diff options
| author | Chandler Carruth <chandlerc@gmail.com> | 2014-02-01 10:38:17 +0000 |
|---|---|---|
| committer | Chandler Carruth <chandlerc@gmail.com> | 2014-02-01 10:38:17 +0000 |
| commit | 6b4cc8b66a2e7a07cc4d1064200c245597743b88 (patch) | |
| tree | 2aee15bb52bead910729a7b8d22dc4aab2b50866 /llvm/lib/MC/MCContext.cpp | |
| parent | da7160d6d2af892889cf051987b025310178245d (diff) | |
| download | bcm5719-llvm-6b4cc8b66a2e7a07cc4d1064200c245597743b88.tar.gz bcm5719-llvm-6b4cc8b66a2e7a07cc4d1064200c245597743b88.zip | |
[inliner] Skip debug intrinsics even earlier in computing the inline
cost so that they don't impact the vector bonus. Fundamentally, counting
unsimplified instructions is just *wrong*; it will continue to introduce
instability as things which do not generate code bizarrely impact
inlining. For example, sufficiently nested inlined functions could turn
off the vector bonus with lifetime markers just like the debug
intrinsics do. =/
This is a short-term tactical fix. Long term, I think we need to remove
the vector bonus entirely. That's a separate patch and discussion
though.
The patch to fix this provided by Dario Domizioli. I've added some
comments about the planned direction and used a heavily pruned form of
debug info intrinsics for the test case. While this debug info doesn't
work or "do" anything useful, it lets us easily test all manner of
interference easily, and I suspect this will not be the last time we
want to craft a pattern where debug info interferes with the inliner in
a problematic way.
llvm-svn: 200609
Diffstat (limited to 'llvm/lib/MC/MCContext.cpp')
0 files changed, 0 insertions, 0 deletions
