Re: [PATCH 1/2] drm/amdgpu: refine amdgpu_bo_create_kernel_at()

[Lang Yu <Lang.Yu@amd.com>]

On 07/27/ , Christian König wrote:
> Am 27.07.23 um 10:15 schrieb Lang Yu:
> > On 07/27/ , Christian König wrote:
> > > Am 27.07.23 um 09:56 schrieb Lang Yu:
> > > > amdgpu_bo_create_kernel_at() is used to create a physical
> > > > contiguous VRAM BO at the specific offset. It calls
> > > > amdgpu_bo_create_reserved() to create a VRAM BO first,
> > > > then frees its old memory and allocates new memory at
> > > > the specific offset. But amdgpu_bo_create_reserved() would
> > > > fail if requested VRAM BO size is too large(>128MB) on
> > > > APU which usually has a small default VRAM size(512MB).
> > > > 
> > > > That is because VRAM fragmentation and DRM_BUDDY_RANGE_ALLOCATION
> > > > is not natively supported by amdgpu_bo_create().
> > > > 
> > > > The approach is using amdgpu_bo_create_reserved() to
> > > > create a BO in CPU domain first, it will always succeed.
> > > > Then use amdgpu_bo_pin_restricted() to move the BO to
> > > > requested domain and location.
> > > That won't work like that.
> > > 
> > > The amdgpu_bo_create_kernel_at() function is used to take over specific
> > > memory areas from the BIOS and *not* to create a large VRAM BO.
> >    Literally, it is creating a VRAM BO.
> 
> No, it doesn't.
> 
> amdgpu_bo_create_kernel_at() creates a BO for a predefined offset in VRAM.
> 
> E.g. the BIOS says to us you can find the table at offsets 0x12345678000 in
> VRAM and during driver load we create a buffer object for this so that we
> can map or copy it away.
> 
> > 
> > > Allocating the initial dummy in the CPU domain is a good idea to avoid
> > > overlap, but you are messing this up quite a bit here and will probably
> > > break tons of stuff.
> >    I don't see it breaks something.amdgpu_bo_create() doesn't support
> >    DRM_BUDDY_RANGE_ALLOCATION. Actually it works, this approach is just using
> >    DRM_BUDDY_RANGE_ALLOCATION to satify such allocation request.
> 
> The big difference between amdgpu_bo_create_kernel_at() and
> amdgpu_bo_create_kernel() is that the allocated VRAM is not initialized to
> zero.
> 
> E.g. we keep the original content the BIOS has placed there for us. With
> your modifications that content would be overwritten.

Thanks, I got it. But it looks like only ttm_bo_type_device would be
initialized to zero. 

	switch (bo->type) {
	case ttm_bo_type_device:
		if (zero_alloc)
			page_flags |= TTM_TT_FLAG_ZERO_ALLOC;
		break;
	case ttm_bo_type_kernel:
		break;
	case ttm_bo_type_sg:
		page_flags |= TTM_TT_FLAG_EXTERNAL;
		break;
	default:
		pr_err("Illegal buffer object type\n");
		return -EINVAL;
	}

Yes, it would be overwritten. That's not expected.

