From: Sourab Gupta <sourab.gu...@intel.com>
This patch adds support for opening multiple concurrent perf streams for
different gpu engines, while having the restriction to open only a single
stream open for a particular gpu engine.
This enables userspace client to open multiple streams, one per engine,
at any time to capture sample data for multiple gpu engines.
Signed-off-by: Sourab Gupta <sourab.gu...@intel.com>
---
drivers/gpu/drm/i915/i915_drv.h | 2 +-
drivers/gpu/drm/i915/i915_perf.c | 65 +++++++++++++++++++++++-----------------
2 files changed, 38 insertions(+), 29 deletions(-)
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index b1c952c..bf65acb 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -2086,7 +2086,7 @@ struct drm_i915_private {
spinlock_t hook_lock;
struct hrtimer poll_check_timer;
- struct i915_perf_stream *exclusive_stream;
+ struct i915_perf_stream *exclusive_stream[I915_NUM_RINGS];
wait_queue_head_t poll_wq[I915_NUM_RINGS];
struct {
diff --git a/drivers/gpu/drm/i915/i915_perf.c b/drivers/gpu/drm/i915/i915_perf.c
index 1d2712d..3eb56d4 100644
--- a/drivers/gpu/drm/i915/i915_perf.c
+++ b/drivers/gpu/drm/i915/i915_perf.c
@@ -1061,7 +1061,7 @@ static void i915_perf_stream_destroy(struct
i915_perf_stream *stream)
{
struct drm_i915_private *dev_priv = stream->dev_priv;
- BUG_ON(stream != dev_priv->perf.exclusive_stream);
+ BUG_ON(stream != dev_priv->perf.exclusive_stream[stream->ring_id]);
if (stream->using_oa) {
dev_priv->perf.oa.ops.disable_metric_set(dev_priv);
@@ -1075,7 +1075,7 @@ static void i915_perf_stream_destroy(struct
i915_perf_stream *stream)
if (stream->cs_mode)
free_command_stream_buf(dev_priv, stream->ring_id);
- dev_priv->perf.exclusive_stream = NULL;
+ dev_priv->perf.exclusive_stream[stream->ring_id] = NULL;
}
static void *vmap_oa_buffer(struct drm_i915_gem_object *obj)
@@ -1434,17 +1434,17 @@ static void gen7_update_oacontrol_locked(struct
drm_i915_private *dev_priv)
{
assert_spin_locked(&dev_priv->perf.hook_lock);
- if (dev_priv->perf.exclusive_stream->enabled) {
+ if (dev_priv->perf.exclusive_stream[RCS]->enabled) {
unsigned long ctx_id = 0;
bool pinning_ok = false;
- if (dev_priv->perf.exclusive_stream->ctx &&
+ if (dev_priv->perf.exclusive_stream[RCS]->ctx &&
dev_priv->perf.oa.specific_ctx_id) {
ctx_id = dev_priv->perf.oa.specific_ctx_id;
pinning_ok = true;
}
- if (dev_priv->perf.exclusive_stream->ctx == NULL ||
+ if (dev_priv->perf.exclusive_stream[RCS]->ctx == NULL ||
pinning_ok) {
bool periodic = dev_priv->perf.oa.periodic;
u32 period_exponent = dev_priv->perf.oa.period_exponent;
@@ -1556,14 +1556,6 @@ static int i915_perf_stream_init(struct i915_perf_stream
*stream,
SAMPLE_TS);
int ret;
- /* To avoid the complexity of having to accurately filter
- * counter reports and marshal to the appropriate client
- * we currently only allow exclusive access */
- if (dev_priv->perf.exclusive_stream) {
- DRM_ERROR("Stream already in use\n");
- return -EBUSY;
- }
-
/* Ctx Id can be sampled in HSW only through command streamer mode */
if (IS_HASWELL(dev_priv->dev) &&
(props->sample_flags & SAMPLE_CTX_ID) && !props->cs_mode) {
@@ -1576,6 +1568,13 @@ static int i915_perf_stream_init(struct i915_perf_stream
*stream,
if (require_oa_unit) {
