Sorry Peter,
I tried your fix up but I still end at the same place with exactly the same error message.
What does that error 11 stand for? I checked gcc (version 2.8.1) and it was working fine so the "error" shouldn't come from there. I tried your fix up starting from the begining, that is deleting all the *.o files and reruning the make command.
Thanks for your help ....
Igor
What does that error 11 stand for? I checked gcc (version 2.8.1) and it was working fine so the "error" shouldn't come from there. I tried your fix up starting from the begining, that is deleting all the *.o files and reruning the make command.
A `make distclean' or just `make clean' will do a pretty good job of removing all *.o files (although I suspect you already knew that).
error 11 is probably signal SIGSEGV - Segmentation violation. You can confirm this by looking in /usr/include/signum.h and seeing which #define equates to 11.
If it is a SIGSEGV then there is the likelihood that a `core' file has been produced somewhere in the gcc directory structure that relates to that crash. If you can find one then load it into gdb and see what the stack backtrace says (a stack backtrace is only sensible when you've turned debugging on i.e. the -g option to gcc).
To examine a core dump file, load gdb.
$ gdb
then load the core file with (assuming it's in the pwd)
(gdb) core-file core
gdb should automatically provide you with the stack backtrace.
I must say that it looks like your problem is caused by a gcc code generation bug, rather than a bug in gpc. Otherwise the same problem would manifest itself on other machines. We should be able to investigate this further once you've supplied that backtrace.
Cheers,
Nick.