we can fetch timeline point fence after expanded. v2: The parameter fence is the result of the function and should come last.

Signed-off-by: Chunming Zhou david1.zhou@amd.com --- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 2 +- drivers/gpu/drm/drm_syncobj.c | 5 +++-- drivers/gpu/drm/v3d/v3d_gem.c | 4 ++-- drivers/gpu/drm/vc4/vc4_gem.c | 2 +- include/drm/drm_syncobj.h | 2 +- 5 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c index 7a625f3989a0..0e8cf088175f 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c @@ -1062,7 +1062,7 @@ static int amdgpu_syncobj_lookup_and_add_to_sync(struct amdgpu_cs_parser *p, { int r; struct dma_fence *fence; - r = drm_syncobj_find_fence(p->filp, handle, &fence); + r = drm_syncobj_find_fence(p->filp, handle, 0, &fence); if (r) return r;

diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c index d4b48fb410a1..2dcb60f4c0f7 100644 --- a/drivers/gpu/drm/drm_syncobj.c +++ b/drivers/gpu/drm/drm_syncobj.c @@ -217,6 +217,7 @@ static int drm_syncobj_assign_null_handle(struct drm_syncobj *syncobj) * drm_syncobj_find_fence - lookup and reference the fence in a sync object * @file_private: drm file private pointer * @handle: sync object handle to lookup. + * @point: timeline point * @fence: out parameter for the fence * * This is just a convenience function that combines drm_syncobj_find() and @@ -227,7 +228,7 @@ static int drm_syncobj_assign_null_handle(struct drm_syncobj *syncobj) * dma_fence_put(). */ int drm_syncobj_find_fence(struct drm_file *file_private, - u32 handle, + u32 handle, u64 point, struct dma_fence **fence) { struct drm_syncobj *syncobj = drm_syncobj_find(file_private, handle); @@ -498,7 +499,7 @@ static int drm_syncobj_export_sync_file(struct drm_file *file_private, if (fd < 0) return fd;

- ret = drm_syncobj_find_fence(file_private, handle, &fence); + ret = drm_syncobj_find_fence(file_private, handle, 0, &fence); if (ret) goto err_put_fd;

diff --git a/drivers/gpu/drm/v3d/v3d_gem.c b/drivers/gpu/drm/v3d/v3d_gem.c index e1fcbb4cd0ae..d25c35c45c33 100644 --- a/drivers/gpu/drm/v3d/v3d_gem.c +++ b/drivers/gpu/drm/v3d/v3d_gem.c @@ -521,12 +521,12 @@ v3d_submit_cl_ioctl(struct drm_device *dev, void *data, kref_init(&exec->refcount);

ret = drm_syncobj_find_fence(file_priv, args->in_sync_bcl, - &exec->bin.in_fence); + 0, &exec->bin.in_fence); if (ret == -EINVAL) goto fail;

ret = drm_syncobj_find_fence(file_priv, args->in_sync_rcl, - &exec->render.in_fence); + 0, &exec->render.in_fence); if (ret == -EINVAL) goto fail;

diff --git a/drivers/gpu/drm/vc4/vc4_gem.c b/drivers/gpu/drm/vc4/vc4_gem.c index 7910b9acedd6..928718b467bd 100644 --- a/drivers/gpu/drm/vc4/vc4_gem.c +++ b/drivers/gpu/drm/vc4/vc4_gem.c @@ -1173,7 +1173,7 @@ vc4_submit_cl_ioctl(struct drm_device *dev, void *data,

if (args->in_sync) { ret = drm_syncobj_find_fence(file_priv, args->in_sync, - &in_fence); + 0, &in_fence); if (ret) goto fail;

diff --git a/include/drm/drm_syncobj.h b/include/drm/drm_syncobj.h index e419c79ba94d..ab9055f943c7 100644 --- a/include/drm/drm_syncobj.h +++ b/include/drm/drm_syncobj.h @@ -134,7 +134,7 @@ struct drm_syncobj *drm_syncobj_find(struct drm_file *file_private, void drm_syncobj_replace_fence(struct drm_syncobj *syncobj, struct dma_fence *fence); int drm_syncobj_find_fence(struct drm_file *file_private, - u32 handle, + u32 handle, u64 point, struct dma_fence **fence); void drm_syncobj_free(struct kref *kref); int drm_syncobj_create(struct drm_syncobj **out_syncobj, uint32_t flags,

