[Intel-gfx] [PATCH 10/15] drm/i915/guc: Get rid of GuC log runtime
Sagar Arun Kamble
sagar.a.kamble at intel.com
Tue Mar 6 09:30:22 UTC 2018
I think keeping the distinction between base/mandatory members and
runtime/optional members is okay as they are handled differently.
What do others think?
another related query, should we alloc/dealloc guc_log work queue in
relay_open/release?
On 2/27/2018 6:22 PM, Michał Winiarski wrote:
> Keeping a separate runtime struct is only making the code dealing with
> relay less clear to read. Let's get rid of it, keeping everything in the
> log instead.
>
> Signed-off-by: Michał Winiarski <michal.winiarski at intel.com>
> Cc: Chris Wilson <chris at chris-wilson.co.uk>
> Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio at intel.com>
> Cc: Sagar Arun Kamble <sagar.a.kamble at intel.com>
> Cc: Michal Wajdeczko <michal.wajdeczko at intel.com>
> ---
> drivers/gpu/drm/i915/intel_guc.c | 14 ++++----
> drivers/gpu/drm/i915/intel_guc_log.c | 68 ++++++++++++++++++------------------
> drivers/gpu/drm/i915/intel_guc_log.h | 13 +++----
> 3 files changed, 46 insertions(+), 49 deletions(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_guc.c b/drivers/gpu/drm/i915/intel_guc.c
> index 41f2c3b3c482..e3b6ae158a12 100644
> --- a/drivers/gpu/drm/i915/intel_guc.c
> +++ b/drivers/gpu/drm/i915/intel_guc.c
> @@ -87,9 +87,9 @@ int intel_guc_init_wq(struct intel_guc *guc)
> * or scheduled later on resume. This way the handling of work
> * item can be kept same between system suspend & rpm suspend.
> */
> - guc->log.runtime.flush_wq = alloc_ordered_workqueue("i915-guc_log",
> - WQ_HIGHPRI | WQ_FREEZABLE);
> - if (!guc->log.runtime.flush_wq) {
> + guc->log.flush_wq = alloc_ordered_workqueue("i915-guc_log",
> + WQ_HIGHPRI | WQ_FREEZABLE);
> + if (!guc->log.flush_wq) {
> DRM_ERROR("Couldn't allocate workqueue for GuC log\n");
> return -ENOMEM;
> }
> @@ -112,7 +112,7 @@ int intel_guc_init_wq(struct intel_guc *guc)
> guc->preempt_wq = alloc_ordered_workqueue("i915-guc_preempt",
> WQ_HIGHPRI);
> if (!guc->preempt_wq) {
> - destroy_workqueue(guc->log.runtime.flush_wq);
> + destroy_workqueue(guc->log.flush_wq);
> DRM_ERROR("Couldn't allocate workqueue for GuC "
> "preemption\n");
> return -ENOMEM;
> @@ -130,7 +130,7 @@ void intel_guc_fini_wq(struct intel_guc *guc)
> USES_GUC_SUBMISSION(dev_priv))
> destroy_workqueue(guc->preempt_wq);
>
> - destroy_workqueue(guc->log.runtime.flush_wq);
> + destroy_workqueue(guc->log.flush_wq);
> }
>
> static int guc_shared_data_create(struct intel_guc *guc)
> @@ -389,8 +389,8 @@ void intel_guc_notification_handler(struct intel_guc *guc)
>
> if (msg & (INTEL_GUC_RECV_MSG_FLUSH_LOG_BUFFER |
> INTEL_GUC_RECV_MSG_CRASH_DUMP_POSTED)) {
> - queue_work(guc->log.runtime.flush_wq,
> - &guc->log.runtime.flush_work);
> + queue_work(guc->log.flush_wq,
> + &guc->log.flush_work);
>
> guc->log.flush_interrupt_count++;
> }
> diff --git a/drivers/gpu/drm/i915/intel_guc_log.c b/drivers/gpu/drm/i915/intel_guc_log.c
> index d2aca10ab986..f95e18be1c5f 100644
> --- a/drivers/gpu/drm/i915/intel_guc_log.c
> +++ b/drivers/gpu/drm/i915/intel_guc_log.c
> @@ -156,10 +156,10 @@ static void guc_move_to_next_buf(struct intel_guc *guc)
> smp_wmb();
>
> /* All data has been written, so now move the offset of sub buffer. */
> - relay_reserve(guc->log.runtime.relay_chan, guc->log.vma->obj->base.size);
> + relay_reserve(guc->log.relay_chan, guc->log.vma->obj->base.size);
>
> /* Switch to the next sub buffer */
> - relay_flush(guc->log.runtime.relay_chan);
> + relay_flush(guc->log.relay_chan);
> }
>
> static void *guc_get_write_buffer(struct intel_guc *guc)
> @@ -173,7 +173,7 @@ static void *guc_get_write_buffer(struct intel_guc *guc)
> * done without using relay_reserve() along with relay_write(). So its
> * better to use relay_reserve() alone.
