path: root/bootstrap/llvm16/exegesis-mcjit.patch

This is a backport from LLVM 17 that fixes a test failure observed on
ppc64:

********************
FAIL: LLVM :: tools/llvm-exegesis/PowerPC/latency-by-opcode-name.s (216 of 47688)
******************** TEST 'LLVM :: tools/llvm-exegesis/PowerPC/latency-by-opcode-name.s' FAILED ********************
PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace.
Stack dump:
0.	Program arguments: /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/build/bin/llvm-exegesis -mtriple=powerpc64le-unknown-linux-gnu -mcpu=pwr8 --benchmark-phase=assemble-measured-code -mode=latency -opcode-name=ADD8
 #0 0x0000000141a994d0 llvm::sys::RunSignalHandlers() /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/Support/Signals.cpp:103:22
 #1 0x0000000141a994d0 SignalHandler(int) /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/Support/Unix/Signals.inc:402:31
 #2 0x00003fff89552444  0x444 __register_frame_info_bases
 #3 0x00003fff89552444 
 #4 0x00003fff89552444 __register_frame_info (+0x444)
 #5 0xaaaaaaaaaaaaaaab 
 #6 0x00003fff8907b700 __register_frame (/usr/lib/libgcc_s.so.1+0x12700)
 #7 0x00003fff8907b760 llvm::RTDyldMemoryManager::registerEHFramesInProcess(unsigned char*, unsigned long) /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RTDyldMemoryManager.cpp:119:19
 #8 0x00003fff8907b760 llvm::RTDyldMemoryManager::registerEHFrames(unsigned char*, unsigned long, unsigned long) /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RTDyldMemoryManager.cpp:131:28
 #9 0x00003fff8907b7c0 llvm::RuntimeDyldELF::registerEHFrames() /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp:227:28
#10 0x0000000142e9d1b0 llvm::RuntimeDyld::registerEHFrames() /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RuntimeDyld.cpp:1435:27
#11 0x0000000142eb4258 llvm::MCJIT::finalizeLoadedModules() /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/MCJIT/MCJIT.cpp:251:24
#12 0x0000000142ea2154 __gthread_mutex_unlock /usr/include/c++/13.3.0/powerpc64-foxkit-linux-musl/bits/gthr-default.h:779:43
#13 0x0000000142ea2154 __gthread_recursive_mutex_unlock /usr/include/c++/13.3.0/powerpc64-foxkit-linux-musl/bits/gthr-default.h:832:33
#14 0x0000000142ea2154 std::recursive_mutex::unlock() /usr/include/c++/13.3.0/mutex:139:39
#15 0x0000000142ea2154 llvm::sys::SmartMutex<false>::unlock() /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/include/llvm/Support/Mutex.h:47:22
#16 0x0000000142ea2154 std::lock_guard<llvm::sys::SmartMutex<false>>::~lock_guard() /usr/include/c++/13.3.0/bits/std_mutex.h:255:25
#17 0x0000000142ea2154 llvm::MCJIT::getFunctionAddress(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char>> const&) /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/MCJIT/MCJIT.cpp:410:1
#18 0x0000000142b6f948 llvm::exegesis::ExecutableFunction::ExecutableFunction(std::unique_ptr<llvm::LLVMTargetMachine, std::default_delete<llvm::LLVMTargetMachine>>, llvm::object::OwningBinary<llvm::object::ObjectFile>&&) /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/tools/llvm-exegesis/lib/Assembler.cpp:319:66
#19 0x0000000142b73a18 llvm::exegesis::BenchmarkRunner::getRunnableConfiguration(llvm::exegesis::BenchmarkCode const&, unsigned int, unsigned int, llvm::exegesis::SnippetRepetitor const&) const /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/tools/llvm-exegesis/lib/BenchmarkRunner.cpp:181:62
