1. Patchew
  2. linux
  3. View series

[PATCH v2] dma-buf/udmabuf: skip redundant cpu sync to fix cacheline EEXIST warning

Mikhail Gavrilov posted 1 patch 1 day, 5 hours ago 
drivers/dma-buf/udmabuf.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
[PATCH v2] dma-buf/udmabuf: skip redundant cpu sync to fix cacheline EEXIST warning
Posted by Mikhail Gavrilov 1 day, 5 hours ago 
When CONFIG_DMA_API_DEBUG_SG is enabled, importing a udmabuf into a DRM
driver (e.g. amdgpu for video playback in GNOME Videos / Showtime)
triggers a spurious warning:

  DMA-API: amdgpu 0000:03:00.0: cacheline tracking EEXIST, \
      overlapping mappings aren't supported
  WARNING: kernel/dma/debug.c:619 at add_dma_entry+0x473/0x5f0

The call chain is:

  amdgpu_cs_ioctl
   -> amdgpu_ttm_backend_bind
    -> dma_buf_map_attachment
     -> [udmabuf] map_udmabuf -> get_sg_table
      -> dma_map_sgtable(dev, sg, direction, 0)  // attrs=0
       -> debug_dma_map_sg -> add_dma_entry -> EEXIST

This happens because udmabuf builds a per-page scatter-gather list via
sg_set_folio().  When begin_cpu_udmabuf() has already created an sg
table mapped for the misc device, and an importer such as amdgpu maps
the same pages for its own device via map_udmabuf(), the DMA debug
infrastructure sees two active mappings whose physical addresses share
cacheline boundaries and warns about the overlap.

The DMA_ATTR_SKIP_CPU_SYNC flag suppresses this check in
add_dma_entry() because it signals that no CPU cache maintenance is
performed at map/unmap time, making the cacheline overlap harmless.

All other major dma-buf exporters already pass this flag:
  - drm_gem_map_dma_buf() passes DMA_ATTR_SKIP_CPU_SYNC
  - amdgpu_dma_buf_map() passes DMA_ATTR_SKIP_CPU_SYNC

The CPU sync at map/unmap time is also redundant for udmabuf:
begin_cpu_udmabuf() and end_cpu_udmabuf() already perform explicit
cache synchronization via dma_sync_sgtable_for_cpu/device() when CPU
access is requested through the dma-buf interface.

Pass DMA_ATTR_SKIP_CPU_SYNC to dma_map_sgtable() and
dma_unmap_sgtable() in udmabuf to suppress the spurious warning and
skip the redundant sync.

Fixes: 284562e1f348 ("udmabuf: implement begin_cpu_access/end_cpu_access hooks")
Cc: stable@vger.kernel.org
Signed-off-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
---

v1 -> v2:
  - Rebased on drm-tip to resolve conflict with folio conversion
    patches. No code change, same two-line fix.

v1: https://lore.kernel.org/all/20260317053653.28888-1-mikhail.v.gavrilov@gmail.com/

 drivers/dma-buf/udmabuf.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c
index 94b26ea706a3..bced421c0d65 100644
--- a/drivers/dma-buf/udmabuf.c
+++ b/drivers/dma-buf/udmabuf.c
@@ -145,7 +145,7 @@ static struct sg_table *get_sg_table(struct device *dev, struct dma_buf *buf,
 	if (ret < 0)
 		goto err_alloc;
 
