linux-kernel - RE: [PATCH -tip V2 00/10] workqueue: break affinity initiatively

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]

Date:   Wed, 23 Dec 2020 20:27:18 +0000
From:   Dexuan Cui <decui@...rosoft.com>
To:     Lai Jiangshan <jiangshanlai@...il.com>,
        Dexuan-Linux Cui <dexuan.linux@...il.com>
CC:     Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Valentin Schneider <valentin.schneider@....com>,
        Peter Zijlstra <peterz@...radead.org>,
        Qian Cai <cai@...hat.com>,
        Vincent Donnefort <vincent.donnefort@....com>,
        Lai Jiangshan <laijs@...ux.alibaba.com>,
        Hillf Danton <hdanton@...a.com>, Tejun Heo <tj@...nel.org>
Subject: RE: [PATCH -tip V2 00/10] workqueue: break affinity initiatively

> From: Lai Jiangshan <jiangshanlai@...il.com>
> Sent: Wednesday, December 23, 2020 7:02 AM
> >
> > Hi,
> > I tested this patchset on today's tip.git's master branch
> > (981316394e35 ("Merge branch 'locking/urgent'")).
> >
> > Every time the kernel boots with 32 CPUs (I'm running the Linux VM on
> > Hyper-V), I get the below warning.
> > (BTW, with 8 or 16 CPUs, I don't see the warning).
> > By printing the cpumasks with "%*pbl", I know the warning happens
> > because:
> > new_mask = 16-31
> > cpu_online_mask= 0-16
> > cpu_active_mask= 0-15
> > p->nr_cpus_allowed=16
> >
> 
> Hello, Dexuan
> 
> Could you omit patch4 of the patchset and test it again, please?
> ("workqueue: don't set the worker's cpumask when kthread_bind_mask()")
> 
> kthread_bind_mask() set the worker task to the pool's cpumask without
> any check. And set_cpus_allowed_ptr() finds that the task's cpumask
> is unchanged (already set by kthread_bind_mask()) and skips all the checks.
> 
> And I found that numa=fake=2U seems broken on cpumask_of_node() in my
> box.
> 
> Thanks,
> Lai

Looks like your analysis is correct: the warning can't repro if I configure all
the 32 vCPUs into 1 virtual NUMA node (and I don't see the message
"smpboot: CPU 16 Converting physical 0 to logical die 1"):

[    1.495440] smp: Bringing up secondary CPUs ...
[    1.499207] x86: Booting SMP configuration:
[    1.503038] .... node  #0, CPUs:        #1  #2  #3  #4  #5  #6  #7 
#8  #9 #10 #11 #12 #13 #14 #15 #16 #17 #18 #19 #20 #21 #22 #23 #24 #25 #26
#27 #28 #29 #30 #31
[    1.531930] smp: Brought up 1 node, 32 CPUs
[    1.538779] smpboot: Max logical packages: 1
[    1.539041] smpboot: Total of 32 processors activated (146859.90 BogoMIPS)

The warning only repros if there are more than 1 node, and it only prints once
for the first vCPU of the second node (i.e. node #1).

With more than 1 node, if I don't use patch4, the warning does not repro.

Thanks,
-- Dexuan