Intel-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: John Harrison <john.c.harrison@intel.com>
To: Matthew Brost <matthew.brost@intel.com>,
	intel-gfx@lists.freedesktop.org,
	 dri-devel@lists.freedesktop.org
Subject: Re: [Intel-gfx] [PATCH 6/7] drm/i915/guc: Optimize CTB writes and reads
Date: Tue, 6 Jul 2021 15:51:00 -0700	[thread overview]
Message-ID: <9834f3c0-a1a3-9b1a-d291-732501b7a013@intel.com> (raw)
In-Reply-To: <20210706222010.101522-7-matthew.brost@intel.com>

On 7/6/2021 15:20, 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) which could result in accesses across the PCIe bus,
> store shadow local copies and only read/write the descriptor values when
> absolutely necessary. Also store the current space in the each channel
> locally.
>
> v2:
>   (Michal)
>    - Add additional sanity checks for head / tail pointers
>    - Use GUC_CTB_HDR_LEN rather than magic 1
> v3:
>   (Michal / John H)
>    - Drop redundant check of head value
>
> 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 | 88 +++++++++++++++--------
>   drivers/gpu/drm/i915/gt/uc/intel_guc_ct.h |  6 ++
>   2 files changed, 65 insertions(+), 29 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 db3e85b89573..4a73a1f03a9b 100644
> --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> @@ -130,6 +130,10 @@ 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);
>   }
>   
> @@ -383,10 +387,8 @@ 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 type;
> @@ -396,25 +398,22 @@ static int ct_write(struct intel_guc_ct *ct,
>   	if (unlikely(desc->status))
>   		goto corrupted;
>   
> -	if (unlikely((tail | head) >= size)) {
> +	GEM_BUG_ON(tail > size);
> +
> +#ifdef CONFIG_DRM_I915_DEBUG_GUC
> +	if (unlikely(tail != READ_ONCE(desc->tail))) {
> +		CT_ERROR(ct, "Tail was modified %u != %u\n",
> +			 desc->tail, ctb->tail);
> +		desc->status |= GUC_CTB_STATUS_MISMATCH;
> +		goto corrupted;
> +	}
> +	if (unlikely((desc->tail | desc->head) >= size)) {
Same arguments below about head apply to tail here. Also, there is no 
#else check on ctb->head?

>   		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;
>   	}
> -
> -	/*
> -	 * 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 header */
> -	if (unlikely(used + len + GUC_CTB_HDR_LEN >= size))
> -		return -ENOSPC;
> +#endif
>   
>   	/*
>   	 * dw0: CT header (including fence)
> @@ -453,7 +452,9 @@ 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 + GUC_CTB_HDR_LEN;
>   
>   	return 0;
>   
> @@ -469,7 +470,7 @@ 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.
> @@ -525,24 +526,35 @@ static inline bool ct_deadlocked(struct intel_guc_ct *ct)
>   	return ret;
>   }
>   
> -static inline bool h2g_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)
>   {
> -	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;
>   
>   	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(!h2g_has_room(ctb, len_dw))) {
> +	if (unlikely(!h2g_has_room(ct, len_dw))) {
>   		if (ct->stall_time == KTIME_MAX)
>   			ct->stall_time = ktime_get();
>   
> @@ -612,7 +624,7 @@ static int ct_send(struct intel_guc_ct *ct,
>   	 */
>   retry:
>   	spin_lock_irqsave(&ctb->lock, flags);
> -	if (unlikely(!h2g_has_room(ctb, len + GUC_CTB_HDR_LEN))) {
> +	if (unlikely(!h2g_has_room(ct, len + GUC_CTB_HDR_LEN))) {
>   		if (ct->stall_time == KTIME_MAX)
>   			ct->stall_time = ktime_get();
>   		spin_unlock_irqrestore(&ctb->lock, flags);
> @@ -732,7 +744,7 @@ 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;
> @@ -747,12 +759,29 @@ 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)) {
> +	GEM_BUG_ON(head > size);
> +
> +#ifdef CONFIG_DRM_I915_DEBUG_GUC
> +	if (unlikely(head != READ_ONCE(desc->head))) {
> +		CT_ERROR(ct, "Head was modified %u != %u\n",
> +			 desc->head, ctb->head);
> +		desc->status |= GUC_CTB_STATUS_MISMATCH;
> +		goto corrupted;
> +	}
> +	if (unlikely((desc->tail | desc->head) >= size)) {
As per comment in other thread, the check on head here is redundant 
because you have already hit a BUG_ON(ctb->head > size) followed by 
CT_ERROR(ctb->head != desc->head). Therefore, you can't get here if 
'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 >= size)) {
> +		CT_ERROR(ct, "Invalid offsets tail=%u (size=%u)\n",
> +			 tail, size);
> +		desc->status |= GUC_CTB_STATUS_OVERFLOW;
> +		goto corrupted;
> +	}
> +#endif
>   
>   	/* tail == head condition indicates empty */
>   	available = tail - head;
> @@ -802,6 +831,7 @@ 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 bee03794c1eb..edd1bba0445d 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

