[PATCH v5 0/4] Add system-wide tunables

[Michael Jeanson]

On 2026-02-12 13:41, DJ Delorie wrote:
> Michael Jeanson <mjeanson@efficios.com> writes:
>> On 2026-02-11 00:02, DJ Delorie wrote:
>>>
>>> This patch series is the initial implementation of system-wide
>>> tunables, which I mentioned a while ago.
>>
>> This is a very open question but what would the expected behavior of
>> tests that use tunables be when a system-wide policy is present?
>>
>> Like for example some rseq tests assume that the "glibc.pthread.rseq"
>> tunable is on by default and some other tests will explicitly use the
>> tunable to disable the feature. If the test suite runs on a system that
>> has a system-wide policy enabling or disabling "glibc.pthread.rseq" with
>> overridability set to "-" should those tests be skipped or failed?
> 
> Those tests could be run in a container, I suppose, so that they are not
> subject to the host's configuration.
> 
> It would be prudent to add checks to such programs to ensure the tunable
> is set as expected, though, and skipped (with a suitable note) if the
> tunable was not set.

So most tests that use GLIBC_TUNABLES should be modified to explictly 
check that the tunables they depend on are set to the expected values?

For example in the rseq-disabled test I could add something like this at 
the beginning of the test:

static void
check_rseq_tunable_disabled (void)
{
   bool rseq_tunable = TUNABLE_GET_FULL (glibc, pthread, rseq, int, NULL);

   if (rseq_tunable)
     FAIL_UNSUPPORTED ("rseq tunable forced to enabled, skipping test");
}