Files
archived-llvm/include/llvm/CodeGen/TargetSubtargetInfo.h
Chandler Carruth fd5a8723ce Introduce the "retpoline" x86 mitigation technique for variant #2 of the speculative execution vulnerabilities disclosed today, specifically identified by CVE-2017-5715, "Branch Target Injection", and is one of the two halves to Spectre..
Summary:
First, we need to explain the core of the vulnerability. Note that this
is a very incomplete description, please see the Project Zero blog post
for details:
https://googleprojectzero.blogspot.com/2018/01/reading-privileged-memory-with-side.html

The basis for branch target injection is to direct speculative execution
of the processor to some "gadget" of executable code by poisoning the
prediction of indirect branches with the address of that gadget. The
gadget in turn contains an operation that provides a side channel for
reading data. Most commonly, this will look like a load of secret data
followed by a branch on the loaded value and then a load of some
predictable cache line. The attacker then uses timing of the processors
cache to determine which direction the branch took *in the speculative
execution*, and in turn what one bit of the loaded value was. Due to the
nature of these timing side channels and the branch predictor on Intel
processors, this allows an attacker to leak data only accessible to
a privileged domain (like the kernel) back into an unprivileged domain.

The goal is simple: avoid generating code which contains an indirect
branch that could have its prediction poisoned by an attacker. In many
cases, the compiler can simply use directed conditional branches and
a small search tree. LLVM already has support for lowering switches in
this way and the first step of this patch is to disable jump-table
lowering of switches and introduce a pass to rewrite explicit indirectbr
sequences into a switch over integers.

However, there is no fully general alternative to indirect calls. We
introduce a new construct we call a "retpoline" to implement indirect
calls in a non-speculatable way. It can be thought of loosely as
a trampoline for indirect calls which uses the RET instruction on x86.
Further, we arrange for a specific call->ret sequence which ensures the
processor predicts the return to go to a controlled, known location. The
retpoline then "smashes" the return address pushed onto the stack by the
call with the desired target of the original indirect call. The result
is a predicted return to the next instruction after a call (which can be
used to trap speculative execution within an infinite loop) and an
actual indirect branch to an arbitrary address.

On 64-bit x86 ABIs, this is especially easily done in the compiler by
using a guaranteed scratch register to pass the target into this device.
For 32-bit ABIs there isn't a guaranteed scratch register and so several
different retpoline variants are introduced to use a scratch register if
one is available in the calling convention and to otherwise use direct
stack push/pop sequences to pass the target address.

This "retpoline" mitigation is fully described in the following blog
post: https://support.google.com/faqs/answer/7625886

We also support a target feature that disables emission of the retpoline
thunk by the compiler to allow for custom thunks if users want them.
These are particularly useful in environments like kernels that
routinely do hot-patching on boot and want to hot-patch their thunk to
different code sequences. They can write this custom thunk and use
`-mretpoline-external-thunk` *in addition* to `-mretpoline`. In this
case, on x86-64 thu thunk names must be:
```
  __llvm_external_retpoline_r11
```
or on 32-bit:
```
  __llvm_external_retpoline_eax
  __llvm_external_retpoline_ecx
  __llvm_external_retpoline_edx
  __llvm_external_retpoline_push
```
And the target of the retpoline is passed in the named register, or in
the case of the `push` suffix on the top of the stack via a `pushl`
instruction.

There is one other important source of indirect branches in x86 ELF
binaries: the PLT. These patches also include support for LLD to
generate PLT entries that perform a retpoline-style indirection.

The only other indirect branches remaining that we are aware of are from
precompiled runtimes (such as crt0.o and similar). The ones we have
found are not really attackable, and so we have not focused on them
here, but eventually these runtimes should also be replicated for
retpoline-ed configurations for completeness.

For kernels or other freestanding or fully static executables, the
compiler switch `-mretpoline` is sufficient to fully mitigate this
particular attack. For dynamic executables, you must compile *all*
libraries with `-mretpoline` and additionally link the dynamic
executable and all shared libraries with LLD and pass `-z retpolineplt`
(or use similar functionality from some other linker). We strongly
recommend also using `-z now` as non-lazy binding allows the
retpoline-mitigated PLT to be substantially smaller.

