From @masak

<Radvendii> is it *remotely* possible that given $i {} is changing the
value of $i?
<masak> nom​: my $i = 5; given $i { $_ = 42 }; say $i
<p6eval> nom 9719f7​: OUTPUT«42␤»
<Radvendii> that's not what i'm saying
<Radvendii> nom​: my $i=5; $i.say; given $i { $i.say};
<p6eval> nom 9719f7​: OUTPUT«5␤5␤»
<Radvendii> and those two are different in my program
<Radvendii> feel free to golf a short example of it.
<Radvendii> masak​: i'll try...
<Radvendii> YAY
<Radvendii> masak​: i distilled the error down
<Radvendii> nom​: sub foo ($a, $f) { if $f { foo('z', 0) }; given $a
{say $a; say $_} }; foo('x', 1)
<p6eval> nom 9719f7​: OUTPUT«z␤z␤z␤x␤»
<masak> Radvendii++
<masak> that is the real deal.
* masak submits rakudobug
<Radvendii> masak​: yay :D
<moritz> perl6​: sub foo ($a, $f) { if $f { foo('z', 0) }; given $a
{say $a; say $_} }; foo('x', 1)
<p6eval> rakudo 9719f7​: OUTPUT«z␤z␤z␤x␤»
<p6eval> ..pugs b927740, niecza v13-381-g4158fa9​: OUTPUT«z␤z␤x␤x␤»
* masak is surprised we still have this kind of lexpad bug
* Radvendii assumes lexpad has to do with lexical scopes?
<moritz> yes
<Radvendii> cool
<masak> Radvendii​: if I had to guess, what you're seeing is the given
block anchoring to the lexpad of the first invocation, and then when
the lookup for $a is done in the second invocation, it finds the wrong
-- the old -- OUTER for the given block.
<masak> at least that's a hypothesis consistent with the data.

From @masak

<Radvendii> nom​: sub foo ($a, $f) { if $f { foo('z', 0) }; {$_=$a; say
$a; say $_} }; foo('x', 1) #does it work explicitly?
<p6eval> nom 9719f7​: OUTPUT«z␤z␤z␤z␤»
<Radvendii> ooh. that's even worse...
<masak> Radvendii​: but it has exactly the same cause.
<Radvendii> masak​: yup... i just realized that
<Radvendii> i thought it maybe had to do with the "given" statement
itself
<masak> me too.
<masak> no, it's just blocks inside reentrant blocks.
* masak updates ticket

From [Unknown Contact. See original ticket]

<Radvendii> nom​: sub foo ($a, $f) { if $f { foo('z', 0) }; {$_=$a; say
$a; say $_} }; foo('x', 1) #does it work explicitly?
<p6eval> nom 9719f7​: OUTPUT«z␤z␤z␤z␤»
<Radvendii> ooh. that's even worse...
<masak> Radvendii​: but it has exactly the same cause.
<Radvendii> masak​: yup... i just realized that
<Radvendii> i thought it maybe had to do with the "given" statement
itself
<masak> me too.
<masak> no, it's just blocks inside reentrant blocks.
* masak updates ticket

@masak - Status changed from 'new' to 'open'

From @masak

<Radvendii> nom​: sub foo ($a, $f) { if $f { foo('z', 0) }; {$_=$a; say
$a; say $_} }; foo('x', 1) #does it work explicitly?
<p6eval> nom 9719f7​: OUTPUT«z␤z␤z␤z␤»
<Radvendii> ooh. that's even worse...
<masak> Radvendii​: but it has exactly the same cause.
<Radvendii> masak​: yup... i just realized that
<Radvendii> i thought it maybe had to do with the "given" statement
itself
<masak> me too.
<masak> no, it's just blocks inside reentrant blocks.
* masak updates ticket

From [Unknown Contact. See original ticket]

<Radvendii> nom​: sub foo ($a, $f) { if $f { foo('z', 0) }; {$_=$a; say
$a; say $_} }; foo('x', 1) #does it work explicitly?
<p6eval> nom 9719f7​: OUTPUT«z␤z␤z␤z␤»
<Radvendii> ooh. that's even worse...
<masak> Radvendii​: but it has exactly the same cause.
<Radvendii> masak​: yup... i just realized that
<Radvendii> i thought it maybe had to do with the "given" statement
itself
<masak> me too.
<masak> no, it's just blocks inside reentrant blocks.
* masak updates ticket

From @usev6

The first example works now as expected​:

$ perl6-m -e 'sub foo ($a, $f) { if $f { foo("z", 0) }; given $a {say $a; say $_} }; foo("x", 1)'
z
z
x
x

The second example works on rakudo.jvm but fails on rakudo.moar​:

$ perl6-j -e 'sub foo ($a, $f) { if $f { foo("z", 0) }; {$_=$a; say $a; say $_} }; foo("x", 1)'
z
z
x
x

$ perl6-m -e 'sub foo ($a, $f) { if $f { foo("z", 0) }; {$_=$a; say $a; say $_} }; foo("x", 1)'
z
z
z
z

I added two tests (the second fudged 'todo' for Moar) to S06-advanced/lexical-subs.t with commit Raku/roast@0867f11ce3

From @masak

A bit late to the game, but...

<masak> here's a shorter version​:
<masak> m​: my $y = 1; sub foo($x) { $y-- && foo("z"); { say $x } }; foo("x")
<camelia> rakudo-moar d179b4​: OUTPUT«z␤z␤»
<masak> again, should say "z␤x␤"
* masak adds the example to RT #​109322

From @jnthn

On Thu Jun 18 13​:49​:36 2015, bartolin@​gmx.de wrote​:

The first example works now as expected​:

$ perl6-m -e 'sub foo ($a, $f) { if $f { foo("z", 0) }; given $a {say
$a; say $_} }; foo("x", 1)'
z
z
x
x

The second example works on rakudo.jvm but fails on rakudo.moar​:

$ perl6-j -e 'sub foo ($a, $f) { if $f { foo("z", 0) }; {$_=$a; say
$a; say $_} }; foo("x", 1)'
z
z
x
x

$ perl6-m -e 'sub foo ($a, $f) { if $f { foo("z", 0) }; {$_=$a; say
$a; say $_} }; foo("x", 1)'
z
z
z
z

I added two tests (the second fudged 'todo' for Moar) to S06-
advanced/lexical-subs.t with commit
Raku/roast@0867f11ce3

After a fix, the test now passes on MoarVM too. Unfudged it.

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