Merge llvm, clang, compiler-rt, libc++, libunwind, lld, lldb and openmp
llvmorg-10.0.0-0-gd32170dbd5b (aka 10.0.0 release), and a number of
follow-ups.

MFC r356479 (by bdragon):

[PowerPC] Fix libllvmminimal build when building from powerpc64 ELFv1.

When bootstrapping on powerpc64 ELFv1, it is necessary to use binutils
ld.bfd from ports for the bootstrap, as this is the only modern linker for
ELFv1 host tools.

As binutils ld.bfd is rather strict in its handling of undefined symbols,
it is necessary to pull in Support/Atomic.cpp to avoid an undefined symbol.

Reviewed by: dim, emaste
Sponsored by: Tag1 Consulting, Inc.
Differential Revision: https://reviews.freebsd.org/D23072

MFC r356930:

Add more Subversion mergeinfo bootstrap information, to hopefully
increase the probability of merging in vendor changes.

MFC r358408 (by brooks):

Merge commit 7214f7a79 from llvm git (by Sam Elliott):

[RISCV] Lower llvm.trap and llvm.debugtrap

Summary:
Until this commit, these have lowered to a call to abort().

`llvm.trap()` now lowers to `unimp`, which should trap on all systems.

`llvm.debugtrap()` now lowers to `ebreak`, which is exactly what this
instruction is for.

Reviewers: asb, luismarques

Reviewed By: asb

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D69390

This fixes miscompilation resulting in linking failures with
INVARIANTS disabled.

Reviewed by: dim
Differential Revision: https://reviews.freebsd.org/D23857

MFC r358851:

Merge llvm, clang, compiler-rt, libc++, libunwind, lld, lldb and openmp
10.0.0-rc3 c290cb61fdc.

Release notes for llvm, clang, lld and libc++ 10.0.0 will become
available here:

https://releases.llvm.org/10.0.0/docs/ReleaseNotes.html
https://releases.llvm.org/10.0.0/tools/clang/docs/ReleaseNotes.html
https://releases.llvm.org/10.0.0/tools/lld/docs/ReleaseNotes.html
https://releases.llvm.org/10.0.0/projects/libcxx/docs/ReleaseNotes.html

PR: 244251

MFC r358854:

Add one additional file to libllvmminimal, to help the ppc64 bootstrap.

Reported by: bdragon
PR: 244251

MFC r358857:

Move another file in libllvm from sources required for world, to sources
required for bootstrap, as the PowerPC builds need this.

Reported by: bdragon
PR: 244251

MFC r359082:

Merge llvm, clang, compiler-rt, libc++, libunwind, lld, lldb and openmp
llvmorg-10.0.0-rc4-5-g52c365aa9ca. The actual release should follow Real
Soon Now.

PR: 244251

MFC r359084:

Merge commit 00925aadb from llvm git (by Fangrui Song):

[ELF][PPC32] Fix canonical PLTs when the order does not match the PLT order

Reviewed By: Bdragon28

Differential Revision: https://reviews.llvm.org/D75394

This is needed to fix miscompiled canonical PLTs on ppc32/lld10.

Requested by: bdragon
Differential Revision: https://reviews.freebsd.org/D24109

MFC r359085:

Merge commit 315f8a55f from llvm git (by Fangrui Song):

[ELF][PPC32] Don't report "relocation refers to a discarded section"
for .got2

Similar to D63182 [ELF][PPC64] Don't report "relocation refers to a
discarded section" for .toc

Reviewed By: Bdragon28

Differential Revision: https://reviews.llvm.org/D75419

This is needed to fix compile errors when building for ppc32/lld10.

Requested by: bdragon
Differential Revision: https://reviews.freebsd.org/D24110

MFC r359086:

Merge commit b8ebc11f0 from llvm git (by Sanjay Patel):

[EarlyCSE] avoid crashing when detecting min/max/abs patterns (PR41083)

As discussed in PR41083:
https://bugs.llvm.org/show_bug.cgi?id=41083
...we can assert/crash in EarlyCSE using the current hashing scheme
and instructions with flags.

