linux-kernel - Re: [PATCH 0/2] arm64/hotplug: Process MEM_OFFLINE and MEM_CANCEL

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, 15 Apr 2020 09:35:33 +0200
From:   David Hildenbrand <david@...hat.com>
To:     Anshuman Khandual <anshuman.khandual@....com>, linux-mm@...ck.org,
        linux-arm-kernel@...ts.infradead.org
Cc:     akpm@...ux-foundation.org, catalin.marinas@....com,
        mark.rutland@....com, Michal Hocko <mhocko@...e.com>,
        Dan Williams <dan.j.williams@...el.com>,
        Yu Zhao <yuzhao@...gle.com>,
        Hsin-Yi Wang <hsinyi@...omium.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Steve Capper <steve.capper@....com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 0/2] arm64/hotplug: Process MEM_OFFLINE and
 MEM_CANCEL_OFFLINE events

On 15.04.20 08:39, Anshuman Khandual wrote:
> This series improves arm64 memory event notifier (hot remove) robustness by
> enabling it to detect potential problems (if any) in the future. But first
> it enumerates memory isolation failure reasons that can be sent across a
> notifier. This series does not go beyond arm64 to explore if these failure
> reason codes could be used in other existing registered memory notifiers.
> Please do let me know if there is any other potential use cases, will be
> happy to incorporate next time around. Also should we add similar failure
> reasons for online_pages() as well ?
> 
> This series has been tested on arm64, boot tested on x86 and build tested
> on multiple other platforms.
> 

I'm sorry, but I have to nack this series. Why?

1. A hotplug notifier should not have to bother why offlining failed. He
received a MEM_GOING_OFFLINE, followed by a MEM_CANCEL_OFFLINE. That's
all he really has to know. Undo what you've done, end of story.

2. Patch 2 just introduces dead code that should never happen unless
something is horribly broken in the core (memory offlined although
nacked from notifier). And, it (for *whatever reason*) thinks it's okay
to bail out if another notifier canceled offlining hotplugged memory.

I fail to see the benefit for core changes and

 4 files changed, 99 insertions(+), 13 deletions(-)

-- 
Thanks,

David / dhildenb