re PR libfortran/21468 (vectorizing libfortran)

2005-11-14  Janne Blomqvist  <jb@gcc.gnu.org>

        PR fortran/21468
        * Makefile.am: Add -ftree-vectorize for compiling matmul.
        * m4/matmul.m4: Add const and restrict to type declarations as
        appropriate.
        * m4/matmull.m4: Likewise.
        * Makefile.in: Regenerated.
        * generated/matmul_*.c: Likewise.

From-SVN: r106898
This commit is contained in:
Janne Blomqvist 2005-11-14 21:48:31 +02:00
parent d0568f737a
commit 8520690170
19 changed files with 283 additions and 246 deletions

View file

@ -1,3 +1,13 @@
2005-11-14 Janne Blomqvist <jb@gcc.gnu.org>
PR fortran/21468
* Makefile.am: Add -ftree-vectorize for compiling matmul.
* m4/matmul.m4: Add const and restrict to type declarations as
appropriate.
* m4/matmull.m4: Likewise.
* Makefile.in: Regenerated.
* generated/matmul_*.c: Likewise.
2005-11-13 Francois-Xavier Coudert <coudert@clipper.ens.fr>
* intrinsics/fget.c: New file.

View file

@ -433,7 +433,7 @@ m4_files= m4/iparm.m4 m4/ifunction.m4 m4/iforeach.m4 m4/all.m4 \
gfor_built_src= $(i_all_c) $(i_any_c) $(i_count_c) $(i_maxloc0_c) \
$(i_maxloc1_c) $(i_maxval_c) $(i_minloc0_c) $(i_minloc1_c) $(i_minval_c) \
$(i_product_c) $(i_sum_c) $(i_dotprod_c) $(i_dotprodl_c) $(i_dotprodc_c) \
$(i_matmul_c) $(i_matmull_c) $(i_transpose_c) $(i_shape_c) $(i_eoshift1_c) \
$(i_matmull_c) $(i_transpose_c) $(i_shape_c) $(i_eoshift1_c) \
$(i_eoshift3_c) $(i_cshift1_c) $(i_reshape_c) $(in_pack_c) $(in_unpack_c) \
$(i_exponent_c) $(i_fraction_c) $(i_nearest_c) $(i_set_exponent_c) \
$(i_pow_c) \
@ -573,11 +573,17 @@ $(gfor_built_specific2_src) \
intrinsics/dprod_r8.f90 \
intrinsics/f2c_specifics.F90
EXTRA_LTLIBRARIES = libmatmul.la
libmatmul_la_SOURCES = $(i_matmul_c)
libmatmul_la_CFLAGS = -ftree-vectorize $(AM_CFLAGS)
BUILT_SOURCES=$(gfor_built_src) $(gfor_built_specific_src) \
$(gfor_built_specific2_src)
$(gfor_built_specific2_src) $(libmatmul_la_SOURCES)
libgfortran_la_SOURCES = $(gfor_src) $(gfor_built_src) $(gfor_io_src) \
$(gfor_helper_src) $(gfor_io_headers) $(gfor_specific_src)
libgfortran_la_LIBADD = libmatmul.la
I_M4_DEPS=m4/iparm.m4
I_M4_DEPS0=$(I_M4_DEPS) m4/iforeach.m4
I_M4_DEPS1=$(I_M4_DEPS) m4/ifunction.m4

