percpu-refcount: add @gfp to percpu_ref_init()

Percpu allocator now supports allocation mask.  Add @gfp to
percpu_ref_init() so that !GFP_KERNEL allocation masks can be used
with percpu_refs too.

This patch doesn't make any functional difference.

v2: blk-mq conversion was missing.  Updated.

Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Benjamin LaHaise <bcrl@kvack.org>
Cc: Li Zefan <lizefan@huawei.com>
Cc: Nicholas A. Bellinger <nab@linux-iscsi.org>
Cc: Jens Axboe <axboe@kernel.dk>
diff --git a/block/blk-mq.c b/block/blk-mq.c
index 5189cb1..702df07 100644
--- a/block/blk-mq.c
+++ b/block/blk-mq.c
@@ -1776,7 +1776,8 @@
 	if (!q)
 		goto err_hctxs;
 
-	if (percpu_ref_init(&q->mq_usage_counter, blk_mq_usage_counter_release))
+	if (percpu_ref_init(&q->mq_usage_counter, blk_mq_usage_counter_release,
+			    GFP_KERNEL))
 		goto err_map;
 
 	setup_timer(&q->timeout, blk_mq_rq_timer, (unsigned long) q);
diff --git a/drivers/target/target_core_tpg.c b/drivers/target/target_core_tpg.c
index fddfae6..4ab6da3 100644
--- a/drivers/target/target_core_tpg.c
+++ b/drivers/target/target_core_tpg.c
@@ -819,7 +819,8 @@
 {
 	int ret;
 
-	ret = percpu_ref_init(&lun->lun_ref, core_tpg_lun_ref_release);
+	ret = percpu_ref_init(&lun->lun_ref, core_tpg_lun_ref_release,
+			      GFP_KERNEL);
 	if (ret < 0)
 		return ret;
 
diff --git a/fs/aio.c b/fs/aio.c
index bd7ec2c..93fbcc0f 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -666,10 +666,10 @@
 
 	INIT_LIST_HEAD(&ctx->active_reqs);
 
-	if (percpu_ref_init(&ctx->users, free_ioctx_users))
+	if (percpu_ref_init(&ctx->users, free_ioctx_users, GFP_KERNEL))
 		goto err;
 
-	if (percpu_ref_init(&ctx->reqs, free_ioctx_reqs))
+	if (percpu_ref_init(&ctx->reqs, free_ioctx_reqs, GFP_KERNEL))
 		goto err;
 
 	ctx->cpu = alloc_percpu(struct kioctx_cpu);
diff --git a/include/linux/percpu-refcount.h b/include/linux/percpu-refcount.h
index 3dfbf23..ee83251 100644
--- a/include/linux/percpu-refcount.h
+++ b/include/linux/percpu-refcount.h
@@ -49,6 +49,7 @@
 #include <linux/kernel.h>
 #include <linux/percpu.h>
 #include <linux/rcupdate.h>
+#include <linux/gfp.h>
 
 struct percpu_ref;
 typedef void (percpu_ref_func_t)(struct percpu_ref *);
@@ -66,7 +67,7 @@
 };
 
 int __must_check percpu_ref_init(struct percpu_ref *ref,
-				 percpu_ref_func_t *release);
+				 percpu_ref_func_t *release, gfp_t gfp);
 void percpu_ref_reinit(struct percpu_ref *ref);
 void percpu_ref_exit(struct percpu_ref *ref);
 void percpu_ref_kill_and_confirm(struct percpu_ref *ref,
diff --git a/kernel/cgroup.c b/kernel/cgroup.c
index 7dc8788..589b4d8 100644
--- a/kernel/cgroup.c
+++ b/kernel/cgroup.c
@@ -1628,7 +1628,7 @@
 		goto out;
 	root_cgrp->id = ret;
 
-	ret = percpu_ref_init(&root_cgrp->self.refcnt, css_release);
+	ret = percpu_ref_init(&root_cgrp->self.refcnt, css_release, GFP_KERNEL);
 	if (ret)
 		goto out;
 
@@ -4487,7 +4487,7 @@
 
 	init_and_link_css(css, ss, cgrp);
 
-	err = percpu_ref_init(&css->refcnt, css_release);
+	err = percpu_ref_init(&css->refcnt, css_release, GFP_KERNEL);
 	if (err)
 		goto err_free_css;
 
@@ -4555,7 +4555,7 @@
 		goto out_unlock;
 	}
 
-	ret = percpu_ref_init(&cgrp->self.refcnt, css_release);
+	ret = percpu_ref_init(&cgrp->self.refcnt, css_release, GFP_KERNEL);
 	if (ret)
 		goto out_free_cgrp;
 
diff --git a/lib/percpu-refcount.c b/lib/percpu-refcount.c
index fe5a334..ff99032 100644
--- a/lib/percpu-refcount.c
+++ b/lib/percpu-refcount.c
@@ -40,6 +40,7 @@
  * percpu_ref_init - initialize a percpu refcount
  * @ref: percpu_ref to initialize
  * @release: function which will be called when refcount hits 0
+ * @gfp: allocation mask to use
  *
  * Initializes the refcount in single atomic counter mode with a refcount of 1;
  * analagous to atomic_set(ref, 1).
@@ -47,11 +48,12 @@
  * Note that @release must not sleep - it may potentially be called from RCU
  * callback context by percpu_ref_kill().
  */
-int percpu_ref_init(struct percpu_ref *ref, percpu_ref_func_t *release)
+int percpu_ref_init(struct percpu_ref *ref, percpu_ref_func_t *release,
+		    gfp_t gfp)
 {
 	atomic_set(&ref->count, 1 + PCPU_COUNT_BIAS);
 
-	ref->pcpu_count_ptr = (unsigned long)alloc_percpu(unsigned);
+	ref->pcpu_count_ptr = (unsigned long)alloc_percpu_gfp(unsigned, gfp);
 	if (!ref->pcpu_count_ptr)
 		return -ENOMEM;