Techwizz/kernel-fxtec-pro1x
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

pids: sys_getpgid: fix unsafe *pid usage, s/tasklist/rcu/

Browse source 
1. sys_getpgid() needs rcu_read_lock() to derive the pgrp _nr, even if
   the task is current, otherwise we can race with another thread which
   does sys_setpgid().

2. Use rcu_read_lock() instead of tasklist_lock when pid != 0, make sure
   that we don't use the NULL pid if the task exits right after successful
   find_task_by_vpid().

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
Oleg Nesterov 2008-04-30 00:54:29 -07:00 committed by Linus Torvalds
parent 1dd768c081
commit 12a3de0a96
1 changed files with 22 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

38
kernel/sys.c
View file

@ -991,31 +991,37 @@ asmlinkage long sys_setpgid(pid_t pid, pid_t pgid)
asmlinkage long sys_getpgid(pid_t pid)
{
if (!pid)
return task_pgrp_vnr(current);
else {
int retval;
struct task_struct *p;
struct task_struct *p;
struct pid *grp;
int retval;
read_lock(&tasklist_lock);
p = find_task_by_vpid(pid);
rcu_read_lock();
if (!pid)
grp = task_pgrp(current);
else {
retval = -ESRCH;
if (p) {
retval = security_task_getpgid(p);
if (!retval)
retval = task_pgrp_vnr(p);
}
read_unlock(&tasklist_lock);
return retval;
p = find_task_by_vpid(pid);
if (!p)
goto out;
grp = task_pgrp(p);
if (!grp)
goto out;
retval = security_task_getpgid(p);
if (retval)
goto out;
}
retval = pid_vnr(grp);
out:
rcu_read_unlock();
return retval;
}
#ifdef __ARCH_WANT_SYS_GETPGRP
asmlinkage long sys_getpgrp(void)
{
/* SMP - assuming writes are word atomic this is fine */
return task_pgrp_vnr(current);
return sys_getpgid(0);
}
#endif