virtio_blk: set the zone write granularitygrafted

virtblk_read_zoned_limits() reads the write granularity that the device
reports in virtio_blk_zoned_characteristics and assigns it to the
physical block size and to io_min, but never to the limit that is named
after it. queue_limits.zone_write_granularity is left at zero, so
blk_validate_zoned_limits() raises it to the logical block size:

	if (lim->zone_write_granularity < lim->logical_block_size)
		lim->zone_write_granularity = lim->logical_block_size;

A device that reports a granularity coarser than its logical block size,
which is what the field exists to express, therefore has it silently
reduced. A 512e host managed disk passed through to a guest reports a
logical block size of 512 and a write granularity of 4096, and the guest
ends up with a zone write granularity of 512.

bio_split_alignment() returns lim->zone_write_granularity if it is non-zero
and bio_split_io_at() may split a bio with as per bio_split_alignment().
This can real to the write getting rejected by the host drive, as the write
is not aligned to the physical block size.

zonefs also takes its block size from bdev_zone_write_granularity(), so it
would incorrectly use 512 on a disk that requires 4096.

sd_zbc_read_zones() sets the limit from the physical block size for the
same reason. NVMe ZNS and null_blk leave it unset, but the fallback
gives the right answer for them, as their write granularity is the
logical block size. virtio carries a separate value that may exceed it.

Set the zone write granularity from the value that the device reports.

Fixes: 95bfec41bd3d ("virtio-blk: add support for zoned block devices")
Signed-off-by: Niklas Cassel <cassel@kernel.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Link: https://patch.msgid.link/20260918140641.2031075-2-cassel@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>

1 files changed, 99 insertions, 0 deletions

diff --git a/fs/fuse/req.c b/fs/fuse/req.c
new file mode 100644
index 000000000..a01ee743d
--- /dev/null
+++ b/fs/fuse/req.c
@@ -0,0 +1,99 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#include "dev.h"
+#include "fuse_i.h"
+
+static int fuse_fill_creds(struct fuse_mount *fm, struct fuse_args *args, struct mnt_idmap *idmap)
+{
+	struct fuse_conn *fc = fm->fc;
+	bool no_idmap = !fm->sb || (fm->sb->s_iflags & SB_I_NOIDMAP);
+	kuid_t fsuid = mapped_fsuid(idmap, fc->user_ns);
+	kgid_t fsgid = mapped_fsgid(idmap, fc->user_ns);
+
+	args->pid = pid_nr_ns(task_pid(current), fc->pid_ns);
+
+	if (args->force) {
+		if (args->nocreds)
+			return 0;
+
+		if (no_idmap) {
+			args->uid = from_kuid_munged(fc->user_ns, current_fsuid());
+			args->gid = from_kgid_munged(fc->user_ns, current_fsgid());
+		} else {
+			args->uid = FUSE_INVALID_UIDGID;
+			args->gid = FUSE_INVALID_UIDGID;
+		}
+		return 0;
+	}
+
+	WARN_ON(args->nocreds);
+	/*
+	 * Keep the old behavior when idmappings support was not
+	 * declared by a FUSE server.
+	 *
+	 * For those FUSE servers who support idmapped mounts, we send UID/GID
+	 * only along with "inode creation" fuse requests, otherwise idmap ==
+	 * &invalid_mnt_idmap and req->in.h.{u,g}id will be equal to
+	 * FUSE_INVALID_UIDGID.
+	 */
+	if (no_idmap) {
+		fsuid = current_fsuid();
+		fsgid = current_fsgid();
+	}
+	args->uid = from_kuid(fc->user_ns, fsuid);
+	args->gid = from_kgid(fc->user_ns, fsgid);
+
+	if (no_idmap && unlikely(args->uid == ((uid_t)-1) || args->gid == ((gid_t)-1)))
+		return -EOVERFLOW;
+
+	return 0;
+}
+
+static int fuse_req_prep(struct fuse_mount *fm, struct fuse_args *args, struct mnt_idmap *idmap)
+{
+	if (!args->force && fm->fc->conn_error)
+		return -ECONNREFUSED;
+
+	return fuse_fill_creds(fm, args, idmap);
+}
+
+ssize_t __fuse_simple_request(struct mnt_idmap *idmap, struct fuse_mount *fm,
+			      struct fuse_args *args)
+{
+	struct fuse_conn *fc = fm->fc;
+	int err = fuse_req_prep(fm, args, idmap);
+
+	if (err)
+		return err;
+
+	return fuse_chan_send(fc->chan, args);
+}
+
+int fuse_simple_background(struct fuse_mount *fm, struct fuse_args *args, gfp_t gfp_flags)
+{
+	struct fuse_conn *fc = fm->fc;
+	int err;
+
+	WARN_ON(args->force && !args->nocreds);
+
+	err = fuse_req_prep(fm, args, &invalid_mnt_idmap);
+	if (err)
+		return err;
+
+	return fuse_chan_send_bg(fc->chan, args, gfp_flags);
+}
+EXPORT_SYMBOL_GPL(fuse_simple_background);
+
+int fuse_simple_notify_reply(struct fuse_mount *fm, struct fuse_args *args, u64 unique)
+{
+	struct fuse_conn *fc = fm->fc;
+	int err;
+
+	WARN_ON(args->force && !args->nocreds);
+
+	err = fuse_req_prep(fm, args, &invalid_mnt_idmap);
+	if (err)
+		return err;
+
+	return fuse_chan_send_notify_reply(fc->chan, args, unique);
+}