[PATCH] D83082: [Alignment][NFC] Use proper getter to retrieve alignment from ConstantInt and ConstantSDNode

[Jessica Clarke via Phabricator via llvm-commits]

jrtc27 added a comment.

In D83082#2131971 <https://reviews.llvm.org/D83082#2131971>, @sbc100 wrote:

> This change seems to have broken one of our tests in emscripten.  I reduced it to this:
>
>   #include <stdlib.h>
>   #include <assert.h>
>  
>   int main() {
>     void * ptr = aligned_alloc(3, 64);
>     assert(ptr == NULL);
>     return 0;
>   }
>


Yeah, this is testing DR 460's new (and accepted) requirements (previously it was undefined):

> If the value of alignment is not a valid alignment supported by the implementation or the value of size is not an integral multiple of alignment the function shall fail by returning a null pointer.

I feel this should probably be handled specially (ie use align(1) or no alignment) in the Clang frontend? Though I wouldn't like to say how things change when you change it to:

  int align = 3;
  void *ptr = aligned_alloc(align, 64);

For sufficiently obfuscated code I can imagine Clang not noticing but a later LLVM pass discovering the alignment is a constant and trying to set the attributes up as such.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D83082/new/

https://reviews.llvm.org/D83082