VK_KHR_timeline_semaphore: This extension introduces a new type of semaphore that has an integer payload identifying a point in a timeline. Such timeline semaphores support the following operations: * CPU query - A host operation that allows querying the payload of the timeline semaphore. * CPU wait - A host operation that allows a blocking wait for a timeline semaphore to reach a specified value. * Device wait - A device operation that allows waiting for a timeline semaphore to reach a specified value. * Device signal - A device operation that allows advancing the timeline semaphore to a specified value.

Since it's a timeline, that means the front time point(PT) always is signaled before the late PT. a. signal PT design: Signal PT fence N depends on PT[N-1] fence and signal opertion fence, when PT[N] fence is signaled, the timeline will increase to value of PT[N]. b. wait PT design: Wait PT fence is signaled by reaching timeline point value, when timeline is increasing, will compare wait PTs value with new timeline value, if PT value is lower than timeline value, then wait PT will be signaled, otherwise keep in list. semaphore wait operation can wait on any point of timeline, so need a RB tree to order them. And wait PT could ahead of signal PT, we need a sumission fence to perform that.

v2: 1. remove unused DRM_SYNCOBJ_CREATE_TYPE_NORMAL. (Christian) 2. move unexposed denitions to .c file. (Daniel Vetter) 3. split up the change to drm_syncobj_find_fence() in a separate patch. (Christian) 4. split up the change to drm_syncobj_replace_fence() in a separate patch. 5. drop the submission_fence implementation and instead use wait_event() for that. (Christian) 6. WARN_ON(point != 0) for NORMAL type syncobj case. (Daniel Vetter)

TODO: 1. CPU query and wait on timeline semaphore. 2. test application (Daniel Vetter)

Signed-off-by: Chunming Zhou david1.zhou@amd.com Cc: Christian Konig christian.koenig@amd.com Cc: Dave Airlie airlied@redhat.com Cc: Daniel Rakos Daniel.Rakos@amd.com Cc: Daniel Vetter daniel@ffwll.ch --- drivers/gpu/drm/drm_syncobj.c | 384 ++++++++++++++++++++++++++++++++-- include/drm/drm_syncobj.h | 28 +++ include/uapi/drm/drm.h | 1 + 3 files changed, 390 insertions(+), 23 deletions(-)

diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c index ab43559398d0..e0d89a84359f 100644 --- a/drivers/gpu/drm/drm_syncobj.c +++ b/drivers/gpu/drm/drm_syncobj.c @@ -56,6 +56,44 @@ #include "drm_internal.h" #include <drm/drm_syncobj.h>

+struct drm_syncobj_stub_fence { + struct dma_fence base; + spinlock_t lock; +}; + +static const char *drm_syncobj_stub_fence_get_name(struct dma_fence *fence) +{ + return "syncobjstub"; +} + +static bool drm_syncobj_stub_fence_enable_signaling(struct dma_fence *fence) +{ + return !dma_fence_is_signaled(fence); +} + +static const struct dma_fence_ops drm_syncobj_stub_fence_ops = { + .get_driver_name = drm_syncobj_stub_fence_get_name, + .get_timeline_name = drm_syncobj_stub_fence_get_name, + .enable_signaling = drm_syncobj_stub_fence_enable_signaling, + .release = NULL, +}; + +struct drm_syncobj_wait_pt { + struct drm_syncobj_stub_fence base; + u64 value; + struct rb_node node; +}; +struct drm_syncobj_signal_pt { + struct drm_syncobj_stub_fence base; + struct dma_fence *signal_fence; + struct dma_fence *pre_pt_base; + struct dma_fence_cb signal_cb; + struct dma_fence_cb pre_pt_cb; + struct drm_syncobj *syncobj; + u64 value; + struct list_head list; +}; + /** * drm_syncobj_find - lookup and reference a sync object. * @file_private: drm file private pointer @@ -137,6 +175,152 @@ void drm_syncobj_remove_callback(struct drm_syncobj *syncobj, spin_unlock(&syncobj->lock); }

+static void drm_syncobj_timeline_signal_wait_pts(struct drm_syncobj *syncobj) +{ + struct rb_node *node = NULL; + struct drm_syncobj_wait_pt *wait_pt = NULL; + + spin_lock(&syncobj->lock); + for(node = rb_first(&syncobj->syncobj_timeline.wait_pt_tree); + node != NULL; ) { + wait_pt = rb_entry(node, struct drm_syncobj_wait_pt, node); + node = rb_next(node); + if (wait_pt->value <= syncobj->syncobj_timeline.timeline) { + dma_fence_signal(&wait_pt->base.base); + rb_erase(&wait_pt->node, + &syncobj->syncobj_timeline.wait_pt_tree); + RB_CLEAR_NODE(&wait_pt->node); + /* kfree(wait_pt) is excuted by fence put */ + dma_fence_put(&wait_pt->base.base); + } else { + /* the loop is from left to right, the later entry value is + * bigger, so don't need to check any more */ + break; + } + } + spin_unlock(&syncobj->lock); +} + + +static void pt_fence_cb(struct drm_syncobj_signal_pt *signal_pt) +{ + struct dma_fence *fence = NULL; + struct drm_syncobj *syncobj; + + fence = signal_pt->signal_fence; + signal_pt->signal_fence = NULL; + dma_fence_put(fence); + fence = signal_pt->pre_pt_base; + signal_pt->pre_pt_base = NULL; + dma_fence_put(fence); + + syncobj = signal_pt->syncobj; + spin_lock(&syncobj->lock); + list_del(&signal_pt->list); + syncobj->syncobj_timeline.timeline = signal_pt->value; + spin_unlock(&syncobj->lock); + /* kfree(signal_pt) will be executed by below fence put */ + dma_fence_put(&signal_pt->base.base); + drm_syncobj_timeline_signal_wait_pts(syncobj); +} +static void pt_signal_fence_func(struct dma_fence *fence, + struct dma_fence_cb *cb) +{ + struct drm_syncobj_signal_pt *signal_pt = + container_of(cb, struct drm_syncobj_signal_pt, signal_cb); + + if (signal_pt->pre_pt_base && + !dma_fence_is_signaled(signal_pt->pre_pt_base)) + return; + + pt_fence_cb(signal_pt); +} +static void pt_pre_fence_func(struct dma_fence *fence, + struct dma_fence_cb *cb) +{ + struct drm_syncobj_signal_pt *signal_pt = + container_of(cb, struct drm_syncobj_signal_pt, pre_pt_cb); + + if (signal_pt->signal_fence && + !dma_fence_is_signaled(signal_pt->pre_pt_base)) + return; + + pt_fence_cb(signal_pt); +} + +static int drm_syncobj_timeline_create_signal_pt(struct drm_syncobj *syncobj, + struct dma_fence *fence, + u64 point) +{ + struct drm_syncobj_signal_pt *signal_pt = + kzalloc(sizeof(struct drm_syncobj_signal_pt), GFP_KERNEL); + struct drm_syncobj_signal_pt *tail_pt; + struct dma_fence *tail_pt_fence = NULL; + int ret = 0; + + if (!signal_pt) + return -ENOMEM; + if (syncobj->syncobj_timeline.signal_point >= point) { + DRM_WARN("A later signal is ready!"); + goto out; + } + if (fence) + dma_fence_get(fence); + spin_lock(&syncobj->lock); + spin_lock_init(&signal_pt->base.lock); + dma_fence_init(&signal_pt->base.base, + &drm_syncobj_stub_fence_ops, + &signal_pt->base.lock, + syncobj->syncobj_timeline.timeline_context, point); + signal_pt->signal_fence = + rcu_dereference_protected(fence, + lockdep_is_held(&fence->lock)); + if (!list_empty(&syncobj->syncobj_timeline.signal_pt_list)) { + tail_pt = list_last_entry(&syncobj->syncobj_timeline.signal_pt_list, + struct drm_syncobj_signal_pt, list); + tail_pt_fence = &tail_pt->base.base; + if (dma_fence_is_signaled(tail_pt_fence)) + tail_pt_fence = NULL; + } + if (tail_pt_fence) + signal_pt->pre_pt_base = + dma_fence_get(rcu_dereference_protected(tail_pt_fence, + lockdep_is_held(&tail_pt_fence->lock))); + + signal_pt->value = point; + syncobj->syncobj_timeline.signal_point = point; + signal_pt->syncobj = syncobj; + INIT_LIST_HEAD(&signal_pt->list); + list_add_tail(&signal_pt->list, &syncobj->syncobj_timeline.signal_pt_list); + spin_unlock(&syncobj->lock); + wake_up_all(&syncobj->syncobj_timeline.wq); + /** + * Every pt is depending on signal fence and previous pt fence, add + * callbacks to them + */ + if (!dma_fence_is_signaled(signal_pt->signal_fence)) + dma_fence_add_callback(signal_pt->signal_fence, + &signal_pt->signal_cb, + pt_signal_fence_func); + else + pt_signal_fence_func(signal_pt->signal_fence, + &signal_pt->signal_cb); + if (signal_pt->pre_pt_base) { + if (!dma_fence_is_signaled(signal_pt->pre_pt_base)) + dma_fence_add_callback(signal_pt->pre_pt_base, + &signal_pt->pre_pt_cb, + pt_pre_fence_func); + else + pt_pre_fence_func(signal_pt->pre_pt_base, + &signal_pt->pre_pt_cb); + } + + return 0; +out: + kfree(signal_pt); + return ret; +} + /** * drm_syncobj_replace_fence - replace fence in a sync object. * @syncobj: Sync object to replace fence in @@ -152,6 +336,10 @@ void drm_syncobj_replace_fence(struct drm_syncobj *syncobj, struct dma_fence *old_fence; struct drm_syncobj_cb *cur, *tmp;

+ if (syncobj->type == DRM_SYNCOBJ_TYPE_TIMELINE) { + drm_syncobj_timeline_create_signal_pt(syncobj, fence, point); + return; + } if (fence) dma_fence_get(fence);

@@ -174,28 +362,6 @@ void drm_syncobj_replace_fence(struct drm_syncobj *syncobj, } EXPORT_SYMBOL(drm_syncobj_replace_fence);

-struct drm_syncobj_stub_fence { - struct dma_fence base; - spinlock_t lock; -}; - -static const char *drm_syncobj_stub_fence_get_name(struct dma_fence *fence) -{ - return "syncobjstub"; -} - -static bool drm_syncobj_stub_fence_enable_signaling(struct dma_fence *fence) -{ - return !dma_fence_is_signaled(fence); -} - -static const struct dma_fence_ops drm_syncobj_stub_fence_ops = { - .get_driver_name = drm_syncobj_stub_fence_get_name, - .get_timeline_name = drm_syncobj_stub_fence_get_name, - .enable_signaling = drm_syncobj_stub_fence_enable_signaling, - .release = NULL, -}; - static int drm_syncobj_assign_null_handle(struct drm_syncobj *syncobj) { struct drm_syncobj_stub_fence *fence; @@ -215,6 +381,121 @@ static int drm_syncobj_assign_null_handle(struct drm_syncobj *syncobj) return 0; }

