https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123697
Bug ID: 123697
Summary: aarch64: ICE when unrolling at -msve-vector-bits=128
Product: gcc
Version: 16.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: tree-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: alfierichards at gcc dot gnu.org
Target Milestone: ---
The following:
```c+++
struct a {
double b() const;
int c[2];
};
double a::b() const { return c[1]; }
a d[0];
double f;
void g() {
for (int e = 0; e < 3; ++e)
f += d[e].b();
}
```
ICE's for `gcc/xgcc -B gcc -march=armv9-a test.cc -mmax-vectorization
-msve-vector-bits=128 -O3`
With:
```
test.cc: In function ‘void g()’:
test.cc:8:6: error: non-trivial conversion in ‘vector_cst’
8 | void g() {
| ^
const vector(4) int[2]
const vector(4) int
# .MEM_21 = VDEF <.MEM_60>
vect_array.19 = { 0, 0, 0, 0 };
```
