do_wait() is a big wait loop, but we set TASK_RUNNING too late; we end up calling potential sleeps before we reset it. Not strictly a bug since we're guaranteed to exit the loop and not call schedule(); put in annotations to quiet might_sleep(). WARNING: CPU: 0 PID: 1 at ../kernel/sched/core.c:7123 __might_sleep+0x7e/0x90() do not call blocking ops when !TASK_RUNNING; state=1 set at [] do_wait+0x88/0x270 Call Trace: [] dump_stack+0x4e/0x7a [] warn_slowpath_common+0x8c/0xc0 [] warn_slowpath_fmt+0x4c/0x50 [] __might_sleep+0x7e/0x90 [] might_fault+0x55/0xb0 [] wait_consider_task+0x90b/0xc10 [] do_wait+0x104/0x270 [] SyS_wait4+0x77/0x100 [] system_call_fastpath+0x16/0x1b Signed-off-by: Peter Zijlstra (Intel) --- include/linux/kernel.h | 2 ++ kernel/exit.c | 5 +++++ 2 files changed, 7 insertions(+) --- a/include/linux/kernel.h +++ b/include/linux/kernel.h @@ -175,10 +175,12 @@ extern int _cond_resched(void); */ # define might_sleep() \ do { __might_sleep(__FILE__, __LINE__, 0); might_resched(); } while (0) +# define fixup_sleep() __set_current_state(TASK_RUNNING) #else static inline void __might_sleep(const char *file, int line, int preempt_offset) { } # define might_sleep() do { might_resched(); } while (0) +# define fixup_sleep() do { } while (0) #endif #define might_sleep_if(cond) do { if (cond) might_sleep(); } while (0) --- a/kernel/exit.c +++ b/kernel/exit.c @@ -997,6 +997,8 @@ static int wait_task_zombie(struct wait_ get_task_struct(p); read_unlock(&tasklist_lock); + fixup_sleep(); + if ((exit_code & 0x7f) == 0) { why = CLD_EXITED; status = exit_code >> 8; @@ -1079,6 +1081,7 @@ static int wait_task_zombie(struct wait_ * thread can reap it because we its state == DEAD/TRACE. */ read_unlock(&tasklist_lock); + fixup_sleep(); retval = wo->wo_rusage ? getrusage(p, RUSAGE_BOTH, wo->wo_rusage) : 0; @@ -1210,6 +1213,7 @@ static int wait_task_stopped(struct wait pid = task_pid_vnr(p); why = ptrace ? CLD_TRAPPED : CLD_STOPPED; read_unlock(&tasklist_lock); + fixup_sleep(); if (unlikely(wo->wo_flags & WNOWAIT)) return wait_noreap_copyout(wo, p, pid, uid, why, exit_code); @@ -1272,6 +1276,7 @@ static int wait_task_continued(struct wa pid = task_pid_vnr(p); get_task_struct(p); read_unlock(&tasklist_lock); + fixup_sleep(); if (!wo->wo_info) { retval = wo->wo_rusage -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/