From: Alexander Atanasov <alexander.atana...@virtuozzo.com> The difference between hlist_unhashed_lockless and hlist_unhashed is that _lockless version uses READ_ONCE to do the check. Since it is used without locks we must switch to the _lockless variant.
Also make locking clusters and adding to inflight_pios return result so we can track if they failed and where. https://virtuozzo.atlassian.net/browse/VSTOR-91821 Signed-off-by: Alexander Atanasov <alexander.atana...@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-map.c | 70 ++++++++++++++++++++++++++++++--------- 1 file changed, 54 insertions(+), 16 deletions(-) diff --git a/drivers/md/dm-ploop-map.c b/drivers/md/dm-ploop-map.c index d2d919b6c544..4d15701cae53 100644 --- a/drivers/md/dm-ploop-map.c +++ b/drivers/md/dm-ploop-map.c @@ -498,17 +498,32 @@ static void ploop_dec_nr_inflight(struct ploop *ploop, struct pio *pio) } } -static void ploop_link_pio(struct hlist_head head[], struct pio *pio, +static int ploop_link_pio(struct hlist_head head[], struct pio *pio, u32 clu, bool exclusive) { struct hlist_head *slot = ploop_htable_slot(head, clu); + struct ploop *ploop = pio->ploop; + + if (exclusive) { + struct pio *pe = ploop_find_pio(head, clu); + + if (pe == pio) + return 1; - if (exclusive) - WARN_ON_ONCE(ploop_find_pio(head, clu) != NULL); + WARN_ON_ONCE(pe != NULL); - BUG_ON(!hlist_unhashed(&pio->hlist_node)); + if (pe) + PL_ERR("clu:%u already exclusively locked\n", clu); + } + + if (!hlist_unhashed_lockless(&pio->hlist_node)) { + PL_ERR("already hashed clu:%u e:%d\n", clu, exclusive); + WARN_ON_ONCE(1); + return 0; + } hlist_add_head(&pio->hlist_node, slot); pio->clu = clu; + return 1; } /* @@ -516,45 +531,55 @@ static void ploop_link_pio(struct hlist_head head[], struct pio *pio, * or from exclusive_bios_rbtree. BIOs from endio_list are requeued * to deferred_list. */ -static void ploop_unlink_pio(struct ploop *ploop, struct pio *pio, +static int ploop_unlink_pio(struct ploop *ploop, struct pio *pio, struct list_head *pio_list) { - BUG_ON(hlist_unhashed(&pio->hlist_node)); + if (hlist_unhashed_lockless(&pio->hlist_node)) { + WARN_ON_ONCE(1); + return 0; + } hlist_del_init(&pio->hlist_node); list_splice_tail_init(&pio->endio_list, pio_list); + return 1; } -static void ploop_add_cluster_lk(struct ploop *ploop, struct pio *pio, u32 clu) +static int ploop_add_cluster_lk(struct ploop *ploop, struct pio *pio, u32 clu) { unsigned long flags; + int ret; spin_lock_irqsave(&ploop->deferred_lock, flags); - ploop_link_pio(ploop->exclusive_pios, pio, clu, true); + ret = ploop_link_pio(ploop->exclusive_pios, pio, clu, true); spin_unlock_irqrestore(&ploop->deferred_lock, flags); + return ret; } -static void ploop_del_cluster_lk(struct ploop *ploop, struct pio *pio) +static int ploop_del_cluster_lk(struct ploop *ploop, struct pio *pio) { LIST_HEAD(pio_list); unsigned long flags; + int ret; spin_lock_irqsave(&ploop->deferred_lock, flags); - ploop_unlink_pio(ploop, pio, &pio_list); + ret = ploop_unlink_pio(ploop, pio, &pio_list); spin_unlock_irqrestore(&ploop->deferred_lock, flags); if (!list_empty(&pio_list)) ploop_dispatch_pios(ploop, NULL, &pio_list); + return ret; } -static void ploop_link_submitting_pio(struct ploop *ploop, struct pio *pio, +static int ploop_link_submitting_pio(struct ploop *ploop, struct pio *pio, u32 clu) { unsigned long flags; + int ret; spin_lock_irqsave(&ploop->inflight_lock, flags); - ploop_link_pio(ploop->inflight_pios, pio, clu, false); + ret = ploop_link_pio(ploop->inflight_pios, pio, clu, false); spin_unlock_irqrestore(&ploop->inflight_lock, flags); + return ret; } static void ploop_unlink_completed_pio(struct ploop *ploop, struct pio *pio) @@ -562,7 +587,7 @@ static void ploop_unlink_completed_pio(struct ploop *ploop, struct pio *pio) LIST_HEAD(pio_list); unsigned long flags; - if (hlist_unhashed(&pio->hlist_node)) + if (hlist_unhashed_lockless(&pio->hlist_node)) return; spin_lock_irqsave(&ploop->inflight_lock, flags); @@ -706,7 +731,8 @@ static int ploop_handle_discard_pio(struct ploop *ploop, struct pio *pio, return 0; } - ploop_add_cluster_lk(ploop, pio, clu); + if (!ploop_add_cluster_lk(ploop, pio, clu)) + PL_ERR("dis clu %d already locked\n", clu); pio->wants_discard_index_cleanup = true; punch_hole: @@ -1429,7 +1455,8 @@ static int ploop_submit_cluster_cow(struct ploop *ploop, unsigned int level, cow->aux_pio = aux_pio; cow->cow_pio = cow_pio; - ploop_add_cluster_lk(ploop, cow_pio, clu); + if (!ploop_add_cluster_lk(ploop, cow_pio, clu)) + PL_ERR("cowclu %d already locked\n", clu); /* Stage #0: read secondary delta full clu */ ploop_map_and_submit_rw(ploop, dst_clu, aux_pio, level); @@ -1706,8 +1733,19 @@ static int ploop_process_one_deferred_bio(struct ploop *ploop, struct pio *pio) &dst_clu, pio); if (!ret) goto executed; + queue: /* pio needs to go to the queue */ - ploop_link_submitting_pio(ploop, pio, clu); + if (!ploop_link_submitting_pio(ploop, pio, clu)) { + PL_ERR("link submitting pio failed\n"); + /* Pio is put into endio_list */ + if (ploop_postpone_if_cluster_locked(ploop, pio, clu)) { + goto handled; + } else { + PL_ERR("link submitting pio postpone failed\n"); + WARN_ON_ONCE(1); + /* fail thru and run it anyway nothing */ + } + } ploop_map_and_submit_rw(ploop, dst_clu, pio, ploop_top_level(ploop)); // return 1; executed: /* pio op started */ -- 2.43.5 _______________________________________________ Devel mailing list Devel@openvz.org https://lists.openvz.org/mailman/listinfo/devel
