From: Tvrtko Ursulin <tvrtko@xxxxxxxxxxx>
With the previous patch having extended the pinned lifetime of
contexts by referencing the previous context from the current
request until the latter is retired (completed by the GPU),
we can now remove usage of execlist retired queue entirely.
This is because the above now guarantees that all execlist
object access requirements are satisfied by this new tracking,
and we can stop taking additional references and stop keeping
request on the execlists retired queue.
The latter was a source of significant scalability issues in
the driver causing performance hits on some tests. Most
dramatical of which was igt/gem_close_race which had run time
in tens of minutes which is now reduced to tens of seconds.
Signed-off-by: Tvrtko Ursulin <tvrtko@xxxxxxxxxxx>
Cc: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
Reviewed-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
Signed-off-by: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
---
drivers/gpu/drm/i915/i915_gem.c | 10 +--------
drivers/gpu/drm/i915/intel_lrc.c | 39 ++++++++++++---------------------
drivers/gpu/drm/i915/intel_lrc.h | 2 +-
drivers/gpu/drm/i915/intel_ringbuffer.h | 1 -
4 files changed, 16 insertions(+), 36 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index d16680b77e13..a88aa7f11a9c 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -2876,13 +2876,7 @@ static void i915_gem_reset_engine_cleanup(struct drm_i915_private *dev_priv,
/* Ensure irq handler finishes or is cancelled. */
tasklet_kill(&engine->irq_tasklet);
- spin_lock_bh(&engine->execlist_lock);
- /* list_splice_tail_init checks for empty lists */
- list_splice_tail_init(&engine->execlist_queue,
- &engine->execlist_retired_req_list);
- spin_unlock_bh(&engine->execlist_lock);
-
- intel_execlists_retire_requests(engine);
+ intel_execlists_cancel_requests(engine);
}
/*
@@ -3006,8 +3000,6 @@ i915_gem_retire_requests(struct drm_device *dev)
spin_lock_bh(&engine->execlist_lock);
idle &= list_empty(&engine->execlist_queue);
spin_unlock_bh(&engine->execlist_lock);
-
- intel_execlists_retire_requests(engine);
}
}
diff --git a/drivers/gpu/drm/i915/intel_lrc.c b/drivers/gpu/drm/i915/intel_lrc.c
index 0c46a5cf4be0..b6f12cb4dba9 100644
--- a/drivers/gpu/drm/i915/intel_lrc.c
+++ b/drivers/gpu/drm/i915/intel_lrc.c
@@ -431,8 +431,8 @@ static void execlists_context_unqueue(struct intel_engine_cs *engine)
/* Same ctx: ignore first request, as second request
* will update tail past first request's workload */
cursor->elsp_submitted = req0->elsp_submitted;
- list_move_tail(&req0->execlist_link,
- &engine->execlist_retired_req_list);
+ list_del(&req0->execlist_link);
+ i915_gem_request_unreference(req0);
req0 = cursor;
} else {
req1 = cursor;
@@ -464,7 +464,7 @@ static void execlists_context_unqueue(struct intel_engine_cs *engine)
}
static unsigned int
-execlists_check_remove_request(struct intel_engine_cs *engine, u32 request_id)
+execlists_check_remove_request(struct intel_engine_cs *engine, u32 ctx_id)
{
struct drm_i915_gem_request *head_req;
@@ -474,19 +474,16 @@ execlists_check_remove_request(struct intel_engine_cs *engine, u32 request_id)
struct drm_i915_gem_request,
execlist_link);
- if (!head_req)
- return 0;
-
- if (unlikely(head_req->ctx_hw_id != request_id))
- return 0;
+ if (WARN_ON(!head_req || (head_req->ctx_hw_id != ctx_id)))
+ return 0;
WARN(head_req->elsp_submitted == 0, "Never submitted head request\n");
if (--head_req->elsp_submitted > 0)
