On Tue, Sep 22, 2026 at 11:43:51AM -0400, Stefan Hajnoczi wrote:
> Linux v6.1 commit 825cf206ed51 ("statx: add direct I/O alignment
> information") added an interface to fetch O_DIRECT alignment values for
> block devices and file systems.
>
> Prefer STATX_DIOALIGN to older interfaces and probing, but keep them as
> fallbacks in case STATX_DIOALIGN cannot provide the information.
>
> STATX_DIOALIGN support as of Linux 6.1:
> - btrfs: no
> - ext4: yes
> - XFS: yes
> - block devices: yes
> - dm-crypt: yes
>
> This change fixes #1290 ("IO alignment probing delivers incorrect
> results on Linux when used with e.g. dm-crypt") on Linux host kernel
> version 6.1 or later by avoiding the problematic probing code in
> raw_probe_alignment(). It would be nice to fix the probing code but I
> have not been able to reproduce it. Stable/LTS distros ship Linux 6.1+,
> so the number of users without STATX_DIOALIGN is becoming small.
>
> Resolves: https://gitlab.com/qemu-project/qemu/-/work_items/1290
> Cc: Eric Biggers <[email protected]>
> Signed-off-by: Stefan Hajnoczi <[email protected]>
> ---
> v3:
> - Rebase and update commit description
> v2:
> - Make sure that XFS_IOC_DIOINFO takes priority over logical blocksize [Eric
> Biggers]
> ---
> block/file-posix.c | 56 +++++++++++++++++++++++++++++++---------------
> 1 file changed, 38 insertions(+), 18 deletions(-)
Ping
>
> diff --git a/block/file-posix.c b/block/file-posix.c
> index 9aad156aad4..0dedd090068 100644
> --- a/block/file-posix.c
> +++ b/block/file-posix.c
> @@ -426,28 +426,48 @@ static void raw_probe_alignment(BlockDriverState *bs,
> int fd, Error **errp)
>
> bs->bl.request_alignment = 0;
> s->buf_align = 0;
> - /* Let's try to use the logical blocksize for the alignment. */
> - if (probe_logical_blocksize(fd, &bs->bl.request_alignment) < 0) {
> - bs->bl.request_alignment = 0;
> +
> +#if defined(__linux__) && defined(STATX_DIOALIGN)
> + struct statx stx;
> +
> + /*
> + * Linux 6.1 introduced an interface for both block devices and file
> + * systems. The system call returns with the STATX_DIOALIGN bit cleared
> + * when the information is unavailable.
> + */
> + if (statx(fd, "", AT_EMPTY_PATH, STATX_DIOALIGN, &stx) == 0 &&
> + (stx.stx_mask & STATX_DIOALIGN)) {
> + bs->bl.request_alignment = stx.stx_dio_offset_align;
> + s->buf_align = stx.stx_dio_mem_align;
> }
> +#endif /* defined(__linux__) && defined(STATX_DIOALIGN) */
>
> #ifdef __linux__
> - /*
> - * The XFS ioctl definitions are shipped in extra packages that might
> - * not always be available. Since we just need the XFS_IOC_DIOINFO ioctl
> - * here, we simply use our own definition instead:
> - */
> - struct xfs_dioattr {
> - uint32_t d_mem;
> - uint32_t d_miniosz;
> - uint32_t d_maxiosz;
> - } da;
> - if (ioctl(fd, _IOR('X', 30, struct xfs_dioattr), &da) >= 0) {
> - bs->bl.request_alignment = da.d_miniosz;
> - /* The kernel returns wrong information for d_mem */
> - /* s->buf_align = da.d_mem; */
> + if (!bs->bl.request_alignment) {
> + /*
> + * The XFS ioctl definitions are shipped in extra packages that might
> + * not always be available. Since we just need the XFS_IOC_DIOINFO
> ioctl
> + * here, we simply use our own definition instead:
> + */
> + struct xfs_dioattr {
> + uint32_t d_mem;
> + uint32_t d_miniosz;
> + uint32_t d_maxiosz;
> + } da;
> + if (ioctl(fd, _IOR('X', 30, struct xfs_dioattr), &da) >= 0) {
> + bs->bl.request_alignment = da.d_miniosz;
> + /* The kernel returns wrong information for d_mem */
> + /* s->buf_align = da.d_mem; */
> + }
> + }
> +#endif /* __linux__ */
> +
> + /* Let's try to use the logical blocksize for the alignment. */
> + if (!bs->bl.request_alignment) {
> + if (probe_logical_blocksize(fd, &bs->bl.request_alignment) < 0) {
> + bs->bl.request_alignment = 0;
> + }
> }
> -#endif
>
> /*
> * If we could not get the sizes so far, we can only guess them. First
> try
> --
> 2.55.0
>