#20 0x0000000141af099c llvm::exegesis::runBenchmarkConfigurations(llvm::exegesis::LLVMState const&, llvm::ArrayRef<llvm::exegesis::BenchmarkCode>, llvm::ArrayRef<std::unique_ptr<llvm::exegesis::SnippetRepetitor const, std::default_delete<llvm::exegesis::SnippetRepetitor const>>>, llvm::exegesis::BenchmarkRunner const&) /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/tools/llvm-exegesis/llvm-exegesis.cpp:376:58
#21 0x0000000141aca060 llvm::exegesis::benchmarkMain() /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/tools/llvm-exegesis/llvm-exegesis.cpp:506:30
#22 0x000000013fbe7144 main /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/tools/llvm-exegesis/llvm-exegesis.cpp:660:1
#23 0x000000013fbea0a8 libc_start_main_stage2 /usr/src/packages/system/musl/src/musl-1.2.3/src/env/__libc_start_main.c:95:2
#24 0x000000013fb105ac __libc_start_main /usr/src/packages/system/musl/src/musl-1.2.3/src/env/__libc_start_main.c:86:9
#25 0x00003fff89494038 _start_c /usr/src/packages/system/musl/src/musl-1.2.3/crt/crt1.c:18:2
#26 0x00003fff894940b4 _start (/lib/ld-musl-powerpc64.so.1+0x220b4)
Stack dump without symbol names (ensure you have llvm-symbolizer in your PATH or set the environment var `LLVM_SYMBOLIZER_PATH` to point to it):
0  llvm-exegesis          0x0000000141a994d0
1                         0x00003fff89552444 __kernel_sigtramp_rt64 + 0
2                         0xaaaaaaaaaaaaaaab __kernel_sigtramp_rt64 + 12297759015719765607
3  libgcc_s.so.1          0x00003fff8907b700 __register_frame_info_bases + 128
4  libgcc_s.so.1          0x00003fff8907b760 __register_frame_info + 32
5  libgcc_s.so.1          0x00003fff8907b7c0 __register_frame + 64
6  llvm-exegesis          0x0000000142e9d1b0
7  llvm-exegesis          0x0000000142eb4258
8  llvm-exegesis          0x0000000142ea2154
9  llvm-exegesis          0x0000000142b6f948
10 llvm-exegesis          0x0000000142b73a18
11 llvm-exegesis          0x0000000141af099c
12 llvm-exegesis          0x0000000141aca060
13 llvm-exegesis          0x000000013fbe7144
14 llvm-exegesis          0x000000013fbea0a8
15 llvm-exegesis          0x000000013fb105ac
16 ld-musl-powerpc64.so.1 0x00003fff89494038
17 ld-musl-powerpc64.so.1 0x00003fff894940b4 __libc_start_main + 100
18 llvm-exegesis          0x000000013fbdcf78
19 llvm-exegesis          0x000000013fbdcf20
FileCheck error: '<stdin>' is empty.
FileCheck command line:  /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/build/bin/FileCheck /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/test/tools/llvm-exegesis/PowerPC/latency-by-opcode-name.s

Tracing with a debug-build libgcc_s yields:

Program received signal SIGSEGV, Segmentation fault.
classify_object_over_fdes (ob=0x3ffff7904700, this_fde=0x400007ae4004, range=0x3fffffffd700) at /home/awilcox/Code/awilfox/gcc-next/system/gcc/src/gcc-13.3.0/libgcc/unwind-dw2-fde.c:732
732	  for (; ! last_fde (ob, this_fde); this_fde = next_fde (this_fde))
(gdb) bt
#0  classify_object_over_fdes (ob=0x3ffff7904700, this_fde=0x400007ae4004, range=0x3fffffffd700)
    at /home/awilcox/Code/awilfox/gcc-next/system/gcc/src/gcc-13.3.0/libgcc/unwind-dw2-fde.c:732
