Skip to content
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

Jump to bottom

Support quad-complex ABS intrinsic #1411

Merged
merged 1 commit into from
Aug 28, 2023
Merged

Support quad-complex ABS intrinsic #1411

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
1 change: 1 addition & 0 deletions runtime/libpgmath/lib/common/CMakeLists.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -229,6 +229,7 @@ set(MTH_CMPLX_SRCS
ctanh.c)

set(MTH_CMPLX_SRCS_QUADFP
cqabs.c
cqdiv.c
cqpowcq.c
cqpowi.c
Expand Down
16 changes: 16 additions & 0 deletions runtime/libpgmath/lib/common/cqabs.c
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,16 @@
/*
* Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
* See https://llvm.org/LICENSE.txt for license information.
* SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
*
*/

#include "mthdecls.h"

/* ----------------------------- long double complex functions: */

QUADFUNC_C(__mth_i_cqabs)
{
LZMPLXARGS_LZ;
LZRETURN_Q(hypotl(real, imag));
}
10 changes: 9 additions & 1 deletion runtime/libpgmath/lib/common/mthdecls.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -193,6 +193,8 @@ quad_complex_t pgmath_cmplxl(float128_t r, float128_t i)
float _f(float real, float imag)
#define DBLFUNC_C_(_f) \
double _f(double real, double imag)
#define QUADFUNC_C_(_f) \
float128_t _f(float128_t real, float128_t imag)

#define CMPLXFUNC_C_(_f) \
void _f(cmplx_t *cmplx, float real, float imag)
Expand Down Expand Up @@ -237,6 +239,9 @@ quad_complex_t pgmath_cmplxl(float128_t r, float128_t i)
#define DBLFUNC_C_C99_(_f) \
double MTHCONCAT__(_f,__MTH_C99_CMPLX_SUFFIX) \
(double_complex_t zarg)
#define QUADFUNC_C_C99_(_f) \
float128_t MTHCONCAT__(_f,__MTH_C99_CMPLX_SUFFIX) \
(quad_complex_t lzarg)

#define CMPLXFUNC_C_C99_(_f) \
float_complex_t MTHCONCAT__(_f,__MTH_C99_CMPLX_SUFFIX) \
Expand Down Expand Up @@ -290,6 +295,7 @@ quad_complex_t pgmath_cmplxl(float128_t r, float128_t i)

#define FLTFUNC_C(_f) FLTFUNC_C_(_f)
#define DBLFUNC_C(_f) DBLFUNC_C_(_f)
#define QUADFUNC_C(_f) QUADFUNC_C_(_f)

#define CMPLXFUNC_C(_f) CMPLXFUNC_C_(_f)
#define CMPLXFUNC_C_C(_f) CMPLXFUNC_C_C_(_f)
Expand Down Expand Up @@ -359,6 +365,7 @@ quad_complex_t pgmath_cmplxl(float128_t r, float128_t i)

#define FLTFUNC_C(_f) FLTFUNC_C_C99_(_f)
#define DBLFUNC_C(_f) DBLFUNC_C_C99_(_f)
#define QUADFUNC_C(_f) QUADFUNC_C_C99_(_f)

#define CMPLXFUNC_C(_f) CMPLXFUNC_C_C99_(_f)
#define CMPLXFUNC_C_C(_f) CMPLXFUNC_C_C_C99_(_f)
Expand Down Expand Up @@ -427,6 +434,7 @@ quad_complex_t pgmath_cmplxl(float128_t r, float128_t i)
*/
#define FLTDECL_C(_f) FLTFUNC_C_(_f) ; FLTFUNC_C_C99_(_f);
#define DBLDECL_C(_f) DBLFUNC_C_(_f) ; DBLFUNC_C_C99_(_f);
#define QUADDECL_C(_f) QUADFUNC_C_(_f) ; QUADFUNC_C_C99_(_f);