ValueTracking's matchSelectPattern() may rely on overflow (nsw, etc)
or other flags when detecting patterns such as min/max/abs composed
of compare+select. But the value numbering / hashing mechanism used
by EarlyCSE intersects those flags to allow more CSE.

Several alternatives to solve this are discussed in the bug report.
This patch avoids the issue by doing simple matching of min/max/abs
patterns that never requires instruction flags. We give up some CSE
power because of that, but that is not expected to result in much
actual performance difference because InstCombine will canonicalize
these patterns when possible. It even has this comment for abs/nabs:

/// Canonicalize all these variants to 1 pattern.
/// This makes CSE more likely.

(And this patch adds PhaseOrdering tests to verify that the expected
transforms are still happening in the standard optimization
pipelines.

I left this code to use ValueTracking's "flavor" enum values, so we
don't have to change the callers' code. If we decide to go back to
using the ValueTracking call (by changing the hashing algorithm
instead), it should be obvious how to replace this chunk.

Differential Revision: https://reviews.llvm.org/D74285

This fixes an assertion when building the math/gsl port on PowerPC64.

Requested by: pkubja

MFC r359087:

Merge commit 585a3cc31 from llvm git (by me):

Fix -Wdeprecated-copy-dtor and -Wdeprecated-dynamic-exception-spec
warnings.

Summary:
The former are like:

libcxx/include/typeinfo:322:11: warning: definition of implicit copy
constructor for 'bad_cast' is deprecated because it has a
user-declared destructor [-Wdeprecated-copy-dtor]
virtual ~bad_cast() _NOEXCEPT;
^
libcxx/include/typeinfo:344:11: note: in implicit copy constructor
for 'std::bad_cast' first required here
throw bad_cast();
^

Fix these by adding an explicitly defaulted copy constructor.

The latter are like:

libcxx/include/codecvt:105:37: warning: dynamic exception
specifications are deprecated [-Wdeprecated-dynamic-exception-spec]
virtual int do_encoding() const throw();
^~~~~~~

Fix these by using the _NOEXCEPT macro instead.

Reviewers: EricWF, mclow.lists, ldionne, #libc

Reviewed By: EricWF, #libc

Subscribers: dexonsmith, libcxx-commits

Tags: #libc

Differential Revision: https://reviews.llvm.org/D76150

This is because we use -Wsystem-headers during buildworld, and the two
warnings above are now triggered by default with clang 10, preventing
most C++ code from compiling without NO_WERROR.

Requested by: brooks
Differential Revision: https://reviews.freebsd.org/D24049

MFC r359333:

Merge commit f0990e104 from llvm git (by Justin Hibbits):

[PowerPC]: e500 target can't use lwsync, use msync instead

The e500 core has a silicon bug that triggers an illegal instruction
program trap on any sync other than msync. Other cores will typically
ignore illegal sync types, and the documentation even implies that
the 'illegal' bits are ignored.

Address this hardware deficiency by only using msync, like the PPC440.

Differential Revision: https://reviews.llvm.org/D76614

Requested by: jhibbits

MFC r359334:

Merge commit 459e8e948 from llvm git (by Justin Hibbits):

[PowerPC]: Don't allow r0 as a target for LD_GOT_TPREL_L/32

Summary:
The linker is free to relax this (relocation R_PPC_GOT_TPREL16)
against R_PPC_TLS, if it sees fit (initial exec to local exec). If r0
is used, this can generate execution-invalid code (converts to 'addi
%rX, %r0, FOO, which translates in PPC-lingo to li %rX, FOO). Forbid
this instead.

This fixes static binaries using locales on FreeBSD/powerpc (tested
on FreeBSD/powerpcspe).

Reviewed By: nemanjai
Differential Revision: https://reviews.llvm.org/D76662

Requested by: jhibbits

MFC r359338:

Merge llvm, clang, compiler-rt, libc++, libunwind, lld, lldb and openmp
llvmorg-10.0.0-0-gd32170dbd5b (aka 10.0.0 release).

