Another shot at fixing the issues with (constant) struct literals and their addresses. See DMD2682, #218, #324. The idea is to separate the notion of const from 'this variable can always be replaced with its initializer' in the frontend. To do that, I introduced Declaration::isSameAsInitializer, which is overridden in VarDeclaration to return false for constants that have a struct literal initializer. So {{{ const S s = S(5); void foo() { auto ps = &s; } // is no longer replaced by void foo() { auto ps = &(S(5)); } }}} To make taking the address of a struct constant with a struct-initializer outside of function scope possible, I made sure that AddrExp::optimize doesn't try to run the argument's optimization with WANTinterpret - that'd again replace the constant with a struct literal temporary.

--- a/dmd/optimize.c	Sun Jun 14 14:28:11 2009 +0200
+++ b/dmd/optimize.c	Sun Jun 14 19:49:58 2009 +0200
@@ -194,7 +194,9 @@
 {   Expression *e;
 
     //printf("AddrExp::optimize(result = %d) %s\n", result, toChars());
-    e1 = e1->optimize(result);
+    // never try to interpret: it could change the semantics by turning
+    // const p = &s; into an something like const p = &(Struct());
+    e1 = e1->optimize(result & ~WANTinterpret);
     // Convert &*ex to ex
     if (e1->op == TOKstar)
     {	Expression *ex;