linux-kernel - Re: [RFC] ubsan: signed integer overflow in mem_cgroup_event

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] [day] [month] [year] [list]

Message-ID: <20170616150057.GQ30580@dhcp22.suse.cz>
Date:   Fri, 16 Jun 2017 17:00:57 +0200
From:   Michal Hocko <mhocko@...nel.org>
To:     Alice Ferrazzi <alicef@...too.org>
Cc:     hannes@...xchg.org, vdavydov.dev@...il.com,
        cgroups@...r.kernel.org, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org,
        Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [RFC] ubsan: signed integer overflow in
 mem_cgroup_event_ratelimit

[CC Andrew]

On Fri 16-06-17 21:26:53, Alice Ferrazzi wrote:
> Hello,
> 
> a user reported a UBSAN signed integer overflow in memcontrol.c
> Shall we change something in mem_cgroup_event_ratelimit()?

It took me quite some staring but it seems the report is correct.
---
>From 5683a96c2bfe694b66c54206aabbc36fbbf621bf Mon Sep 17 00:00:00 2001
From: Michal Hocko <mhocko@...e.com>
Date: Fri, 16 Jun 2017 16:56:55 +0200
Subject: [PATCH] mm, memcg: fix potential undefined behavior in
 mem_cgroup_event_ratelimit

Alice has reported the following UBSAN splat:
kernel: UBSAN: Undefined behaviour in mm/memcontrol.c:661:17
kernel: signed integer overflow:
kernel: -2147483644 - 2147483525 cannot be represented in type 'long int'
kernel: CPU: 1 PID: 11758 Comm: mybibtex2filena Tainted: P           O 4.9.25-gentoo #4
kernel: Hardware name: XXXXXX, BIOS YYYYYY
kernel: e9a3bd64 d1f444f2 00000007 e9a3bd94 7fffff85 e9a3bd74 d1fc8ffe e9a3bd74
kernel: d2b4ef1c e9a3bdf8 d1fc934b d28b15c0 e9a3bd98 0000002d e9a3bdc0 d2b4ef1c
kernel: 0000002d 00000002 3431322d 33383437 00343436 d1700ca2 00000000 ecb4effc
kernel: Call Trace:
kernel: [<d1f444f2>] dump_stack+0x59/0x87
kernel: [<d1fc8ffe>] ubsan_epilogue+0xe/0x40
kernel: [<d1fc934b>] handle_overflow+0xbb/0xf0
kernel: [<d1700ca2>] ? update_curr+0xe2/0x500
kernel: [<d1fc93b2>] __ubsan_handle_sub_overflow+0x12/0x20
kernel: [<d196a553>] memcg_check_events.isra.36+0x223/0x360
kernel: [<d1f44281>] ? cpumask_any_but+0x31/0x60
kernel: [<d19709c5>] mem_cgroup_commit_charge+0x55/0x140
kernel: [<d1925b42>] ? ptep_clear_flush+0x72/0xb0
kernel: [<d19017de>] wp_page_copy+0x34e/0xb80
kernel: [<d19037a6>] do_wp_page+0x1e6/0x1300
kernel: [<d16f0350>] ? check_preempt_curr+0x110/0x230
kernel: [<d1695de6>] ? kmap_atomic_prot+0x126/0x210
kernel: [<d1909b3b>] handle_mm_fault+0x88b/0x1990
kernel: [<d16a1905>] ? _do_fork+0x155/0x5b0
kernel: [<d1689e3e>] __do_page_fault+0x2de/0x8a0
kernel: [<d16a1e27>] ? SyS_clone+0x27/0x30
kernel: [<d168a400>] ? __do_page_fault+0x8a0/0x8a0
kernel: [<d168a41a>] do_page_fault+0x1a/0x20
kernel: [<d265a35b>] error_code+0x67/0x6c

the reason is that we subtract two signed types. Let's fix this by truly
mimicing time_after and cast the result of the subtraction.

Reported-by: Alice Ferrazzi <alicef@...too.org>
Signed-off-by: Michal Hocko <mhocko@...e.com>
---
 mm/memcontrol.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index e3fe4d0913b3..544d47e5cbbd 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -631,7 +631,7 @@ static bool mem_cgroup_event_ratelimit(struct mem_cgroup *memcg,
 	val = __this_cpu_read(memcg->stat->nr_page_events);
 	next = __this_cpu_read(memcg->stat->targets[target]);
 	/* from time_after() in jiffies.h */
-	if ((long)next - (long)val < 0) {
+	if ((long)(next - val) < 0) {
 		switch (target) {
 		case MEM_CGROUP_TARGET_THRESH:
 			next = val + THRESHOLDS_EVENTS_TARGET;
-- 
2.11.0

-- 
Michal Hocko
SUSE Labs