fork: Stop protecting back_fork_cleanup_cgroup_lock with CONFIG_NUMA
authorEric W. Biederman <ebiederm@xmission.com>
Mon, 20 Dec 2021 16:20:14 +0000 (10:20 -0600)
committerEric W. Biederman <ebiederm@xmission.com>
Mon, 20 Dec 2021 16:50:51 +0000 (10:50 -0600)
commit6692c98c7df53502adb8b8b73ab9bcbd399f7a06
treea4f0e5eaff118a71b726e2aa0675568fab9ac8e5
parent1fb466dff904e4a72282af336f2c355f011eec61
fork: Stop protecting back_fork_cleanup_cgroup_lock with CONFIG_NUMA

Mark Brown <broonie@kernel.org> reported:

> This is also causing further build errors including but not limited to:
>
> /tmp/next/build/kernel/fork.c: In function 'copy_process':
> /tmp/next/build/kernel/fork.c:2106:4: error: label 'bad_fork_cleanup_threadgroup_lock' used but not defined
>  2106 |    goto bad_fork_cleanup_threadgroup_lock;
>       |    ^~~~

It turns out that I messed up and was depending upon a label protected
by an ifdef.  Move the label out of the ifdef as the ifdef around the label
no longer makes sense (if it ever did).

Link: https://lkml.kernel.org/r/YbugCP144uxXvRsk@sirena.org.uk
Fixes: 40966e316f86 ("kthread: Ensure struct kthread is present for all kthreads")
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
kernel/fork.c