Message-ID: <570D3B3A.8080204@hpe.com>
Date: Tue, 12 Apr 2016 14:15:22 -0400
From: Waiman Long <waiman.long@hpe.com>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:10.0.12) Gecko/20130109 Thunderbird/10.0.12
MIME-Version: 1.0
To: Tejun Heo <tj@kernel.org>
CC: "Theodore Ts'o" <tytso@mit.edu>, Andreas Dilger <adilger.kernel@dilger.ca>,
        Christoph Lameter <cl@linux.com>, <linux-ext4@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>, Scott J Norton <scott.norton@hpe.com>,
        Douglas Hatch <doug.hatch@hpe.com>,
        Toshimitsu Kani <toshi.kani@hpe.com>
Subject: Re: [PATCH v2 2/4] percpu_stats: Enable 64-bit counts in 32-bit architectures
References: <1460132182-11690-1-git-send-email-Waiman.Long@hpe.com> <1460132182-11690-3-git-send-email-Waiman.Long@hpe.com> <20160408164747.GM24661@htj.duckdns.org> <5707EB44.9020703@hpe.com> <20160408174628.GN24661@htj.duckdns.org> <5707FC52.3070707@hpe.com> <20160411221700.GP24661@htj.duckdns.org>
In-Reply-To: <20160411221700.GP24661@htj.duckdns.org>
Content-Type: text/plain; charset="ISO-8859-1"; format=flowed
Content-Transfer-Encoding: 7bit
SpamDiagnosticOutput: 1:23
SpamDiagnosticMetadata: NSPM
X-MS-Exchange-CrossTenant-OriginalArrivalTime: 12 Apr 2016 18:15:27.4656 (UTC)
X-MS-Exchange-CrossTenant-FromEntityHeader: Hosted
X-MS-Exchange-Transport-CrossTenantHeadersStamped: AT5PR84MB0306
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org
Content-Length: 1028
Lines: 25

On 04/11/2016 06:17 PM, Tejun Heo wrote:
> Hello, Waiman.
>
> On Fri, Apr 08, 2016 at 02:45:38PM -0400, Waiman Long wrote:
>> The percpu_stats construct allows minimal overhead in maintaining statistics
>> counts. The percpu_counter construct, on the other hand, has a higher
>> performance overhead and a bit more complex to set up and tear down when
> If you're referring to the preemption on/off, as I wrote before, we'll
> probably be able to improve that with this_cpu_add_return so that the
> only extra overhead is an easily predictable branch which is extremely
> cheap.  It's better to improve common constructs anyway.
>
>> more than one statistics counts are needed. In fact, my first draft of the
> And yeah, it can be cumbersome to set up and tear down multiple
> percpu_counters.  If there are enough consumers, we can extend
> percpu_counter to handle multiple counters, right?
>
> Thanks.
>

I have updated my patch to use percpu_counter instead. Thanks for all 
the comments and review so far.

Cheers,
Longman