From 5d53f2c26f29e917b401e8bb539ba85e52725e8c Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Mon, 23 May 2016 16:23:50 -0700 Subject: [PATCH 1/1] wait/ptrace: assume __WALL if the child is traced commit bf959931ddb88c4e4366e96dd22e68fa0db9527c upstream. The following program (simplified version of generated by syzkaller) #include #include #include #include #include void *thread_func(void *arg) { ptrace(PTRACE_TRACEME, 0,0,0); return 0; } int main(void) { pthread_t thread; if (fork()) return 0; while (getppid() != 1) ; pthread_create(&thread, NULL, thread_func, NULL); pthread_join(thread, NULL); return 0; } creates an unreapable zombie if /sbin/init doesn't use __WALL. This is not a kernel bug, at least in a sense that everything works as expected: debugger should reap a traced sub-thread before it can reap the leader, but without __WALL/__WCLONE do_wait() ignores sub-threads. Unfortunately, it seems that /sbin/init in most (all?) distributions doesn't use it and we have to change the kernel to avoid the problem. Note also that most init's use sys_waitid() which doesn't allow __WALL, so the necessary user-space fix is not that trivial. This patch just adds the "ptrace" check into eligible_child(). To some degree this matches the "tsk->ptrace" in exit_notify(), ->exit_signal is mostly ignored when the tracee reports to debugger. Or WSTOPPED, the tracer doesn't need to set this flag to wait for the stopped tracee. This obviously means the user-visible change: __WCLONE and __WALL no longer have any meaning for debugger. And I can only hope that this won't break something, but at least strace/gdb won't suffer. We could make a more conservative change. Say, we can take __WCLONE into account, or !thread_group_leader(). But it would be nice to not complicate these historical/confusing checks. Signed-off-by: Oleg Nesterov Reported-by: Dmitry Vyukov Cc: Denys Vlasenko Cc: Jan Kratochvil Cc: "Michael Kerrisk (man-pages)" Cc: Pedro Alves Cc: Roland McGrath Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds [bwh: Backported to 3.2: adjust context] Signed-off-by: Ben Hutchings --- kernel/exit.c | 29 ++++++++++++++++++++--------- 1 file changed, 20 insertions(+), 9 deletions(-) diff --git a/kernel/exit.c b/kernel/exit.c index fde15f96b158..ee5a8786c3d2 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -1138,17 +1138,28 @@ static int eligible_pid(struct wait_opts *wo, struct task_struct *p) task_pid_type(p, wo->wo_type) == wo->wo_pid; } -static int eligible_child(struct wait_opts *wo, struct task_struct *p) +static int +eligible_child(struct wait_opts *wo, bool ptrace, struct task_struct *p) { if (!eligible_pid(wo, p)) return 0; - /* Wait for all children (clone and not) if __WALL is set; - * otherwise, wait for clone children *only* if __WCLONE is - * set; otherwise, wait for non-clone children *only*. (Note: - * A "clone" child here is one that reports to its parent - * using a signal other than SIGCHLD.) */ - if (((p->exit_signal != SIGCHLD) ^ !!(wo->wo_flags & __WCLONE)) - && !(wo->wo_flags & __WALL)) + + /* + * Wait for all children (clone and not) if __WALL is set or + * if it is traced by us. + */ + if (ptrace || (wo->wo_flags & __WALL)) + return 1; + + /* + * Otherwise, wait for clone children *only* if __WCLONE is set; + * otherwise, wait for non-clone children *only*. + * + * Note: a "clone" child here is one that reports to its parent + * using a signal other than SIGCHLD, or a non-leader thread which + * we can only see if it is traced by us. + */ + if ((p->exit_signal != SIGCHLD) ^ !!(wo->wo_flags & __WCLONE)) return 0; return 1; @@ -1525,7 +1536,7 @@ static int wait_task_continued(struct wait_opts *wo, struct task_struct *p) static int wait_consider_task(struct wait_opts *wo, int ptrace, struct task_struct *p) { - int ret = eligible_child(wo, p); + int ret = eligible_child(wo, ptrace, p); if (!ret) return ret; -- 2.39.2