+static struct drm_syncobj_wait_pt * +drm_syncobj_timeline_lookup_wait_pt(struct drm_syncobj *syncobj, u64 point) +{ + struct rb_node *node = syncobj->syncobj_timeline.wait_pt_tree.rb_node; + struct drm_syncobj_wait_pt *wait_pt = NULL; + + + spin_lock(&syncobj->lock); + while(node) { + int result = point - wait_pt->value; + + wait_pt = rb_entry(node, struct drm_syncobj_wait_pt, node); + if (result < 0) + node = node->rb_left; + else if (result > 0) + node = node->rb_right; + else + break; + } + spin_unlock(&syncobj->lock); + + return wait_pt; +} + +static struct drm_syncobj_wait_pt * +drm_syncobj_timeline_create_wait_pt(struct drm_syncobj *syncobj, u64 point) +{ + struct drm_syncobj_wait_pt *wait_pt; + struct rb_node **new = &(syncobj->syncobj_timeline.wait_pt_tree.rb_node), *parent = NULL; + + wait_pt = kzalloc(sizeof(*wait_pt), GFP_KERNEL); + if (!wait_pt) + return NULL; + spin_lock_init(&wait_pt->base.lock); + dma_fence_init(&wait_pt->base.base, + &drm_syncobj_stub_fence_ops, + &wait_pt->base.lock, + syncobj->syncobj_timeline.timeline_context, point); + wait_pt->value = point; + + /* wait pt must be in an order, so that we can easily lookup and signal + * it */ + spin_lock(&syncobj->lock); + if (point <= syncobj->syncobj_timeline.timeline) + dma_fence_signal(&wait_pt->base.base); + while(*new) { + struct drm_syncobj_wait_pt *this = + rb_entry(*new, struct drm_syncobj_wait_pt, node); + int result = wait_pt->value - this->value; + + parent = *new; + if (result < 0) + new = &((*new)->rb_left); + else if (result > 0) + new = &((*new)->rb_right); + else + goto exist; + } + + rb_link_node(&wait_pt->node, parent, new); + rb_insert_color(&wait_pt->node, &syncobj->syncobj_timeline.wait_pt_tree); + spin_unlock(&syncobj->lock); + return wait_pt; +exist: + spin_unlock(&syncobj->lock); + dma_fence_put(&wait_pt->base.base); + wait_pt = drm_syncobj_timeline_lookup_wait_pt(syncobj, point); + return wait_pt; +} + +static struct dma_fence * +drm_syncobj_timeline_point_get(struct drm_syncobj *syncobj, u64 point, u64 flag) +{ + struct drm_syncobj_wait_pt *wait_pt; + + /* already signaled, simply return a signaled stub fence */ + if (point <= syncobj->syncobj_timeline.timeline) { + struct drm_syncobj_stub_fence *fence; + + fence = kzalloc(sizeof(*fence), GFP_KERNEL); + if (fence == NULL) + return NULL; + + spin_lock_init(&fence->lock); + dma_fence_init(&fence->base, &drm_syncobj_stub_fence_ops, + &fence->lock, 0, 0); + dma_fence_signal(&fence->base); + return &fence->base; + } + + /* check if the wait pt exists */ + wait_pt = drm_syncobj_timeline_lookup_wait_pt(syncobj, point); + if (!wait_pt) { + /* This is a new wait pt, so create it */ + wait_pt = drm_syncobj_timeline_create_wait_pt(syncobj, point); + if (!wait_pt) + return NULL; + } + if (wait_pt) { + struct dma_fence *fence; + int ret = + wait_event_interruptible_timeout(syncobj->syncobj_timeline.wq, + wait_pt->value <= syncobj->syncobj_timeline.signal_point, + msecs_to_jiffies(10000)); /* wait 10s */ + + if (ret <= 0) + return NULL; + rcu_read_lock(); + fence = dma_fence_get_rcu(&wait_pt->base.base); + rcu_read_unlock(); + return fence; + } + return NULL; +} + /** * drm_syncobj_find_fence - lookup and reference the fence in a sync object * @file_private: drm file private pointer @@ -239,7 +520,17 @@ int drm_syncobj_find_fence(struct drm_file *file_private, if (!syncobj) return -ENOENT;

- *fence = drm_syncobj_fence_get(syncobj); + if (syncobj->type == DRM_SYNCOBJ_TYPE_NORMAL) { + /* NORMAL syncobj doesn't care point value */ + WARN_ON(point != 0); + *fence = drm_syncobj_fence_get(syncobj); + } else if (syncobj->type == DRM_SYNCOBJ_TYPE_TIMELINE) { + *fence = drm_syncobj_timeline_point_get(syncobj, point, + DRM_SYNCOBJ_WAIT_FLAGS_WAIT_FOR_SUBMIT); + } else { + DRM_ERROR("Don't support this type syncobj\n"); + *fence = NULL; + } if (!*fence) { ret = -EINVAL; } @@ -248,6 +539,34 @@ int drm_syncobj_find_fence(struct drm_file *file_private, } EXPORT_SYMBOL(drm_syncobj_find_fence);

