/* * metadata/data stored in disk with 4k size unit (a block) regardless * underneath hardware sector size. only works with PAGE_SIZE == 4096
*/ #define BLOCK_SECTORS (8) #define BLOCK_SECTOR_SHIFT (3)
/* * log->max_free_space is min(1/4 disk size, 10G reclaimable space). * * In write through mode, the reclaim runs every log->max_free_space. * This can prevent the recovery scans for too long
*/ #define RECLAIM_MAX_FREE_SPACE (10 * 1024 * 1024 * 2) /* sector */ #define RECLAIM_MAX_FREE_SPACE_SHIFT (2)
/* wake up reclaim thread periodically */ #define R5C_RECLAIM_WAKEUP_INTERVAL (30 * HZ) /* start flush with these full stripes */ #define R5C_FULL_STRIPE_FLUSH_BATCH(conf) (conf->max_nr_stripes / 4) /* reclaim stripes in groups */ #define R5C_RECLAIM_STRIPE_GROUP (NR_STRIPE_HASH_LOCKS * 2)
/* * We only need 2 bios per I/O unit to make progress, but ensure we * have a few more available to not get too tight.
*/ #define R5L_POOL_SIZE 4
staticchar *r5c_journal_mode_str[] = {"write-through", "write-back"}; /* * raid5 cache state machine * * With the RAID cache, each stripe works in two phases: * - caching phase * - writing-out phase * * These two phases are controlled by bit STRIPE_R5C_CACHING: * if STRIPE_R5C_CACHING == 0, the stripe is in writing-out phase * if STRIPE_R5C_CACHING == 1, the stripe is in caching phase * * When there is no journal, or the journal is in write-through mode, * the stripe is always in writing-out phase. * * For write-back journal, the stripe is sent to caching phase on write * (r5c_try_caching_write). r5c_make_stripe_write_out() kicks off * the write-out phase by clearing STRIPE_R5C_CACHING. * * Stripes in caching phase do not write the raid disks. Instead, all * writes are committed from the log device. Therefore, a stripe in * caching phase handles writes as: * - write to log device * - return IO * * Stripes in writing-out phase handle writes as: * - calculate parity * - write pending data and parity to journal * - write data and parity to raid disks * - return IO for pending writes
*/
struct r5l_log { struct md_rdev *rdev;
u32 uuid_checksum;
sector_t device_size; /* log device size, round to
* BLOCK_SECTORS */
sector_t max_free_space; /* reclaim run if free space is at
* this size */
sector_t last_checkpoint; /* log tail. where recovery scan
* starts from */
u64 last_cp_seq; /* log tail sequence */
sector_t log_start; /* log head. where new data appends */
u64 seq; /* log head sequence */
sector_t next_checkpoint;
struct mutex io_mutex; struct r5l_io_unit *current_io; /* current io_unit accepting new data */
spinlock_t io_list_lock; struct list_head running_ios; /* io_units which are still running, * and have not yet been completely
* written to the log */ struct list_head io_end_ios; /* io_units which have been completely * written to the log but not yet written
* to the RAID */ struct list_head flushing_ios; /* io_units which are waiting for log
* cache flush */ struct list_head finished_ios; /* io_units which settle down in log disk */ struct bio flush_bio;
struct md_thread __rcu *reclaim_thread; unsignedlong reclaim_target; /* number of space that need to be * reclaimed. if it's 0, reclaim spaces * used by io_units which are in * IO_UNIT_STRIPE_END state (eg, reclaim * doesn't wait for specific io_unit * switching to IO_UNIT_STRIPE_END
* state) */
wait_queue_head_t iounit_wait;
struct list_head no_space_stripes; /* pending stripes, log has no space */
spinlock_t no_space_stripes_lock;
bool need_cache_flush;
/* for r5c_cache */ enum r5c_journal_mode r5c_journal_mode;
/* all stripes in r5cache, in the order of seq at sh->log_start */ struct list_head stripe_in_journal_list;
/* to submit async io_units, to fulfill ordering of flush */ struct work_struct deferred_io_work; /* to disable write back during in degraded mode */ struct work_struct disable_writeback_work;
/* to for chunk_aligned_read in writeback mode, details below */
spinlock_t tree_lock; struct radix_tree_root big_stripe_tree;
};
/* * Enable chunk_aligned_read() with write back cache. * * Each chunk may contain more than one stripe (for example, a 256kB * chunk contains 64 4kB-page, so this chunk contain 64 stripes). For * chunk_aligned_read, these stripes are grouped into one "big_stripe". * For each big_stripe, we count how many stripes of this big_stripe * are in the write back cache. These data are tracked in a radix tree * (big_stripe_tree). We use radix_tree item pointer as the counter. * r5c_tree_index() is used to calculate keys for the radix tree. * * chunk_aligned_read() calls r5c_big_stripe_cached() to look up * big_stripe of each chunk in the tree. If this big_stripe is in the * tree, chunk_aligned_read() aborts. This look up is protected by * rcu_read_lock(). * * It is necessary to remember whether a stripe is counted in * big_stripe_tree. Instead of adding new flag, we reuses existing flags: * STRIPE_R5C_PARTIAL_STRIPE and STRIPE_R5C_FULL_STRIPE. If either of these * two flags are set, the stripe is counted in big_stripe_tree. This * requires moving set_bit(STRIPE_R5C_PARTIAL_STRIPE) to * r5c_try_caching_write(); and moving clear_bit of * STRIPE_R5C_PARTIAL_STRIPE and STRIPE_R5C_FULL_STRIPE to * r5c_finish_stripe_write_out().