When manually apply similar transformations to `-mretpoline` to the
Linux kernel we observed very small performance hits to applications
running typical workloads, and relatively minor hits (approximately 2%)
even for extremely syscall-heavy applications. This is largely due to
the small number of indirect branches that occur in performance
sensitive paths of the kernel.

When using these patches on statically linked applications, especially
C++ applications, you should expect to see a much more dramatic
performance hit. For microbenchmarks that are switch, indirect-, or
virtual-call heavy we have seen overheads ranging from 10% to 50%.

However, real-world workloads exhibit substantially lower performance
impact. Notably, techniques such as PGO and ThinLTO dramatically reduce
the impact of hot indirect calls (by speculatively promoting them to
direct calls) and allow optimized search trees to be used to lower
switches. If you need to deploy these techniques in C++ applications, we
*strongly* recommend that you ensure all hot call targets are statically
linked (avoiding PLT indirection) and use both PGO and ThinLTO. Well
tuned servers using all of these techniques saw 5% - 10% overhead from
the use of retpoline.

We will add detailed documentation covering these components in
subsequent patches, but wanted to make the core functionality available
as soon as possible. Happy for more code review, but we'd really like to
get these patches landed and backported ASAP for obvious reasons. We're
planning to backport this to both 6.0 and 5.0 release streams and get
a 5.0 release with just this cherry picked ASAP for distros and vendors.

This patch is the work of a number of people over the past month: Eric, Reid,
Rui, and myself. I'm mailing it out as a single commit due to the time
sensitive nature of landing this and the need to backport it. Huge thanks to
everyone who helped out here, and everyone at Intel who helped out in
discussions about how to craft this. Also, credit goes to Paul Turner (at
Google, but not an LLVM contributor) for much of the underlying retpoline
design.

Reviewers: echristo, rnk, ruiu, craig.topper, DavidKreitzer

Subscribers: sanjoy, emaste, mcrosier, mgorny, mehdi_amini, hiraditya, llvm-commits

Differential Revision: https://reviews.llvm.org/D41723

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@323155 91177308-0d34-0410-b5e6-96231b3b80d8
2018-01-22 22:05:25 +00:00

