On Tue, Mar 17, 2026 at 10:37 AM Mikhail Gavrilov <[email protected]> wrote: > > 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: [email protected] > Signed-off-by: Mikhail Gavrilov <[email protected]> > --- > 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 94b8ecb892bb..9c6f8785a28a 100644 > --- a/drivers/dma-buf/udmabuf.c > +++ b/drivers/dma-buf/udmabuf.c > @@ -162,7 +162,7 @@ static struct sg_table *get_sg_table(struct device *dev, > struct dma_buf *buf, > sg_set_folio(sgl, ubuf->folios[i], PAGE_SIZE, > ubuf->offsets[i]); > > - 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; > @@ -177,7 +177,7 @@ 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 >
Gentle ping on this patch. To summarize the review so far: Vivek: "Looks OK to me" Christian: Acked-by Note: while my separate dma-debug patch [1] would also suppress this specific warning on x86, the udmabuf fix here is still valuable as it removes the redundant CPU sync at map/unmap time and aligns udmabuf with other dma-buf exporters that already pass DMA_ATTR_SKIP_CPU_SYNC. [1] https://lore.kernel.org/all/[email protected]/ Gerd, could you take a look when you have a chance? -- Best Regards, Mike Gavrilov.
