| Message ID | 20240911071319.72356-1-jz531210@gmail.com |
|---|---|
| State | New |
| Headers | show |
| Series | RISC-V: Use builtin for fma and fmaf | expand |
Ping. 在 2024/9/11 15:13, Julian Zhu 写道: > The built-in functions `builtin_{fma, fmaf}` are sufficient to generate correct `fmadd.d`/`fmadd.s` instructions on RISC-V. > > Signed-off-by: Julian Zhu <jz531210@gmail.com> > --- > sysdeps/riscv/rvd/math-use-builtins-fma.h | 4 +++ > sysdeps/riscv/rvd/s_fma.c | 37 ----------------------- > sysdeps/riscv/rvf/s_fmaf.c | 31 ------------------- > 3 files changed, 4 insertions(+), 68 deletions(-) > create mode 100644 sysdeps/riscv/rvd/math-use-builtins-fma.h > delete mode 100644 sysdeps/riscv/rvd/s_fma.c > delete mode 100644 sysdeps/riscv/rvf/s_fmaf.c > > diff --git a/sysdeps/riscv/rvd/math-use-builtins-fma.h b/sysdeps/riscv/rvd/math-use-builtins-fma.h > new file mode 100644 > index 0000000000..eede75aa41 > --- /dev/null > +++ b/sysdeps/riscv/rvd/math-use-builtins-fma.h > @@ -0,0 +1,4 @@ > +#define USE_FMA_BUILTIN 1 > +#define USE_FMAF_BUILTIN 1 > +#define USE_FMAL_BUILTIN 0 > +#define USE_FMAF128_BUILTIN 0 > diff --git a/sysdeps/riscv/rvd/s_fma.c b/sysdeps/riscv/rvd/s_fma.c > deleted file mode 100644 > index fb4b6fb982..0000000000 > --- a/sysdeps/riscv/rvd/s_fma.c > +++ /dev/null > @@ -1,37 +0,0 @@ > -/* Double precision floating point fused multiply-add. RISC-V version. > - Copyright (C) 2017-2024 Free Software Foundation, Inc. > - This file is part of the GNU C Library. > - > - The GNU C Library is free software; you can redistribute it and/or > - modify it under the terms of the GNU Lesser General Public > - License as published by the Free Software Foundation; either > - version 2.1 of the License, or (at your option) any later version. > - > - The GNU C Library is distributed in the hope that it will be useful, > - but WITHOUT ANY WARRANTY; without even the implied warranty of > - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > - Lesser General Public License for more details. > - > - You should have received a copy of the GNU Lesser General Public > - License along with the GNU C Library. If not, see > - <https://www.gnu.org/licenses/>. */ > - > -#define NO_MATH_REDIRECT > -#define dfmal __hide_dfmal > -#define f32xfmaf64 __hide_f32xfmaf64 > -#include <math.h> > -#undef dfmal > -#undef f32xfmaf64 > -#include <fenv.h> > -#include <ieee754.h> > -#include <libm-alias-double.h> > -#include <math-narrow-alias.h> > - > -double > -__fma (double x, double y, double z) > -{ > - asm ("fmadd.d %0, %1, %2, %3" : "=f" (x) : "f" (x), "f" (y), "f" (z)); > - return x; > -} > -libm_alias_double (__fma, fma) > -libm_alias_double_narrow (__fma, fma) > diff --git a/sysdeps/riscv/rvf/s_fmaf.c b/sysdeps/riscv/rvf/s_fmaf.c > deleted file mode 100644 > index 1f1ad3ff41..0000000000 > --- a/sysdeps/riscv/rvf/s_fmaf.c > +++ /dev/null > @@ -1,31 +0,0 @@ > -/* fmaf(). RISC-V version. > - Copyright (C) 2017-2024 Free Software Foundation, Inc. > - This file is part of the GNU C Library. > - > - The GNU C Library is free software; you can redistribute it and/or > - modify it under the terms of the GNU Lesser General Public > - License as published by the Free Software Foundation; either > - version 2.1 of the License, or (at your option) any later version. > - > - The GNU C Library is distributed in the hope that it will be useful, > - but WITHOUT ANY WARRANTY; without even the implied warranty of > - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > - Lesser General Public License for more details. > - > - You should have received a copy of the GNU Lesser General Public > - License along with the GNU C Library. If not, see > - <https://www.gnu.org/licenses/>. */ > - > -#define NO_MATH_REDIRECT > -#include <math.h> > -#include <fenv.h> > -#include <ieee754.h> > -#include <libm-alias-float.h> > - > -float > -__fmaf (float x, float y, float z) > -{ > - asm ("fmadd.s %0, %1, %2, %3" : "=f" (x) : "f" (x), "f" (y), "f" (z)); > - return x; > -} > -libm_alias_float (__fma, fma)
