Bug 6902 – Different "pure nothrow int()" types

--- void main() { static assert(is(typeof({ return int.init; // int, long, real, etc. }))); int f() pure nothrow { assert(0); } alias int T() pure nothrow; static if(is(typeof(&f) DT == delegate)) { static assert(is(DT* == T*)); // ok // Error: static assert (is(pure nothrow int() == pure nothrow int())) is false static assert(is(DT == T)); } } ---

I thought we got rid of those crazy function types. Seems that they've survived in alias declarations somehow. The declaration of T shouldn't compile.

Good reducing! I found a function/delegate literal type inference bug. I have done Phobos unit tests with my patch, and this resolves Phobos build breaking in Windows. http://d.puremagic.com/test-results/test_data.ghtml?dataid=113170 Please wait the patch a while.

https://github.com/D-Programming-Language/dmd/pull/498 https://github.com/D-Programming-Language/phobos/pull/322

https://github.com/D-Programming-Language/dmd/commit/3af76245c7626a1489322bb6fc8353b1c4aa2e35 https://github.com/D-Programming-Language/phobos/commit/b7a9f446dfb9e299354318437d1e6a5752d5feca

*** Issue 6600 has been marked as a duplicate of this issue. ***

cat > bug.d << CODE void a() { function void() { }(); } void b() { static void safe_nothrow() @safe nothrow pure { } auto f = &safe_nothrow; enum mangle = typeof(*f).mangleof; static assert(mangle == "FNaNbNfZv", mangle); } CODE dmd -c bug ---- I've actually found a new failing test case for this.

*** Issue 3796 has been marked as a duplicate of this issue. ***

(In reply to comment #6) > cat > bug.d << CODE > void a() > { > function void() { }(); > } > > void b() > { > static void safe_nothrow() @safe nothrow pure { } > auto f = &safe_nothrow; > enum mangle = typeof(*f).mangleof; > static assert(mangle == "FNaNbNfZv", mangle); > } > CODE > > dmd -c bug > > ---- > > I've actually found a new failing test case for this. It is a dup of bug 8504. Pull #1096 will fix the issue.