| Message ID | 1423133245-23953-4-git-send-email-nicholas.hoath@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, Feb 05, 2015 at 10:47:18AM +0000, Nick Hoath wrote: > From: "Hoath, Nicholas" <nicholas.hoath@intel.com> > > Add: > WaDisablePartialInstShootdown Just an editor note: that's not really additional information compared to the subject of the patch. Also subject message could be a bit more direct and mention SKL: drm/i915/skl: Implement WaDisablePartialInstShootdown > v1: Dont add WaDisableThreadStallDopClockGating as not SKL WA. (Found by Damien Lespiau) v2, v1 was the original patch. Commit messages should also be wrapped like emails, so 90 chars is too long. Reviewed-by: Damien Lespiau <damien.lespiau@intel.com> > Signed-off-by: Nick Hoath <nicholas.hoath@intel.com> > --- > drivers/gpu/drm/i915/intel_ringbuffer.c | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c > index fbfe65a..b869f1c 100644 > --- a/drivers/gpu/drm/i915/intel_ringbuffer.c > +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c > @@ -875,6 +875,13 @@ static int chv_init_workarounds(struct intel_engine_cs *ring) > > static int gen9_init_workarounds(struct intel_engine_cs *ring) > { > + struct drm_device *dev = ring->dev; > + struct drm_i915_private *dev_priv = dev->dev_private; > + > + /* WaDisablePartialInstShootdown:skl */ > + WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, > + PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE); > + > return 0; > } > > -- > 2.1.1 >
On Thu, Feb 05, 2015 at 05:51:46PM +0000, Damien Lespiau wrote: > On Thu, Feb 05, 2015 at 10:47:18AM +0000, Nick Hoath wrote: > > From: "Hoath, Nicholas" <nicholas.hoath@intel.com> > > > > Add: > > WaDisablePartialInstShootdown > > Just an editor note: that's not really additional information compared > to the subject of the patch. Also subject message could be a bit more > direct and mention SKL: > > drm/i915/skl: Implement WaDisablePartialInstShootdown Well it's gen9 but yeah. The commit message body should explain the commit (e.g. more details on impact), but for w/a that's only really required if there's been an outside report. > > v1: Dont add WaDisableThreadStallDopClockGating as not SKL WA. (Found by Damien Lespiau) > > v2, v1 was the original patch. Commit messages should also be wrapped > like emails, so 90 chars is too long. > > Reviewed-by: Damien Lespiau <damien.lespiau@intel.com> Frobbed commit while applying. -Daniel > > > Signed-off-by: Nick Hoath <nicholas.hoath@intel.com> > > --- > > drivers/gpu/drm/i915/intel_ringbuffer.c | 7 +++++++ > > 1 file changed, 7 insertions(+) > > > > diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c > > index fbfe65a..b869f1c 100644 > > --- a/drivers/gpu/drm/i915/intel_ringbuffer.c > > +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c > > @@ -875,6 +875,13 @@ static int chv_init_workarounds(struct intel_engine_cs *ring) > > > > static int gen9_init_workarounds(struct intel_engine_cs *ring) > > { > > + struct drm_device *dev = ring->dev; > > + struct drm_i915_private *dev_priv = dev->dev_private; > > + > > + /* WaDisablePartialInstShootdown:skl */ > > + WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, > > + PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE); > > + > > return 0; > > } > > > > -- > > 2.1.1 > > > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@lists.freedesktop.org > http://lists.freedesktop.org/mailman/listinfo/intel-gfx