*/
/* * radix tree requests lowest 2 bits of data pointer to be 2b'00. * So it is necessary to left shift the counter by 2 bits before using it * as data pointer of the tree.
*/ #define R5C_RADIX_COUNT_SHIFT 2
/* * an IO range starts from a meta data block and end at the next meta data * block. The io unit's the meta data block tracks data/parity followed it. io * unit is written to log disk with normal write, as we always flush log disk * first and then start move data to raid disks, there is no requirement to * write io unit with FLUSH/FUA
*/ struct r5l_io_unit { struct r5l_log *log;
struct page *meta_page; /* store meta block */ int meta_offset; /* current offset in meta_page */
struct bio *current_bio;/* current_bio accepting new data */
atomic_t pending_stripe;/* how many stripes not flushed to raid */
u64 seq; /* seq number of the metablock */
sector_t log_start; /* where the io_unit starts */
sector_t log_end; /* where the io_unit ends */ struct list_head log_sibling; /* log->running_ios */ struct list_head stripe_list; /* stripes added to the io_unit */
int state; bool need_split_bio; struct bio *split_bio;
unsignedint has_flush:1; /* include flush request */ unsignedint has_fua:1; /* include fua request */ unsignedint has_null_flush:1; /* include null flush request */ unsignedint has_flush_payload:1; /* include flush payload */ /* * io isn't sent yet, flush/fua request can only be submitted till it's * the first IO in running_ios list
*/ unsignedint io_deferred:1;
/* Check whether we should flush some stripes to free up stripe cache */ void r5c_check_stripe_cache_usage(struct r5conf *conf)
{ int total_cached; struct r5l_log *log = READ_ONCE(conf->log);
/* * The following condition is true for either of the following: * - stripe cache pressure high: * total_cached > 3/4 min_nr_stripes || * empty_inactive_list_nr > 0 * - stripe cache pressure moderate: * total_cached > 1/2 min_nr_stripes
*/ if (total_cached > conf->min_nr_stripes * 1 / 2 ||
atomic_read(&conf->empty_inactive_list_nr) > 0)
r5l_wake_reclaim(log, 0);
}
/* * flush cache when there are R5C_FULL_STRIPE_FLUSH_BATCH or more full * stripes in the cache
*/ void r5c_check_cached_full_stripe(struct r5conf *conf)
{ struct r5l_log *log = READ_ONCE(conf->log);
if (!r5c_is_writeback(log)) return;
/* * wake up reclaim for R5C_FULL_STRIPE_FLUSH_BATCH cached stripes * or a full stripe (chunk size / 4k stripes).
*/ if (atomic_read(&conf->r5c_cached_full_stripes) >=
min(R5C_FULL_STRIPE_FLUSH_BATCH(conf),
conf->chunk_sectors >> RAID5_STRIPE_SHIFT(conf)))
r5l_wake_reclaim(log, 0);
}
/* * Total log space (in sectors) needed to flush all data in cache * * To avoid deadlock due to log space, it is necessary to reserve log * space to flush critical stripes (stripes that occupying log space near * last_checkpoint). This function helps check how much log space is * required to flush all cached stripes. * * To reduce log space requirements, two mechanisms are used to give cache * flush higher priorities: * 1. In handle_stripe_dirtying() and schedule_reconstruction(), * stripes ALREADY in journal can be flushed w/o pending writes; * 2. In r5l_write_stripe() and r5c_cache_data(), stripes NOT in journal * can be delayed (r5l_add_no_space_stripe). * * In cache flush, the stripe goes through 1 and then 2. For a stripe that * already passed 1, flushing it requires at most (conf->max_degraded + 1) * pages of journal space. For stripes that has not passed 1, flushing it * requires (conf->raid_disks + 1) pages of journal space. There are at * most (conf->group_cnt + 1) stripe that passed 1. So total journal space * required to flush all cached stripes (in pages) is: * * (stripe_in_journal_count - group_cnt - 1) * (max_degraded + 1) + * (group_cnt + 1) * (raid_disks + 1) * or * (stripe_in_journal_count) * (max_degraded + 1) + * (group_cnt + 1) * (raid_disks - max_degraded)
*/ static sector_t r5c_log_required_to_flush_cache(struct r5conf *conf)
{ struct r5l_log *log = READ_ONCE(conf->log);
/* * evaluate log space usage and update R5C_LOG_TIGHT and R5C_LOG_CRITICAL * * R5C_LOG_TIGHT is set when free space on the log device is less than 3x of * reclaim_required_space. R5C_LOG_CRITICAL is set when free space on the log * device is less than 2x of reclaim_required_space.
*/ staticinlinevoid r5c_update_log_state(struct r5l_log *log)
{ struct r5conf *conf = log->rdev->mddev->private;
sector_t free_space;
sector_t reclaim_space; bool wake_reclaim = false;
/* * Put the stripe into writing-out phase by clearing STRIPE_R5C_CACHING. * This function should only be called in write-back mode.
*/ void r5c_make_stripe_write_out(struct stripe_head *sh)
{ struct r5conf *conf = sh->raid_conf; struct r5l_log *log = READ_ONCE(conf->log);
if (!test_and_set_bit(STRIPE_PREREAD_ACTIVE, &sh->state))
atomic_inc(&conf->preread_active_stripes);
}
staticvoid r5c_handle_data_cached(struct stripe_head *sh)
{ int i;
for (i = sh->disks; i--; ) if (test_and_clear_bit(R5_Wantwrite, &sh->dev[i].flags)) {
set_bit(R5_InJournal, &sh->dev[i].flags);
clear_bit(R5_LOCKED, &sh->dev[i].flags);
}
clear_bit(STRIPE_LOG_TRAPPED, &sh->state);
}
/* * this journal write must contain full parity, * it may also contain some data pages
*/ staticvoid r5c_handle_parity_cached(struct stripe_head *sh)
{ int i;
for (i = sh->disks; i--; ) if (test_bit(R5_InJournal, &sh->dev[i].flags))
set_bit(R5_Wantwrite, &sh->dev[i].flags);
}
/* * Setting proper flags after writing (or flushing) data and/or parity to the * log device. This is called from r5l_log_endio() or r5l_log_flush_endio().
*/ staticvoid r5c_finish_cache_stripe(struct stripe_head *sh)
{ struct r5l_log *log = READ_ONCE(sh->raid_conf->log);
if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH) {
BUG_ON(test_bit(STRIPE_R5C_CACHING, &sh->state)); /* * Set R5_InJournal for parity dev[pd_idx]. This means * all data AND parity in the journal. For RAID 6, it is * NOT necessary to set the flag for dev[qd_idx], as the * two parities are written out together.
*/
set_bit(R5_InJournal, &sh->dev[sh->pd_idx].flags);
} elseif (test_bit(STRIPE_R5C_CACHING, &sh->state)) {
r5c_handle_data_cached(sh);
} else {
r5c_handle_parity_cached(sh);
set_bit(R5_InJournal, &sh->dev[sh->pd_idx].flags);
}
}
/* * if the io doesn't not have null_flush or flush payload, * it is not safe to access it after releasing io_list_lock. * Therefore, it is necessary to check the condition with * the lock held.
*/
has_null_flush = io->has_null_flush;
has_flush_payload = io->has_flush_payload;
if (log->need_cache_flush && !list_empty(&io->stripe_list))
r5l_move_to_end_ios(log); else
r5l_log_run_stripes(log); if (!list_empty(&log->running_ios)) { /* * FLUSH/FUA io_unit is deferred because of ordering, now we * can dispatch it
*/
io_deferred = list_first_entry(&log->running_ios, struct r5l_io_unit, log_sibling); if (io_deferred->io_deferred)
schedule_work(&log->deferred_io_work);
}
/* * In case of journal device failures, submit_bio will get error * and calls endio, then active stripes will continue write * process. Therefore, it is not necessary to check Faulty bit * of journal device here. * * We can't check split_bio after current_bio is submitted. If * io->split_bio is null, after current_bio is submitted, current_bio * might already be completed and the io_unit is freed. We submit * split_bio first to avoid the issue.
*/ if (io->split_bio) { if (io->has_flush)
io->split_bio->bi_opf |= REQ_PREFLUSH; if (io->has_fua)
io->split_bio->bi_opf |= REQ_FUA;
submit_bio(io->split_bio);
}
if (io->has_flush)
io->current_bio->bi_opf |= REQ_PREFLUSH; if (io->has_fua)
io->current_bio->bi_opf |= REQ_FUA;
submit_bio(io->current_bio);
}
/* deferred io_unit will be dispatched here */ staticvoid r5l_submit_io_async(struct work_struct *work)
{ struct r5l_log *log = container_of(work, struct r5l_log,
deferred_io_work); struct r5l_io_unit *io = NULL; unsignedlong flags;
spin_lock_irqsave(&log->io_list_lock, flags); if (!list_empty(&log->running_ios)) {
io = list_first_entry(&log->running_ios, struct r5l_io_unit,
log_sibling); if (!io->io_deferred)
io = NULL; else
io->io_deferred = 0;
}
spin_unlock_irqrestore(&log->io_list_lock, flags); if (io)
r5l_do_submit_io(log, io);
}
r5c_update_log_state(log); /* * If we filled up the log device start from the beginning again, * which will require a new bio. * * Note: for this to work properly the log size needs to me a multiple * of BLOCK_SECTORS.