PR: 244251

MFC r359506 (by emaste):

lldb: stop excluding bindings/ subdir

With liblua in the tree we should be able to enable lldb's lua
scripting. We'll need the files in bindings/, so start by allowing them
to come in with the next import.

Approved by: dim
Sponsored by: The FreeBSD Foundation

MFC r359578:

Merge once more from ^/vendor/llvm-project/release-10.x, to get the
lldb/bindings directory, which will be used to provide lua bindings for
lldb.

Requested by: emaste

MFC r359826:

Merge commit 30588a739 from llvm git (by Erich Keane):

Make target features check work with ctor and dtor-

The problem was reported in PR45468, applying target features to an
always_inline constructor/destructor runs afoul of GlobalDecl
construction assert when checking for target-feature compatibility.

The core problem is fixed by using the version of the check that
takes a FunctionDecl rather than the GlobalDecl. However, while
writing the test, I discovered that source locations weren't properly
set for this check on ctors/dtors. This patch also fixes constructors
and CALLED destructors.

Unfortunately, it doesn't seem too possible to get a meaningful
source location for a 'cleanup' destructor, so those are still
'frontend' level errors unfortunately. A fixme was added to the test
to cover that situation.

This should fix 'Assertion failed: (!isa<CXXConstructorDecl>(D) && "Use
other ctor with ctor decls!"), function Init, file
/usr/src/contrib/llvm-project/clang/include/clang/AST/GlobalDecl.h, line
45' when compiling the security/botan2 port.

PR: 245550

MFC r359981:

Revert commit a9ad65a2b from llvm git (by Nemanja Ivanovic):

[PowerPC] Change default for unaligned FP access for older subtargets

This is a fix for https://bugs.llvm.org/show_bug.cgi?id=40554

Some CPU's trap to the kernel on unaligned floating point access and
there are kernels that do not handle the interrupt. The program then
fails with a SIGBUS according to the PR. This just switches the
default for unaligned access to only allow it on recent server CPUs
that are known to allow this.

Differential revision: https://reviews.llvm.org/D71954

This upstream commit causes a compiler hang when building certain ports
(e.g. security/nss, multimedia/x264) for powerpc64. The hang has been
reported in https://bugs.llvm.org/show_bug.cgi?id=45186, but in the mean
time it is more convenient to revert the commit.

Requested by: jhibbits

MFC r359994:

Revert commit b6cf400aa fro llvm git (by Nemanja Ivanovic):

Fix bots after a9ad65a2b34f

In the last commit, I neglected to initialize the new subtarget
feature I added which caused failures on a few bots. This should fix
that.

This unbreaks the build after r359981, which reverted upstream commit
a9ad65a2b34f.

Reported by: jhibbits (and jenkins :)

MFC r360129:

Merge commit ce5173c0e from llvm git (by Reid Kleckner):

Use FinishThunk to finish musttail thunks

FinishThunk, and the invariant of setting and then unsetting
CurCodeDecl, was added in 7f416cc42638 (2015). The invariant didn't
exist when I added this musttail codepath in ab2090d10765 (2014).
Recently in 28328c3771, I started using this codepath on non-Windows
platforms, and users reported problems during release testing
(PR44987).

The issue was already present for users of EH on i686-windows-msvc,
so I added a test for that case as well.

Reviewed By: hans

Differential Revision: https://reviews.llvm.org/D76444

This should fix 'Assertion failed: (!empty() && "popping exception stack
when not empty"), function popTerminate, file
/usr/src/contrib/llvm-project/clang/lib/CodeGen/CGCleanup.h, line 583'
when building the net-p2p/libtorrent-rasterbar

PR: 244830
Reported by: jbeich, yuri

MFC r360134:

Merge commit 64b31d96d from llvm git (by Nemanja Ivanovic):

[PowerPC] Do not attempt to reuse load for 64-bit FP_TO_UINT without
FPCVT

We call the function that attempts to reuse the conversion without
checking whether the target matches the constraints that the callee
expects. This patch adds the check prior to the call.