> */
> - return relay_reserve(guc->log.runtime.relay_chan, 0);
> + return relay_reserve(guc->log.relay_chan, 0);
> }
>
> static bool guc_check_log_buf_overflow(struct intel_guc *guc,
> @@ -224,11 +224,11 @@ static void guc_read_update_log_buffer(struct intel_guc *guc)
> void *src_data, *dst_data;
> bool new_overflow;
>
> - if (WARN_ON(!guc->log.runtime.buf_addr))
> + if (WARN_ON(!guc->log.buf_addr))
> return;
>
> /* Get the pointer to shared GuC log buffer */
> - log_buf_state = src_data = guc->log.runtime.buf_addr;
> + log_buf_state = src_data = guc->log.buf_addr;
>
> /* Get the pointer to local buffer to store the logs */
> log_buf_snapshot_state = dst_data = guc_get_write_buffer(guc);
> @@ -316,16 +316,16 @@ static void guc_read_update_log_buffer(struct intel_guc *guc)
> static void capture_logs_work(struct work_struct *work)
> {
> struct intel_guc *guc =
> - container_of(work, struct intel_guc, log.runtime.flush_work);
> + container_of(work, struct intel_guc, log.flush_work);
>
> - mutex_lock(&guc->log.runtime.lock);
> + mutex_lock(&guc->log.lock);
> guc_log_capture_logs(guc);
> - mutex_unlock(&guc->log.runtime.lock);
> + mutex_unlock(&guc->log.lock);
> }
>
> -static bool guc_log_has_runtime(struct intel_guc *guc)
> +static bool guc_log_relay_enabled(struct intel_guc *guc)
> {
> - return guc->log.runtime.buf_addr != NULL;
> + return guc->log.buf_addr != NULL;
> }
>
> static int guc_log_map(struct intel_guc *guc)
> @@ -354,23 +354,23 @@ static int guc_log_map(struct intel_guc *guc)
> }
> mutex_unlock(&dev_priv->drm.struct_mutex);
>
> - guc->log.runtime.buf_addr = vaddr;
> + guc->log.buf_addr = vaddr;
>
> return 0;
> }
>
> static void guc_log_unmap(struct intel_guc *guc)
> {
> - lockdep_assert_held(&guc->log.runtime.lock);
> + lockdep_assert_held(&guc->log.lock);
>
> i915_gem_object_unpin_map(guc->log.vma->obj);
> - guc->log.runtime.buf_addr = NULL;
> + guc->log.buf_addr = NULL;
> }
>
> void intel_guc_log_init_early(struct intel_guc *guc)
> {
> - mutex_init(&guc->log.runtime.lock);
> - INIT_WORK(&guc->log.runtime.flush_work, capture_logs_work);
> + mutex_init(&guc->log.lock);
> + INIT_WORK(&guc->log.flush_work, capture_logs_work);
> }
>
> int guc_log_relay_create(struct intel_guc *guc)
> @@ -380,7 +380,7 @@ int guc_log_relay_create(struct intel_guc *guc)
> size_t n_subbufs, subbuf_size;
> int ret;
>
> - lockdep_assert_held(&guc->log.runtime.lock);
> + lockdep_assert_held(&guc->log.lock);
>
> /* Keep the size of sub buffers same as shared log buffer */
> subbuf_size = GUC_LOG_SIZE;
> @@ -410,7 +410,7 @@ int guc_log_relay_create(struct intel_guc *guc)
> }
>
> GEM_BUG_ON(guc_log_relay_chan->subbuf_size < subbuf_size);
> - guc->log.runtime.relay_chan = guc_log_relay_chan;
> + guc->log.relay_chan = guc_log_relay_chan;
>
> return 0;
>
> @@ -422,10 +422,10 @@ int guc_log_relay_create(struct intel_guc *guc)
>
> void guc_log_relay_destroy(struct intel_guc *guc)
> {
> - lockdep_assert_held(&guc->log.runtime.lock);
> + lockdep_assert_held(&guc->log.lock);
>
> - relay_close(guc->log.runtime.relay_chan);
> - guc->log.runtime.relay_chan = NULL;
> + relay_close(guc->log.relay_chan);
> + guc->log.relay_chan = NULL;
> }
>
> static void guc_log_capture_logs(struct intel_guc *guc)
> @@ -525,7 +525,7 @@ int intel_guc_log_level_set(struct intel_guc *guc, u64 val)
> GUC_LOG_LEVEL_TO_VERBOSITY(val) > GUC_LOG_VERBOSITY_MAX)
> return -EINVAL;
