[Libclc-dev] [PATCH 1/1] hypot: Port from amd builtins
Aaron Watry via Libclc-dev
libclc-dev at lists.llvm.org
Mon Apr 9 13:44:02 PDT 2018
On Mon, Apr 9, 2018 at 1:18 PM, Jan Vesely via Libclc-dev
<libclc-dev at lists.llvm.org> wrote:
> Use only subnormal path.
> Passes CTS on carrizo and turks.
> Signed-off-by: Jan Vesely <jan.vesely at rutgers.edu>
> ---
> generic/include/math/clc_hypot.h | 5 ++
> generic/lib/SOURCES | 1 +
> generic/lib/math/clc_hypot.cl | 98 ++++++++++++++++++++++++++++++++++++++++
> generic/lib/math/hypot.cl | 5 +-
> generic/lib/math/hypot.inc | 3 --
> 5 files changed, 108 insertions(+), 4 deletions(-)
> create mode 100644 generic/include/math/clc_hypot.h
> create mode 100644 generic/lib/math/clc_hypot.cl
> delete mode 100644 generic/lib/math/hypot.inc
>
> diff --git a/generic/include/math/clc_hypot.h b/generic/include/math/clc_hypot.h
> new file mode 100644
> index 0000000..66901e5
> --- /dev/null
> +++ b/generic/include/math/clc_hypot.h
> @@ -0,0 +1,5 @@
> +#define __CLC_FUNCTION __clc_hypot
> +#define __CLC_BODY <clc/math/binary_decl_tt.inc>
> +#include <clc/math/gentype.inc>
> +#undef __CLC_BODY
> +#undef __CLC_FUNCTION
> diff --git a/generic/lib/SOURCES b/generic/lib/SOURCES
> index dfa9d0c..36be3c0 100644
> --- a/generic/lib/SOURCES
> +++ b/generic/lib/SOURCES
> @@ -121,6 +121,7 @@ math/half_rsqrt.cl
> math/half_sin.cl
> math/half_sqrt.cl
> math/half_tan.cl
> +math/clc_hypot.cl
> math/hypot.cl
> math/ilogb.cl
> math/clc_ldexp.cl
> diff --git a/generic/lib/math/clc_hypot.cl b/generic/lib/math/clc_hypot.cl
> new file mode 100644
> index 0000000..6954e36
> --- /dev/null
> +++ b/generic/lib/math/clc_hypot.cl
> @@ -0,0 +1,98 @@
> +/*
> + * Copyright (c) 2014 Advanced Micro Devices, Inc.
> + *
> + * Permission is hereby granted, free of charge, to any person obtaining a copy
> + * of this software and associated documentation files (the "Software"), to deal
> + * in the Software without restriction, including without limitation the rights
> + * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
> + * copies of the Software, and to permit persons to whom the Software is
> + * furnished to do so, subject to the following conditions:
> + *
> + * The above copyright notice and this permission notice shall be included in
> + * all copies or substantial portions of the Software.
> + *
> + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
> + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
> + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
> + * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
> + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
> + * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
> + * THE SOFTWARE.
> + */
> +
> +#include <clc/clc.h>
> +#include <math/clc_hypot.h>
> +
> +#include "config.h"
> +#include "math.h"
> +#include "../clcmacro.h"
> +
> +// Returns sqrt(x*x + y*y) with no overflow or underflow unless the result warrants it
> +_CLC_DEF _CLC_OVERLOAD float __clc_hypot(float x, float y)
> +{
> + uint ux = as_uint(x);
> + uint aux = ux & EXSIGNBIT_SP32;
> + uint uy = as_uint(y);
> + uint auy = uy & EXSIGNBIT_SP32;
> + float retval;
Indentation is off here.
