From: Christoph Lameter <cl@linux.com>
To: Tejun Heo <tj@kernel.org>
Cc: akpm@linuxfoundation.org, rostedt@goodmis.org,
linux-kernel@vger.kernel.org, Ingo Molnar <mingo@kernel.org>,
Peter Zijlstra <peterz@infradead.org>,
Thomas Gleixner <tglx@linutronix.de>,
netdev@vger.kernel.org, edumazet@google.com,
"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 04/48] net: Replace __this_cpu_inc in route.c with raw_cpu_inc
Date: Fri, 14 Feb 2014 14:18:45 -0600 [thread overview]
Message-ID: <20140214201904.326059608@linux.com> (raw)
In-Reply-To: 20140214201841.826179349@linux.com
[-- Attachment #1: preempt_rt_cache_stat --]
[-- Type: text/plain, Size: 3608 bytes --]
[Patch depends on another patch in this series that introduces raw_cpu_ops]
The RT_CACHE_STAT_INC macro triggers the new preemption checks
for __this_cpu ops.
I do not see any other synchronization that would allow the use
of a __this_cpu operation here however in commit
dbd2915ce87e811165da0717f8e159276ebb803e Andrew justifies
the use of raw_smp_processor_id() here because "we do not care"
about races. In the past we agreed that the price of disabling
interrupts here to get consistent counters would be too high.
These counters may be inaccurate due to race conditions.
The use of __this_cpu op improves the situation already from what commit
dbd2915ce87e811165da0717f8e159276ebb803e did since the single instruction
emitted on x86 does not allow the race to occur anymore. However,
non x86 platforms could still experience a race here.
Trace:
[ 1277.189084] __this_cpu_add operation in preemptible [00000000] code: avahi-daemon/1193
[ 1277.189085] caller is __this_cpu_preempt_check+0x38/0x60
[ 1277.189086] CPU: 1 PID: 1193 Comm: avahi-daemon Tainted: GF 3.12.0-rc4+ #187
[ 1277.189087] Hardware name: FUJITSU CELSIUS W530 Power/D3227-A1, BIOS V4.6.5.4 R1.10.0 for D3227-A1x 09/16/2013
[ 1277.189088] 0000000000000001 ffff8807ef78fa00 ffffffff816d5a57 ffff8807ef78ffd8
[ 1277.189089] ffff8807ef78fa30 ffffffff8137359c ffff8807ef78fba0 ffff88079f822b40
[ 1277.189091] 0000000020000000 ffff8807ee32c800 ffff8807ef78fa70 ffffffff813735f8
[ 1277.189093] Call Trace:
[ 1277.189094] [<ffffffff816d5a57>] dump_stack+0x4e/0x82
[ 1277.189096] [<ffffffff8137359c>] check_preemption_disabled+0xec/0x110
[ 1277.189097] [<ffffffff813735f8>] __this_cpu_preempt_check+0x38/0x60
[ 1277.189098] [<ffffffff81610d65>] __ip_route_output_key+0x575/0x8c0
[ 1277.189100] [<ffffffff816110d7>] ip_route_output_flow+0x27/0x70
[ 1277.189101] [<ffffffff81616c80>] ? ip_copy_metadata+0x1a0/0x1a0
[ 1277.189102] [<ffffffff81640b15>] udp_sendmsg+0x825/0xa20
[ 1277.189104] [<ffffffff811b4aa9>] ? do_sys_poll+0x449/0x5d0
[ 1277.189105] [<ffffffff8164c695>] inet_sendmsg+0x85/0xc0
[ 1277.189106] [<ffffffff815c6e3c>] sock_sendmsg+0x9c/0xd0
[ 1277.189108] [<ffffffff813735f8>] ? __this_cpu_preempt_check+0x38/0x60
[ 1277.189109] [<ffffffff815c7550>] ? move_addr_to_kernel+0x40/0xa0
[ 1277.189111] [<ffffffff815c71ec>] ___sys_sendmsg+0x37c/0x390
[ 1277.189112] [<ffffffff8136613a>] ? string.isra.3+0x3a/0xd0
[ 1277.189113] [<ffffffff8136613a>] ? string.isra.3+0x3a/0xd0
[ 1277.189115] [<ffffffff81367b54>] ? vsnprintf+0x364/0x650
[ 1277.189116] [<ffffffff81367ee9>] ? snprintf+0x39/0x40
[ 1277.189118] [<ffffffff813735f8>] ? __this_cpu_preempt_check+0x38/0x60
[ 1277.189119] [<ffffffff815c7ff9>] __sys_sendmsg+0x49/0x90
[ 1277.189121] [<ffffffff815c8052>] SyS_sendmsg+0x12/0x20
