[PATCH] D120499: [NVPTX] Fix nvvm.match.sync*.i64 intrinsics return type (i64 -> i32)

Artem Belevich via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Thu Feb 24 11:17:11 PST 2022


tra added a comment.

Good catch. Thank you for the fix.



================
Comment at: clang/include/clang/Basic/BuiltinsNVPTX.def:477
 TARGET_BUILTIN(__nvvm_match_any_sync_i32, "UiUiUi", "", PTX60)
-TARGET_BUILTIN(__nvvm_match_any_sync_i64, "WiUiWi", "", PTX60)
+TARGET_BUILTIN(__nvvm_match_any_sync_i64, "UiUiWi", "", PTX60)
 // These return a pair {value, predicate}, which requires custom lowering.
----------------
I've also noticed that the PTX spec also says `Requires sm_70 or higher.`, so we may want to fix the constraint, too. 


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D120499/new/

https://reviews.llvm.org/D120499



More information about the cfe-commits mailing list