AMD-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Chunming Zhou <zhoucm1-5C7GfCeVMHo@public.gmane.org>
To: "Christian König"
	<ckoenig.leichtzumerken-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>,
	amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org,
	dri-devel-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org
Subject: Re: [PATCH 8/8] drm/amdgpu: use the new TTM bytes moved counter v2
Date: Mon, 20 Nov 2017 10:38:01 +0800	[thread overview]
Message-ID: <f6551228-2785-9bbd-8d16-eba95b8d23dc@amd.com> (raw)
In-Reply-To: <20171117104935.4150-9-christian.koenig-5C7GfCeVMHo@public.gmane.org>

  Reviewed-by: Chunming Zhou <david1.zhou@amd.com> for the series.


btw: any new patch based on this for enabling eviction and swapout for 
per-vm-bo?

Regards,
David Zhou
On 2017年11月17日 18:49, Christian König wrote:
> Instead of the global statistics use the per context bytes moved counter.
>
> v2: rebased
>
> Signed-off-by: Christian König <christian.koenig@amd.com>
> Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c     |  9 +++------
>   drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 10 +++-------
>   2 files changed, 6 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> index 41994b87c76e..bea5bc64bf7b 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c
> @@ -344,7 +344,6 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
>   {
>   	struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
>   	struct ttm_operation_ctx ctx = { true, false };
> -	u64 initial_bytes_moved, bytes_moved;
>   	uint32_t domain;
>   	int r;
>   
> @@ -374,15 +373,13 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
>   
>   retry:
>   	amdgpu_ttm_placement_from_domain(bo, domain);
> -	initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
>   	r = ttm_bo_validate(&bo->tbo, &bo->placement, &ctx);
> -	bytes_moved = atomic64_read(&adev->num_bytes_moved) -
> -		      initial_bytes_moved;
> -	p->bytes_moved += bytes_moved;
> +
> +	p->bytes_moved += ctx.bytes_moved;
>   	if (adev->mc.visible_vram_size < adev->mc.real_vram_size &&
>   	    bo->tbo.mem.mem_type == TTM_PL_VRAM &&
>   	    bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT)
> -		p->bytes_moved_vis += bytes_moved;
> +		p->bytes_moved_vis += ctx.bytes_moved;
>   
>   	if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
>   		domain = bo->allowed_domains;
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> index 15027f751e07..dc0a8be98043 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> @@ -331,7 +331,6 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
>   	struct amdgpu_bo *bo;
>   	enum ttm_bo_type type;
>   	unsigned long page_align;
> -	u64 initial_bytes_moved, bytes_moved;
>   	size_t acc_size;
>   	int r;
>   
> @@ -406,22 +405,19 @@ static int amdgpu_bo_do_create(struct amdgpu_device *adev,
>   	bo->tbo.bdev = &adev->mman.bdev;
>   	amdgpu_ttm_placement_from_domain(bo, domain);
>   
> -	initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
> -	/* Kernel allocation are uninterruptible */
>   	r = ttm_bo_init_reserved(&adev->mman.bdev, &bo->tbo, size, type,
>   				 &bo->placement, page_align, &ctx, NULL,
>   				 acc_size, sg, resv, &amdgpu_ttm_bo_destroy);
>   	if (unlikely(r != 0))
>   		return r;
>   
> -	bytes_moved = atomic64_read(&adev->num_bytes_moved) -
> -		      initial_bytes_moved;
>   	if (adev->mc.visible_vram_size < adev->mc.real_vram_size &&
>   	    bo->tbo.mem.mem_type == TTM_PL_VRAM &&
>   	    bo->tbo.mem.start < adev->mc.visible_vram_size >> PAGE_SHIFT)
> -		amdgpu_cs_report_moved_bytes(adev, bytes_moved, bytes_moved);
> +		amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved,
> +					     ctx.bytes_moved);
>   	else
> -		amdgpu_cs_report_moved_bytes(adev, bytes_moved, 0);
> +		amdgpu_cs_report_moved_bytes(adev, ctx.bytes_moved, 0);
>   
>   	if (kernel)
>   		bo->tbo.priority = 1;

_______________________________________________
amd-gfx mailing list
amd-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/amd-gfx

  parent reply	other threads:[~2017-11-20  2:38 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-11-17 10:49 Operation context for TTM Christian König
2017-11-17 10:49 ` [PATCH 1/8] drm/ttm: add operation ctx to ttm_bo_validate v2 Christian König
     [not found]   ` <20171117104935.4150-2-christian.koenig-5C7GfCeVMHo@public.gmane.org>
2017-11-20 10:14     ` Daniel Vetter
2017-11-17 10:49 ` [PATCH 2/8] drm/ttm: use an operation ctx for ttm_bo_init_reserved Christian König
2017-11-17 10:49 ` [PATCH 4/8] drm/ttm: use the operation context inside TTM Christian König
2017-11-17 10:49 ` [PATCH 7/8] drm/amdgpu: forward operation context to ttm_bo_mem_space Christian König
     [not found] ` <20171117104935.4150-1-christian.koenig-5C7GfCeVMHo@public.gmane.org>
2017-11-17 10:49   ` [PATCH 3/8] drm/ttm: use an operation context for ttm_bo_mem_space v2 Christian König
2017-11-17 10:49   ` [PATCH 5/8] drm/ttm: add context to driver move callback as well Christian König
2017-11-17 10:49   ` [PATCH 6/8] drm/ttm: add number of bytes moved to the operation context Christian König
2017-11-17 10:49   ` [PATCH 8/8] drm/amdgpu: use the new TTM bytes moved counter v2 Christian König
     [not found]     ` <20171117104935.4150-9-christian.koenig-5C7GfCeVMHo@public.gmane.org>
2017-11-20  2:38       ` Chunming Zhou [this message]
2017-11-17 14:55   ` Operation context for TTM Michel Dänzer
2017-11-17 20:26   ` Felix Kuehling
2017-11-20  4:35   ` Dieter Nützel
     [not found]     ` <cb11aafa8199f35d8ab93f7199c1f1c0-0hun7QTegEsDD4udEopG9Q@public.gmane.org>
2017-11-21  2:49       ` Dieter Nützel

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=f6551228-2785-9bbd-8d16-eba95b8d23dc@amd.com \
    --to=zhoucm1-5c7gfcevmho@public.gmane.org \
    --cc=amd-gfx-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org \
    --cc=ckoenig.leichtzumerken-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
    --cc=dri-devel-PD4FTy7X32lNgt0PjOBp9y5qC8QIuHrW@public.gmane.org \
    /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