*/ if (log->log_start == 0)
io->need_split_bio = true;
/* * payload_flush requires extra writes to the journal. * To avoid handling the extra IO in quiesce, just skip * flush_payload
*/ if (conf->quiesce) return;
if (r5l_get_meta(log, meta_size)) {
mutex_unlock(&log->io_mutex); return;
}
/* current implementation is one stripe per flush payload */
io = log->current_io;
payload = page_address(io->meta_page) + io->meta_offset;
payload->header.type = cpu_to_le16(R5LOG_PAYLOAD_FLUSH);
payload->header.flags = cpu_to_le16(0);
payload->size = cpu_to_le32(sizeof(__le64));
payload->flush_stripes[0] = cpu_to_le64(sect);
io->meta_offset += meta_size; /* multiple flush payloads count as one pending_stripe */ if (!io->has_flush_payload) {
io->has_flush_payload = 1;
atomic_inc(&io->pending_stripe);
}
mutex_unlock(&log->io_mutex);
}
staticint r5l_log_stripe(struct r5l_log *log, struct stripe_head *sh, int data_pages, int parity_pages)
{ int i; int meta_size; int ret; struct r5l_io_unit *io;
ret = r5l_get_meta(log, meta_size); if (ret) return ret;
io = log->current_io;
if (test_and_clear_bit(STRIPE_R5C_PREFLUSH, &sh->state))
io->has_flush = 1;
for (i = 0; i < sh->disks; i++) { if (!test_bit(R5_Wantwrite, &sh->dev[i].flags) ||
test_bit(R5_InJournal, &sh->dev[i].flags)) continue; if (i == sh->pd_idx || i == sh->qd_idx) continue; if (test_bit(R5_WantFUA, &sh->dev[i].flags) &&
log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_BACK) {
io->has_fua = 1; /* * we need to flush journal to make sure recovery can * reach the data with fua flag
*/
io->has_flush = 1;
}
r5l_append_payload_meta(log, R5LOG_PAYLOAD_DATA,
raid5_compute_blocknr(sh, i, 0),
sh->dev[i].log_checksum, 0, false);
r5l_append_payload_page(log, sh->dev[i].page);
}
/* add stripe to no_space_stripes, and then wake up reclaim */ staticinlinevoid r5l_add_no_space_stripe(struct r5l_log *log, struct stripe_head *sh)
{
spin_lock(&log->no_space_stripes_lock);
list_add_tail(&sh->log_list, &log->no_space_stripes);
spin_unlock(&log->no_space_stripes_lock);
}
/* * running in raid5d, where reclaim could wait for raid5d too (when it flushes * data from log to raid disks), so we shouldn't wait for reclaim here
*/ int r5l_write_stripe(struct r5l_log *log, struct stripe_head *sh)
{ struct r5conf *conf = sh->raid_conf; int write_disks = 0; int data_pages, parity_pages; int reserve; int i; int ret = 0; bool wake_reclaim = false;
if (!log) return -EAGAIN; /* Don't support stripe batch */ if (sh->log_io || !test_bit(R5_Wantwrite, &sh->dev[sh->pd_idx].flags) ||
test_bit(STRIPE_SYNCING, &sh->state)) { /* the stripe is written to log, we start writing it to raid */
clear_bit(STRIPE_LOG_TRAPPED, &sh->state); return -EAGAIN;
}
if (!test_bit(R5_Wantwrite, &sh->dev[i].flags) ||
test_bit(R5_InJournal, &sh->dev[i].flags)) continue;
write_disks++; /* checksum is already calculated in last run */ if (test_bit(STRIPE_LOG_TRAPPED, &sh->state)) continue;
addr = kmap_local_page(sh->dev[i].page);
sh->dev[i].log_checksum = crc32c(log->uuid_checksum,
addr, PAGE_SIZE);
kunmap_local(addr);
}
parity_pages = 1 + !!(sh->qd_idx >= 0);
data_pages = write_disks - parity_pages;
set_bit(STRIPE_LOG_TRAPPED, &sh->state); /* * The stripe must enter state machine again to finish the write, so * don't delay.
*/
clear_bit(STRIPE_DELAYED, &sh->state);
atomic_inc(&sh->count);
mutex_lock(&log->io_mutex); /* meta + data */
reserve = (1 + write_disks) << (PAGE_SHIFT - 9);
if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH) { if (!r5l_has_free_space(log, reserve)) {
r5l_add_no_space_stripe(log, sh);
wake_reclaim = true;
} else {
ret = r5l_log_stripe(log, sh, data_pages, parity_pages); if (ret) {
spin_lock_irq(&log->io_list_lock);
list_add_tail(&sh->log_list,
&log->no_mem_stripes);
spin_unlock_irq(&log->io_list_lock);
}
}
} else { /* R5C_JOURNAL_MODE_WRITE_BACK */ /* * log space critical, do not process stripes that are * not in cache yet (sh->log_start == MaxSector).