int format_size;
+
+ /* Only allow exclusive access per stream */
+ if (dev_priv->perf.exclusive_stream[RCS]) {
+ DRM_ERROR("Stream:0 already in use\n");
+ return -EBUSY;
+ }
+
if (!dev_priv->perf.oa.ops.init_oa_buffer) {
DRM_ERROR("OA unit not supported\n");
return -ENODEV;
@@ -1673,6 +1672,13 @@ static int i915_perf_stream_init(struct i915_perf_stream
*stream,
}
if (props->cs_mode) {
+ /* Only allow exclusive access per stream */
+ if (dev_priv->perf.exclusive_stream[props->ring_id]) {
+ DRM_ERROR("Stream:%d already in use\n", props->ring_id);
+ ret = -EBUSY;
+ goto cs_error;
+ }
+
/*
* The only time we should allow enabling CS mode if it's not
* strictly required, is if SAMPLE_CTX_ID has been requested
@@ -1709,7 +1715,7 @@ static int i915_perf_stream_init(struct i915_perf_stream
*stream,
goto cs_error;
}
- dev_priv->perf.exclusive_stream = stream;
+ dev_priv->perf.exclusive_stream[stream->ring_id] = stream;
stream->destroy = i915_perf_stream_destroy;
stream->enable = i915_perf_stream_enable;
@@ -1751,8 +1757,8 @@ static void i915_oa_context_pin_notify_locked(struct
drm_i915_private *dev_priv,
dev_priv->perf.oa.ops.update_hw_ctx_id_locked == NULL)
return;
- if (dev_priv->perf.exclusive_stream &&
- dev_priv->perf.exclusive_stream->ctx == context) {
+ if (dev_priv->perf.exclusive_stream[RCS] &&
+ dev_priv->perf.exclusive_stream[RCS]->ctx == context) {
struct drm_i915_gem_object *obj =
context->legacy_hw_ctx.rcs_state;
u32 ctx_id = i915_gem_obj_ggtt_offset(obj);
@@ -1820,8 +1826,8 @@ void i915_oa_legacy_ctx_switch_notify(struct
drm_i915_gem_request *req)
if (dev_priv->perf.oa.ops.legacy_ctx_switch_unlocked == NULL)
return;
- if (dev_priv->perf.exclusive_stream &&
- dev_priv->perf.exclusive_stream->enabled) {
+ if (dev_priv->perf.exclusive_stream[RCS] &&
+ dev_priv->perf.exclusive_stream[RCS]->enabled) {
/* XXX: We don't take a lock here and this may run
* async with respect to stream methods. Notably we
@@ -1944,23 +1950,26 @@ static ssize_t i915_perf_read(struct file *file,
return ret;
}
-static enum hrtimer_restart poll_check_timer_cb(struct hrtimer *hrtimer)
+static void wake_up_perf_streams(void *data, async_cookie_t cookie)
{
+ struct drm_i915_private *dev_priv = data;
struct i915_perf_stream *stream;
- struct drm_i915_private *dev_priv =
- container_of(hrtimer, typeof(*dev_priv),
- perf.poll_check_timer);
-
- /* No need to protect the streams list here, since the hrtimer is
- * disabled before the stream is removed from list, and currently a
- * single exclusive_stream is supported.
- * XXX: revisit this when multiple concurrent streams are supported.
- */
+ mutex_lock(&dev_priv->perf.streams_lock);
list_for_each_entry(stream, &dev_priv->perf.streams, link) {
if (stream_have_data__unlocked(stream))
wake_up(&dev_priv->perf.poll_wq[stream->ring_id]);
}
+ mutex_unlock(&dev_priv->perf.streams_lock);
+}
+
+static enum hrtimer_restart poll_check_timer_cb(struct hrtimer *hrtimer)
+{
+ struct drm_i915_private *dev_priv =
+ container_of(hrtimer, typeof(*dev_priv),
+ perf.poll_check_timer);
+
+ async_schedule(wake_up_perf_streams, dev_priv);
hrtimer_forward_now(hrtimer, ns_to_ktime(POLL_PERIOD));
--
1.9.1
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
