[PATCH] cgroup: simplify code for cgroup_subtree_control_write()

[PATCH] cgroup: simplify code for cgroup_subtree_control_write()

[Peng Wang]

Process "enable" and "disable" earlier to simplify code.

Signed-off-by: Peng Wang <rocking@whu.edu.cn>
---
 kernel/cgroup/cgroup.c | 23 ++++++++---------------
 1 file changed, 8 insertions(+), 15 deletions(-)

diff --git a/kernel/cgroup/cgroup.c b/kernel/cgroup/cgroup.c
index bf9dbffd46b1..e49b8bde5c99 100644
--- a/kernel/cgroup/cgroup.c
+++ b/kernel/cgroup/cgroup.c
@@ -3269,23 +3269,21 @@ static ssize_t cgroup_subtree_control_write(struct kernfs_open_file *of,
 	if (!cgrp)
 		return -ENODEV;
 
+	enable &= ~(cgrp->subtree_control);
+	disable &= cgrp->subtree_control;
+
+	if (!enable && !disable) {
+		ret = 0;
+		goto out_unlock;
+	}
+
 	for_each_subsys(ss, ssid) {
 		if (enable & (1 << ssid)) {
-			if (cgrp->subtree_control & (1 << ssid)) {
-				enable &= ~(1 << ssid);
-				continue;
-			}
-
 			if (!(cgroup_control(cgrp) & (1 << ssid))) {
 				ret = -ENOENT;
 				goto out_unlock;
 			}
 		} else if (disable & (1 << ssid)) {
-			if (!(cgrp->subtree_control & (1 << ssid))) {
-				disable &= ~(1 << ssid);
-				continue;
-			}
-
 			/* a child has it enabled? */
 			cgroup_for_each_live_child(child, cgrp) {
 				if (child->subtree_control & (1 << ssid)) {
@@ -3296,11 +3294,6 @@ static ssize_t cgroup_subtree_control_write(struct kernfs_open_file *of,
 		}
 	}
 
-	if (!enable && !disable) {
-		ret = 0;
-		goto out_unlock;
-	}
-
 	ret = cgroup_vet_subtree_control_enable(cgrp, enable);
 	if (ret)
 		goto out_unlock;
-- 
2.19.1

Re: [PATCH] cgroup: simplify code for cgroup_subtree_control_write()

[Tejun Heo]

On Mon, Jul 08, 2019 at 09:01:32PM +0800, Peng Wang wrote:
> Process "enable" and "disable" earlier to simplify code.

I don't think this is correct and even if it were the value of this
change is close to none, so nack on this one.

Thanks.

-- 
tejun

RE: [PATCH] cgroup: simplify code for cgroup_subtree_control_write()

[Peng Wang]

On Monday, July 8, 2019 10:41 PM, Tejun Heo wrote:
> On Mon, Jul 08, 2019 at 09:01:32PM +0800, Peng Wang wrote:
> > Process "enable" and "disable" earlier to simplify code.
> 
> I don't think this is correct and even if it were the value of this
> change is close to none, so nack on this one.

OK. 
Although I have tested this patch for common case, It's just somewhat trivial.

Thanks for your time :-).

> 
> Thanks.
> 
> --
> tejun