ARC: Increase parallel eviction batching

Before parallel eviction implementation zfs_arc_evict_batch_limit
caused loop exits after evicting 10 headers.  The cost of it is not
big and well motivated.  Now though taskq task exit after the same
10 headers is much more expensive.  To cover the context switch
overhead of taskq introduce another level of batching, controlled
by zfs_arc_evict_batches_limit tunable, used only for parallel
eviction.

My tests including 36 parallel reads with 4KB recordsize that shown
1.4GB/s (~460K blocks/s) before with heavy arc_evict_lock contention,
now show 6.5GB/s (~1.6M blocks/s) without arc_evict_lock contention.

Reviewed-by: Tony Hutter <hutter2@llnl.gov>
Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Signed-off-by: Alexander Motin <alexander.motin@TrueNAS.com>
Closes #17970

man/man4/zfs.4

@@ -771,6 +771,12 @@ Number ARC headers to evict per sub-list before proceeding to another sub-list.
 This batch-style operation prevents entire sub-lists from being evicted at once
 but comes at a cost of additional unlocking and locking.
 .
+.It Sy zfs_arc_evict_batches_limit Ns = Ns Sy 5 Pq uint
+Number of
+.Sy zfs_arc_evict_batch_limit
+batches to process per parallel eviction task under heavy load to reduce number
+of context switches.
+.
 .It Sy zfs_arc_evict_threads Ns = Ns Sy 0 Pq int
 Sets the number of ARC eviction threads to be used.
 .Pp