linux-kernel - Re: [PATCH mm-unstable] mm/demotion: Assign correct memory type for multiple dax devices with the same node affinity

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] [day] [month] [year] [list]

Message-ID: <87y1v3zj81.fsf@yhuang6-desk2.ccr.corp.intel.com>
Date:   Thu, 01 Sep 2022 14:45:02 +0800
From:   "Huang, Ying" <ying.huang@...el.com>
To:     Aneesh Kumar K V <aneesh.kumar@...ux.ibm.com>
Cc:     linux-mm@...ck.org, akpm@...ux-foundation.org,
        Wei Xu <weixugc@...gle.com>, Yang Shi <shy828301@...il.com>,
        Davidlohr Bueso <dave@...olabs.net>,
        Tim C Chen <tim.c.chen@...el.com>,
        Michal Hocko <mhocko@...nel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Hesham Almatary <hesham.almatary@...wei.com>,
        Dave Hansen <dave.hansen@...el.com>,
        Jonathan Cameron <Jonathan.Cameron@...wei.com>,
        Alistair Popple <apopple@...dia.com>,
        Dan Williams <dan.j.williams@...el.com>,
        Johannes Weiner <hannes@...xchg.org>, jvgediya.oss@...il.com,
        Bharata B Rao <bharata@....com>
Subject: Re: [PATCH mm-unstable] mm/demotion: Assign correct memory type for
 multiple dax devices with the same node affinity

Aneesh Kumar K V <aneesh.kumar@...ux.ibm.com> writes:

> On 9/1/22 11:45 AM, Huang, Ying wrote:
>> "Aneesh Kumar K.V" <aneesh.kumar@...ux.ibm.com> writes:
>> 
>>> With multiple dax devices having the same node affinity, the kernel wrongly assigned
>>> default_dram memory type to some devices after the memory hotplug operation. Fix this by
>>> not clearing node_memory_types on the dax device remove.
>> 
>> Sorry for late reply.
>> 
>> Just for confirmation.  There are multiple dax devices in one NUMA node?
>> 
>> If you can show the bug reproducing steps, that will make it even easier
>> to understand.
>> 
>
> NUMA nodes are assigned per region and you can have multiple devdax namespace with same NUMA node affinity. 

I think that the patch description will be clearer if we added the above
line into it.

> dax0.1 and dax0.2 are examples. To recreate you can follow the below steps
>
>
> root@...ntu-guest:/sys/devices/system/node/node3# ls
> compact  cpumap    meminfo   memory34  memory36  memory38  memory41  memory43  memory45  memory47  memory50  memory52  memory54  numastat   uevent
> cpulist  distance  memory33  memory35  memory37  memory39  memory42  memory44  memory46  memory49  memory51  memory53  memory55  subsystem  vmstat
> root@...ntu-guest:/sys/devices/system/node/node3#
> root@...ntu-guest:/sys/devices/system/node/node3# for mem in memory*; do echo 0 > $mem/online; done
> root@...ntu-guest:/sys/devices/system/node/node3# cd /sys/bus/dax/drivers
> root@...ntu-guest:/sys/bus/dax/drivers# echo dax0.0  > kmem/unbind
> root@...ntu-guest:/sys/bus/dax/drivers# cd /sys/devices/system/node/node3/
> root@...ntu-guest:/sys/devices/system/node/node3# ls
> compact  cpumap    meminfo   memory42  memory44  memory46  memory49  memory51  memory53  memory55  subsystem  vmstat
> cpulist  distance  memory41  memory43  memory45  memory47  memory50  memory52  memory54  numastat  uevent
> root@...ntu-guest:/sys/devices/system/node/node3# for mem in memory*; do echo 1 > $mem/online; done

Thanks a lot for detailed information.

Best Regards,
Huang, Ying