[Devel] [PATCH RHEL9 COMMIT] dm-ploop: move preparations of pios into the caller from worker
Konstantin Khorenko
khorenko at virtuozzo.com
Mon Jan 27 16:12:36 MSK 2025
The commit is pushed to "branch-rh9-5.14.0-427.44.1.vz9.80.x-ovz" and will appear at git at bitbucket.org:openvz/vzkernel.git
after rh9-5.14.0-427.44.1.vz9.80.6
------>
commit c376cd1bb787bcc7ead8db618cb38ae654ef50af
Author: Alexander Atanasov <alexander.atanasov at virtuozzo.com>
Date: Fri Jan 24 17:35:45 2025 +0200
dm-ploop: move preparations of pios into the caller from worker
Prepare pios earlier in preparation to try to execute them earlier.
Convert more places to use lock less lists.
https://virtuozzo.atlassian.net/browse/VSTOR-91820
Signed-off-by: Alexander Atanasov <alexander.atanasov at virtuozzo.com>
======
Patchset description:
ploop: optimistations and scalling
Ploop processes requsts in a different threads in parallel
where possible which results in significant improvement in
performance and makes further optimistations possible.
Known bugs:
- delayed metadata writeback is not working and is missing error handling
- patch to disable it until fixed
- fast path is not working - causes rcu lockups - patch to disable it
Further improvements:
- optimize md pages lookups
Alexander Atanasov (50):
dm-ploop: md_pages map all pages at creation time
dm-ploop: Use READ_ONCE/WRITE_ONCE to access md page data
dm-ploop: fsync after all pios are sent
dm-ploop: move md status to use proper bitops
dm-ploop: convert wait_list and wb_batch_llist to use lockless lists
dm-ploop: convert enospc handling to use lockless lists
dm-ploop: convert suspended_pios list to use lockless list
dm-ploop: convert the rest of the lists to use llist variant
dm-ploop: combine processing of pios thru prepare list and remove
fsync worker
dm-ploop: move from wq to kthread
dm-ploop: move preparations of pios into the caller from worker
dm-ploop: fast path execution for reads
dm-ploop: do not use a wrapper for set_bit to make a page writeback
dm-ploop: BAT use only one list for writeback
dm-ploop: make md writeback timeout to be per page
dm-ploop: add interface to disable bat writeback delay
dm-ploop: convert wb_batch_list to lockless variant
dm-ploop: convert high_prio to status
dm-ploop: split cow processing into two functions
dm-ploop: convert md page rw lock to spin lock
dm-ploop: convert bat_rwlock to bat_lock spinlock
dm-ploop: prepare bat updates under bat_lock
dm-ploop: make ploop_bat_write_complete ready for parallel pio
completion
dm-ploop: make ploop_submit_metadata_writeback return number of
requests sent
dm-ploop: introduce pio runner threads
dm-ploop: add pio list ids to be used when passing pios to runners
dm-ploop: process pios via runners
dm-ploop: disable metadata writeback delay
dm-ploop: disable fast path
dm-ploop: use lockless lists for chained cow updates list
dm-ploop: use lockless lists for data ready pios
dm-ploop: give runner threads better name
dm-ploop: resize operation - add holes bitmap locking
dm-ploop: remove unnecessary operations
dm-ploop: use filp per thread
dm-ploop: catch if we try to advance pio past bio end
dm-ploop: support REQ_FUA for data pios
dm-ploop: proplerly access nr_bat_entries
dm-ploop: fix locking and improve error handling when submitting pios
dm-ploop: fix how ENOTBLK is handled
dm-ploop: sync when suspended or stopping
dm-ploop: rework bat completion logic
dm-ploop: rework logic in pio processing
dm-ploop: end fsync pios in parallel
dm-ploop: make filespace preallocations async
dm-ploop: resubmit enospc pios from dispatcher thread
dm-ploop: dm-ploop: simplify discard completion
