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, 153 insertions, 0 deletions

diff --git a/drivers/clocksource/timer-goldfish.c b/drivers/clocksource/timer-goldfish.c
new file mode 100644
index 000000000..3fe12c7ee
--- /dev/null
+++ b/drivers/clocksource/timer-goldfish.c
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0
+
+#include <linux/interrupt.h>
+#include <linux/ioport.h>
+#include <linux/clocksource.h>
+#include <linux/clockchips.h>
+#include <linux/module.h>
+#include <linux/slab.h>
+#include <linux/goldfish.h>
+#include <clocksource/timer-goldfish.h>
+
+struct goldfish_timer {
+	struct clocksource cs;
+	struct clock_event_device ced;
+	struct resource res;
+	void __iomem *base;
+};
+
+static struct goldfish_timer *ced_to_gf(struct clock_event_device *ced)
+{
+	return container_of(ced, struct goldfish_timer, ced);
+}
+
+static struct goldfish_timer *cs_to_gf(struct clocksource *cs)
+{
+	return container_of(cs, struct goldfish_timer, cs);
+}
+
+static u64 goldfish_timer_read(struct clocksource *cs)
+{
+	struct goldfish_timer *timerdrv = cs_to_gf(cs);
+	void __iomem *base = timerdrv->base;
+	u32 time_low, time_high;
+	u64 ticks;
+
+	/*
+	 * time_low: get low bits of current time and update time_high
+	 * time_high: get high bits of time at last time_low read
+	 */
+	time_low = gf_ioread32(base + TIMER_TIME_LOW);
+	time_high = gf_ioread32(base + TIMER_TIME_HIGH);
+
+	ticks = ((u64)time_high << 32) | time_low;
+
+	return ticks;
+}
+
+static int goldfish_timer_set_oneshot(struct clock_event_device *evt)
+{
+	struct goldfish_timer *timerdrv = ced_to_gf(evt);
+	void __iomem *base = timerdrv->base;
+
+	gf_iowrite32(0, base + TIMER_ALARM_HIGH);
+	gf_iowrite32(0, base + TIMER_ALARM_LOW);
+	gf_iowrite32(1, base + TIMER_IRQ_ENABLED);
+
+	return 0;
+}
+
+static int goldfish_timer_shutdown(struct clock_event_device *evt)
+{
+	struct goldfish_timer *timerdrv = ced_to_gf(evt);
+	void __iomem *base = timerdrv->base;
+
+	gf_iowrite32(0, base + TIMER_IRQ_ENABLED);
+
+	return 0;
+}
+
+static int goldfish_timer_next_event(unsigned long delta,
+				     struct clock_event_device *evt)
+{
+	struct goldfish_timer *timerdrv = ced_to_gf(evt);
+	void __iomem *base = timerdrv->base;
+	u64 now;
+
+	now = goldfish_timer_read(&timerdrv->cs);
+
+	now += delta;
+
+	gf_iowrite32(upper_32_bits(now), base + TIMER_ALARM_HIGH);
+	gf_iowrite32(lower_32_bits(now), base + TIMER_ALARM_LOW);
+
+	return 0;
+}
+
+static irqreturn_t goldfish_timer_irq(int irq, void *dev_id)
+{
+	struct goldfish_timer *timerdrv = dev_id;
+	struct clock_event_device *evt = &timerdrv->ced;
+	void __iomem *base = timerdrv->base;
+
+	gf_iowrite32(1, base + TIMER_CLEAR_INTERRUPT);
+
+	evt->event_handler(evt);
+
+	return IRQ_HANDLED;
+}
+
+int __init goldfish_timer_init(int irq, void __iomem *base)
+{
+	struct goldfish_timer *timerdrv;
+	int ret;
+
+	timerdrv = kzalloc_obj(*timerdrv);
+	if (!timerdrv)
+		return -ENOMEM;
+
+	timerdrv->base = base;
+
+	timerdrv->ced = (struct clock_event_device){
+		.name			= "goldfish_timer",
+		.features		= CLOCK_EVT_FEAT_ONESHOT,
+		.set_state_shutdown	= goldfish_timer_shutdown,
+		.set_state_oneshot      = goldfish_timer_set_oneshot,
+		.set_next_event		= goldfish_timer_next_event,
+	};
+
+	timerdrv->res = (struct resource){
+		.name  = "goldfish_timer",
+		.start = (unsigned long)base,
+		.end   = (unsigned long)base + 0xfff,
+	};
+
+	ret = request_resource(&iomem_resource, &timerdrv->res);
+	if (ret) {
+		pr_err("Cannot allocate '%s' resource\n", timerdrv->res.name);
+		return ret;
+	}
+
+	timerdrv->cs = (struct clocksource){
+		.name		= "goldfish_timer",
+		.rating		= 400,
+		.read		= goldfish_timer_read,
+		.mask		= CLOCKSOURCE_MASK(64),
+		.flags		= 0,
+		.max_idle_ns	= LONG_MAX,
+	};
+
+	clocksource_register_hz(&timerdrv->cs, NSEC_PER_SEC);
+
+	ret = request_irq(irq, goldfish_timer_irq, IRQF_TIMER,
+			  "goldfish_timer", timerdrv);
+	if (ret) {
+		pr_err("Couldn't register goldfish-timer interrupt\n");
+		return ret;
+	}
+
+	clockevents_config_and_register(&timerdrv->ced, NSEC_PER_SEC,
+					1, 0xffffffff);
+
+	return 0;
+}