[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <tip-bf29ed1567b67854dc13504f685c45a2ea9b2081@git.kernel.org>
Date: Sun, 17 Sep 2017 10:53:35 -0700
From: tip-bot for Thomas Garnier <tipbot@...or.com>
To: linux-tip-commits@...r.kernel.org
Cc: hpa@...or.com, linux@...linux.org.uk, arnd@...db.de,
luto@...capital.net, panand@...hat.com, wad@...omium.org,
catalin.marinas@....com, linux-kernel@...r.kernel.org,
viro@...iv.linux.org.uk, Dave.Martin@....com, thgarnie@...gle.com,
will.deacon@....com, dave.hansen@...el.com, dhowells@...hat.com,
tglx@...utronix.de, mingo@...nel.org, yhs@...com,
keescook@...omium.org
Subject: [tip:core/urgent] syscalls: Use CHECK_DATA_CORRUPTION for
addr_limit_user_check
Commit-ID: bf29ed1567b67854dc13504f685c45a2ea9b2081
Gitweb: http://git.kernel.org/tip/bf29ed1567b67854dc13504f685c45a2ea9b2081
Author: Thomas Garnier <thgarnie@...gle.com>
AuthorDate: Thu, 7 Sep 2017 08:30:44 -0700
Committer: Thomas Gleixner <tglx@...utronix.de>
CommitDate: Sun, 17 Sep 2017 19:45:32 +0200
syscalls: Use CHECK_DATA_CORRUPTION for addr_limit_user_check
Use CHECK_DATA_CORRUPTION instead of BUG_ON to provide more flexibility
on address limit failures. By default, send a SIGKILL signal to kill the
current process preventing exploitation of a bad address limit.
Make the TIF_FSCHECK flag optional so ARM can use this function.
Signed-off-by: Thomas Garnier <thgarnie@...gle.com>
Signed-off-by: Kees Cook <keescook@...omium.org>
Signed-off-by: Thomas Gleixner <tglx@...utronix.de>
Cc: Pratyush Anand <panand@...hat.com>
Cc: Dave Martin <Dave.Martin@....com>
Cc: Will Drewry <wad@...omium.org>
Cc: Arnd Bergmann <arnd@...db.de>
Cc: Catalin Marinas <catalin.marinas@....com>
Cc: Will Deacon <will.deacon@....com>
Cc: Russell King <linux@...linux.org.uk>
Cc: Andy Lutomirski <luto@...capital.net>
Cc: David Howells <dhowells@...hat.com>
Cc: Dave Hansen <dave.hansen@...el.com>
Cc: Al Viro <viro@...iv.linux.org.uk>
Cc: linux-api@...r.kernel.org
Cc: Yonghong Song <yhs@...com>
Cc: linux-arm-kernel@...ts.infradead.org
Link: http://lkml.kernel.org/r/1504798247-48833-2-git-send-email-keescook@chromium.org
---
include/linux/syscalls.h | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index 95606a2..a78186d 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -221,21 +221,25 @@ static inline int is_syscall_trace_event(struct trace_event_call *tp_event)
} \
static inline long SYSC##name(__MAP(x,__SC_DECL,__VA_ARGS__))
-#ifdef TIF_FSCHECK
/*
* Called before coming back to user-mode. Returning to user-mode with an
* address limit different than USER_DS can allow to overwrite kernel memory.
*/
static inline void addr_limit_user_check(void)
{
-
+#ifdef TIF_FSCHECK
if (!test_thread_flag(TIF_FSCHECK))
return;
+#endif
- BUG_ON(!segment_eq(get_fs(), USER_DS));
+ if (CHECK_DATA_CORRUPTION(!segment_eq(get_fs(), USER_DS),
+ "Invalid address limit on user-mode return"))
+ force_sig(SIGKILL, current);
+
+#ifdef TIF_FSCHECK
clear_thread_flag(TIF_FSCHECK);
-}
#endif
+}
asmlinkage long sys32_quotactl(unsigned int cmd, const char __user *special,
qid_t id, void __user *addr);
Powered by blists - more mailing lists