[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20140823163504.GA5333@linux.vnet.ibm.com>
Date: Sat, 23 Aug 2014 09:35:04 -0700
From: "Paul E. McKenney" <paulmck@...ux.vnet.ibm.com>
To: Fengguang Wu <fengguang.wu@...el.com>
Cc: Pranith Kumar <bobby.prani@...il.com>,
Dave Hansen <dave.hansen@...el.com>,
LKML <linux-kernel@...r.kernel.org>, lkp@...org
Subject: Re: [rcu] 0eb885afb20: -25.8% softirqs.RCU
On Thu, Aug 21, 2014 at 09:50:35PM +0800, Fengguang Wu wrote:
> Hi Pranith,
>
> FYI, these changes look nice:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git rcu/dev
> commit 0eb885afb20317016670c1c5dd3a436c91e1e910 ("rcu: Use rcu_gp_kthread_wake() to wake up grace period kthreads")
>
> test case: lkp-a04/aim7/400-disk_rr
Good show, Pranith!!!
Thanx, Paul
> It's an in-memory aim7/disk_rr test.
> lkp-a04 is an Atom micro server with 8G Memory.
>
> 5db6a289a3a0a69 0eb885afb20317016670c1c5d
> --------------- -------------------------
> 150860 ± 1% -25.8% 111937 ± 1% TOTAL softirqs.RCU
> 2809 ± 1% -12.5% 2457 ± 1% TOTAL vmstat.system.cs
> 4202 ± 1% -6.9% 3911 ± 1% TOTAL vmstat.system.in
> 212421 ± 2% -7.7% 196035 ± 3% TOTAL time.involuntary_context_switches
>
> vmstat.system.in
>
> 4300 ++-------------------------------------------------------------------+
> | .*.. *.. .*.. * |
> 4250 *+ .*.. + .*. * *.. .. : |
> 4200 ++ *..*..*. *.. + *..*. : .. * : |
> | * : .* : .* |
> 4150 ++ : .*..*. *. |
> 4100 ++ *. |
> | |
> 4050 ++ |
> 4000 ++ |
> | O O O O O |
> 3950 O+ O O O O O O O O O O |
> 3900 ++ O O O O |
> | O |
> 3850 ++-------------------------O-----------------------------------O-----O
>
>
> vmstat.system.cs
>
> 3000 ++-------------------------------------------------------------------+
> | |
> 2900 ++ *.. |
> *..*..*..*..*..*..*.. .. *..*..*..*..*.. .*.. .*.. |
> 2800 ++ * .*. *. .* |
> | *..*..*. *. |
> 2700 ++ |
> | |
> 2600 ++ |
> | O O |
> 2500 O+ O O O O O O O O O O O |
> | O O O O O |
> 2400 ++ O O O O
> | O |
> 2300 ++-------------------------------------------------------------------+
>
>
> [*] bisect-good sample
> [O] bisect-bad sample
>
>
> Disclaimer:
> Results have been estimated based on internal Intel analysis and are provided
> for informational purposes only. Any difference in system hardware or software
> design or configuration may affect actual performance.
>
> Thanks,
> Fengguang
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists