Re: [avr-gcc-list] Updates needed to AVR test files

[Andy H]

Hi Wouter!

At this moment there are 407 failures. I started at about 900.
There are pending patches that will reduce this further, so the near 
horizon is about 300 failures.

Very roughly 50% of remaining are real  bugs in avr-gcc. Most are 
reported already (which just goes to show why we need to use testsuite).
The others failures are problems are with tests. But until I go thru 
each one and check, I cannot be sure.

There are pending patches that will reduce this further, so the near 
horizon is about 300 failures.

When a test is fixed, often more tests are run - so total number of 
tests increase. One patch might fix 1 failure or 10 failures or add some 
failures. Of course gcc testsuite is always growing. At this time we 
have 43,000 tests performed.

I think board file should be patched on Dejagnu, But I sure someone can 
take care of creating  WinAVR patch for me ;-)

Mike Stein graciously posts AVR test results on gcc mailing list. You 
can check progress there. It is normally about 2 days old. Make sure you 
look at 4.4 test results as he posts other versions and branches to.

Andy


Wouter van Gulik wrote:
> Hi Andy / list,
>
> I was just curious, how many issues are still left? IIRC the last time you
> posted some result you got ~50 torture execution and ~500 compile issues.
>
> IIUC this should ideally be patched to the boardfile in the WinAVR CVS
> repository or is there an other place where avr board files life these days
> (in a GCC repository)?
>
> Wouter
>
>   
> > -----Oorspronkelijk bericht-----
> > Van: address@hidden [mailto:avr-gcc-
> > address@hidden Namens Andy H
> > Verzonden: zondag 1 juni 2008 1:35
> > Aan: AVR-GCC; Paulo Marques; Mike Stein; Weddington, Eric; Anatoly Sokolov
> > Onderwerp: Re: [avr-gcc-list] Updates needed to AVR test files
> >
> > Small typo - in COMPLEX_INT
> >
> > Should be
> >
> > set COMPAT_SKIPS [list {VA} {COMPLEX_INT}]
> >
> > not plural
> >
> > Duh!
> > Andy
> >
> >
> > Andy H wrote:
> >     
> > > BTW you can also just define these in environment and leave board file
> > > unchanged
> > >
> > > set COMPAT_SKIPS [list {VA} {COMPLEX_INTS}]
> > > set COMPAT_OPTIONS [list [list {-Os -mcall-prologues} {-Os
> > > -mcall-prologues}]]
> > >
> > > Andy
> > >
> > > Andy H wrote:
> > >       
> > > > There are a couple of changes needed to AVR test files to pass a few
> > > > tests.
> > > >
> > > > Compatibility tests default to no optimization and maximum tests -
> > > > this can easily overflow 128K code area.
> > > > Add these lines to end board file (mine is called
> > > > atmega128-simnew.exp). They set environment vars that control these
> > > > tests and get many more to work. (Some still need other fixes).
> > > >
> > > > # Restrict compatibility tests. And optimise to reduce size.
> > > > set COMPAT_SKIPS [list {VA} {COMPLEX_INTS}]
> > > > set COMPAT_OPTIONS [list [list {-Os -mcall-prologues} {-Os
> > > > -mcall-prologues}]]
> > > >
> > > > Dummy io/exit/abort  file exit.c  has unused parameter stream. The
> > > > warning created  then causes a failure in some tests. Hack as follows
> > > > to create
> > > > dummy reference to stream, thus removing the warning.
> > > >
> > > > int putchar_exit_c(char c, FILE *stream)
> > > > {
> > > >    *((volatile unsigned char *) STDIO_PORT) = c;
> > > >    stream = NULL;
> > > >    return 0;
> > > > }
> > > >
> > > >
> > > > best regards
> > > > Andy
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > _______________________________________________
> > > > AVR-GCC-list mailing list
> > > > address@hidden
> > > > http://lists.nongnu.org/mailman/listinfo/avr-gcc-list
> > > >         
> > >       
> > _______________________________________________
> > AVR-GCC-list mailing list
> > address@hidden
> > http://lists.nongnu.org/mailman/listinfo/avr-gcc-list
> >     
>
>
>
> _______________________________________________
> AVR-GCC-list mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/avr-gcc-list
>