Since obj->frontbuffer is no longer protected by the struct_mutex, as we are processing the execbuf, it may be removed. Mark the intel_frontbuffer as rcu protected, and so acquire a reference to the struct as we track activity upon it.
Closes: https://gitlab.freedesktop.org/drm/intel/issues/827 Fixes: 8e7cb1799b4f ("drm/i915: Extract intel_frontbuffer active tracking") Signed-off-by: Chris Wilson chris@chris-wilson.co.uk Cc: Matthew Auld matthew.auld@intel.com Cc: stable@vger.kernel.org # v5.4+ --- drivers/gpu/drm/i915/display/intel_display.c | 2 +- .../gpu/drm/i915/display/intel_frontbuffer.c | 16 ++++----- .../gpu/drm/i915/display/intel_frontbuffer.h | 34 +++++++++++++++++-- drivers/gpu/drm/i915/display/intel_overlay.c | 17 +++++++--- drivers/gpu/drm/i915/gem/i915_gem_clflush.c | 3 +- drivers/gpu/drm/i915/gem/i915_gem_domain.c | 4 +-- drivers/gpu/drm/i915/gem/i915_gem_object.c | 26 +++++++++++++- drivers/gpu/drm/i915/gem/i915_gem_object.h | 23 ++++++++++++- .../gpu/drm/i915/gem/i915_gem_object_types.h | 2 +- drivers/gpu/drm/i915/i915_gem.c | 10 +++--- drivers/gpu/drm/i915/i915_vma.c | 10 ++++-- 11 files changed, 116 insertions(+), 31 deletions(-)
diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c index 64e4bfb0dfc9..e18ee1f17d6e 100644 --- a/drivers/gpu/drm/i915/display/intel_display.c +++ b/drivers/gpu/drm/i915/display/intel_display.c @@ -15186,7 +15186,7 @@ intel_prepare_plane_fb(struct drm_plane *plane, return ret;
fb_obj_bump_render_priority(obj); - intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_DIRTYFB); + i915_gem_object_flush_frontbuffer(obj, ORIGIN_DIRTYFB);
if (!new_plane_state->uapi.fence) { /* implicit fencing */ struct dma_fence *fence; diff --git a/drivers/gpu/drm/i915/display/intel_frontbuffer.c b/drivers/gpu/drm/i915/display/intel_frontbuffer.c index 84b164f31895..6cb02c912acc 100644 --- a/drivers/gpu/drm/i915/display/intel_frontbuffer.c +++ b/drivers/gpu/drm/i915/display/intel_frontbuffer.c @@ -229,11 +229,11 @@ static void frontbuffer_release(struct kref *ref) vma->display_alignment = I915_GTT_MIN_ALIGNMENT; spin_unlock(&obj->vma.lock);
- obj->frontbuffer = NULL; + RCU_INIT_POINTER(obj->frontbuffer, NULL); spin_unlock(&to_i915(obj->base.dev)->fb_tracking.lock);
i915_gem_object_put(obj); - kfree(front); + kfree_rcu(front, rcu); }
struct intel_frontbuffer * @@ -242,11 +242,7 @@ intel_frontbuffer_get(struct drm_i915_gem_object *obj) struct drm_i915_private *i915 = to_i915(obj->base.dev); struct intel_frontbuffer *front;
- spin_lock(&i915->fb_tracking.lock); - front = obj->frontbuffer; - if (front) - kref_get(&front->ref); - spin_unlock(&i915->fb_tracking.lock); + front = __intel_frontbuffer_get(obj); if (front) return front;
@@ -262,13 +258,13 @@ intel_frontbuffer_get(struct drm_i915_gem_object *obj) i915_active_may_sleep(frontbuffer_retire));
spin_lock(&i915->fb_tracking.lock); - if (obj->frontbuffer) { + if (rcu_access_pointer(obj->frontbuffer)) { kfree(front); - front = obj->frontbuffer; + front = rcu_dereference_protected(obj->frontbuffer, true); kref_get(&front->ref); } else { i915_gem_object_get(obj); - obj->frontbuffer = front; + rcu_assign_pointer(obj->frontbuffer, front); } spin_unlock(&i915->fb_tracking.lock);
diff --git a/drivers/gpu/drm/i915/display/intel_frontbuffer.h b/drivers/gpu/drm/i915/display/intel_frontbuffer.h index adc64d61a4a5..ae4a1fff9f41 100644 --- a/drivers/gpu/drm/i915/display/intel_frontbuffer.h +++ b/drivers/gpu/drm/i915/display/intel_frontbuffer.h @@ -27,10 +27,10 @@ #include <linux/atomic.h> #include <linux/kref.h>
+#include "gem/i915_gem_object_types.h" #include "i915_active.h"
struct drm_i915_private; -struct drm_i915_gem_object;
