openkylin
/
linux
mirror of https://gitee.com/openkylin/linux.git
9
0
Fork 0
Code Issues Projects Releases Wiki Activity

drm/i915: Convert the file mutex into a spinlock 

Daniel Vetter pointed out that in this case is would be clearer and
cleaner to use a spinlock instead of a mutex to protect the per-file
request list manipulation. Make it so.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
This commit is contained in:
Chris Wilson 2010-09-26 11:03:27 +01:00
parent 447da18742
commit 1c25595f8d
3 changed files with 17 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
drivers/gpu/drm/i915/i915_dma.c
View File

@ -2173,8 +2173,8 @@ int i915_driver_open(struct drm_device *dev, struct drm_file *file)
file->driver_priv = file_priv;
spin_lock_init(&file_priv->mm.lock);
INIT_LIST_HEAD(&file_priv->mm.request_list);
mutex_init(&file_priv->mutex);
return 0;
}

2
drivers/gpu/drm/i915/i915_drv.h
View File

@ -851,8 +851,8 @@ struct drm_i915_gem_request {
};
struct drm_i915_file_private {
struct mutex mutex;
struct {
struct spinlock lock;
struct list_head request_list;
} mm;
};

28
drivers/gpu/drm/i915/i915_gem.c
View File

@ -1694,11 +1694,11 @@ i915_add_request(struct drm_device *dev,
list_add_tail(&request->list, &ring->request_list);
if (file_priv) {
mutex_lock(&file_priv->mutex);
spin_lock(&file_priv->mm.lock);
request->file_priv = file_priv;
list_add_tail(&request->client_list,
&file_priv->mm.request_list);
mutex_unlock(&file_priv->mutex);
spin_unlock(&file_priv->mm.lock);
}
if (!dev_priv->mm.suspended) {
@ -1733,11 +1733,15 @@ i915_retire_commands(struct drm_device *dev, struct intel_ring_buffer *ring)
static inline void
i915_gem_request_remove_from_client(struct drm_i915_gem_request *request)
{
if (request->file_priv) {
mutex_lock(&request->file_priv->mutex);
list_del(&request->client_list);
mutex_unlock(&request->file_priv->mutex);
}
struct drm_i915_file_private *file_priv = request->file_priv;
if (!file_priv)
return;
spin_lock(&file_priv->mm.lock);
list_del(&request->client_list);
request->file_priv = NULL;
spin_unlock(&file_priv->mm.lock);
}
static void i915_gem_reset_ring_lists(struct drm_i915_private *dev_priv,
@ -3464,7 +3468,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
u32 seqno = 0;
int ret;
mutex_lock(&file_priv->mutex);
spin_lock(&file_priv->mm.lock);
list_for_each_entry(request, &file_priv->mm.request_list, client_list) {
if (time_after_eq(request->emitted_jiffies, recent_enough))
break;
@ -3472,7 +3476,7 @@ i915_gem_ring_throttle(struct drm_device *dev, struct drm_file *file)
ring = request->ring;
seqno = request->seqno;
}
mutex_unlock(&file_priv->mutex);
spin_unlock(&file_priv->mm.lock);
if (seqno == 0)
return 0;
@ -4974,8 +4978,7 @@ void i915_gem_release(struct drm_device *dev, struct drm_file *file)
* later retire_requests won't dereference our soon-to-be-gone
* file_priv.
*/
mutex_lock(&dev->struct_mutex);
mutex_lock(&file_priv->mutex);
spin_lock(&file_priv->mm.lock);
while (!list_empty(&file_priv->mm.request_list)) {
struct drm_i915_gem_request *request;
@ -4985,8 +4988,7 @@ void i915_gem_release(struct drm_device *dev, struct drm_file *file)
list_del(&request->client_list);
request->file_priv = NULL;
}
mutex_unlock(&file_priv->mutex);
mutex_unlock(&dev->struct_mutex);
spin_unlock(&file_priv->mm.lock);
}
static int