Regards,
Lang

> Regards,
> Christian.
> 
> > 
> >    Reagrds,
> >    Lang
> > > Regards,
> > > Christian.
> > > 
> > > 
> > > > Signed-off-by: Lang Yu <Lang.Yu@amd.com>
> > > > ---
> > > >    drivers/gpu/drm/amd/amdgpu/amdgpu_object.c | 30 ++++++++--------------
> > > >    drivers/gpu/drm/amd/amdgpu/amdgpu_object.h |  3 ++-
> > > >    drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c    | 15 +++++------
> > > >    drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c   |  2 +-
> > > >    4 files changed, 21 insertions(+), 29 deletions(-)
> > > > 
> > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> > > > index ff73cc11d47e..331daa47a444 100644
> > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.c
> > > > @@ -358,6 +358,7 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
> > > >     * @offset: offset of the BO
> > > >     * @size: size of the BO
> > > >     * @bo_ptr:  used to initialize BOs in structures
> > > > + * @gpu_addr: GPU addr of the pinned BO
> > > >     * @cpu_addr: optional CPU address mapping
> > > >     *
> > > >     * Creates a kernel BO at a specific offset in VRAM.
> > > > @@ -367,42 +368,33 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
> > > >     */
> > > >    int amdgpu_bo_create_kernel_at(struct amdgpu_device *adev,
> > > >    			       uint64_t offset, uint64_t size,
> > > > -			       struct amdgpu_bo **bo_ptr, void **cpu_addr)
> > > > +			       struct amdgpu_bo **bo_ptr,
> > > > +			       u64 *gpu_addr, void **cpu_addr)
> > > >    {
> > > > -	struct ttm_operation_ctx ctx = { false, false };
> > > > -	unsigned int i;
> > > >    	int r;
> > > >    	offset &= PAGE_MASK;
> > > >    	size = ALIGN(size, PAGE_SIZE);
> > > >    	r = amdgpu_bo_create_reserved(adev, size, PAGE_SIZE,
> > > > -				      AMDGPU_GEM_DOMAIN_VRAM, bo_ptr, NULL,
> > > > -				      cpu_addr);
> > > > +				      AMDGPU_GEM_DOMAIN_CPU,
> > > > +				      bo_ptr, NULL, cpu_addr);
> > > >    	if (r)
> > > >    		return r;
> > > >    	if ((*bo_ptr) == NULL)
> > > >    		return 0;
> > > > -	/*
> > > > -	 * Remove the original mem node and create a new one at the request
> > > > -	 * position.
> > > > -	 */
> > > > -	if (cpu_addr)
> > > > -		amdgpu_bo_kunmap(*bo_ptr);
> > > > -
> > > > -	ttm_resource_free(&(*bo_ptr)->tbo, &(*bo_ptr)->tbo.resource);
> > > > +	amdgpu_bo_unpin(*bo_ptr);
> > > > -	for (i = 0; i < (*bo_ptr)->placement.num_placement; ++i) {
> > > > -		(*bo_ptr)->placements[i].fpfn = offset >> PAGE_SHIFT;
> > > > -		(*bo_ptr)->placements[i].lpfn = (offset + size) >> PAGE_SHIFT;
> > > > -	}
> > > > -	r = ttm_bo_mem_space(&(*bo_ptr)->tbo, &(*bo_ptr)->placement,
> > > > -			     &(*bo_ptr)->tbo.resource, &ctx);
> > > > +	r = amdgpu_bo_pin_restricted(*bo_ptr, AMDGPU_GEM_DOMAIN_VRAM,
> > > > +				     offset, offset + size);
> > > >    	if (r)
> > > >    		goto error;
> > > > +	if (gpu_addr)
> > > > +		*gpu_addr = amdgpu_bo_gpu_offset(*bo_ptr);
> > > > +
> > > >    	if (cpu_addr) {
> > > >    		r = amdgpu_bo_kmap(*bo_ptr, cpu_addr);
> > > >    		if (r)
> > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
> > > > index 5d3440d719e4..8f5b5664a1b6 100644
> > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
> > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_object.h
> > > > @@ -315,7 +315,8 @@ int amdgpu_bo_create_kernel(struct amdgpu_device *adev,
> > > >    			    u64 *gpu_addr, void **cpu_addr);
> > > >    int amdgpu_bo_create_kernel_at(struct amdgpu_device *adev,
> > > >    			       uint64_t offset, uint64_t size,
> > > > -			       struct amdgpu_bo **bo_ptr, void **cpu_addr);
> > > > +			       struct amdgpu_bo **bo_ptr,
> > > > +			       u64 *gpu_addr, void **cpu_addr);
> > > >    int amdgpu_bo_create_user(struct amdgpu_device *adev,
> > > >    			  struct amdgpu_bo_param *bp,
> > > >    			  struct amdgpu_bo_user **ubo_ptr);
> > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
> > > > index 7c6dd3de1867..a210c243dac0 100644
> > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
> > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
> > > > @@ -1619,7 +1619,7 @@ static int amdgpu_ttm_fw_reserve_vram_init(struct amdgpu_device *adev)
> > > >    					  adev->mman.fw_vram_usage_start_offset,
> > > >    					  adev->mman.fw_vram_usage_size,
> > > >    					  &adev->mman.fw_vram_usage_reserved_bo,
> > > > -					  &adev->mman.fw_vram_usage_va);
> > > > +					  NULL, &adev->mman.fw_vram_usage_va);
> > > >    }
> > > >    /**
> > > > @@ -1644,7 +1644,7 @@ static int amdgpu_ttm_drv_reserve_vram_init(struct amdgpu_device *adev)
> > > >    					  adev->mman.drv_vram_usage_start_offset,
> > > >    					  adev->mman.drv_vram_usage_size,
> > > >    					  &adev->mman.drv_vram_usage_reserved_bo,
> > > > -					  &adev->mman.drv_vram_usage_va);
> > > > +					  NULL, &adev->mman.drv_vram_usage_va);
> > > >    }
> > > >    /*
> > > > @@ -1729,8 +1729,7 @@ static int amdgpu_ttm_reserve_tmr(struct amdgpu_device *adev)
> > > >    		ret = amdgpu_bo_create_kernel_at(adev,
> > > >    						 ctx->c2p_train_data_offset,
> > > >    						 ctx->train_data_size,
> > > > -						 &ctx->c2p_bo,
> > > > -						 NULL);
> > > > +						 &ctx->c2p_bo, NULL, NULL);
> > > >    		if (ret) {
> > > >    			DRM_ERROR("alloc c2p_bo failed(%d)!\n", ret);
> > > >    			amdgpu_ttm_training_reserve_vram_fini(adev);
> > > > @@ -1742,7 +1741,7 @@ static int amdgpu_ttm_reserve_tmr(struct amdgpu_device *adev)
> > > >    	if (!adev->gmc.is_app_apu) {
> > > >    		ret = amdgpu_bo_create_kernel_at(
> > > >    			adev, adev->gmc.real_vram_size - reserve_size,
> > > > -			reserve_size, &adev->mman.fw_reserved_memory, NULL);
> > > > +			reserve_size, &adev->mman.fw_reserved_memory, NULL, NULL);
> > > >    		if (ret) {
> > > >    			DRM_ERROR("alloc tmr failed(%d)!\n", ret);
> > > >    			amdgpu_bo_free_kernel(&adev->mman.fw_reserved_memory,
> > > > @@ -1885,14 +1884,14 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
> > > >    		r = amdgpu_bo_create_kernel_at(adev, 0,
> > > >    					       adev->mman.stolen_vga_size,
> > > >    					       &adev->mman.stolen_vga_memory,
> > > > -					       NULL);
> > > > +					       NULL, NULL);
> > > >    		if (r)
> > > >    			return r;
> > > >    		r = amdgpu_bo_create_kernel_at(adev, adev->mman.stolen_vga_size,
> > > >    					       adev->mman.stolen_extended_size,
> > > >    					       &adev->mman.stolen_extended_memory,
> > > > -					       NULL);
> > > > +					       NULL, NULL);
> > > >    		if (r)
> > > >    			return r;
> > > > @@ -1901,7 +1900,7 @@ int amdgpu_ttm_init(struct amdgpu_device *adev)
> > > >    					       adev->mman.stolen_reserved_offset,
> > > >    					       adev->mman.stolen_reserved_size,
> > > >    					       &adev->mman.stolen_reserved_memory,
> > > > -					       NULL);
> > > > +					       NULL, NULL);
> > > >    		if (r)
> > > >    			return r;
> > > >    	} else {
> > > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
> > > > index 41aa853a07d2..b93b42b916ce 100644
> > > > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
> > > > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
> > > > @@ -397,7 +397,7 @@ static void amdgpu_virt_ras_reserve_bps(struct amdgpu_device *adev)
> > > >    		 */
> > > >    		if (amdgpu_bo_create_kernel_at(adev, bp << AMDGPU_GPU_PAGE_SHIFT,
> > > >    					       AMDGPU_GPU_PAGE_SIZE,
> > > > -					       &bo, NULL))
> > > > +					       &bo, NULL, NULL))
> > > >    			DRM_DEBUG("RAS WARN: reserve vram for retired page %llx fail\n", bp);
> > > >    		data->bps_bo[i] = bo;
>