Fixes: https://bugs.llvm.org/show_bug.cgi?id=43976

Differential revision: https://reviews.llvm.org/D77564

This should fix 'Assertion failed: ((Op.getOpcode() == ISD::FP_TO_SINT
|| Subtarget.hasFPCVT()) && "i64 FP_TO_UINT is supported only with
FPCVT"), function LowerFP_TO_INTForReuse, file
/usr/src/contrib/llvm/lib/Target/PowerPC/PPCISelLowering.cpp, line 7276'
when building the devel/libslang2 port (and a few others) for PowerPC64.

Requested by: pkubaj

MFC r360350:

Tentatively apply https://reviews.llvm.org/D78877 (by Dave Green):

[ARM] Only produce qadd8b under hasV6Ops

When compiling for a arm5te cpu from clang, the +dsp attribute is
set. This meant we could try and generate qadd8 instructions where we
would end up having no pattern. I've changed the condition here to be
hasV6Ops && hasDSP, which is what other parts of ARMISelLowering seem
to use for similar instructions.

Fixed PR45677.

This fixes "fatal error: error in backend: Cannot select: t37: i32 =
ARMISD::QADD8b t43, t44" when compiling sys/dev/sound/pcm/feeder_mixer.c
for armv5. For some reason we do not encounter this on head, but this
error popped up while building universes for stable/12.

MFC r360697:

In r358396 I merged llvm upstream commit 2e24219d3, which fixed "error:
unsupported relocation on symbol" when assembling arm 'adr' pseudo
instructions. However, the upstream commit did not take big-endian arm
into account.

Applying the same changes to the big-endian handling is straightforward,
thanks to Andrew Turner and Peter Smith for the hint. This will also be
submitted upstream.

Merge clang, llvm, lld, lldb, compiler-rt and libc++ 6.0.0 release, and
several follow-up fixes.

MFC r327952:

Upgrade our copies of clang, llvm, lld, lldb, compiler-rt and libc++ to
6.0.0 (branches/release_60 r321788). Upstream has branched for the
6.0.0 release, which should be in about 6 weeks. Please report bugs and
regressions, so we can get them into the release.

Please note that from 3.5.0 onwards, clang, llvm and lldb require C++11
support to build; see UPDATING for more information.

MFC r328010:

Pull in r322473 from upstream llvm trunk (by Andrei Elovikov):

[LV] Don't call recordVectorLoopValueForInductionCast for
newly-created IV from a trunc.

Summary:
This method is supposed to be called for IVs that have casts in their
use-def chains that are completely ignored after vectorization under
PSE. However, for truncates of such IVs the same InductionDescriptor
is used during creation/widening of both original IV based on PHINode
and new IV based on TruncInst.

This leads to unintended second call to
recordVectorLoopValueForInductionCast with a VectorLoopVal set to the
newly created IV for a trunc and causes an assert due to attempt to
store new information for already existing entry in the map. This is
wrong and should not be done.

Fixes PR35773.

Reviewers: dorit, Ayal, mssimpso

Reviewed By: dorit

Subscribers: RKSimon, dim, dcaballe, hsaito, llvm-commits, hiraditya

Differential Revision: https://reviews.llvm.org/D41913

This should fix "Vector value already set for part" assertions when
building the net/iodine and sysutils/daa2iso ports.

Reported by: jbeich
PR: 224867, 224868

MFC r328090:

Pull in r322623 from upstream llvm trunk (by Andrew V. Tischenko):

Allow usage of X86-prefixes as separate instrs.
Differential Revision: https://reviews.llvm.org/D42102

This should fix parse errors when x86 prefixes (such as 'lock' and
'rep') are followed by various non-mnemonic tokens, e.g. comments, .byte
directives and labels.

PR: 224669, 225054

MFC r328091:

Revert r327340, as the workaround for rep prefixes followed by .byte
directives is no longer needed after r328090.

MFC r328141 (by emaste):

lld: Fix for ld.lld does not accept "AT" syntax for declaring LMA region

AT> lma_region expression allows to specify the memory region
for section load address.

Should fix [upstream LLVM] PR35684.

LLVM review: https://reviews.llvm.org/D41397

Obtained from: LLVM r322359 by George Rimar

MFC r328143 (by emaste):

lld: Handle parsing AT(ADDR(.foo-bar)).

