Skip to content

sse2neon -> SIMDe and added non-SIMD version #597

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 6 commits into from
Apr 26, 2020
Merged

sse2neon -> SIMDe and added non-SIMD version #597

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
9e6fdd3
Changed sse2neon with SIMDe. Added building non-SIMD version.
mbrcic Apr 20, 2020
c172df7
fix for Neon
mbrcic Apr 20, 2020
f47e8a5
SIMDe made optional. Include paths changes for SIMDe.
mbrcic Apr 20, 2020
3e16e4e
Added documentation entry for added functionality, simde and no_simd.
mbrcic Apr 20, 2020
2b3403f
fix for Neon after test.
mbrcic Apr 21, 2020
66db9da
changed preprocessor conditionals for SIMDe
mbrcic Apr 22, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions .gitmodules
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,3 @@
[submodule "lib/simde"]
path = lib/simde
url = https://github.com/nemequ/simde.git
38 changes: 28 additions & 10 deletions Makefile
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,26 +1,30 @@
CFLAGS= -g -Wall -O2 -Wc++-compat #-Wextra
CFLAGS= -g -Wall -O2 -Wc++-compat #-Wextra
Copy link
Contributor

@junaruga junaruga Apr 20, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Unnecessary difference. 1 space here.

Copy link
Contributor Author

@mbrcic mbrcic Apr 20, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I've removed the space in the following commit.

CPPFLAGS= -DHAVE_KALLOC
INCLUDES=
OBJS= kthread.o kalloc.o misc.o bseq.o sketch.o sdust.o options.o index.o chain.o align.o hit.o map.o format.o pe.o esterr.o splitidx.o ksw2_ll_sse.o
INCLUDES= -Ilib/simde/simde
Copy link
Contributor

@junaruga junaruga Apr 20, 2020 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This could be -Ilib/simde rather than -Ilib/simde/simde, modifying the loaded header paths. Because for example #include <simde/x86/sse2.h> is the right call by the simde system library (such as libsimde-dev in Debian) or simde-devel in Fedora and EPEL (CentOS).

Copy link
Contributor Author

@mbrcic mbrcic Apr 20, 2020

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I've changed to this in the following commit.

OBJS= kthread.o kalloc.o misc.o bseq.o sketch.o sdust.o options.o index.o chain.o align.o hit.o map.o format.o pe.o esterr.o splitidx.o
PROG= minimap2
PROG_EXTRA= sdust minimap2-lite
LIBS= -lm -lz -lpthread

ifeq ($(no_simd),) # if no_simd is not defined
ifeq ($(arm_neon),) # if arm_neon is not defined
OBJS+=ksw2_ll_sse.o
ifeq ($(sse2only),) # if sse2only is not defined
OBJS+=ksw2_extz2_sse41.o ksw2_extd2_sse41.o ksw2_exts2_sse41.o ksw2_extz2_sse2.o ksw2_extd2_sse2.o ksw2_exts2_sse2.o ksw2_dispatch.o
else # if sse2only is defined
OBJS+=ksw2_extz2_sse.o ksw2_extd2_sse.o ksw2_exts2_sse.o
endif
else # if arm_neon is defined
OBJS+=ksw2_extz2_neon.o ksw2_extd2_neon.o ksw2_exts2_neon.o
INCLUDES+=-Isse2neon
OBJS+=ksw2_ll_sse.o ksw2_extz2_neon.o ksw2_extd2_neon.o ksw2_exts2_neon.o
ifeq ($(aarch64),) #if aarch64 is not defined
CFLAGS+=-D_FILE_OFFSET_BITS=64 -mfpu=neon -fsigned-char
else #if aarch64 is defined
CFLAGS+=-D_FILE_OFFSET_BITS=64 -fsigned-char
endif
endif
else
OBJS+=ksw2_ll_nosimd.o ksw2_extz2_nosimd.o ksw2_extd2_nosimd.o ksw2_exts2_nosimd.o
endif

ifneq ($(asan),)
CFLAGS+=-fsanitize=address
Expand Down Expand Up @@ -56,10 +60,8 @@ sdust:sdust.c kalloc.o kalloc.h kdq.h kvec.h kseq.h ketopt.h sdust.h

# SSE-specific targets on x86/x86_64

ifeq ($(arm_neon),) # if arm_neon is defined, compile this target with the default setting (i.e. no -msse2)
ksw2_ll_sse.o:ksw2_ll_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) -msse2 $(CPPFLAGS) $(INCLUDES) $< -o $@
endif

ksw2_extz2_sse41.o:ksw2_extz2_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) -msse4.1 $(CPPFLAGS) -DKSW_CPU_DISPATCH $(INCLUDES) $< -o $@
Expand All @@ -83,15 +85,31 @@ ksw2_dispatch.o:ksw2_dispatch.c ksw2.h
$(CC) -c $(CFLAGS) -msse4.1 $(CPPFLAGS) -DKSW_CPU_DISPATCH $(INCLUDES) $< -o $@

# NEON-specific targets on ARM
ksw2_ll_neon.o:ksw2_ll_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) -DSIMDE_ENABLE_NATIVE_ALIASES $(CPPFLAGS) $(INCLUDES) $< -o $@

ksw2_extz2_neon.o:ksw2_extz2_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ $(INCLUDES) $< -o $@
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ -DSIMDE_ENABLE_NATIVE_ALIASES $(INCLUDES) $< -o $@

ksw2_extd2_neon.o:ksw2_extd2_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ $(INCLUDES) $< -o $@
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ -DSIMDE_ENABLE_NATIVE_ALIASES $(INCLUDES) $< -o $@

ksw2_exts2_neon.o:ksw2_exts2_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ $(INCLUDES) $< -o $@
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ -DSIMDE_ENABLE_NATIVE_ALIASES $(INCLUDES) $< -o $@

# no-SIMD version

ksw2_ll_nosimd.o:ksw2_ll_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DSIMDE_ENABLE_NATIVE_ALIASES -DSIMDE_NO_NATIVE $(INCLUDES) $< -o $@

ksw2_extz2_nosimd.o:ksw2_extz2_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ -DSIMDE_ENABLE_NATIVE_ALIASES -DSIMDE_NO_NATIVE $(INCLUDES) $< -o $@

ksw2_extd2_nosimd.o:ksw2_extd2_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ -DSIMDE_ENABLE_NATIVE_ALIASES -DSIMDE_NO_NATIVE $(INCLUDES) $< -o $@

ksw2_exts2_nosimd.o:ksw2_exts2_sse.c ksw2.h kalloc.h
$(CC) -c $(CFLAGS) $(CPPFLAGS) -DKSW_SSE2_ONLY -D__SSE2__ -DSIMDE_ENABLE_NATIVE_ALIASES -DSIMDE_NO_NATIVE $(INCLUDES) $< -o $@

# other non-file targets

Expand Down
6 changes: 4 additions & 2 deletions ksw2_extd2_sse.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,14 +4,16 @@
#include "ksw2.h"

#ifdef __SSE2__
#include <emmintrin.h>
//#include <emmintrin.h>
#include <x86/sse2.h>

#ifdef KSW_SSE2_ONLY
#undef __SSE4_1__
#endif

#ifdef __SSE4_1__
#include <smmintrin.h>
//#include <smmintrin.h>
#include <x86/sse4.1.h>
#endif

#ifdef KSW_CPU_DISPATCH
Expand Down
7 changes: 4 additions & 3 deletions ksw2_exts2_sse.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4,14 +4,15 @@
#include "ksw2.h"

#ifdef __SSE2__
#include <emmintrin.h>

//#include <emmintrin.h>
#include <x86/sse2.h>
#ifdef KSW_SSE2_ONLY
#undef __SSE4_1__
#endif

#ifdef __SSE4_1__
#include <smmintrin.h>
//#include <smmintrin.h>
#include <x86/sse4.1.h>
#endif

#ifdef KSW_CPU_DISPATCH
Expand Down
6 changes: 4 additions & 2 deletions ksw2_extz2_sse.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,14 +3,16 @@
#include "ksw2.h"

#ifdef __SSE2__
#include <emmintrin.h>
//#include <emmintrin.h>
#include <x86/sse2.h>

#ifdef KSW_SSE2_ONLY
#undef __SSE4_1__
#endif

#ifdef __SSE4_1__
#include <smmintrin.h>
//#include <smmintrin.h>
#include <x86/sse4.1.h>
#endif

#ifdef KSW_CPU_DISPATCH
Expand Down
3 changes: 2 additions & 1 deletion ksw2_ll_sse.c
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,7 +1,8 @@
#include <stdlib.h>
#include <stdint.h>
#include <string.h>
#include <emmintrin.h>
//#include <emmintrin.h>
#include <x86/sse2.h>
#include "ksw2.h"

#ifdef __GNUC__
Expand Down
1 change: 1 addition & 0 deletions lib/simde
Open in desktop
Submodule simde added at b30129