Skip to content

Commit

Permalink
Add narrowing multiply functions.
Browse files Browse the repository at this point in the history 
This patch adds the narrowing multiply functions from TS 18661-1 to
glibc's libm: fmul, fmull, dmull, f32mulf64, f32mulf32x, f32xmulf64
for all configurations; f32mulf64x, f32mulf128, f64mulf64x,
f64mulf128, f32xmulf64x, f32xmulf128, f64xmulf128 for configurations
with _Float64x and _Float128; __nldbl_dmull for ldbl-opt.

The changes are mostly essentially the same as for the narrowing add
functions, so the description of those generally applies to this patch
as well.  f32xmulf64 for i386 cannot use precision control as used for
add and subtract, because that would result in double rounding for
subnormal results, so that uses round-to-odd with long double
intermediate result instead.  The soft-fp support involves adding a
new FP_TRUNC_COOKED since soft-fp multiplication uses cooked inputs
and outputs.

Tested for x86_64, x86, mips64 (all three ABIs, both hard and soft
float) and powerpc, and with build-many-glibcs.py.

	* math/Makefile (libm-narrow-fns): Add mul.
	(libm-test-funcs-narrow): Likewise.
	* math/Versions (GLIBC_2.28): Add narrowing multiply functions.
	* math/bits/mathcalls-narrow.h (mul): Use __MATHCALL_NARROW.
	* math/gen-auto-libm-tests.c (test_functions): Add mul.
	* math/math-narrow.h (CHECK_NARROW_MUL): New macro.
	(NARROW_MUL_ROUND_TO_ODD): Likewise.
	(NARROW_MUL_TRIVIAL): Likewise.
	* soft-fp/op-common.h (FP_TRUNC_COOKED): Likewise.
	* sysdeps/ieee754/float128/float128_private.h (__fmull): New
	macro.
	(__dmull): Likewise.
	* sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Add fmul and
	dmul.
	(CFLAGS-nldbl-dmul.c): New variable.
	(CFLAGS-nldbl-fmul.c): Likewise.
	* sysdeps/ieee754/ldbl-opt/Versions (GLIBC_2.28): Add
	__nldbl_dmull.
	* sysdeps/ieee754/ldbl-opt/nldbl-compat.h (__nldbl_dmull): New
	prototype.
	* manual/arith.texi (Misc FP Arithmetic): Document fmul, fmull,
	dmull, fMmulfN, fMmulfNx, fMxmulfN and fMxmulfNx.
	* math/auto-libm-test-in: Add tests of mul.
	* math/auto-libm-test-out-narrow-mul: New generated file.
	* math/libm-test-narrow-mul.inc: New file.
	* sysdeps/i386/fpu/s_f32xmulf64.c: Likewise.
	* sysdeps/ieee754/dbl-64/s_f32xmulf64.c: Likewise.
	* sysdeps/ieee754/dbl-64/s_fmul.c: Likewise.
	* sysdeps/ieee754/float128/s_f32mulf128.c: Likewise.
	* sysdeps/ieee754/float128/s_f64mulf128.c: Likewise.
	* sysdeps/ieee754/float128/s_f64xmulf128.c: Likewise.
	* sysdeps/ieee754/ldbl-128/s_dmull.c: Likewise.
	* sysdeps/ieee754/ldbl-128/s_f64xmulf128.c: Likewise.
	* sysdeps/ieee754/ldbl-128/s_fmull.c: Likewise.
	* sysdeps/ieee754/ldbl-128ibm/s_dmull.c: Likewise.
	* sysdeps/ieee754/ldbl-128ibm/s_fmull.c: Likewise.
	* sysdeps/ieee754/ldbl-96/s_dmull.c: Likewise.
	* sysdeps/ieee754/ldbl-96/s_fmull.c: Likewise.
	* sysdeps/ieee754/ldbl-opt/nldbl-dmul.c: Likewise.
	* sysdeps/ieee754/ldbl-opt/nldbl-fmul.c: Likewise.
	* sysdeps/ieee754/soft-fp/s_dmull.c: Likewise.
	* sysdeps/ieee754/soft-fp/s_fmul.c: Likewise.
	* sysdeps/ieee754/soft-fp/s_fmull.c: Likewise.
	* sysdeps/powerpc/fpu/libm-test-ulps: Update.
	* sysdeps/mach/hurd/i386/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/aarch64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/alpha/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/arm/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/hppa/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/m68k/coldfire/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/m68k/m680x0/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/microblaze/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/mips/mips32/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/mips/mips64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/nios2/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist: Likewise.
	* sysdeps/unix/sysv/linux/powerpc/powerpc64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/riscv/rv64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/s390/s390-32/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/s390/s390-64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/sh/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/sparc/sparc32/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/sparc/sparc64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise.
	* sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.
  • Loading branch information
@jsm28
jsm28 committed May 16, 2018
1 parent a15529f commit 69a0146
Show file tree
Hide file tree
Showing 60 changed files with 12,631 additions and 4 deletions.
72 changes: 72 additions & 0 deletions ChangeLog
View file
Original file line number Diff line number Diff line change
@@ -1,3 +1,75 @@
2018-05-16 Joseph Myers <[email protected]>

* math/Makefile (libm-narrow-fns): Add mul.
(libm-test-funcs-narrow): Likewise.
* math/Versions (GLIBC_2.28): Add narrowing multiply functions.
* math/bits/mathcalls-narrow.h (mul): Use __MATHCALL_NARROW.
* math/gen-auto-libm-tests.c (test_functions): Add mul.
* math/math-narrow.h (CHECK_NARROW_MUL): New macro.
(NARROW_MUL_ROUND_TO_ODD): Likewise.
(NARROW_MUL_TRIVIAL): Likewise.
* soft-fp/op-common.h (FP_TRUNC_COOKED): Likewise.
* sysdeps/ieee754/float128/float128_private.h (__fmull): New
macro.
(__dmull): Likewise.
* sysdeps/ieee754/ldbl-opt/Makefile (libnldbl-calls): Add fmul and
dmul.
(CFLAGS-nldbl-dmul.c): New variable.
(CFLAGS-nldbl-fmul.c): Likewise.
* sysdeps/ieee754/ldbl-opt/Versions (GLIBC_2.28): Add
__nldbl_dmull.
* sysdeps/ieee754/ldbl-opt/nldbl-compat.h (__nldbl_dmull): New
prototype.
* manual/arith.texi (Misc FP Arithmetic): Document fmul, fmull,
dmull, fMmulfN, fMmulfNx, fMxmulfN and fMxmulfNx.
* math/auto-libm-test-in: Add tests of mul.
* math/auto-libm-test-out-narrow-mul: New generated file.
* math/libm-test-narrow-mul.inc: New file.
* sysdeps/i386/fpu/s_f32xmulf64.c: Likewise.
* sysdeps/ieee754/dbl-64/s_f32xmulf64.c: Likewise.
* sysdeps/ieee754/dbl-64/s_fmul.c: Likewise.
* sysdeps/ieee754/float128/s_f32mulf128.c: Likewise.
* sysdeps/ieee754/float128/s_f64mulf128.c: Likewise.
* sysdeps/ieee754/float128/s_f64xmulf128.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_dmull.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_f64xmulf128.c: Likewise.
* sysdeps/ieee754/ldbl-128/s_fmull.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_dmull.c: Likewise.
* sysdeps/ieee754/ldbl-128ibm/s_fmull.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_dmull.c: Likewise.
* sysdeps/ieee754/ldbl-96/s_fmull.c: Likewise.
* sysdeps/ieee754/ldbl-opt/nldbl-dmul.c: Likewise.
* sysdeps/ieee754/ldbl-opt/nldbl-fmul.c: Likewise.
* sysdeps/ieee754/soft-fp/s_dmull.c: Likewise.
* sysdeps/ieee754/soft-fp/s_fmul.c: Likewise.
* sysdeps/ieee754/soft-fp/s_fmull.c: Likewise.
* sysdeps/powerpc/fpu/libm-test-ulps: Update.
* sysdeps/mach/hurd/i386/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/aarch64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/alpha/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/arm/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/hppa/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/i386/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/ia64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/m68k/coldfire/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/m68k/m680x0/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/microblaze/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/mips/mips32/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/mips/mips64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/nios2/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/fpu/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc32/nofpu/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/libm-le.abilist: Likewise.
* sysdeps/unix/sysv/linux/powerpc/powerpc64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/riscv/rv64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/s390/s390-32/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/s390/s390-64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/sh/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/sparc/sparc32/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/sparc/sparc64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/64/libm.abilist: Likewise.
* sysdeps/unix/sysv/linux/x86_64/x32/libm.abilist: Likewise.

2018-05-14 H.J. Lu <[email protected]>

* sysdeps/i386/crti.S (_init): Replace PREINIT_FUNCTION@PLT
Expand Down
3 changes: 3 additions & 0 deletions NEWS
View file
Original file line number Diff line number Diff line change
Expand Up @@ -18,6 +18,9 @@ Major new features:
- fsub, fsubl, dsubl and corresponding fMsubfN, fMsubfNx, fMxsubfN and
fMxsubfNx functions.

- fmul, fmull, dmull and corresponding fMmulfN, fMmulfNx, fMxmulfN and
fMxmulfNx functions.

* Nominative and genitive month names are now supported for the Catalan and
Czech languages. The Catalan and Greek languages now support abbreviated
alternative month names.
Expand Down
19 changes: 19 additions & 0 deletions manual/arith.texi
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2229,6 +2229,25 @@ function without any intermediate rounding to the type of the
arguments.
@end deftypefun

@deftypefun float fmul (double @var{x}, double @var{y})
@deftypefunx float fmull (long double @var{x}, long double @var{y})
@deftypefunx double dmull (long double @var{x}, long double @var{y})
@deftypefunx _FloatM fMmulfN (_Float@var{N} @var{x}, _Float@var{N} @var{y})
@deftypefunx _FloatM fMmulfNx (_Float@var{N}x @var{x}, _Float@var{N}x @var{y})
@deftypefunx _FloatMx fMxmulfN (_Float@var{N} @var{x}, _Float@var{N} @var{y})
@deftypefunx _FloatMx fMxmulfNx (_Float@var{N}x @var{x}, _Float@var{N}x @var{y})
@standards{TS 18661-1:2014, math.h}
@standardsx{fMmulfN, TS 18661-3:2015, math.h}
@standardsx{fMmulfNx, TS 18661-3:2015, math.h}
@standardsx{fMxmulfN, TS 18661-3:2015, math.h}
@standardsx{fMxmulfNx, TS 18661-3:2015, math.h}
@safety{@prelim{}@mtsafe{}@assafe{}@acsafe{}}
These functions, from TS 18661-1:2014 and TS 18661-3:2015, return
@math{@var{x} * @var{y}}, rounded once to the return type of the
function without any intermediate rounding to the type of the
arguments.
@end deftypefun

@node Complex Numbers
@section Complex Numbers
@pindex complex.h
Expand Down
4 changes: 2 additions & 2 deletions math/Makefile
View file
Original file line number Diff line number Diff line change
Expand Up @@ -89,7 +89,7 @@ libm-compat-calls = \
w_lgammaF_r_compat w_lgammaF_compat2 w_expF_compat \
w_lgamma_compatF k_standardF

