From 06fbca713e8e4a04c3506a64978969be580cd077 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Tue, 7 Apr 2015 16:20:36 +0100 Subject: drm/i915: Split the batch pool by engine I woke up one morning and found 50k objects sitting in the batch pool and every search seemed to iterate the entire list... Painting the screen in oils would provide a more fluid display. One issue with the current design is that we only check for retirements on the current ring when preparing to submit a new batch. This means that we can have thousands of "active" batches on another ring that we have to walk over. The simplest way to avoid that is to split the pools per ring and then our LRU execution ordering will also ensure that the inactive buffers remain at the front. v2: execlists still requires duplicate code. v3: execlists requires more duplicate code Signed-off-by: Chris Wilson Reviewed-by: Tvrtko Ursulin Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_batch_pool.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'drivers/gpu/drm/i915/i915_gem_batch_pool.c') diff --git a/drivers/gpu/drm/i915/i915_gem_batch_pool.c b/drivers/gpu/drm/i915/i915_gem_batch_pool.c index 21f3356cc0ab..1287abf55b84 100644 --- a/drivers/gpu/drm/i915/i915_gem_batch_pool.c +++ b/drivers/gpu/drm/i915/i915_gem_batch_pool.c @@ -96,8 +96,9 @@ i915_gem_batch_pool_get(struct i915_gem_batch_pool *pool, list_for_each_entry_safe(tmp, next, &pool->cache_list, batch_pool_list) { + /* The batches are strictly LRU ordered */ if (tmp->active) - continue; + break; /* While we're looping, do some clean up */ if (tmp->madv == __I915_MADV_PURGED) { -- cgit v1.2.3-59-g8ed1b