#1  0x00003ffff7b47700 in register_pc_range_for_object (ob=0x3ffff7904700, begin=<optimized out>)
    at /home/awilcox/Code/awilfox/gcc-next/system/gcc/src/gcc-13.3.0/libgcc/unwind-dw2-fde.c:118
#2  __register_frame_info_bases (dbase=<optimized out>, tbase=<optimized out>, ob=0x3ffff7904700, begin=<optimized out>)
    at /home/awilcox/Code/awilfox/gcc-next/system/gcc/src/gcc-13.3.0/libgcc/unwind-dw2-fde.c:144
#3  __register_frame_info_bases (begin=<optimized out>, ob=0x3ffff7904700, tbase=<optimized out>, dbase=<optimized out>)
    at /home/awilcox/Code/awilfox/gcc-next/system/gcc/src/gcc-13.3.0/libgcc/unwind-dw2-fde.c:126
#4  0x00003ffff7b47760 in __register_frame_info (begin=<optimized out>, ob=<optimized out>)
    at /home/awilcox/Code/awilfox/gcc-next/system/gcc/src/gcc-13.3.0/libgcc/unwind-dw2-fde.c:159
#5  0x00003ffff7b477c0 in __register_frame (begin=0x3ffff7ae4000)
    at /home/awilcox/Code/awilfox/gcc-next/system/gcc/src/gcc-13.3.0/libgcc/unwind-dw2-fde.c:172
#6  0x00000001038da4f0 in llvm::RTDyldMemoryManager::registerEHFramesInProcess (Size=<optimized out>, Addr=<optimized out>)
    at /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RTDyldMemoryManager.cpp:119
#7  llvm::RTDyldMemoryManager::registerEHFrames (this=0x3ffff78b1700, Addr=<optimized out>, LoadAddr=<optimized out>,
    Size=<optimized out>)
    at /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RTDyldMemoryManager.cpp:131
#8  0x00000001038f1598 in llvm::RuntimeDyldELF::registerEHFrames (this=0x3ffff7b0e550)
    at /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp:227
#9  0x00000001038df494 in llvm::RuntimeDyld::registerEHFrames (this=<optimized out>)
    at /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/RuntimeDyld/RuntimeDyld.cpp:1435
#10 0x00000001035acc88 in llvm::MCJIT::finalizeLoadedModules (this=0x3ffff7aef0b0)
    at /home/awilcox/Code/awilfox/rust-next/bootstrap/llvm16/src/llvm-project-16.0.6.src/llvm/lib/ExecutionEngine/MCJIT/MCJIT.cpp:251

The reason is because it is trying to interpret a little-endian FDE as a
native (big-endian) one and failing.  This was noted upstream and fixed
in the LLVM 17 branch but it was never backported to a 16 release.

From 27f37db76a79e746561ad2545ff6ccf0d50cc38c Mon Sep 17 00:00:00 2001
From: Pavel Kosov <kpdev42@gmail.com>
Date: Thu, 8 Jun 2023 19:14:00 +0300
Subject: [PATCH] [llvm-exegesis] Use MCJIT only for execution

Initially, llvm-exegesis was generating the benchmark code for the
host CPU to execute it inside its own process. Thus, MCJIT was reused
for fetching function's bytes to fill the assembled_snippet field in
the benchmark report.

Later, the --mtriple and --benchmark-phase command line options were
introduced that are handy for testing snippet generation even if
snippet execution is not possible. In that setup, MCJIT is asked to
parse an object file for a foreign CPU or operating system that is
probably not guaranteed to succeed and was actually observed to fail
in https://reviews.llvm.org/D145763.

This commit implements a much simplified function's code fetching,
assuming the benchmark function is the only function in the object file
and it spans across the entire text section (note that MCJIT-based code
has more or less the same assumption - see TrackingSectionMemoryManager
class).

~~~

Huawei RRI, OS Lab

Reviewed By: courbet

Differential Revision: https://reviews.llvm.org/D148921
---
 .../AArch64/print-assembled-snippet.s         | 31 +++++++++++++++++
 .../Mips/print-assembled-snippet.s            | 23 +++++++++++++
 .../PowerPC/print-assembled-snippet.s         | 15 ++++++++
 .../X86/print-assembled-snippet.s             | 32 +++++++++++++++++
 llvm/tools/llvm-exegesis/lib/Assembler.cpp    | 34 +++++++++++++++++++
 llvm/tools/llvm-exegesis/lib/Assembler.h      |  4 +++
 .../llvm-exegesis/lib/BenchmarkRunner.cpp     |  8 ++---
 7 files changed, 143 insertions(+), 4 deletions(-)
 create mode 100644 llvm/test/tools/llvm-exegesis/AArch64/print-assembled-snippet.s
 create mode 100644 llvm/test/tools/llvm-exegesis/Mips/print-assembled-snippet.s
 create mode 100644 llvm/test/tools/llvm-exegesis/PowerPC/print-assembled-snippet.s
 create mode 100644 llvm/test/tools/llvm-exegesis/X86/print-assembled-snippet.s

diff --git a/llvm/test/tools/llvm-exegesis/AArch64/print-assembled-snippet.s b/llvm/test/tools/llvm-exegesis/AArch64/print-assembled-snippet.s
new file mode 100644
index 0000000000000..e8b312d7a6ae9
--- /dev/null
+++ llvm/test/tools/llvm-exegesis/AArch64/print-assembled-snippet.s
@@ -0,0 +1,31 @@
+# Check that "assembled_snippet" is fetched correctly from object file.
+# Feel free to update the snippet in this test if the code generated by the
+# snippet repetitor was changed intentionally and it is still fetched correctly.
+
+# RUN: llvm-exegesis --mtriple=aarch64-linux-gnu --mcpu=generic --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-LINUX
+
+# RUN: llvm-exegesis --mtriple=aarch64-windows-gnu --mcpu=generic --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-WINDOWS-GNU
+
+# RUN: llvm-exegesis --mtriple=aarch64-windows-msvc --mcpu=generic --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-WINDOWS-MSVC
+
+# LLVM-EXEGESIS-DEFREG X0 0
+# LLVM-EXEGESIS-DEFREG X1 0
+add x0, x0, x1
+
+# CHECK-LINUX: cpu_name:          generic
+# CHECK-LINUX: llvm_triple:       aarch64-unknown-linux-gnu
+# CHECK-LINUX: assembled_snippet: 000080D2010080D20000018B0000018B0000018B0000018BC0035FD6{{$}}
+
+# CHECK-WINDOWS-GNU: cpu_name:          generic
+# CHECK-WINDOWS-GNU: llvm_triple:       aarch64-unknown-windows-gnu
+# CHECK-WINDOWS-GNU: assembled_snippet: 000080D2010080D20000018B0000018B0000018B0000018BC0035FD6{{$}}
+
+# CHECK-WINDOWS-MSVC: cpu_name:          generic
+# CHECK-WINDOWS-MSVC: llvm_triple:       aarch64-unknown-windows-msvc
+# CHECK-WINDOWS-MSVC: assembled_snippet: 000080D2010080D20000018B0000018B0000018B0000018BC0035FD6{{$}}
diff --git a/llvm/test/tools/llvm-exegesis/Mips/print-assembled-snippet.s b/llvm/test/tools/llvm-exegesis/Mips/print-assembled-snippet.s
new file mode 100644
index 0000000000000..f0844c4145a2b
--- /dev/null
+++ llvm/test/tools/llvm-exegesis/Mips/print-assembled-snippet.s
@@ -0,0 +1,23 @@
+# Check that "assembled_snippet" is fetched correctly from object file.
+# Feel free to update the snippet in this test if the code generated by the
+# snippet repetitor was changed intentionally and it is still fetched correctly.
+
+# RUN: llvm-exegesis --mtriple=mips-linux-gnu --mcpu=generic --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-LINUX
+
+# RUN: llvm-exegesis --mtriple=mips64-linux-gnu --mcpu=generic --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-LINUX-64
+
+# LLVM-EXEGESIS-DEFREG A0 0
+# LLVM-EXEGESIS-DEFREG A1 0
+add $2, $5, $4
+
+# CHECK-LINUX: cpu_name:             generic
+# CHECK-LINUX: llvm_triple:          mips-unknown-linux-gnu
+# CHECK-LINUX: assembled_snippet:    340400003405000000A4102000A4102000A4102000A4102003E00008{{$}}
+
+# CHECK-LINUX-64: cpu_name:           generic
+# CHECK-LINUX-64: llvm_triple:        mips64-unknown-linux-gnu
+# CHECK-LINUX-64: assembled_snippet:  340400003405000000A4102000A4102000A4102000A4102003E00008{{$}}
diff --git a/llvm/test/tools/llvm-exegesis/PowerPC/print-assembled-snippet.s b/llvm/test/tools/llvm-exegesis/PowerPC/print-assembled-snippet.s
new file mode 100644
index 0000000000000..75cad37a768eb
--- /dev/null
+++ llvm/test/tools/llvm-exegesis/PowerPC/print-assembled-snippet.s
@@ -0,0 +1,15 @@
+# Check that "assembled_snippet" is fetched correctly from object file.
