GerbilSoft/zlib-ng
1
Fork 0
mirror of https://github.com/GerbilSoft/zlib-ng.git synced 2025-06-18 11:35:35 -04:00
Code Issues Actions Packages Projects Releases Wiki Activity
19b3b7f6e9
zlib-ng/win32/Makefile.msc
Adam Stylinski 724dc0cfb4 Explicit SSE2 vectorization of Chorba CRC method 
The version that's currently in the generic implementation for 32768
byte buffers leverages the stack. It manages to autovectorize but
unfortunately the trips to the stack hurt its performance for CPUs which
need this the most. This version is explicitly SIMD vectorized and
doesn't use trips to the stack.  In my testing it's ~10% faster than the
"small" variant, and about 42% faster than the "32768" variant.
2025-03-28 20:43:59 +01:00

299 lines
12 KiB
Makefile
Raw Blame History

# Makefile for zlib using Microsoft (Visual) C
# zlib is copyright (C) 1995-2024 Jean-loup Gailly and Mark Adler
#
# Usage:
# nmake -f win32/Makefile.msc (standard build)
# nmake -f win32/Makefile.msc LOC=-DFOO (nonstandard build)
# The toplevel directory of the source tree.
#
TOP = .
# optional build flags
LOC =
# variables
STATICLIB = zlib.lib
SHAREDLIB = zlib1.dll
IMPLIB = zdll.lib
SYMBOL_PREFIX =
CC = cl
CXX = cl
LD = link
AR = lib
RC = rc
CP = copy /y
INCLUDES = -I$(TOP) -I$(TOP)/arch/x86 -I$(TOP)/arch/generic
CFLAGS = -nologo -MD -W3 -O2 -Oy- -Zi -Fd"zlib" $(LOC) $(INCLUDES)
CXXFLAGS = -nologo -EHsc -MD -W3 -O2 -Oy- -Zi -Fd"zlib" $(LOC) $(INCLUDES)
WFLAGS = \
-D_CRT_SECURE_NO_DEPRECATE \
-D_CRT_NONSTDC_NO_DEPRECATE \
-DX86_FEATURES \
-DX86_PCLMULQDQ_CRC \
-DX86_SSE2 \
-DX86_SSE42 \
-DX86_SSSE3 \
-DX86_AVX2
LDFLAGS = -nologo -debug -incremental:no -opt:ref -manifest
ARFLAGS = -nologo
RCFLAGS = /dWIN32 /r
DEFFILE = zlib.def
RCFILE = zlib1.rc
RESFILE = zlib1.res
WITH_GZFILEOP = yes
ZLIB_COMPAT =
SUFFIX =
OBJS = \
adler32.obj \
adler32_c.obj \
adler32_avx2.obj \
adler32_avx512.obj \
adler32_avx512_vnni.obj \
adler32_sse42.obj \
adler32_ssse3.obj \
adler32_fold_c.obj \
chunkset_c.obj \
chunkset_avx2.obj \
chunkset_sse2.obj \
chunkset_ssse3.obj \
chorba_sse2.obj \
compare256_c.obj \
compare256_avx2.obj \
compare256_sse2.obj \
compress.obj \
cpu_features.obj \
crc32.obj \
crc32_braid_c.obj \
crc32_braid_comb.obj \
crc32_c.obj \
crc32_chorba_c.obj \
crc32_fold_c.obj \
crc32_pclmulqdq.obj \
deflate.obj \
deflate_fast.obj \
deflate_huff.obj \
deflate_medium.obj \
deflate_quick.obj \
deflate_rle.obj \
deflate_slow.obj \
deflate_stored.obj \
functable.obj \
infback.obj \
inflate.obj \
inftrees.obj \
insert_string.obj \
insert_string_roll.obj \
slide_hash_c.obj \
slide_hash_avx2.obj \
slide_hash_sse2.obj \
trees.obj \
uncompr.obj \
zutil.obj \
x86_features.obj \
#
!if "$(ZLIB_COMPAT)" != ""
WITH_GZFILEOP = yes
WFLAGS = $(WFLAGS) -DZLIB_COMPAT
DEFFILE = zlibcompat.def
!else
STATICLIB = zlib-ng.lib
SHAREDLIB = zlib-ng1.dll
IMPLIB = zngdll.lib
DEFFILE = zlib-ng.def
RCFILE = zlib-ng1.rc
RESFILE = zlib-ng1.res
SUFFIX = -ng
!endif
!if "$(WITH_GZFILEOP)" != ""
WFLAGS = $(WFLAGS) -DWITH_GZFILEOP
OBJS = $(OBJS) gzlib.obj gzread.obj gzwrite.obj
!endif
# targets
all: $(STATICLIB) $(SHAREDLIB) $(IMPLIB) \
example.exe minigzip.exe example_d.exe minigzip_d.exe
!if "$(SYMBOL_PREFIX)" != ""
zlib_name_mangling$(SUFFIX).h: zlib_name_mangling$(SUFFIX).h.in
cscript $(TOP)\win32\replace.vbs $(TOP)\zlib_name_mangling$(SUFFIX).h.in zlib_name_mangling$(SUFFIX).h "@ZLIB_SYMBOL_PREFIX@" "$(SYMBOL_PREFIX)"
!else
zlib_name_mangling$(SUFFIX).h: zlib_name_mangling.h.empty
$(CP) $(TOP)\zlib_name_mangling.h.empty zlib_name_mangling$(SUFFIX).h
!endif
zlib$(SUFFIX).h: zlib$(SUFFIX).h.in
cscript $(TOP)\win32\replace.vbs $(TOP)\zlib$(SUFFIX).h.in zlib$(SUFFIX).h "@ZLIB_SYMBOL_PREFIX@" "$(SYMBOL_PREFIX)"
gzread.c: gzread.c.in
cscript $(TOP)\win32\replace.vbs $(TOP)\gzread.c.in gzread.c "@ZLIB_SYMBOL_PREFIX@" "$(SYMBOL_PREFIX)"
zconf: $(TOP)/zconf$(SUFFIX).h.in $(TOP)/zlib$(SUFFIX).h $(TOP)/zlib_name_mangling$(SUFFIX).h
$(CP) $(TOP)\zconf$(SUFFIX).h.in $(TOP)\zconf$(SUFFIX).h
$(TOP)/win32/$(DEFFILE): $(TOP)/win32/$(DEFFILE).in
cscript $(TOP)\win32\replace.vbs $(TOP)/win32/$(DEFFILE).in $(TOP)/win32/$(DEFFILE) "@ZLIB_SYMBOL_PREFIX@" "$(SYMBOL_PREFIX)"
$(STATICLIB): zconf $(OBJS)
$(AR) $(ARFLAGS) -out:$@ $(OBJS)
$(IMPLIB): $(SHAREDLIB)
$(SHAREDLIB): zconf $(TOP)/win32/$(DEFFILE) $(OBJS) $(RESFILE)
$(LD) $(LDFLAGS) -def:$(TOP)/win32/$(DEFFILE) -dll -implib:$(IMPLIB) \
-out:$@ $(OBJS) $(RESFILE)
if exist $@.manifest \
mt -nologo -manifest $@.manifest -outputresource:$@;2
depcheck.exe: depcheck.obj
$(LD) $(LDFLAGS) depcheck.obj
if exist $@.manifest \
mt -nologo -manifest $@.manifest -outputresource:$@;1
example.exe: example.obj gzlib2.obj gzread2.obj gzwrite2.obj $(STATICLIB)
$(LD) $(LDFLAGS) example.obj gzlib2.obj gzread2.obj gzwrite2.obj $(STATICLIB)
if exist $@.manifest \
mt -nologo -manifest $@.manifest -outputresource:$@;1
minigzip.exe: minigzip.obj gzlib2.obj gzread2.obj gzwrite2.obj $(STATICLIB)
$(LD) $(LDFLAGS) minigzip.obj gzlib2.obj gzread2.obj gzwrite2.obj $(STATICLIB)
if exist $@.manifest \
mt -nologo -manifest $@.manifest -outputresource:$@;1
example_d.exe: example.obj gzlib2.obj gzread2.obj gzwrite2.obj $(IMPLIB)
$(LD) $(LDFLAGS) -out:$@ example.obj gzlib2.obj gzread2.obj gzwrite2.obj $(IMPLIB)
if exist $@.manifest \
mt -nologo -manifest $@.manifest -outputresource:$@;1
minigzip_d.exe: minigzip.obj gzlib2.obj gzread2.obj gzwrite2.obj $(IMPLIB)
$(LD) $(LDFLAGS) -out:$@ minigzip.obj gzlib2.obj gzread2.obj gzwrite2.obj $(IMPLIB)
if exist $@.manifest \
mt -nologo -manifest $@.manifest -outputresource:$@;1
{$(TOP)}.c.obj:
$(CC) -c $(WFLAGS) $(CFLAGS) $<
gzlib2.obj: gzlib.c $(TOP)/zbuild.h $(TOP)/gzguts.h $(TOP)/zutil_p.h
$(CC) -c $(WFLAGS) $(CFLAGS) -DWITH_GZFILEOP -Fogzlib2.obj gzlib.c
gzread2.obj: gzread.c $(TOP)/zbuild.h $(TOP)/gzguts.h $(TOP)/zutil_p.h
$(CC) -c $(WFLAGS) $(CFLAGS) -DWITH_GZFILEOP -Fogzread2.obj gzread.c
gzwrite2.obj: gzwrite.c $(TOP)/zbuild.h $(TOP)/gzguts.h $(TOP)/zutil_p.h
$(CC) -c $(WFLAGS) $(CFLAGS) -DWITH_GZFILEOP -Fogzwrite2.obj gzwrite.c
{$(TOP)/arch/x86}.c.obj:
$(CC) -c -I$(TOP) $(WFLAGS) $(CFLAGS) $<
{$(TOP)/arch/generic}.c.obj:
$(CC) -c -I$(TOP) $(WFLAGS) $(CFLAGS) $<
{$(TOP)/test}.c.obj:
$(CC) -c -I$(TOP) $(WFLAGS) $(CFLAGS) -DWITH_GZFILEOP $<
$(TOP)/zconf$(SUFFIX).h: zconf
{$(TOP)/win32}.cpp.obj:
$(CXX) -c -I$(TOP) $(WFLAGS) $(CXXFLAGS) $<
adler32.obj: $(TOP)/adler32.c $(TOP)/zbuild.h $(TOP)/functable.h $(TOP)/adler32_p.h
adler32_c.obj: $(TOP)/arch/generic/adler32_c.c $(TOP)/zbuild.h $(TOP)/functable.h $(TOP)/adler32_p.h
adler32_avx2.obj: $(TOP)/arch/x86/adler32_avx2.c $(TOP)/zbuild.h $(TOP)/adler32_p.h $(TOP)/arch/x86/adler32_avx2_p.h $(TOP)/arch/x86/x86_intrins.h
adler32_avx512.obj: $(TOP)/arch/x86/adler32_avx512.c $(TOP)/zbuild.h $(TOP)/arch_functions.h $(TOP)/adler32_p.h $(TOP)/arch/x86/adler32_avx512_p.h $(TOP)/arch/x86/x86_intrins.h
adler32_avx512_vnni.obj: $(TOP)/arch/x86/adler32_avx512_vnni.c $(TOP)/zbuild.h $(TOP)/arch_functions.h $(TOP)/adler32_p.h $(TOP)/arch/x86/adler32_avx512_p.h \
$(TOP)/arch/x86/adler32_avx2_p.h $(TOP)/arch/x86/x86_intrins.h
adler32_sse42.obj: $(TOP)/arch/x86/adler32_sse42.c $(TOP)/zbuild.h $(TOP)/adler32_p.h \
$(TOP)/arch/x86/adler32_ssse3_p.h
adler32_ssse3.obj: $(TOP)/arch/x86/adler32_ssse3.c $(TOP)/zbuild.h $(TOP)/adler32_p.h \
$(TOP)/arch/x86/adler32_ssse3_p.h
adler32_fold_c.obj: $(TOP)/arch/generic/adler32_fold_c.c $(TOP)/zbuild.h $(TOP)/functable.h
chorba_sse2.obj: $(TOP)/arch/x86/chorba_sse2.c $(TOP)/zbuild.h $(TOP)/crc32_braid_tbl.h $(TOP)/crc32_braid_p.h
chunkset_c.obj: $(TOP)/arch/generic/chunkset_c.c $(TOP)/zbuild.h $(TOP)/chunkset_tpl.h $(TOP)/inffast_tpl.h
chunkset_avx2.obj: $(TOP)/arch/x86/chunkset_avx2.c $(TOP)/zbuild.h $(TOP)/chunkset_tpl.h $(TOP)/inffast_tpl.h $(TOP)/arch/generic/chunk_permute_table.h
chunkset_sse2.obj: $(TOP)/arch/x86/chunkset_sse2.c $(TOP)/zbuild.h $(TOP)/chunkset_tpl.h $(TOP)/inffast_tpl.h
chunkset_ssse3.obj: $(TOP)/arch/x86/chunkset_ssse3.c $(TOP)/zbuild.h $(TOP)/chunkset_tpl.h $(TOP)/inffast_tpl.h $(TOP)/arch/generic/chunk_permute_table.h
compare256_c.obj: $(TOP)/arch/generic/compare256_c.c $(TOP)/zbuild.h $(TOP)/zmemory.h $(TOP)/deflate.h $(TOP)/fallback_builtins.h $(TOP)/match_tpl.h
compare256_avx2.obj: $(TOP)/arch/x86/compare256_avx2.c $(TOP)/zbuild.h $(TOP)/zmemory.h $(TOP)/deflate.h $(TOP)/fallback_builtins.h $(TOP)/match_tpl.h
compare256_sse2.obj: $(TOP)/arch/x86/compare256_sse2.c $(TOP)/zbuild.h $(TOP)/zmemory.h $(TOP)/deflate.h $(TOP)/fallback_builtins.h $(TOP)/match_tpl.h
compress.obj: $(TOP)/compress.c $(TOP)/zbuild.h $(TOP)/zutil.h
cpu_features.obj: $(TOP)/cpu_features.c $(TOP)/cpu_features.h $(TOP)/zbuild.h
crc32.obj: $(TOP)/crc32.c $(TOP)/zbuild.h $(TOP)/functable.h $(TOP)/crc32_braid_tbl.h
crc32_braid_c.obj: $(TOP)/arch/generic/crc32_braid_c.c $(TOP)/zbuild.h $(TOP)/crc32_braid_p.h $(TOP)/crc32_braid_tbl.h
crc32_braid_comb.obj: $(TOP)/crc32_braid_comb.c $(TOP)/zutil.h $(TOP)/crc32_braid_p.h $(TOP)/crc32_braid_tbl.h $(TOP)/crc32_braid_comb_p.h
crc32_c.obj: $(TOP)/arch/generic/crc32_c.c $(TOP)/zbuild.h $(TOP)/crc32_braid_p.h
crc32_chorba_c.obj: $(TOP)/arch/generic/crc32_chorba_c.c $(TOP)/zbuild.h $(TOP)/crc32_braid_p.h $(TOP)/crc32_braid_tbl.h
crc32_fold_c.obj: $(TOP)/arch/generic/crc32_fold_c.c $(TOP)/zbuild.h $(TOP)/crc32.h $(TOP)/functable.h $(TOP)/zutil.h
crc32_pclmulqdq.obj: $(TOP)/arch/x86/crc32_pclmulqdq.c $(TOP)/arch/x86/crc32_pclmulqdq_tpl.h
deflate.obj: $(TOP)/deflate.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h
deflate_fast.obj: $(TOP)/deflate_fast.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h
deflate_huff.obj: $(TOP)/deflate_huff.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h
deflate_medium.obj: $(TOP)/deflate_medium.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h
deflate_quick.obj: $(TOP)/deflate_quick.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h $(TOP)/trees_emit.h $(TOP)/zmemory.h
deflate_rle.obj: $(TOP)/deflate_rle.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h $(TOP)/compare256_rle.h
deflate_slow.obj: $(TOP)/deflate_slow.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h
deflate_stored.obj: $(TOP)/deflate_stored.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/deflate_p.h $(TOP)/functable.h
functable.obj: $(TOP)/functable.c $(TOP)/zbuild.h $(TOP)/functable.h $(TOP)/cpu_features.h $(TOP)/arch/x86/x86_features.h $(TOP)/arch_functions.h
gzlib.obj: $(TOP)/gzlib.c $(TOP)/zbuild.h $(TOP)/gzguts.h $(TOP)/zutil_p.h
gzread.obj: $(TOP)/gzread.c $(TOP)/zbuild.h $(TOP)/gzguts.h $(TOP)/zutil_p.h
gzwrite.obj: $(TOP)/gzwrite.c $(TOP)/zbuild.h $(TOP)/gzguts.h $(TOP)/zutil_p.h
infback.obj: $(TOP)/infback.c $(TOP)/zbuild.h $(TOP)/zutil.h $(TOP)/inftrees.h $(TOP)/inflate.h $(TOP)/inflate_p.h $(TOP)/functable.h
inflate.obj: $(TOP)/inflate.c $(TOP)/zbuild.h $(TOP)/zutil.h $(TOP)/inftrees.h $(TOP)/inflate.h $(TOP)/inflate_p.h $(TOP)/functable.h $(TOP)/inffixed_tbl.h
inftrees.obj: $(TOP)/inftrees.c $(TOP)/zbuild.h $(TOP)/zutil.h $(TOP)/inftrees.h
insert_string.obj: $(TOP)/insert_string.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/insert_string_tpl.h
insert_string_roll.obj: $(TOP)/insert_string_roll.c $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/insert_string_tpl.h
slide_hash_c.obj: $(TOP)/arch/generic/slide_hash_c.c $(TOP)/zbuild.h $(TOP)/deflate.h
slide_hash_avx2.obj: $(TOP)/arch/x86/slide_hash_avx2.c $(TOP)/zbuild.h $(TOP)/deflate.h
slide_hash_sse2.obj: $(TOP)/arch/x86/slide_hash_sse2.c $(TOP)/zbuild.h $(TOP)/deflate.h
trees.obj: $(TOP)/trees.c $(TOP)/trees.h $(TOP)/trees_emit.h $(TOP)/zbuild.h $(TOP)/deflate.h $(TOP)/trees_tbl.h
uncompr.obj: $(TOP)/uncompr.c $(TOP)/zbuild.h $(TOP)/zutil.h
zutil.obj: $(TOP)/zutil.c $(TOP)/zbuild.h $(TOP)/zutil.h $(TOP)/zutil_p.h
$(RESFILE): $(TOP)/win32/$(RCFILE)
$(RC) $(RCFLAGS) /fo$@ $(TOP)/win32/$(RCFILE)
# testing
depcheck: depcheck.exe
depcheck win32\Makefile.msc .
depcheck win32\Makefile.arm .
depcheck win32\Makefile.a64 .
test: example.exe minigzip.exe depcheck
example
echo hello world | minigzip | minigzip -d
testdll: example_d.exe minigzip_d.exe
example_d
echo hello world | minigzip_d | minigzip_d -d
depcheck.obj: $(TOP)/win32/depcheck.cpp
example.obj: $(TOP)/test/example.c $(TOP)/zbuild.h $(TOP)/zlib$(SUFFIX).h $(TOP)/deflate.h $(TOP)/test/test_shared_ng.h
minigzip.obj: $(TOP)/test/minigzip.c $(TOP)/zbuild.h $(TOP)/zlib$(SUFFIX).h
# cleanup
clean:
-del $(STATICLIB)
-del $(SHAREDLIB)
-del $(IMPLIB)
-del *.obj
-del *.res
-del *.exp
-del *.exe
-del *.pdb
-del *.manifest
distclean: clean
-del zconf$(SUFFIX).h
-del zlib$(SUFFIX).h
-del zlib_name_mangling$(SUFFIX).h
-del $(TOP)\win32\zlib.def
-del $(TOP)\win32\zlibcompat.def
-del $(TOP)\win32\zlib-ng.def
-del gzread.c
Reference in New Issue View Git Blame Copy Permalink