The patch titled
fix sys_unshare()+SEM_UNDO: add support for CLONE_SYSVSEM
has been removed from the -mm tree. Its filename was
fix-sys_unsharesem_undo-add-support-for-clone_sysvsem.patch
This patch was dropped because an updated version will be merged
The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/
------------------------------------------------------
Subject: fix sys_unshare()+SEM_UNDO: add support for CLONE_SYSVSEM
From: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
sys_unshare(CLONE_NEWIPC) doesn't handle the undo lists properly, this can
cause a kernel memory corruption. CLONE_NEWIPC must detach from the existing
undo lists.
Fix, part 1: add support for sys_unshare(CLONE_SYSVSEM)
Signed-off-by: Manfred Spraul <manfred@xxxxxxxxxxxxxxxx>
Cc: "Serge E. Hallyn" <serue@xxxxxxxxxx>
Cc: "Eric W. Biederman" <ebiederm@xxxxxxxxxxxx>
Cc: Pavel Emelyanov <xemul@xxxxxxxxxx>
Cc: Sukadev Bhattiprolu <sukadev@xxxxxxxxxx>
Cc: "Pierre Peiffer" <peifferp@xxxxxxxxx>
Cc: Nadia Derbey <Nadia.Derbey@xxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
ipc/sem.c | 1 +
kernel/fork.c | 18 ++++++++++++++----
2 files changed, 15 insertions(+), 4 deletions(-)
diff -puN ipc/sem.c~fix-sys_unsharesem_undo-add-support-for-clone_sysvsem ipc/sem.c
--- a/ipc/sem.c~fix-sys_unsharesem_undo-add-support-for-clone_sysvsem
+++ a/ipc/sem.c
@@ -1298,6 +1298,7 @@ void exit_sem(struct task_struct *tsk)
undo_list = tsk->sysvsem.undo_list;
if (!undo_list)
return;
+ tsk->sysvsem.undo_list = NULL;
if (!atomic_dec_and_test(&undo_list->refcnt))
return;
diff -puN kernel/fork.c~fix-sys_unsharesem_undo-add-support-for-clone_sysvsem kernel/fork.c
--- a/kernel/fork.c~fix-sys_unsharesem_undo-add-support-for-clone_sysvsem
+++ a/kernel/fork.c
@@ -1675,13 +1675,17 @@ static int unshare_fd(unsigned long unsh
}
/*
- * Unsharing of semundo for tasks created with CLONE_SYSVSEM is not
- * supported yet
+ * Unsharing of semundo for tasks created with CLONE_SYSVSEM doesn't require
+ * any allocations: it means that the task leaves the existing undo lists,
+ * just like sys_exit(). The new undo lists are allocated on demand in the
+ * ipc syscalls.
+ * new_ulistp is set to a non-NULL value, the caller expects that on success.
*/
static int unshare_semundo(unsigned long unshare_flags, struct sem_undo_list **new_ulistp)
{
- if (unshare_flags & CLONE_SYSVSEM)
- return -EINVAL;
+ if (unshare_flags & CLONE_SYSVSEM) {
+ *new_ulistp = (void*)1;
+ }
return 0;
}
@@ -1731,6 +1735,12 @@ asmlinkage long sys_unshare(unsigned lon
goto bad_unshare_cleanup_semundo;
if (new_fs || new_mm || new_fd || new_ulist || new_nsproxy) {
+ if (unshare_flags & CLONE_SYSVSEM) {
+ /*
+ * CLONE_SYSVSEM is equivalent to sys_exit().
+ */
+ exit_sem(current);
+ }
if (new_nsproxy) {
switch_task_namespaces(current, new_nsproxy);
_
Patches currently in -mm which might be from manfred@xxxxxxxxxxxxxxxx are
fix-sys_unsharesem_undo-add-support-for-clone_sysvsem.patch
fix-sys_unsharesem_undo-perform-an-implicit-clone_sysvsem-in-clone_newipc.patch
mm-debug-dump-pageframes-on-bad_page.patch
slab-leaks3-default-y.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