return 0;
- list_move_tail(&head_req->execlist_link,
- &engine->execlist_retired_req_list);
+ list_del(&head_req->execlist_link);
+ i915_gem_request_unreference(head_req);
return 1;
}
@@ -590,9 +587,6 @@ static void execlists_context_queue(struct drm_i915_gem_request *request)
struct drm_i915_gem_request *cursor;
int num_elements = 0;
- intel_lr_context_pin(request->ctx, request->engine);
- i915_gem_request_reference(request);
-
spin_lock_bh(&engine->execlist_lock);
list_for_each_entry(cursor, &engine->execlist_queue, execlist_link)
@@ -609,11 +603,12 @@ static void execlists_context_queue(struct drm_i915_gem_request *request)
if (request->ctx == tail_req->ctx) {
WARN(tail_req->elsp_submitted != 0,
"More than 2 already-submitted reqs queued\n");
- list_move_tail(&tail_req->execlist_link,
- &engine->execlist_retired_req_list);
+ list_del(&tail_req->execlist_link);
+ i915_gem_request_unreference(tail_req);
}
}
+ i915_gem_request_reference(request);
list_add_tail(&request->execlist_link, &engine->execlist_queue);
request->ctx_hw_id = request->ctx->hw_id;
if (num_elements == 0)
@@ -1011,23 +1006,18 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
return 0;
}
-void intel_execlists_retire_requests(struct intel_engine_cs *engine)
+void intel_execlists_cancel_requests(struct intel_engine_cs *engine)
{
struct drm_i915_gem_request *req, *tmp;
- struct list_head retired_list;
+ LIST_HEAD(cancel_list);
WARN_ON(!mutex_is_locked(&engine->dev->struct_mutex));
- if (list_empty(&engine->execlist_retired_req_list))
- return;
- INIT_LIST_HEAD(&retired_list);
spin_lock_bh(&engine->execlist_lock);
- list_replace_init(&engine->execlist_retired_req_list, &retired_list);
+ list_replace_init(&engine->execlist_queue, &cancel_list);
spin_unlock_bh(&engine->execlist_lock);
- list_for_each_entry_safe(req, tmp, &retired_list, execlist_link) {
- intel_lr_context_unpin(req->ctx, engine);
-
+ list_for_each_entry_safe(req, tmp, &cancel_list, execlist_link) {
list_del(&req->execlist_link);
i915_gem_request_unreference(req);
}
@@ -2119,7 +2109,6 @@ logical_ring_init(struct drm_device *dev, struct intel_engine_cs *engine)
INIT_LIST_HEAD(&engine->buffers);
INIT_LIST_HEAD(&engine->execlist_queue);
- INIT_LIST_HEAD(&engine->execlist_retired_req_list);
spin_lock_init(&engine->execlist_lock);
tasklet_init(&engine->irq_tasklet,
diff --git a/drivers/gpu/drm/i915/intel_lrc.h b/drivers/gpu/drm/i915/intel_lrc.h
index 8bea937973f6..4b1c896b5019 100644
--- a/drivers/gpu/drm/i915/intel_lrc.h
+++ b/drivers/gpu/drm/i915/intel_lrc.h
@@ -119,6 +119,6 @@ int intel_execlists_submission(struct i915_execbuffer_params *params,
struct drm_i915_gem_execbuffer2 *args,
struct list_head *vmas);
-void intel_execlists_retire_requests(struct intel_engine_cs *engine);
+void intel_execlists_cancel_requests(struct intel_engine_cs *engine);
#endif /* _INTEL_LRC_H_ */
diff --git a/drivers/gpu/drm/i915/intel_ringbuffer.h b/drivers/gpu/drm/i915/intel_ringbuffer.h
index 2ade194bbea9..527549dbeb3c 100644
--- a/drivers/gpu/drm/i915/intel_ringbuffer.h
+++ b/drivers/gpu/drm/i915/intel_ringbuffer.h
@@ -269,7 +269,6 @@ struct intel_engine_cs {
struct tasklet_struct irq_tasklet;
spinlock_t execlist_lock; /* used inside tasklet, use spin_lock_bh */
struct list_head execlist_queue;
- struct list_head execlist_retired_req_list;
unsigned int fw_domains;
unsigned int next_context_status_buffer;
unsigned int idle_lite_restore_wa;
--
2.8.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