enum fb_op_origin { ORIGIN_GTT, @@ -45,6 +45,7 @@ struct intel_frontbuffer { atomic_t bits; struct i915_active write; struct drm_i915_gem_object *obj; + struct rcu_head rcu; };
void intel_frontbuffer_flip_prepare(struct drm_i915_private *i915, @@ -54,6 +55,35 @@ void intel_frontbuffer_flip_complete(struct drm_i915_private *i915, void intel_frontbuffer_flip(struct drm_i915_private *i915, unsigned frontbuffer_bits);
+void intel_frontbuffer_put(struct intel_frontbuffer *front); + +static inline struct intel_frontbuffer * +__intel_frontbuffer_get(const struct drm_i915_gem_object *obj) +{ + struct intel_frontbuffer *front; + + if (likely(!rcu_access_pointer(obj->frontbuffer))) + return NULL; + + rcu_read_lock(); + do { + front = rcu_dereference(obj->frontbuffer); + if (!front) + break; + + if (!kref_get_unless_zero(&front->ref)) + continue; + + if (front == rcu_access_pointer(obj->frontbuffer)) + break; + + intel_frontbuffer_put(front); + } while (1); + rcu_read_unlock(); + + return front; +} + struct intel_frontbuffer * intel_frontbuffer_get(struct drm_i915_gem_object *obj);
@@ -119,6 +149,4 @@ void intel_frontbuffer_track(struct intel_frontbuffer *old, struct intel_frontbuffer *new, unsigned int frontbuffer_bits);
-void intel_frontbuffer_put(struct intel_frontbuffer *front); - #endif /* __INTEL_FRONTBUFFER_H__ */ diff --git a/drivers/gpu/drm/i915/display/intel_overlay.c b/drivers/gpu/drm/i915/display/intel_overlay.c index 2a44b3be2600..6097594468a9 100644 --- a/drivers/gpu/drm/i915/display/intel_overlay.c +++ b/drivers/gpu/drm/i915/display/intel_overlay.c @@ -279,12 +279,21 @@ static void intel_overlay_flip_prepare(struct intel_overlay *overlay, struct i915_vma *vma) { enum pipe pipe = overlay->crtc->pipe; + struct intel_frontbuffer *from, *to;
WARN_ON(overlay->old_vma);
- intel_frontbuffer_track(overlay->vma ? overlay->vma->obj->frontbuffer : NULL, - vma ? vma->obj->frontbuffer : NULL, - INTEL_FRONTBUFFER_OVERLAY(pipe)); + if (overlay->vma) + from = intel_frontbuffer_get(overlay->vma->obj); + if (vma) + to = intel_frontbuffer_get(vma->obj); + + intel_frontbuffer_track(from, to, INTEL_FRONTBUFFER_OVERLAY(pipe)); + + if (to) + intel_frontbuffer_put(to); + if (from) + intel_frontbuffer_put(from);
intel_frontbuffer_flip_prepare(overlay->i915, INTEL_FRONTBUFFER_OVERLAY(pipe)); @@ -764,7 +773,7 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay, ret = PTR_ERR(vma); goto out_pin_section; } - intel_frontbuffer_flush(new_bo->frontbuffer, ORIGIN_DIRTYFB); + i915_gem_object_flush_frontbuffer(new_bo, ORIGIN_DIRTYFB);
if (!overlay->active) { u32 oconfig; diff --git a/drivers/gpu/drm/i915/gem/i915_gem_clflush.c b/drivers/gpu/drm/i915/gem/i915_gem_clflush.c index 5448efa77710..34be4c0ee7c5 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_clflush.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_clflush.c @@ -20,7 +20,8 @@ static void __do_clflush(struct drm_i915_gem_object *obj) { GEM_BUG_ON(!i915_gem_object_has_pages(obj)); drm_clflush_sg(obj->mm.pages); - intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU); + + i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU); }
static int clflush_work(struct dma_fence_work *base) diff --git a/drivers/gpu/drm/i915/gem/i915_gem_domain.c b/drivers/gpu/drm/i915/gem/i915_gem_domain.c index 65f1851e2863..0cc40e77bbd2 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_domain.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_domain.c @@ -558,7 +558,7 @@ i915_gem_set_domain_ioctl(struct drm_device *dev, void *data, i915_gem_object_unlock(obj);
if (write_domain) - intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU);
out_unpin: i915_gem_object_unpin_pages(obj); @@ -678,7 +678,7 @@ int i915_gem_object_prepare_write(struct drm_i915_gem_object *obj, }
