bug 142 -- fixed but dstress still failing

[Brad Roberts]

Thomas Kuehne wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Frits van Bommel schrieb am 2007-01-19:
>> Lionello Lunesu wrote:
>>> Brad Roberts wrote:
>>>> http://d.puremagic.com/issues/show_bug.cgi?id=142 is marked as fixed 
>>>> as of 0.158.  However, looking at the dstress failures for 1.00, I see 
>>>> bug_template_610_[ABDE] all failing.  The error message has changed in 
>>>> more recent releases from what it used to be:
>>>>
>>>> dmd version 0.146 (something pre-0.158 that I had installed already)
>>>> $ ../dmd/146/dmd/bin/dmd -I../dmd/146/dmd/src/phobos bug_template_610_A.d
>>>> bug_template_610_A.d(10): template 'A(alias T)' is not a variable
>>>> bug_template_610_A.d(10): function expected before (), not A(alias T) 
>>>> of type int
>>>>
>>>> dmd version 1.00
>>>> $ dmd bug_template_610_A.d
>>>> bug_template_610_A.d(10): template 
>>>> dstress.nocompile.b.bug_template_610_A.A(alias T) does not match any 
>>>> template declaration
>>>> bug_template_610_A.d(10): template 
>>>> dstress.nocompile.b.bug_template_610_A.A(alias T) cannot deduce 
>>>> template function from argument types (int)
>>>>
>>>> $ cat -n bug_template_610_A.d
>>>>      1 module dstress.nocompile.b.bug_template_610_A;
>>>>      2
>>>>      3 template A(alias T) {
>>>>      4         void A(T){
>>>>      5         }
>>>>      6 }
>>>>      7
>>>>      8 void main(){
>>>>      9         int i;
>>>>     10         A(i);
>>>>     11 }
>>>>
>>> Should that code even work? I'm no template guru (far from it), but I 
>>> have no idea what should be happening. The template should generate a 
>>> function called A, which in turn takes a parameter of type T, with T 
>>> being "i"? I don't get it.
>> No it shouldn't work. That's why the module name includes *nocompile*, I 
>> would think. The bug was that the compiler had an assertion failure when 
>> compiling that code (look at the summary), not that it didn't compile.
>> I'm not sure why dstress lists it as "FAIL" though. I'd think it should 
>> be "XFAIL"...
>>
>> Thomas, can you shed any light on this?
> 
> The above code is missing a crucial information:
> 
>  9  // __DSTRESS_ELINE__ 14
> 10
> 11  module dstress.nocompile.b.bug_template_610_A;
> 12
> 13  template A(alias T) {
> 14    void A(T){
> 15    }
> 16  }
> 17
> 18  void main(){
> 19    int i;
> 20    A(i);
> 21  }
> 
> The template declaration in line 13 matches the request in line 20 but
> fails to instantiate due to line 14. If you look at DMD's output
> however, only line 20 is mentioned.
> 
> bug_template_610_A.d(20): template dstress.nocompile.b.bug_template_610_A.A(alias T) does not match any template declaration
> bug_template_610_A.d(20): template dstress.nocompile.b.bug_template_610_A.A(alias T) cannot deduce template function from argument types (int)
> 
> Thomas
> 
> 
> -----BEGIN PGP SIGNATURE-----
> 
> iD8DBQFFse2PLK5blCcjpWoRAiIUAJ0Q4ZRv0HLxNyjQG8UIgaOteaA4BwCdGPF9
> uggdR/pfiBLaMLakMo4OFQE=
> =uUET
> -----END PGP SIGNATURE-----

Ok, so, um.. apologies for being obtuse.  Is dmd still broken in your 
mind?  If so, then either a new bug needs to be filed or bug 142 needs 
to be re-opened with details as to what's still wrong.  Would you make 
the choice and perform the action? :)

Thanks,
Brad