Thread (38 messages) 38 messages, 1 author, 2019-10-26
STALE2420d

[PATCH AUTOSEL 4.19 34/59] net: avoid possible false sharing in sk_leave_memory_pressure()

From: Sasha Levin <sashal@kernel.org>
Date: 2019-10-26 13:20:35
Also in: lkml, stable
Subsystem: networking [general], networking [sockets], the rest · Maintainers: "David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, Kuniyuki Iwashima, Willem de Bruijn, Linus Torvalds

From: Eric Dumazet <edumazet@google.com>

[ Upstream commit 503978aca46124cd714703e180b9c8292ba50ba7 ]

As mentioned in https://github.com/google/ktsan/wiki/READ_ONCE-and-WRITE_ONCE#it-may-improve-performance
a C compiler can legally transform :

if (memory_pressure && *memory_pressure)
        *memory_pressure = 0;

to :

if (memory_pressure)
        *memory_pressure = 0;

Fixes: 0604475119de ("tcp: add TCPMemoryPressuresChrono counter")
Fixes: 180d8cd942ce ("foundations of per-cgroup memory pressure controlling.")
Fixes: 3ab224be6d69 ("[NET] CORE: Introducing new memory accounting interface.")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <redacted>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/core/sock.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/core/sock.c b/net/core/sock.c
index f881eea1c4a47..fa38a2d3cb106 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2179,8 +2179,8 @@ static void sk_leave_memory_pressure(struct sock *sk)
 	} else {
 		unsigned long *memory_pressure = sk->sk_prot->memory_pressure;
 
-		if (memory_pressure && *memory_pressure)
-			*memory_pressure = 0;
+		if (memory_pressure && READ_ONCE(*memory_pressure))
+			WRITE_ONCE(*memory_pressure, 0);
 	}
 }
 
-- 
2.20.1
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help