[PATCH] D136560: llvm-reduce: Try to turn calls into something else

[Arthur Eubanks via Phabricator via llvm-commits]

aeubanks accepted this revision.
aeubanks added inline comments.
This revision is now accepted and ready to land.


================
Comment at: llvm/tools/llvm-reduce/deltas/ReduceOpcodes.cpp:194
+    case 3:
+      return B.CreateIntrinsic(Intrinsic::fshl, {CB->getType()}, Arguments);
+    default:
----------------
arsenm wrote:
> aeubanks wrote:
> > IMO a call to an intrinsic is less reduced than a call to some arbitrary external function.
> > Perhaps it's finally time to create a flag for reducing IR optimizing for codegen?
> An intrinsic has known behavior, an arbitrary external call doesn't. From a semantics perspective, that's more understandable 
from a "I want to submit this IR reproducer as a test case" standpoint, calling an intrinsic as opposed to an arbitrary call is weirder. there's very little code that deals specifically with `fshl` at least in the middle-end, but a lot that deals with generic calls. so picking some random intrinsic is weird

but sure, we can go with this and see if people (I) get annoyed or not


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D136560/new/

https://reviews.llvm.org/D136560