[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1e53da45-0892-42dc-b837-4b25640762d4@linux.dev>
Date: Thu, 15 Aug 2024 22:53:16 +0800
From: Wen Yang <wen.yang@...ux.dev>
To: Jens Axboe <axboe@...nel.dk>, Mateusz Guzik <mjguzik@...il.com>
Cc: Christian Brauner <brauner@...nel.org>, Jan Kara <jack@...e.cz>,
Alexander Viro <viro@...iv.linux.org.uk>, Dylan Yudaken <dylany@...com>,
David Woodhouse <dwmw@...zon.co.uk>, Paolo Bonzini <pbonzini@...hat.com>,
Dave Young <dyoung@...hat.com>, kernel test robot <lkp@...el.com>,
linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [RESEND PATCH v2] eventfd: introduce ratelimited wakeup for
non-semaphore eventfd
On 2024/8/15 00:50, Jens Axboe wrote:
> On 8/14/24 10:15 AM, Wen Yang wrote:
>>
>>
>> On 2024/8/11 18:26, Mateusz Guzik wrote:
>>> On Sun, Aug 11, 2024 at 04:59:54PM +0800, Wen Yang wrote:
>>>> For the NON-SEMAPHORE eventfd, a write (2) call adds the 8-byte integer
>>>> value provided in its buffer to the counter, while a read (2) returns the
>>>> 8-byte value containing the value and resetting the counter value to 0.
>>>> Therefore, the accumulated value of multiple writes can be retrieved by a
>>>> single read.
>>>>
>>>> However, the current situation is to immediately wake up the read thread
>>>> after writing the NON-SEMAPHORE eventfd, which increases unnecessary CPU
>>>> overhead. By introducing a configurable rate limiting mechanism in
>>>> eventfd_write, these unnecessary wake-up operations are reduced.
>>>>
>>>>
>>> [snip]
>>>
>>>> # ./a.out -p 2 -s 3
>>>> The original cpu usage is as follows:
>>>> 09:53:38 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
>>>> 09:53:40 PM 2 47.26 0.00 52.74 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>> 09:53:40 PM 3 44.72 0.00 55.28 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>>
>>>> 09:53:40 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
>>>> 09:53:42 PM 2 45.73 0.00 54.27 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>> 09:53:42 PM 3 46.00 0.00 54.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>>
>>>> 09:53:42 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
>>>> 09:53:44 PM 2 48.00 0.00 52.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>> 09:53:44 PM 3 45.50 0.00 54.50 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>>
>>>> Then enable the ratelimited wakeup, eg:
>>>> # ./a.out -p 2 -s 3 -r1000 -c2
>>>>
>>>> Observing a decrease of over 20% in CPU utilization (CPU # 3, 54% ->30%), as shown below:
>>>> 10:02:32 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
>>>> 10:02:34 PM 2 53.00 0.00 47.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>> 10:02:34 PM 3 30.81 0.00 30.81 0.00 0.00 0.00 0.00 0.00 0.00 38.38
>>>>
>>>> 10:02:34 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
>>>> 10:02:36 PM 2 48.50 0.00 51.50 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>> 10:02:36 PM 3 30.20 0.00 30.69 0.00 0.00 0.00 0.00 0.00 0.00 39.11
>>>>
>>>> 10:02:36 PM CPU %usr %nice %sys %iowait %irq %soft %steal %guest %gnice %idle
>>>> 10:02:38 PM 2 45.00 0.00 55.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
>>>> 10:02:38 PM 3 27.08 0.00 30.21 0.00 0.00 0.00 0.00 0.00 0.00 42.71
>>>>
>>>>
>>>
>>> Where are these stats from? Is this from your actual program you coded
>>> the feature for?
>>>
>>> The program you inlined here does next to nothing in userspace and
>>> unsurprisingly the entire thing is dominated by kernel time, regardless
>>> of what event rate can be achieved.
>>>
>>> For example I got: /a.out -p 2 -s 3 5.34s user 60.85s system 99% cpu 66.19s (1:06.19) total
>>>
>>> Even so, looking at perf top shows me that a significant chunk is
>>> contention stemming from calls to poll -- perhaps the overhead will
>>> sufficiently go down if you epoll instead?
>>
>> We have two threads here, one publishing and one subscribing, running
>> on CPUs 2 and 3 respectively. If we further refine and collect
>> performance data on CPU 2, we will find that a large amount of CPU is
>> consumed on the spin lock of the wake-up logic of event write, for
>> example:
>
> This is hardly surprising - you've got probably the worst kind of
> producer/consumer setup here, with the producer on one CPU, and the
> consumer on another. You force this relationship by pinning both of
> them. Then you have a queue in between, and locking that needs to be
> acquired on both sides.
>
Thank you for pointing it out.
We bind the CPU here to highlight this issue.
In fact, setting cpumask to -1 still remains the same:
./a.out -p -1 -s -1
9.27% [kernel] [k] _raw_spin_lock_irq
6.23% [kernel] [k] vfs_write
And another test program using libzmq also did not bind the CPU:
https://github.com/taskset/tests/blob/master/src/test.c
We can indeed solve this problem in user mode by using methods such as
shared memory, periodic data reading, atomic variables, etc. instead of
eventfd.
But since eventfd has already provided *NON-SEMAPHORE* , could you also
guide us to further utilize it and make it more comprehensive?
Especially linux is increasingly being used in automotive scenarios.
--
Best wishes,
Wen
Powered by blists - more mailing lists