*/ if (test_bit(R5C_LOG_CRITICAL, &conf->cache_state) &&
sh->log_start == MaxSector) {
r5l_add_no_space_stripe(log, sh);
wake_reclaim = true;
reserve = 0;
} elseif (!r5l_has_free_space(log, reserve)) { if (sh->log_start == log->last_checkpoint)
BUG(); else
r5l_add_no_space_stripe(log, sh);
} else {
ret = r5l_log_stripe(log, sh, data_pages, parity_pages); if (ret) {
spin_lock_irq(&log->io_list_lock);
list_add_tail(&sh->log_list,
&log->no_mem_stripes);
spin_unlock_irq(&log->io_list_lock);
}
}
}
mutex_unlock(&log->io_mutex); if (wake_reclaim)
r5l_wake_reclaim(log, reserve); return 0;
}
int r5l_handle_flush_request(struct r5l_log *log, struct bio *bio)
{ if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH) { /* * in write through (journal only) * we flush log disk cache first, then write stripe data to * raid disks. So if bio is finished, the log disk cache is * flushed already. The recovery guarantees we can recovery * the bio from log disk, so we don't need to flush again
*/ if (bio->bi_iter.bi_size == 0) {
bio_endio(bio); return 0;
}
bio->bi_opf &= ~REQ_PREFLUSH;
} else { /* write back (with cache) */ if (bio->bi_iter.bi_size == 0) {
mutex_lock(&log->io_mutex);
r5l_get_meta(log, 0);
bio_list_add(&log->current_io->flush_barriers, bio);
log->current_io->has_flush = 1;
log->current_io->has_null_flush = 1;
atomic_inc(&log->current_io->pending_stripe);
r5l_submit_current_io(log);
mutex_unlock(&log->io_mutex); return 0;
}
} return -EAGAIN;
}
/* This will run after log space is reclaimed */ staticvoid r5l_run_no_space_stripes(struct r5l_log *log)
{ struct stripe_head *sh;
spin_lock(&log->no_space_stripes_lock); while (!list_empty(&log->no_space_stripes)) {
sh = list_first_entry(&log->no_space_stripes, struct stripe_head, log_list);
list_del_init(&sh->log_list);
set_bit(STRIPE_HANDLE, &sh->state);
raid5_release_stripe(sh);
}
spin_unlock(&log->no_space_stripes_lock);
}
/* * calculate new last_checkpoint * for write through mode, returns log->next_checkpoint * for write back, returns log_start of first sh in stripe_in_journal_list
*/ static sector_t r5c_calculate_new_cp(struct r5conf *conf)
{ struct stripe_head *sh; struct r5l_log *log = READ_ONCE(conf->log);
sector_t new_cp; unsignedlong flags;
if (log->r5c_journal_mode == R5C_JOURNAL_MODE_WRITE_THROUGH) return log->next_checkpoint;
spin_lock_irqsave(&log->stripe_in_journal_lock, flags); if (list_empty(&log->stripe_in_journal_list)) { /* all stripes flushed */
spin_unlock_irqrestore(&log->stripe_in_journal_lock, flags); return log->next_checkpoint;
}
sh = list_first_entry(&log->stripe_in_journal_list, struct stripe_head, r5c);
new_cp = sh->log_start;
spin_unlock_irqrestore(&log->stripe_in_journal_lock, flags); return new_cp;
}
/* * Starting dispatch IO to raid. * io_unit(meta) consists of a log. There is one situation we want to avoid. A * broken meta in the middle of a log causes recovery can't find meta at the * head of log. If operations require meta at the head persistent in log, we * must make sure meta before it persistent in log too. A case is: * * stripe data/parity is in log, we start write stripe to raid disks. stripe * data/parity must be persistent in log before we do the write to raid disks. * * The solution is we restrictly maintain io_unit list order. In this case, we * only write stripes of an io_unit to raid disks till the io_unit is the first * one whose data/parity is in log.
