linux-kernel - Re: [PATCH 2/3] x86/vdso: Allow clock specific mult and shift values

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: <82a61daf-f6f9-8be0-2157-e9f9d7ba1cdf@arm.com>
Date:   Mon, 15 Apr 2019 11:15:56 +0100
From:   Vincenzo Frascino <vincenzo.frascino@....com>
To:     Thomas Gleixner <tglx@...utronix.de>,
        Huw Davies <huw@...eweavers.com>
Cc:     linux kernel <linux-kernel@...r.kernel.org>,
        Andy Lutomirski <luto@...nel.org>
Subject: Re: [PATCH 2/3] x86/vdso: Allow clock specific mult and shift values

Hi Thomas,

On 15/04/2019 10:51, Thomas Gleixner wrote:
> On Mon, 15 Apr 2019, Huw Davies wrote:
> 
>> On Sun, Apr 14, 2019 at 12:53:32PM +0200, Thomas Gleixner wrote:
>>> So it stays in the same cache line, but as we move the VDSO to generic
>>> code, the mask field needs to stay and this will make basetime[CLOCK_MONOTONIC] 
>>> overlap into the next cache line.
>>
>> Thanks for the great review; this is a good point.
>>
>>> See https://lkml.kernel.org/r/alpine.DEB.2.21.1902231727060.1666@nanos.tec.linutronix.de
>>> for an alternate solution to this problem, which avoids this and just gives
>>> CLOCK_MONOTONIC_RAW a separate storage space alltogether.
>>
>> I can certainly do this for the x86 vdso.  Would that be useful or
>> should I wait for Vincenzo's work on the generic vdso first?
> 
> Depends. If Vincenzo comes along with his new version soon, then you might
> get this for free :)
> 
> Vincenzo, what's the state of your work?
>

I am mostly done with the development, the only thing missing is the integration
of the generic update_vsyscall. After this is complete, I will need to do some
testing and extract the performance numbers.

Considering that I will be on Easter holiday from this Wednesday till the end of
April, I think v6 will be ready around second week of May.

> Thanks,
> 
> 	tglx
> 

-- 
Regards,
Vincenzo