Re: [RFC PATCH v2 4/4] x86/vdso: Add __vdso_sgx_enter_enclave() to wrap SGX enclave transitions

From: Andy Lutomirski <luto@kernel.org>
To: "Christopherson, Sean J" <sean.j.christopherson@intel.com>
Cc: Andrew Lutomirski <luto@kernel.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	Ingo Molnar <mingo@redhat.com>, Borislav Petkov <bp@alien8.de>,
	X86 ML <x86@kernel.org>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	Peter Zijlstra <peterz@infradead.org>,
	"H. Peter Anvin" <hpa@zytor.com>,
	LKML <linux-kernel@vger.kernel.org>,
	Jarkko Sakkinen <jarkko.sakkinen@linux.intel.com>,
	Josh Triplett <josh@joshtriplett.org>,
	linux-sgx@vger.kernel.org,
	Haitao Huang <haitao.huang@linux.intel.com>,
	Jethro Beekman <jethro@fortanix.com>,
	"Dr. Greg Wettstein" <greg@enjellic.com>
Subject: Re: [RFC PATCH v2 4/4] x86/vdso: Add __vdso_sgx_enter_enclave() to wrap SGX enclave transitions
Date: Tue, 11 Dec 2018 15:12:14 -0800	[thread overview]
Message-ID: <CALCETrWFHn0QrLmnUuxNC+rZ5LDdYmpVz8PSJk5Zmp+sWFM7zw@mail.gmail.com> (raw)
In-Reply-To: <20181211220010.GH14731@linux.intel.com>

On Tue, Dec 11, 2018 at 2:00 PM Sean Christopherson
<sean.j.christopherson@intel.com> wrote:
>
> On Tue, Dec 11, 2018 at 12:04:15PM -0800, Andy Lutomirski wrote:
> > On Tue, Dec 11, 2018 at 11:31 AM Sean Christopherson
> > <sean.j.christopherson@intel.com> wrote:
> > >
> > > On Fri, Dec 07, 2018 at 03:33:57PM -0800, Andy Lutomirski wrote:
> > > > On Fri, Dec 7, 2018 at 1:26 PM Sean Christopherson
> > > > <sean.j.christopherson@intel.com> wrote:
> > > > >
> > > > > Running a checksum on the stack for every exit doesn't seem like it'd
> > > > > be worth the effort, especially since this type of bug should be quite
> > > > > rare, at least in production environments.
> > > > >
> > > > > If we want to pursue the checksum idea I think the easiest approach
> > > > > would be to combine it with an exit_handler and do a simple check on
> > > > > the handler.  It'd be minimal overhead in the fast path and would flag
> > > > > cases where invoking exit_handle() would explode, while deferring all
> > > > > other checks to the user.
> > > >
> > > > How about this variant?
> > > >
> > > > #define MAGIC 0xaaaabbbbccccddddul
> > > > #define RETADDR_HASH ((unsigned long)__builtin_return_address(0) ^ MAGIC)
> > > >
> > > > void foo(void)
> > > > {
> > > >     volatile unsigned long hash = RETADDR_HASH;
> > > >
> > > >     /* placeholder for your actual code */
> > > >     asm volatile ("nop");
> > > >
> > > >     if (hash != RETADDR_HASH)
> > > >         asm volatile ("ud2");
> > > > }
> > > >
> > > > But I have a real argument for dropping exit_handler: in this new age
> > > > of Spectre, the indirect call is a retpoline, and it's therefore quite
> > > > slow.
> > >
> > > Technically slower, but would the extra CALL+RET pair even be noticeable
> > > in the grand scheme of SGX?
> >
> > But it's CALL, CALL, MOV to overwrite return address, intentionally
> > midpredicted RET, and RET because Spectre.  That whole sequence seems
> > to be several tens of cycles, so it's a lot worse than just CALL+RET.
> > Whether it's noticeable overall is a fair question, though.
>
> I was thinking of the case where the handler re-entered the enclave vs.
> leaving and re-calling the vDSO, which would be RET+CALL and some other
> stuff.

Fair enough, although the case where we do an EENTER, an AEP, a kernel
entry, an IRET, and an ERESUME will be so slow that the CALL+RET seems
even less relevant.  The EENTER+EEXIT case at least avoids the round
trip through x86's amazingly performant exception handling mechanism
:)