check_unshare_flags: kill the bogus CLONE_SIGHAND/sig->count check
Oleg Nesterov [Wed, 26 May 2010 21:43:21 +0000 (14:43 -0700)]
check_unshare_flags(CLONE_SIGHAND) adds CLONE_THREAD to *flags_ptr if the
task is multithreaded to ensure unshare_thread() will fail.

Not only this is a bit strange way to return the error, this is absolutely
meaningless.  If signal->count > 1 then sighand->count must be also > 1,
and unshare_sighand() will fail anyway.

In fact, all CLONE_THREAD/SIGHAND/VM checks inside sys_unshare() do not
look right.  Fortunately this code doesn't really work anyway.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Balbir Singh <balbir@linux.vnet.ibm.com>
Acked-by: Roland McGrath <roland@redhat.com>
Cc: Veaceslav Falico <vfalico@redhat.com>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kernel/fork.c

index 7701470..40cd099 100644 (file)
@@ -1518,14 +1518,6 @@ static void check_unshare_flags(unsigned long *flags_ptr)
                *flags_ptr |= CLONE_SIGHAND;
 
        /*
-        * If unsharing signal handlers and the task was created
-        * using CLONE_THREAD, then must unshare the thread
-        */
-       if ((*flags_ptr & CLONE_SIGHAND) &&
-           (atomic_read(&current->signal->count) > 1))
-               *flags_ptr |= CLONE_THREAD;
-
-       /*
         * If unsharing namespace, must also unshare filesystem information.
         */
        if (*flags_ptr & CLONE_NEWNS)