#define CMPLXDECL_C(_f) CMPLXFUNC_C_(_f) ; CMPLXFUNC_C_C99_(_f);
#define CMPLXDECL_C_C(_f) CMPLXFUNC_C_C_(_f) ; CMPLXFUNC_C_C_C99_(_f);
Expand Down Expand Up @@ -718,7 +726,7 @@ ZMPLXDECL_Z(__mth_i_cdsqrt);
ZMPLXDECL_Z(__mth_i_cdtan);
ZMPLXDECL_Z(__mth_i_cdtanh);


QUADDECL_C(__mth_i_cqabs);
__CDECL LZMPLXDECL_LZ_LZ(__mth_i_cqdiv);
__CDECL LZMPLXDECL_LZ_LZ(__mth_i_cqpowcq);
LZMPLXDECL_LZ_I(__mth_i_cqpowi);
Expand Down
20 changes: 20 additions & 0 deletions test/f90_correct/inc/qc_abs_01.mk
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
# Test intrinsics function abs with quad precision complex.

$(TEST): run

build: $(SRC)/$(TEST).f08
-$(RM) $(TEST).$(EXESUFFIX) core *.d *.mod FOR*.DAT FTN* ftn* fort.*
@echo ------------------------------------ building test $@
-$(FC) -c $(FFLAGS) $(SRC)/check_mod.F90 -o check_mod.$(OBJX)
-$(FC) -c $(FFLAGS) $(LDFLAGS) $(SRC)/$(TEST).f08 -o $(TEST).$(OBJX)
-$(FC) $(FFLAGS) $(LDFLAGS) $(TEST).$(OBJX) check_mod.$(OBJX) $(LIBS) -o $(TEST).$(EXESUFFIX)

run:
@echo ------------------------------------ executing test $(TEST)
$(TEST).$(EXESUFFIX)

verify: ;
20 changes: 20 additions & 0 deletions test/f90_correct/inc/qc_abs_02.mk
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
# Test intrinsics function abs with quad precision complex.

$(TEST): run

build: $(SRC)/$(TEST).f08
-$(RM) $(TEST).$(EXESUFFIX) core *.d *.mod FOR*.DAT FTN* ftn* fort.*
@echo ------------------------------------ building test $@
-$(FC) -c $(FFLAGS) $(SRC)/check_mod.F90 -o check_mod.$(OBJX)
-$(FC) -c $(FFLAGS) $(LDFLAGS) $(SRC)/$(TEST).f08 -o $(TEST).$(OBJX)
-$(FC) $(FFLAGS) $(LDFLAGS) $(TEST).$(OBJX) check_mod.$(OBJX) $(LIBS) -o $(TEST).$(EXESUFFIX)

run:
@echo ------------------------------------ executing test $(TEST)
$(TEST).$(EXESUFFIX)

verify: ;
20 changes: 20 additions & 0 deletions test/f90_correct/inc/qc_abs_03.mk
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
# Test intrinsics function abs with quad precision complex.

$(TEST): run

build: $(SRC)/$(TEST).f08
-$(RM) $(TEST).$(EXESUFFIX) core *.d *.mod FOR*.DAT FTN* ftn* fort.*
@echo ------------------------------------ building test $@
-$(FC) -c $(FFLAGS) $(SRC)/check_mod.F90 -o check_mod.$(OBJX)
-$(FC) -c $(FFLAGS) $(LDFLAGS) $(SRC)/$(TEST).f08 -o $(TEST).$(OBJX)
-$(FC) $(FFLAGS) $(LDFLAGS) $(TEST).$(OBJX) check_mod.$(OBJX) $(LIBS) -o $(TEST).$(EXESUFFIX)

run:
@echo ------------------------------------ executing test $(TEST)
$(TEST).$(EXESUFFIX)

verify: ;
20 changes: 20 additions & 0 deletions test/f90_correct/inc/qc_abs_04.mk
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
# Test intrinsics function abs with quad precision complex.

$(TEST): run

build: $(SRC)/$(TEST).f08
-$(RM) $(TEST).$(EXESUFFIX) core *.d *.mod FOR*.DAT FTN* ftn* fort.*
@echo ------------------------------------ building test $@
-$(FC) -c $(FFLAGS) $(SRC)/check_mod.F90 -o check_mod.$(OBJX)
-$(FC) -c $(FFLAGS) $(LDFLAGS) $(SRC)/$(TEST).f08 -o $(TEST).$(OBJX)
-$(FC) $(FFLAGS) $(LDFLAGS) $(TEST).$(OBJX) check_mod.$(OBJX) $(LIBS) -o $(TEST).$(EXESUFFIX)

run:
@echo ------------------------------------ executing test $(TEST)
$(TEST).$(EXESUFFIX)

verify: ;
20 changes: 20 additions & 0 deletions test/f90_correct/inc/qc_abs_05.mk
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,20 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
# Test intrinsics function abs with quad precision complex.

$(TEST): run

build: $(SRC)/$(TEST).f08
-$(RM) $(TEST).$(EXESUFFIX) core *.d *.mod FOR*.DAT FTN* ftn* fort.*
@echo ------------------------------------ building test $@
-$(FC) -c $(FFLAGS) $(SRC)/check_mod.F90 -o check_mod.$(OBJX)
-$(FC) -c $(FFLAGS) $(LDFLAGS) $(SRC)/$(TEST).f08 -o $(TEST).$(OBJX)
-$(FC) $(FFLAGS) $(LDFLAGS) $(TEST).$(OBJX) check_mod.$(OBJX) $(LIBS) -o $(TEST).$(EXESUFFIX)

run:
@echo ------------------------------------ executing test $(TEST)
$(TEST).$(EXESUFFIX)

verify: ;
10 changes: 10 additions & 0 deletions test/f90_correct/lit/qc_abs_01.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

# Shared lit script for each tests. Run bash commands that run tests with make.

# REQUIRES: quadfp
# RUN: KEEP_FILES=%keep FLAGS=%flags TEST_SRC=%s MAKE_FILE_DIR=%S/.. bash %S/runmake | tee %t
# RUN: cat %t | FileCheck %S/runmake
10 changes: 10 additions & 0 deletions test/f90_correct/lit/qc_abs_02.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

# Shared lit script for each tests. Run bash commands that run tests with make.

# REQUIRES: quadfp
# RUN: KEEP_FILES=%keep FLAGS=%flags TEST_SRC=%s MAKE_FILE_DIR=%S/.. bash %S/runmake | tee %t
# RUN: cat %t | FileCheck %S/runmake
10 changes: 10 additions & 0 deletions test/f90_correct/lit/qc_abs_03.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

# Shared lit script for each tests. Run bash commands that run tests with make.

# REQUIRES: quadfp
# RUN: KEEP_FILES=%keep FLAGS=%flags TEST_SRC=%s MAKE_FILE_DIR=%S/.. bash %S/runmake | tee %t
# RUN: cat %t | FileCheck %S/runmake
10 changes: 10 additions & 0 deletions test/f90_correct/lit/qc_abs_04.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

# Shared lit script for each tests. Run bash commands that run tests with make.

# REQUIRES: quadfp
# RUN: KEEP_FILES=%keep FLAGS=%flags TEST_SRC=%s MAKE_FILE_DIR=%S/.. bash %S/runmake | tee %t
# RUN: cat %t | FileCheck %S/runmake
10 changes: 10 additions & 0 deletions test/f90_correct/lit/qc_abs_05.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
#
# Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
# See https://llvm.org/LICENSE.txt for license information.
# SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception

# Shared lit script for each tests. Run bash commands that run tests with make.

# REQUIRES: quadfp
# RUN: KEEP_FILES=%keep FLAGS=%flags TEST_SRC=%s MAKE_FILE_DIR=%S/.. bash %S/runmake | tee %t
# RUN: cat %t | FileCheck %S/runmake
93 changes: 93 additions & 0 deletions test/f90_correct/src/check_mod.F90
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,6 +15,9 @@ module check_mod
module procedure checkr16
#endif
module procedure checkc4, checkc8, checkc1
#ifdef __flang_quadfp__
module procedure checkc16
#endif
module procedure checkcptr, checkcptr2d, checkbytes, checkdt
end interface

