[llvm-dev] Status of Intel JCC Mitigations and Next Steps

Wed Mar 25 19:07:24 PDT 2020

I agree that there’re many assembly codes written over many years. They didn’t have consistent rules. So we have to be conservative not to break them. If there’re too many concerns about the prefix padding, maybe we have to give up. Anyway, we still have nop padding which works fine so far and gets most of the performance. It has less concern to break the assembler syntax or semantics. Meanwhile, there’re some heuristics to handle some known idioms to keep the code safe.

From: Eli Friedman <efriedma at quicinc.com>
Sent: Thursday, March 26, 2020 8:57 AM
To: Philip Reames <listmail at philipreames.com>; Eric Christopher <echristo at gmail.com>
Cc: Luo, Yuanke <yuanke.luo at intel.com>; llvm-dev <llvm-dev at lists.llvm.org>; Zhang, Annita <annita.zhang at intel.com>; Topper, Craig <craig.topper at intel.com>
Subject: RE: [llvm-dev] Status of Intel JCC Mitigations and Next Steps

> Why is one idiomatic construct supported, but another not?

I don’t think the line has to be at precisely “the given line describes an instruction, therefore we can insert a nop before it”.  It doesn’t make sense to say a rule is fundamentally unreasonable simply because it doesn’t agree with that.  Assembly has been written by many people over many years, and the rules haven’t always been the same.  Because of that, writing something a x86 CPU would consider a single instruction over multiple lines is a somewhat established practice.

I think it’s reasonable to distinguish between inserting padding between two instructions we understand, versus inserting padding between an instruction we understand and arbitrary data we don’t understand.

-Eli

From: Philip Reames <listmail at philipreames.com<mailto:listmail at philipreames.com>>
Sent: Wednesday, March 25, 2020 3:54 PM
To: Eli Friedman <efriedma at quicinc.com<mailto:efriedma at quicinc.com>>; Eric Christopher <echristo at gmail.com<mailto:echristo at gmail.com>>
Cc: Luo, Yuanke <yuanke.luo at intel.com<mailto:yuanke.luo at intel.com>>; llvm-dev <llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>>; Zhang, Annita <annita.zhang at intel.com<mailto:annita.zhang at intel.com>>; Craig Topper <craig.topper at intel.com<mailto:craig.topper at intel.com>>
Subject: [EXT] Re: [llvm-dev] Status of Intel JCC Mitigations and Next Steps

On 3/25/20 2:22 PM, Eli Friedman wrote:
I agree we shouldn’t try to guess what the user is trying to do.  There shouldn’t be an unbounded set of heuristic rules; “documented” implies some sort of promise of stability in addition to the actual text in the manual.  And we shouldn’t try to guess whether the user’s code cares about the length of a specific instruction.

I think you’re creating a false dichotomy here, though.  There’s some space between “the assembler can rewrite any instruction” and “the assembler should guess what the user is doing”.  The patches you’re referring to are dealing with a very specific issue: there are multiple ways users specify prefixes for certain instructions.  Sometimes the prefixes are integrated into the instruction.  Sometimes they’re written separately, like “rep; stosb”. And sometimes users end up using raw “.byte” directives.  So we add a rule: if we see something that looks like a prefix, followed by an instruction, we assume the combination might actually be a prefixed instruction, and don’t rewrite it.  This is a heuristic in some sense: we can’t prove whether the prefixed instruction will ever execute at runtime.  But adding prefixes to the instruction is likely problematic.

I see this as very much a slipper slope problem.  As you mention, I am trying to draw a hard line, but that's mostly because I can't figure out what another reasonable line to draw might be.  Why is one idiomatic construct supported, but another not?  Why is "label: nop; mov $(rax), %rax" (with the expectation a fault occurs at label + 1 byte) any less reasonable?  If we say that's reasonable, and should just work, well, we literally can't implement prefix padding at all.  Unless maybe there's something I'm missing here?

