linux-kernel - Re: [PATCH] locking/paravirt: Fix performance regression on core bonded vCPU

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] [day] [month] [year] [list]

Message-ID: <1472097d-327f-ac46-acf5-76ef3571922a@suse.com>
Date:   Thu, 3 Nov 2022 15:09:08 +0100
From:   Juergen Gross <jgross@...e.com>
To:     johnnyaiai <arafatms@...look.com>
Cc:     bp@...en8.de, dave.hansen@...ux.intel.com, hpa@...or.com,
        johnnyaiai@...cent.com, linux-kernel@...r.kernel.org,
        mingo@...hat.com, tglx@...utronix.de,
        virtualization@...ts.linux-foundation.org
Subject: Re: [PATCH] locking/paravirt: Fix performance regression on core
 bonded vCPU

On 03.11.22 14:58, johnnyaiai wrote:
> Thanks for reply! I think nopvspin parameters controls
> pvspinlock or native spinlock. a vm guest always running
> on virtspin wheather nopvspin sets or not when EXITS_HALT
> not supported by hypervisor. So provide a missing parameter
> 'novirtspin'.

Your patch is doing:

     static_branch_disable(&virt_spin_lock_key);

in case your parameter has been specified.

When running as Xen guest the related coding is doing:

     static_branch_disable(&virt_spin_lock_key);

Only when running as KVM guest there is a difference, but this could be changed
by modifying kvm_spinlock_init().

Having two parameters with the same or very similar semantics is not a good
idea IMO.


Juergen

Download attachment "OpenPGP_0xB0DE9DD628BF132F.asc" of type "application/pgp-keys" (3099 bytes)

Download attachment "OpenPGP_signature" of type "application/pgp-signature" (496 bytes)