linux-kernel - Re: [PATCH v2 3/4] arch/x86: Optionally flush L1D on context switch

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] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 08 Apr 2020 01:49:40 +0200
From:   Thomas Gleixner <tglx@...utronix.de>
To:     "Singh\, Balbir" <sblbir@...zon.com>,
        "keescook\@chromium.org" <keescook@...omium.org>
Cc:     "linux-kernel\@vger.kernel.org" <linux-kernel@...r.kernel.org>,
        "tony.luck\@intel.com" <tony.luck@...el.com>,
        "benh\@kernel.crashing.org" <benh@...nel.crashing.org>,
        "jpoimboe\@redhat.com" <jpoimboe@...hat.com>,
        "x86\@kernel.org" <x86@...nel.org>,
        "dave.hansen\@intel.com" <dave.hansen@...el.com>
Subject: Re: [PATCH v2 3/4] arch/x86: Optionally flush L1D on context switch

"Singh, Balbir" <sblbir@...zon.com> writes:
> On Tue, 2020-04-07 at 11:26 -0700, Kees Cook wrote:
>> On Mon, Apr 06, 2020 at 01:19:45PM +1000, Balbir Singh wrote:
>> > Add arch specific prctl()'s to opt-in to the L1D cache on context switch
>> > out, the existing mechanisms of tracking prev_mm via cpu_tlbstate is
>> > reused. cond_ibpb() is refactored and renamed into cond_mitigation().
>> 
>> I still think this should be a generic prctl(). If there is a strong
>> reason not to do this, can it be described in the commit log here?
>
> I can move to prctl() if that is what you prefer, the prctl() can then do arch
> specific things. I thought in my question around would other arch's like to do
> this, I did not hear anything specific, but I am happy to convert the
> interface over.

Yes, please. It's just a matter of time that other architectures find
this useful. L1D attacks are not restricted to x86 AFAICT.

Thanks,

        tglx