linux-kernel - Re: [PATCH 1/3] x86/umip: Check that the instruction opcode is at least two bytes

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: <20250920100712.GBaM580B8WQipOrD2f@fat_crate.local>
Date: Sat, 20 Sep 2025 12:07:12 +0200
From: Borislav Petkov <bp@...en8.de>
To: Sean Christopherson <seanjc@...gle.com>
Cc: Masami Hiramatsu <mhiramat@...nel.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	Ingo Molnar <mingo@...hat.com>,
	Dave Hansen <dave.hansen@...ux.intel.com>, x86@...nel.org,
	linux-kernel@...r.kernel.org, Dan Snyder <dansnyder@...gle.com>
Subject: Re: [PATCH 1/3] x86/umip: Check that the instruction opcode is at
 least two bytes

On Fri, Sep 19, 2025 at 02:24:58PM -0700, Sean Christopherson wrote:
> The "weirdness" I am referring to is purely speculative; what I was trying to say
> is that I deliberate went with a "bad" check on nbytes, i.e. it really should be
> "insn->opcode.nbytes == 2".  But I didn't want to risk breaking some bizarre
> userspace that happened to be relying on a quirk of the kernel's decoder (I
> haven't dug into the decoder, so I genuinely have/had no idea what all could
> happen).

Yeah, after yesterday, my todo list has one more item - to dig into the
decoder and see what's going on there.

For example, in this particular case, the decoder should report an error when
it cannot decode the insn instead of emulating a totally different insn...

-- 
Regards/Gruss,
    Boris.

https://people.kernel.org/tglx/notes-about-netiquette