View file

@ -37,12 +37,12 @@ POST_UNINSTALL = :
build_triplet = @build@
host_triplet = @host@
target_triplet = @target@
DIST_COMMON = $(am__configure_deps) $(srcdir)/../config.guess \
$(srcdir)/../config.sub $(srcdir)/../install-sh \
$(srcdir)/../ltmain.sh $(srcdir)/../missing \
$(srcdir)/../mkinstalldirs $(srcdir)/Makefile.am \
$(srcdir)/Makefile.in $(srcdir)/config.h.in \
$(top_srcdir)/configure ChangeLog
DIST_COMMON = $(am__configure_deps) $(srcdir)/../compile \
$(srcdir)/../config.guess $(srcdir)/../config.sub \
$(srcdir)/../install-sh $(srcdir)/../ltmain.sh \
$(srcdir)/../missing $(srcdir)/../mkinstalldirs \
$(srcdir)/Makefile.am $(srcdir)/Makefile.in \
$(srcdir)/config.h.in $(top_srcdir)/configure ChangeLog
subdir = .
ACLOCAL_M4 = $(top_srcdir)/aclocal.m4
am__aclocal_m4_deps = $(top_srcdir)/../config/lead-dot.m4 \
@ -65,7 +65,7 @@ am__strip_dir = `echo $$p | sed -e 's|^.*/||'`;
am__installdirs = "$(DESTDIR)$(toolexeclibdir)"
toolexeclibLTLIBRARIES_INSTALL = $(INSTALL)
LTLIBRARIES = $(toolexeclib_LTLIBRARIES)
libgfortran_la_LIBADD =
libgfortran_la_DEPENDENCIES = libmatmul.la
am__objects_1 = compile_options.lo environ.lo error.lo fpu.lo main.lo \
memory.lo pause.lo stop.lo string.lo select.lo
am__objects_2 = all_l4.lo all_l8.lo all_l16.lo
@ -116,33 +116,30 @@ am__objects_13 = dotprod_i4.lo dotprod_i8.lo dotprod_i16.lo \
am__objects_14 = dotprod_l4.lo dotprod_l8.lo dotprod_l16.lo
am__objects_15 = dotprod_c4.lo dotprod_c8.lo dotprod_c10.lo \
dotprod_c16.lo
am__objects_16 = matmul_i4.lo matmul_i8.lo matmul_i16.lo matmul_r4.lo \
matmul_r8.lo matmul_r10.lo matmul_r16.lo matmul_c4.lo \
matmul_c8.lo matmul_c10.lo matmul_c16.lo
am__objects_17 = matmul_l4.lo matmul_l8.lo matmul_l16.lo
am__objects_18 = transpose_i4.lo transpose_i8.lo transpose_i16.lo \
am__objects_16 = matmul_l4.lo matmul_l8.lo matmul_l16.lo
am__objects_17 = transpose_i4.lo transpose_i8.lo transpose_i16.lo \
transpose_c4.lo transpose_c8.lo transpose_c10.lo \
transpose_c16.lo
am__objects_19 = shape_i4.lo shape_i8.lo shape_i16.lo
am__objects_20 = eoshift1_4.lo eoshift1_8.lo eoshift1_16.lo
am__objects_21 = eoshift3_4.lo eoshift3_8.lo eoshift3_16.lo
am__objects_22 = cshift1_4.lo cshift1_8.lo cshift1_16.lo
am__objects_23 = reshape_i4.lo reshape_i8.lo reshape_i16.lo \
am__objects_18 = shape_i4.lo shape_i8.lo shape_i16.lo
am__objects_19 = eoshift1_4.lo eoshift1_8.lo eoshift1_16.lo
am__objects_20 = eoshift3_4.lo eoshift3_8.lo eoshift3_16.lo
am__objects_21 = cshift1_4.lo cshift1_8.lo cshift1_16.lo
am__objects_22 = reshape_i4.lo reshape_i8.lo reshape_i16.lo \
reshape_c4.lo reshape_c8.lo reshape_c10.lo reshape_c16.lo
am__objects_24 = in_pack_i4.lo in_pack_i8.lo in_pack_i16.lo \
am__objects_23 = in_pack_i4.lo in_pack_i8.lo in_pack_i16.lo \
in_pack_c4.lo in_pack_c8.lo in_pack_c10.lo in_pack_c16.lo
am__objects_25 = in_unpack_i4.lo in_unpack_i8.lo in_unpack_i16.lo \
am__objects_24 = in_unpack_i4.lo in_unpack_i8.lo in_unpack_i16.lo \
in_unpack_c4.lo in_unpack_c8.lo in_unpack_c10.lo \
in_unpack_c16.lo
am__objects_26 = exponent_r4.lo exponent_r8.lo exponent_r10.lo \
am__objects_25 = exponent_r4.lo exponent_r8.lo exponent_r10.lo \
exponent_r16.lo
am__objects_27 = fraction_r4.lo fraction_r8.lo fraction_r10.lo \
am__objects_26 = fraction_r4.lo fraction_r8.lo fraction_r10.lo \
fraction_r16.lo
am__objects_28 = nearest_r4.lo nearest_r8.lo nearest_r10.lo \
am__objects_27 = nearest_r4.lo nearest_r8.lo nearest_r10.lo \
nearest_r16.lo
am__objects_29 = set_exponent_r4.lo set_exponent_r8.lo \
am__objects_28 = set_exponent_r4.lo set_exponent_r8.lo \
set_exponent_r10.lo set_exponent_r16.lo
am__objects_30 = pow_i4_i4.lo pow_i8_i4.lo pow_i16_i4.lo pow_r4_i4.lo \
am__objects_29 = pow_i4_i4.lo pow_i8_i4.lo pow_i16_i4.lo pow_r4_i4.lo \
pow_r8_i4.lo pow_r10_i4.lo pow_r16_i4.lo pow_c4_i4.lo \
pow_c8_i4.lo pow_c10_i4.lo pow_c16_i4.lo pow_i4_i8.lo \
pow_i8_i8.lo pow_i16_i8.lo pow_r4_i8.lo pow_r8_i8.lo \
@ -151,7 +148,7 @@ am__objects_30 = pow_i4_i4.lo pow_i8_i4.lo pow_i16_i4.lo pow_r4_i4.lo \
pow_i16_i16.lo pow_r4_i16.lo pow_r8_i16.lo pow_r10_i16.lo \
pow_r16_i16.lo pow_c4_i16.lo pow_c8_i16.lo pow_c10_i16.lo \
pow_c16_i16.lo
am__objects_31 = $(am__objects_2) $(am__objects_3) $(am__objects_4) \
am__objects_30 = $(am__objects_2) $(am__objects_3) $(am__objects_4) \
$(am__objects_5) $(am__objects_6) $(am__objects_7) \
$(am__objects_8) $(am__objects_9) $(am__objects_10) \
$(am__objects_11) $(am__objects_12) $(am__objects_13) \
@ -160,11 +157,11 @@ am__objects_31 = $(am__objects_2) $(am__objects_3) $(am__objects_4) \
$(am__objects_20) $(am__objects_21) $(am__objects_22) \
$(am__objects_23) $(am__objects_24) $(am__objects_25) \
$(am__objects_26) $(am__objects_27) $(am__objects_28) \
$(am__objects_29) $(am__objects_30)
am__objects_32 = close.lo file_pos.lo format.lo inquire.lo \
$(am__objects_29)
am__objects_31 = close.lo file_pos.lo format.lo inquire.lo \
list_read.lo lock.lo open.lo read.lo size_from_kind.lo \
