redonkable
/
remarkable-linux
1
0
Fork 0
Code Releases Activity

exit: wait: don't use zombie->real_parent 

1. wait_task_zombie() uses p->real_parent to get psig/siglock. This is
   correct but needs tasklist_lock, ->real_parent can exit.

   We can use "current" instead. This is our natural child, its parent
   must be our sub-thread.

2. Read psig/sig outside of ->siglock, ->signal is no longer protected
   by this lock.

3. Fix the outdated comments about tasklist_lock. We can not race with
   __exit_signal(), the whole thread group is dead, nobody but us can
   call it.

   Also clarify the usage of ->stats_lock and ->siglock.

Note: thread_group_cputime_adjusted() is sub-optimal in this case, we
probably want to export cputime_adjust() to avoid thread_group_cputime().
The comment says "all threads" but there are no other threads.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Aaron Tomlin <atomlin@redhat.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Sterling Alexander <stalexan@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
wifi-calibration
Oleg Nesterov 2014-12-10 15:54:48 -08:00 committed by Linus Torvalds
parent f6507f83bc
commit f953ccd006
1 changed files with 11 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
kernel/exit.c
View File

@ -1010,8 +1010,8 @@ static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p)
* Check thread_group_leader() to exclude the traced sub-threads.
*/
if (state == EXIT_DEAD && thread_group_leader(p)) {
struct signal_struct *psig;
struct signal_struct *sig;
struct signal_struct *sig = p->signal;
struct signal_struct *psig = current->signal;
unsigned long maxrss;
cputime_t tgutime, tgstime;
@ -1023,21 +1023,20 @@ static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p)
* accumulate in the parent's signal_struct c* fields.
*
* We don't bother to take a lock here to protect these
* p->signal fields, because they are only touched by
* __exit_signal, which runs with tasklist_lock
* write-locked anyway, and so is excluded here. We do
* need to protect the access to parent->signal fields,
* as other threads in the parent group can be right
* here reaping other children at the same time.
* p->signal fields because the whole thread group is dead
* and nobody can change them.
*
* psig->stats_lock also protects us from our sub-theads
* which can reap other children at the same time. Until
* we change k_getrusage()-like users to rely on this lock
* we have to take ->siglock as well.
*
* We use thread_group_cputime_adjusted() to get times for
* the thread group, which consolidates times for all threads
* in the group including the group leader.
*/
thread_group_cputime_adjusted(p, &tgutime, &tgstime);
spin_lock_irq(&p->real_parent->sighand->siglock);
psig = p->real_parent->signal;
sig = p->signal;
spin_lock_irq(&current->sighand->siglock);
write_seqlock(&psig->stats_lock);
psig->cutime += tgutime + sig->cutime;
psig->cstime += tgstime + sig->cstime;
@ -1062,7 +1061,7 @@ static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p)
task_io_accounting_add(&psig->ioac, &p->ioac);
task_io_accounting_add(&psig->ioac, &sig->ioac);
write_sequnlock(&psig->stats_lock);
spin_unlock_irq(&p->real_parent->sighand->siglock);
spin_unlock_irq(&current->sighand->siglock);
}
/*