Re: [Intel-gfx] [RFC PATCH 38/97] drm/i915/guc: Optimize CTB writes and reads

[Michal Wajdeczko <michal.wajdeczko@intel.com>]

On 06.05.2021 21:13, Matthew Brost wrote:
> CTB writes are now in the path of command submission and should be
> optimized for performance. Rather than reading CTB descriptor values
> (e.g. head, tail, size) which could result in accesses across the PCIe

size was removed from the descriptor in 25/97

> bus, store shadow local copies and only read/write the descriptor
> values when absolutely necessary.

maybe worth to add some words about caching available space ?

> 
> Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
> Signed-off-by: Matthew Brost <matthew.brost@intel.com>
> ---
>  drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c | 78 +++++++++++++----------
>  drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h |  6 ++
>  2 files changed, 52 insertions(+), 32 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> index 4eab319d61be..77dfbc94dcc3 100644
> --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> @@ -127,6 +127,10 @@ static void guc_ct_buffer_desc_init(struct guc_ct_buffer_desc *desc)
>  static void guc_ct_buffer_reset(struct intel_guc_ct_buffer *ctb)
>  {
>  	ctb->broken = false;
> +	ctb->tail = 0;
> +	ctb->head = 0;
> +	ctb->space = CIRC_SPACE(ctb->tail, ctb->head, ctb->size);
> +
>  	guc_ct_buffer_desc_init(ctb->desc);
>  }
>  
> @@ -371,10 +375,8 @@ static int ct_write(struct intel_guc_ct *ct,
>  {
>  	struct intel_guc_ct_buffer *ctb = &ct->ctbs.send;
>  	struct guc_ct_buffer_desc *desc = ctb->desc;
> -	u32 head = desc->head;
> -	u32 tail = desc->tail;
> +	u32 tail = ctb->tail;
>  	u32 size = ctb->size;
> -	u32 used;
>  	u32 header;
>  	u32 hxg;
>  	u32 *cmds = ctb->cmds;
> @@ -386,25 +388,14 @@ static int ct_write(struct intel_guc_ct *ct,
>  	if (unlikely(desc->status))
>  		goto corrupted;
>  
> -	if (unlikely((tail | head) >= size)) {
> +#ifdef CONFIG_DRM_I915_DEBUG_GUC
> +	if (unlikely((desc->tail | desc->head) >= size)) {
>  		CT_ERROR(ct, "Invalid offsets head=%u tail=%u (size=%u)\n",
> -			 head, tail, size);
> +			 desc->head, desc->tail, size);
>  		desc->status |= GUC_CTB_STATUS_OVERFLOW;
>  		goto corrupted;

nit: as we are now caching tail value, we can start comparing it with
the value in descriptor and report GUC_CTB_STATUS_MISMATCH if needed

>  	}
> -
> -	/*
> -	 * tail == head condition indicates empty. GuC FW does not support
> -	 * using up the entire buffer to get tail == head meaning full.
> -	 */
> -	if (tail < head)
> -		used = (size - head) + tail;
> -	else
> -		used = tail - head;
> -
> -	/* make sure there is a space including extra dw for the fence */
> -	if (unlikely(used + len + 1 >= size))
> -		return -ENOSPC;
> +#endif
>  
>  	/*
>  	 * dw0: CT header (including fence)
> @@ -444,7 +435,9 @@ static int ct_write(struct intel_guc_ct *ct,
>  	write_barrier(ct);
>  
>  	/* now update descriptor */
> +	ctb->tail = tail;
>  	WRITE_ONCE(desc->tail, tail);
> +	ctb->space -= len + 1;
>  
>  	return 0;
>  
> @@ -460,7 +453,7 @@ static int ct_write(struct intel_guc_ct *ct,
>   * @req:	pointer to pending request
>   * @status:	placeholder for status
>   *
> - * For each sent request, Guc shall send bac CT response message.
> + * For each sent request, GuC shall send back CT response message.
>   * Our message handler will update status of tracked request once
>   * response message with given fence is received. Wait here and
>   * check for valid response status value.
> @@ -508,24 +501,35 @@ static inline bool ct_deadlocked(struct intel_guc_ct *ct)
>  	return ret;
>  }
>  
> -static inline bool ctb_has_room(struct intel_guc_ct_buffer *ctb, u32 len_dw)
> +static inline bool h2g_has_room(struct intel_guc_ct *ct, u32 len_dw)

this function was introduced moment ago ...
can we minimize number of changes between patches ?

>  {
> -	struct guc_ct_buffer_desc *desc = ctb->desc;
> -	u32 head = READ_ONCE(desc->head);
> +	struct intel_guc_ct_buffer *ctb = &ct->ctbs.send;
> +	u32 head;
>  	u32 space;
>  
> -	space = CIRC_SPACE(desc->tail, head, ctb->size);
> +	if (ctb->space >= len_dw)
> +		return true;
> +
> +	head = READ_ONCE(ctb->desc->head);
> +	if (unlikely(head > ctb->size)) {
> +		CT_ERROR(ct, "Corrupted descriptor head=%u tail=%u size=%u\n",
> +			  ctb->desc->head, ctb->desc->tail, ctb->size);
> +		ctb->desc->status |= GUC_CTB_STATUS_OVERFLOW;
> +		ctb->broken = true;
> +		return false;
> +	}
> +
> +	space = CIRC_SPACE(ctb->tail, head, ctb->size);
> +	ctb->space = space;

shouldn't we update our ctb->head with new head value?

