linux-kernel - Re: [RFC v2 10/10] mm: Introduce CONFIG_MEM_RANGE

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20170531161919.GC28615@linux-80c1.suse>
Date:   Wed, 31 May 2017 09:19:19 -0700
From:   Davidlohr Bueso <dave@...olabs.net>
To:     Laurent Dufour <ldufour@...ux.vnet.ibm.com>
Cc:     linux-mm@...ck.org, akpm@...ux-foundation.org,
        Jan Kara <jack@...e.cz>,
        "Kirill A . Shutemov" <kirill@...temov.name>,
        Michal Hocko <mhocko@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Mel Gorman <mgorman@...hsingularity.net>,
        Andi Kleen <andi@...stfloor.org>, haren@...ux.vnet.ibm.com,
        aneesh.kumar@...ux.vnet.ibm.com, khandual@...ux.vnet.ibm.com,
        paulmck@...ux.vnet.ibm.com, linux-kernel@...r.kernel.org
Subject: Re: [RFC v2 10/10] mm: Introduce CONFIG_MEM_RANGE_LOCK

On Wed, 24 May 2017, Laurent Dufour wrote:

>A new configuration variable is introduced to activate the use of
>range lock instead of semaphore to protect per process memory layout.
>
>This range lock is replacing the use of a semaphore for mmap_sem.
>
>Currently only available for X86_64 and PPC64 architectures.
>
>By default this option is turned off and requires the EXPERT mode
>since it is not yet complete.

Just fyi I find this option quite useful for dev and debugging purposes,
however it should not exist once any of this is seriously considered
for merging. The reason being is that fundamentally such internals
should not be exposed to configuration options. We either get it right
for everybody, or we don't.

I'm currently running lots of tests to see the overhead in real workloads
on different boxes. While I hope that my artificial testing somewhat
resembles some patterns, this will be the real deal.

Thanks,
Davidlohr