[ 1277.189122] [<ffffffff816e4fd3>] tracesys+0xe1/0xe6
Cc: netdev@vger.kernel.org
Cc: edumazet@google.com
Acked-by: David S. Miller <davem@davemloft.net>
Acked-by: Ingo Molnar <mingo@kernel.org>
Signed-off-by: Christoph Lameter <cl@linux.com>
Index: linux/net/ipv4/route.c
===================================================================
--- linux.orig/net/ipv4/route.c 2014-01-30 14:40:44.276650882 -0600
+++ linux/net/ipv4/route.c 2014-01-30 14:40:44.276650882 -0600
@@ -194,7 +194,7 @@
EXPORT_SYMBOL(ip_tos2prio);
static DEFINE_PER_CPU(struct rt_cache_stat, rt_cache_stat);
-#define RT_CACHE_STAT_INC(field) __this_cpu_inc(rt_cache_stat.field)
+#define RT_CACHE_STAT_INC(field) raw_cpu_inc(rt_cache_stat.field)
#ifdef CONFIG_PROC_FS
static void *rt_cache_seq_start(struct seq_file *seq, loff_t *pos)
next prev parent reply other threads:[~2014-02-14 20:19 UTC|newest]
Thread overview: 81+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-02-14 20:18 [PATCH 00/48] percpu: Consistent per cpu operations V4 Christoph Lameter
2014-02-14 20:18 ` [PATCH 01/48] percpu: Add raw_cpu_ops Christoph Lameter
2014-02-14 20:18 ` [PATCH 02/48] mm: Use raw_cpu ops for determining current NUMA node Christoph Lameter
2014-02-14 20:18 ` [PATCH 03/48] modules: Use raw_cpu_write for initialization of per cpu refcount Christoph Lameter
2014-02-14 20:18 ` Christoph Lameter [this message]
2014-02-14 20:18 ` [PATCH 05/48] percpu: Add preemption checks to __this_cpu ops Christoph Lameter
2014-03-04 22:27 ` Andrew Morton
2014-03-04 23:27 ` Steven Rostedt
2014-03-05 3:27 ` Christoph Lameter
2014-03-05 21:34 ` Andrew Morton
2014-02-14 20:18 ` [PATCH 06/48] mm: Replace __get_cpu_var uses with this_cpu_ptr Christoph Lameter
2014-02-14 20:18 ` [PATCH 07/48] tracing: " Christoph Lameter
2014-02-14 20:18 ` [PATCH 08/48] percpu: Replace __get_cpu_var " Christoph Lameter
2014-02-14 20:18 ` [PATCH 09/48] kernel misc: Replace __get_cpu_var uses Christoph Lameter
2014-02-14 20:18 ` [PATCH 10/48] drivers/char/random: " Christoph Lameter
2014-02-14 20:18 ` [PATCH 11/48] drivers/cpuidle: Replace __get_cpu_var uses for address calculation Christoph Lameter
2014-02-14 20:18 ` [PATCH 12/48] drivers/oprofile: " Christoph Lameter
2014-02-14 20:18 ` [PATCH 13/48] drivers/leds: Replace __get_cpu_var use through this_cpu_ptr Christoph Lameter
2014-02-14 20:18 ` [PATCH 14/48] drivers/clocksource: Replace __get_cpu_var used for address calculation Christoph Lameter
2014-02-14 20:18 ` [PATCH 15/48] parisc: Replace __get_cpu_var uses " Christoph Lameter
2014-02-14 20:18 ` [PATCH 16/48] metag: " Christoph Lameter
2014-02-14 20:18 ` [PATCH 17/48] drivers/net/ethernet/tile: " Christoph Lameter
2014-02-14 20:18 ` [PATCH 18/48] drivers/net/ethernet/tile: __get_cpu_var call introduced in 3.14 Christoph Lameter
2014-02-14 20:19 ` [PATCH 19/48] tilegx: Another case of get_cpu_var Christoph Lameter
2014-02-14 20:19 ` [PATCH 20/48] time: Replace __get_cpu_var uses Christoph Lameter
2014-02-15 11:33 ` Thomas Gleixner
2014-02-14 20:19 ` [PATCH 21/48] scheduler: Replace __get_cpu_var with this_cpu_ptr Christoph Lameter
2014-02-14 20:19 ` [PATCH 22/48] tick-sched: Fix two new uses of __get_cpu_ptr Christoph Lameter
2014-02-15 11:33 ` Thomas Gleixner
2014-02-14 20:19 ` [PATCH 23/48] block: Replace __this_cpu_ptr with raw_cpu_ptr Christoph Lameter
2014-02-14 20:19 ` [PATCH 24/48] rcu: Replace __this_cpu_ptr uses " Christoph Lameter