The problem we had with it is that anything inside an AT is an
expression, so we failed to parse the section name because of the - in
it.

Requested by: royger
Obtained from: LLVM r322801 by Rafael Espindola

MFC r328144 (by emaste):

lld: Fix incorrect physical address on self-referencing AT command.

When a section placement (AT) command references the section itself,
the physical address of the section in the ELF header was calculated
incorrectly due to alignment happening right after the location
pointer's value was captured.

The problem was diagnosed and the first version of the patch written
by Erick Reyes.

Obtained from: LLVM r322421 by Rafael Espindola

MFC r328145:

Pull in r322016 from upstream llvm trunk (by Sanjay Patel):

[ValueTracking] remove overzealous assert

The test is derived from a failing fuzz test:
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=5008

Credit to @rksimon for pointing out the problem.

This should fix "Bad flavor while matching min/max" errors when building
the graphics/libsixel and science/kst2 ports.

Reported by: jbeich
PR: 225268, 225269

MFC r328146:

Pull in r322106 from upstream llvm trunk (by Alexey Bataev):

[COST]Fix PR35865: Fix cost model evaluation for shuffle on X86.

Summary:
If the vector type is transformed to non-vector single type, the
compile may crash trying to get vector information about non-vector
type.

Reviewers: RKSimon, spatel, mkuper, hfinkel

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D41862

This should fix "Not a vector MVT!" errors when building the
games/dhewm3 port.

Reported by: jbeich
PR: 225271

MFC r328286 (by emaste):

lld: Don't mark a shared library as needed because of a lazy symbol.

Obtained from: LLVM r323221 by Rafael Esp?ndola

MFC r328381:

Upgrade our copies of clang, llvm, lld, lldb, compiler-rt and libc++ to
6.0.0 (branches/release_60 r323338).

PR: 224669

MFC r328513:

Pull in r322245 from upstream clang trunk (by Craig Topper):

[X86] Make -mavx512f imply -mfma and -mf16c in the frontend like it
does in the backend.

Similarly, make -mno-fma and -mno-f16c imply -mno-avx512f.

Withou this "-mno-sse -mavx512f" ends up with avx512f being enabled
in the frontend but disabled in the backend.

Reported by: pawel
PR: 225488

MFC r328542 (by emaste):

lld: Use lookup instead of find. NFC, just simpler.

Obtained from: LLVM r323395 by Rafael Espindola

MFC r328543 (by emaste):

lld: Only lookup LMARegion once. NFC.

This is similar to how we handle MemRegion.

Obtained from: LLVM r323396 by Rafael Espindola

MFC r328544 (by emaste):

lld: Remove MemRegionOffset. NFC.

We can just use a member variable in MemoryRegion.

Obtained from: LLVM r323399 by Rafael Espindola

MFC r328545 (by emaste):

lld: Simplify. NFC.

Obtained from: LLVM r323440 by Rafael Espindola

MFC r328546 (by emaste):

lld: Improve LMARegion handling.

This fixes the crash reported at [LLVM] PR36083.

The issue is that we were trying to put all the sections in the same
PT_LOAD and crashing trying to write past the end of the file.

This also adds accounting for used space in LMARegion, without it all
3 PT_LOADs would have the same physical address.

Obtained from: LLVM r323449 by Rafael Espindola

MFC r328547 (by emaste):

lld: Move LMAOffset from the OutputSection to the PhdrEntry. NFC.

If two sections are in the same PT_LOAD, their relatives offsets,
virtual address and physical addresses are all the same.

[Rafael] initially wanted to have a single global LMAOffset, on the
assumption that every ELF file was in practiced loaded contiguously in
both physical and virtual memory.

Unfortunately that is not the case. The linux kernel has:

LOAD 0x200000 0xffffffff81000000 0x0000000001000000 0xced000 0xced000 R E 0x200000
LOAD 0x1000000 0xffffffff81e00000 0x0000000001e00000 0x15f000 0x15f000 RW 0x200000
LOAD 0x1200000 0x0000000000000000 0x0000000001f5f000 0x01b198 0x01b198 RW 0x200000
LOAD 0x137b000 0xffffffff81f7b000 0x0000000001f7b000 0x116000 0x1ec000 RWE 0x200000

The delta for all but the third PT_LOAD is the same:
0xffffffff80000000. [Rafael] thinks the 3rd one is a hack for implementing
per cpu data, but we can't break that.

