Re: Patch "drm/amdkfd: Move dma unmapping after TLB flush" has been added to the 6.6-stable tree

Re: Patch "drm/amdkfd: Move dma unmapping after TLB flush" has been added to the 6.6-stable tree

[Felix Kuehling]

This patch introduced a regression. If you want to backport it, I'd recommend including this fix as well:

commit 9c29282ecbeeb1b43fced3055c6a5bb244b9390b
Author: Lang Yu <Lang.Yu@amd.com>
Date:   Thu Jan 11 12:27:07 2024 +0800

    drm/amdkfd: reserve the BO before validating it
    
    Fix a warning.
    
    v2: Avoid unmapping attachment repeatedly when ERESTARTSYS.
    
    v3: Lock the BO before accessing ttm->sg to avoid race conditions.(Felix)
    
    [   41.708711] WARNING: CPU: 0 PID: 1463 at drivers/gpu/drm/ttm/ttm_bo.c:846 ttm_bo_validate+0x146/0x1b0 [ttm]
    [   41.708989] Call Trace:
    [   41.708992]  <TASK>
    [   41.708996]  ? show_regs+0x6c/0x80
    [   41.709000]  ? ttm_bo_validate+0x146/0x1b0 [ttm]
    [   41.709008]  ? __warn+0x93/0x190
    [   41.709014]  ? ttm_bo_validate+0x146/0x1b0 [ttm]
    [   41.709024]  ? report_bug+0x1f9/0x210
    [   41.709035]  ? handle_bug+0x46/0x80
    [   41.709041]  ? exc_invalid_op+0x1d/0x80
    [   41.709048]  ? asm_exc_invalid_op+0x1f/0x30
    [   41.709057]  ? amdgpu_amdkfd_gpuvm_dmaunmap_mem+0x2c/0x80 [amdgpu]
    [   41.709185]  ? ttm_bo_validate+0x146/0x1b0 [ttm]
    [   41.709197]  ? amdgpu_amdkfd_gpuvm_dmaunmap_mem+0x2c/0x80 [amdgpu]
    [   41.709337]  ? srso_alias_return_thunk+0x5/0x7f
    [   41.709346]  kfd_mem_dmaunmap_attachment+0x9e/0x1e0 [amdgpu]
    [   41.709467]  amdgpu_amdkfd_gpuvm_dmaunmap_mem+0x56/0x80 [amdgpu]
    [   41.709586]  kfd_ioctl_unmap_memory_from_gpu+0x1b7/0x300 [amdgpu]
    [   41.709710]  kfd_ioctl+0x1ec/0x650 [amdgpu]
    [   41.709822]  ? __pfx_kfd_ioctl_unmap_memory_from_gpu+0x10/0x10 [amdgpu]
    [   41.709945]  ? srso_alias_return_thunk+0x5/0x7f
    [   41.709949]  ? tomoyo_file_ioctl+0x20/0x30
    [   41.709959]  __x64_sys_ioctl+0x9c/0xd0
    [   41.709967]  do_syscall_64+0x3f/0x90
    [   41.709973]  entry_SYSCALL_64_after_hwframe+0x6e/0xd8
    
    Fixes: 101b8104307e ("drm/amdkfd: Move dma unmapping after TLB flush")
    Signed-off-by: Lang Yu <Lang.Yu@amd.com>
    Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
    Signed-off-by: Alex Deucher <alexander.deucher@amd.com>

Regards,
  Felix

On 2024-08-20 8:00, Sasha Levin wrote:
> This is a note to let you know that I've just added the patch titled
> 
>     drm/amdkfd: Move dma unmapping after TLB flush
> 
> to the 6.6-stable tree which can be found at:
>     http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
> 
> The filename of the patch is:
>      drm-amdkfd-move-dma-unmapping-after-tlb-flush.patch
> and it can be found in the queue-6.6 subdirectory.
> 
> If you, or anyone else, feels it should not be added to the stable tree,
> please let <stable@vger.kernel.org> know about it.
> 
> 
> 
> commit 23f8ef0f6e5deee5814fda6ec2e2ee4c2f19a384
> Author: Philip Yang <Philip.Yang@amd.com>
> Date:   Mon Sep 11 14:44:22 2023 -0400
> 
>     drm/amdkfd: Move dma unmapping after TLB flush
>     
>     [ Upstream commit 101b8104307eac734f2dfa4d3511430b0b631c73 ]
>     
>     Otherwise GPU may access the stale mapping and generate IOMMU
>     IO_PAGE_FAULT.
>     
>     Move this to inside p->mutex to prevent multiple threads mapping and
>     unmapping concurrently race condition.
>     
>     After kfd_mem_dmaunmap_attachment is removed from unmap_bo_from_gpuvm,
>     kfd_mem_dmaunmap_attachment is called if failed to map to GPUs, and
>     before free the mem attachment in case failed to unmap from GPUs.
>     
>     Signed-off-by: Philip Yang <Philip.Yang@amd.com>
>     Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
>     Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
>     Signed-off-by: Sasha Levin <sashal@kernel.org>
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
> index 2fe9860725bd9..5e4fb33b97351 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h
> @@ -303,6 +303,7 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(struct amdgpu_device *adev,
>  					  struct kgd_mem *mem, void *drm_priv);
>  int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu(
>  		struct amdgpu_device *adev, struct kgd_mem *mem, void *drm_priv);
> +void amdgpu_amdkfd_gpuvm_dmaunmap_mem(struct kgd_mem *mem, void *drm_priv);
>  int amdgpu_amdkfd_gpuvm_sync_memory(
>  		struct amdgpu_device *adev, struct kgd_mem *mem, bool intr);
>  int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_mem *mem,
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
> index 62c1dc9510a41..c2d1d57a6c668 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c
> @@ -733,7 +733,7 @@ kfd_mem_dmaunmap_sg_bo(struct kgd_mem *mem,
>  	enum dma_data_direction dir;
>  
>  	if (unlikely(!ttm->sg)) {
> -		pr_err("SG Table of BO is UNEXPECTEDLY NULL");
> +		pr_debug("SG Table of BO is NULL");
>  		return;
>  	}
>  
> @@ -1202,8 +1202,6 @@ static void unmap_bo_from_gpuvm(struct kgd_mem *mem,
>  	amdgpu_vm_clear_freed(adev, vm, &bo_va->last_pt_update);
>  
>  	amdgpu_sync_fence(sync, bo_va->last_pt_update);
> -
> -	kfd_mem_dmaunmap_attachment(mem, entry);
>  }
>  
>  static int update_gpuvm_pte(struct kgd_mem *mem,
> @@ -1258,6 +1256,7 @@ static int map_bo_to_gpuvm(struct kgd_mem *mem,
>  
>  update_gpuvm_pte_failed:
>  	unmap_bo_from_gpuvm(mem, entry, sync);
> +	kfd_mem_dmaunmap_attachment(mem, entry);
>  	return ret;
>  }
>  
> @@ -1862,8 +1861,10 @@ int amdgpu_amdkfd_gpuvm_free_memory_of_gpu(
>  		mem->va + bo_size * (1 + mem->aql_queue));
>  
>  	/* Remove from VM internal data structures */
> -	list_for_each_entry_safe(entry, tmp, &mem->attachments, list)
> +	list_for_each_entry_safe(entry, tmp, &mem->attachments, list) {
> +		kfd_mem_dmaunmap_attachment(mem, entry);
>  		kfd_mem_detach(entry);
> +	}
>  
>  	ret = unreserve_bo_and_vms(&ctx, false, false);
>  
> @@ -2037,6 +2038,23 @@ int amdgpu_amdkfd_gpuvm_map_memory_to_gpu(
>  	return ret;
>  }
>  
> +void amdgpu_amdkfd_gpuvm_dmaunmap_mem(struct kgd_mem *mem, void *drm_priv)
> +{
> +	struct kfd_mem_attachment *entry;
> +	struct amdgpu_vm *vm;
> +
> +	vm = drm_priv_to_vm(drm_priv);
> +
> +	mutex_lock(&mem->lock);
> +
> +	list_for_each_entry(entry, &mem->attachments, list) {
> +		if (entry->bo_va->base.vm == vm)
> +			kfd_mem_dmaunmap_attachment(mem, entry);
> +	}
> +
> +	mutex_unlock(&mem->lock);
> +}
> +
>  int amdgpu_amdkfd_gpuvm_unmap_memory_from_gpu(
>  		struct amdgpu_device *adev, struct kgd_mem *mem, void *drm_priv)
>  {
> diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
> index d33ba4fe9ad5b..045280c2b607c 100644
> --- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
> +++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
> @@ -1432,17 +1432,21 @@ static int kfd_ioctl_unmap_memory_from_gpu(struct file *filep,
>  			goto sync_memory_failed;
>  		}
>  	}
> -	mutex_unlock(&p->mutex);
>  
> -	if (flush_tlb) {
> -		/* Flush TLBs after waiting for the page table updates to complete */
> -		for (i = 0; i < args->n_devices; i++) {
> -			peer_pdd = kfd_process_device_data_by_id(p, devices_arr[i]);
> -			if (WARN_ON_ONCE(!peer_pdd))
> -				continue;
> +	/* Flush TLBs after waiting for the page table updates to complete */
> +	for (i = 0; i < args->n_devices; i++) {
> +		peer_pdd = kfd_process_device_data_by_id(p, devices_arr[i]);
> +		if (WARN_ON_ONCE(!peer_pdd))
> +			continue;
> +		if (flush_tlb)
>  			kfd_flush_tlb(peer_pdd, TLB_FLUSH_HEAVYWEIGHT);
> -		}
> +
> +		/* Remove dma mapping after tlb flush to avoid IO_PAGE_FAULT */
> +		amdgpu_amdkfd_gpuvm_dmaunmap_mem(mem, peer_pdd->drm_priv);
>  	}
> +
> +	mutex_unlock(&p->mutex);
> +
>  	kfree(devices_arr);
>  
>  	return 0;

Re: Patch "drm/amdkfd: Move dma unmapping after TLB flush" has been added to the 6.6-stable tree

[Greg KH]

On Tue, Aug 20, 2024 at 11:58:30AM -0400, Felix Kuehling wrote:
> This patch introduced a regression. If you want to backport it, I'd recommend including this fix as well:
> 
> commit 9c29282ecbeeb1b43fced3055c6a5bb244b9390b

Wrong git id :(

Or right one, but the amd graphics tree is so messed up it's hard to
tell, there's a different commit with the same content but different id.

{sigh}

greg k-h