262 lines
10 KiB
C++

//===- llvm/CodeGen/TargetSubtargetInfo.h - Target Information --*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
//
// This file describes the subtarget options of a Target machine.
//
//===----------------------------------------------------------------------===//
#ifndef LLVM_CODEGEN_TARGETSUBTARGETINFO_H
#define LLVM_CODEGEN_TARGETSUBTARGETINFO_H
#include "llvm/ADT/ArrayRef.h"
#include "llvm/ADT/SmallVector.h"
#include "llvm/ADT/StringRef.h"
#include "llvm/CodeGen/PBQPRAConstraint.h"
#include "llvm/CodeGen/ScheduleDAGMutation.h"
#include "llvm/CodeGen/SchedulerRegistry.h"
#include "llvm/MC/MCSubtargetInfo.h"
#include "llvm/Support/CodeGen.h"
#include <memory>
#include <vector>
namespace llvm {
class CallLowering;
class InstrItineraryData;
struct InstrStage;
class InstructionSelector;
class LegalizerInfo;
class MachineInstr;
struct MachineSchedPolicy;
struct MCReadAdvanceEntry;
struct MCWriteLatencyEntry;
struct MCWriteProcResEntry;
class RegisterBankInfo;
class SDep;
class SelectionDAGTargetInfo;
struct SubtargetFeatureKV;
struct SubtargetInfoKV;
class SUnit;
class TargetFrameLowering;
class TargetInstrInfo;
class TargetLowering;
class TargetRegisterClass;
class TargetRegisterInfo;
class TargetSchedModel;
class Triple;
//===----------------------------------------------------------------------===//
///
/// TargetSubtargetInfo - Generic base class for all target subtargets. All
/// Target-specific options that control code generation and printing should
/// be exposed through a TargetSubtargetInfo-derived class.
///
class TargetSubtargetInfo : public MCSubtargetInfo {
protected: // Can only create subclasses...
TargetSubtargetInfo(const Triple &TT, StringRef CPU, StringRef FS,
ArrayRef<SubtargetFeatureKV> PF,
ArrayRef<SubtargetFeatureKV> PD,
const SubtargetInfoKV *ProcSched,
const MCWriteProcResEntry *WPR,
const MCWriteLatencyEntry *WL,
const MCReadAdvanceEntry *RA, const InstrStage *IS,
const unsigned *OC, const unsigned *FP);
public:
// AntiDepBreakMode - Type of anti-dependence breaking that should
// be performed before post-RA scheduling.
using AntiDepBreakMode = enum { ANTIDEP_NONE, ANTIDEP_CRITICAL, ANTIDEP_ALL };
using RegClassVector = SmallVectorImpl<const TargetRegisterClass *>;
TargetSubtargetInfo() = delete;
TargetSubtargetInfo(const TargetSubtargetInfo &) = delete;
TargetSubtargetInfo &operator=(const TargetSubtargetInfo &) = delete;
~TargetSubtargetInfo() override;
virtual bool isXRaySupported() const { return false; }
// Interfaces to the major aspects of target machine information:
//
// -- Instruction opcode and operand information
// -- Pipelines and scheduling information
// -- Stack frame information
// -- Selection DAG lowering information
// -- Call lowering information
//
// N.B. These objects may change during compilation. It's not safe to cache
// them between functions.
virtual const TargetInstrInfo *getInstrInfo() const { return nullptr; }
virtual const TargetFrameLowering *getFrameLowering() const {
return nullptr;
}
virtual const TargetLowering *getTargetLowering() const { return nullptr; }
virtual const SelectionDAGTargetInfo *getSelectionDAGInfo() const {
return nullptr;
}
virtual const CallLowering *getCallLowering() const { return nullptr; }
// FIXME: This lets targets specialize the selector by subtarget (which lets
// us do things like a dedicated avx512 selector). However, we might want
// to also specialize selectors by MachineFunction, which would let us be
// aware of optsize/optnone and such.
virtual const InstructionSelector *getInstructionSelector() const {
return nullptr;
}
virtual unsigned getHwMode() const { return 0; }
/// Target can subclass this hook to select a different DAG scheduler.
virtual RegisterScheduler::FunctionPassCtor
getDAGScheduler(CodeGenOpt::Level) const {
return nullptr;
}
virtual const LegalizerInfo *getLegalizerInfo() const { return nullptr; }
/// getRegisterInfo - If register information is available, return it. If
/// not, return null.
virtual const TargetRegisterInfo *getRegisterInfo() const { return nullptr; }
/// If the information for the register banks is available, return it.
/// Otherwise return nullptr.
virtual const RegisterBankInfo *getRegBankInfo() const { return nullptr; }
/// getInstrItineraryData - Returns instruction itinerary data for the target
/// or specific subtarget.
virtual const InstrItineraryData *getInstrItineraryData() const {
return nullptr;
}
/// Resolve a SchedClass at runtime, where SchedClass identifies an
/// MCSchedClassDesc with the isVariant property. This may return the ID of
/// another variant SchedClass, but repeated invocation must quickly terminate
/// in a nonvariant SchedClass.
virtual unsigned resolveSchedClass(unsigned SchedClass,
const MachineInstr *MI,
const TargetSchedModel *SchedModel) const {
return 0;
}
/// \brief True if the subtarget should run MachineScheduler after aggressive
/// coalescing.
///
/// This currently replaces the SelectionDAG scheduler with the "source" order
/// scheduler (though see below for an option to turn this off and use the
/// TargetLowering preference). It does not yet disable the postRA scheduler.
virtual bool enableMachineScheduler() const;
/// \brief Support printing of [latency:throughput] comment in output .S file.
virtual bool supportPrintSchedInfo() const { return false; }
/// \brief True if the machine scheduler should disable the TLI preference
/// for preRA scheduling with the source level scheduler.
virtual bool enableMachineSchedDefaultSched() const { return true; }
/// \brief True if the subtarget should enable joining global copies.
///
/// By default this is enabled if the machine scheduler is enabled, but
/// can be overridden.
virtual bool enableJoinGlobalCopies() const;
/// True if the subtarget should run a scheduler after register allocation.
///
/// By default this queries the PostRAScheduling bit in the scheduling model
/// which is the preferred way to influence this.
virtual bool enablePostRAScheduler() const;
/// \brief True if the subtarget should run the atomic expansion pass.
virtual bool enableAtomicExpand() const;
/// True if the subtarget should run the indirectbr expansion pass.
virtual bool enableIndirectBrExpand() const;
/// \brief Override generic scheduling policy within a region.
///
/// This is a convenient way for targets that don't provide any custom
/// scheduling heuristics (no custom MachineSchedStrategy) to make
/// changes to the generic scheduling policy.
virtual void overrideSchedPolicy(MachineSchedPolicy &Policy,
unsigned NumRegionInstrs) const {}
// \brief Perform target specific adjustments to the latency of a schedule
// dependency.
virtual void adjustSchedDependency(SUnit *def, SUnit *use, SDep &dep) const {}
// For use with PostRAScheduling: get the anti-dependence breaking that should
// be performed before post-RA scheduling.
virtual AntiDepBreakMode getAntiDepBreakMode() const { return ANTIDEP_NONE; }
// For use with PostRAScheduling: in CriticalPathRCs, return any register
// classes that should only be considered for anti-dependence breaking if they
// are on the critical path.
virtual void getCriticalPathRCs(RegClassVector &CriticalPathRCs) const {
return CriticalPathRCs.clear();
}
// \brief Provide an ordered list of schedule DAG mutations for the post-RA
// scheduler.
virtual void getPostRAMutations(
std::vector<std::unique_ptr<ScheduleDAGMutation>> &Mutations) const {
}
// \brief Provide an ordered list of schedule DAG mutations for the machine
// pipeliner.
virtual void getSMSMutations(
std::vector<std::unique_ptr<ScheduleDAGMutation>> &Mutations) const {
}
// For use with PostRAScheduling: get the minimum optimization level needed
// to enable post-RA scheduling.
virtual CodeGenOpt::Level getOptLevelToEnablePostRAScheduler() const {
return CodeGenOpt::Default;
}
/// \brief True if the subtarget should run the local reassignment
/// heuristic of the register allocator.
/// This heuristic may be compile time intensive, \p OptLevel provides
/// a finer grain to tune the register allocator.
virtual bool enableRALocalReassignment(CodeGenOpt::Level OptLevel) const;
/// \brief True if the subtarget should consider the cost of local intervals
/// created by a split candidate when choosing the best split candidate. This
/// heuristic may be compile time intensive.
virtual bool enableAdvancedRASplitCost() const;
/// \brief Enable use of alias analysis during code generation (during MI
/// scheduling, DAGCombine, etc.).
virtual bool useAA() const;
/// \brief Enable the use of the early if conversion pass.
virtual bool enableEarlyIfConversion() const { return false; }
/// \brief Return PBQPConstraint(s) for the target.
///
/// Override to provide custom PBQP constraints.
virtual std::unique_ptr<PBQPRAConstraint> getCustomPBQPConstraints() const {
return nullptr;
}
/// Enable tracking of subregister liveness in register allocator.
/// Please use MachineRegisterInfo::subRegLivenessEnabled() instead where
/// possible.
virtual bool enableSubRegLiveness() const { return false; }
/// Returns string representation of scheduler comment
std::string getSchedInfoStr(const MachineInstr &MI) const override;
std::string getSchedInfoStr(MCInst const &MCI) const override;
/// This is called after a .mir file was loaded.
virtual void mirFileLoaded(MachineFunction &MF) const;
};
} // end namespace llvm
#endif // LLVM_CODEGEN_TARGETSUBTARGETINFO_H