* [PATCH] sys_unshare: remove a broken CLONE_SIGHAND code
@ 2006-10-23 15:53 Oleg Nesterov
2006-10-24 9:49 ` Peter Zijlstra
0 siblings, 1 reply; 2+ messages in thread
From: Oleg Nesterov @ 2006-10-23 15:53 UTC (permalink / raw)
To: Andrew Morton; +Cc: Janak Desai, Peter Zijlstra, Alan Cox, linux-kernel
sys_unshare(CLONE_SIGHAND) is broken, the code under 'if (new_sigh)' is
never executed but very wrong. Just remove it to avoid a confusion,
task_lock() has nothing to do with ->sighand changing.
Also, change the comment in unshare_sighand(). Yes, CLONE_THREAD implies
CLONE_SIGHAND, but still it looks confusing. Also, we don't need to check
current->sighand != NULL.
Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
--- rc2-mm2/kernel/fork.c~ 2006-10-22 19:28:17.000000000 +0400
+++ rc2-mm2/kernel/fork.c 2006-10-23 19:23:19.000000000 +0400
@@ -1523,15 +1523,13 @@ static int unshare_mnt_namespace(unsigne
}
/*
- * Unsharing of sighand for tasks created with CLONE_SIGHAND is not
- * supported yet
+ * Unsharing of sighand is not supported yet
*/
static int unshare_sighand(unsigned long unshare_flags, struct sighand_struct **new_sighp)
{
struct sighand_struct *sigh = current->sighand;
- if ((unshare_flags & CLONE_SIGHAND) &&
- (sigh && atomic_read(&sigh->count) > 1))
+ if ((unshare_flags & CLONE_SIGHAND) && atomic_read(&sigh->count) > 1)
return -EINVAL;
else
return 0;
@@ -1605,7 +1603,7 @@ asmlinkage long sys_unshare(unsigned lon
int err = 0;
struct fs_struct *fs, *new_fs = NULL;
struct mnt_namespace *ns, *new_ns = NULL;
- struct sighand_struct *sigh, *new_sigh = NULL;
+ struct sighand_struct *new_sigh = NULL;
struct mm_struct *mm, *new_mm = NULL, *active_mm = NULL;
struct files_struct *fd, *new_fd = NULL;
struct sem_undo_list *new_ulist = NULL;
@@ -1650,7 +1648,7 @@ asmlinkage long sys_unshare(unsigned lon
}
}
- if (new_fs || new_ns || new_sigh || new_mm || new_fd || new_ulist ||
+ if (new_fs || new_ns || new_mm || new_fd || new_ulist ||
new_uts || new_ipc) {
task_lock(current);
@@ -1672,12 +1670,6 @@ asmlinkage long sys_unshare(unsigned lon
new_ns = ns;
}
- if (new_sigh) {
- sigh = current->sighand;
- rcu_assign_pointer(current->sighand, new_sigh);
- new_sigh = sigh;
- }
-
if (new_mm) {
mm = current->mm;
active_mm = current->active_mm;
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH] sys_unshare: remove a broken CLONE_SIGHAND code
2006-10-23 15:53 [PATCH] sys_unshare: remove a broken CLONE_SIGHAND code Oleg Nesterov
@ 2006-10-24 9:49 ` Peter Zijlstra
0 siblings, 0 replies; 2+ messages in thread
From: Peter Zijlstra @ 2006-10-24 9:49 UTC (permalink / raw)
To: Oleg Nesterov; +Cc: Andrew Morton, Janak Desai, Alan Cox, linux-kernel
On Mon, 2006-10-23 at 19:53 +0400, Oleg Nesterov wrote:
> sys_unshare(CLONE_SIGHAND) is broken, the code under 'if (new_sigh)' is
> never executed but very wrong. Just remove it to avoid a confusion,
> task_lock() has nothing to do with ->sighand changing.
>
> Also, change the comment in unshare_sighand(). Yes, CLONE_THREAD implies
> CLONE_SIGHAND, but still it looks confusing. Also, we don't need to check
> current->sighand != NULL.
>
> Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Good thing, this confused me when figuring out the ->siglock rules, its
tricky enough without being put off balance by dead code.
> --- rc2-mm2/kernel/fork.c~ 2006-10-22 19:28:17.000000000 +0400
> +++ rc2-mm2/kernel/fork.c 2006-10-23 19:23:19.000000000 +0400
> @@ -1523,15 +1523,13 @@ static int unshare_mnt_namespace(unsigne
> }
>
> /*
> - * Unsharing of sighand for tasks created with CLONE_SIGHAND is not
> - * supported yet
> + * Unsharing of sighand is not supported yet
> */
> static int unshare_sighand(unsigned long unshare_flags, struct sighand_struct **new_sighp)
> {
> struct sighand_struct *sigh = current->sighand;
>
> - if ((unshare_flags & CLONE_SIGHAND) &&
> - (sigh && atomic_read(&sigh->count) > 1))
> + if ((unshare_flags & CLONE_SIGHAND) && atomic_read(&sigh->count) > 1)
> return -EINVAL;
> else
> return 0;
> @@ -1605,7 +1603,7 @@ asmlinkage long sys_unshare(unsigned lon
> int err = 0;
> struct fs_struct *fs, *new_fs = NULL;
> struct mnt_namespace *ns, *new_ns = NULL;
> - struct sighand_struct *sigh, *new_sigh = NULL;
> + struct sighand_struct *new_sigh = NULL;
> struct mm_struct *mm, *new_mm = NULL, *active_mm = NULL;
> struct files_struct *fd, *new_fd = NULL;
> struct sem_undo_list *new_ulist = NULL;
> @@ -1650,7 +1648,7 @@ asmlinkage long sys_unshare(unsigned lon
> }
> }
>
> - if (new_fs || new_ns || new_sigh || new_mm || new_fd || new_ulist ||
> + if (new_fs || new_ns || new_mm || new_fd || new_ulist ||
> new_uts || new_ipc) {
>
> task_lock(current);
> @@ -1672,12 +1670,6 @@ asmlinkage long sys_unshare(unsigned lon
> new_ns = ns;
> }
>
> - if (new_sigh) {
> - sigh = current->sighand;
> - rcu_assign_pointer(current->sighand, new_sigh);
> - new_sigh = sigh;
> - }
> -
> if (new_mm) {
> mm = current->mm;
> active_mm = current->active_mm;
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2006-10-24 9:49 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-10-23 15:53 [PATCH] sys_unshare: remove a broken CLONE_SIGHAND code Oleg Nesterov
2006-10-24 9:49 ` Peter Zijlstra
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox