[Gc] MacOS 10.4 failure

S M Ryan wyrmwif at rawbw.com
Mon Feb 12 14:46:59 PST 2007


I've been thowing chunks of code and finally collection seems running  
on a stub once I comment out one of my routines. So it may be my  
fault after all (after somehow evading the block smash code and a  
separate library for block overruns).


On Feb 12, 2007, at 2:32 PM, Andreas Tobler wrote:

> S M Ryan wrote:
>> I'm getting aborts, apparently in darwin_stop_world. I thought I  
>> applied MACHINE_THREAD_STATE patch, but I've lost track of what  
>> I'm supposed to be compiling. I've included the state of things  
>> here. If some kind soul could point to me the right place if I'm  
>> out of date. Currently I have to GC_no_gc =1 to get anything done.
>
> If you have a chance to use the cvs version, please try. If not, I  
> can try to backport the cvs mods to 6.8.
>
> There is one patch pending for darwin* which should bring  
> MPROTECT_VDB support. But this is minor I think.
>
> http://sourceforge.net/cvs/?group_id=144171
>
> cvs -d:pserver:anonymous at bdwgc.cvs.sourceforge.net:/cvsroot/bdwgc  
> login
>
> cvs -z3 -d:pserver:anonymous at bdwgc.cvs.sourceforge.net:/cvsroot/ 
> bdwgc co -P bdwgc
>
> Hans, would this be useful, a backport ?
>
> Andreas
>



More information about the Gc mailing list