AMD-GFX Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Luben Tuikov <luben.tuikov@amd.com>
To: Liu ChengZhe <ChengZhe.Liu@amd.com>, amd-gfx@lists.freedesktop.org
Cc: "Jack Xiao" <Jack.Xiao@amd.com>, "Feifei Xu" <Feifei.Xu@amd.com>,
	"Kevin Wang" <Kevin1.Wang@amd.com>,
	"Deucher Alexander" <Alexander.Deucher@amd.com>,
	"Xiaojie Yuan" <xiaojie.yuan@amd.com>,
	"Christian König" <christian.koenig@amd.com>,
	"Hawking Zhang" <Hawking.Zhang@amd.com>
Subject: Re: [PATCH] drm/amdgpu: Skip some registers config for SRIOV
Date: Fri, 7 Aug 2020 21:29:41 -0400	[thread overview]
Message-ID: <7c1eaca0-8bbf-8d9f-1c2e-36a0d6047c42@amd.com> (raw)
In-Reply-To: <20200807084839.730118-1-ChengZhe.Liu@amd.com>

Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>

On 2020-08-07 04:48, Liu ChengZhe wrote:
> Some registers are not accessible to virtual function setup, so
> skip their initialization when in VF-SRIOV mode.
> 
> v2: move SRIOV VF check into specify functions;
> modify commit description and comment.
> 
> Signed-off-by: Liu ChengZhe <ChengZhe.Liu@amd.com>
> ---
>  drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c | 19 +++++++++++++++++++
>  drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c  | 19 +++++++++++++++++++
>  2 files changed, 38 insertions(+)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c
> index 1f6112b7fa49..80c906a0383f 100644
> --- a/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c
> +++ b/drivers/gpu/drm/amd/amdgpu/gfxhub_v2_1.c
> @@ -182,6 +182,12 @@ static void gfxhub_v2_1_init_cache_regs(struct amdgpu_device *adev)
>  {
>  	uint32_t tmp;
>  
> +	/* These registers are not accessible to VF-SRIOV.
> +	 * The PF will program them instead.
> +	 */
> +	if (amdgpu_sriov_vf(adev))
> +		return;
> +
>  	/* Setup L2 cache */
>  	tmp = RREG32_SOC15(GC, 0, mmGCVM_L2_CNTL);
>  	tmp = REG_SET_FIELD(tmp, GCVM_L2_CNTL, ENABLE_L2_CACHE, 1);
> @@ -237,6 +243,12 @@ static void gfxhub_v2_1_enable_system_domain(struct amdgpu_device *adev)
>  
>  static void gfxhub_v2_1_disable_identity_aperture(struct amdgpu_device *adev)
>  {
> +	/* These registers are not accessible to VF-SRIOV.
> +	 * The PF will program them instead.
> +	 */
> +	if (amdgpu_sriov_vf(adev))
> +		return;
> +
>  	WREG32_SOC15(GC, 0, mmGCVM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_LO32,
>  		     0xFFFFFFFF);
>  	WREG32_SOC15(GC, 0, mmGCVM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_HI32,
> @@ -373,6 +385,13 @@ void gfxhub_v2_1_set_fault_enable_default(struct amdgpu_device *adev,
>  					  bool value)
>  {
>  	u32 tmp;
> +
> +	/* These registers are not accessible to VF-SRIOV.
> +	 * The PF will program them instead.
> +	 */
> +	if (amdgpu_sriov_vf(adev))
> +		return;
> +
>  	tmp = RREG32_SOC15(GC, 0, mmGCVM_L2_PROTECTION_FAULT_CNTL);
>  	tmp = REG_SET_FIELD(tmp, GCVM_L2_PROTECTION_FAULT_CNTL,
>  			    RANGE_PROTECTION_FAULT_ENABLE_DEFAULT, value);
> diff --git a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c
> index d83912901f73..8acb3b625afe 100644
> --- a/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/mmhub_v2_0.c
> @@ -181,6 +181,12 @@ static void mmhub_v2_0_init_cache_regs(struct amdgpu_device *adev)
>  {
>  	uint32_t tmp;
>  
> +	/* These registers are not accessible to VF-SRIOV.
> +	 * The PF will program them instead.
> +	 */
> +	if (amdgpu_sriov_vf(adev))
> +		return;
> +
>  	/* Setup L2 cache */
>  	tmp = RREG32_SOC15(MMHUB, 0, mmMMVM_L2_CNTL);
>  	tmp = REG_SET_FIELD(tmp, MMVM_L2_CNTL, ENABLE_L2_CACHE, 1);
> @@ -236,6 +242,12 @@ static void mmhub_v2_0_enable_system_domain(struct amdgpu_device *adev)
>  
>  static void mmhub_v2_0_disable_identity_aperture(struct amdgpu_device *adev)
>  {
> +	/* These registers are not accessible to VF-SRIOV.
> +	 * The PF will program them instead.
> +	 */
> +	if (amdgpu_sriov_vf(adev))
> +		return;
> +
>  	WREG32_SOC15(MMHUB, 0,
>  		     mmMMVM_L2_CONTEXT1_IDENTITY_APERTURE_LOW_ADDR_LO32,
>  		     0xFFFFFFFF);
> @@ -365,6 +377,13 @@ void mmhub_v2_0_gart_disable(struct amdgpu_device *adev)
>  void mmhub_v2_0_set_fault_enable_default(struct amdgpu_device *adev, bool value)
>  {
>  	u32 tmp;
> +
> +	/* These registers are not accessible to VF-SRIOV.
> +	 * The PF will program them instead.
> +	 */
> +	if (amdgpu_sriov_vf(adev))
> +		return;
> +
>  	tmp = RREG32_SOC15(MMHUB, 0, mmMMVM_L2_PROTECTION_FAULT_CNTL);
>  	tmp = REG_SET_FIELD(tmp, MMVM_L2_PROTECTION_FAULT_CNTL,
>  			    RANGE_PROTECTION_FAULT_ENABLE_DEFAULT, value);
> 

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

  reply	other threads:[~2020-08-08  1:29 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-08-07  8:48 [PATCH] drm/amdgpu: Skip some registers config for SRIOV Liu ChengZhe
2020-08-08  1:29 ` Luben Tuikov [this message]
  -- strict thread matches above, loose matches on Subject: below --
2020-08-06  7:40 Liu ChengZhe
2020-08-06 19:46 ` Luben Tuikov

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=7c1eaca0-8bbf-8d9f-1c2e-36a0d6047c42@amd.com \
    --to=luben.tuikov@amd.com \
    --cc=Alexander.Deucher@amd.com \
    --cc=ChengZhe.Liu@amd.com \
    --cc=Feifei.Xu@amd.com \
    --cc=Hawking.Zhang@amd.com \
    --cc=Jack.Xiao@amd.com \
    --cc=Kevin1.Wang@amd.com \
    --cc=amd-gfx@lists.freedesktop.org \
    --cc=christian.koenig@amd.com \
    --cc=xiaojie.yuan@amd.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