x86/dma/amd_gart: Look at dev->coherent_dma_mask instead of GFP_DMA

We want to phase out looking at the magic GFP_DMA flag in the DMA mapping
routines, so switch the gart driver to use the dev->coherent_dma_mask
instead, which is used to select the GFP_DMA flag in the caller.

Tested-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Joerg Roedel <joro@8bytes.org>
Cc: Jon Mason <jdmason@kudzu.us>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Muli Ben-Yehuda <mulix@mulix.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: iommu@lists.linux-foundation.org
Link: http://lkml.kernel.org/r/20180319103826.12853-6-hch@lst.de
Signed-off-by: Ingo Molnar <mingo@kernel.org>

diff --git a/arch/x86/kernel/amd_gart_64.c b/arch/x86/kernel/amd_gart_64.c
index 52e3abc..79ac6ca 100644 (file)
--- a/arch/x86/kernel/amd_gart_64.c
+++ b/arch/x86/kernel/amd_gart_64.c
@@ -484,7 +484,7 @@ gart_alloc_coherent(struct device *dev, size_t size, dma_addr_t *dma_addr,
        unsigned long align_mask;
        struct page *page;
 
-       if (force_iommu && !(flag & GFP_DMA)) {
+       if (force_iommu && dev->coherent_dma_mask > DMA_BIT_MASK(24)) {
                flag &= ~(__GFP_DMA | __GFP_HIGHMEM | __GFP_DMA32);
                page = alloc_pages(flag | __GFP_ZERO, get_order(size));
                if (!page)