*/ void r5l_flush_stripe_to_raid(struct r5l_log *log)
{ bool do_flush;
if (!log || !log->need_cache_flush) return;
spin_lock_irq(&log->io_list_lock); /* flush bio is running */ if (!list_empty(&log->flushing_ios)) {
spin_unlock_irq(&log->io_list_lock); return;
}
list_splice_tail_init(&log->io_end_ios, &log->flushing_ios);
do_flush = !list_empty(&log->flushing_ios);
spin_unlock_irq(&log->io_list_lock);
mddev = log->rdev->mddev; /* * Discard could zero data, so before discard we must make sure * superblock is updated to new log tail. Updating superblock (either * directly call md_update_sb() or depend on md thread) must hold * reconfig mutex. On the other hand, raid5_quiesce is called with * reconfig_mutex hold. The first step of raid5_quiesce() is waiting * for all IO finish, hence waiting for reclaim thread, while reclaim * thread is calling this function and waiting for reconfig mutex. So * there is a deadlock. We workaround this issue with a trylock. * FIXME: we could miss discard if we can't take reconfig mutex
*/
set_mask_bits(&mddev->sb_flags, 0,
BIT(MD_SB_CHANGE_DEVS) | BIT(MD_SB_CHANGE_PENDING)); if (!mddev_trylock(mddev)) return;
md_update_sb(mddev, 1);
mddev_unlock(mddev);
/* * r5c_flush_stripe moves stripe from cached list to handle_list. When called, * the stripe must be on r5c_cached_full_stripes or r5c_cached_partial_stripes. * * must hold conf->device_lock
*/ staticvoid r5c_flush_stripe(struct r5conf *conf, struct stripe_head *sh)
{
BUG_ON(list_empty(&sh->lru));
BUG_ON(!test_bit(STRIPE_R5C_CACHING, &sh->state));
BUG_ON(test_bit(STRIPE_HANDLE, &sh->state));
/* * The stripe is not ON_RELEASE_LIST, so it is safe to call * raid5_release_stripe() while holding conf->device_lock
*/
BUG_ON(test_bit(STRIPE_ON_RELEASE_LIST, &sh->state));
lockdep_assert_held(&conf->device_lock);
if (test_bit(STRIPE_R5C_PARTIAL_STRIPE, &sh->state))
atomic_inc(&conf->r5c_flushing_partial_stripes); else
atomic_inc(&conf->r5c_flushing_full_stripes);
raid5_release_stripe(sh);
}
/* * if num == 0, flush all full stripes * if num > 0, flush all full stripes. If less than num full stripes are * flushed, flush some partial stripes until totally num stripes are * flushed or there is no more cached stripes.
*/ void r5c_flush_cache(struct r5conf *conf, int num)
{ int count; struct stripe_head *sh, *next;
lockdep_assert_held(&conf->device_lock); if (!READ_ONCE(conf->log)) return;
if (total_cached > conf->min_nr_stripes * 3 / 4 ||
atomic_read(&conf->empty_inactive_list_nr) > 0) /* * if stripe cache pressure high, flush all full stripes and * some partial stripes
*/
stripes_to_flush = R5C_RECLAIM_STRIPE_GROUP; elseif (total_cached > conf->min_nr_stripes * 1 / 2 ||
atomic_read(&conf->r5c_cached_full_stripes) - flushing_full >
R5C_FULL_STRIPE_FLUSH_BATCH(conf)) /* * if stripe cache pressure moderate, or if there is many full * stripes,flush all full stripes
*/
stripes_to_flush = 0; else /* no need to flush */
stripes_to_flush = -1;
/* if log space is tight, flush stripes on stripe_in_journal_list */ if (test_bit(R5C_LOG_TIGHT, &conf->cache_state)) {
spin_lock_irqsave(&log->stripe_in_journal_lock, flags);
spin_lock(&conf->device_lock);
list_for_each_entry(sh, &log->stripe_in_journal_list, r5c) { /* * stripes on stripe_in_journal_list could be in any * state of the stripe_cache state machine. In this * case, we only want to flush stripe on * r5c_cached_full/partial_stripes. The following * condition makes sure the stripe is on one of the * two lists.
*/ if (!list_empty(&sh->lru) &&
!test_bit(STRIPE_HANDLE, &sh->state) &&
atomic_read(&sh->count) == 0) {
r5c_flush_stripe(conf, sh); if (count++ >= R5C_RECLAIM_STRIPE_GROUP) break;
}
}
spin_unlock(&conf->device_lock);
spin_unlock_irqrestore(&log->stripe_in_journal_lock, flags);
}
if (!test_bit(R5C_LOG_CRITICAL, &conf->cache_state))
r5l_run_no_space_stripes(log);
spin_lock_irq(&log->io_list_lock);
write_super = r5l_reclaimable_space(log) > log->max_free_space ||
reclaim_target != 0 || !list_empty(&log->no_space_stripes); /* * move proper io_unit to reclaim list. We should not change the order. * reclaimable/unreclaimable io_unit can be mixed in the list, we * shouldn't reuse space of an unreclaimable io_unit
*/ while (1) {
reclaimable = r5l_reclaimable_space(log); if (reclaimable >= reclaim_target ||
(list_empty(&log->running_ios) &&
list_empty(&log->io_end_ios) &&
list_empty(&log->flushing_ios) &&
list_empty(&log->finished_ios))) break;
/* * write_super will flush cache of each raid disk. We must write super * here, because the log area might be reused soon and we don't want to * confuse recovery
*/
r5l_write_super_and_discard_space(log, next_checkpoint);
if (!log) return;
r5c_do_reclaim(conf);
r5l_do_reclaim(log);
}
void r5l_wake_reclaim(struct r5l_log *log, sector_t space)
{ unsignedlong target; unsignedlongnew = (unsignedlong)space; /* overflow in theory */
if (!log) return;
target = READ_ONCE(log->reclaim_target); do { if (new < target) return;
} while (!try_cmpxchg(&log->reclaim_target, &target, new));
md_wakeup_thread(log->reclaim_thread);
}
/* don't allow write if journal disk is missing */ if (!log) return test_bit(MD_HAS_JOURNAL, &conf->mddev->flags); else return test_bit(Faulty, &log->rdev->flags);
}
#define R5L_RECOVERY_PAGE_POOL_SIZE 256
struct r5l_recovery_ctx { struct page *meta_page; /* current meta */
sector_t meta_total_blocks; /* total size of current meta and data */
sector_t pos; /* recovery position */
u64 seq; /* recovery position seq */ int data_parity_stripes; /* number of data_parity stripes */ int data_only_stripes; /* number of data_only stripes */ struct list_head cached_list;
/* * read ahead page pool (ra_pool) * in recovery, log is read sequentially. It is not efficient to * read every page with sync_page_io(). The read ahead page pool * reads multiple pages with one IO, so further log read can * just copy data from the pool.
