gh-90536: Add support for the BOLT post-link binary optimizer (gh-95908)

* Add support for the BOLT post-link binary optimizer

Using [bolt](https://github.com/llvm/llvm-project/tree/main/bolt)
provides a fairly large speedup without any code or functionality
changes. It provides roughly a 1% speedup on pyperformance, and a
4% improvement on the Pyston web macrobenchmarks.

It is gated behind an `--enable-bolt` configure arg because not all
toolchains and environments are supported. It has been tested on a
Linux x86_64 toolchain, using llvm-bolt built from the LLVM 14.0.6
sources (their binary distribution of this version did not include bolt).

Compared to [a previous attempt](https://github.com/faster-cpython/ideas/issues/224),
this commit uses bolt's preferred "instrumentation" approach, as well as adds some non-PIE
flags which enable much better optimizations from bolt.

The effects of this change are a bit more dependent on CPU microarchitecture
than other changes, since it optimizes i-cache behavior which seems
to be a bit more variable between architectures. The 1%/4% numbers
were collected on an Intel Skylake CPU, and on an AMD Zen 3 CPU I
got a slightly larger speedup (2%/4%), and on a c6i.xlarge EC2 instance
I got a slightly lower speedup (1%/3%).

The low speedup on pyperformance is not entirely unexpected, because
BOLT improves i-cache behavior, and the benchmarks in the pyperformance
suite are small and tend to fit in i-cache.

This change uses the existing pgo profiling task (`python -m test --pgo`),
though I was able to measure about a 1% macrobenchmark improvement by
using the macrobenchmarks as the training task. I personally think that
both the PGO and BOLT tasks should be updated to use macrobenchmarks,
but for the sake of splitting up the work this PR uses the existing pgo task.

* Simplify the build flags

* Add a NEWS entry

* Update Makefile.pre.in

Co-authored-by: Dong-hee Na <donghee.na92@gmail.com>

* Update configure.ac

Co-authored-by: Dong-hee Na <donghee.na92@gmail.com>

* Add myself to ACKS

* Add docs

* Other review comments

* fix tab/space issue

* Make it more clear that --enable-bolt is experimental

* Add link to bolt's github page

Co-authored-by: Dong-hee Na <donghee.na92@gmail.com>
This commit is contained in:
Kevin Modzelewski 2022-08-18 17:33:54 -04:00 committed by GitHub
parent 22a95cb511
commit 214eb2cce5
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
7 changed files with 351 additions and 1 deletions

View File

@ -191,7 +191,8 @@ Performance options
-------------------
Configuring Python using ``--enable-optimizations --with-lto`` (PGO + LTO) is
recommended for best performance.
recommended for best performance. The experimental ``--enable-bolt`` flag can
also be used to improve performance.
.. cmdoption:: --enable-optimizations
@ -231,6 +232,24 @@ recommended for best performance.
.. versionadded:: 3.11
To use ThinLTO feature, use ``--with-lto=thin`` on Clang.
.. cmdoption:: --enable-bolt
Enable usage of the `BOLT post-link binary optimizer
<https://github.com/llvm/llvm-project/tree/main/bolt>` (disabled by
default).
BOLT is part of the LLVM project but is not always included in their binary
distributions. This flag requires that ``llvm-bolt`` and ``merge-fdata``
are available.
BOLT is still a fairly new project so this flag should be considered
experimental for now. Because this tool operates on machine code its success
is dependent on a combination of the build environment + the other
optimization configure args + the CPU architecture, and not all combinations
are supported.
.. versionadded:: 3.12
.. cmdoption:: --with-computed-gotos
Enable computed gotos in evaluation loop (enabled by default on supported

View File

@ -133,6 +133,10 @@ Optimizations
It reduces object size by 8 or 16 bytes on 64bit platform. (:pep:`623`)
(Contributed by Inada Naoki in :gh:`92536`.)
* Added experimental support for using the BOLT binary optimizer in the build
process, which improves performance by 1-5%.
(Contributed by Kevin Modzelewski in :gh:`90536`.)
CPython bytecode changes
========================

View File

@ -640,6 +640,16 @@ profile-opt: profile-run-stamp
-rm -f profile-clean-stamp
$(MAKE) @DEF_MAKE_RULE@ CFLAGS_NODIST="$(CFLAGS_NODIST) $(PGO_PROF_USE_FLAG)" LDFLAGS_NODIST="$(LDFLAGS_NODIST)"
bolt-opt: @PREBOLT_RULE@
rm -f *.fdata
@LLVM_BOLT@ ./$(BUILDPYTHON) -instrument -instrumentation-file-append-pid -instrumentation-file=$(abspath $(BUILDPYTHON).bolt) -o $(BUILDPYTHON).bolt_inst
./$(BUILDPYTHON).bolt_inst $(PROFILE_TASK) || true
@MERGE_FDATA@ $(BUILDPYTHON).*.fdata > $(BUILDPYTHON).fdata
@LLVM_BOLT@ ./$(BUILDPYTHON) -o $(BUILDPYTHON).bolt -data=$(BUILDPYTHON).fdata -update-debug-sections -reorder-blocks=ext-tsp -reorder-functions=hfsort+ -split-functions=3 -icf=1 -inline-all -split-eh -reorder-functions-use-hot-size -peepholes=all -jump-tables=aggressive -inline-ap -indirect-call-promotion=all -dyno-stats -use-gnu-stack -frame-opt=hot
rm -f *.fdata
rm -f $(BUILDPYTHON).bolt_inst
mv $(BUILDPYTHON).bolt $(BUILDPYTHON)
# Compile and run with gcov
.PHONY=coverage coverage-lcov coverage-report
coverage:

View File

@ -1212,6 +1212,7 @@ Gideon Mitchell
Tim Mitchell
Zubin Mithra
Florian Mladitsch
Kevin Modzelewski
Doug Moen
Jakub Molinski
Juliette Monsel

View File

@ -0,0 +1,2 @@
Use the BOLT post-link optimizer to improve performance, particularly on
medium-to-large applications.