Expand Down Expand Up @@ -57,6 +60,7 @@ module check_mod
character(80) :: fmt20="'res (0x',z2.2,') exp (0x',z2.2,')')"
#ifdef __flang_quadfp__
character(160) :: fmt21="'res ',f0.33,' (0x',z33.33,') exp ',f0.33,' (0x',z33.33,')')"
character(320) :: fmt22="'res ',2(f0.33,1x),2('(0x',z33.33,') '),'exp ',2(f0.33,1x),2('(0x',z33.33,') '))"
#endif

contains
Expand Down Expand Up @@ -933,6 +937,95 @@ subroutine checkc8(reslt, expct, np, atoler, rtoler, ulptoler, ieee)
endif
return
end subroutine checkc8
#ifdef __flang_quadfp__
! complex*32
subroutine checkc16(reslt, expct, np, atoler, rtoler, ulptoler, ieee)
!dir$ ignore_tkr (r) reslt, expct
complex*32, dimension(*) :: reslt
complex*32, dimension(*) :: expct
integer :: np
complex*32, optional :: atoler, rtoler, ulptoler
logical, optional :: ieee
integer i, tests_passed, tests_failed, tests_tolerated
real*16 abserror, relerror
real*16 rres, rexp, rx
logical ieee_on, anytolerated

anytolerated = present(atoler) .or. present(rtoler) .or. present(ulptoler)
ieee_on = .false.
if (present(ieee)) ieee_on = ieee

tests_passed = 0
tests_failed = 0
tests_tolerated = 0

do i = 1, np
if (ieee_on) then
rres = qreal(reslt(i))
rexp = qreal(expct(i))
irri = ieeecheckcases(rres, rexp)
rres = qimag(reslt(i))
rexp = qimag(expct(i))
icri = ieeecheckcases(rres, rexp)
if ((irri.eq.1) .or. (icri.eq.1)) then
goto 100
else if ((irri.eq.2) .and. (icri.eq.2)) then
tests_passed = tests_passed + 1
cycle
end if
end if

if (expct(i) .eq. reslt(i)) then
tests_passed = tests_passed + 1
cycle
end if

abserror = cqabs(expct(i) - reslt(i))
if (present(atoler)) then
if (abserror .gt. cqabs(atoler)) goto 100
end if

if (present(rtoler)) then
relerror = abserror / qmax(cqabs(expct(i)),ieee_next_after(0.0_16,1.0_16))
if (relerror .gt. cqabs(rtoler)) goto 100
end if

if (present(ulptoler)) then
rres = qreal(reslt(i))
rexp = qreal(expct(i))
if (ulperror(rres,rexp) .gt. qreal(ulptoler)) goto 100
rres = qimag(reslt(i))
rexp = qimag(expct(i))
if (ulperror(rres,rexp) .gt. qimag(ulptoler)) goto 100
end if

if (anytolerated) then ! Some tolerances, so if here we've passed
tests_passed = tests_passed + 1
tests_tolerated = tests_tolerated + 1
if (tests_tolerated .le. 100) then
write(6,fmt=fmt02//fmt22) i, reslt(i),reslt(i), expct(i),expct(i)
end if
cycle
end if

100 tests_failed = tests_failed + 1 ! No tolerances, here we've failed
if (tests_failed .le. 100) then
write(6,fmt=fmt03//fmt22) i, reslt(i),reslt(i), expct(i),expct(i)
end if
enddo

if (tests_failed .eq. 0) then
if (tests_tolerated .eq. 0) then
write(6,fmt=fmt04) np, tests_passed
else
write(6,fmt=fmt05) np, tests_passed, tests_tolerated
endif
else
write(6,fmt=fmt06) np, tests_passed, tests_failed
endif
return
end subroutine checkc16
#endif

! Now character*1
subroutine checkc1(reslt, expct, np, atoler, rtoler)
Expand Down
Loading
Loading