NVMe driver and other applications depend on the data offset to operate correctly. Currently when unaligned data is mapped via SWIOTLB, the data is mapped as slab aligned with the SWIOTLB. When booting with --swiotlb=force option and using NVMe as interface, running mkfs.xfs on Rhel fails because of the unalignment issue. This patch makes sure the mapped data preserves its offset of the orginal address. Tested on latest kernel that this patch fixes the issue.
Signed-off-by: Jianxiong Gao <jx...@google.com> Acked-by: David Rientjes <rient...@google.com> --- kernel/dma/swiotlb.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/kernel/dma/swiotlb.c b/kernel/dma/swiotlb.c index 781b9dca197c..56a35e71b3fd 100644 --- a/kernel/dma/swiotlb.c +++ b/kernel/dma/swiotlb.c @@ -483,6 +483,12 @@ phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr, max_slots = mask + 1 ? ALIGN(mask + 1, 1 << IO_TLB_SHIFT) >> IO_TLB_SHIFT : 1UL << (BITS_PER_LONG - IO_TLB_SHIFT); + + /* + * We need to keep the offset when mapping, so adding the offset + * to the total set we need to allocate in SWIOTLB + */ + alloc_size += offset_in_page(orig_addr); /* * For mappings greater than or equal to a page, we limit the stride @@ -567,6 +573,11 @@ phys_addr_t swiotlb_tbl_map_single(struct device *hwdev, phys_addr_t orig_addr, */ for (i = 0; i < nslots; i++) io_tlb_orig_addr[index+i] = orig_addr + (i << IO_TLB_SHIFT); + /* + * When keeping the offset of the original data, we need to advance + * the tlb_addr by the offset of orig_addr. + */ + tlb_addr += orig_addr & (PAGE_SIZE - 1); if (!(attrs & DMA_ATTR_SKIP_CPU_SYNC) && (dir == DMA_TO_DEVICE || dir == DMA_BIDIRECTIONAL)) swiotlb_bounce(orig_addr, tlb_addr, mapping_size, DMA_TO_DEVICE); -- 2.27.0