linux-kernel - [PATCH 11/16] misc: use ACCESS

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <1258555922-2064-11-git-send-email-jslaby@novell.com>
Date:	Wed, 18 Nov 2009 15:51:57 +0100
From:	Jiri Slaby <jslaby@...ell.com>
To:	jirislaby@...il.com
Cc:	mingo@...e.hu, nhorman@...driver.com, sfr@...b.auug.org.au,
	linux-kernel@...r.kernel.org, akpm@...ux-foundation.org,
	marcin.slusarz@...il.com, tglx@...utronix.de, mingo@...hat.com,
	hpa@...or.com, torvalds@...ux-foundation.org,
	Jiri Slaby <jslaby@...ell.com>,
	James Morris <jmorris@...ei.org>,
	Heiko Carstens <heiko.carstens@...ibm.com>,
	Roland Dreier <rolandd@...co.com>,
	Sean Hefty <sean.hefty@...el.com>,
	Hal Rosenstock <hal.rosenstock@...il.com>,
	linux-rdma@...r.kernel.org
Subject: [PATCH 11/16] misc: use ACCESS_ONCE for rlimits

Make sure compiler won't do weird things with limits. E.g. fetching
them twice may return 2 different values after writable limits are
implemented.

Signed-off-by: Jiri Slaby <jslaby@...ell.com>
Cc: James Morris <jmorris@...ei.org>
Cc: Heiko Carstens <heiko.carstens@...ibm.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>
Cc: Ingo Molnar <mingo@...e.hu>
Cc: Roland Dreier <rolandd@...co.com>
Cc: Sean Hefty <sean.hefty@...el.com>
Cc: Hal Rosenstock <hal.rosenstock@...il.com>
Cc: linux-rdma@...r.kernel.org
---
 drivers/infiniband/core/umem.c                 |    3 ++-
 drivers/infiniband/hw/ipath/ipath_user_pages.c |    4 ++--
 ipc/mqueue.c                                   |    4 ++--
 ipc/shm.c                                      |    4 ++--
 4 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/infiniband/core/umem.c b/drivers/infiniband/core/umem.c
index 6f7c096..90d806b 100644
--- a/drivers/infiniband/core/umem.c
+++ b/drivers/infiniband/core/umem.c
@@ -136,7 +136,8 @@ struct ib_umem *ib_umem_get(struct ib_ucontext *context, unsigned long addr,
 	down_write(&current->mm->mmap_sem);
 
 	locked     = npages + current->mm->locked_vm;
-	lock_limit = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur >> PAGE_SHIFT;
+	lock_limit = ACCESS_ONCE(current->signal->
+			rlim[RLIMIT_MEMLOCK].rlim_cur) >> PAGE_SHIFT;
 
 	if ((locked > lock_limit) && !capable(CAP_IPC_LOCK)) {
 		ret = -ENOMEM;
diff --git a/drivers/infiniband/hw/ipath/ipath_user_pages.c b/drivers/infiniband/hw/ipath/ipath_user_pages.c
index 82878e3..f7db156 100644
--- a/drivers/infiniband/hw/ipath/ipath_user_pages.c
+++ b/drivers/infiniband/hw/ipath/ipath_user_pages.c
@@ -59,8 +59,8 @@ static int __get_user_pages(unsigned long start_page, size_t num_pages,
 	size_t got;
 	int ret;
 
-	lock_limit = current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur >>
-		PAGE_SHIFT;
+	lock_limit = ACCESS_ONCE(current->signal->
+			rlim[RLIMIT_MEMLOCK].rlim_cur) >> PAGE_SHIFT;
 
 	if (num_pages > lock_limit) {
 		ret = -ENOMEM;
diff --git a/ipc/mqueue.c b/ipc/mqueue.c
index ee9d697..d04869f 100644
--- a/ipc/mqueue.c
+++ b/ipc/mqueue.c
@@ -152,8 +152,8 @@ static struct inode *mqueue_get_inode(struct super_block *sb,
 
 			spin_lock(&mq_lock);
 			if (u->mq_bytes + mq_bytes < u->mq_bytes ||
-		 	    u->mq_bytes + mq_bytes >
-			    p->signal->rlim[RLIMIT_MSGQUEUE].rlim_cur) {
+			    u->mq_bytes + mq_bytes > ACCESS_ONCE(p->signal->
+			    rlim[RLIMIT_MSGQUEUE].rlim_cur)) {
 				spin_unlock(&mq_lock);
 				goto out_inode;
 			}
diff --git a/ipc/shm.c b/ipc/shm.c
index 464694e..99de87c 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -761,8 +761,8 @@ SYSCALL_DEFINE3(shmctl, int, shmid, int, cmd, struct shmid_ds __user *, buf)
 			if (euid != shp->shm_perm.uid &&
 			    euid != shp->shm_perm.cuid)
 				goto out_unlock;
-			if (cmd == SHM_LOCK &&
-			    !current->signal->rlim[RLIMIT_MEMLOCK].rlim_cur)
+			if (cmd == SHM_LOCK && !ACCESS_ONCE(current->signal->
+					rlim[RLIMIT_MEMLOCK].rlim_cur))
 				goto out_unlock;
 		}
 
-- 
1.6.4.2

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/