linux-kernel - Re: [PATCH RFC v1 02/11] x86/fpu/xstate: Introduce xstate order table and accessor macro

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: <Z8HuMFQLex_cVn8j@gmail.com>
Date: Fri, 28 Feb 2025 18:11:12 +0100
From: Ingo Molnar <mingo@...nel.org>
To: "Chang S. Bae" <chang.seok.bae@...el.com>
Cc: Andrew Cooper <andrew.cooper3@...rix.com>, bp@...en8.de,
	dave.hansen@...el.com, dave.hansen@...ux.intel.com,
	linux-kernel@...r.kernel.org, mingo@...hat.com, tglx@...utronix.de,
	x86@...nel.org, Linus Torvalds <torvalds@...ux-foundation.org>
Subject: Re: [PATCH RFC v1 02/11] x86/fpu/xstate: Introduce xstate order
 table and accessor macro


* Chang S. Bae <chang.seok.bae@...el.com> wrote:

> > > I propose a new addition, an extension of functionality: if a new 
> > > CPUID bit indicates it, and a new MSR is written, XFEATURES bit 3 
> > > becomes active again - and the MPX area is now used by AVX. 
> > > Obviously only AVX-aware host and guest kernels would enable AVX.
> > 
> > Erm, s/AVX/APX ...
> 
> Just thought of another aspect of this:
> 
> I'm curious about how core dumps should handle this. Initially, an 
> xfeature mask was added to the software-reserved area [1] to indicate 
> which xfeatures were present in the layout. More recently, a new note 
> [2] was introduced to expose CPUID-reported size and offset 
> information, helping tools like GDB. From an offline interpretation 
> standpoint, I think these bits will become ambiguous without further 
> extensions.
> 
> [1] commit 5b3efd500854 ("x86, ptrace: regset extensions to support xstate")
> [2] commit ba386777a30b ("x86/elf: Add a new FPU buffer layout info to x86
> core files")

Okay, I guess I agree and you guys are right, the MPX/APX ambiguity is 
probably not worth it.

Still not happy about xfeature_noncompact_order[], but I guess that is 
the price if we want to reuse the MPX area. :-/

Thanks,

	Ingo