linux-kernel - Re: [PATCH] x86: Align TLB invalidation info

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-next>] [day] [month] [year] [list]

Date:   Wed, 31 Jan 2018 21:38:46 -0800
From:   Nadav Amit <nadav.amit@...il.com>
To:     Dave Hansen <dave.hansen@...ux.intel.com>
Cc:     the arch/x86 maintainers <x86@...nel.org>,
        Andy Lutomirski <luto@...nel.org>,
        "H. Peter Anvin" <hpa@...or.com>,
        Peter Zijlstra <peterz@...radead.org>,
        LKML <linux-kernel@...r.kernel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>
Subject: Re: [PATCH] x86: Align TLB invalidation info

Dave Hansen <dave.hansen@...ux.intel.com> wrote:

> On 01/31/2018 01:09 PM, Nadav Amit wrote:
>>> You also don't have to exhaustively test this, but I'd love to see at
>>> least a sanity check with a microbenchmark (or something) that, yes,
>>> this does help *something*.  Maybe it makes the remote
>>> flush_tlb_func_common() run faster because it's pulling in fewer lines,
>>> or maybe you can even detect fewer misses in there.
>> I agree that with the whole Meltdown/Spectre entry-cost it might not even be
>> measurable, at least on small ( < 2 sockets) machines. But I do not think it
>> worth profiling. Basically, AFAIK, all the data structures that are used for
>> inter-processor communication by the kernel are aligned, and this is an
>> exception.
> 
> I'm certainly not nak'ing this.  I think your patch is likely a good
> idea.  But, could you please take ten or twenty minutes to go see if
> practice matches your assumptions?  I'd really appreciate it.  If you
> can't measure it, then no biggie.

[CC’ing the mailing list]

Per your request, I measured it (which perhaps I should have done before). I
caused a misalignment intentionally by adding some padding to flush_tlb_info
and compared it with an aligned version.

I used ftrace to measure the execution time of flush_tlb_func_remote() on a
2-socket Haswell machine, using a microbenchmark I wrote for some research
project.

It turns out that your skepticism may be correct - In both cases the
function execution time is roughly 400ns (2% improvement on the aligned case
which is probably noise).

So it is up to you whether you want to discard the patch.

Regards,
Nadav