diff mbox series

[5.16,38/43] bpf: Fix possible race in inc_misses_counter

Message ID	20220204091918.402198487@linuxfoundation.org
State	New
Headers	show Return-Path: <stable-owner@kernel.org> From: Greg Kroah-Hartman <gregkh@linuxfoundation.org> To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>, stable@vger.kernel.org, He Fengqing <hefengqing@huawei.com>, John Fastabend <john.fastabend@gmail.com>, Alexei Starovoitov <ast@kernel.org> Subject: [PATCH 5.16 38/43] bpf: Fix possible race in inc_misses_counter Date: Fri, 4 Feb 2022 10:22:45 +0100 Message-Id: <20220204091918.402198487@linuxfoundation.org> In-Reply-To: <20220204091917.166033635@linuxfoundation.org> References: <20220204091917.166033635@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Precedence: bulk
Series	None \| expand [5.16,04/43] net: ipa: prevent concurrent replenish [5.16,05/43] drm/vc4: hdmi: Make sure the device is powered with CEC [5.16,08/43] net: phy: Fix qca8081 with speeds lower than 2.5Gb/s [5.16,09/43] ovl: dont fail copy up if no fileattr support on upper [5.16,10/43] lockd: fix server crash on reboot of client holding lock [5.16,12/43] net/mlx5e: IPsec: Fix crypto offload for non TCP/UDP encapsulated traffic [5.16,14/43] net/mlx5e: TC, Reject rules with drop and modify hdr action [5.16,15/43] net/mlx5: Bridge, take rtnl lock in init error handler [5.16,18/43] net/mlx5: Use del_timer_sync in fw reset flow of halting poll [5.16,19/43] net/mlx5e: Fix module EEPROM query [5.16,20/43] net/mlx5e: TC, Reject rules with forward and drop actions [5.16,24/43] net/mlx5e: Avoid field-overflowing memcpy() [5.16,25/43] net/mlx5e: Fix wrong calculation of header index in HW_GRO [5.16,26/43] net/mlx5e: Fix broken SKB allocation in HW-GRO [5.16,28/43] net/mlx5e: Avoid implicit modify hdr for decap drop rule [5.16,29/43] ipheth: fix EOVERFLOW in ipheth_rcvbulk_callback [5.16,32/43] net: amd-xgbe: ensure to reset the tx_timer_active flag [5.16,34/43] fanotify: Fix stale file descriptor in copy_event_to_user() [5.16,35/43] net: sched: fix use-after-free in tc_new_tfilter() [5.16,38/43] bpf: Fix possible race in inc_misses_counter [5.16,40/43] e1000e: Handshake with CSME starts from ADL platforms [5.16,42/43] tcp: fix mem under-charging with zerocopy sendmsg()

Commit Message

Greg KH Feb. 4, 2022, 9:22 a.m. UTC

From: He Fengqing <hefengqing@huawei.com>

commit 0e3135d3bfa5dfb658145238d2bc723a8e30c3a3 upstream.

It seems inc_misses_counter() suffers from same issue fixed in
the commit d979617aa84d ("bpf: Fixes possible race in update_prog_stats()
for 32bit arches"):
As it can run while interrupts are enabled, it could
be re-entered and the u64_stats syncp could be mangled.

Fixes: 9ed9e9ba2337 ("bpf: Count the number of times recursion was prevented")
Signed-off-by: He Fengqing <hefengqing@huawei.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/r/20220122102936.1219518-1-hefengqing@huawei.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 kernel/bpf/trampoline.c |    5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff mbox series

Patch

--- a/kernel/bpf/trampoline.c
+++ b/kernel/bpf/trampoline.c
@@ -542,11 +542,12 @@  static __always_inline u64 notrace bpf_p
 static void notrace inc_misses_counter(struct bpf_prog *prog)
 {
 	struct bpf_prog_stats *stats;
+	unsigned int flags;
 
 	stats = this_cpu_ptr(prog->stats);
-	u64_stats_update_begin(&stats->syncp);
+	flags = u64_stats_update_begin_irqsave(&stats->syncp);
 	u64_stats_inc(&stats->misses);
-	u64_stats_update_end(&stats->syncp);
+	u64_stats_update_end_irqrestore(&stats->syncp, flags);
 }
 
 /* The logic is similar to bpf_prog_run(), but with an explicit