Now that we switch the pool to read-only mode when the data device runs
out of space it causes active writers to get IO errors once we resume
after resizing the data device.
If no_free_space is set, save bios to the 'retry_on_resume_list' and
requeue them on resume (once the data or metadata device may have been
resized).
With this patch the resize_io test passes again (on slower storage):
dmtest run --suite thin-provisioning -n /resize_io/
Later patches fix some subtle races associated with the pool mode
transitions done as part of the pool's -ENOSPC handling. These races
are exposed on fast storage (e.g. PCIe SSD).
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Acked-by: Joe Thornber <ejt@redhat.com>
struct pool_features pf;
bool low_water_triggered:1; /* A dm event has been sent */
struct pool_features pf;
bool low_water_triggered:1; /* A dm event has been sent */
- bool no_free_space:1; /* A -ENOSPC warning has been issued */
+ bool no_free_space:1; /* bios will be requeued if set */
struct dm_bio_prison *prison;
struct dm_kcopyd_client *copier;
struct dm_bio_prison *prison;
struct dm_kcopyd_client *copier;
spin_unlock_irqrestore(&pool->lock, flags);
}
spin_unlock_irqrestore(&pool->lock, flags);
}
+static void handle_unserviceable_bio(struct pool *pool, struct bio *bio)
+{
+ /*
+ * When pool is read-only, no cell locking is needed because
+ * nothing is changing.
+ */
+ WARN_ON_ONCE(get_pool_mode(pool) != PM_READ_ONLY);
+
+ if (pool->no_free_space)
+ retry_on_resume(bio);
+ else
+ bio_io_error(bio);
+}
+
static void retry_bios_on_resume(struct pool *pool, struct dm_bio_prison_cell *cell)
{
struct bio *bio;
static void retry_bios_on_resume(struct pool *pool, struct dm_bio_prison_cell *cell)
{
struct bio *bio;
cell_release(pool, cell, &bios);
while ((bio = bio_list_pop(&bios)))
cell_release(pool, cell, &bios);
while ((bio = bio_list_pop(&bios)))
+ handle_unserviceable_bio(pool, bio);
}
static void process_discard(struct thin_c *tc, struct bio *bio)
}
static void process_discard(struct thin_c *tc, struct bio *bio)
switch (r) {
case 0:
if (lookup_result.shared && (rw == WRITE) && bio->bi_size)
switch (r) {
case 0:
if (lookup_result.shared && (rw == WRITE) && bio->bi_size)
+ handle_unserviceable_bio(tc->pool, bio);
else {
inc_all_io_entry(tc->pool, bio);
remap_and_issue(tc, bio, lookup_result.block);
else {
inc_all_io_entry(tc->pool, bio);
remap_and_issue(tc, bio, lookup_result.block);
case -ENODATA:
if (rw != READ) {
case -ENODATA:
if (rw != READ) {
+ handle_unserviceable_bio(tc->pool, bio);
if (get_pool_mode(tc->pool) == PM_READ_ONLY) {
/*
* This block isn't provisioned, and we have no way
if (get_pool_mode(tc->pool) == PM_READ_ONLY) {
/*
* This block isn't provisioned, and we have no way
- * of doing so. Just error it.
+ handle_unserviceable_bio(tc->pool, bio);
return DM_MAPIO_SUBMITTED;
}
/* fall through */
return DM_MAPIO_SUBMITTED;
}
/* fall through */