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, 95 insertions, 0 deletions

diff --git a/include/asm-generic/iomap.h b/include/asm-generic/iomap.h
new file mode 100644
index 000000000..9f3f25d7f
--- /dev/null
+++ b/include/asm-generic/iomap.h
@@ -0,0 +1,95 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __GENERIC_IO_H
+#define __GENERIC_IO_H
+
+#include <linux/linkage.h>
+#include <asm/byteorder.h>
+
+/*
+ * These are the "generic" interfaces for doing new-style
+ * memory-mapped or PIO accesses. Architectures may do
+ * their own arch-optimized versions, these just act as
+ * wrappers around the old-style IO register access functions:
+ * read[bwl]/write[bwl]/in[bwl]/out[bwl]
+ *
+ * Don't include this directly, include it from <asm/io.h>.
+ */
+
+/*
+ * Read/write from/to an (offsettable) iomem cookie. It might be a PIO
+ * access or a MMIO access, these functions don't care. The info is
+ * encoded in the hardware mapping set up by the mapping functions
+ * (or the cookie itself, depending on implementation and hw).
+ *
+ * The generic routines just encode the PIO/MMIO as part of the
+ * cookie, and coldly assume that the MMIO IO mappings are not
+ * in the low address range. Architectures for which this is not
+ * true can't use this generic implementation.
+ */
+extern unsigned int ioread8(const void __iomem *);
+extern unsigned int ioread16(const void __iomem *);
+extern unsigned int ioread16be(const void __iomem *);
+extern unsigned int ioread32(const void __iomem *);
+extern unsigned int ioread32be(const void __iomem *);
+
+extern u64 __ioread64_lo_hi(const void __iomem *addr);
+extern u64 __ioread64_hi_lo(const void __iomem *addr);
+extern u64 __ioread64be_lo_hi(const void __iomem *addr);
+extern u64 __ioread64be_hi_lo(const void __iomem *addr);
+
+extern void iowrite8(u8, void __iomem *);
+extern void iowrite16(u16, void __iomem *);
+extern void iowrite16be(u16, void __iomem *);
+extern void iowrite32(u32, void __iomem *);
+extern void iowrite32be(u32, void __iomem *);
+
+extern void __iowrite64_lo_hi(u64 val, void __iomem *addr);
+extern void __iowrite64_hi_lo(u64 val, void __iomem *addr);
+extern void __iowrite64be_lo_hi(u64 val, void __iomem *addr);
+extern void __iowrite64be_hi_lo(u64 val, void __iomem *addr);
+
+/*
+ * "string" versions of the above. Note that they
+ * use native byte ordering for the accesses (on
+ * the assumption that IO and memory agree on a
+ * byte order, and CPU byteorder is irrelevant).
+ *
+ * They do _not_ update the port address. If you
+ * want MMIO that copies stuff laid out in MMIO
+ * memory across multiple ports, use "memcpy_toio()"
+ * and friends.
+ */
+extern void ioread8_rep(const void __iomem *port, void *buf, unsigned long count);
+extern void ioread16_rep(const void __iomem *port, void *buf, unsigned long count);
+extern void ioread32_rep(const void __iomem *port, void *buf, unsigned long count);
+
+extern void iowrite8_rep(void __iomem *port, const void *buf, unsigned long count);
+extern void iowrite16_rep(void __iomem *port, const void *buf, unsigned long count);
+extern void iowrite32_rep(void __iomem *port, const void *buf, unsigned long count);
+
+#ifdef CONFIG_HAS_IOPORT_MAP
+/* Create a virtual mapping cookie for an IO port range */
+extern void __iomem *ioport_map(unsigned long port, unsigned int nr);
+extern void ioport_unmap(void __iomem *);
+#endif
+
+#ifndef ioremap_wc
+#define ioremap_wc ioremap
+#endif
+
+#ifndef ioremap_wt
+#define ioremap_wt ioremap
+#endif
+
+#ifndef ioremap_np
+/* See the comment in asm-generic/io.h about ioremap_np(). */
+#define ioremap_np ioremap_np
+static inline void __iomem *ioremap_np(phys_addr_t offset, size_t size)
+{
+	return NULL;
+}
+#endif
+
+#include <asm-generic/pci_iomap.h>
+
+#endif