https://issues.dlang.org/show_bug.cgi?id=17758
Issue ID: 17758
Summary: assignment of different vector types should require
cast
Product: D
Version: D2
Hardware: x86_64
OS: Linux
Status: NEW
Severity: enhancement
Priority: P1
Component: dmd
Assignee: [email protected]
Reporter: [email protected]
cat > bug.d << CODE
void test()
{
__vector(float[4]) f4;
__vector(uint[4]) u4;
__vector(void[16]) v16;
f4 = cast(typeof(f4)) u4; // OK
f4 = u4; // should fail
v16 = f4; // OK
f4 = v16; // should fail
f4 = cast(typeof(f4)) v16; // OK
}
CODE
dmd -c bug
It's somewhat broken that vectors of different types can be assigned without
any casting. This is OK for implicit conversions to void16 vectors, but not for
conversions from void16.
--