On 11/09/24 04:13, Julian Zhu wrote: > The built-in functions `builtin_{fma, fmaf}` are sufficient to generate correct `fmadd.d`/`fmadd.s` instructions on RISC-V. > > Signed-off-by: Julian Zhu <jz531210@gmail.com> LGTM, thanks. Reviewed-by: Adhemerval Zanella <adhemerval.zanella@linaro.org> > --- > sysdeps/riscv/rvd/math-use-builtins-fma.h | 4 +++ > sysdeps/riscv/rvd/s_fma.c | 37 ----------------------- > sysdeps/riscv/rvf/s_fmaf.c | 31 ------------------- > 3 files changed, 4 insertions(+), 68 deletions(-) > create mode 100644 sysdeps/riscv/rvd/math-use-builtins-fma.h > delete mode 100644 sysdeps/riscv/rvd/s_fma.c > delete mode 100644 sysdeps/riscv/rvf/s_fmaf.c > > diff --git a/sysdeps/riscv/rvd/math-use-builtins-fma.h b/sysdeps/riscv/rvd/math-use-builtins-fma.h > new file mode 100644 > index 0000000000..eede75aa41 > --- /dev/null > +++ b/sysdeps/riscv/rvd/math-use-builtins-fma.h > @@ -0,0 +1,4 @@ > +#define USE_FMA_BUILTIN 1 > +#define USE_FMAF_BUILTIN 1 > +#define USE_FMAL_BUILTIN 0 > +#define USE_FMAF128_BUILTIN 0 > diff --git a/sysdeps/riscv/rvd/s_fma.c b/sysdeps/riscv/rvd/s_fma.c > deleted file mode 100644 > index fb4b6fb982..0000000000 > --- a/sysdeps/riscv/rvd/s_fma.c > +++ /dev/null > @@ -1,37 +0,0 @@ > -/* Double precision floating point fused multiply-add. RISC-V version. > - Copyright (C) 2017-2024 Free Software Foundation, Inc. > - This file is part of the GNU C Library. > - > - The GNU C Library is free software; you can redistribute it and/or > - modify it under the terms of the GNU Lesser General Public > - License as published by the Free Software Foundation; either > - version 2.1 of the License, or (at your option) any later version. > - > - The GNU C Library is distributed in the hope that it will be useful, > - but WITHOUT ANY WARRANTY; without even the implied warranty of > - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > - Lesser General Public License for more details. > - > - You should have received a copy of the GNU Lesser General Public > - License along with the GNU C Library. If not, see > - <https://www.gnu.org/licenses/>. */ > - > -#define NO_MATH_REDIRECT > -#define dfmal __hide_dfmal > -#define f32xfmaf64 __hide_f32xfmaf64 > -#include <math.h> > -#undef dfmal > -#undef f32xfmaf64 > -#include <fenv.h> > -#include <ieee754.h> > -#include <libm-alias-double.h> > -#include <math-narrow-alias.h> > - > -double > -__fma (double x, double y, double z) > -{ > - asm ("fmadd.d %0, %1, %2, %3" : "=f" (x) : "f" (x), "f" (y), "f" (z)); > - return x; > -} > -libm_alias_double (__fma, fma) > -libm_alias_double_narrow (__fma, fma) > diff --git a/sysdeps/riscv/rvf/s_fmaf.c b/sysdeps/riscv/rvf/s_fmaf.c > deleted file mode 100644 > index 1f1ad3ff41..0000000000 > --- a/sysdeps/riscv/rvf/s_fmaf.c > +++ /dev/null > @@ -1,31 +0,0 @@ > -/* fmaf(). RISC-V version. > - Copyright (C) 2017-2024 Free Software Foundation, Inc. > - This file is part of the GNU C Library. > - > - The GNU C Library is free software; you can redistribute it and/or > - modify it under the terms of the GNU Lesser General Public > - License as published by the Free Software Foundation; either > - version 2.1 of the License, or (at your option) any later version. > - > - The GNU C Library is distributed in the hope that it will be useful, > - but WITHOUT ANY WARRANTY; without even the implied warranty of > - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > - Lesser General Public License for more details. > - > - You should have received a copy of the GNU Lesser General Public > - License along with the GNU C Library. If not, see > - <https://www.gnu.org/licenses/>. */ > - > -#define NO_MATH_REDIRECT > -#include <math.h> > -#include <fenv.h> > -#include <ieee754.h> > -#include <libm-alias-float.h> > - > -float > -__fmaf (float x, float y, float z) > -{ > - asm ("fmadd.s %0, %1, %2, %3" : "=f" (x) : "f" (x), "f" (y), "f" (z)); > - return x; > -} > -libm_alias_float (__fma, fma)