+# Feel free to update the snippet in this test if the code generated by the
+# snippet repetitor was changed intentionally and it is still fetched correctly.
+
+# RUN: llvm-exegesis --mtriple=powerpc64le-linux-gnu --mcpu=pwr8 --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-LINUX-64-LE
+
+# LLVM-EXEGESIS-DEFREG R3 0
+# LLVM-EXEGESIS-DEFREG R4 0
+add 3, 4, 3
+
+# CHECK-LINUX-64-LE: cpu_name:           pwr8
+# CHECK-LINUX-64-LE: llvm_triple:        powerpc64le-unknown-linux-gnu
+# CHECK-LINUX-64-LE: assembled_snippet:  0000603800008038141A647C141A647C141A647C141A647C2000804E000000000000000000000000{{$}}
diff --git a/llvm/test/tools/llvm-exegesis/X86/print-assembled-snippet.s b/llvm/test/tools/llvm-exegesis/X86/print-assembled-snippet.s
new file mode 100644
index 0000000000000..532901d0e0430
--- /dev/null
+++ llvm/test/tools/llvm-exegesis/X86/print-assembled-snippet.s
@@ -0,0 +1,32 @@
+# Check that "assembled_snippet" is fetched correctly from object file.
+# Feel free to update the snippet in this test if the code generated by the
+# snippet repetitor was changed intentionally and it is still fetched correctly.
+
+# RUN: llvm-exegesis --mtriple=x86_64-linux-gnu --mcpu=x86-64 --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-LINUX
+
+# RUN: llvm-exegesis --mtriple=x86_64-windows-gnu --mcpu=x86-64 --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-WINDOWS-GNU
+
+# RUN: llvm-exegesis --mtriple=x86_64-windows-msvc --mcpu=x86-64 --benchmark-phase=prepare-and-assemble-snippet \
+# RUN:               --snippets-file=%s --mode=latency --repetition-mode=duplicate \
+# RUN: | FileCheck %s --check-prefix=CHECK-WINDOWS-MSVC
+
+# LLVM-EXEGESIS-DEFREG RDI 0
+# LLVM-EXEGESIS-DEFREG RSI 0
+# LLVM-EXEGESIS-DEFREG RAX 0
+leaq    (%rdi,%rsi), %rax
+
+# CHECK-LINUX: cpu_name:          x86-64
+# CHECK-LINUX: llvm_triple:       x86_64-unknown-linux-gnu
+# CHECK-LINUX: assembled_snippet: 48BF000000000000000048BE000000000000000048B80000000000000000488D0437488D0437488D0437488D0437C3{{$}}
+
+# CHECK-WINDOWS-GNU: cpu_name:          x86-64
+# CHECK-WINDOWS-GNU: llvm_triple:       x86_64-unknown-windows-gnu
+# CHECK-WINDOWS-GNU: assembled_snippet: 565748BF000000000000000048BE000000000000000048B80000000000000000488D0437488D0437488D0437488D04375F5EC3{{$}}
+
+# CHECK-WINDOWS-MSVC: cpu_name:          x86-64
+# CHECK-WINDOWS-MSVC: llvm_triple:       x86_64-unknown-windows-msvc
+# CHECK-WINDOWS-MSVC: assembled_snippet: 565748BF000000000000000048BE000000000000000048B80000000000000000488D0437488D0437488D0437488D04375F5EC3{{$}}
diff --git a/llvm/tools/llvm-exegesis/lib/Assembler.cpp b/llvm/tools/llvm-exegesis/lib/Assembler.cpp
index 07465c71afc70..8e561f587cc2a 100644
--- llvm/tools/llvm-exegesis/lib/Assembler.cpp
+++ llvm/tools/llvm-exegesis/lib/Assembler.cpp
@@ -25,6 +25,7 @@
 #include "llvm/MC/MCInstrInfo.h"
 #include "llvm/Support/Alignment.h"
 #include "llvm/Support/MemoryBuffer.h"
+#include "llvm/Support/raw_ostream.h"
 
 namespace llvm {
 namespace exegesis {
@@ -323,5 +324,38 @@ ExecutableFunction::ExecutableFunction(
       StringRef(reinterpret_cast<const char *>(FunctionAddress), CodeSize);
 }
 
+Error getBenchmarkFunctionBytes(const StringRef InputData,
+                                std::vector<uint8_t> &Bytes) {
+  const auto Holder = getObjectFromBuffer(InputData);
+  const auto *Obj = Holder.getBinary();
+  // See RuntimeDyldImpl::loadObjectImpl(Obj) for much more complete
+  // implementation.