> + int c = aux > auy;
> + ux = c ? aux : auy;
> + uy = c ? auy : aux;
> +
> + int xexp = clamp((int)(ux >> EXPSHIFTBITS_SP32) - EXPBIAS_SP32, -126, 126);
> + float fx_exp = as_float((xexp + EXPBIAS_SP32) << EXPSHIFTBITS_SP32);
> + float fi_exp = as_float((-xexp + EXPBIAS_SP32) << EXPSHIFTBITS_SP32);
> + float fx = as_float(ux) * fi_exp;
> + float fy = as_float(uy) * fi_exp;
> + retval = sqrt(mad(fx, fx, fy*fy)) * fx_exp;
> +
> + retval = ux > PINFBITPATT_SP32 | uy == 0 ? as_float(ux) : retval;
> + retval = ux == PINFBITPATT_SP32 | uy == PINFBITPATT_SP32 ? as_float(PINFBITPATT_SP32) : retval;
> + return retval;
> +}
> +_CLC_BINARY_VECTORIZE(_CLC_DEF _CLC_OVERLOAD, float, __clc_hypot, float, float)
> +
> +#ifdef cl_khr_fp64
> +_CLC_DEF _CLC_OVERLOAD double __clc_hypot(double x, double y)
> +{
> + ulong ux = as_ulong(x) & ~SIGNBIT_DP64;
> + int xexp = ux >> EXPSHIFTBITS_DP64;
> + x = as_double(ux);
> +
> + ulong uy = as_ulong(y) & ~SIGNBIT_DP64;
> + int yexp = uy >> EXPSHIFTBITS_DP64;
> + y = as_double(uy);
> +
> + int c = xexp > EXPBIAS_DP64 + 500 | yexp > EXPBIAS_DP64 + 500;
> + double preadjust = c ? 0x1.0p-600 : 1.0;
> + double postadjust = c ? 0x1.0p+600 : 1.0;
> +
> + c = xexp < EXPBIAS_DP64 - 500 | yexp < EXPBIAS_DP64 - 500;
> + preadjust = c ? 0x1.0p+600 : preadjust;
> + postadjust = c ? 0x1.0p-600 : postadjust;
> +
> + double ax = x * preadjust;
> + double ay = y * preadjust;
> +
> + // The post adjust may overflow, but this can't be avoided in any case
> + double r = sqrt(fma(ax, ax, ay*ay)) * postadjust;
> +
> + // If the difference in exponents between x and y is large
> + double s = x + y;
> + c = abs(xexp - yexp) > MANTLENGTH_DP64 + 1;
> + r = c ? s : r;
> +
> + // Check for NaN
Trailing whitespace here.
Otherwise, looks good to me. Passes CTS on my RX 580.
--Aaron
> + //c = x != x | y != y;
> + c = isnan(x) | isnan(y);
> + r = c ? as_double(QNANBITPATT_DP64) : r;
> +
> + // If either is Inf, we must return Inf
> + c = x == as_double(PINFBITPATT_DP64) | y == as_double(PINFBITPATT_DP64);
> + r = c ? as_double(PINFBITPATT_DP64) : r;
> +
> + return r;
> +}
> +
> +_CLC_BINARY_VECTORIZE(_CLC_DEF _CLC_OVERLOAD, double, __clc_hypot, double, double)
> +#endif
> diff --git a/generic/lib/math/hypot.cl b/generic/lib/math/hypot.cl
> index 8be351e..8339ec7 100644
> --- a/generic/lib/math/hypot.cl
> +++ b/generic/lib/math/hypot.cl
> @@ -1,4 +1,7 @@
> #include <clc/clc.h>
>
> -#define __CLC_BODY <hypot.inc>
> +#include <math/clc_hypot.h>
> +
> +#define __CLC_FUNC hypot
> +#define __CLC_BODY <clc_sw_binary.inc>
> #include <clc/math/gentype.inc>
> diff --git a/generic/lib/math/hypot.inc b/generic/lib/math/hypot.inc
> deleted file mode 100644
> index 036cee7..0000000
> --- a/generic/lib/math/hypot.inc
> +++ /dev/null
> @@ -1,3 +0,0 @@
> -_CLC_OVERLOAD _CLC_DEF __CLC_GENTYPE hypot(__CLC_GENTYPE x, __CLC_GENTYPE y) {
> - return sqrt(x*x + y*y);
> -}
> --
> 2.14.3
>
> _______________________________________________
> Libclc-dev mailing list
> Libclc-dev at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/libclc-dev
More information about the Libclc-dev
mailing list