From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-fsdevel-owner@vger.kernel.org>
Received: from mx1.redhat.com ([209.132.183.28]:50270 "EHLO mx1.redhat.com"
        rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
        id S1726864AbfAGQTQ (ORCPT <rfc822;linux-fsdevel@vger.kernel.org>);
        Mon, 7 Jan 2019 11:19:16 -0500
Subject: Re: [PATCH 2/2] /proc/stat: Add sysctl parameter to control irq
 counts latency
To: Matthew Wilcox <willy@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>,
        Alexey Dobriyan <adobriyan@gmail.com>,
        Luis Chamberlain <mcgrof@kernel.org>,
        Kees Cook <keescook@chromium.org>,
        Jonathan Corbet <corbet@lwn.net>, linux-kernel@vger.kernel.org,
        linux-doc@vger.kernel.org, linux-fsdevel@vger.kernel.org,
        Davidlohr Bueso <dave@stgolabs.net>,
        Miklos Szeredi <miklos@szeredi.hu>,
        Daniel Colascione <dancol@google.com>,
        Dave Chinner <david@fromorbit.com>,
        Randy Dunlap <rdunlap@infradead.org>
References: <1546873978-27797-1-git-send-email-longman@redhat.com>
 <1546873978-27797-3-git-send-email-longman@redhat.com>
 <20190107155840.GY6310@bombadil.infradead.org>
 <c2ef7231-fe1f-9463-0b06-e88f990aa865@redhat.com>
 <20190107161442.GZ6310@bombadil.infradead.org>
From: Waiman Long <longman@redhat.com>
Message-ID: <f8eefa83-f85e-0ba8-dcb9-a76a12449228@redhat.com>
Date: Mon, 7 Jan 2019 11:19:14 -0500
MIME-Version: 1.0
In-Reply-To: <20190107161442.GZ6310@bombadil.infradead.org>
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit
Content-Language: en-US
Sender: linux-fsdevel-owner@vger.kernel.org
List-ID: <linux-fsdevel.vger.kernel.org>

On 01/07/2019 11:14 AM, Matthew Wilcox wrote:
> On Mon, Jan 07, 2019 at 11:07:47AM -0500, Waiman Long wrote:
>>> Why are you caching the _output_ of calling sprintf(), rather than caching the
>>> values of each interrupt?
>>>
>> It is just faster to dump the whole string buffer than redoing the
>> number formatting each time when the values don't change. I can cache
>> the individual sums instead if it is the preferred by most.
> But it also consumes more memory.  Can you gather some measurements to
> find out what the performance difference is if you cache the values
> instead of the strings?

I allocate 11 bytes (10 digits + space) for each unsigned int IRQ value.
I needs 4 bytes for the unsigned value itself. So it is a saving of 7
bytes. With 4k irqs, it will be 28k.

I will run some measurements of caching the values versus saving the
string later this week.

Thanks,
Longman