With our new attributes-based approach to handling allocator functions,
we need to differentiate between "is this literally free()" and "is this
a function which frees the given pointer." This change introduces the
new function to check for the latter category and moves the easy cases
throughout the codebase to check that instead of literally-free().
Details
Details
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo
Unit Tests
Unit Tests
Time | Test | |
---|---|---|
60,030 ms | x64 debian > MLIR.Examples/standalone::test.toy |
Event Timeline
Comment Actions
This affects tests a bit more than I'd expect - I guess these are test issues where we weren't calling the builtin free for one reason or another. Would it be better to fix up the tests so that they do the right things without this change beforehand to make it obvious that this isn't changing behaviour in a weird way?
Comment Actions
Maybe. Do you have some exemplars in mind here that change in ways you don't expect? I've been looking at this patch so long it just seems obviously-the-way-it-is to me now...for good or for ill.