linux-kernel - Re: [LKP] Re: [perf/x86] 81ec3f3c4c: will-it-scale.per_process

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20200224003301.GA5061@shbuild999.sh.intel.com>
Date:   Mon, 24 Feb 2020 08:33:01 +0800
From:   Feng Tang <feng.tang@...el.com>
To:     Linus Torvalds <torvalds@...ux-foundation.org>
Cc:     Jiri Olsa <jolsa@...hat.com>,
        Peter Zijlstra <peterz@...radead.org>,
        kernel test robot <rong.a.chen@...el.com>,
        Ingo Molnar <mingo@...nel.org>,
        Vince Weaver <vincent.weaver@...ne.edu>,
        Jiri Olsa <jolsa@...nel.org>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Arnaldo Carvalho de Melo <acme@...nel.org>,
        Arnaldo Carvalho de Melo <acme@...hat.com>,
        "Naveen N. Rao" <naveen.n.rao@...ux.vnet.ibm.com>,
        Ravi Bangoria <ravi.bangoria@...ux.ibm.com>,
        Stephane Eranian <eranian@...gle.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        LKML <linux-kernel@...r.kernel.org>, lkp@...ts.01.org,
        andi.kleen@...el.com, "Huang, Ying" <ying.huang@...el.com>
Subject: Re: [LKP] Re: [perf/x86] 81ec3f3c4c: will-it-scale.per_process_ops
 -5.5% regression

Hi Linus,

On Sun, Feb 23, 2020 at 09:37:06AM -0800, Linus Torvalds wrote:
> On Sun, Feb 23, 2020 at 6:11 AM Feng Tang <feng.tang@...el.com> wrote:
> >
> > I tried to use perf-c2c on one platform (not the one that show
> > the 5.5% regression), and found the main "hitm" points to the
> > "root_user" global data, as there is a task for each CPU doing
> > the signal stress test, and both __sigqueue_alloc() and
> > __sigqueue_free() will call get_user() and free_uid() to inc/dec
> > this root_user's refcount.
> 
> What's around it for you?
> 
> There might be that 'uidhash_lock' spinlock right next to it, and
> maybe that exacerbates the issue?

The system map shows:

	ffffffff8225b520 d __syscall_meta__ptrace
	ffffffff8225b560 d args__ptrace
	ffffffff8225b580 d types__ptrace
	ffffffff8225b5c0 D root_user
	ffffffff8225b680 D init_user_ns
	ffffffff8225b880 d ratelimit_state.56624
	ffffffff8225b8c0 d event_exit__sigsuspend

I also searched the uidhack_lock, 
	ffffffff82b04c80 b uidhash_lock
 
> > Then I added some alignement inside struct "user_struct" (for
> > "root_user"), then the -5.5% is gone, with a +2.6% instead.
> 
> Do you actually need to align things inside the struct, or is it
> sufficient to just align the structure itself?

Initially I justed add the ____cacheline_aligned after the definition
of struct 'user_struct', which only makes the following 'init_user_ns'
aligned, and test result doesn't show much change.

Then I added
	
	 struct user_struct {
	 +       char dummy[0] ____cacheline_aligned;

to make 'root_user' itself aligned. 


> IOW, is the cache conflicts _within_ the user_struct itself, or is it
> with some nearby data (like that uidhash_lock or whatever?)

>From the perf c2c data, and the source code checking, the conflicts
only happens for root_user.__count, and root_user.sigpending, as
all running tasks are accessing this global data for get/put and
other operations.

> > One thing I don't understand is, this -5.5% only happens in
> > one 2 sockets, 96C/192T Cascadelake platform, as we've run
> > the same test on several different platforms. In therory,
> > the false sharing may also take effect?
> 
> Is that the biggest machine you have access to?
> 
> Maybe it just isn't noticeable with smaller core counts. A lot of
> conflict loads tend to have "exponential" behavior - when things get
> overloaded, performance plummets because it just makes things worse as
> everybody gets slower at that contention point and now it gets even
> more contended...

No, it's not the biggest, I tried another machine 'Xeon Phi(TM) CPU 7295',
which has 72C/288T, and the regression is not seen. This is the part
confusing me :)

Also I've tried to limit the concurrent task number from 192 to 96/48/24/6/1,
and the regression number did get smaller following the task decrease. 

Thanks,
Feng
 
>              Linus