| commit | 4ef8cc277cd2e78923fa64a26c167483d40d7ab2 | |
| author | Paul Bissonnette <paulbiss@fb.com> | |
| Mon, 8 Jun 2015 20:48:49 +0000 (8 13:48 -0700) | ||
| committer | hhvm-bot <hhvm-bot@fb.com> | |
| Mon, 8 Jun 2015 21:02:16 +0000 (8 14:02 -0700) | ||
| tree | 36869839bf525a220f71c0b06ea29b35d1b07ad2 | treesnapshot (tar.gz zip) |
| parent | b014c3a45cdb46274b2ec754aa70057cf1bcd766 | commitdiff |
Allow inlined 86ctors
Summary: These are basically NOPs, so we should probably inline them (if you
declare an empty constructor it gets inlined). Also updates some state tracking
to record FPushCtorD frames.
I didn't see much of a change in perlab but this seems like something we might
want to do anyway.
https://our.intern.facebook.com/intern/perflab/details/910450
https://our.intern.facebook.com/intern/perflab/details/910452
https://our.intern.facebook.com/intern/perflab/details/910453
Reviewed By: @jdelong
Differential Revision: D1540133
Summary: These are basically NOPs, so we should probably inline them (if you
declare an empty constructor it gets inlined). Also updates some state tracking
to record FPushCtorD frames.
I didn't see much of a change in perlab but this seems like something we might
want to do anyway.
https://our.intern.facebook.com/intern/perflab/details/910450
https://our.intern.facebook.com/intern/perflab/details/910452
https://our.intern.facebook.com/intern/perflab/details/910453
Reviewed By: @jdelong
Differential Revision: D1540133