dm-ploop: use GFP_ATOMIC instead of GFP_NOIO
dm-ploop: fix locks used in mixed context
dm-ploop: fix how current flags are managed inside threads
Andrey Zhadchenko (13):
dm-ploop: do not flush after metadata writes
dm-ploop: set IOCB_DSYNC on all FUA requests
dm-ploop: remove extra ploop_cluster_is_in_top_delta()
dm-ploop: introduce per-md page locking
dm-ploop: reduce BAT accesses on discard completion
dm-ploop: simplify llseek
dm-ploop: speed up ploop_prepare_bat_update()
dm-ploop: make new allocations immediately visible in BAT
dm-ploop: drop ploop_cluster_is_in_top_delta()
dm-ploop: do not wait for BAT update for non-FUA requests
dm-ploop: add delay for metadata writeback
dm-ploop: submit all postponed metadata on REQ_OP_FLUSH
dm-ploop: handle REQ_PREFLUSH
Feature: dm-ploop: ploop target driver
---
drivers/md/dm-ploop-cmd.c | 13 +-----
drivers/md/dm-ploop-map.c | 101 +++++++++++++++++++++++++++-------------------
drivers/md/dm-ploop.h | 2 +-
3 files changed, 62 insertions(+), 54 deletions(-)
diff --git a/drivers/md/dm-ploop-cmd.c b/drivers/md/dm-ploop-cmd.c
index 3d9818003a08..1cd215800a16 100644
--- a/drivers/md/dm-ploop-cmd.c
+++ b/drivers/md/dm-ploop-cmd.c
@@ -123,7 +123,6 @@ ALLOW_ERROR_INJECTION(ploop_inflight_bios_ref_switch, ERRNO);
static void ploop_resume_submitting_pios(struct ploop *ploop)
{
- LIST_HEAD(list);
struct llist_node *suspended_pending;
spin_lock_irq(&ploop->deferred_lock);
@@ -132,16 +131,8 @@ static void ploop_resume_submitting_pios(struct ploop *ploop)
spin_unlock_irq(&ploop->deferred_lock);
suspended_pending = llist_del_all(&ploop->llsuspended_pios);
- if (suspended_pending) {
- struct llist_node *pos, *t;
-
- llist_for_each_safe(pos, t, suspended_pending) {
- struct pio *pio = list_entry((struct list_head *)pos, typeof(*pio), list);
-
- list_add(&pio->list, &list);
- }
- ploop_submit_embedded_pios(ploop, &list);
- }
+ if (suspended_pending)
+ ploop_submit_embedded_pios(ploop, suspended_pending);
}
static int ploop_suspend_submitting_pios(struct ploop *ploop)
diff --git a/drivers/md/dm-ploop-map.c b/drivers/md/dm-ploop-map.c
index 76ef88d563a3..897b4cb04746 100644
--- a/drivers/md/dm-ploop-map.c
+++ b/drivers/md/dm-ploop-map.c
@@ -143,18 +143,11 @@ static void ploop_init_prq_and_embedded_pio(struct ploop *ploop,
void ploop_enospc_timer(struct timer_list *timer)
{
struct ploop *ploop = from_timer(ploop, timer, enospc_timer);
- LIST_HEAD(list);
- struct llist_node *pos, *t;
struct llist_node *enospc_pending = llist_del_all(&ploop->enospc_pios);
if (enospc_pending) {
enospc_pending = llist_reverse_order(enospc_pending);
- llist_for_each_safe(pos, t, enospc_pending) {
- struct pio *pio = list_entry((struct list_head *)pos, typeof(*pio), list);
-
- list_add(&pio->list, &list);
- }
- ploop_submit_embedded_pios(ploop, &list);
+ ploop_submit_embedded_pios(ploop, enospc_pending);
}
}
@@ -303,11 +296,15 @@ static struct pio *ploop_split_and_chain_pio(struct ploop *ploop,
ALLOW_ERROR_INJECTION(ploop_split_and_chain_pio, NULL);
static int ploop_split_pio_to_list(struct ploop *ploop, struct pio *pio,
- struct list_head *ret_list)
+ struct llist_head *ret_list)
{
u32 clu_size = CLU_SIZE(ploop);
+ struct llist_node *pos, *t;
struct pio *split;
- LIST_HEAD(list);
+ LLIST_HEAD(llist);
+ struct llist_node *lltmp;
+
+ WARN_ON(!pio->bi_iter.bi_size);
while (1) {
loff_t start = to_bytes(pio->bi_iter.bi_sector);
@@ -325,17 +322,24 @@ static int ploop_split_pio_to_list(struct ploop *ploop, struct pio *pio,
if (!split)
goto err;
- list_add_tail(&split->list, &list);
+ llist_add((struct llist_node *)(&split->list), &llist);
}
- list_splice_tail(&list, ret_list);