transfer.lo unit.lo unix.lo write.lo
am__objects_33 = associated.lo abort.lo args.lo bessel.lo \
am__objects_32 = associated.lo abort.lo args.lo bessel.lo \
c99_functions.lo chdir.lo cpu_time.lo cshift0.lo ctime.lo \
date_and_time.lo env.lo erf.lo eoshift0.lo eoshift2.lo \
etime.lo exit.lo fget.lo flush.lo fnum.lo ftell.lo gerror.lo \
@ -177,8 +174,8 @@ am__objects_33 = associated.lo abort.lo args.lo bessel.lo \
system_clock.lo time.lo transpose_generic.lo tty.lo umask.lo \
unlink.lo unpack_generic.lo in_pack_generic.lo \
in_unpack_generic.lo normalize.lo
am__objects_34 =
am__objects_35 = _abs_c4.lo _abs_c8.lo _abs_c10.lo _abs_c16.lo \
am__objects_33 =
am__objects_34 = _abs_c4.lo _abs_c8.lo _abs_c10.lo _abs_c16.lo \
_abs_i4.lo _abs_i8.lo _abs_i16.lo _abs_r4.lo _abs_r8.lo \
_abs_r10.lo _abs_r16.lo _exp_r4.lo _exp_r8.lo _exp_r10.lo \
_exp_r16.lo _exp_c4.lo _exp_c8.lo _exp_c10.lo _exp_c16.lo \
@ -198,20 +195,29 @@ am__objects_35 = _abs_c4.lo _abs_c8.lo _abs_c10.lo _abs_c16.lo \
_conjg_c4.lo _conjg_c8.lo _conjg_c10.lo _conjg_c16.lo \
_aint_r4.lo _aint_r8.lo _aint_r10.lo _aint_r16.lo _anint_r4.lo \
_anint_r8.lo _anint_r10.lo _anint_r16.lo
am__objects_36 = _sign_i4.lo _sign_i8.lo _sign_i16.lo _sign_r4.lo \
am__objects_35 = _sign_i4.lo _sign_i8.lo _sign_i16.lo _sign_r4.lo \
_sign_r8.lo _sign_r10.lo _sign_r16.lo _dim_i4.lo _dim_i8.lo \
_dim_i16.lo _dim_r4.lo _dim_r8.lo _dim_r10.lo _dim_r16.lo \
_atan2_r4.lo _atan2_r8.lo _atan2_r10.lo _atan2_r16.lo \
_mod_i4.lo _mod_i8.lo _mod_i16.lo _mod_r4.lo _mod_r8.lo
am__objects_37 = $(am__objects_35) $(am__objects_36) dprod_r8.lo \
am__objects_36 = $(am__objects_34) $(am__objects_35) dprod_r8.lo \
f2c_specifics.lo
am_libgfortran_la_OBJECTS = $(am__objects_1) $(am__objects_31) \
$(am__objects_32) $(am__objects_33) $(am__objects_34) \
$(am__objects_37)
am_libgfortran_la_OBJECTS = $(am__objects_1) $(am__objects_30) \
$(am__objects_31) $(am__objects_32) $(am__objects_33) \
$(am__objects_36)
libgfortran_la_OBJECTS = $(am_libgfortran_la_OBJECTS)
libgfortranbegin_la_LIBADD =
am_libgfortranbegin_la_OBJECTS = fmain.lo
libgfortranbegin_la_OBJECTS = $(am_libgfortranbegin_la_OBJECTS)
libmatmul_la_LIBADD =
am__objects_37 = libmatmul_la-matmul_i4.lo libmatmul_la-matmul_i8.lo \
libmatmul_la-matmul_i16.lo libmatmul_la-matmul_r4.lo \
libmatmul_la-matmul_r8.lo libmatmul_la-matmul_r10.lo \
libmatmul_la-matmul_r16.lo libmatmul_la-matmul_c4.lo \
libmatmul_la-matmul_c8.lo libmatmul_la-matmul_c10.lo \
libmatmul_la-matmul_c16.lo
am_libmatmul_la_OBJECTS = $(am__objects_37)
libmatmul_la_OBJECTS = $(am_libmatmul_la_OBJECTS)
DEFAULT_INCLUDES = -I. -I$(srcdir) -I.
depcomp =
am__depfiles_maybe =
@ -233,9 +239,10 @@ LINK = $(LIBTOOL) --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) \
$(AM_LDFLAGS) $(LDFLAGS) -o $@
FCCOMPILE = $(FC) $(AM_FCFLAGS) $(FCFLAGS)
LTFCCOMPILE = $(LIBTOOL) --mode=compile $(FC) $(AM_FCFLAGS) $(FCFLAGS)
SOURCES = $(libgfortran_la_SOURCES) $(libgfortranbegin_la_SOURCES)
SOURCES = $(libgfortran_la_SOURCES) $(libgfortranbegin_la_SOURCES) \
$(libmatmul_la_SOURCES)
DIST_SOURCES = $(libgfortran_la_SOURCES) \
$(libgfortranbegin_la_SOURCES)
$(libgfortranbegin_la_SOURCES) $(libmatmul_la_SOURCES)
MULTISRCTOP =
MULTIBUILDTOP =
MULTIDIRS =
@ -774,7 +781,7 @@ m4_files = m4/iparm.m4 m4/ifunction.m4 m4/iforeach.m4 m4/all.m4 \
gfor_built_src = $(i_all_c) $(i_any_c) $(i_count_c) $(i_maxloc0_c) \
$(i_maxloc1_c) $(i_maxval_c) $(i_minloc0_c) $(i_minloc1_c) $(i_minval_c) \
$(i_product_c) $(i_sum_c) $(i_dotprod_c) $(i_dotprodl_c) $(i_dotprodc_c) \
$(i_matmul_c) $(i_matmull_c) $(i_transpose_c) $(i_shape_c) $(i_eoshift1_c) \
$(i_matmull_c) $(i_transpose_c) $(i_shape_c) $(i_eoshift1_c) \
$(i_eoshift3_c) $(i_cshift1_c) $(i_reshape_c) $(in_pack_c) $(in_unpack_c) \
$(i_exponent_c) $(i_fraction_c) $(i_nearest_c) $(i_set_exponent_c) \
$(i_pow_c) \
@ -915,12 +922,16 @@ $(gfor_built_specific2_src) \
intrinsics/dprod_r8.f90 \
intrinsics/f2c_specifics.F90
EXTRA_LTLIBRARIES = libmatmul.la
libmatmul_la_SOURCES = $(i_matmul_c)
libmatmul_la_CFLAGS = -ftree-vectorize $(AM_CFLAGS)
BUILT_SOURCES = $(gfor_built_src) $(gfor_built_specific_src) \
$(gfor_built_specific2_src)
$(gfor_built_specific2_src) $(libmatmul_la_SOURCES)
libgfortran_la_SOURCES = $(gfor_src) $(gfor_built_src) $(gfor_io_src) \
$(gfor_helper_src) $(gfor_io_headers) $(gfor_specific_src)
libgfortran_la_LIBADD = libmatmul.la
I_M4_DEPS = m4/iparm.m4
I_M4_DEPS0 = $(I_M4_DEPS) m4/iforeach.m4
I_M4_DEPS1 = $(I_M4_DEPS) m4/ifunction.m4
@ -1011,6 +1022,8 @@ libgfortran.la: $(libgfortran_la_OBJECTS) $(libgfortran_la_DEPENDENCIES)
$(FCLINK) -rpath $(toolexeclibdir) $(libgfortran_la_LDFLAGS) $(libgfortran_la_OBJECTS) $(libgfortran_la_LIBADD) $(LIBS)
libgfortranbegin.la: $(libgfortranbegin_la_OBJECTS) $(libgfortranbegin_la_DEPENDENCIES)
$(LINK) -rpath $(toolexeclibdir) $(libgfortranbegin_la_LDFLAGS) $(libgfortranbegin_la_OBJECTS) $(libgfortranbegin_la_LIBADD) $(LIBS)