On 06/02/2015 08:52, Daniel Vetter wrote: > On Thu, Feb 05, 2015 at 05:51:46PM +0000, Damien Lespiau wrote: >> On Thu, Feb 05, 2015 at 10:47:18AM +0000, Nick Hoath wrote: >>> From: "Hoath, Nicholas" <nicholas.hoath@intel.com> >>> >>> Add: >>> WaDisablePartialInstShootdown >> >> Just an editor note: that's not really additional information compared >> to the subject of the patch. Also subject message could be a bit more >> direct and mention SKL: >> >> drm/i915/skl: Implement WaDisablePartialInstShootdown > > Well it's gen9 but yeah. The commit message body should explain the > commit (e.g. more details on impact), but for w/a that's only really > required if there's been an outside report. So an empty body is ok if the subject has sufficient information? > >>> v1: Dont add WaDisableThreadStallDopClockGating as not SKL WA. (Found by Damien Lespiau) >> >> v2, v1 was the original patch. Commit messages should also be wrapped Ah, I thought we were software engineers so started counting at 0 :) >> like emails, so 90 chars is too long. >> >> Reviewed-by: Damien Lespiau <damien.lespiau@intel.com> > > Frobbed commit while applying. > -Daniel >> >>> Signed-off-by: Nick Hoath <nicholas.hoath@intel.com> >>> --- >>> drivers/gpu/drm/i915/intel_ringbuffer.c | 7 +++++++ >>> 1 file changed, 7 insertions(+) >>> >>> diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c >>> index fbfe65a..b869f1c 100644 >>> --- a/drivers/gpu/drm/i915/intel_ringbuffer.c >>> +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c >>> @@ -875,6 +875,13 @@ static int chv_init_workarounds(struct intel_engine_cs *ring) >>> >>> static int gen9_init_workarounds(struct intel_engine_cs *ring) >>> { >>> + struct drm_device *dev = ring->dev; >>> + struct drm_i915_private *dev_priv = dev->dev_private; >>> + >>> + /* WaDisablePartialInstShootdown:skl */ >>> + WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, >>> + PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE); >>> + >>> return 0; >>> } >>> >>> -- >>> 2.1.1 >>> >> _______________________________________________ >> Intel-gfx mailing list >> Intel-gfx@lists.freedesktop.org >> http://lists.freedesktop.org/mailman/listinfo/intel-gfx >
On Fri, Feb 06, 2015 at 09:36:08AM +0000, Nick Hoath wrote: > On 06/02/2015 08:52, Daniel Vetter wrote: > >On Thu, Feb 05, 2015 at 05:51:46PM +0000, Damien Lespiau wrote: > >>On Thu, Feb 05, 2015 at 10:47:18AM +0000, Nick Hoath wrote: > >>>From: "Hoath, Nicholas" <nicholas.hoath@intel.com> > >>> > >>>Add: > >>>WaDisablePartialInstShootdown > >> > >>Just an editor note: that's not really additional information compared > >>to the subject of the patch. Also subject message could be a bit more > >>direct and mention SKL: > >> > >> drm/i915/skl: Implement WaDisablePartialInstShootdown > > > >Well it's gen9 but yeah. The commit message body should explain the > >commit (e.g. more details on impact), but for w/a that's only really > >required if there's been an outside report. > So an empty body is ok if the subject has sufficient information? Yeah for hw wa there's often not more to say. Except when some details about the impact are known which are relevant (e.g. existing bug report or not wa for a feature not yet used in upstream). Generally an empty commit message is a bit thin, but just repeating the summar doesn't add value. -Daniel
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.c b/drivers/gpu/drm/i915/intel_ringbuffer.c index fbfe65a..b869f1c 100644 --- a/drivers/gpu/drm/i915/intel_ringbuffer.c +++ b/drivers/gpu/drm/i915/intel_ringbuffer.c @@ -875,6 +875,13 @@ static int chv_init_workarounds(struct intel_engine_cs *ring) static int gen9_init_workarounds(struct intel_engine_cs *ring) { + struct drm_device *dev = ring->dev; + struct drm_i915_private *dev_priv = dev->dev_private; + + /* WaDisablePartialInstShootdown:skl */ + WA_SET_BIT_MASKED(GEN8_ROW_CHICKEN, + PARTIAL_INSTRUCTION_SHOOTDOWN_DISABLE); + return 0; }