- User Since
- Jul 8 2015, 10:26 AM (123 w, 4 d)
Wed, Nov 15
Tue, Nov 14
- Add a release note, and rename the C APIs so they use 'WithLoc' instead of '2'.
Thanks for the review Alex. There hasn't been any more feedback so I'll commit this soon.
After some consideration, I no longer think that we should provide an update tool or take the one-shot approach.
Mon, Nov 13
- Address review feedback from Adrian.
- Pass DebugLoc by value, and add a comment explaining what the PrintTo test helper does.
- Address review feedback from Quentin and Adrian.
- Add test coverage for the logic in GenerateIVChain, and for more users of LSRInstance::Expand.
- Upload a diff with context.
Fri, Nov 10
Thu, Nov 9
- Consider logical nots and some binary comparison operators in the hash, per an offline conversation with @bogner
Tue, Nov 7
This looks good to me, but it'd be worth getting another +1.
Mon, Nov 6
Fri, Nov 3
Wed, Nov 1
- Clarify comment.
- Handle loop nesting, conditions, and out-of-order control flow.
- Improve test coverage. Add a format compatibility test, and check that functions which were previously hashed the same way get different hashes now.
Tue, Oct 31
Apart from Alex's point, this patch is missing a compatibility test for v5 of the profile format, and is missing coverage for the new entries in HashType.
Mon, Oct 30
Fri, Oct 27
Thu, Oct 26
- Tighten test case to show the dbg loc on return instructions.
Wed, Oct 25
- Simplify the checks by not hardcoding a line number.
Tue, Oct 24
@mgrang thanks, should be fixed in r316490.
Mon, Oct 23
Oct 18 2017
- Address review feedback, and set aside the dwarfdump changes for later.
Oct 13 2017
D38913 should make this unnecessary.
Sounds good. This doesn't seem too controversial, since it just takes us back to the old behavior on all platforms except Darwin. I'll wait an hour or so before committing in case there are any more comments.
Ah, sorry I missed the path issue. This looks good.
Oct 12 2017
Looks great. Thank you very much!
llvm-profdata is tightly coupled with the host compiler: while this setup may work if you get lucky, I don't think it's resilient to changes in libProfData. Also, using the instrumented llvm-profdata will be slow and create extra profiles.
- Update to check against a whitelist of supported targets.
Oct 10 2017
For those following along, Alex worked out that this doesn't affect apple-clang 802. We took a closer look and found that the build break just affects clang-900, and was introduced in this r290889. The fix (r293167) didn't make it into clang-900. Adding a warning here wouldn't be the right solution, since it would be better to just cherry pick r293167.
Oct 4 2017
I'm not sure how to test the warning against anything but the macOS SDK. When I tried, I hit a -Wincompatible-sysroot issue. I can leave those changes out of this patch if we want to be more conservative.
Oct 3 2017
Landed as r314749
Great, thank you! Lgtm.
It seems easy enough to add a regression test for this, and I think we should.
Oct 2 2017
Could you upload a diff with some source context (e.g 'git diff -U1000')?
Oct 1 2017
Sep 29 2017
Sep 28 2017
LGTM, thanks for the pointer.
Sep 27 2017
I'm inclined to just accept this because it looks simple and harmless. Just to be sure, could you point me to any explanation of why LSan is disabled on Darwin in the first place? CC'ing @gottesmm, since he's looked at using LSan on Darwin before.
LGTM with two changes. Thanks!
Sep 26 2017
Thanks. Could you update the textual coverage logic as well and re-upload the diff with context lines (e.g git diff -U10000)?
Sep 25 2017
- Remove some noisy changes.
Sep 22 2017
- Tighten up lit test.