libmatmul.la: $(libmatmul_la_OBJECTS) $(libmatmul_la_DEPENDENCIES)
$(LINK) $(libmatmul_la_LDFLAGS) $(libmatmul_la_OBJECTS) $(libmatmul_la_LIBADD) $(LIBS)
mostlyclean-compile:
-rm -f *.$(OBJEXT)
@ -1870,39 +1883,6 @@ dotprod_c10.lo: generated/dotprod_c10.c
dotprod_c16.lo: generated/dotprod_c16.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o dotprod_c16.lo `test -f 'generated/dotprod_c16.c' || echo '$(srcdir)/'`generated/dotprod_c16.c
matmul_i4.lo: generated/matmul_i4.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_i4.lo `test -f 'generated/matmul_i4.c' || echo '$(srcdir)/'`generated/matmul_i4.c
matmul_i8.lo: generated/matmul_i8.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_i8.lo `test -f 'generated/matmul_i8.c' || echo '$(srcdir)/'`generated/matmul_i8.c
matmul_i16.lo: generated/matmul_i16.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_i16.lo `test -f 'generated/matmul_i16.c' || echo '$(srcdir)/'`generated/matmul_i16.c
matmul_r4.lo: generated/matmul_r4.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_r4.lo `test -f 'generated/matmul_r4.c' || echo '$(srcdir)/'`generated/matmul_r4.c
matmul_r8.lo: generated/matmul_r8.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_r8.lo `test -f 'generated/matmul_r8.c' || echo '$(srcdir)/'`generated/matmul_r8.c
matmul_r10.lo: generated/matmul_r10.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_r10.lo `test -f 'generated/matmul_r10.c' || echo '$(srcdir)/'`generated/matmul_r10.c
matmul_r16.lo: generated/matmul_r16.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_r16.lo `test -f 'generated/matmul_r16.c' || echo '$(srcdir)/'`generated/matmul_r16.c
matmul_c4.lo: generated/matmul_c4.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_c4.lo `test -f 'generated/matmul_c4.c' || echo '$(srcdir)/'`generated/matmul_c4.c
matmul_c8.lo: generated/matmul_c8.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_c8.lo `test -f 'generated/matmul_c8.c' || echo '$(srcdir)/'`generated/matmul_c8.c
matmul_c10.lo: generated/matmul_c10.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_c10.lo `test -f 'generated/matmul_c10.c' || echo '$(srcdir)/'`generated/matmul_c10.c
matmul_c16.lo: generated/matmul_c16.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_c16.lo `test -f 'generated/matmul_c16.c' || echo '$(srcdir)/'`generated/matmul_c16.c
matmul_l4.lo: generated/matmul_l4.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o matmul_l4.lo `test -f 'generated/matmul_l4.c' || echo '$(srcdir)/'`generated/matmul_l4.c
@ -2389,6 +2369,39 @@ in_unpack_generic.lo: runtime/in_unpack_generic.c
normalize.lo: runtime/normalize.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) -c -o normalize.lo `test -f 'runtime/normalize.c' || echo '$(srcdir)/'`runtime/normalize.c
libmatmul_la-matmul_i4.lo: generated/matmul_i4.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_i4.lo `test -f 'generated/matmul_i4.c' || echo '$(srcdir)/'`generated/matmul_i4.c
libmatmul_la-matmul_i8.lo: generated/matmul_i8.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_i8.lo `test -f 'generated/matmul_i8.c' || echo '$(srcdir)/'`generated/matmul_i8.c
libmatmul_la-matmul_i16.lo: generated/matmul_i16.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_i16.lo `test -f 'generated/matmul_i16.c' || echo '$(srcdir)/'`generated/matmul_i16.c
libmatmul_la-matmul_r4.lo: generated/matmul_r4.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_r4.lo `test -f 'generated/matmul_r4.c' || echo '$(srcdir)/'`generated/matmul_r4.c
libmatmul_la-matmul_r8.lo: generated/matmul_r8.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_r8.lo `test -f 'generated/matmul_r8.c' || echo '$(srcdir)/'`generated/matmul_r8.c
libmatmul_la-matmul_r10.lo: generated/matmul_r10.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_r10.lo `test -f 'generated/matmul_r10.c' || echo '$(srcdir)/'`generated/matmul_r10.c
libmatmul_la-matmul_r16.lo: generated/matmul_r16.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_r16.lo `test -f 'generated/matmul_r16.c' || echo '$(srcdir)/'`generated/matmul_r16.c
libmatmul_la-matmul_c4.lo: generated/matmul_c4.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_c4.lo `test -f 'generated/matmul_c4.c' || echo '$(srcdir)/'`generated/matmul_c4.c
libmatmul_la-matmul_c8.lo: generated/matmul_c8.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_c8.lo `test -f 'generated/matmul_c8.c' || echo '$(srcdir)/'`generated/matmul_c8.c
libmatmul_la-matmul_c10.lo: generated/matmul_c10.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_c10.lo `test -f 'generated/matmul_c10.c' || echo '$(srcdir)/'`generated/matmul_c10.c
libmatmul_la-matmul_c16.lo: generated/matmul_c16.c
$(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(libmatmul_la_CFLAGS) $(CFLAGS) -c -o libmatmul_la-matmul_c16.lo `test -f 'generated/matmul_c16.c' || echo '$(srcdir)/'`generated/matmul_c16.c
.f90.o:
$(FCCOMPILE) -c -o $@ $<

