linux-kernel - Re: [PATCH] KVM: X86: Limit timer frequency with more smaller interval

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: <20180430123149.mroow2pdhglqesrn@starbug-mbp.local>
Date:   Mon, 30 Apr 2018 13:31:49 +0100
From:   Darren Kenny <darren.kenny@...cle.com>
To:     Wanpeng Li <kernellwp@...il.com>
Cc:     linux-kernel@...r.kernel.org, kvm@...r.kernel.org,
        Paolo Bonzini <pbonzini@...hat.com>,
        Radim Krčmář <rkrcmar@...hat.com>,
        Anthoine Bourgeois <anthoine.bourgeois@...de-group.com>
Subject: Re: [PATCH] KVM: X86: Limit timer frequency with more smaller
 interval

Hi Wanpeng Li,

On Sun, Apr 29, 2018 at 07:38:20PM -0700, Wanpeng Li wrote:
>From: Wanpeng Li <wanpengli@...cent.com>
>
>Anthoine reported:
> The period used by Windows change over time but it can be 1 milliseconds
> or less. I saw the limit_periodic_timer_frequency print so 500
> microseconds is sometimes reached.
>
>This patchs limits timer frequency with more smaller interval 200ms(5000Hz)
>to leave some headroom as Paolo suggested since Windows 10 changed the
>scheduler tick limit from 1024 Hz to 2048 Hz.

I would suggest re-writing this slightly, removing the 'this patch'
as some people suggested in other threads, and maybe some other
small re-wording, e.g.:

  As suggested by Paolo, lower the timer frequency limit to a
  smaller interval of 200 ms (5000 Hz) to leave some headroom. This
  is required due to Windows 10 changing the scheduler tick limit
  from 1024 Hz to 2048 Hz.

Also, in the subject line, maybe write it as 'Lower timer frequency
limit to 200ms'?

>
>Reported-by: Anthoine Bourgeois <anthoine.bourgeois@...de-group.com>
>Suggested-by: Paolo Bonzini <pbonzini@...hat.com>
>Cc: Paolo Bonzini <pbonzini@...hat.com>
>Cc: Radim Krčmář <rkrcmar@...hat.com>
>Cc: Anthoine Bourgeois <anthoine.bourgeois@...de-group.com>
>Signed-off-by: Wanpeng Li <wanpengli@...cent.com>

With those changes (or similar):

Reviewed-by: Darren Kenny <darren.kenny@...cle.com>

>---
> arch/x86/kvm/x86.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
>diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
>index 51ecd38..dc47073 100644
>--- a/arch/x86/kvm/x86.c
>+++ b/arch/x86/kvm/x86.c
>@@ -114,7 +114,7 @@ module_param(ignore_msrs, bool, S_IRUGO | S_IWUSR);
> static bool __read_mostly report_ignored_msrs = true;
> module_param(report_ignored_msrs, bool, S_IRUGO | S_IWUSR);
>
>-unsigned int min_timer_period_us = 500;
>+unsigned int min_timer_period_us = 200;
> module_param(min_timer_period_us, uint, S_IRUGO | S_IWUSR);
>
> static bool __read_mostly kvmclock_periodic_sync = true;
>-- 
>2.7.4
>