| Message ID | 20201007120329.17076-1-ville.syrjala@linux.intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [1/3] drm/i915: Mark ininitial fb obj as WT on eLLC machines to avoid rcu lockup during fbdev init | expand |
See subject, s/ininitial/iniital/ Quoting Ville Syrjala (2020-10-07 13:03:27) > From: Ville Syrjälä <ville.syrjala@linux.intel.com> > > Currently we leave the cache_level of the initial fb obj > set to NONE. This means on eLLC machines the first pin_to_display() > will try to switch it to WT which requires a vma unbind+bind. > If that happens during the fbdev initialization rcu does not > seem operational which causes the unbind to get stuck. To > most appearances this looks like a dead machine on boot. > > Avoid the unbind by already marking the object cache_level > as WT when creating it. We still do an excplicit ggtt pin > which will rewrite the PTEs anyway, so they will match whatever > cache level we set. > > Cc: <stable@vger.kernel.org> # v5.7+ > Suggested-by: Chris Wilson <chris@chris-wilson.co.uk> > Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2381 > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> > --- > drivers/gpu/drm/i915/display/intel_display.c | 8 ++++++++ > 1 file changed, 8 insertions(+) > > diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c > index 907e1d155443..00c08600c60a 100644 > --- a/drivers/gpu/drm/i915/display/intel_display.c > +++ b/drivers/gpu/drm/i915/display/intel_display.c > @@ -3445,6 +3445,14 @@ initial_plane_vma(struct drm_i915_private *i915, > if (IS_ERR(obj)) > return NULL; > > + /* > + * Mark it WT ahead of time to avoid changing the > + * cache_level during fbdev initialization. The > + * unbind there would get stuck waiting for rcu. > + */ > + i915_gem_object_set_cache_coherency(obj, HAS_WT(i915) ? > + I915_CACHE_WT : I915_CACHE_NONE); Ok, I've been worrying about whether there were any more side-effects, but I think it all comes out in the wash. The proof is definitely in the eating, and we will know soon enough if we break someone's virtual terminal. Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> -Chris
On Tue, Oct 13, 2020 at 04:47:49PM +0100, Chris Wilson wrote: > See subject, s/ininitial/iniital/ > > Quoting Ville Syrjala (2020-10-07 13:03:27) > > From: Ville Syrjälä <ville.syrjala@linux.intel.com> > > > > Currently we leave the cache_level of the initial fb obj > > set to NONE. This means on eLLC machines the first pin_to_display() > > will try to switch it to WT which requires a vma unbind+bind. > > If that happens during the fbdev initialization rcu does not > > seem operational which causes the unbind to get stuck. To > > most appearances this looks like a dead machine on boot. > > > > Avoid the unbind by already marking the object cache_level > > as WT when creating it. We still do an excplicit ggtt pin > > which will rewrite the PTEs anyway, so they will match whatever > > cache level we set. > > > > Cc: <stable@vger.kernel.org> # v5.7+ > > Suggested-by: Chris Wilson <chris@chris-wilson.co.uk> > > Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/2381 > > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com> > > --- > > drivers/gpu/drm/i915/display/intel_display.c | 8 ++++++++ > > 1 file changed, 8 insertions(+) > > > > diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c > > index 907e1d155443..00c08600c60a 100644 > > --- a/drivers/gpu/drm/i915/display/intel_display.c > > +++ b/drivers/gpu/drm/i915/display/intel_display.c > > @@ -3445,6 +3445,14 @@ initial_plane_vma(struct drm_i915_private *i915, > > if (IS_ERR(obj)) > > return NULL; > > > > + /* > > + * Mark it WT ahead of time to avoid changing the > > + * cache_level during fbdev initialization. The > > + * unbind there would get stuck waiting for rcu. > > + */ > > + i915_gem_object_set_cache_coherency(obj, HAS_WT(i915) ? > > + I915_CACHE_WT : I915_CACHE_NONE); > > Ok, I've been worrying about whether there were any more side-effects, > but I think it all comes out in the wash. The proof is definitely in the > eating, and we will know soon enough if we break someone's virtual > terminal. At least it seems to work on my CFL with eLLC caching enabled. > > Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Ta.
diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c index 907e1d155443..00c08600c60a 100644 --- a/drivers/gpu/drm/i915/display/intel_display.c +++ b/drivers/gpu/drm/i915/display/intel_display.c @@ -3445,6 +3445,14 @@ initial_plane_vma(struct drm_i915_private *i915, if (IS_ERR(obj)) return NULL; + /* + * Mark it WT ahead of time to avoid changing the + * cache_level during fbdev initialization. The + * unbind there would get stuck waiting for rcu. + */ + i915_gem_object_set_cache_coherency(obj, HAS_WT(i915) ? + I915_CACHE_WT : I915_CACHE_NONE); + switch (plane_config->tiling) { case I915_TILING_NONE: break;