linux-kernel - Re: [PATCH v2] mm/hotplug: fix offline undo_isolate_page

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]

Date:   Thu, 14 Mar 2019 09:22:19 +0100
From:   Michal Hocko <mhocko@...nel.org>
To:     Oscar Salvador <osalvador@...e.de>
Cc:     Qian Cai <cai@....pw>, akpm@...ux-foundation.org, vbabka@...e.cz,
        linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] mm/hotplug: fix offline undo_isolate_page_range()

On Thu 14-03-19 09:09:22, Oscar Salvador wrote:
> On Wed, Mar 13, 2019 at 10:31:33AM -0400, Qian Cai wrote:
> > Also, after calling the "useless" undo_isolate_page_range() here, it
> > reaches the point of no returning by notifying MEM_OFFLINE. Those pages
> > will be marked as MIGRATE_MOVABLE again once onlining. The only thing
> > left to do is to decrease the number of isolated pageblocks zone
> > counter which would make some paths of the page allocation slower that
> > the above commit introduced. A memory block is usually at most 1GiB in
> > size, so an "int" should be enough to represent the number of pageblocks
> > in a block. Fix an incorrect comment along the way.
> 
> Well, x86_64's memblocks can be up to 2GB depending on the memory we got.
> Plus the fact that alloc_contig_range can be used to isolate 16GB-hugetlb
> pages on powerpc, and that could be a lot of pageblocks.
> 
> While an "int" could still hold, I think we should use "long" just to be
> more future-proof.

I would rather not touch that for now. If we really need to change the
type then do it when it is really needed. I pressume more things would
have to be changed anyway.

-- 
Michal Hocko
SUSE Labs