On 5/7/24 12:54 AM, Ryan Roberts wrote:
On 05/05/2024 23:13, John Hubbard wrote:
...
diff --git a/tools/testing/selftests/mqueue/mq_perf_tests.c b/tools/testing/selftests/mqueue/mq_perf_tests.c
index 5c16159d0bcd..fb898850867c 100644
--- a/tools/testing/selftests/mqueue/mq_perf_tests.c
+++ b/tools/testing/selftests/mqueue/mq_perf_tests.c
@@ -323,7 +323,8 @@ void *fake_cont_thread(void *arg)
void *cont_thread(void *arg)
{
char buff[MSG_SIZE];
- int i, priority;
+ int i;
+ unsigned int priority;
for (i = 0; i < num_cpus_to_pin; i++)
if (cpu_threads[i] == pthread_self())
@@ -425,7 +426,8 @@ struct test test2[] = {
void *perf_test_thread(void *arg)
{
char buff[MSG_SIZE];
- int prio_out, prio_in;
+ int prio_out;
It feels a bit odd for prio_out and prio_in to have different types. I don't
have any prior familiararity with these tests but looks like they are ultimately
the parameters of mq_send() and mq_receive() which both define them as unsigned
ints. Perhaps both should be converted?
This makes sense, and I recall wondering about it. Looking at it again,
I see why didn't go that far: there is a mini-unit test manager inside,
passing around priorities that are signed, throughout:
struct test {
char *desc; void (*func)(int *);
};
...
void inc_prio(int *prio) {
if (++*prio == mq_prio_max)
*prio = 0;
}
However, I can probably fix up everything to match up. Given that you've
called it out, I'll go ahead with that approach. Iit will be quite a few
changes but they will all be trivial too.
thanks,
--
John Hubbard
NVIDIA
+ unsigned int prio_in;
int i;
clockid_t clock;
pthread_t *t;
base-commit: f462ae0edd3703edd6f22fe41d336369c38b884b
prerequisite-patch-id: b901ece2a5b78503e2fb5480f20e304d36a0ea27