out: - intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU); obj->mm.dirty = true; /* return with the pages pinned */ return 0; diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.c b/drivers/gpu/drm/i915/gem/i915_gem_object.c index 16d611db9ca6..ddc82a7a34ff 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.c +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.c @@ -313,7 +313,7 @@ i915_gem_object_flush_write_domain(struct drm_i915_gem_object *obj, } spin_unlock(&obj->vma.lock);
- intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU); break;
case I915_GEM_DOMAIN_WC: @@ -333,6 +333,30 @@ i915_gem_object_flush_write_domain(struct drm_i915_gem_object *obj, obj->write_domain = 0; }
+void __i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj, + enum fb_op_origin origin) +{ + struct intel_frontbuffer *front; + + front = __intel_frontbuffer_get(obj); + if (front) { + intel_frontbuffer_flush(front, origin); + intel_frontbuffer_put(front); + } +} + +void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj, + enum fb_op_origin origin) +{ + struct intel_frontbuffer *front; + + front = __intel_frontbuffer_get(obj); + if (front) { + intel_frontbuffer_invalidate(front, origin); + intel_frontbuffer_put(front); + } +} + void i915_gem_init__objects(struct drm_i915_private *i915) { INIT_WORK(&i915->mm.free_work, __i915_gem_free_work); diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h b/drivers/gpu/drm/i915/gem/i915_gem_object.h index a1eb7c0b23ac..858f8bf49a04 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h @@ -13,8 +13,8 @@
#include <drm/i915_drm.h>
+#include "display/intel_frontbuffer.h" #include "i915_gem_object_types.h" - #include "i915_gem_gtt.h"
void i915_gem_init__objects(struct drm_i915_private *i915); @@ -471,4 +471,25 @@ int i915_gem_object_wait_priority(struct drm_i915_gem_object *obj, unsigned int flags, const struct i915_sched_attr *attr);
+void __i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj, + enum fb_op_origin origin); +void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj, + enum fb_op_origin origin); + +static inline void +i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj, + enum fb_op_origin origin) +{ + if (unlikely(rcu_access_pointer(obj->frontbuffer))) + __i915_gem_object_flush_frontbuffer(obj, origin); +} + +static inline void +i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj, + enum fb_op_origin origin) +{ + if (unlikely(rcu_access_pointer(obj->frontbuffer))) + __i915_gem_object_invalidate_frontbuffer(obj, origin); +} + #endif diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h index 2d404e6f63df..88e268633fdc 100644 --- a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h +++ b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h @@ -173,7 +173,7 @@ struct drm_i915_gem_object { */ u16 write_domain;
- struct intel_frontbuffer *frontbuffer; + struct intel_frontbuffer __rcu *frontbuffer;
/** Current tiling stride for the object, if it's tiled. */ unsigned int tiling_and_stride; diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 5eeef1ef7448..f19c678ebefc 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -200,7 +200,7 @@ i915_gem_phys_pwrite(struct drm_i915_gem_object *obj, * We manually control the domain here and pretend that it * remains coherent i.e. in the GTT domain, like shmem_pwrite. */ - intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU);
if (copy_from_user(vaddr, user_data, args->size)) return -EFAULT; @@ -208,7 +208,7 @@ i915_gem_phys_pwrite(struct drm_i915_gem_object *obj, drm_clflush_virt_range(vaddr, args->size); intel_gt_chipset_flush(&to_i915(obj->base.dev)->gt);
- intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU); return 0; }
@@ -628,7 +628,7 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj, goto out_unpin; }
- intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_invalidate_frontbuffer(obj, ORIGIN_CPU);
user_data = u64_to_user_ptr(args->data_ptr); offset = args->offset; @@ -672,7 +672,7 @@ i915_gem_gtt_pwrite_fast(struct drm_i915_gem_object *obj, }
intel_gt_flush_ggtt_writes(ggtt->vm.gt); - intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU);
i915_gem_object_unlock_fence(obj, fence); out_unpin: @@ -761,7 +761,7 @@ i915_gem_shmem_pwrite(struct drm_i915_gem_object *obj, offset = 0; }
- intel_frontbuffer_flush(obj->frontbuffer, ORIGIN_CPU); + i915_gem_object_flush_frontbuffer(obj, ORIGIN_CPU); i915_gem_object_unlock_fence(obj, fence);
return ret; diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c index 878975b37a45..8df0bf85f800 100644 --- a/drivers/gpu/drm/i915/i915_vma.c +++ b/drivers/gpu/drm/i915/i915_vma.c @@ -1148,8 +1148,14 @@ int i915_vma_move_to_active(struct i915_vma *vma, return err;
if (flags & EXEC_OBJECT_WRITE) { - if (intel_frontbuffer_invalidate(obj->frontbuffer, ORIGIN_CS)) - i915_active_add_request(&obj->frontbuffer->write, rq); + struct intel_frontbuffer *front; + + front = __intel_frontbuffer_get(obj); + if (unlikely(front)) { + if (intel_frontbuffer_invalidate(front, ORIGIN_CS)) + i915_active_add_request(&front->write, rq); + intel_frontbuffer_put(front); + }
dma_resv_add_excl_fence(vma->resv, &rq->fence); obj->write_domain = I915_GEM_DOMAIN_RENDER;
Quoting Chris Wilson (2019-12-16 18:17:17)
Since obj->frontbuffer is no longer protected by the struct_mutex, as we are processing the execbuf, it may be removed. Mark the intel_frontbuffer as rcu protected, and so acquire a reference to the struct as we track activity upon it.
Closes: https://gitlab.freedesktop.org/drm/intel/issues/827 Fixes: 8e7cb1799b4f ("drm/i915: Extract intel_frontbuffer active tracking") Signed-off-by: Chris Wilson chris@chris-wilson.co.uk Cc: Matthew Auld matthew.auld@intel.com Cc: stable@vger.kernel.org # v5.4+
<SNIP>
@@ -54,6 +55,35 @@ void intel_frontbuffer_flip_complete(struct drm_i915_private *i915, void intel_frontbuffer_flip(struct drm_i915_private *i915, unsigned frontbuffer_bits); +void intel_frontbuffer_put(struct intel_frontbuffer *front);
+static inline struct intel_frontbuffer * +__intel_frontbuffer_get(const struct drm_i915_gem_object *obj) +{
struct intel_frontbuffer *front;
if (likely(!rcu_access_pointer(obj->frontbuffer)))
return NULL;
rcu_read_lock();
do {
front = rcu_dereference(obj->frontbuffer);
if (!front)
break;
if (!kref_get_unless_zero(&front->ref))
continue;
if (front == rcu_access_pointer(obj->frontbuffer))
break;
intel_frontbuffer_put(front);
} while (1);
rcu_read_unlock();
return front;
+}
Understood that there can only be so many writers, so it's not technically blocking, but the constructs looks like it was designed to spin.
It would look more appropriate without the loop:
if (unlikely(!kref_get_unless_zero(&front->ref))) goto retry;
<SNIP>
+void __i915_gem_object_flush_frontbuffer(struct drm_i915_gem_object *obj,
enum fb_op_origin origin)
+{
struct intel_frontbuffer *front;
front = __intel_frontbuffer_get(obj);
if (front) {
intel_frontbuffer_flush(front, origin);
intel_frontbuffer_put(front);
}
+}
+void __i915_gem_object_invalidate_frontbuffer(struct drm_i915_gem_object *obj,
enum fb_op_origin origin)
+{
struct intel_frontbuffer *front;
front = __intel_frontbuffer_get(obj);
if (front) {
intel_frontbuffer_invalidate(front, origin);
intel_frontbuffer_put(front);
}
+}
Could be de-duped as by taking the vfunc as parameter, but that's just by coincidence that parameters match.
The rest checks out, so with the loop removed:
Reviewed-by: Joonas Lahtinen joonas.lahtinen@linux.intel.com
Regards, Joonas
linux-stable-mirror@lists.linaro.org