2014-02-16 16:17 ` Paul E. McKenney
2014-02-14 20:19 ` [PATCH 25/48] watchdog: Replace __raw_get_cpu_var uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 26/48] net: Replace get_cpu_var through this_cpu_ptr Christoph Lameter
2014-02-14 20:19 ` [PATCH 27/48] md: Replace __this_cpu_ptr with raw_cpu_ptr Christoph Lameter
2014-02-14 20:19 ` [PATCH 28/48] irqchips: Replace __this_cpu_ptr uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 29/48] x86: Replace __get_cpu_var uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 30/48] x86: Change __get_cpu_var calls introduced in 3.14 Christoph Lameter
2014-02-14 20:19 ` [PATCH 31/48] uv: Replace __get_cpu_var Christoph Lameter
2014-03-04 23:02 ` Andrew Morton
2014-03-04 23:42 ` Steven Rostedt
2014-03-04 23:47 ` Andrew Morton
2014-03-05 0:18 ` H. Peter Anvin
2014-03-05 3:31 ` Christoph Lameter
2014-03-05 4:00 ` Andrew Morton
2014-03-05 15:35 ` Christoph Lameter
2014-03-05 21:57 ` Christoph Lameter
2014-03-06 2:53 ` Mike Travis
2014-03-07 18:16 ` Christoph Lameter
2014-02-14 20:19 ` [PATCH 32/48] arm: Replace __this_cpu_ptr with raw_cpu_ptr Christoph Lameter
2014-02-14 20:19 ` [PATCH 33/48] MIPS: Replace __get_cpu_var uses in FPU emulator Christoph Lameter
2014-02-14 20:19 ` [PATCH 34/48] mips: Replace __get_cpu_var uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 35/48] s390: rename __this_cpu_ptr to raw_cpu_ptr Christoph Lameter
2014-02-14 20:19 ` [PATCH 36/48] s390: Replace __get_cpu_var uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 37/48] s390: Handle new __get_cpu_var calls added in 3.14 Christoph Lameter
2014-02-14 20:19 ` [PATCH 38/48] ia64: Replace __get_cpu_var uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 39/48] powerpc: " Christoph Lameter
2014-02-15 3:50 ` Benjamin Herrenschmidt
2014-02-15 4:26 ` Steven Rostedt
2014-02-15 7:54 ` Mike Galbraith
2014-02-15 10:00 ` Benjamin Herrenschmidt
2014-02-15 11:29 ` Mike Galbraith
2014-02-15 9:59 ` Benjamin Herrenschmidt
2014-02-15 9:42 ` Peter Zijlstra
2014-02-15 10:01 ` Benjamin Herrenschmidt
2014-02-15 12:07 ` Andreas Schwab
2014-02-15 15:45 ` Peter Zijlstra
2014-02-15 18:12 ` David Woodhouse
2014-02-15 20:32 ` Benjamin Herrenschmidt
2014-02-15 20:52 ` Benjamin Herrenschmidt
2014-02-14 20:19 ` [PATCH 40/48] powerpc: Handle new __get_cpu_var calls in 3.14 Christoph Lameter
2014-02-14 20:19 ` [PATCH 41/48] sparc: Replace __get_cpu_var uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 42/48] tile: " Christoph Lameter
2014-02-14 20:19 ` [PATCH 43/48] blackfin: " Christoph Lameter
2014-02-14 20:19 ` [PATCH 44/48] avr32: Replace __get_cpu_var with __this_cpu_write Christoph Lameter
2014-02-14 20:19 ` [PATCH 45/48] alpha: Replace __get_cpu_var Christoph Lameter
2014-02-14 20:19 ` [PATCH 46/48] sh: Replace __get_cpu_var uses Christoph Lameter
2014-02-14 20:19 ` [PATCH 47/48] Remove __get_cpu_var and __raw_get_cpu_var macros [only in 3.16] Christoph Lameter
2014-02-14 20:19 ` [PATCH 48/48] percpu: Remove __this_cpu_ptr Christoph Lameter
2014-03-04 22:27 ` [PATCH 00/48] percpu: Consistent per cpu operations V4 Andrew Morton
2014-03-05 3:29 ` Christoph Lameter
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140214201904.326059608@linux.com \
--to=cl@linux.com \
--cc=akpm@linuxfoundation.org \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
--cc=tj@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox