Hi just one more time :)

The same situation as in previous two posts. Snipet would not compile with -inline flag and would without it.

module core;

struct Struct
{
	real foo()
	{
		return 0;;
	}
	
	void bar(out Struct Q)
	{
		if (foo < 0)
			Q = *this;			
	}
}

int main(char[][] args)
{
	return 0;	
}

D:\proj\dtest>dmd -release -inline core.d
Internal error: ..\ztc\cod3.c 736

Does somebody knows are all bugs for Walter has similiar priority or bugs of this kind has lower priority because they does appear in retail compilations only?

-- 
Victor (aka nail) Nakoryakov
nail-mail<at>mail<dot>ru

Krasnoznamensk, Moscow, Russia

Victor Nakoryakov wrote:
> Hi just one more time :)
> 
> The same situation as in previous two posts. Snipet would not compile with -inline flag and would without it.
> 
> module core;
> 
> struct Struct
> {
>     real foo()
>     {
>         return 0;;
>     }
>         void bar(out Struct Q)
>     {
>         if (foo < 0)
>             Q = *this;               }
> }
> 
> int main(char[][] args)
> {
>     return 0;   }
> 
> D:\proj\dtest>dmd -release -inline core.d
> Internal error: ..\ztc\cod3.c 736
> 
> Does somebody knows are all bugs for Walter has similiar priority or bugs of this kind has lower priority because they does appear in retail compilations only?
> 

Oh, I forgot to write: note that foo() has double ;; in the return statement. If I'll delete one then bug would disappear.

-- 
Victor (aka nail) Nakoryakov
nail-mail<at>mail<dot>ru

Krasnoznamensk, Moscow, Russia

>Does somebody knows are all bugs for Walter has similiar priority or bugs of this kind has lower priority because they does appear in retail compilations only?

I can't say for sure, but from looking at the recent changelog it seems that the focus is on dmd crashes/internal errors, and on incorrect code generation. I would guess bugs involving optimization switches, error messages and the like are given a lower priority.

Nick

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Victor Nakoryakov schrieb am Thu, 23 Jun 2005 21:30:52 +0400:
> Hi just one more time :)
>
> The same situation as in previous two posts. Snipet would not compile with -inline flag and would without it.
>
> module core;
>
> struct Struct
> {
> 	real foo()
> 	{
> 		return 0;;
> 	}
> 
> 	void bar(out Struct Q)
> 	{
> 		if (foo < 0)
> 			Q = *this;
> 	}
> }
>
> int main(char[][] args)
> {
> 	return 0;
> }
>
> D:\proj\dtest>dmd -release -inline core.d
> Internal error: ..\ztc\cod3.c 736
>
> Does somebody knows are all bugs for Walter has similiar priority or bugs of this kind has lower priority because they does appear in retail compilations only?

Added to DStress as http://dstress.kuehne.cn/run/b/bug_cod3_736_A.d http://dstress.kuehne.cn/run/b/bug_cod3_736_B.d

Thomas


-----BEGIN PGP SIGNATURE-----

iD8DBQFCu8SR3w+/yD4P9tIRArh1AJ9NKc8CPpaqldFYNY0vpvPI7gdwRgCePOML
DxrWan9HYvhg/8TpjJLgi7c=
=En6U
-----END PGP SIGNATURE-----

"Nick" <Nick_member@pathlink.com> wrote in message news:d9f1r8$26p$1@digitaldaemon.com...
> >Does somebody knows are all bugs for Walter has similiar priority or bugs of this kind has lower priority because they does appear in retail compilations only?
>
> I can't say for sure, but from looking at the recent changelog it seems
that the
> focus is on dmd crashes/internal errors, and on incorrect code generation.
I
> would guess bugs involving optimization switches, error messages and the
like
> are given a lower priority.

There aren't any hard and fast rules, but silently generating bad code gets highest priority, and compiler crashes get next highest. Bugs that break formerly working code gets higher priority than bugs that have existed all along. Easy to fix bugs get higher priority than hard ones. Bugs where the bug reporter did a nice job isolating down a simple test case get higher priority than ones that come with hundreds of files and a megabyte of source <g>.