3792 – Regression(1.053) "non-constant expression" for a template inside a struct using a struct initializer

The conversions of scalars to Fixed type now fails at compile time.
There are several similar bugs to which it might be connected.
I join a reduced tango and xf independent test case.
Possibly connected to this xf.omg.core.CoordSys (that includes LinearAlgebra) segfaults.

Created attachment 566 [details] 
a testcase that fails

Reduced:
struct S {
	int x;
	
	template T(int val) {
		const S T = { val };
	}
}
const x = S.T!(0);
// arst.d(9): Error: non-constant expression (S).T

This bug is listed as fixed in 1.057, but while the reduced case by Matti Niemenmaa is indeed fixed, the original testcase still fails (I am pointing it out so that it stays open...)

Fixed dmd 1.057

As I wrote in my previous comment, the attached testcase still fails...

Reduced test case for the case that still fails. It's the same, except that the struct member is explicitly named in the struct initializer.
struct S {
 int x;
 template T(int val) {
 const S T = { x: val };
 }
}
const x = S.T!(0);

The D2 version of this code never compiled without error.
struct S {
 int x;
 template T(int val) {
 enum S T = { x: val };
 }
}
const x = S.T!(0);
It fails to compile because init.c, StructInitializer::toExpression() returns NULL if any of the fields have names ( if (field.data[i]) goto Lno; )
And this is just because it's never been implemented.

D1 fix: https://github.com/D-Programming-Language/dmd/commit/da0159d02d0e4ecb7ad1afa4bc6da402e677846f 

Fixed DMD2.053