Obtained from: LLVM r323456 by Rafael Espindola

MFC r328548 (by emaste):

lld: Put the header in the first PT_LOAD even if that PT_LOAD has a LMAExpr

The root problem is that we were creating a PT_LOAD just for the header.
That was technically valid, but inconvenient: we should not be making
the ELF discontinuous.

The solution is to allow a section with LMAExpr to be added to a PT_LOAD
if that PT_LOAD doesn't already have a LMAExpr.

LLVM PR: 36017
Obtained from: LLVM r323625 by Rafael Espindola

MFC r328594 (by emaste):

Pull in r322108 from upstream llvm trunk (by Rafael Esp?ndola):

Make one of the emitFill methods non virtual. NFC.

This is just preparatory work to fix [LLVM] PR35858.

MFC r328595 (by emaste):

Pull in r322123 from upstream llvm trunk (by Rafael Esp?ndola):

Don't create MCFillFragment directly.

Instead use higher level APIs that take care of most bookkeeping.

MFC r328596 (by emaste):

Pull in r322131 from upstream llvm trunk (by Rafael Esp?ndola):

Use a MCExpr for the size of MCFillFragment.

This allows the size to be found during ralaxation. This fixes
[LLVM] pr35858.

Requested by: royger

MFC r328753:

Upgrade our copies of clang, llvm, lld, lldb, compiler-rt and libc++ to
6.0.0 (branches/release_60 r323948).

PR: 224669

MFC r328817:

Upgrade our copies of clang, llvm, lld, lldb, compiler-rt and libc++ to
6.0.0 (branches/release_60 r324090).

This introduces retpoline support, with the -mretpoline flag. The
upstream initial commit message (r323155 by Chandler Carruth) contains
quite a bit of explanation. Quoting:

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 typic al 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

PR: 224669

MFC r329033:

Pull in r324594 from upstream clang trunk (by Alexander Ivchenko):

Fix for #31362 - ms_abi is implemented incorrectly for values >=16
bytes.

Summary:
This patch is a fix for following issue:
https://bugs.llvm.org/show_bug.cgi?id=31362 The problem was caused by
front end lowering C calling conventions without taking into account
calling conventions enforced by attribute. In this case win64cc was
no correctly lowered on targets other than Windows.

Reviewed By: rnk (Reid Kleckner)

Differential Revision: https://reviews.llvm.org/D43016

Author: belickim <mateusz.belicki@intel.com>

This fixes clang 6.0.0 assertions when building the emulators/wine and
emulators/wine-devel ports, and should also make it use the correct
Windows calling conventions. Bump __FreeBSD_version to make the fix
easy to detect.

PR: 224863

MFC r329223:

Pull in r323998 from upstream clang trunk (by Richard Smith):

PR36157: When injecting an implicit function declaration in C89, find
the right DeclContext rather than injecting it wherever we happen to
be.

This avoids creating functions whose DeclContext is a struct or
similar.

This fixes assertion failures when parsing certain not-completely-valid
struct declarations.

Reported by: ae
PR: 225862

MFC r329410:

Upgrade our copies of clang, llvm, lld, lldb, compiler-rt and libc++ to
6.0.0 (branches/release_60 r325330).

PR: 224669

MFC r329983:

Upgrade our copies of clang, llvm, lld, lldb, compiler-rt and libc++ to
6.0.0 (branches/release_60 r325932). This corresponds to 6.0.0 rc3.

PR: 224669

MFC r330384:

Upgrade our copies of clang, llvm, lld, lldb, compiler-rt and libc++ to
6.0.0 release (upstream r326565).

Release notes for llvm, clang and lld will be available here soon:
<http://releases.llvm.org/6.0.0/docs/ReleaseNotes.html>
<http://releases.llvm.org/6.0.0/tools/clang/docs/ReleaseNotes.html>
<http://releases.llvm.org/6.0.0/tools/lld/docs/ReleaseNotes.html>

Relnotes: yes
PR: 224669

MFC r330686:

Pull in r326882 from upstream llvm trunk (by Sjoerd Meijer):

[ARM] Fix for PR36577

Don't PerformSHLSimplify if the given node is used by a node that
also uses a constant because we may get stuck in an infinite combine
loop.

bugzilla: https://bugs.llvm.org/show_bug.cgi?id=36577

Patch by Sam Parker.

Differential Revision: https://reviews.llvm.org/D44097

This fixes a hang when compiling one particular file in java/openjdk8
for armv6 and armv7.

Reported by: swills
PR: 226388

MFC r331065:

Pull in r327638 from upstream llvm trunk (by Matthew Simpson):

[ConstantFolding, InstSimplify] Handle more vector GEPs

This patch addresses some additional cases where the compiler crashes
upon encountering vector GEPs. This should fix PR36116.

Differential Revision: https://reviews.llvm.org/D44219
Reference: https://bugs.llvm.org/show_bug.cgi?id=36116

This fixes an assertion when building the emulators/snes9x port.

Reported by: jbeich
PR: 225471

MFC r331066:

Pull in r321999 from upstream clang trunk (by Ivan A. Kosarev):

[CodeGen] Fix TBAA info for accesses to members of base classes

Resolves:
Bug 35724 - regression (r315984): fatal error: error in backend:
Broken function found (Did not see access type in access path!)
https://bugs.llvm.org/show_bug.cgi?id=35724

Differential Revision: https://reviews.llvm.org/D41547

This fixes "Did not see access type in access path" fatal errors when
building the devel/gdb port (version 8.1).

Reported by: jbeich
PR: 226658

MFC r331366:

Pull in r327101 from upstream llvm trunk (by Rafael Espindola):

Don't treat .symver as a regular alias definition.

This patch starts simplifying the handling of .symver.

For now it just moves the responsibility for creating an alias down to
the streamer. With that the asm streamer can pass a .symver unchanged,
which is nice since gas cannot parse "foo@bar = zed".

In a followup I hope to move the handling down to the writer so that
we don't need special hacks for avoiding breaking names with @@@ on
windows.

Pull in r327160 from upstream llvm trunk (by Rafael Espindola):

Delay creating an alias for @@@.

With this we only create an alias for @@@ once we know if it should
use @ or @@. This avoids last minutes renames and hacks to handle MS
names.

This only handles the ELF writer. LTO still has issues with @@@
aliases.

Pull in r327928 from upstream llvm trunk (by Vitaly Buka):

Object: Move attribute calculation into RecordStreamer. NFC

Summary: Preparation for D44274

Reviewers: pcc, espindola

Subscribers: hiraditya

Differential Revision: https://reviews.llvm.org/D44276

Pull in r327930 from upstream llvm trunk (by Vitaly Buka):

Object: Fix handling of @@@ in .symver directive

Summary:
name@@@nodename is going to be replaced with name@@nodename if symbols is
defined in the assembled file, or name@nodename if undefined.
https://sourceware.org/binutils/docs/as/Symver.html

Fixes PR36623

Reviewers: pcc, espindola

Subscribers: mehdi_amini, hiraditya

Differential Revision: https://reviews.llvm.org/D44274

Together, these changes fix handling of @@@ in .symver directives when
doing Link Time Optimization.

Reported by: Shawn Webb <shawn.webb@hardenedbsd.org>

MFC r331731:

Pull in r328738 from upstream lld trunk (by Rafael Espindola):

Strip @VER suffices from the LTO output.

This fixes pr36623.

The problem is that we have to parse versions out of names before LTO
so that LTO can use that information.

When we get the LTO produced .o files, we replace the previous symbols
with the LTO produced ones, but they still have @ in their names.

We could just trim the name directly, but calling parseSymbolVersion
to do it is simpler.

This is a follow-up to r331366, since we discovered that lld could
append version strings to symbols twice, when using Link Time
Optimization.