linux-kernel - Re: [PATCH v3 04/10] x86/fpu/xstate: Correct guest fpstate size calculation

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: <Z86PgkOXRfNFkoBX@intel.com>
Date: Mon, 10 Mar 2025 15:06:42 +0800
From: Chao Gao <chao.gao@...el.com>
To: "Chang S. Bae" <chang.seok.bae@...el.com>
CC: <tglx@...utronix.de>, <dave.hansen@...el.com>, <x86@...nel.org>,
	<seanjc@...gle.com>, <pbonzini@...hat.com>, <linux-kernel@...r.kernel.org>,
	<kvm@...r.kernel.org>, <peterz@...radead.org>, <rick.p.edgecombe@...el.com>,
	<weijiang.yang@...el.com>, <john.allen@....com>, <bp@...en8.de>
Subject: Re: [PATCH v3 04/10] x86/fpu/xstate: Correct guest fpstate size
 calculation

On Sun, Mar 09, 2025 at 10:21:12PM -0700, Chang S. Bae wrote:
>On 3/9/2025 6:33 PM, Chao Gao wrote:
>> 
>> This is fixed by the patch 3.
>
>Well, take a look at your changelog — the context is quite different. I don't
>think it'S mergeable without a rewrite. Also, this should be a standalone fix
>to complement the recent tip-tree changes.

Should patch 2 be posted separately?

Because current tip/master branch has:

	gfpu->xfeatures		= fpu_user_cfg.default_features;
	gfpu->perm		= fpu_user_cfg.default_features;

Adjusting only fpu_guest->features raises the question: why isn't gfpu->perm
adjusted as well?

If patch 2 should go first, I don't think it's necessary to post patches 2-3
separately as maintainers can easily pick up patches 1-3 when they are in good
shape.

Regarding the changelog, I am uncertain what's quite different in the context.
It seems both you and I are talking about the inconsistency between
gfpu->xfeatures and fpstate->xfeatures. Did I miss something obvious?