Path: news.eternal-september.org!eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: "Chris M. Thomasson" <chris.m.thomasson.1@gmail.com>
Newsgroups: comp.lang.c++
Subject: Re: Futexes ain't fast
Date: Thu, 29 Aug 2024 12:25:01 -0700
Organization: A noiseless patient Spider
Lines: 58
Message-ID: <vaqhud$2r8p$8@dont-email.me>
References: <van414$3fe9t$1@raubtier-asyl.eternal-september.org>
 <vanrkt$3j1oq$1@dont-email.me>
 <vaor7n$3r4u5$1@raubtier-asyl.eternal-september.org>
 <vaosod$3ra5a$2@dont-email.me> <vaosr5$3ra5a$3@dont-email.me>
 <vapkh5$3ukli$1@raubtier-asyl.eternal-september.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Thu, 29 Aug 2024 21:25:02 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="98c6004928f46df0abe90729ccf8af51";
	logging-data="93465"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/MpHDXfaQPwRdnt4FiFzkZ6oZM+jGxn+M="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:1fIhpSuxYrqiV0/Vtse9grROtuo=
In-Reply-To: <vapkh5$3ukli$1@raubtier-asyl.eternal-september.org>
Content-Language: en-US

On 8/29/2024 4:03 AM, Bonita Montero wrote:
> Am 29.08.2024 um 06:18 schrieb Chris M. Thomasson:
>> On 8/28/2024 9:17 PM, Chris M. Thomasson wrote:
>>> On 8/28/2024 8:51 PM, Bonita Montero wrote:
>>>> Am 28.08.2024 um 20:52 schrieb Chris M. Thomasson:
>>>>> On 8/28/2024 5:09 AM, Bonita Montero wrote:
>>>>>> I tested the operating-system specific mutex (CRITICAL_SECTION Or
>>>>>> pthread_mutext_t) against a futex and a std::mutex. I guessed 
>>>>>> std::mutex
>>>>>> uses th operating system specific mutex internally, but the times 
>>>>>> varied
>>>>>> so much across Windows and Linux that I gues that std::mutex used at
>>>>>> least a differently parametrized operating system mutex or maybe even
>>>>>> completely own code.
>>>>>> This are the times and each line has a further contender:
>>>>> [...]
>>>>>>              while( futex.exchange( true, memory_order_acquire ) )
>>>>>>                  futex.wait( true, memory_order_relaxed );
>>>>>>              futex.exchange( false, memory_order_release );
>>>>>>              futex.notify_one();
>>>>>>          } );
>>>>> [...]
>>>>>
>>>>> A wait bit would help out here... ;^) Afaict, this is a rather 
>>>>> "naive" use of futexes. In you use case here, the exchange to 
>>>>> unlock can be a simple atomic store with release semantics. Also, 
>>>>> try to think about calling notify_* only when you absolutely need 
>>>>> to... :^)
>>>>
>>>> Show me your code ...
>>>>
>>>
>>> I did a while back for a futex mutex, win32 impl iirc... Do I have to 
>>> find the older code?
>>
>> Lets see... It's a bit hard to find for some reason:
>>
>> https://groups.google.com/g/comp.lang.c++/c/1MZvhswJ6DQ/m/qyaYH-i0CgAJ
> 
> This includes your futex:
> 
>                os-mutex  C++-futex   CT-Futex std::mutex
>            1      4.458      5.288      2.932      6.213
[...]
>           23    1016.16    2828.65    905.289    2798.37
>           24    1096.64    2919.04    946.814    2914.23
>           25    1161.21    3052.88    996.146       3093
>           26    1256.53    3094.02    1034.94    3220.38
>           27    1379.79    3323.04    1079.41    3390.07
>           28    1446.68    3533.71    1106.27    3516.21
>           29    1514.12     3618.5       1160     3624.3
>           30    1601.25     3761.2    1218.03     3765.6
>           31    1673.59    3913.48    1259.65    3947.46
>           32    1827.81    4134.67    1319.16    4061.73

The futex mutex seems to scale a little better...? 1319.16 v 1827.81 for 
os mutex?