261
configure generated vendored
View File

@ -887,6 +887,9 @@ LLVM_PROF_FILE
LLVM_PROF_MERGER
PGO_PROF_USE_FLAG
PGO_PROF_GEN_FLAG
MERGE_FDATA
LLVM_BOLT
PREBOLT_RULE
LLVM_AR_FOUND
LLVM_AR
PROFILE_TASK
@ -1049,6 +1052,7 @@ enable_pystats
with_assertions
enable_optimizations
with_lto
enable_bolt
with_address_sanitizer
with_memory_sanitizer
with_undefined_behavior_sanitizer
@ -1774,6 +1778,8 @@ Optional Features:
--enable-pystats enable internal statistics gathering (default is no)
--enable-optimizations enable expensive, stable optimizations (PGO, etc.)
(default is no)
--enable-bolt enable usage of the llvm-bolt post-link optimizer
(default is no)
--enable-loadable-sqlite-extensions
support loadable extensions in the sqlite3 module,
see Doc/library/sqlite3.rst (default is no)
@ -7878,6 +7884,261 @@ $as_echo "$as_me: llvm-ar found via xcrun: ${LLVM_AR}" >&6;}
LDFLAGS_NODIST="$LDFLAGS_NODIST $LTOFLAGS"
fi
# Enable bolt flags
Py_BOLT='false'
{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for --enable-bolt" >&5
$as_echo_n "checking for --enable-bolt... " >&6; }
# Check whether --enable-bolt was given.
if test "${enable_bolt+set}" = set; then :
enableval=$enable_bolt;
if test "$enableval" != no
then
Py_BOLT='true'
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: yes" >&5
$as_echo "yes" >&6; };
else
Py_BOLT='false'
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
$as_echo "no" >&6; };
fi
else
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
$as_echo "no" >&6; }
fi
if test "$Py_BOLT" = 'true' ; then
PREBOLT_RULE="${DEF_MAKE_ALL_RULE}"
DEF_MAKE_ALL_RULE="bolt-opt"
DEF_MAKE_RULE="build_all"
# These flags are required for bolt to work:
CFLAGS_NODIST="$CFLAGS_NODIST -fno-reorder-blocks-and-partition"
LDFLAGS_NODIST="$LDFLAGS_NODIST -Wl,--emit-relocs"
# These flags are required to get good performance from bolt:
CFLAGS_NODIST="$CFLAGS_NODIST -fno-pie"
# We want to add these no-pie flags to linking executables but not shared libraries:
LINKCC="$LINKCC -fno-pie -no-pie"
# Designate the DWARF version into 4 since the LLVM-BOLT does not support DWARF5 yet.
CFLAGS="$CFLAGS -gdwarf-4"
LDFLAGS="$LDFLAGS -gdwarf-4"
if test -n "$ac_tool_prefix"; then
# Extract the first word of "${ac_tool_prefix}llvm-bolt", so it can be a program name with args.
set dummy ${ac_tool_prefix}llvm-bolt; ac_word=$2
{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5
$as_echo_n "checking for $ac_word... " >&6; }
if ${ac_cv_path_LLVM_BOLT+:} false; then :
$as_echo_n "(cached) " >&6
else
case $LLVM_BOLT in
[\\/]* | ?:[\\/]*)
ac_cv_path_LLVM_BOLT="$LLVM_BOLT" # Let the user override the test with a path.
;;
*)
as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
for as_dir in ${llvm_path}
do
IFS=$as_save_IFS
test -z "$as_dir" && as_dir=.
for ac_exec_ext in '' $ac_executable_extensions; do
if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
ac_cv_path_LLVM_BOLT="$as_dir/$ac_word$ac_exec_ext"
$as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5
break 2
fi
done
done
IFS=$as_save_IFS
;;
esac
fi
LLVM_BOLT=$ac_cv_path_LLVM_BOLT
if test -n "$LLVM_BOLT"; then
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $LLVM_BOLT" >&5
$as_echo "$LLVM_BOLT" >&6; }
else
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
$as_echo "no" >&6; }
fi
fi
if test -z "$ac_cv_path_LLVM_BOLT"; then
ac_pt_LLVM_BOLT=$LLVM_BOLT
# Extract the first word of "llvm-bolt", so it can be a program name with args.
set dummy llvm-bolt; ac_word=$2
{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5
$as_echo_n "checking for $ac_word... " >&6; }
if ${ac_cv_path_ac_pt_LLVM_BOLT+:} false; then :
$as_echo_n "(cached) " >&6
else
case $ac_pt_LLVM_BOLT in
[\\/]* | ?:[\\/]*)
ac_cv_path_ac_pt_LLVM_BOLT="$ac_pt_LLVM_BOLT" # Let the user override the test with a path.
;;
*)
as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
for as_dir in ${llvm_path}
do
IFS=$as_save_IFS
test -z "$as_dir" && as_dir=.
for ac_exec_ext in '' $ac_executable_extensions; do
if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
ac_cv_path_ac_pt_LLVM_BOLT="$as_dir/$ac_word$ac_exec_ext"
$as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5
break 2
fi
done
done
IFS=$as_save_IFS
;;
esac
fi
ac_pt_LLVM_BOLT=$ac_cv_path_ac_pt_LLVM_BOLT
if test -n "$ac_pt_LLVM_BOLT"; then
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_pt_LLVM_BOLT" >&5
$as_echo "$ac_pt_LLVM_BOLT" >&6; }
else
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
$as_echo "no" >&6; }
fi
if test "x$ac_pt_LLVM_BOLT" = x; then
LLVM_BOLT="''"
else
case $cross_compiling:$ac_tool_warned in
yes:)
{ $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5
$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;}
ac_tool_warned=yes ;;
esac
LLVM_BOLT=$ac_pt_LLVM_BOLT
fi
else
LLVM_BOLT="$ac_cv_path_LLVM_BOLT"
fi
if test -n "${LLVM_BOLT}" -a -x "${LLVM_BOLT}"
then
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: \"Found llvm-bolt\"" >&5
$as_echo "\"Found llvm-bolt\"" >&6; }
else
as_fn_error $? "llvm-bolt is required for a --enable-bolt build but could not be found." "$LINENO" 5
fi
if test -n "$ac_tool_prefix"; then
# Extract the first word of "${ac_tool_prefix}merge-fdata", so it can be a program name with args.
set dummy ${ac_tool_prefix}merge-fdata; ac_word=$2
{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5
$as_echo_n "checking for $ac_word... " >&6; }
if ${ac_cv_path_MERGE_FDATA+:} false; then :
$as_echo_n "(cached) " >&6
else
case $MERGE_FDATA in
[\\/]* | ?:[\\/]*)
ac_cv_path_MERGE_FDATA="$MERGE_FDATA" # Let the user override the test with a path.
;;
*)
as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
for as_dir in ${llvm_path}
do
IFS=$as_save_IFS
test -z "$as_dir" && as_dir=.
for ac_exec_ext in '' $ac_executable_extensions; do
if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
ac_cv_path_MERGE_FDATA="$as_dir/$ac_word$ac_exec_ext"
$as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5
break 2
fi
done
done
IFS=$as_save_IFS
;;
esac
fi
MERGE_FDATA=$ac_cv_path_MERGE_FDATA
if test -n "$MERGE_FDATA"; then
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $MERGE_FDATA" >&5
$as_echo "$MERGE_FDATA" >&6; }
else
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
$as_echo "no" >&6; }
fi
fi
if test -z "$ac_cv_path_MERGE_FDATA"; then
ac_pt_MERGE_FDATA=$MERGE_FDATA
# Extract the first word of "merge-fdata", so it can be a program name with args.
set dummy merge-fdata; ac_word=$2
{ $as_echo "$as_me:${as_lineno-$LINENO}: checking for $ac_word" >&5
$as_echo_n "checking for $ac_word... " >&6; }
if ${ac_cv_path_ac_pt_MERGE_FDATA+:} false; then :
$as_echo_n "(cached) " >&6
else
case $ac_pt_MERGE_FDATA in
[\\/]* | ?:[\\/]*)
ac_cv_path_ac_pt_MERGE_FDATA="$ac_pt_MERGE_FDATA" # Let the user override the test with a path.
;;
*)
as_save_IFS=$IFS; IFS=$PATH_SEPARATOR
for as_dir in ${llvm_path}
do
IFS=$as_save_IFS
test -z "$as_dir" && as_dir=.
for ac_exec_ext in '' $ac_executable_extensions; do
if as_fn_executable_p "$as_dir/$ac_word$ac_exec_ext"; then
ac_cv_path_ac_pt_MERGE_FDATA="$as_dir/$ac_word$ac_exec_ext"
$as_echo "$as_me:${as_lineno-$LINENO}: found $as_dir/$ac_word$ac_exec_ext" >&5
break 2
fi
done
done
IFS=$as_save_IFS
;;
esac
fi
ac_pt_MERGE_FDATA=$ac_cv_path_ac_pt_MERGE_FDATA
if test -n "$ac_pt_MERGE_FDATA"; then
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: $ac_pt_MERGE_FDATA" >&5
$as_echo "$ac_pt_MERGE_FDATA" >&6; }
else
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: no" >&5
$as_echo "no" >&6; }
fi
if test "x$ac_pt_MERGE_FDATA" = x; then
MERGE_FDATA="''"
else
case $cross_compiling:$ac_tool_warned in
yes:)
{ $as_echo "$as_me:${as_lineno-$LINENO}: WARNING: using cross tools not prefixed with host triplet" >&5
$as_echo "$as_me: WARNING: using cross tools not prefixed with host triplet" >&2;}
ac_tool_warned=yes ;;
esac
MERGE_FDATA=$ac_pt_MERGE_FDATA
fi
else
MERGE_FDATA="$ac_cv_path_MERGE_FDATA"
fi
if test -n "${MERGE_FDATA}" -a -x "${MERGE_FDATA}"
then
{ $as_echo "$as_me:${as_lineno-$LINENO}: result: \"Found merge-fdata\"" >&5
$as_echo "\"Found merge-fdata\"" >&6; }
else
as_fn_error $? "merge-fdata is required for a --enable-bolt build but could not be found." "$LINENO" 5
fi
fi
# Enable PGO flags.

