Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <d0032404-e627-2d51-9c7b-42f5cdf29929@infradead.org>
Date:   Thu, 23 Mar 2023 21:30:14 -0700
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.8.0
Subject: Re: [PATCH RFC] Documentation: Add document for false sharing
Content-Language: en-US
To:     Feng Tang <feng.tang@intel.com>
Cc:     Jonathan Corbet <corbet@lwn.net>, linux-doc@vger.kernel.org,
        linux-kernel@vger.kernel.org,
        Arnaldo Carvalho de Melo <acme@redhat.com>,
        Joe Mario <jmario@redhat.com>, Ingo Molnar <mingo@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Eric Dumazet <edumazet@google.com>,
        Shakeel Butt <shakeelb@google.com>, dave.hansen@intel.com,
        ying.huang@intel.com, tim.c.chen@intel.com, andi.kleen@intel.com
References: <20230323082626.100718-1-feng.tang@intel.com>
 <dcb0c139-e40e-f448-ad99-245020025862@infradead.org>
 <ZB0S/gfCLGP44HME@feng-clx>
From:   Randy Dunlap <rdunlap@infradead.org>
In-Reply-To: <ZB0S/gfCLGP44HME@feng-clx>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi--

On 3/23/23 20:03, Feng Tang wrote:
> Hi Randy,
> 
> Thank you for the thorough reviews!
> 
> On Thu, Mar 23, 2023 at 09:49:02AM -0700, Randy Dunlap wrote:
>> Hi,
>>
>> Lots of good/interesting info here.
>>
>> On 3/23/23 01:26, Feng Tang wrote:
>>> From: "Tang, Feng" <feng.tang@intel.com>


>>> +
>>> +'refcount' is modified frequently, but 'name' is set once at object
>>> +creation time and is never modified.  When many CPUs access 'foo' at
>>> +the same time, and 'refcount' is only bumped by one CPU frequently,
>>> +while 'name' is read by all other CPUs, which have to reload the whole
>>> +cache line over and over, even though the 'name' is never changed.
>>
>> That last "sentence" is not a sentence.
> 
> How about:
> 
> "
> When many CPUs access 'foo' at the same time, with 'refcount' being only
> bumped by one CPU frequently and 'name' being read by other CPUs, all
> those reading CPUs have to reload the whole cache line over and over
> due to the 'sharing', even though 'name' is never changed.
> "
> ?

Yes, that's good. Thanks.

-- 
~Randy