[PATCH] loop, zloop: fix dma_alignment for large or unreported limits

[PATCH] loop, zloop: fix dma_alignment for large or unreported limits

[Keith Busch]

From: Keith Busch <kbusch@kernel.org>

A file system sets STATX_DIOALIGN with zeroed alignments when the file
can't be used for direct I/O. The zero underflowed to UINT_MAX and
triggered a queue limits validation warning. Fall back to the block
device's limits when dio_mem_align isn't reported.

A file system with a block size larger than PAGE_SIZE may also report a
memory alignment that can't be expressed as a queue limit. File systems
fall back to buffered I/O for requests that don't meet their alignment,
so cap the reported limit to the largest possible value.

Fixes: 6c8dec275ccc ("loop: set dma_alignment from the backing file for direct I/O")
Fixes: c5059c1af2bd ("zloop: set dma_alignment from the backing files for direct I/O")
Reported-by: syzbot+ac00e7bf7ac8c91af921@syzkaller.appspotmail.com
Closes: https://syzkaller.appspot.com/bug?extid=ac00e7bf7ac8c91af921
Signed-off-by: Keith Busch <kbusch@kernel.org>
---
 drivers/block/loop.c  | 8 +++++---
 drivers/block/zloop.c | 8 +++++---
 2 files changed, 10 insertions(+), 6 deletions(-)

diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index 8639fa34b8470..cbf1e2ce27126 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c
@@ -458,12 +458,14 @@ static void loop_update_dio_alignment(struct loop_device *lo)
 	 * Use the dio alignment of the file system if provided.  The incomoing
 	 * request's bio_vec is forwarded to the backing file unchanged, so its
 	 * required memory alignment becomes the device's dma_alignment when
-	 * used for direct-io.
+	 * used for direct-io.  The file system reports zeroed alignments if the
+	 * file can't be used for direct-io at all, so fall back to the block
+	 * device limits in that case.
 	 */
 	if (!vfs_getattr(&file->f_path, &st, STATX_DIOALIGN, 0) &&
-	    (st.result_mask & STATX_DIOALIGN)) {
+	    (st.result_mask & STATX_DIOALIGN) && st.dio_mem_align) {
 		lo->lo_min_dio_size = st.dio_offset_align;
-		lo->lo_dio_mem_align = st.dio_mem_align - 1;
+		lo->lo_dio_mem_align = min(st.dio_mem_align - 1, PAGE_SIZE - 1);
 		return;
 	}
 
diff --git a/drivers/block/zloop.c b/drivers/block/zloop.c
index 4323ac108cae8..f0ca221524db6 100644
--- a/drivers/block/zloop.c
+++ b/drivers/block/zloop.c
@@ -1042,12 +1042,14 @@ static int zloop_get_block_size(struct zloop_device *zlo,
 	 * Use the dio alignment of the file system if provided.  The incoming
 	 * request's bio_vec is forwarded to the backing file unchanged, so its
 	 * required memory alignment becomes the device's dma_alignment when
-	 * used for direct-io.
+	 * used for direct-io.  The file system reports zeroed alignments if the
+	 * file can't be used for direct-io at all, so fall back to the block
+	 * device limits in that case.
 	 */
 	if (!vfs_getattr(&zone->file->f_path, &st, STATX_DIOALIGN, 0) &&
-	    (st.result_mask & STATX_DIOALIGN)) {
+	    (st.result_mask & STATX_DIOALIGN) && st.dio_mem_align) {
 		zlo->block_size = st.dio_offset_align;
-		zlo->dio_mem_align = st.dio_mem_align - 1;
+		zlo->dio_mem_align = min(st.dio_mem_align - 1, PAGE_SIZE - 1);
 	} else if (sb_bdev) {
 		zlo->block_size = bdev_physical_block_size(sb_bdev);
 		zlo->dio_mem_align = bdev_dma_alignment(sb_bdev);
-- 
2.53.0-Meta

Re: [PATCH] loop, zloop: fix dma_alignment for large or unreported limits

[Christoph Hellwig]

On Mon, Aug 10, 2026 at 09:42:17AM -0700, Keith Busch wrote:
> From: Keith Busch <kbusch@kernel.org>
> 
> A file system sets STATX_DIOALIGN with zeroed alignments when the file
> can't be used for direct I/O. The zero underflowed to UINT_MAX and
> triggered a queue limits validation warning. Fall back to the block
> device's limits when dio_mem_align isn't reported.
> 
> A file system with a block size larger than PAGE_SIZE may also report a
> memory alignment that can't be expressed as a queue limit. File systems
> fall back to buffered I/O for requests that don't meet their alignment,
> so cap the reported limit to the largest possible value.

In general having separate patches per driver would seem nicer than a
batched up one like this, but the changes do look good:

Reviewed-by: Christoph Hellwig <hch@lst.de>