-	ret = dma_map_sgtable(dev, sg, direction, 0);
+	ret = dma_map_sgtable(dev, sg, direction, DMA_ATTR_SKIP_CPU_SYNC);
 	if (ret < 0)
 		goto err_map;
 	return sg;
@@ -160,7 +160,7 @@ static struct sg_table *get_sg_table(struct device *dev, struct dma_buf *buf,
 static void put_sg_table(struct device *dev, struct sg_table *sg,
 			 enum dma_data_direction direction)
 {
-	dma_unmap_sgtable(dev, sg, direction, 0);
+	dma_unmap_sgtable(dev, sg, direction, DMA_ATTR_SKIP_CPU_SYNC);
 	sg_free_table(sg);
 	kfree(sg);
 }
-- 
2.53.0
RE: [PATCH v2] dma-buf/udmabuf: skip redundant cpu sync to fix cacheline EEXIST warning
Posted by Kasireddy, Vivek 10 hours ago 
> Subject: [PATCH v2] dma-buf/udmabuf: skip redundant cpu sync to fix
> cacheline EEXIST warning
> 
> When CONFIG_DMA_API_DEBUG_SG is enabled, importing a udmabuf
> into a DRM
> driver (e.g. amdgpu for video playback in GNOME Videos / Showtime)
> triggers a spurious warning:
> 
>   DMA-API: amdgpu 0000:03:00.0: cacheline tracking EEXIST, \
>       overlapping mappings aren't supported
>   WARNING: kernel/dma/debug.c:619 at add_dma_entry+0x473/0x5f0
> 
> The call chain is:
> 
>   amdgpu_cs_ioctl
>    -> amdgpu_ttm_backend_bind
>     -> dma_buf_map_attachment
>      -> [udmabuf] map_udmabuf -> get_sg_table
>       -> dma_map_sgtable(dev, sg, direction, 0)  // attrs=0
>        -> debug_dma_map_sg -> add_dma_entry -> EEXIST
> 
> This happens because udmabuf builds a per-page scatter-gather list via
> sg_set_folio().  When begin_cpu_udmabuf() has already created an sg
> table mapped for the misc device, and an importer such as amdgpu
> maps
> the same pages for its own device via map_udmabuf(), the DMA debug
> infrastructure sees two active mappings whose physical addresses share
> cacheline boundaries and warns about the overlap.
> 
> The DMA_ATTR_SKIP_CPU_SYNC flag suppresses this check in
> add_dma_entry() because it signals that no CPU cache maintenance is
> performed at map/unmap time, making the cacheline overlap harmless.
> 
> All other major dma-buf exporters already pass this flag:
>   - drm_gem_map_dma_buf() passes DMA_ATTR_SKIP_CPU_SYNC
>   - amdgpu_dma_buf_map() passes DMA_ATTR_SKIP_CPU_SYNC
> 
> The CPU sync at map/unmap time is also redundant for udmabuf:
> begin_cpu_udmabuf() and end_cpu_udmabuf() already perform explicit
> cache synchronization via dma_sync_sgtable_for_cpu/device() when
> CPU
> access is requested through the dma-buf interface.
> 
> Pass DMA_ATTR_SKIP_CPU_SYNC to dma_map_sgtable() and
> dma_unmap_sgtable() in udmabuf to suppress the spurious warning
> and
> skip the redundant sync.
> 
> Fixes: 284562e1f348 ("udmabuf: implement
> begin_cpu_access/end_cpu_access hooks")
> Cc: stable@vger.kernel.org
> Signed-off-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
> ---
> 
> v1 -> v2:
>   - Rebased on drm-tip to resolve conflict with folio conversion
>     patches. No code change, same two-line fix.
> 
> v1: https://lore.kernel.org/all/20260317053653.28888-1-
> mikhail.v.gavrilov@gmail.com/
> 
>  drivers/dma-buf/udmabuf.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c
> index 94b26ea706a3..bced421c0d65 100644
> --- a/drivers/dma-buf/udmabuf.c
> +++ b/drivers/dma-buf/udmabuf.c
> @@ -145,7 +145,7 @@ static struct sg_table *get_sg_table(struct device
> *dev, struct dma_buf *buf,
>  	if (ret < 0)
>  		goto err_alloc;
> 
> -	ret = dma_map_sgtable(dev, sg, direction, 0);
> +	ret = dma_map_sgtable(dev, sg, direction,
> DMA_ATTR_SKIP_CPU_SYNC);
>  	if (ret < 0)
>  		goto err_map;
>  	return sg;
> @@ -160,7 +160,7 @@ static struct sg_table *get_sg_table(struct device
> *dev, struct dma_buf *buf,
>  static void put_sg_table(struct device *dev, struct sg_table *sg,
>  			 enum dma_data_direction direction)
>  {
> -	dma_unmap_sgtable(dev, sg, direction, 0);
> +	dma_unmap_sgtable(dev, sg, direction,
> DMA_ATTR_SKIP_CPU_SYNC);
Acked-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
Will push this one to drm-misc-next soon.

Thanks,
Vivek
>  	sg_free_table(sg);
>  	kfree(sg);
>  }
> --
> 2.53.0

Patchew 2.1-devel-ea86937

© 2016 - 2026 Red Hat, Inc.