View File

@ -1917,6 +1917,59 @@ if test "$Py_LTO" = 'true' ; then
LDFLAGS_NODIST="$LDFLAGS_NODIST $LTOFLAGS"
fi
# Enable bolt flags
Py_BOLT='false'
AC_MSG_CHECKING(for --enable-bolt)
AC_ARG_ENABLE(bolt, AS_HELP_STRING(
[--enable-bolt],
[enable usage of the llvm-bolt post-link optimizer (default is no)]),
[
if test "$enableval" != no
then
Py_BOLT='true'
AC_MSG_RESULT(yes);
else
Py_BOLT='false'
AC_MSG_RESULT(no);
fi],
[AC_MSG_RESULT(no)])
AC_SUBST(PREBOLT_RULE)
if test "$Py_BOLT" = 'true' ; then
PREBOLT_RULE="${DEF_MAKE_ALL_RULE}"
DEF_MAKE_ALL_RULE="bolt-opt"
DEF_MAKE_RULE="build_all"
# These flags are required for bolt to work:
CFLAGS_NODIST="$CFLAGS_NODIST -fno-reorder-blocks-and-partition"
LDFLAGS_NODIST="$LDFLAGS_NODIST -Wl,--emit-relocs"
# These flags are required to get good performance from bolt:
CFLAGS_NODIST="$CFLAGS_NODIST -fno-pie"
# We want to add these no-pie flags to linking executables but not shared libraries:
LINKCC="$LINKCC -fno-pie -no-pie"
# Designate the DWARF version into 4 since the LLVM-BOLT does not support DWARF5 yet.
CFLAGS="$CFLAGS -gdwarf-4"
LDFLAGS="$LDFLAGS -gdwarf-4"
AC_SUBST(LLVM_BOLT)
AC_PATH_TOOL(LLVM_BOLT, llvm-bolt, '', ${llvm_path})
if test -n "${LLVM_BOLT}" -a -x "${LLVM_BOLT}"
then
AC_MSG_RESULT("Found llvm-bolt")
else
AC_MSG_ERROR([llvm-bolt is required for a --enable-bolt build but could not be found.])
fi
AC_SUBST(MERGE_FDATA)
AC_PATH_TOOL(MERGE_FDATA, merge-fdata, '', ${llvm_path})
if test -n "${MERGE_FDATA}" -a -x "${MERGE_FDATA}"
then
AC_MSG_RESULT("Found merge-fdata")
else
AC_MSG_ERROR([merge-fdata is required for a --enable-bolt build but could not be found.])
fi
fi
# Enable PGO flags.
AC_SUBST(PGO_PROF_GEN_FLAG)
AC_SUBST(PGO_PROF_USE_FLAG)