+
+  // Find the only function in the object file.
+  SmallVector<object::SymbolRef, 1> Functions;
+  for (auto &Sym : Obj->symbols()) {
+    auto SymType = Sym.getType();
+    if (SymType && *SymType == object::SymbolRef::Type::ST_Function)
+      Functions.push_back(Sym);
+  }
+  if (Functions.size() != 1)
+    return make_error<Failure>("Exactly one function expected");
+
+  // Find the containing section - it is assumed to contain only this function.
+  auto SectionOrErr = Functions.front().getSection();
+  if (!SectionOrErr || *SectionOrErr == Obj->section_end())
+    return make_error<Failure>("Section not found");
+
+  auto Address = Functions.front().getAddress();
+  if (!Address || *Address != SectionOrErr.get()->getAddress())
+    return make_error<Failure>("Unexpected layout");
+
+  auto ContentsOrErr = SectionOrErr.get()->getContents();
+  if (!ContentsOrErr)
+    return ContentsOrErr.takeError();
+  Bytes.assign(ContentsOrErr->begin(), ContentsOrErr->end());
+  return Error::success();
+}
+
 } // namespace exegesis
 } // namespace llvm
diff --git a/llvm/tools/llvm-exegesis/lib/Assembler.h b/llvm/tools/llvm-exegesis/lib/Assembler.h
index 2a83344b751e5..cd1e398e4a564 100644
--- llvm/tools/llvm-exegesis/lib/Assembler.h
+++ llvm/tools/llvm-exegesis/lib/Assembler.h
@@ -121,6 +121,10 @@ struct ExecutableFunction {
   StringRef FunctionBytes;
 };
 
+// Copies benchmark function's bytes from benchmark object.
+Error getBenchmarkFunctionBytes(const StringRef InputData,
+                                std::vector<uint8_t> &Bytes);
+
 // Creates a void(int8*) MachineFunction.
 MachineFunction &createVoidVoidPtrMachineFunction(StringRef FunctionID,
                                                   Module *Module,
diff --git a/llvm/tools/llvm-exegesis/lib/BenchmarkRunner.cpp b/llvm/tools/llvm-exegesis/lib/BenchmarkRunner.cpp
index e1df2940a9b7e..f934b8ec4c6cc 100644
--- llvm/tools/llvm-exegesis/lib/BenchmarkRunner.cpp
+++ llvm/tools/llvm-exegesis/lib/BenchmarkRunner.cpp
@@ -177,10 +177,10 @@ BenchmarkRunner::getRunnableConfiguration(
                                    LoopBodySizeForSnippet);
     if (Error E = Snippet.takeError())
       return std::move(E);
-    const ExecutableFunction EF(State.createTargetMachine(),
-                                getObjectFromBuffer(*Snippet));
-    const auto FnBytes = EF.getFunctionBytes();
-    llvm::append_range(InstrBenchmark.AssembledSnippet, FnBytes);
+
+    if (auto Err = getBenchmarkFunctionBytes(*Snippet,
+                                             InstrBenchmark.AssembledSnippet))
+      return std::move(Err);
   }
 
   // Assemble NumRepetitions instructions repetitions of the snippet for