[clang] [AMDGPU] Mark int64_t and intmax_t as LP64 (PR #214865)

Matt Arsenault via cfe-commits cfe-commits at lists.llvm.org
Sat Aug 8 01:41:46 PDT 2026


================
@@ -87,7 +87,7 @@ __attribute__((device)) void test_non_volatile_parameter32(__UINT32_TYPE__ *ptr)
 // CHECK-NEXT:    [[TMP7:%.*]] = atomicrmw udec_wrap ptr [[TMP4]], i64 [[TMP6]] syncscope("workgroup") seq_cst, align 8, !amdgpu.no.fine.grained.memory [[META4]]
 // CHECK-NEXT:    store i64 [[TMP7]], ptr [[RES_ASCAST]], align 8
 // CHECK-NEXT:    ret void
-// GCN-LABEL: @_Z29test_non_volatile_parameter64Py(
+// GCN-LABEL: @_Z29test_non_volatile_parameter64Pm(
----------------
arsenm wrote:

seems like there should be a mor targeted test 

https://github.com/llvm/llvm-project/pull/214865


More information about the cfe-commits mailing list