Wei Yongjun fixed a long standing bug in the ring buffer startup test.

If for some unknown reason, the kthread that is created fails to be
 created, the return from kthread_create() is an PTR_ERR and not a NULL.
 The test incorrectly checks for NULL instead of an error.
 -----BEGIN PGP SIGNATURE-----
 
 iQExBAABCAAbBQJY5mOWFBxyb3N0ZWR0QGdvb2RtaXMub3JnAAoJEMm5BfJq2Y3L
 XcsH/iBX7Kf4ta/0Jo4+sR4+HeDmWNPVBTwlei+dvMfaK1rWDgW6hbwSJg3geUwN
 d2zL/o7uCWbXubO9sjeCX2n+ecUiUcJRheewfdm0KzaPH387ofdUd24yz3DNDNcl
 /yaZMmeApjpHJjJWxoH5TUSF/yliC2FvjHYWxgEx9qhrzldLk/r5qAealj2tKl1Q
 1cgSQEgXf5n6Wg0onBuR2JiMOo3+4lXh+pIpO1Dupalhj7cC91HatDDYrNmGRIWR
 qucf3iQLoD/m88bgpxsRortkQ09NfVJExxzIPliVoYF8VwtzL+77XD81EdgvLdTs
 WP+CAoMFk83fkuXK7Vg1HZZa5zg=
 =Z0D5
 -----END PGP SIGNATURE-----

Merge tag 'trace-v4.11-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace

Pull tracing fix from Steven Rostedt:
 "Wei Yongjun fixed a long standing bug in the ring buffer startup test.

  If for some unknown reason, the kthread that is created fails to be
  created, the return from kthread_create() is an PTR_ERR and not a
  NULL. The test incorrectly checks for NULL instead of an error"

* tag 'trace-v4.11-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
  ring-buffer: Fix return value check in test_ringbuffer()
This commit is contained in:
Linus Torvalds 2017-04-06 13:12:12 -07:00
commit 4691f4a6d4

View File

@ -4826,9 +4826,9 @@ static __init int test_ringbuffer(void)
rb_data[cpu].cnt = cpu;
rb_threads[cpu] = kthread_create(rb_test, &rb_data[cpu],
"rbtester/%d", cpu);
if (WARN_ON(!rb_threads[cpu])) {
if (WARN_ON(IS_ERR(rb_threads[cpu]))) {
pr_cont("FAILED\n");
ret = -1;
ret = PTR_ERR(rb_threads[cpu]);
goto out_free;
}
@ -4838,9 +4838,9 @@ static __init int test_ringbuffer(void)
/* Now create the rb hammer! */
rb_hammer = kthread_run(rb_hammer_test, NULL, "rbhammer");
if (WARN_ON(!rb_hammer)) {
if (WARN_ON(IS_ERR(rb_hammer))) {
pr_cont("FAILED\n");
ret = -1;
ret = PTR_ERR(rb_hammer);
goto out_free;
}