>  
>  	return space >= len_dw;
>  }
>  
>  static int has_room_nb(struct intel_guc_ct *ct, u32 len_dw)
>  {
> -	struct intel_guc_ct_buffer *ctb = &ct->ctbs.send;
> -
>  	lockdep_assert_held(&ct->ctbs.send.lock);
>  
> -	if (unlikely(!ctb_has_room(ctb, len_dw))) {
> +	if (unlikely(!h2g_has_room(ct, len_dw))) {
>  		if (ct->stall_time == KTIME_MAX)
>  			ct->stall_time = ktime_get();
>  
> @@ -593,11 +597,11 @@ static int ct_send(struct intel_guc_ct *ct,
>  	 * rare.
>  	 */
>  retry:
> -	spin_lock_irqsave(&ct->ctbs.send.lock, flags);
> -	if (unlikely(!ctb_has_room(ctb, len + 1))) {
> +	spin_lock_irqsave(&ctb->lock, flags);
> +	if (unlikely(!h2g_has_room(ct, len + 1))) {
>  		if (ct->stall_time == KTIME_MAX)
>  			ct->stall_time = ktime_get();
> -		spin_unlock_irqrestore(&ct->ctbs.send.lock, flags);
> +		spin_unlock_irqrestore(&ctb->lock, flags);

this ...

>  
>  		if (unlikely(ct_deadlocked(ct)))
>  			return -EDEADLK;
> @@ -620,7 +624,7 @@ static int ct_send(struct intel_guc_ct *ct,
>  
>  	err = ct_write(ct, action, len, fence, 0);
>  
> -	spin_unlock_irqrestore(&ct->ctbs.send.lock, flags);
> +	spin_unlock_irqrestore(&ctb->lock, flags);

and this likely could be done in earlier patch

>  
>  	if (unlikely(err))
>  		goto unlink;
> @@ -708,7 +712,7 @@ static int ct_read(struct intel_guc_ct *ct, struct ct_incoming_msg **msg)
>  {
>  	struct intel_guc_ct_buffer *ctb = &ct->ctbs.recv;
>  	struct guc_ct_buffer_desc *desc = ctb->desc;
> -	u32 head = desc->head;
> +	u32 head = ctb->head;
>  	u32 tail = desc->tail;
>  	u32 size = ctb->size;
>  	u32 *cmds = ctb->cmds;
> @@ -723,12 +727,21 @@ static int ct_read(struct intel_guc_ct *ct, struct ct_incoming_msg **msg)
>  	if (unlikely(desc->status))
>  		goto corrupted;
>  
> -	if (unlikely((tail | head) >= size)) {
> +#ifdef CONFIG_DRM_I915_DEBUG_GUC
> +	if (unlikely((desc->tail | desc->head) >= size)) {
>  		CT_ERROR(ct, "Invalid offsets head=%u tail=%u (size=%u)\n",
>  			 head, tail, size);
>  		desc->status |= GUC_CTB_STATUS_OVERFLOW;
>  		goto corrupted;
>  	}
> +#else
> +	if (unlikely((tail | ctb->head) >= size)) {

we are in control of cached 'ctb->head' so it shall never be > size

> +		CT_ERROR(ct, "Invalid offsets head=%u tail=%u (size=%u)\n",
> +			 head, tail, size);
> +		desc->status |= GUC_CTB_STATUS_OVERFLOW;
> +		goto corrupted;
> +	}
> +#endif
>  
>  	/* tail == head condition indicates empty */
>  	available = tail - head;
> @@ -778,6 +791,7 @@ static int ct_read(struct intel_guc_ct *ct, struct ct_incoming_msg **msg)
>  	}
>  	CT_DEBUG(ct, "received %*ph\n", 4 * len, (*msg)->msg);
>  
> +	ctb->head = head;
>  	/* now update descriptor */
>  	WRITE_ONCE(desc->head, head);
>  
> diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h
> index 55ef7c52472f..9924335e2ee6 100644
> --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h
> +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h
> @@ -33,6 +33,9 @@ struct intel_guc;
>   * @desc: pointer to the buffer descriptor
>   * @cmds: pointer to the commands buffer
>   * @size: size of the commands buffer in dwords
> + * @head: local shadow copy of head in dwords
> + * @tail: local shadow copy of tail in dwords
> + * @space: local shadow copy of space in dwords
>   * @broken: flag to indicate if descriptor data is broken
>   */
>  struct intel_guc_ct_buffer {
> @@ -40,6 +43,9 @@ struct intel_guc_ct_buffer {
>  	struct guc_ct_buffer_desc *desc;
>  	u32 *cmds;
>  	u32 size;
> +	u32 tail;
> +	u32 head;
> +	u32 space;
>  	bool broken;
>  };
>  
> 
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx