linux-kernel - Re: [PATCH] x86/pkeys: Manually set X86_FEATURE

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: <fbf2b74e-bc37-985a-11f0-03e5ae5e390c@intel.com>
Date:   Thu, 27 Feb 2020 10:20:36 -0800
From:   Jacob Keller <jacob.e.keller@...el.com>
To:     Borislav Petkov <bp@...en8.de>
Cc:     Sean Christopherson <sean.j.christopherson@...el.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>, x86@...nel.org,
        "H. Peter Anvin" <hpa@...or.com>, linux-kernel@...r.kernel.org,
        Dave Hansen <dave.hansen@...ux.intel.com>
Subject: Re: [PATCH] x86/pkeys: Manually set X86_FEATURE_OSPKE to preserve
 existing changes



On 2/27/2020 10:01 AM, Borislav Petkov wrote:
> On Wed, Feb 26, 2020 at 04:16:13PM -0800, Jacob Keller wrote:
>> I tested this and it resolves my report! Thanks for a timely fix.
> 
> Adding your Tested-by.
> 

Yep, thanks.

>> I agree with the analysis. Perhaps it would make sense in the long term
>> to find a solution where get_cpu_cap can remember what was cleared for
>> each CPU and restore those? It already does this using the global
>> variables...
> 
> get_cpu_cap() remembers if you use setup_force_cpu_cap() but I agree
> that that whole feature bit handling is a bit, hm, "strange". It had its
> raisins.

Right. Nothing quite equivalent to that for per-CPU changes though.

> 
> FWIW, we had started cleaning those up but then the security nightmares
> happened so on the backburner it went...
> 
> Thx.
> 

Completely understandable. Especially since changes here are tricky to
get right, this being a case in point.

Thanks,
Jake