Fix 8044 - Print name of enum passed a tmpl param

[Geod24]

The fix was actually quite trivial. Looking at the code flow, I doubt this actually change any other diagnostic, but let's see what the CI says.

[dlang-bot]

Thanks for your pull request, @Geod24!

Bugzilla references

Auto-close	Bugzilla	Severity	Description
✓	8044	enhancement	Print names, not casted values when using enum template parameter

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub run digger -- build "master + dmd#11841"

[Geod24]

I doubt this actually change any other diagnostic

It does, but for the best!

[Geod24]

This should also fix https://issues.dlang.org/show_bug.cgi?id=14067 (marked as a duplicate now) and so improve the documentation.

[Geod24]

-----------------------------------------------------------
ERROR: Newly generated header file (/home/circleci/dmd/generated/linux/release/64/frontend.h) doesn't match with the reference header file (/home/circleci/dmd/src/dmd/frontend.h)

DETAILS:

diff --git a/home/circleci/dmd/src/dmd/frontend.h b/home/circleci/dmd/generated/linux/release/64/frontend.h
index a845385d4..86c99e3aa 100644
--- a/home/circleci/dmd/src/dmd/frontend.h
+++ b/home/circleci/dmd/generated/linux/release/64/frontend.h
@@ -1703,7 +1703,7 @@ struct ParameterList
     Array<Parameter*>* parameters;
     StorageClass stc;
     VarArg varargs;
-    ParameterList(Array<Parameter*>* parameters, VarArg varargs = cast(VarArg)cast(ubyte)0u, StorageClass stc = 0);
+    ParameterList(Array<Parameter*>* parameters, VarArg varargs = VarArg.none, StorageClass stc = 0);
     size_t length();
     Parameter* opIndex(size_t i);
     ParameterList() :

===============

The header generation depended on that bug ? :|

[ibuclaw]

The header generation depended on that bug ? :|

Strictly for dmd, I don't think we require any of these automatic constructors, as initialisation is handed over to the D layer.

[UplinkCoder]

If I recall correctly there are still cases in which you'll see cast(EnumType)value
and the comment out assert(0) was there for that reason, to help debug such cases.

[Geod24]

If I recall correctly there are still cases in which you'll see cast(EnumType)value

If you find any, please report. As can be seen from the diff in the testsuite, many, if not all of them, should disappear. I left the cast in to be sure no regression is introduced.

and the comment out assert(0) was there for that reason, to help debug such cases.

If someone want to debug such case, adding back and assert or printf is trivial. There's no need to have this assert here as it's not something frequently needed as opposed to the other debugging printf (e.g. printf in semantic routines).

[MoonlightSentinel]

Reduced test for the dtoh segfault:

enum E { m }

extern (C++) void foo(ulong = E.m) {}

Working on it.

[Geod24]

This should be good to ggo