>
> - mutex_lock(&guc->log.runtime.lock);
> + mutex_lock(&guc->log.lock);
>
> if (i915_modparams.guc_log_level == val) {
> ret = 0;
> @@ -542,7 +542,7 @@ int intel_guc_log_level_set(struct intel_guc *guc, u64 val)
> i915_modparams.guc_log_level = val;
>
> out_unlock:
> - mutex_unlock(&guc->log.runtime.lock);
> + mutex_unlock(&guc->log.lock);
>
> return ret;
> }
> @@ -551,9 +551,9 @@ int intel_guc_log_relay_open(struct intel_guc *guc)
> {
> int ret;
>
> - mutex_lock(&guc->log.runtime.lock);
> + mutex_lock(&guc->log.lock);
>
> - if (guc_log_has_runtime(guc)) {
> + if (guc_log_relay_enabled(guc)) {
> ret = -EEXIST;
> goto out_unlock;
> }
> @@ -576,7 +576,7 @@ int intel_guc_log_relay_open(struct intel_guc *guc)
> if (ret)
> goto out_relay;
>
> - mutex_unlock(&guc->log.runtime.lock);
> + mutex_unlock(&guc->log.lock);
>
> guc_log_flush_irq_enable(guc);
>
> @@ -585,14 +585,14 @@ int intel_guc_log_relay_open(struct intel_guc *guc)
> * the flush notification. This means that we need to unconditionally
> * flush on relay enabling, since GuC only notifies us once.
> */
> - queue_work(guc->log.runtime.flush_wq, &guc->log.runtime.flush_work);
> + queue_work(guc->log.flush_wq, &guc->log.flush_work);
>
> return 0;
>
> out_relay:
> guc_log_relay_destroy(guc);
> out_unlock:
> - mutex_unlock(&guc->log.runtime.lock);
> + mutex_unlock(&guc->log.lock);
>
> return ret;
> }
> @@ -605,27 +605,27 @@ void intel_guc_log_relay_flush(struct intel_guc *guc)
> * Before initiating the forceful flush, wait for any pending/ongoing
> * flush to complete otherwise forceful flush may not actually happen.
> */
> - flush_work(&guc->log.runtime.flush_work);
> + flush_work(&guc->log.flush_work);
>
> intel_runtime_pm_get(dev_priv);
> guc_log_flush(guc);
> intel_runtime_pm_put(dev_priv);
>
> - mutex_lock(&guc->log.runtime.lock);
> + mutex_lock(&guc->log.lock);
> /* GuC would have updated log buffer by now, so capture it */
> guc_log_capture_logs(guc);
> - mutex_unlock(&guc->log.runtime.lock);
> + mutex_unlock(&guc->log.lock);
> }
>
> void intel_guc_log_relay_close(struct intel_guc *guc)
> {
> - GEM_BUG_ON(!guc_log_has_runtime(guc));
> + GEM_BUG_ON(!guc_log_relay_enabled(guc));
>
> guc_log_flush_irq_disable(guc);
> - flush_work(&guc->log.runtime.flush_work);
> + flush_work(&guc->log.flush_work);
>
> - mutex_lock(&guc->log.runtime.lock);
> + mutex_lock(&guc->log.lock);
> guc_log_unmap(guc);
> guc_log_relay_destroy(guc);
> - mutex_unlock(&guc->log.runtime.lock);
> + mutex_unlock(&guc->log.lock);
> }
> diff --git a/drivers/gpu/drm/i915/intel_guc_log.h b/drivers/gpu/drm/i915/intel_guc_log.h
> index df91f12a36ed..11f49ef41f01 100644
> --- a/drivers/gpu/drm/i915/intel_guc_log.h
> +++ b/drivers/gpu/drm/i915/intel_guc_log.h
> @@ -42,14 +42,11 @@ struct intel_guc;
> struct intel_guc_log {
> u32 flags;
> struct i915_vma *vma;
> - /* The runtime stuff gets created only when GuC logging gets enabled */
> - struct {
> - void *buf_addr;
> - struct workqueue_struct *flush_wq;
> - struct work_struct flush_work;
> - struct rchan *relay_chan;
> - struct mutex lock;
> - } runtime;
> + void *buf_addr;
> + struct workqueue_struct *flush_wq;
> + struct work_struct flush_work;
> + struct rchan *relay_chan;
> + struct mutex lock;
> /* logging related stats */
> u32 capture_miss_count;
> u32 flush_interrupt_count;
--
Thanks,
Sagar
More information about the Intel-gfx
mailing list