diff --git a/sysdeps/riscv/rvd/math-use-builtins-fma.h b/sysdeps/riscv/rvd/math-use-builtins-fma.h new file mode 100644 index 0000000000..eede75aa41 --- /dev/null +++ b/sysdeps/riscv/rvd/math-use-builtins-fma.h @@ -0,0 +1,4 @@ +#define USE_FMA_BUILTIN 1 +#define USE_FMAF_BUILTIN 1 +#define USE_FMAL_BUILTIN 0 +#define USE_FMAF128_BUILTIN 0 diff --git a/sysdeps/riscv/rvd/s_fma.c b/sysdeps/riscv/rvd/s_fma.c deleted file mode 100644 index fb4b6fb982..0000000000 --- a/sysdeps/riscv/rvd/s_fma.c +++ /dev/null @@ -1,37 +0,0 @@ -/* Double precision floating point fused multiply-add. RISC-V version. - Copyright (C) 2017-2024 Free Software Foundation, Inc. - This file is part of the GNU C Library. - - The GNU C Library is free software; you can redistribute it and/or - modify it under the terms of the GNU Lesser General Public - License as published by the Free Software Foundation; either - version 2.1 of the License, or (at your option) any later version. - - The GNU C Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public - License along with the GNU C Library. If not, see - <https://www.gnu.org/licenses/>. */ - -#define NO_MATH_REDIRECT -#define dfmal __hide_dfmal -#define f32xfmaf64 __hide_f32xfmaf64 -#include <math.h> -#undef dfmal -#undef f32xfmaf64 -#include <fenv.h> -#include <ieee754.h> -#include <libm-alias-double.h> -#include <math-narrow-alias.h> - -double -__fma (double x, double y, double z) -{ - asm ("fmadd.d %0, %1, %2, %3" : "=f" (x) : "f" (x), "f" (y), "f" (z)); - return x; -} -libm_alias_double (__fma, fma) -libm_alias_double_narrow (__fma, fma) diff --git a/sysdeps/riscv/rvf/s_fmaf.c b/sysdeps/riscv/rvf/s_fmaf.c deleted file mode 100644 index 1f1ad3ff41..0000000000 --- a/sysdeps/riscv/rvf/s_fmaf.c +++ /dev/null @@ -1,31 +0,0 @@ -/* fmaf(). RISC-V version. - Copyright (C) 2017-2024 Free Software Foundation, Inc. - This file is part of the GNU C Library. - - The GNU C Library is free software; you can redistribute it and/or - modify it under the terms of the GNU Lesser General Public - License as published by the Free Software Foundation; either - version 2.1 of the License, or (at your option) any later version. - - The GNU C Library is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - Lesser General Public License for more details. - - You should have received a copy of the GNU Lesser General Public - License along with the GNU C Library. If not, see - <https://www.gnu.org/licenses/>. */ - -#define NO_MATH_REDIRECT -#include <math.h> -#include <fenv.h> -#include <ieee754.h> -#include <libm-alias-float.h> - -float -__fmaf (float x, float y, float z) -{ - asm ("fmadd.s %0, %1, %2, %3" : "=f" (x) : "f" (x), "f" (y), "f" (z)); - return x; -} -libm_alias_float (__fma, fma)
The built-in functions `builtin_{fma, fmaf}` are sufficient to generate correct `fmadd.d`/`fmadd.s` instructions on RISC-V. Signed-off-by: Julian Zhu <jz531210@gmail.com> --- sysdeps/riscv/rvd/math-use-builtins-fma.h | 4 +++ sysdeps/riscv/rvd/s_fma.c | 37 ----------------------- sysdeps/riscv/rvf/s_fmaf.c | 31 ------------------- 3 files changed, 4 insertions(+), 68 deletions(-) create mode 100644 sysdeps/riscv/rvd/math-use-builtins-fma.h delete mode 100644 sysdeps/riscv/rvd/s_fma.c delete mode 100644 sysdeps/riscv/rvf/s_fmaf.c