llvm-project/clang/test/CodeGen/excess-embed-data.c
Mariya Podchishchaeva 59169036ca
[clang] Fix crash when #embed data does not fit into an array (#129567)
Tune SemaInit code handling #embed to take into account how many array
elements remains to initialize.
Also issue a warning/error message when the array/struct is at the end
but there is still #embed data left.

Fixes https://github.com/llvm/llvm-project/issues/128987
2025-03-05 11:04:52 +01:00

56 lines
1.8 KiB
C

// RUN: %clang_cc1 -std=c23 -emit-llvm -triple x86_64 %s -o - | FileCheck %s
struct S {
int arr[3];
};
struct S1 {
struct S s;
};
// CHECK: @[[BConst:.*]] = private unnamed_addr constant [2 x i32] [i32 47, i32 47]
// CHECK: @[[DConst:.*]] = private unnamed_addr constant [2 x i8] c"//"
// CHECK: @[[SConst:.*]] = private unnamed_addr constant %struct.S { [3 x i32] [i32 47, i32 47, i32 32] }
// CHECK: @[[S1Const:.*]] = private unnamed_addr constant %struct.S1 { %struct.S { [3 x i32] [i32 47, i32 47, i32 32] } }
void cases(int x) {
int a[3] = {x, x,
#embed __FILE__
};
int b[2] = {
#embed __FILE__
};
char d[2] = {
#embed __FILE__
};
struct S s = {
#embed __FILE__
, x
};
struct S1 s1 = {
#embed __FILE__
, x
};
}
// CHECK: define dso_local void @cases(i32 noundef %[[X:.*]])
// CHECK: %[[A:.*]] = alloca [3 x i32]
// CHECK: %[[B:.*]] = alloca [2 x i32]
// CHECK: %[[D:.*]] = alloca [2 x i8]
// CHECK: %[[S:.*]] = alloca %struct.S
// CHECK: %[[S1:.*]] = alloca %struct.S1
// CHECK: %[[LX:.*]] = load i32, ptr %[[X]].addr
// CHECK: store i32 %[[LX]], ptr %[[A]]
// CHECK: %[[GEP1:.*]] = getelementptr inbounds i32, ptr %[[A]], i64 1
// CHECK: %[[LX1:.*]] = load i32, ptr %[[X]].addr
// CHECK: store i32 %1, ptr %arrayinit.element
// CHECK: %[[GEP1:.*]] = getelementptr inbounds i32, ptr %[[A]], i64 2
// CHECK: store i32 47, ptr %[[GEP1]]
// CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 4 %[[B]], ptr align 4 @[[BConst]], i64 8, i1 false)
// CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 1 %[[D]], ptr align 1 @[[DConst]], i64 2, i1 false)
// CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 4 %[[S]], ptr align 4 @[[SConst]], i64 12, i1 false)
// CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 4 %[[S1]], ptr align 4 @[[S1Const]], i64 12, i1 false)