From his comment in PR12168, Eli seems confused about the alignment assumptions
we're making. We only ignore implicit casts. Add a testcase showing that
we get the right behavior with explicit casts.
git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@151994 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/test/CodeGen/builtin-memfns.c b/test/CodeGen/builtin-memfns.c
index e4389f6..72d3406 100644
--- a/test/CodeGen/builtin-memfns.c
+++ b/test/CodeGen/builtin-memfns.c
@@ -52,9 +52,13 @@
// CHECK: @test7
// PR12094
int test7(int *p) {
+ struct snd_pcm_hw_params_t* hwparams; // incomplete type.
+
// CHECK: call void @llvm.memset{{.*}}256, i32 4, i1 false)
__builtin_memset(p, 0, 256); // Should be alignment = 4
- struct snd_pcm_hw_params_t* hwparams; // incomplete type.
+
+ // CHECK: call void @llvm.memset{{.*}}256, i32 1, i1 false)
+ __builtin_memset((char*)p, 0, 256); // Should be alignment = 1
__builtin_memset(hwparams, 0, 256); // No crash alignment = 1
// CHECK: call void @llvm.memset{{.*}}256, i32 1, i1 false)