+static void drm_syncobj_timeline_fini(struct drm_syncobj *syncobj, + struct drm_syncobj_timeline *syncobj_timeline) +{ + struct rb_node *node = NULL; + struct drm_syncobj_wait_pt *wait_pt = NULL; + struct drm_syncobj_signal_pt *signal_pt = NULL, *tmp; + + spin_lock(&syncobj->lock); + for(node = rb_first(&syncobj_timeline->wait_pt_tree); + node != NULL; ) { + wait_pt = rb_entry(node, struct drm_syncobj_wait_pt, node); + node = rb_next(node); + rb_erase(&wait_pt->node, + &syncobj_timeline->wait_pt_tree); + RB_CLEAR_NODE(&wait_pt->node); + /* kfree(wait_pt) is excuted by fence put */ + dma_fence_put(&wait_pt->base.base); + } + list_for_each_entry_safe(signal_pt, tmp, + &syncobj_timeline->signal_pt_list, list) { + list_del(&signal_pt->list); + dma_fence_put(signal_pt->signal_fence); + dma_fence_put(signal_pt->pre_pt_base); + dma_fence_put(&signal_pt->base.base); + } + spin_unlock(&syncobj->lock); +} + /** * drm_syncobj_free - free a sync object. * @kref: kref to free. @@ -260,10 +579,23 @@ void drm_syncobj_free(struct kref *kref) struct drm_syncobj, refcount); drm_syncobj_replace_fence(syncobj, 0, NULL); + drm_syncobj_timeline_fini(syncobj, &syncobj->syncobj_timeline); kfree(syncobj); } EXPORT_SYMBOL(drm_syncobj_free);

+static void drm_syncobj_timeline_init(struct drm_syncobj_timeline + *syncobj_timeline) +{ + syncobj_timeline->timeline_context = dma_fence_context_alloc(1); + syncobj_timeline->timeline = 0; + syncobj_timeline->signal_point = 0; + init_waitqueue_head(&syncobj_timeline->wq); + + syncobj_timeline->wait_pt_tree = RB_ROOT; + INIT_LIST_HEAD(&syncobj_timeline->signal_pt_list); +} + /** * drm_syncobj_create - create a new syncobj * @out_syncobj: returned syncobj @@ -289,6 +621,12 @@ int drm_syncobj_create(struct drm_syncobj **out_syncobj, uint32_t flags, kref_init(&syncobj->refcount); INIT_LIST_HEAD(&syncobj->cb_list); spin_lock_init(&syncobj->lock); + if (flags & DRM_SYNCOBJ_CREATE_TYPE_TIMELINE) { + syncobj->type = DRM_SYNCOBJ_TYPE_TIMELINE; + drm_syncobj_timeline_init(&syncobj->syncobj_timeline); + } else { + syncobj->type = DRM_SYNCOBJ_TYPE_NORMAL; + }

if (flags & DRM_SYNCOBJ_CREATE_SIGNALED) { ret = drm_syncobj_assign_null_handle(syncobj); diff --git a/include/drm/drm_syncobj.h b/include/drm/drm_syncobj.h index 425432b85a87..fb4e7edf90bf 100644 --- a/include/drm/drm_syncobj.h +++ b/include/drm/drm_syncobj.h @@ -30,6 +30,25 @@

struct drm_syncobj_cb;

+enum drm_syncobj_type { + DRM_SYNCOBJ_TYPE_NORMAL, + DRM_SYNCOBJ_TYPE_TIMELINE +}; + +struct drm_syncobj_timeline { + wait_queue_head_t wq; + u64 timeline_context; + /** + * @timeline: syncobj timeline + */ + u64 timeline; + u64 signal_point; + + + struct rb_root wait_pt_tree; + struct list_head signal_pt_list; +}; + /** * struct drm_syncobj - sync object. * @@ -40,6 +59,15 @@ struct drm_syncobj { * @refcount: Reference count of this object. */ struct kref refcount; + /** + * @type: indicate syncobj type + */ + enum drm_syncobj_type type; + /** + * @syncobj_timeline: timeline + */ + struct drm_syncobj_timeline syncobj_timeline; + /** * @fence: * NULL or a pointer to the fence bound to this object. diff --git a/include/uapi/drm/drm.h b/include/uapi/drm/drm.h index 300f336633f2..cebdb2541eb7 100644 --- a/include/uapi/drm/drm.h +++ b/include/uapi/drm/drm.h @@ -717,6 +717,7 @@ struct drm_prime_handle { struct drm_syncobj_create { __u32 handle; #define DRM_SYNCOBJ_CREATE_SIGNALED (1 << 0) +#define DRM_SYNCOBJ_CREATE_TYPE_TIMELINE (1 << 1) __u32 flags; };

