linux-kernel - Re: [PATCH RFC v1 06/11] x86/fpu/mpx: Remove MPX xstate component support

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: <2ac2d1e7-d04b-443a-8fff-7aa3f436dcce@intel.com>
Date: Tue, 1 Apr 2025 10:13:58 -0700
From: "Chang S. Bae" <chang.seok.bae@...el.com>
To: Ingo Molnar <mingo@...nel.org>
CC: <linux-kernel@...r.kernel.org>, <x86@...nel.org>, <tglx@...utronix.de>,
	<mingo@...hat.com>, <bp@...en8.de>, <dave.hansen@...ux.intel.com>
Subject: Re: [PATCH RFC v1 06/11] x86/fpu/mpx: Remove MPX xstate component
 support

On 3/20/2025 4:47 PM, Chang S. Bae wrote:
> 
> In the meantime, kernel memory usage could be improved by treating MPX 
> as VCPU-only, aligning with the ongoing supervisor CET state handling 
> work [3].

Sorry, I made a naive assumption when mentioning this earlier.

Since MPX is a user feature, excluding it from the host fpstate while 
exposing it to the guest would also require enforcement in userspace. 
The only viable option seems to be switching XCR0. However, this would 
introduce significant kernel code complexity (as you’re aware) with a 
marginal benefit — saving just per-task 128 bytes on legacy systems. 
Unlike guest-exclusive supervisor states, this trade-off appears to be 
too costly.

Thanks,
Chang