[Gcl-devel] strange new gcl fork error on hppa

[Camm Maguire]

Greetings!  gclcvs has a new FTBFS error on hppa due to a failing
system() call to a subprocess:

fork(Process 19167 attached
)                                  = 19167
[pid 19166] waitpid(19167, Process 19166 suspended
 <unfinished ...>
[pid 19167] rt_sigaction(SIGINT, {SIG_DFL}, NULL, 8) = 0
[pid 19167] rt_sigaction(SIGQUIT, {SIG_DFL}, NULL, 8) = 0
[pid 19167] rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0
[pid 19167] execve("/bin/sh", ["sh", "-c", ".//rsym 
/home/camm/gclcvs-2.7.0/"...], [/* 19 vars */]) = -1 EFAULT (Bad address)
[pid 19167] +++ killed by SIGKILL +++

system returns 9 corresponding to the sigkill.  The following program
has no problem:

int main() {printf("%d\n",system("./rsym 
/home/camm/gclcvs-2.7.0/unixport/raw_pre_gcl rsym2"));}

so there appears to be a forking address conflit.  gclcvs uses the
brk() system call, which succeeds earlier in the trace:

brk(0)                                  = 0x6c3000
brk(0x9c4000)                           = 0x9c4000

Any advice most appreciated.

Take care,

-- 
Camm Maguire                                        address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah