Re: [RFC 04/10] x86/mm: Only flush indirect branches when switching into non dumpable process

From: "David Woodhouse" <dwmw2@infradead.org>
To: "Peter Zijlstra" <peterz@infradead.org>, hjl.tools@gmail.com
Cc: "KarimAllah Ahmed" <karahmed@amazon.de>,
	linux-kernel@vger.kernel.org, "Andi Kleen" <ak@linux.intel.com>,
	"Andrea Arcangeli" <aarcange@redhat.com>,
	"Andy Lutomirski" <luto@kernel.org>,
	"Arjan van de Ven" <arjan@linux.intel.com>,
	"Ashok Raj" <ashok.raj@intel.com>,
	"Asit Mallick" <asit.k.mallick@intel.com>,
	"Borislav Petkov" <bp@suse.de>,
	"Dan Williams" <dan.j.williams@intel.com>,
	"Dave Hansen" <dave.hansen@intel.com>,
	"David Woodhouse" <dwmw@amazon.co.uk>,
	"Greg Kroah-Hartman" <gregkh@linuxfoundation.org>,
	"H . Peter Anvin" <hpa@zytor.com>,
	"Ingo Molnar" <mingo@redhat.com>,
	"Janakarajan Natarajan" <janakarajan.natarajan@amd.com>,
	"Joerg Roedel" <joro@8bytes.org>,
	"Jun Nakajima" <jun.nakajima@intel.com>,
	"Laura Abbott" <labbott@redhat.com>,
	"Linus Torvalds" <torvalds@linux-foundation.org>,
	"Masami Hiramatsu" <mhiramat@kernel.org>,
	"Paolo Bonzini" <pbonzini@redhat.com>,
	"\"Radim Krčmář\"" <rkrcmar@redhat.com>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Tim Chen" <tim.c.chen@linux.intel.com>,
	"Tom Lendacky" <thomas.lendacky@amd.com>,
	kvm@vger.kernel.org, x86@kernel.org
Subject: Re: [RFC 04/10] x86/mm: Only flush indirect branches when switching into non dumpable process
Date: Sun, 21 Jan 2018 12:04:03 -0000	[thread overview]
Message-ID: <477cc452f17665440978ae1e227861ca.squirrel@twosheds.infradead.org> (raw)
In-Reply-To: <20180121112224.GH2269@hirez.programming.kicks-ass.net>

> On Sat, Jan 20, 2018 at 08:22:55PM +0100, KarimAllah Ahmed wrote:
>> From: Tim Chen <tim.c.chen@linux.intel.com>
>>
>> Flush indirect branches when switching into a process that marked
>> itself non dumpable.  This protects high value processes like gpg
>> better, without having too high performance overhead.
>
> So if I understand it right, this is only needed if the 'other'
> executable itself is susceptible to spectre. If say someone audited gpg
> for spectre-v1 and build it with retpoline, it would be safe to not
> issue the IBPB, right?

Spectre V2 not v1. V1 is separate.
For V2 retpoline is enough... as long as all the libraries have it too.

> So would it make sense to provide an ELF flag / personality thing such
> that userspace can indicate its spectre-safe?

Yes, Arjan and I were pondering that yesterday; it probably does make
sense. Also for allowing a return to userspace after vmexit, if the army
process itself is so marked.

> I realize that this is all future work, because so far auditing for v1
> is a lot of pain (we need better tools), but would it be something that
> makes sense in the longer term?

It's *only* retpoline so it isn't actually that much. Although I'm wary of
Cc'ing HJ on such thoughts because he seems to never sleep and always
respond promptly with "OK I did that... " :)

If we did systematically do this in userspace we'd probably want to do
external thunks there too, and a flag in the auxvec to tell it not to
bother (for IBRS_ALL etc.).

-- 
dwmw2