[v1] futex: Use correct exit on failure from futex_hash_allocate_default()

[PATCH] futex: Use correct exit on failure from futex_hash_allocate_default()

Posted by Sebastian Andrzej Siewior 4 months, 3 weeks ago

copy_process() uses the wrong error exit path from
futex_hash_allocate_default().
After exiting from futex_hash_allocate_default(), neither tasklist_lock
nor siglock has been acquired. The exit label bad_fork_core_free unlocks
both of these locks which is wrong.

The previous label, bad_fork_cancel_cgroup, is the correct exit.
sched_cgroup_fork() did not allocate any resources that need to freed.

Use bad_fork_cancel_cgroup on error exit from
futex_hash_allocate_default().

Fixes: 7c4f75a21f636 ("futex: Allow automatic allocation of process wide futex hash")
Reported-by: syzbot+80cb3cc5c14fad191a10@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/68cb1cbd.050a0220.2ff435.0599.GAE@google.com
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
---

That private-futex code was marked BROKEN in v6.16 and re-enabled in
v6.17. It could use
  56180dd20c19e ("futex: Use RCU-based per-CPU reference counting instead of rcuref_t")

as Fixes: instead to avoid backporting to v6.16.

 kernel/fork.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/fork.c b/kernel/fork.c
index c4ada32598bd5..6ca8689a83b5b 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -2295,7 +2295,7 @@ __latent_entropy struct task_struct *copy_process(
 	if (need_futex_hash_allocate_default(clone_flags)) {
 		retval = futex_hash_allocate_default();
 		if (retval)
-			goto bad_fork_core_free;
+			goto bad_fork_cancel_cgroup;
 		/*
 		 * If we fail beyond this point we don't free the allocated
 		 * futex hash map. We assume that another thread will be created
-- 
2.51.0

Re: [PATCH] futex: Use correct exit on failure from futex_hash_allocate_default()

Posted by Steven Rostedt 4 months, 3 weeks ago

On Thu, 18 Sep 2025 15:09:45 +0200
Sebastian Andrzej Siewior <bigeasy@linutronix.de> wrote:

> copy_process() uses the wrong error exit path from
> futex_hash_allocate_default().
> After exiting from futex_hash_allocate_default(), neither tasklist_lock
> nor siglock has been acquired. The exit label bad_fork_core_free unlocks
> both of these locks which is wrong.
> 
> The previous label, bad_fork_cancel_cgroup, is the correct exit.
> sched_cgroup_fork() did not allocate any resources that need to freed.
> 
> Use bad_fork_cancel_cgroup on error exit from
> futex_hash_allocate_default().

	if (need_futex_hash_allocate_default(clone_flags)) {
		retval = futex_hash_allocate_default();
		if (retval)
			goto bad_fork_core_free;
		[..]
	}
	[..]
	write_lock_irq(&tasklist_lock);
	[..]
	klp_copy_process(p);

	sched_core_fork(p);

	spin_lock(&current->sighand->siglock);

	[..]

 bad_fork_core_free:
	sched_core_free(p);
	spin_unlock(&current->sighand->siglock);
	write_unlock_irq(&tasklist_lock);
 bad_fork_cancel_cgroup:
	cgroup_cancel_fork(p, args);

Yep, looks bad to me!

Reviewed-by: Steven Rostedt (Google) <rostedt@goodmis.org>

-- Steve


> 
> Fixes: 7c4f75a21f636 ("futex: Allow automatic allocation of process wide futex hash")
> Reported-by: syzbot+80cb3cc5c14fad191a10@syzkaller.appspotmail.com
> Closes: https://lore.kernel.org/all/68cb1cbd.050a0220.2ff435.0599.GAE@google.com
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
>

[tip: locking/urgent] futex: Use correct exit on failure from futex_hash_allocate_default()

Posted by tip-bot2 for Sebastian Andrzej Siewior 4 months, 2 weeks ago

The following commit has been merged into the locking/urgent branch of tip:

Commit-ID:     4ec3c15462b9f44562f45723a92e2807746ba7d1
Gitweb:        https://git.kernel.org/tip/4ec3c15462b9f44562f45723a92e2807746ba7d1
Author:        Sebastian Andrzej Siewior <bigeasy@linutronix.de>
AuthorDate:    Thu, 18 Sep 2025 15:09:45 +02:00
Committer:     Thomas Gleixner <tglx@linutronix.de>
CommitterDate: Wed, 24 Sep 2025 09:20:02 +02:00

futex: Use correct exit on failure from futex_hash_allocate_default()

copy_process() uses the wrong error exit path from futex_hash_allocate_default().
After exiting from futex_hash_allocate_default(), neither tasklist_lock
nor siglock has been acquired. The exit label bad_fork_core_free unlocks
both of these locks which is wrong.

The next exit label, bad_fork_cancel_cgroup, is the correct exit.
sched_cgroup_fork() did not allocate any resources that need to freed.

Use bad_fork_cancel_cgroup on error exit from futex_hash_allocate_default().

Fixes: 7c4f75a21f636 ("futex: Allow automatic allocation of process wide futex hash")
Reported-by: syzbot+80cb3cc5c14fad191a10@syzkaller.appspotmail.com
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Steven Rostedt (Google) <rostedt@goodmis.org>
Closes: https://lore.kernel.org/all/68cb1cbd.050a0220.2ff435.0599.GAE@google.com
---
 kernel/fork.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/fork.c b/kernel/fork.c
index c4ada32..6ca8689 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -2295,7 +2295,7 @@ __latent_entropy struct task_struct *copy_process(
 	if (need_futex_hash_allocate_default(clone_flags)) {
 		retval = futex_hash_allocate_default();
 		if (retval)
-			goto bad_fork_core_free;
+			goto bad_fork_cancel_cgroup;
 		/*
 		 * If we fail beyond this point we don't free the allocated
 		 * futex hash map. We assume that another thread will be created