  reply	other threads:[~2021-07-06 22:51 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-07-06 22:20 [Intel-gfx] [PATCH 0/7] CT changes required for GuC submission Matthew Brost
2021-07-06 22:10 ` [Intel-gfx] ✗ Fi.CI.CHECKPATCH: warning for CT changes required for GuC submission (rev3) Patchwork
2021-07-06 22:11 ` [Intel-gfx] ✗ Fi.CI.SPARSE: " Patchwork
2021-07-06 22:20 ` [Intel-gfx] [PATCH 1/7] drm/i915/guc: Relax CTB response timeout Matthew Brost
2021-07-06 22:20 ` [Intel-gfx] [PATCH 2/7] drm/i915/guc: Improve error message for unsolicited CT response Matthew Brost
2021-07-06 22:20 ` [Intel-gfx] [PATCH 3/7] drm/i915/guc: Increase size of CTB buffers Matthew Brost
2021-07-06 22:20 ` [Intel-gfx] [PATCH 4/7] drm/i915/guc: Add non blocking CTB send function Matthew Brost
2021-07-06 22:20 ` [Intel-gfx] [PATCH 5/7] drm/i915/guc: Add stall timer to " Matthew Brost
2021-07-06 22:20 ` [Intel-gfx] [PATCH 6/7] drm/i915/guc: Optimize CTB writes and reads Matthew Brost
2021-07-06 22:51   ` John Harrison [this message]
2021-07-07 17:50     ` Matthew Brost
2021-07-07 18:19       ` John Harrison
2021-07-07 18:56         ` Matthew Brost
2021-07-07 20:21           ` John Harrison
2021-07-07 20:23             ` Matthew Brost
2021-07-06 22:20 ` [Intel-gfx] [PATCH 7/7] drm/i915/guc: Module load failure test for CT buffer creation Matthew Brost
2021-07-06 22:38 ` [Intel-gfx] ✗ Fi.CI.BAT: failure for CT changes required for GuC submission (rev3) Patchwork
2021-07-07 19:09 ` [Intel-gfx] [PATCH 06/56] drm/i915/guc: Optimize CTB writes and reads Matthew Brost
2021-07-07 20:30   ` Michal Wajdeczko
2021-07-07 23:25 ` [Intel-gfx] [PATCH 06/7] " Matthew Brost
2021-07-08 13:23   ` Michal Wajdeczko
2021-07-08  0:30 ` [Intel-gfx] [PATCH 0/2] Introduce set_parallel2 extension Matthew Brost
2021-07-08  0:30   ` [Intel-gfx] [PATCH 2/2] REVIEW: Full tree diff against internal/internal Matthew Brost
2021-07-08  2:18 ` [Intel-gfx] ✗ Fi.CI.BUILD: failure for CT changes required for GuC submission (rev4) Patchwork
  -- strict thread matches above, loose matches on Subject: below --
2021-07-08 16:20 [Intel-gfx] [PATCH 0/7] CT changes required for GuC submission Matthew Brost
2021-07-08 16:20 ` [Intel-gfx] [PATCH 6/7] drm/i915/guc: Optimize CTB writes and reads Matthew Brost
2021-07-01 17:15 [Intel-gfx] [PATCH 0/7] CT changes required for GuC submission Matthew Brost
2021-07-01 17:15 ` [Intel-gfx] [PATCH 6/7] drm/i915/guc: Optimize CTB writes and reads Matthew Brost
2021-07-06 19:00   ` John Harrison
2021-07-06 19:12     ` Michal Wajdeczko
2021-07-06 19:19       ` John Harrison
2021-07-06 19:33         ` Michal Wajdeczko
2021-07-06 21:22           ` Matthew Brost
2021-07-06 22:41           ` John Harrison
2021-06-27 23:14 [Intel-gfx] [PATCH 0/7] CT changes required for GuC submission Matthew Brost
2021-06-27 23:14 ` [Intel-gfx] [PATCH 6/7] drm/i915/guc: Optimize CTB writes and reads Matthew Brost

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=9834f3c0-a1a3-9b1a-d291-732501b7a013@intel.com \
    --to=john.c.harrison@intel.com \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=matthew.brost@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox