mirror of
https://sourceware.org/git/glibc.git
synced 2024-11-25 22:40:05 +00:00
6a5d6ea128
We have a single thread that runs a no-op initialization once and then repeatedly runs checks of the initialization (i.e., an acquire load and conditional jump) in a tight loop. This gives us, on average, the best-case latency of pthread_once (the initialization is the exactly-once slow path, and we're not looking at initialization-related synchronization overheads in this case).
158 lines
4.7 KiB
Makefile
158 lines
4.7 KiB
Makefile
# Copyright (C) 2013-2014 Free Software Foundation, Inc.
|
||
# This file is part of the GNU C Library.
|
||
|
||
# The GNU C Library is free software; you can redistribute it and/or
|
||
# modify it under the terms of the GNU Lesser General Public
|
||
# License as published by the Free Software Foundation; either
|
||
# version 2.1 of the License, or (at your option) any later version.
|
||
|
||
# The GNU C Library is distributed in the hope that it will be useful,
|
||
# but WITHOUT ANY WARRANTY; without even the implied warranty of
|
||
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
||
# Lesser General Public License for more details.
|
||
|
||
# You should have received a copy of the GNU Lesser General Public
|
||
# License along with the GNU C Library; if not, see
|
||
# <http://www.gnu.org/licenses/>.
|
||
|
||
|
||
# Makefile for benchmark tests. The only useful target here is `bench`.
|
||
# Add benchmark functions in alphabetical order.
|
||
|
||
subdir := benchtests
|
||
|
||
include ../Makeconfig
|
||
bench := acos acosh asin asinh atan atanh cos cosh exp exp2 ffs ffsll \
|
||
log log2 modf pow pthread_once rint sin sincos sinh sqrt tan tanh
|
||
|
||
# String function benchmarks.
|
||
string-bench := bcopy bzero memccpy memchr memcmp memcpy memmem memmove \
|
||
mempcpy memset rawmemchr stpcpy stpncpy strcasecmp strcasestr \
|
||
strcat strchr strchrnul strcmp strcpy strcspn strlen \
|
||
strncasecmp strncat strncmp strncpy strnlen strpbrk strrchr \
|
||
strspn strstr strcpy_chk stpcpy_chk memrchr strsep strtok
|
||
string-bench-all := $(string-bench)
|
||
|
||
stdlib-bench := strtod
|
||
|
||
benchset := $(string-bench-all) $(stdlib-bench)
|
||
|
||
CFLAGS-bench-ffs.c += -fno-builtin
|
||
CFLAGS-bench-ffsll.c += -fno-builtin
|
||
|
||
LDLIBS-bench-acos = -lm
|
||
LDLIBS-bench-acosh = -lm
|
||
LDLIBS-bench-asin = -lm
|
||
LDLIBS-bench-asinh = -lm
|
||
LDLIBS-bench-atan = -lm
|
||
LDLIBS-bench-atanh = -lm
|
||
LDLIBS-bench-cos = -lm
|
||
LDLIBS-bench-cosh = -lm
|
||
LDLIBS-bench-exp = -lm
|
||
LDLIBS-bench-exp2 = -lm
|
||
LDLIBS-bench-log = -lm
|
||
LDLIBS-bench-log2 = -lm
|
||
LDLIBS-bench-pow = -lm
|
||
LDLIBS-bench-pthread_once = -lpthread
|
||
LDLIBS-bench-rint = -lm
|
||
LDLIBS-bench-sin = -lm
|
||
LDLIBS-bench-sinh = -lm
|
||
LDLIBS-bench-sqrt = -lm
|
||
LDLIBS-bench-tan = -lm
|
||
LDLIBS-bench-tanh = -lm
|
||
LDLIBS-bench-sincos = -lm
|
||
|
||
|
||
|
||
# Rules to build and execute the benchmarks. Do not put any benchmark
|
||
# parameters beyond this point.
|
||
|
||
# We don't want the benchmark programs to run in parallel since that could
|
||
# affect their performance.
|
||
.NOTPARALLEL:
|
||
|
||
include ../Rules
|
||
|
||
binaries-bench := $(addprefix $(objpfx)bench-,$(bench))
|
||
binaries-benchset := $(addprefix $(objpfx)bench-,$(benchset))
|
||
|
||
# The default duration: 10 seconds.
|
||
ifndef BENCH_DURATION
|
||
BENCH_DURATION := 10
|
||
endif
|
||
|
||
CPPFLAGS-nonlib += -DDURATION=$(BENCH_DURATION)
|
||
|
||
# Use clock_gettime to measure performance of functions. The default is to use
|
||
# HP_TIMING if it is available.
|
||
ifdef USE_CLOCK_GETTIME
|
||
CPPFLAGS-nonlib += -DUSE_CLOCK_GETTIME
|
||
endif
|
||
|
||
DETAILED_OPT :=
|
||
|
||
ifdef DETAILED
|
||
DETAILED_OPT := -d
|
||
endif
|
||
|
||
# This makes sure CPPFLAGS-nonlib and CFLAGS-nonlib are passed
|
||
# for all these modules.
|
||
cpp-srcs-left := $(binaries-benchset:=.c) $(binaries-bench:=.c)
|
||
lib := nonlib
|
||
include $(patsubst %,$(..)cppflags-iterator.mk,$(cpp-srcs-left))
|
||
|
||
bench-deps := bench-skeleton.c bench-timing.h Makefile
|
||
|
||
run-bench = $(test-wrapper-env) \
|
||
GCONV_PATH=$(common-objpfx)iconvdata LC_ALL=C \
|
||
$($*-ENV) $(rtld-prefix) $${run}
|
||
|
||
timing-type := $(objpfx)bench-timing-type
|
||
|
||
bench-clean:
|
||
rm -f $(binaries-bench) $(addsuffix .o,$(binaries-bench))
|
||
rm -f $(binaries-benchset) $(addsuffix .o,$(binaries-benchset))
|
||
rm -f $(timing-type) $(addsuffix .o,$(timing-type))
|
||
|
||
bench: $(timing-type) bench-set bench-func
|
||
|
||
bench-set: $(binaries-benchset)
|
||
for run in $^; do \
|
||
echo "Running $${run}"; \
|
||
$(run-bench) > $${run}.out; \
|
||
done
|
||
|
||
# Build and execute the benchmark functions. This target generates JSON
|
||
# formatted bench.out. Each of the programs produce independent JSON output,
|
||
# so one could even execute them individually and process it using any JSON
|
||
# capable language or tool.
|
||
bench-func: $(binaries-bench)
|
||
{ echo "{"; \
|
||
$(timing-type); \
|
||
echo " ,\"functions\": {"; \
|
||
for run in $^; do \
|
||
if ! [ "x$${run}" = "x$<" ]; then \
|
||
echo ","; \
|
||
fi; \
|
||
echo "Running $${run}" >&2; \
|
||
$(run-bench) $(DETAILED_OPT); \
|
||
done; \
|
||
echo " }"; \
|
||
echo "}"; } > $(objpfx)bench.out-tmp; \
|
||
if [ -f $(objpfx)bench.out ]; then \
|
||
mv -f $(objpfx)bench.out $(objpfx)bench.out.old; \
|
||
fi; \
|
||
mv -f $(objpfx)bench.out-tmp $(objpfx)bench.out
|
||
|
||
$(timing-type) $(binaries-bench) $(binaries-benchset): %: %.o \
|
||
$(sort $(filter $(common-objpfx)lib%,$(link-libc))) \
|
||
$(addprefix $(csu-objpfx),start.o) $(+preinit) $(+postinit)
|
||
$(+link)
|
||
|
||
$(objpfx)bench-%.c: %-inputs $(bench-deps)
|
||
{ if [ -n "$($*-INCLUDE)" ]; then \
|
||
cat $($*-INCLUDE); \
|
||
fi; \
|
||
$(.)scripts/bench.py $(patsubst %-inputs,%,$<); } > $@-tmp
|
||
mv -f $@-tmp $@
|