From: Jason Gunthorpe Date: Mon, 23 Jan 2023 20:35:57 +0000 (-0400) Subject: iommu/dma: Use the gfp parameter in __iommu_dma_alloc_noncontiguous() X-Git-Tag: v6.6.7~3431^2~4^2~6 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=96d57808808595d33f98cef5b3c0f75dde6a72f2;p=platform%2Fkernel%2Flinux-starfive.git iommu/dma: Use the gfp parameter in __iommu_dma_alloc_noncontiguous() This function does an allocation of a buffer to return to the caller and then goes on to allocate some internal memory, eg the scatterlist and IOPTEs. Instead of hard wiring GFP_KERNEL and a wrong GFP_ATOMIC, continue to use the passed in gfp flags for all of the allocations. Clear the zone and policy bits that are only relevant for the buffer allocation before re-using them for internal allocations. Auditing says this is never called from an atomic context, so the GFP_ATOMIC is the incorrect flag. Reviewed-by: Kevin Tian Signed-off-by: Jason Gunthorpe Link: https://lore.kernel.org/r/4-v3-76b587fe28df+6e3-iommu_map_gfp_jgg@nvidia.com Signed-off-by: Joerg Roedel --- diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c index 72cfa24..c99e4bc 100644 --- a/drivers/iommu/dma-iommu.c +++ b/drivers/iommu/dma-iommu.c @@ -822,7 +822,14 @@ static struct page **__iommu_dma_alloc_noncontiguous(struct device *dev, if (!iova) goto out_free_pages; - if (sg_alloc_table_from_pages(sgt, pages, count, 0, size, GFP_KERNEL)) + /* + * Remove the zone/policy flags from the GFP - these are applied to the + * __iommu_dma_alloc_pages() but are not used for the supporting + * internal allocations that follow. + */ + gfp &= ~(__GFP_DMA | __GFP_DMA32 | __GFP_HIGHMEM | __GFP_COMP); + + if (sg_alloc_table_from_pages(sgt, pages, count, 0, size, gfp)) goto out_free_iova; if (!(ioprot & IOMMU_CACHE)) { @@ -834,7 +841,7 @@ static struct page **__iommu_dma_alloc_noncontiguous(struct device *dev, } ret = iommu_map_sg(domain, iova, sgt->sgl, sgt->orig_nents, ioprot, - GFP_ATOMIC); + gfp); if (ret < 0 || ret < size) goto out_free_sg;