percpu_ida: fix a live lock

steal_tags only happens when free tags is more than half of the total
tags.  This is too strict and can cause live lock. I found that if one
cpu has free tags, but other cpu can't steal (thread is bound to
specific cpus), threads which want to allocate tags are always
sleeping. I found this when I run next patch, but this could happen
without it I think.

I did performance test too with null_blk. Two cases (each cpu has enough
percpu tags, or total tags are limited), no performance changes were
observed.

Signed-off-by: Shaohua Li <shli@fusionio.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

diff --git a/lib/percpu_ida.c b/lib/percpu_ida.c
index 9d054bf91d0f3cd278494dedd3114dfd3b292bcb..85f43b1718d4451115909cb6aa5e20d387c3fd6e 100644 (file)
--- a/lib/percpu_ida.c
+++ b/lib/percpu_ida.c
@@ -54,9 +54,7 @@ static inline void move_tags(unsigned *dst, unsigned *dst_nr,
 /*
  * Try to steal tags from a remote cpu's percpu freelist.
  *
- * We first check how many percpu freelists have tags - we don't steal tags
- * unless enough percpu freelists have tags on them that it's possible more than
- * half the total tags could be stuck on remote percpu freelists.
+ * We first check how many percpu freelists have tags
  *
  * Then we iterate through the cpus until we find some tags - we don't attempt
  * to find the "best" cpu to steal from, to keep cacheline bouncing to a
@@ -69,8 +67,7 @@ static inline void steal_tags(struct percpu_ida *pool,
        struct percpu_ida_cpu *remote;
 
        for (cpus_have_tags = cpumask_weight(&pool->cpus_have_tags);
-            cpus_have_tags * pool->percpu_max_size > pool->nr_tags / 2;
-            cpus_have_tags--) {
+            cpus_have_tags; cpus_have_tags--) {
                cpu = cpumask_next(cpu, &pool->cpus_have_tags);
 
                if (cpu >= nr_cpu_ids) {