- list_add_tail(&pio->list, ret_list);
+ INIT_LIST_HEAD(&pio->list);
+ llist_add((struct llist_node *)(&pio->list), &llist);
+ lltmp = llist_reverse_order(llist_del_all(&llist));
+ pio->list.next = NULL;
+ llist_add_batch(lltmp, (struct llist_node *)(&pio->list), ret_list);
+
return 0;
err:
- while ((pio = ploop_pio_list_pop(&list)) != NULL) {
+ llist_for_each_safe(pos, t, llist.first) {
+ pio = list_entry((struct list_head *)pos, typeof(*pio), list);
pio->bi_status = BLK_STS_RESOURCE;
+ INIT_LIST_HEAD(&pio->list);
ploop_pio_endio(pio);
}
+
return -ENOMEM;
}
ALLOW_ERROR_INJECTION(ploop_split_pio_to_list, ERRNO);
@@ -347,7 +351,7 @@ static void ploop_schedule_work(struct ploop *ploop)
static void ploop_dispatch_pio(struct ploop *ploop, struct pio *pio)
{
- struct llist_head *list = (struct llist_head *)&ploop->pios[pio->queue_list_id];
+ struct llist_head *list = &ploop->pios[pio->queue_list_id];
WARN_ON_ONCE(pio->queue_list_id >= PLOOP_LIST_COUNT);
@@ -1632,12 +1636,11 @@ ALLOW_ERROR_INJECTION(ploop_create_bvec_from_rq, NULL);
static void ploop_prepare_one_embedded_pio(struct ploop *ploop,
struct pio *pio,
- struct list_head *deferred_pios)
+ struct llist_head *lldeferred_pios)
{
struct ploop_rq *prq = pio->endio_cb_data;
struct request *rq = prq->rq;
struct bio_vec *bvec = NULL;
- LIST_HEAD(list);
int ret;
if (rq->bio != rq->biotail) {
@@ -1656,16 +1659,18 @@ static void ploop_prepare_one_embedded_pio(struct ploop *ploop,
pio->bi_iter.bi_idx = 0;
pio->bi_iter.bi_bvec_done = 0;
} else {
- /* Single bio already provides bvec array */
+ /* Single bio already provides bvec array
+ * bvec is updated to the correct on submit
+ * it is different after partial IO
+ */
bvec = rq->bio->bi_io_vec;
-
pio->bi_iter = rq->bio->bi_iter;
}
pio->bi_iter.bi_sector = ploop_rq_pos(ploop, rq);
pio->bi_io_vec = bvec;
pio->queue_list_id = PLOOP_LIST_DEFERRED;
- ret = ploop_split_pio_to_list(ploop, pio, deferred_pios);
+ ret = ploop_split_pio_to_list(ploop, pio, lldeferred_pios);
if (ret)
goto err_nomem;
@@ -1677,7 +1682,7 @@ static void ploop_prepare_one_embedded_pio(struct ploop *ploop,
static void ploop_prepare_embedded_pios(struct ploop *ploop,
struct llist_node *pios,
- struct list_head *deferred_pios)
+ struct llist_head *deferred_pios)
{
struct pio *pio;
struct llist_node *pos, *t;
@@ -1694,12 +1699,17 @@ static void ploop_prepare_embedded_pios(struct ploop *ploop,
}
static void ploop_process_deferred_pios(struct ploop *ploop,
- struct list_head *pios)
+ struct llist_head *pios)
{
struct pio *pio;
- while ((pio = ploop_pio_list_pop(pios)) != NULL)
+ struct llist_node *pos, *t;
+
+ llist_for_each_safe(pos, t, pios->first) {
+ pio = list_entry((struct list_head *)pos, typeof(*pio), list);
+ INIT_LIST_HEAD(&pio->list); /* until type is changed */
ploop_process_one_deferred_bio(ploop, pio);
+ }
}
static void ploop_process_one_discard_pio(struct ploop *ploop, struct pio *pio)
@@ -1798,19 +1808,13 @@ static void ploop_submit_metadata_writeback(struct ploop *ploop)
}
}
-static void process_ploop_fsync_work(struct ploop *ploop)
+static void process_ploop_fsync_work(struct ploop *ploop, struct llist_node *llflush_pios)
{
struct file *file;
struct pio *pio;
int ret;
- struct llist_node *llflush_pios;
struct llist_node *pos, *t;
- llflush_pios = llist_del_all(&ploop->pios[PLOOP_LIST_FLUSH]);
-
- if (!llflush_pios)
- return;
-
file = ploop_top_delta(ploop)->file;
/* All flushes are done as one */
ret = vfs_fsync(file, 0);
@@ -1829,12 +1833,13 @@ static void process_ploop_fsync_work(struct ploop *ploop)
void do_ploop_run_work(struct ploop *ploop)
{
- LIST_HEAD(deferred_pios);
+ LLIST_HEAD(deferred_pios);
