drm/i915: Make the lock for pageflips interruptible
authorChris Wilson <chris@chris-wilson.co.uk>
Wed, 23 May 2012 10:13:58 +0000 (11:13 +0100)
committerDaniel Vetter <daniel.vetter@ffwll.ch>
Fri, 20 Jul 2012 10:21:41 +0000 (12:21 +0200)
As we take the struct_mutex lock to access the command-stream, there is
a possibility that we may need to wait for a GPU hang and so should make
the lock both interruptible and error-checking.

References: https://bugs.freedesktop.org/show_bug.cgi?id=50069
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
drivers/gpu/drm/i915/intel_display.c

index 80fe02d2d733c7326fbded035765e4422a4b5160..c9b69e3b2c29f0bc3361cf16e29d82ef25390548 100644 (file)
@@ -6486,7 +6486,9 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc,
        intel_fb = to_intel_framebuffer(fb);
        obj = intel_fb->obj;
 
-       mutex_lock(&dev->struct_mutex);
+       ret = i915_mutex_lock_interruptible(dev);
+       if (ret)
+               goto cleanup;
 
        /* Reference the objects for the scheduled work. */
        drm_gem_object_reference(&work->old_fb_obj->base);
@@ -6521,6 +6523,7 @@ cleanup_pending:
        drm_gem_object_unreference(&obj->base);
        mutex_unlock(&dev->struct_mutex);
 
+cleanup:
        spin_lock_irqsave(&dev->event_lock, flags);
        intel_crtc->unpin_work = NULL;
        spin_unlock_irqrestore(&dev->event_lock, flags);