*/ struct page *ra_pool[R5L_RECOVERY_PAGE_POOL_SIZE]; struct bio_vec ra_bvec[R5L_RECOVERY_PAGE_POOL_SIZE];
sector_t pool_offset; /* offset of first page in the pool */ int total_pages; /* total allocated pages */ int valid_pages; /* pages with valid data */
};
if (!page) break;
ctx->ra_pool[ctx->total_pages] = page;
ctx->total_pages += 1;
}
if (ctx->total_pages == 0) return -ENOMEM;
ctx->pool_offset = 0; return 0;
}
staticvoid r5l_recovery_free_ra_pool(struct r5l_log *log, struct r5l_recovery_ctx *ctx)
{ int i;
for (i = 0; i < ctx->total_pages; ++i)
put_page(ctx->ra_pool[i]);
}
/* * fetch ctx->valid_pages pages from offset * In normal cases, ctx->valid_pages == ctx->total_pages after the call. * However, if the offset is close to the end of the journal device, * ctx->valid_pages could be smaller than ctx->total_pages
*/ staticint r5l_recovery_fetch_ra_pool(struct r5l_log *log, struct r5l_recovery_ctx *ctx,
sector_t offset)
{ struct bio bio; int ret;
if (offset == 0) /* reached end of the device */ break;
}
ret = submit_bio_wait(&bio);
bio_uninit(&bio); return ret;
}
/* * try read a page from the read ahead page pool, if the page is not in the * pool, call r5l_recovery_fetch_ra_pool
*/ staticint r5l_recovery_read_page(struct r5l_log *log, struct r5l_recovery_ctx *ctx, struct page *page,
sector_t offset)
{ int ret;
if (offset < ctx->pool_offset ||
offset >= ctx->pool_offset + ctx->valid_pages * BLOCK_SECTORS) {
ret = r5l_recovery_fetch_ra_pool(log, ctx, offset); if (ret) return ret;
}
/* * r5l_recovery_load_data and r5l_recovery_load_parity uses flag R5_Wantwrite * to mark valid (potentially not flushed) data in the journal. * * We already verified checksum in r5l_recovery_verify_data_checksum_for_mb, * so there should not be any mismatch here.
*/ staticvoid r5l_recovery_load_data(struct r5l_log *log, struct stripe_head *sh, struct r5l_recovery_ctx *ctx, struct r5l_payload_data_parity *payload,
sector_t log_offset)
{ struct mddev *mddev = log->rdev->mddev; struct r5conf *conf = mddev->private; int dd_idx;
for (disk_index = 0; disk_index < sh->disks; disk_index++) { if (!test_bit(R5_Wantwrite, &sh->dev[disk_index].flags)) continue; if (disk_index == sh->qd_idx || disk_index == sh->pd_idx) continue;
data_count++;
}
/* * stripes that only have parity must have been flushed * before the crash that we are now recovering from, so * there is nothing more to recovery.