libm-narrow-fns = add sub
libm-narrow-fns = add mul sub
libm-narrow-types-basic = s_fF s_f32xFf64
libm-narrow-types-ldouble-yes = s_fFl s_dFl
libm-narrow-types-float128-yes = s_f32Ff128 s_f64Ff128 s_f64xFf128
Expand Down Expand Up @@ -284,7 +284,7 @@ libm-test-funcs-noauto = canonicalize ceil cimag conj copysign cproj creal \
scalbln scalbn setpayload setpayloadsig signbit \
significand totalorder totalordermag trunc ufromfp \
ufromfpx
libm-test-funcs-narrow = add sub
libm-test-funcs-narrow = add mul sub
libm-test-funcs-all = $(libm-test-funcs-auto) $(libm-test-funcs-noauto)
libm-test-c-auto = $(foreach f,$(libm-test-funcs-auto),libm-test-$(f).c)
libm-test-c-noauto = $(foreach f,$(libm-test-funcs-noauto),libm-test-$(f).c)
Expand Down
5 changes: 5 additions & 0 deletions math/Versions
View file
Original file line number Diff line number Diff line change
Expand Up @@ -555,12 +555,17 @@ libm {
# Functions not involving _Float64x or _Float128, for all configurations.
fadd; faddl; daddl;
f32addf32x; f32addf64; f32xaddf64;
fmul; fmull; dmull;
f32mulf32x; f32mulf64; f32xmulf64;
fsub; fsubl; dsubl;
f32subf32x; f32subf64; f32xsubf64;
# Functions involving _Float64x or _Float128, for some configurations.
f32addf64x; f32addf128;
f32xaddf64x; f32xaddf128; f64addf64x; f64addf128;
f64xaddf128;
f32mulf64x; f32mulf128;
f32xmulf64x; f32xmulf128; f64mulf64x; f64mulf128;
f64xmulf128;
f32subf64x; f32subf128;
f32xsubf64x; f32xsubf128; f64subf64x; f64subf128;
f64xsubf128;
Expand Down
48 changes: 48 additions & 0 deletions math/auto-libm-test-in
View file
Original file line number Diff line number Diff line change
Expand Up @@ -6445,6 +6445,54 @@ log2 min
log2 min_subnorm
log2 max

mul 0 0
mul 0 -0
mul -0 0
mul -0 -0
mul max max
mul max -max
mul -max max
mul -max -max
mul min min
mul min -min
mul -min min
mul -min -min
mul min_subnorm min_subnorm
mul min_subnorm -min_subnorm
mul -min_subnorm min_subnorm
mul -min_subnorm -min_subnorm
mul 1 2
mul 1 -2
mul -1 2
mul -1 -2
mul 100.5 0.75
mul 100.5 -0.75
mul -100.5 0.75
mul -100.5 -0.75
# Cases where the product of two values in a wider format is very
# close to half way between two representable values in a narrower
# format, so that double rounding would sometimes yield the wrong
# result. These examples have products of the form
# (2^a + 1)(2^b + 1)/2^(a+b), where a is the width of the narrower
# format, b is greater than the width of the wider format and factors
# have been rearranged between the terms so that both terms do fit
# within the wider format.
# a = 24, b = 54
mul 0x5000005p-24 0xccccccccccccdp-54
# a = 24, b = 65
mul 0x3000003p-24 0xaaaaaaaaaaaaaaabp-65
# a = 24, b = 114
mul 0x5000005p-24 0xcccccccccccccccccccccccccccdp-114
# a = 53, b = 65
mul 0x60000000000003p-53 0xaaaaaaaaaaaaaaabp-65
# a = 53, b = 114
mul 0xa0000000000005p-53 0xcccccccccccccccccccccccccccdp-114
# a = 64, b = 114
mul 0x50000000000000005p-64 0xcccccccccccccccccccccccccccdp-114
# Similar, for double rounding to 64 bit of a product of 53-bit values.
# This product equals 2^64 + 2^11 + 1.
mul 97689974585 188829449

pow 0 0
pow 0 -0
pow -0 0
Expand Down
Loading

0 comments on commit 69a0146

Please sign in to comment.