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_c10 (gfc_array_c10 * retarray, gfc_array_c10 * a, gfc_array_c10 * b);
extern void matmul_c10 (gfc_array_c10 * const restrict retarray,
gfc_array_c10 * const restrict a, gfc_array_c10 * const restrict b);
export_proto(matmul_c10);
void
matmul_c10 (gfc_array_c10 * retarray, gfc_array_c10 * a, gfc_array_c10 * b)
matmul_c10 (gfc_array_c10 * const restrict retarray,
gfc_array_c10 * const restrict a, gfc_array_c10 * const restrict b)
{
GFC_COMPLEX_10 *abase;
GFC_COMPLEX_10 *bbase;
GFC_COMPLEX_10 *dest;
const GFC_COMPLEX_10 * restrict abase;
const GFC_COMPLEX_10 * restrict bbase;
GFC_COMPLEX_10 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_c10 (gfc_array_c10 * retarray, gfc_array_c10 * a, gfc_array_c10 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_c10 (gfc_array_c10 * retarray, gfc_array_c10 * a, gfc_array_c10 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_COMPLEX_10 *bbase_y;
GFC_COMPLEX_10 *dest_y;
GFC_COMPLEX_10 *abase_n;
const GFC_COMPLEX_10 * restrict bbase_y;
GFC_COMPLEX_10 * restrict dest_y;
const GFC_COMPLEX_10 * restrict abase_n;
GFC_COMPLEX_10 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_c16 (gfc_array_c16 * retarray, gfc_array_c16 * a, gfc_array_c16 * b);
extern void matmul_c16 (gfc_array_c16 * const restrict retarray,
gfc_array_c16 * const restrict a, gfc_array_c16 * const restrict b);
export_proto(matmul_c16);
void
matmul_c16 (gfc_array_c16 * retarray, gfc_array_c16 * a, gfc_array_c16 * b)
matmul_c16 (gfc_array_c16 * const restrict retarray,
gfc_array_c16 * const restrict a, gfc_array_c16 * const restrict b)
{
GFC_COMPLEX_16 *abase;
GFC_COMPLEX_16 *bbase;
GFC_COMPLEX_16 *dest;
const GFC_COMPLEX_16 * restrict abase;
const GFC_COMPLEX_16 * restrict bbase;
GFC_COMPLEX_16 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_c16 (gfc_array_c16 * retarray, gfc_array_c16 * a, gfc_array_c16 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_c16 (gfc_array_c16 * retarray, gfc_array_c16 * a, gfc_array_c16 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_COMPLEX_16 *bbase_y;
GFC_COMPLEX_16 *dest_y;
GFC_COMPLEX_16 *abase_n;
const GFC_COMPLEX_16 * restrict bbase_y;
GFC_COMPLEX_16 * restrict dest_y;
const GFC_COMPLEX_16 * restrict abase_n;
GFC_COMPLEX_16 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_c4 (gfc_array_c4 * retarray, gfc_array_c4 * a, gfc_array_c4 * b);
extern void matmul_c4 (gfc_array_c4 * const restrict retarray,
gfc_array_c4 * const restrict a, gfc_array_c4 * const restrict b);
export_proto(matmul_c4);
void
matmul_c4 (gfc_array_c4 * retarray, gfc_array_c4 * a, gfc_array_c4 * b)
matmul_c4 (gfc_array_c4 * const restrict retarray,
gfc_array_c4 * const restrict a, gfc_array_c4 * const restrict b)
{
GFC_COMPLEX_4 *abase;
GFC_COMPLEX_4 *bbase;
GFC_COMPLEX_4 *dest;
const GFC_COMPLEX_4 * restrict abase;
const GFC_COMPLEX_4 * restrict bbase;
GFC_COMPLEX_4 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_c4 (gfc_array_c4 * retarray, gfc_array_c4 * a, gfc_array_c4 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_c4 (gfc_array_c4 * retarray, gfc_array_c4 * a, gfc_array_c4 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_COMPLEX_4 *bbase_y;
GFC_COMPLEX_4 *dest_y;
GFC_COMPLEX_4 *abase_n;
const GFC_COMPLEX_4 * restrict bbase_y;
GFC_COMPLEX_4 * restrict dest_y;
const GFC_COMPLEX_4 * restrict abase_n;
GFC_COMPLEX_4 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_c8 (gfc_array_c8 * retarray, gfc_array_c8 * a, gfc_array_c8 * b);
extern void matmul_c8 (gfc_array_c8 * const restrict retarray,
gfc_array_c8 * const restrict a, gfc_array_c8 * const restrict b);
export_proto(matmul_c8);
void
matmul_c8 (gfc_array_c8 * retarray, gfc_array_c8 * a, gfc_array_c8 * b)
matmul_c8 (gfc_array_c8 * const restrict retarray,
gfc_array_c8 * const restrict a, gfc_array_c8 * const restrict b)
{
GFC_COMPLEX_8 *abase;
GFC_COMPLEX_8 *bbase;
GFC_COMPLEX_8 *dest;
const GFC_COMPLEX_8 * restrict abase;
const GFC_COMPLEX_8 * restrict bbase;
GFC_COMPLEX_8 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_c8 (gfc_array_c8 * retarray, gfc_array_c8 * a, gfc_array_c8 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_c8 (gfc_array_c8 * retarray, gfc_array_c8 * a, gfc_array_c8 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_COMPLEX_8 *bbase_y;
GFC_COMPLEX_8 *dest_y;
GFC_COMPLEX_8 *abase_n;
const GFC_COMPLEX_8 * restrict bbase_y;
GFC_COMPLEX_8 * restrict dest_y;
const GFC_COMPLEX_8 * restrict abase_n;
GFC_COMPLEX_8 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_i16 (gfc_array_i16 * retarray, gfc_array_i16 * a, gfc_array_i16 * b);
extern void matmul_i16 (gfc_array_i16 * const restrict retarray,
gfc_array_i16 * const restrict a, gfc_array_i16 * const restrict b);
export_proto(matmul_i16);
void
matmul_i16 (gfc_array_i16 * retarray, gfc_array_i16 * a, gfc_array_i16 * b)
matmul_i16 (gfc_array_i16 * const restrict retarray,
gfc_array_i16 * const restrict a, gfc_array_i16 * const restrict b)
{
GFC_INTEGER_16 *abase;
GFC_INTEGER_16 *bbase;
GFC_INTEGER_16 *dest;
const GFC_INTEGER_16 * restrict abase;
const GFC_INTEGER_16 * restrict bbase;
GFC_INTEGER_16 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_i16 (gfc_array_i16 * retarray, gfc_array_i16 * a, gfc_array_i16 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_i16 (gfc_array_i16 * retarray, gfc_array_i16 * a, gfc_array_i16 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_INTEGER_16 *bbase_y;
GFC_INTEGER_16 *dest_y;
GFC_INTEGER_16 *abase_n;
const GFC_INTEGER_16 * restrict bbase_y;
GFC_INTEGER_16 * restrict dest_y;
const GFC_INTEGER_16 * restrict abase_n;
GFC_INTEGER_16 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_i4 (gfc_array_i4 * retarray, gfc_array_i4 * a, gfc_array_i4 * b);
extern void matmul_i4 (gfc_array_i4 * const restrict retarray,
gfc_array_i4 * const restrict a, gfc_array_i4 * const restrict b);
export_proto(matmul_i4);
void
matmul_i4 (gfc_array_i4 * retarray, gfc_array_i4 * a, gfc_array_i4 * b)
matmul_i4 (gfc_array_i4 * const restrict retarray,
gfc_array_i4 * const restrict a, gfc_array_i4 * const restrict b)
{
GFC_INTEGER_4 *abase;
GFC_INTEGER_4 *bbase;
GFC_INTEGER_4 *dest;
const GFC_INTEGER_4 * restrict abase;
const GFC_INTEGER_4 * restrict bbase;
GFC_INTEGER_4 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_i4 (gfc_array_i4 * retarray, gfc_array_i4 * a, gfc_array_i4 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_i4 (gfc_array_i4 * retarray, gfc_array_i4 * a, gfc_array_i4 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_INTEGER_4 *bbase_y;
GFC_INTEGER_4 *dest_y;
GFC_INTEGER_4 *abase_n;
const GFC_INTEGER_4 * restrict bbase_y;
GFC_INTEGER_4 * restrict dest_y;
const GFC_INTEGER_4 * restrict abase_n;
GFC_INTEGER_4 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_i8 (gfc_array_i8 * retarray, gfc_array_i8 * a, gfc_array_i8 * b);
extern void matmul_i8 (gfc_array_i8 * const restrict retarray,
gfc_array_i8 * const restrict a, gfc_array_i8 * const restrict b);
export_proto(matmul_i8);
void
matmul_i8 (gfc_array_i8 * retarray, gfc_array_i8 * a, gfc_array_i8 * b)
matmul_i8 (gfc_array_i8 * const restrict retarray,
gfc_array_i8 * const restrict a, gfc_array_i8 * const restrict b)
{
GFC_INTEGER_8 *abase;
GFC_INTEGER_8 *bbase;
GFC_INTEGER_8 *dest;
const GFC_INTEGER_8 * restrict abase;
const GFC_INTEGER_8 * restrict bbase;
GFC_INTEGER_8 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_i8 (gfc_array_i8 * retarray, gfc_array_i8 * a, gfc_array_i8 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_i8 (gfc_array_i8 * retarray, gfc_array_i8 * a, gfc_array_i8 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_INTEGER_8 *bbase_y;
GFC_INTEGER_8 *dest_y;
GFC_INTEGER_8 *abase_n;
const GFC_INTEGER_8 * restrict bbase_y;
GFC_INTEGER_8 * restrict dest_y;
const GFC_INTEGER_8 * restrict abase_n;
GFC_INTEGER_8 bbase_yn;
if (rystride == ycount)

View file

@ -38,15 +38,17 @@ Boston, MA 02110-1301, USA. */
/* Dimensions: retarray(x,y) a(x, count) b(count,y).
Either a or b can be rank 1. In this case x or y is 1. */
extern void matmul_l16 (gfc_array_l16 *, gfc_array_l4 *, gfc_array_l4 *);
extern void matmul_l16 (gfc_array_l16 * const restrict,
gfc_array_l4 * const restrict, gfc_array_l4 * const restrict);
export_proto(matmul_l16);
void
matmul_l16 (gfc_array_l16 * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
matmul_l16 (gfc_array_l16 * const restrict retarray,
gfc_array_l4 * const restrict a, gfc_array_l4 * const restrict b)
{
GFC_INTEGER_4 *abase;
GFC_INTEGER_4 *bbase;
GFC_LOGICAL_16 *dest;
const GFC_INTEGER_4 * restrict abase;
const GFC_INTEGER_4 * restrict bbase;
GFC_LOGICAL_16 * restrict dest;
index_type rxstride;
index_type rystride;
index_type xcount;
@ -56,8 +58,8 @@ matmul_l16 (gfc_array_l16 * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
index_type x;
index_type y;
GFC_INTEGER_4 *pa;
GFC_INTEGER_4 *pb;
const GFC_INTEGER_4 * restrict pa;
const GFC_INTEGER_4 * restrict pb;
index_type astride;
index_type bstride;
index_type count;

View file

@ -38,15 +38,17 @@ Boston, MA 02110-1301, USA. */
/* Dimensions: retarray(x,y) a(x, count) b(count,y).
Either a or b can be rank 1. In this case x or y is 1. */
extern void matmul_l4 (gfc_array_l4 *, gfc_array_l4 *, gfc_array_l4 *);
extern void matmul_l4 (gfc_array_l4 * const restrict,
gfc_array_l4 * const restrict, gfc_array_l4 * const restrict);
export_proto(matmul_l4);
void
matmul_l4 (gfc_array_l4 * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
matmul_l4 (gfc_array_l4 * const restrict retarray,
gfc_array_l4 * const restrict a, gfc_array_l4 * const restrict b)
{
GFC_INTEGER_4 *abase;
GFC_INTEGER_4 *bbase;
GFC_LOGICAL_4 *dest;
const GFC_INTEGER_4 * restrict abase;
const GFC_INTEGER_4 * restrict bbase;
GFC_LOGICAL_4 * restrict dest;
index_type rxstride;
index_type rystride;
index_type xcount;
@ -56,8 +58,8 @@ matmul_l4 (gfc_array_l4 * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
index_type x;
index_type y;
GFC_INTEGER_4 *pa;
GFC_INTEGER_4 *pb;
const GFC_INTEGER_4 * restrict pa;
const GFC_INTEGER_4 * restrict pb;
index_type astride;
index_type bstride;
index_type count;

View file

@ -38,15 +38,17 @@ Boston, MA 02110-1301, USA. */
/* Dimensions: retarray(x,y) a(x, count) b(count,y).
Either a or b can be rank 1. In this case x or y is 1. */
extern void matmul_l8 (gfc_array_l8 *, gfc_array_l4 *, gfc_array_l4 *);
extern void matmul_l8 (gfc_array_l8 * const restrict,
gfc_array_l4 * const restrict, gfc_array_l4 * const restrict);
export_proto(matmul_l8);
void
matmul_l8 (gfc_array_l8 * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
matmul_l8 (gfc_array_l8 * const restrict retarray,
gfc_array_l4 * const restrict a, gfc_array_l4 * const restrict b)
{
GFC_INTEGER_4 *abase;
GFC_INTEGER_4 *bbase;
GFC_LOGICAL_8 *dest;
const GFC_INTEGER_4 * restrict abase;
const GFC_INTEGER_4 * restrict bbase;
GFC_LOGICAL_8 * restrict dest;
index_type rxstride;
index_type rystride;
index_type xcount;
@ -56,8 +58,8 @@ matmul_l8 (gfc_array_l8 * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
index_type x;
index_type y;
GFC_INTEGER_4 *pa;
GFC_INTEGER_4 *pb;
const GFC_INTEGER_4 * restrict pa;
const GFC_INTEGER_4 * restrict pb;
index_type astride;
index_type bstride;
index_type count;

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_r10 (gfc_array_r10 * retarray, gfc_array_r10 * a, gfc_array_r10 * b);
extern void matmul_r10 (gfc_array_r10 * const restrict retarray,
gfc_array_r10 * const restrict a, gfc_array_r10 * const restrict b);
export_proto(matmul_r10);
void
matmul_r10 (gfc_array_r10 * retarray, gfc_array_r10 * a, gfc_array_r10 * b)
matmul_r10 (gfc_array_r10 * const restrict retarray,
gfc_array_r10 * const restrict a, gfc_array_r10 * const restrict b)
{
GFC_REAL_10 *abase;
GFC_REAL_10 *bbase;
GFC_REAL_10 *dest;
const GFC_REAL_10 * restrict abase;
const GFC_REAL_10 * restrict bbase;
GFC_REAL_10 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_r10 (gfc_array_r10 * retarray, gfc_array_r10 * a, gfc_array_r10 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_r10 (gfc_array_r10 * retarray, gfc_array_r10 * a, gfc_array_r10 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_REAL_10 *bbase_y;
GFC_REAL_10 *dest_y;
GFC_REAL_10 *abase_n;
const GFC_REAL_10 * restrict bbase_y;
GFC_REAL_10 * restrict dest_y;
const GFC_REAL_10 * restrict abase_n;
GFC_REAL_10 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_r16 (gfc_array_r16 * retarray, gfc_array_r16 * a, gfc_array_r16 * b);
extern void matmul_r16 (gfc_array_r16 * const restrict retarray,
gfc_array_r16 * const restrict a, gfc_array_r16 * const restrict b);
export_proto(matmul_r16);
void
matmul_r16 (gfc_array_r16 * retarray, gfc_array_r16 * a, gfc_array_r16 * b)
matmul_r16 (gfc_array_r16 * const restrict retarray,
gfc_array_r16 * const restrict a, gfc_array_r16 * const restrict b)
{
GFC_REAL_16 *abase;
GFC_REAL_16 *bbase;
GFC_REAL_16 *dest;
const GFC_REAL_16 * restrict abase;
const GFC_REAL_16 * restrict bbase;
GFC_REAL_16 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_r16 (gfc_array_r16 * retarray, gfc_array_r16 * a, gfc_array_r16 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_r16 (gfc_array_r16 * retarray, gfc_array_r16 * a, gfc_array_r16 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_REAL_16 *bbase_y;
GFC_REAL_16 *dest_y;
GFC_REAL_16 *abase_n;
const GFC_REAL_16 * restrict bbase_y;
GFC_REAL_16 * restrict dest_y;
const GFC_REAL_16 * restrict abase_n;
GFC_REAL_16 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_r4 (gfc_array_r4 * retarray, gfc_array_r4 * a, gfc_array_r4 * b);
extern void matmul_r4 (gfc_array_r4 * const restrict retarray,
gfc_array_r4 * const restrict a, gfc_array_r4 * const restrict b);
export_proto(matmul_r4);
void
matmul_r4 (gfc_array_r4 * retarray, gfc_array_r4 * a, gfc_array_r4 * b)
matmul_r4 (gfc_array_r4 * const restrict retarray,
gfc_array_r4 * const restrict a, gfc_array_r4 * const restrict b)
{
GFC_REAL_4 *abase;
GFC_REAL_4 *bbase;
GFC_REAL_4 *dest;
const GFC_REAL_4 * restrict abase;
const GFC_REAL_4 * restrict bbase;
GFC_REAL_4 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_r4 (gfc_array_r4 * retarray, gfc_array_r4 * a, gfc_array_r4 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_r4 (gfc_array_r4 * retarray, gfc_array_r4 * a, gfc_array_r4 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_REAL_4 *bbase_y;
GFC_REAL_4 *dest_y;
GFC_REAL_4 *abase_n;
const GFC_REAL_4 * restrict bbase_y;
GFC_REAL_4 * restrict dest_y;
const GFC_REAL_4 * restrict abase_n;
GFC_REAL_4 bbase_yn;
if (rystride == ycount)

