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/nilfs2/alloc.h b/fs/nilfs2/alloc.h
new file mode 100644
index 000000000..046d876ea
--- /dev/null
+++ b/fs/nilfs2/alloc.h
@@ -0,0 +1,99 @@
+/* SPDX-License-Identifier: GPL-2.0+ */
+/*
+ * Persistent object (dat entry/disk inode) allocator/deallocator
+ *
+ * Copyright (C) 2006-2008 Nippon Telegraph and Telephone Corporation.
+ *
+ * Originally written by Koji Sato.
+ * Two allocators were unified by Ryusuke Konishi and Amagai Yoshiji.
+ */
+
+#ifndef _NILFS_ALLOC_H
+#define _NILFS_ALLOC_H
+
+#include <linux/types.h>
+#include <linux/buffer_head.h>
+#include <linux/fs.h>
+
+/**
+ * nilfs_palloc_entries_per_group - get the number of entries per group
+ * @inode: inode of metadata file using this allocator
+ *
+ * The number of entries per group is defined by the number of bits
+ * that a bitmap block can maintain.
+ *
+ * Return: Number of entries per group.
+ */
+static inline unsigned long
+nilfs_palloc_entries_per_group(const struct inode *inode)
+{
+	return 1UL << (inode->i_blkbits + 3 /* log2(8 = CHAR_BITS) */);
+}
+
+int nilfs_palloc_init_blockgroup(struct inode *, unsigned int);
+int nilfs_palloc_get_entry_block(struct inode *, __u64, int,
+				 struct buffer_head **);
+size_t nilfs_palloc_entry_offset(const struct inode *inode, __u64 nr,
+				 const struct buffer_head *bh);
+
+int nilfs_palloc_count_max_entries(struct inode *, u64, u64 *);
+
+/**
+ * struct nilfs_palloc_req - persistent allocator request and reply
+ * @pr_entry_nr: entry number (vblocknr or inode number)
+ * @pr_desc_bh: buffer head of the buffer containing block group descriptors
+ * @pr_bitmap_bh: buffer head of the buffer containing a block group bitmap
+ * @pr_entry_bh: buffer head of the buffer containing translation entries
+ */
+struct nilfs_palloc_req {
+	__u64 pr_entry_nr;
+	struct buffer_head *pr_desc_bh;
+	struct buffer_head *pr_bitmap_bh;
+	struct buffer_head *pr_entry_bh;
+};
+
+int nilfs_palloc_prepare_alloc_entry(struct inode *inode,
+				     struct nilfs_palloc_req *req, bool wrap);
+void nilfs_palloc_commit_alloc_entry(struct inode *,
+				     struct nilfs_palloc_req *);
+void nilfs_palloc_abort_alloc_entry(struct inode *, struct nilfs_palloc_req *);
+void nilfs_palloc_commit_free_entry(struct inode *, struct nilfs_palloc_req *);
+int nilfs_palloc_prepare_free_entry(struct inode *, struct nilfs_palloc_req *);
+void nilfs_palloc_abort_free_entry(struct inode *, struct nilfs_palloc_req *);
+int nilfs_palloc_freev(struct inode *, __u64 *, size_t);
+
+#define nilfs_set_bit_atomic		ext2_set_bit_atomic
+#define nilfs_clear_bit_atomic		ext2_clear_bit_atomic
+#define nilfs_find_next_zero_bit	find_next_zero_bit_le
+#define nilfs_find_next_bit		find_next_bit_le
+
+/**
+ * struct nilfs_bh_assoc - block offset and buffer head association
+ * @blkoff: block offset
+ * @bh: buffer head
+ */
+struct nilfs_bh_assoc {
+	unsigned long blkoff;
+	struct buffer_head *bh;
+};
+
+/**
+ * struct nilfs_palloc_cache - persistent object allocator cache
+ * @lock: cache protecting lock
+ * @prev_desc: blockgroup descriptors cache
+ * @prev_bitmap: blockgroup bitmap cache
+ * @prev_entry: translation entries cache
+ */
+struct nilfs_palloc_cache {
+	spinlock_t lock;
+	struct nilfs_bh_assoc prev_desc;
+	struct nilfs_bh_assoc prev_bitmap;
+	struct nilfs_bh_assoc prev_entry;
+};
+
+void nilfs_palloc_setup_cache(struct inode *inode,
+			      struct nilfs_palloc_cache *cache);
+void nilfs_palloc_clear_cache(struct inode *inode);
+void nilfs_palloc_destroy_cache(struct inode *inode);
+
+#endif	/* _NILFS_ALLOC_H */