linux-kernel - Re: [v6 08/15] mm: zero struct pages during initialization

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: <f81f5eaa-109e-666e-7020-84c090721a56@oracle.com>
Date:   Fri, 11 Aug 2017 12:03:38 -0400
From:   Pasha Tatashin <pasha.tatashin@...cle.com>
To:     Michal Hocko <mhocko@...nel.org>
Cc:     linux-kernel@...r.kernel.org, sparclinux@...r.kernel.org,
        linux-mm@...ck.org, linuxppc-dev@...ts.ozlabs.org,
        linux-s390@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
        x86@...nel.org, kasan-dev@...glegroups.com, borntraeger@...ibm.com,
        heiko.carstens@...ibm.com, davem@...emloft.net,
        willy@...radead.org, ard.biesheuvel@...aro.org,
        will.deacon@....com, catalin.marinas@....com, sam@...nborg.org
Subject: Re: [v6 08/15] mm: zero struct pages during initialization

> I believe this deserves much more detailed explanation why this is safe.
> What actually prevents any pfn walker from seeing an uninitialized
> struct page? Please make your assumptions explicit in the commit log so
> that we can check them independently.

There is nothing prevents pfn walkers from walk over any struct pages 
deferred and non-deferred. However, during boot before deferred pages 
are initialized we have just a few places that do that, and all of those 
cases are fixed in this patchset.

> Also this is done with some purpose which is the perfmance, right? You
> have mentioned that in the cover letter but if somebody is going to read
> through git logs this wouldn't be obvious from the specific commit.
> So add that information here as well. Especially numbers will be
> interesting.

I will add more performance data to this patch comment.