Signed-off-by: Chunming Zhou david1.zhou@amd.com --- drivers/gpu/drm/drm_syncobj.c | 5 +++-- drivers/gpu/drm/i915/i915_gem_execbuffer.c | 4 +++- include/drm/drm_syncobj.h | 2 ++ 3 files changed, 8 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c index ff0b1fdfd9a8..0be7dd0a0e7d 100644 --- a/drivers/gpu/drm/drm_syncobj.c +++ b/drivers/gpu/drm/drm_syncobj.c @@ -500,8 +500,8 @@ drm_syncobj_timeline_point_get(struct drm_syncobj *syncobj, u64 point, u64 flags * contains a reference to the fence, which must be released by calling * dma_fence_put(). */ -static int drm_syncobj_search_fence(struct drm_syncobj *syncobj, u64 point, - u64 flags, struct dma_fence **fence) +int drm_syncobj_search_fence(struct drm_syncobj *syncobj, u64 point, + u64 flags, struct dma_fence **fence) { int ret = 0;

@@ -530,6 +530,7 @@ static int drm_syncobj_search_fence(struct drm_syncobj *syncobj, u64 point, drm_syncobj_put(syncobj); return ret; } +EXPORT_SYMBOL(drm_syncobj_search_fence); int drm_syncobj_find_fence(struct drm_file *file_private, u32 handle, u64 point, struct dma_fence **fence) { diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 7209dd832d39..bb20d318c9d6 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -2182,7 +2182,9 @@ await_fence_array(struct i915_execbuffer *eb, if (!(flags & I915_EXEC_FENCE_WAIT)) continue;

- fence = drm_syncobj_fence_get(syncobj); + drm_syncobj_search_fence(syncobj, 0, + DRM_SYNCOBJ_WAIT_FLAGS_WAIT_FOR_SUBMIT, + &fence); if (!fence) return -EINVAL;

diff --git a/include/drm/drm_syncobj.h b/include/drm/drm_syncobj.h index fb4e7edf90bf..ec098d4798b8 100644 --- a/include/drm/drm_syncobj.h +++ b/include/drm/drm_syncobj.h @@ -170,5 +170,7 @@ int drm_syncobj_create(struct drm_syncobj **out_syncobj, uint32_t flags, int drm_syncobj_get_handle(struct drm_file *file_private, struct drm_syncobj *syncobj, u32 *handle); int drm_syncobj_get_fd(struct drm_syncobj *syncobj, int *p_fd); +int drm_syncobj_search_fence(struct drm_syncobj *syncobj, u64 point, + u64 flags, struct dma_fence **fence);

#endif

Patches #1-#4 are Reviewed-by: Christian König christian.koenig@amd.com

I would squash patches #5-#8 into one and make sure that the patch is a clean replacement of the existing functionality with the new one.

For example patch #5 still has some code move for the stub fence which now seems unnecessary.

Regards, Christian.

Am 24.08.2018 um 14:12 schrieb Chunming Zhou:

That is certainly totally nonsense. dma_fence_enable_sw_signaling() is the function who is calling this callback.

Signed-off-by: Chunming Zhou david1.zhou@amd.com Cc: Jason Ekstrand jason@jlekstrand.net Reviewed-by: Christian König christian.koenig@amd.com Acked-by: Daniel Vetter daniel@ffwll.ch

---

drivers/gpu/drm/drm_syncobj.c | 1 - 1 file changed, 1 deletion(-)

diff --git a/drivers/gpu/drm/drm_syncobj.c b/drivers/gpu/drm/drm_syncobj.c index 3a8837c49639..d17ed75ac7e2 100644 --- a/drivers/gpu/drm/drm_syncobj.c +++ b/drivers/gpu/drm/drm_syncobj.c @@ -184,7 +184,6 @@ static const char *drm_syncobj_null_fence_get_name(struct dma_fence *fence)

static bool drm_syncobj_null_fence_enable_signaling(struct dma_fence *fence) {

• dma_fence_enable_sw_signaling(fence); return !dma_fence_is_signaled(fence); }