struct llist_node *llembedded_pios;
struct llist_node *lldeferred_pios;
struct llist_node *lldiscard_pios;
struct llist_node *llcow_pios;
struct llist_node *llresubmit;
+ struct llist_node *llflush_pios;
unsigned int old_flags = current->flags;
current->flags |= PF_IO_THREAD|PF_LOCAL_THROTTLE|PF_MEMALLOC_NOIO;
@@ -1842,24 +1847,23 @@ void do_ploop_run_work(struct ploop *ploop)
llembedded_pios = llist_del_all(&ploop->pios[PLOOP_LIST_PREPARE]);
lldeferred_pios = llist_del_all(&ploop->pios[PLOOP_LIST_DEFERRED]);
+ llresubmit = llist_del_all(&ploop->llresubmit_pios);
lldiscard_pios = llist_del_all(&ploop->pios[PLOOP_LIST_DISCARD]);
llcow_pios = llist_del_all(&ploop->pios[PLOOP_LIST_COW]);
- llresubmit = llist_del_all(&ploop->llresubmit_pios);
/* add old deferred back to the list */
if (lldeferred_pios) {
struct llist_node *pos, *t;
- struct pio *pio;
-
+ /* Add one by one we need last for batch add */
llist_for_each_safe(pos, t, lldeferred_pios) {
- pio = list_entry((struct list_head *)pos, typeof(*pio), list);
- INIT_LIST_HEAD(&pio->list);
- list_add(&pio->list, &deferred_pios);
+ llist_add(pos, &deferred_pios);
}
}
ploop_prepare_embedded_pios(ploop, llembedded_pios, &deferred_pios);
+ llflush_pios = llist_del_all(&ploop->pios[PLOOP_LIST_FLUSH]);
+
if (llresubmit)
ploop_process_resubmit_pios(ploop, llist_reverse_order(llresubmit));
@@ -1873,8 +1877,9 @@ void do_ploop_run_work(struct ploop *ploop)
ploop_submit_metadata_writeback(ploop);
- if (!llist_empty(&ploop->pios[PLOOP_LIST_FLUSH]))
- process_ploop_fsync_work(ploop);
+ if (llflush_pios)
+ process_ploop_fsync_work(ploop, llist_reverse_order(llflush_pios));
+
current->flags = old_flags;
}
@@ -1917,6 +1922,9 @@ static void ploop_submit_embedded_pio(struct ploop *ploop, struct pio *pio)
{
struct ploop_rq *prq = pio->endio_cb_data;
struct request *rq = prq->rq;
+ LLIST_HEAD(deferred_pios);
+ struct pio *spio;
+ struct llist_node *pos, *t;
if (blk_rq_bytes(rq)) {
pio->queue_list_id = PLOOP_LIST_PREPARE;
@@ -1931,17 +1939,26 @@ static void ploop_submit_embedded_pio(struct ploop *ploop, struct pio *pio)
}
ploop_inc_nr_inflight(ploop, pio);
- llist_add((struct llist_node *)(&pio->list), &ploop->pios[PLOOP_LIST_PREPARE]);
+ ploop_prepare_one_embedded_pio(ploop, pio, &deferred_pios);
+
+ llist_for_each_safe(pos, t, deferred_pios.first) {
+ spio = list_entry((struct list_head *)pos, typeof(*pio), list);
+ llist_add((struct llist_node *)(&spio->list), &ploop->pios[PLOOP_LIST_DEFERRED]);
+ }
ploop_schedule_work(ploop);
}
-void ploop_submit_embedded_pios(struct ploop *ploop, struct list_head *list)
+void ploop_submit_embedded_pios(struct ploop *ploop, struct llist_node *list)
{
struct pio *pio;
+ struct llist_node *pos, *t;
- while ((pio = ploop_pio_list_pop(list)) != NULL)
+ llist_for_each_safe(pos, t, list) {
+ pio = list_entry((struct list_head *)pos, typeof(*pio), list);
+ INIT_LIST_HEAD(&pio->list);
ploop_submit_embedded_pio(ploop, pio);
+ }
}
int ploop_clone_and_map(struct dm_target *ti, struct request *rq,
diff --git a/drivers/md/dm-ploop.h b/drivers/md/dm-ploop.h
index 64d8eac4ef84..1e5927365d81 100644
--- a/drivers/md/dm-ploop.h
+++ b/drivers/md/dm-ploop.h
@@ -569,7 +569,7 @@ extern int ploop_add_delta(struct ploop *ploop, u32 level,
extern int ploop_check_delta_length(struct ploop *ploop, struct file *file,
loff_t *file_size);
extern void ploop_submit_embedded_pios(struct ploop *ploop,
- struct list_head *list);
+ struct llist_node *list);
extern void ploop_dispatch_pios(struct ploop *ploop, struct pio *pio,
struct list_head *pio_list);
extern void do_ploop_work(struct work_struct *ws);
More information about the Devel
mailing list