LOCK COUNTERS.(9)

💅 Lock Counters. 💅

okay bestie, welcome to Lock Counters.(9) in the Kernel API Manual — this one is short and honestly kinda chic. A lock_counter is intended to keep all of the locks for the blocks in the recovery journal. That's the whole assignment. 💖

All the per-zone counters are kept in a single array, and that array is arranged by zone — i.e. zone 0's lock 0 is at index 0, zone 0's lock 1 is at index 1, and zone 1's lock 0 is at index 'locks'. The way this arrangement is set up is intended to minimize cache-line contention for counters from different zones. lowkey iconic layout behavior. ✨

The locks are implemented as a single object instead of as a lock counter per lock. Why? two reasons, both valid: it affords this opportunity to reduce cache line contention, and it also eliminates the need to have a completion per lock. no extra completion per lock, we love that for us. 🎀

Lock sets are laid out with the set for recovery journal first, followed by the logical zones, and then the physical zones. that's the order, periodt. 🦋

🌸 SEE ALSO 🌸

Kernel file: ./drivers/md/dm-vdo/recovery-journal.c — yes, that exact cutie. 🫶

The whole related-pages squad is literally:

• get_zone_count_ptr(9), get_counter(9), is_journal_zone_locked(9), vdo_release_recovery_journal_block_reference(9), pop_free_list(9)
• is_block_dirty(9), is_block_empty(9), is_block_full(9), assert_on_journal_thread(9), continue_waiter(9)
• has_block_waiters(9), suspend_lock_counter(9), check_for_drain_complete(9), notify_recovery_journal_of_read_only_mode(9), enter_journal_read_only_mode(9)
• vdo_get_recovery_journal_current_sequence_number(9), get_recovery_journal_head(9), compute_recovery_count_byte(9), check_slab_journal_commit_threshold(9), finish_reaping(9)
• complete_reaping(9), handle_flush_error(9), initialize_journal_state(9), vdo_get_recovery_journal_length(9), reap_recovery_journal_callback(9)
• initialize_lock_counter(9), set_journal_tail(9), initialize_recovery_block(9), vdo_decode_recovery_journal(9), vdo_free_recovery_journal(9)
• vdo_initialize_recovery_journal_post_repair(9), vdo_get_journal_block_map_data_blocks_used(9), vdo_get_recovery_journal_thread_id(9), vdo_open_recovery_journal(9), vdo_record_recovery_journal(9)
• get_block_header(9), set_active_sector(9), advance_tail(9), initialize_lock_count(9), prepare_to_assign_entry(9)
• schedule_block_write(9), release_journal_block_reference(9), assign_entry(9), recycle_journal_block(9), continue_committed_waiter(9)
• notify_commit_waiters(9), recycle_journal_blocks(9), complete_write(9), add_queued_recovery_entries(9), write_block(9)
• write_blocks(9), vdo_add_recovery_journal_entry(9), is_lock_locked(9), reap_recovery_journal(9), vdo_acquire_recovery_journal_block_reference(9)
• vdo_release_journal_entry_lock(9), vdo_drain_recovery_journal(9), resume_lock_counter(9), vdo_resume_recovery_journal(9), vdo_get_recovery_journal_logical_blocks_used(9)
• vdo_get_recovery_journal_statistics(9), dump_recovery_block(9), vdo_dump_recovery_journal_statistics(9)

mwah 💋 — this page comes from ./drivers/md/dm-vdo, dated October 2026, and it's literally Lock Counters.(9). bye bestie, go slay those locks 🌈👑