| Message ID | 20181010114136.23957-1-rafael.tinoco@linaro.org |
|---|---|
| State | Accepted |
| Commit | 3c1f4f5c4a86cbff015f046d86ef6e03f1b262ba |
| Headers | show |
| Series | [v2,1/2] futex/futex_wake04.c: fix hugepages setup for test | expand |
diff --git a/testcases/kernel/syscalls/futex/futex_wake04.c b/testcases/kernel/syscalls/futex/futex_wake04.c index 3c7038317..1ae9599ab 100644 --- a/testcases/kernel/syscalls/futex/futex_wake04.c +++ b/testcases/kernel/syscalls/futex/futex_wake04.c @@ -76,14 +76,17 @@ static void setup(void) tst_tmpdir(); SAFE_FILE_SCANF(NULL, PATH_NR_HUGEPAGES, "%ld", &orig_hugepages); - SAFE_FILE_PRINTF(NULL, PATH_NR_HUGEPAGES, "%d", 1); + + if (!orig_hugepages) + SAFE_FILE_PRINTF(NULL, PATH_NR_HUGEPAGES, "%d", 1); TEST_PAUSE; } static void cleanup(void) { - SAFE_FILE_PRINTF(NULL, PATH_NR_HUGEPAGES, "%ld", orig_hugepages); + if (!orig_hugepages) + SAFE_FILE_PRINTF(NULL, PATH_NR_HUGEPAGES, "%d", 0); tst_rmdir(); }
Test was changing number of hugepages during setup()/cleanup() phase, despite system having (or not) already configured hugepages. This was causing some overhead - of destroying (or creating) hugepages - during the test execution, and also changing amount of available hugepages to the rest of the system. Now, if system has > 0 hugepages available, the test doesn't touch it. Link: https://bugs.linaro.org/show_bug.cgi?id=3984 Signed-off-by: Rafael David Tinoco <rafael.tinoco@linaro.org> --- testcases/kernel/syscalls/futex/futex_wake04.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-)