On Tue, Jan 16, 2024 at 09:29:19AM -0800, Juan Escamilla wrote: > The sysfs file is named 'enabled', thus users might want to know the > true state of the RC6 instead of only the indication if the RC6 > should be enabled. > > Let's use rc6.enable directly instead of rc6.supported. For a moment I got concerned about this case here: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_128839v2/shard-tglu-4/igt@i915_pm_rc6_residency@rc6-idle@xxxxxxxxxxxxx but then I noticed it pass the test requirement for RC6: (i915_pm_rc6_residency:5167) DEBUG: Test requirement passed: __pmu_wait_for_rc6(fd) Which also anyway is using the residency directly. Then, I'm also happy with the tests that really uses this like the accuracy and rc6_disable. So, Reviewed-by: Rodrigo Vivi <rodrigo.vivi@xxxxxxxxx> > > Signed-off-by: Juan Escamilla <jcescami@xxxxxxxx> > --- > drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c b/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c > index 2d3c4dab6d21..c0b202223940 100644 > --- a/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c > +++ b/drivers/gpu/drm/i915/gt/intel_gt_sysfs_pm.c > @@ -182,7 +182,7 @@ static ssize_t rc6_enable_show(struct kobject *kobj, > { > struct intel_gt *gt = intel_gt_sysfs_get_drvdata(kobj, attr->attr.name); > > - return sysfs_emit(buff, "%x\n", gt->rc6.supported); > + return sysfs_emit(buff, "%x\n", gt->rc6.enabled); > } > > static ssize_t rc6_enable_dev_show(struct device *dev, > @@ -191,7 +191,7 @@ static ssize_t rc6_enable_dev_show(struct device *dev, > { > struct intel_gt *gt = intel_gt_sysfs_get_drvdata(&dev->kobj, attr->attr.name); > > - return sysfs_emit(buff, "%x\n", gt->rc6.supported); > + return sysfs_emit(buff, "%x\n", gt->rc6.enabled); > } > > static u32 __rc6_residency_ms_show(struct intel_gt *gt) > -- > 2.43.0 >