We could take a different tack on this: instead of giving up pure “strictness”, we could instead give the user a diagnostic on patterns we recognize.  But forcing users to rewrite “rep;stosb”->“rep stosb” doesn’t seem productive.
-Eli

From: Philip Reames <listmail at philipreames.com><mailto:listmail at philipreames.com>
Sent: Wednesday, March 25, 2020 1:34 PM
To: Eric Christopher <echristo at gmail.com><mailto:echristo at gmail.com>; Eli Friedman <efriedma at quicinc.com><mailto:efriedma at quicinc.com>
Cc: Luo, Yuanke <yuanke.luo at intel.com><mailto:yuanke.luo at intel.com>; llvm-dev <llvm-dev at lists.llvm.org><mailto:llvm-dev at lists.llvm.org>; Zhang, Annita <annita.zhang at intel.com><mailto:annita.zhang at intel.com>; Craig Topper <craig.topper at intel.com><mailto:craig.topper at intel.com>
Subject: [EXT] Re: [llvm-dev] Status of Intel JCC Mitigations and Next Steps

The slightly unexpected bit for me in these responses is the willingness to accept layout changes if documented.  Let me lay out some options, which of these seem reasonable?

1) Default to automatic padding, provide option to disable, document where padding is inserted.

2) Default to not automatically padding, provide option to enable, document where padding would be inserted if enabled.

And orthogonality to the above, two interpretations of each:

a) auto padding is allowed to break common idioms (because you enabled it explicitly or because you can disable).  If it does, you are expected to either simply not use it, or possibly use fine grained directives in source.

b) auto padding conservative even when explicitly enabled.  One implication of this scheme is that different versions of the assembler will almost by definition have to tweak insertion as we find new problematic idioms.  This creates both documentation difficulties, and user confusion.

So, four choices total.  Which seem reasonable?

Personally, I'd have no problem w/ 2a.  Any of the other variants concern me.

Philip

On 3/25/20 1:08 PM, Eric Christopher wrote:
FWIW I'm with Eli here if you need any more data points.

-eric

On Tue, Mar 24, 2020 at 8:21 PM Eli Friedman via llvm-dev <llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>> wrote:
Changing the length of a sequence of assembly instructions will break someone’s code at some point.  The length of a sequence of instructions is known, in general, and people will write code to take advantage of that. For example, I’ve seen assembly code using something like Duff’s device, except that instead of using a jump table, it just computed the destination as “base+n*caselength”.  Given that, I don’t think it’s reasonable to hide this mechanism from user control.

We definitely should not have any undocumented or unpredictable behavior in the assembler.  The actual instruction bytes matter.  That said, I’m not sure there’s a strong line between “automagic” and “explicit”, as long as the rules are documented.

-Eli

From: llvm-dev <llvm-dev-bounces at lists.llvm.org<mailto:llvm-dev-bounces at lists.llvm.org>> On Behalf Of Philip Reames via llvm-dev
Sent: Tuesday, March 24, 2020 3:55 PM
To: llvm-dev <llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>>
Cc: Luo, Yuanke <yuanke.luo at intel.com<mailto:yuanke.luo at intel.com>>; Zhang, Annita <annita.zhang at intel.com<mailto:annita.zhang at intel.com>>; Craig Topper <craig.topper at intel.com<mailto:craig.topper at intel.com>>
Subject: [EXT] [llvm-dev] Status of Intel JCC Mitigations and Next Steps

TLDR - We have a choice to make about assembler support, and a disagreement about how to move forward.  Community input needed.

Background

Intel has a hardware bug in Skylake and later whose mitigation requires padding of branches to avoid performance degradation.  Background here: https://www.intel.com/content/dam/support/us/en/documents/processors/mitigations-jump-conditional-code-erratum.pdf

We now have in tree support for alignment of such branches via nop padding, and limited support for padding existing instructions with either prefixes or larger immediate values.  This has survived several days of dedicated testing and appears to be reasonably robust.  The padding support applies both to branch alignment for the mitigation, but also normal align directives.

