers/gpu/drm/xe/display/xe_display.c
> b/drivers/gpu/drm/xe/display/xe_display.c
> index 78cccbe289475..8b83dcff72e17 100644
> --- a/drivers/gpu/drm/xe/display/xe_display.c
> +++ b/drivers/gpu/drm/xe/display/xe_display.c
> @@ -23,6 +23,7 @@
> #include "intel_display_types.h"
> #include "intel_dmc.h"
> #include "intel_dp.h"
> +#include "intel_encoder.h"
> #include "intel_fbdev.h"
> #include "intel_hdcp.h"
> #include "intel_hotplug.h"
> @@ -270,21 +271,6 @@ void xe_display_irq_postinstall(struct xe_device *xe,
> struct xe_gt *gt)
> gen11_de_irq_postinstall(xe);
> }
>
> -static void intel_suspend_encoders(struct xe_device *xe)
> -{
> - struct drm_device *dev = >drm;
> - struct intel_encoder *encoder;
> -
> - if (has_display(xe))
> - return;
> -
> - drm_modeset_lock_all(dev);
> - for_each_intel_encoder(dev, encoder)
> - if (encoder->suspend)
> - encoder->suspend(encoder);
> - drm_modeset_unlock_all(dev);
> -}
> -
> static bool suspend_to_idle(void)
> {
> #if IS_ENABLED(CONFIG_ACPI_SLEEP)
> @@ -315,7 +301,7 @@ void xe_display_pm_suspend(struct xe_device *xe, bool
> runtime)
>
> intel_hpd_cancel_work(xe);
>
> - intel_suspend_encoders(xe);
> + intel_encoder_suspend_all(>display);
>
> intel_opregion_suspend(xe, s2idle ? PCI_D1 : PCI_D3cold);
>
> --
> 2.43.3
>