View file

@ -48,15 +48,17 @@ Boston, MA 02110-1301, USA. */
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_r8 (gfc_array_r8 * retarray, gfc_array_r8 * a, gfc_array_r8 * b);
extern void matmul_r8 (gfc_array_r8 * const restrict retarray,
gfc_array_r8 * const restrict a, gfc_array_r8 * const restrict b);
export_proto(matmul_r8);
void
matmul_r8 (gfc_array_r8 * retarray, gfc_array_r8 * a, gfc_array_r8 * b)
matmul_r8 (gfc_array_r8 * const restrict retarray,
gfc_array_r8 * const restrict a, gfc_array_r8 * const restrict b)
{
GFC_REAL_8 *abase;
GFC_REAL_8 *bbase;
GFC_REAL_8 *dest;
const GFC_REAL_8 * restrict abase;
const GFC_REAL_8 * restrict bbase;
GFC_REAL_8 * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -105,12 +107,10 @@ matmul_r8 (gfc_array_r8 * retarray, gfc_array_r8 * a, gfc_array_r8 * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -175,9 +175,9 @@ matmul_r8 (gfc_array_r8 * retarray, gfc_array_r8 * a, gfc_array_r8 * b)
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
GFC_REAL_8 *bbase_y;
GFC_REAL_8 *dest_y;
GFC_REAL_8 *abase_n;
const GFC_REAL_8 * restrict bbase_y;
GFC_REAL_8 * restrict dest_y;
const GFC_REAL_8 * restrict abase_n;
GFC_REAL_8 bbase_yn;
if (rystride == ycount)

View file

@ -49,15 +49,17 @@ include(iparm.m4)dnl
C(I,J) = C(I,J)+A(I,K)*B(K,J)
*/
extern void matmul_`'rtype_code (rtype * retarray, rtype * a, rtype * b);
extern void matmul_`'rtype_code (rtype * const restrict retarray,
rtype * const restrict a, rtype * const restrict b);
export_proto(matmul_`'rtype_code);
void
matmul_`'rtype_code (rtype * retarray, rtype * a, rtype * b)
matmul_`'rtype_code (rtype * const restrict retarray,
rtype * const restrict a, rtype * const restrict b)
{
rtype_name *abase;
rtype_name *bbase;
rtype_name *dest;
const rtype_name * restrict abase;
const rtype_name * restrict bbase;
rtype_name * restrict dest;
index_type rxstride, rystride, axstride, aystride, bxstride, bystride;
index_type x, y, n, count, xcount, ycount;
@ -106,12 +108,10 @@ matmul_`'rtype_code (rtype * retarray, rtype * a, rtype * b)
retarray->offset = 0;
}
abase = a->data;
bbase = b->data;
dest = retarray->data;
if (retarray->dim[0].stride == 0)
retarray->dim[0].stride = 1;
/* This prevents constifying the input arguments. */
if (a->dim[0].stride == 0)
a->dim[0].stride = 1;
if (b->dim[0].stride == 0)
@ -177,9 +177,9 @@ sinclude(`matmul_asm_'rtype_code`.m4')dnl
if (rxstride == 1 && axstride == 1 && bxstride == 1)
{
rtype_name *bbase_y;
rtype_name *dest_y;
rtype_name *abase_n;
const rtype_name * restrict bbase_y;
rtype_name * restrict dest_y;
const rtype_name * restrict abase_n;
rtype_name bbase_yn;
if (rystride == ycount)

View file

@ -39,15 +39,17 @@ include(iparm.m4)dnl
/* Dimensions: retarray(x,y) a(x, count) b(count,y).
Either a or b can be rank 1. In this case x or y is 1. */
extern void matmul_`'rtype_code (rtype *, gfc_array_l4 *, gfc_array_l4 *);
extern void matmul_`'rtype_code (rtype * const restrict,
gfc_array_l4 * const restrict, gfc_array_l4 * const restrict);
export_proto(matmul_`'rtype_code);
void
matmul_`'rtype_code (rtype * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
matmul_`'rtype_code (rtype * const restrict retarray,
gfc_array_l4 * const restrict a, gfc_array_l4 * const restrict b)
{
GFC_INTEGER_4 *abase;
GFC_INTEGER_4 *bbase;
rtype_name *dest;
const GFC_INTEGER_4 * restrict abase;
const GFC_INTEGER_4 * restrict bbase;
rtype_name * restrict dest;
index_type rxstride;
index_type rystride;
index_type xcount;
@ -57,8 +59,8 @@ matmul_`'rtype_code (rtype * retarray, gfc_array_l4 * a, gfc_array_l4 * b)
index_type x;
index_type y;
GFC_INTEGER_4 *pa;
GFC_INTEGER_4 *pb;
const GFC_INTEGER_4 * restrict pa;
const GFC_INTEGER_4 * restrict pb;
index_type astride;
index_type bstride;
index_type count;