[PATCH 4/9] drm/amdgpu: Add a bitmask in amdgpu_ctx_mgr

[Kuehling, Felix]

On 2018-12-06 6:32 a.m., Rex Zhu wrote:
> used to manager the reserverd vm space.
>
> Signed-off-by: Rex Zhu <Rex.Zhu at amd.com>
> ---
>  drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c | 8 ++++++--
>  drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h | 4 +++-
>  drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 6 +++++-
>  3 files changed, 14 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
> index 8edf54b..8802ff2 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.c
> @@ -529,10 +529,14 @@ int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx,
>  	return 0;
>  }
>  
> -void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr)
> +int amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr)
>  {
>  	mutex_init(&mgr->lock);
>  	idr_init(&mgr->ctx_handles);
> +	mgr->resv_vm_bitmap = kzalloc(DIV_ROUND_UP(AMDGPU_VM_MAX_NUM_CTX, BITS_PER_BYTE), GFP_KERNEL);

The bitmap is in (unsigned long) units. So you should round to multiples
of (unsigned long), not multiples of byte. Something like this:

mgr->resv_vm_bitmap = kzalloc(DIV_ROUND_UP(AMDGPU_VM_MAX_NUM_CTX,
BITS_PER_BYTE * sizeof(unsigned long)) * sizeof(unsigned long), GFP_KERNEL);

Regards,
  Felix


> +	if (unlikely(!mgr->resv_vm_bitmap))
> +		return -ENOMEM;
> +	return 0;
>  }
>  
>  void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr)
> @@ -601,7 +605,7 @@ void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr)
>  		if (kref_put(&ctx->refcount, amdgpu_ctx_fini) != 1)
>  			DRM_ERROR("ctx %p is still alive\n", ctx);
>  	}
> -
> +	kfree(mgr->resv_vm_bitmap);
>  	idr_destroy(&mgr->ctx_handles);
>  	mutex_destroy(&mgr->lock);
>  }
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
> index b3b012c..94ac951 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ctx.h
> @@ -38,6 +38,7 @@ struct amdgpu_ctx_entity {
>  struct amdgpu_ctx {
>  	struct kref			refcount;
>  	struct amdgpu_device		*adev;
> +
>  	unsigned			reset_counter;
>  	unsigned			reset_counter_query;
>  	uint32_t			vram_lost_counter;
> @@ -56,6 +57,7 @@ struct amdgpu_ctx_mgr {
>  	struct mutex		lock;
>  	/* protected by lock */
>  	struct idr		ctx_handles;
> +	unsigned long		*resv_vm_bitmap;
>  };
>  
>  extern const unsigned int amdgpu_ctx_num_entities[AMDGPU_HW_IP_NUM];
> @@ -80,7 +82,7 @@ int amdgpu_ctx_ioctl(struct drm_device *dev, void *data,
>  int amdgpu_ctx_wait_prev_fence(struct amdgpu_ctx *ctx,
>  			       struct drm_sched_entity *entity);
>  
> -void amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr);
> +int amdgpu_ctx_mgr_init(struct amdgpu_ctx_mgr *mgr);
>  void amdgpu_ctx_mgr_entity_fini(struct amdgpu_ctx_mgr *mgr);
>  void amdgpu_ctx_mgr_entity_flush(struct amdgpu_ctx_mgr *mgr);
>  void amdgpu_ctx_mgr_fini(struct amdgpu_ctx_mgr *mgr);
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
> index 52e4e90..338a091 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c
> @@ -988,11 +988,15 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
>  	mutex_init(&fpriv->bo_list_lock);
>  	idr_init(&fpriv->bo_list_handles);
>  
> -	amdgpu_ctx_mgr_init(&fpriv->ctx_mgr);
> +	if (amdgpu_ctx_mgr_init(&fpriv->ctx_mgr))
> +		goto error_ctx_mgr;
>  
>  	file_priv->driver_priv = fpriv;
>  	goto out_suspend;
>  
> +error_ctx_mgr:
> +	idr_destroy(&fpriv->bo_list_handles);
> +	mutex_destroy(&fpriv->bo_list_lock);
>  error_vm:
>  	amdgpu_vm_fini(adev, &fpriv->vm);
>