From @masak

<masak> rakudo​: sub foo(&foo = &foo) { say "OH HAI"; foo }; foo
<p6eval> rakudo 4c31fb​: OUTPUT«Null PMC access in isa()␤in sub foo [...]
<hoelzro> pmichaud​: it should just alert you
<masak> hah!
* masak submits rakudobug
<masak> rakudo​: sub foo(&foo = &foo) {}; foo
<p6eval> rakudo 4c31fb​: OUTPUT«Null PMC access in isa()␤in sub foo [...]
<moritz_> rakudo​: sub foo(&foo = &foo) {};
<p6eval> rakudo 4c31fb​: ( no output )
<moritz_> ah, happens while calling
<masak> aye.
<KyleHa> Is that supposed to live?
<masak> I think so.
<masak> KyleHa​: the first &foo is the param, the second is the sub.

From @jnthn

On Mon Jul 27 08​:11​:54 2009, masak wrote​:

<masak> rakudo​: sub foo(&foo = &foo) { say "OH HAI"; foo }; foo
<p6eval> rakudo 4c31fb​: OUTPUT«Null PMC access in isa()␤in sub foo [...]
<hoelzro> pmichaud​: it should just alert you
<masak> hah!
* masak submits rakudobug
<masak> rakudo​: sub foo(&foo = &foo) {}; foo
<p6eval> rakudo 4c31fb​: OUTPUT«Null PMC access in isa()␤in sub foo [...]
<moritz_> rakudo​: sub foo(&foo = &foo) {};
<p6eval> rakudo 4c31fb​: ( no output )
<moritz_> ah, happens while calling
<masak> aye.
<KyleHa> Is that supposed to live?
<masak> I think so.
<masak> KyleHa​: the first &foo is the param, the second is the sub.

It now lives​:

00​:30 <@​jnthn> rakudo​: sub foo(&foo = &foo) {}; foo
00​:30 <+p6eval> rakudo 346e76​: ( no output )

Since the &foo inside the sub is a new lexical variable, you essentially
end up doing like my $a = $a which is fine but useless...

00​:30 <@​jnthn> rakudo​: sub foo(&foo = &foo) {say &foo}; foo
00​:30 <+p6eval> rakudo 346e76​: OUTPUT«Any()␤»

Anyway, it's what I'd expect. Giving to moritz++ for spectest.

Thanks,

Jonathan

The RT System itself - Status changed from 'new' to 'open'

From @moritz

Tested in S06-signature/code.t.

@moritz - Status changed from 'open' to 'resolved'

From @kyleha

This is an automatically generated mail to inform you that tests are now available in t/spec/S06-signature/code.t

commit 6d18247f0f0e02ee8084016b843522245cc6c5bb
Author​: moritz <moritz@​c213334d-75ef-0310-aa23-eaa082d1ae64>
Date​: Wed Apr 21 20​:30​:51 2010 +0000

  [t/spec] test for RT #​67932
 
  git-svn-id​: http://svn.pugscode.org/pugs@&#8203;30432 c213334d-75ef-0310-aa23-eaa082d1ae64

diff --git a/t/spec/S06-signature/code.t b/t/spec/S06-signature/code.t
index 9a54662..733970b 100644
--- a/t/spec/S06-signature/code.t
+++ b/t/spec/S06-signature/code.t
@@ -1,6 +1,6 @@
 use v6;
 use Test;
-plan 6;
+plan 8;
 
 # TODO: move this test to closure-parameters.t if it works in the future
 
@@ -37,4 +37,15 @@ is tester2({ 'block' }), 1, 'Can pass a block to a &parameter';
             "Can't pass something that isn't typed as returning Callable";
 }
 
+# RT #67932
+{
+    my $tracker;
+    sub foo(&foo = &foo) {
+        $tracker = &foo
+    };
+    lives_ok { foo },
+        'can call a sub with a code object defaulting to something of its own name';
+    ok !$tracker.defined, 'the inner &foo is undefined (scoping)';
+}
+
 # vim: ft=perl6