On 8 February 2012 04:09, Stas Malyshev <smalys...@sugarcrm.com> wrote: >> Specifically, how about the attached patch (also available at >> http://www.adamharvey.name/patches/osx-max_execution_time-test-patch.diff, >> in case the mailing list decides to strip it)? This only patches the >> test to XFAIL this test on OS X, which makes the PHP_5_4 test suite >> run without any unexpected failures for me on a default build. >> >> In the longer term, I'll still work up a patch for trunk that at least >> attempts to deal with this on a ZE level, but if the test patch is OK >> for 5.4, I can apply that straightaway. > > > I'm not sure it's a proper approach - as the correct way should be to fix it > so it works properly on Mac in the core, and then we'd have to re-merge the > tests again. Since the code of the test is identical for both, I'm not sure > splitting the test is the right idea just because it fails on one platform.
We're in violent agreement. I don't think it's the right approach at all — the right fix is to either fall back on ITIMER_REAL or come up with some other approach on OS X — but I'm trying to come up with a stopgap for 5.4.0 to make sure the test is XFAILed where we already know it fails as part of the "let's actually make the test suite meaningful" push, since we've been living with this in 5.3 for ages anyway and I don't think this is anywhere near critical enough to tinker with 5.4 during a code freeze. Adam -- PHP Internals - PHP Runtime Development Mailing List To unsubscribe, visit: http://www.php.net/unsub.php
