[Mono-devel-list] mono-1.0b2, Solaris x86

[Guenter Feldmann]

Hi Paolo,

thanks for your quick response. Changing 'libc.so.6' in file
data/config into 'libc.so' helped. The installation succeeded
and the compiler woks now!


> > + #ifdef sun    // Solaris x86
> > + #  define MONO_ARCH_USE_SIGACTION 1
> > + struct sigcontext {
> > +         unsigned short gs, __gsh;
> [...]
> > + };
> > + #endif  // sun, Solaris x86
> You should instead include the header where sigcontext is defined in
> Solaris, instead of making you own copy here.

Solaris x86 does not have a struct sigcontext. Only offset constants are
supplied. But using these would require a lot of #ifdef's.


> > GC Warning: Insufficient space for /proc read
> 
> libgc issue: you should report it to the libgc mailing list.

The GC problem existed only in Solaris8. With Solaris9 it
vanished.

> > - What is the easiest way to test if exception handling works?
> 
> run make rcheck in mini and/or make test in mono/tests.

'make test' gives me the following results:

[ ... ]
Testing checked.exe... failed 256 (1) signal (0).
[ ... ]
Testing exception.exe... failed 256 (1) signal (0).
Testing exception2.exe... failed 256 (1) signal (0).
Testing exception3.exe... failed 256 (1) signal (0).
Testing exception4.exe... failed 256 (1) signal (0).
Testing exception5.exe... failed 256 (1) signal (0).
[ ... ]
Testing jit-float.exe... failed 256 (1) signal (0).
[ ... ]
Testing appdomain-unload.exe... ^Cfailed 33280 (130) signal (0)
   <At this point the check hung!>
[ ... ]
Testing delegate1.exe... failed 256 (1) signal (0).
[ ... ]
Testing nonvirt.exe... failed 256 (1) signal (0).
[ ... ]
Testing ckfiniteTest.exe... failed 256 (1) signal (0).
[ ... ]
165 test(s) passed. 11 test(s) failed.


Do these failures [failed 256 (1) signal (0)] indicate a single 
systematic bug?

-- Guenter