diff options
| author | Ingo Molnar <mingo@elte.hu> | 2002-12-14 19:44:20 -0800 |
|---|---|---|
| committer | Ingo Molnar <mingo@elte.hu> | 2002-12-14 19:44:20 -0800 |
| commit | b9daa0066d792983da59154af3ae486eff9b9aa1 (patch) | |
| tree | c571dbad3d99584edbfdcc5dbe9d751f0c224c28 /fs/exec.c | |
| parent | f3ce0064cbb6bbada20942729aa439cfd33da301 (diff) | |
[PATCH] threaded coredumps, tcore-fixes-2.5.51-A0
This fixes one more threaded-coredumps detail reported by the glibc
people: all threads taken down by the coredump code should report the
proper exit code. We can do this rather easily via the group_exit
mechanism. 'Other' threads used to report SIGKILL, which was highly
confusing as the shell often displayed the 'Killed' message instead of a
'Segmentation fault' message.
Another missing bit was the 0x80 bit set in the exit status for all
threads, if the coredump was successful. (it's safe to set this bit in
->sig->group_exit_code in an unlocked way because all threads are
artificially descheduled by the coredump code.)
Diffstat (limited to 'fs/exec.c')
| -rw-r--r-- | fs/exec.c | 5 |
1 files changed, 4 insertions, 1 deletions
diff --git a/fs/exec.c b/fs/exec.c index 4c0937c43144..1699f5bdf92e 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1268,7 +1268,7 @@ static void coredump_wait(struct mm_struct *mm) BUG_ON(mm->core_waiters); } -int do_coredump(long signr, struct pt_regs * regs) +int do_coredump(long signr, int exit_code, struct pt_regs * regs) { char corename[CORENAME_MAX_SIZE + 1]; struct mm_struct *mm = current->mm; @@ -1288,6 +1288,8 @@ int do_coredump(long signr, struct pt_regs * regs) } mm->dumpable = 0; init_completion(&mm->core_done); + current->sig->group_exit = 1; + current->sig->group_exit_code = exit_code; coredump_wait(mm); if (current->rlim[RLIMIT_CORE].rlim_cur < binfmt->min_coredump) @@ -1314,6 +1316,7 @@ int do_coredump(long signr, struct pt_regs * regs) retval = binfmt->core_dump(signr, regs, file); + current->sig->group_exit_code |= 0x80; close_fail: filp_close(file, NULL); fail_unlock: |