The original patches proposed a somewhat different approach than we've ended up taking - primarily because of memory overhead concerns.  However, there was also a deeper disagreement on the original review threads (D70157 and others) which was never settled, and we seem to be at a point where this needs attention.  In short, the question is how assembler support should be handled.

The Choice

The problematic use case comes when assembling user provided .s files.  (Instead of the more restricted output of the compiler.)  Our basic choice is do we want to force a new directive syntax (and thus a source code change to use the new feature), or attempt to automatically infer where it's safe to insert padding?

The options as I see them:

  *   Assembler directives w/explicit opt in - In this model, assembler input is assumed to only enable padding in regions where it is safe to do so.
  *   Automagic assembler - In this model, the assembler is responsible for inferring where it is legal to pad without breaking user expectations.

(I'll stop and disclaim that I'm strongly in favor of the former.  I've tried to describe the pros/cons of each, but my perspective is definitely biased.)

The difference between the two is a huge amount of complexity, and a very fundamental correctness risk.  The basic problem is that assemblers have to handle unconstrained inputs, and IMO, the semantics of assembler as used in practice is so under specified that it's really hard to infer semantics in any useful way.  As a couple of examples, is the fault behavior of an instruction well defined?  Is the label near an instruction used by the signal handler?  Is that data byte just before an instruction actually decoded as part of the instruction?

The benefit of the later option is that existing assembly files can be used without modification.  This is a huge advantage in terms of ease of mitigation for existing code bases.  It's also the approach the original patch sets for GCC took.

In the original review thread(s), I had taken the position that we should reject the automagic assembler based on the correctness concerns mentioned.  I had thought the consensus in the review was clearly in that direction as well, but this has recently come up again.  Given that, I wanted to open it to a wider audience.

Why am I pushing for a decision now?

There are two major reasons.  First, there have recently been a couple of patches posted and landed (D76176, and D76052) building towards the automagic assembler variant.  And second, I've started getting review comments (https://reviews.llvm.org/D76398#1930383) which block forward progress on generalized padding support assuming the automagic interpretation.  Implementing the automatic assembler variant for prefix and immediate padding adds substantial complexity and I would very much like not to bother with if I don't have to.

Current implementation details

We have support in the integrated assembler only for autopadding suppression.  This allows a LLVM based compiler to effectively apply padding selectively.  In particular, we've instrumented lowering from MI to MC (X86MCInstLowering.cpp) to selectively disable padding around constructs which are thought to be problematic.  We do not have an agreed upon syntax for this in assembler; the code that got checked in is modeled closely around the last seriously discussed variant (see below).  This support is able to use all of the padding variants: nop, prefix, and immediate.

We also have limited support in the assembler for not inserting nops between fragments where doing so would break known idioms.  The list of such idioms is, IMO, ad hoc.  This assembler support does not include prefix or immediate padding.

Philip

p.s. For those interested, here's roughly what the last round of assembler syntax I remember being discussed looked like.

.autopadding
.noautopadding

These two directives would respectively enable and disable automatic padding of instructions within the region defined.  It's presumed to be legal to insert nops between instructions, modify encodings, or otherwise adjust offsets of instruction boundaries within the region to achieve target specific desired alignments.  Similarly, it's presumed not to be legal to change relative offsets outside an explicitly enabled region.  (Except for existing cases - e.g. relaxation of branches, etc...)

The assembler would provide a command line flag which conceptually wrapped the whole file in a pair of enable/disable directives.

We'd previously discussed a variant with push/pop semantics and more fine grained control over alignment requests, but I believe we decided that was overkill in the end.  (I walked away with that impression based on the integrated assembler work at least.)

_______________________________________________
LLVM Developers mailing list
llvm-dev at lists.llvm.org<mailto:llvm-dev at lists.llvm.org>
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20200326/5d720ee5/attachment.html>