*/ if (data_count == 0) goto out;
for (disk_index = 0; disk_index < sh->disks; disk_index++) { if (!test_bit(R5_Wantwrite, &sh->dev[disk_index].flags)) continue;
/* in case device is broken */
rdev = conf->disks[disk_index].rdev; if (rdev) {
atomic_inc(&rdev->nr_pending);
sync_page_io(rdev, sh->sector, PAGE_SIZE,
sh->dev[disk_index].page, REQ_OP_WRITE, false);
rdev_dec_pending(rdev, rdev->mddev);
}
rrdev = conf->disks[disk_index].replacement; if (rrdev) {
atomic_inc(&rrdev->nr_pending);
sync_page_io(rrdev, sh->sector, PAGE_SIZE,
sh->dev[disk_index].page, REQ_OP_WRITE, false);
rdev_dec_pending(rrdev, rrdev->mddev);
}
}
ctx->data_parity_stripes++;
out:
r5l_recovery_reset_stripe(sh);
}
/* * before loading data to stripe cache, we need verify checksum for all data, * if there is mismatch for any data page, we drop all data in the mata block
*/ staticint
r5l_recovery_verify_data_checksum_for_mb(struct r5l_log *log, struct r5l_recovery_ctx *ctx)
{ struct mddev *mddev = log->rdev->mddev; struct r5conf *conf = mddev->private; struct r5l_meta_block *mb = page_address(ctx->meta_page);
sector_t mb_offset = sizeof(struct r5l_meta_block);
sector_t log_offset = r5l_ring_add(log, ctx->pos, BLOCK_SECTORS); struct page *page; struct r5l_payload_data_parity *payload; struct r5l_payload_flush *payload_flush;
page = alloc_page(GFP_KERNEL); if (!page) return -ENOMEM;
/* * Analyze all data/parity pages in one meta block * Returns: * 0 for success * -EINVAL for unknown playload type * -EAGAIN for checksum mismatch of data page * -ENOMEM for run out of memory (alloc_page failed or run out of stripes)
*/ staticint
r5c_recovery_analyze_meta_block(struct r5l_log *log, struct r5l_recovery_ctx *ctx, struct list_head *cached_stripe_list)
{ struct mddev *mddev = log->rdev->mddev; struct r5conf *conf = mddev->private; struct r5l_meta_block *mb; struct r5l_payload_data_parity *payload; struct r5l_payload_flush *payload_flush; int mb_offset;
sector_t log_offset;
sector_t stripe_sect; struct stripe_head *sh; int ret;
/* * for mismatch in data blocks, we will drop all data in this mb, but * we will still read next mb for other data with FLUSH flag, as * io_unit could finish out of order.
*/
ret = r5l_recovery_verify_data_checksum_for_mb(log, ctx); if (ret == -EINVAL) return -EAGAIN; elseif (ret) return ret; /* -ENOMEM duo to alloc_page() failed */
/* * Load the stripe into cache. The stripe will be written out later by * the stripe cache state machine.
*/ staticvoid r5c_recovery_load_one_stripe(struct r5l_log *log, struct stripe_head *sh)
{ struct r5dev *dev; int i;
for (i = sh->disks; i--; ) {
dev = sh->dev + i; if (test_and_clear_bit(R5_Wantwrite, &dev->flags)) {
set_bit(R5_InJournal, &dev->flags);
set_bit(R5_UPTODATE, &dev->flags);
}
}
}
/* * Scan through the log for all to-be-flushed data * * For stripes with data and parity, namely Data-Parity stripe * (STRIPE_R5C_CACHING == 0), we simply replay all the writes. * * For stripes with only data, namely Data-Only stripe * (STRIPE_R5C_CACHING == 1), we load them to stripe cache state machine. * * For a stripe, if we see data after parity, we should discard all previous * data and parity for this stripe, as these data are already flushed to * the array. * * At the end of the scan, we return the new journal_tail, which points to * first data-only stripe on the journal device, or next invalid meta block.
*/ staticint r5c_recovery_flush_log(struct r5l_log *log, struct r5l_recovery_ctx *ctx)
{ struct stripe_head *sh; int ret = 0;
/* scan through the log */ while (1) { if (r5l_recovery_read_meta_block(log, ctx)) break;
ret = r5c_recovery_analyze_meta_block(log, ctx,
&ctx->cached_list); /* * -EAGAIN means mismatch in data block, in this case, we still * try scan the next metablock
*/ if (ret && ret != -EAGAIN) break; /* ret == -EINVAL or -ENOMEM */
ctx->seq++;
ctx->pos = r5l_ring_add(log, ctx->pos, ctx->meta_total_blocks);
}
if (ret == -ENOMEM) {
r5c_recovery_drop_stripes(&ctx->cached_list, ctx); return ret;
}
/* * we did a recovery. Now ctx.pos points to an invalid meta block. New * log will start here. but we can't let superblock point to last valid * meta block. The log might looks like: * | meta 1| meta 2| meta 3| * meta 1 is valid, meta 2 is invalid. meta 3 could be valid. If * superblock points to meta 1, we write a new valid meta 2n. if crash * happens again, new recovery will start from meta 1. Since meta 2n is * valid now, recovery will think meta 3 is valid, which is wrong. * The solution is we create a new meta in meta2 with its seq == meta * 1's seq + 10000 and let superblock points to meta2. The same recovery * will not think meta 3 is a valid meta, because its seq doesn't match
*/
/*
--> --------------------
--> maximum size reached
--> --------------------
Messung V0.5
¤ Dauer der Verarbeitung: 0.45 Sekunden
(vorverarbeitet)
¤
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.