From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 1283 invoked by alias); 3 Jul 2013 20:21:03 -0000 Mailing-List: contact libc-ports-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Post: List-Help: , Sender: libc-ports-owner@sourceware.org Received: (qmail 1226 invoked by uid 89); 3 Jul 2013 20:20:57 -0000 X-Spam-SWARE-Status: No, score=-1.4 required=5.0 tests=AWL,BAYES_00,DATE_IN_PAST_03_06,RP_MATCHES_RCVD,TW_SF autolearn=no version=3.3.1 Received: from usmamail.tilera.com (HELO USMAMAIL.TILERA.COM) (12.216.194.151) by sourceware.org (qpsmtpd/0.84/v0.84-167-ge50287c) with ESMTP; Wed, 03 Jul 2013 20:20:50 +0000 Received: from farm-0002.internal.tilera.com (10.2.0.32) by USMAEXCH2.tad.internal.tilera.com (10.3.0.33) with Microsoft SMTP Server (TLS) id 14.0.722.0; Wed, 3 Jul 2013 16:20:48 -0400 Received: (from cmetcalf@localhost) by farm-0002.internal.tilera.com (8.14.4/8.12.11/Submit) id r63KKmb3007920; Wed, 3 Jul 2013 16:20:48 -0400 Message-ID: <201307032020.r63KKmb3007920@farm-0002.internal.tilera.com> From: Chris Metcalf Date: Wed, 03 Jul 2013 20:21:00 -0000 Subject: [PATCH] tile: use soft-fp for fma() and fmaf() To: MIME-Version: 1.0 Content-Type: text/plain X-SW-Source: 2013-07/txt/msg00016.txt.bz2 The sfp-machine.h is based on the gcc version, but extended with required new macros by comparison with other architectures and by investigating the hardware support for FP on tile. --- 2013-07-03 Chris Metcalf * sysdeps/tile/s_fma.c: Use soft-fp version instead of ieee754. * sysdeps/tile/s_fmaf.c: Likewise. * sysdeps/tile/sfp-machine.h: New file. ports/ChangeLog.tile | 6 ++ ports/sysdeps/tile/s_fma.c | 5 +-- ports/sysdeps/tile/s_fmaf.c | 5 +-- ports/sysdeps/tile/sfp-machine.h | 97 ++++++++++++++++++++++++++++++++++++++ 4 files changed, 105 insertions(+), 8 deletions(-) create mode 100644 ports/sysdeps/tile/sfp-machine.h diff --git a/ports/sysdeps/tile/s_fma.c b/ports/sysdeps/tile/s_fma.c index c84a9bb..d9613fa 100644 --- a/ports/sysdeps/tile/s_fma.c +++ b/ports/sysdeps/tile/s_fma.c @@ -1,4 +1 @@ -/* tile does not support the rounding modes required by the ieee754/dbl-64 - implementation, but it's still better than the generic implementation. */ - -#include +#include diff --git a/ports/sysdeps/tile/s_fmaf.c b/ports/sysdeps/tile/s_fmaf.c index 84e9adc..aa5c9b2 100644 --- a/ports/sysdeps/tile/s_fmaf.c +++ b/ports/sysdeps/tile/s_fmaf.c @@ -1,4 +1 @@ -/* tile does not support the rounding modes required by the ieee754/dbl-64 - implementation, but it's still better than the generic implementation. */ - -#include +#include diff --git a/ports/sysdeps/tile/sfp-machine.h b/ports/sysdeps/tile/sfp-machine.h new file mode 100644 index 0000000..3ddc4b4 --- /dev/null +++ b/ports/sysdeps/tile/sfp-machine.h @@ -0,0 +1,97 @@ +/* Machine-dependent software floating-point definitions, tile version. + Copyright (C) 2013 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 + . */ + +#include + +#define _FP_W_TYPE_SIZE __WORDSIZE +#define _FP_W_TYPE unsigned long +#define _FP_WS_TYPE signed long +#define _FP_I_TYPE long + +#if _FP_W_TYPE_SIZE == 64 + +#define _FP_MUL_MEAT_S(R,X,Y) \ + _FP_MUL_MEAT_1_imm(_FP_WFRACBITS_S,R,X,Y) +#define _FP_MUL_MEAT_D(R,X,Y) \ + _FP_MUL_MEAT_1_wide(_FP_WFRACBITS_D,R,X,Y,umul_ppmm) +#define _FP_MUL_MEAT_Q(R,X,Y) \ + _FP_MUL_MEAT_2_wide(_FP_WFRACBITS_Q,R,X,Y,umul_ppmm) + +#define _FP_MUL_MEAT_DW_S(R,X,Y) \ + _FP_MUL_MEAT_DW_1_imm(_FP_WFRACBITS_S,R,X,Y) +#define _FP_MUL_MEAT_DW_D(R,X,Y) \ + _FP_MUL_MEAT_DW_1_wide(_FP_WFRACBITS_D,R,X,Y,umul_ppmm) +#define _FP_MUL_MEAT_DW_Q(R,X,Y) \ + _FP_MUL_MEAT_DW_2_wide_3mul(_FP_WFRACBITS_Q,R,X,Y,umul_ppmm) + +#define _FP_DIV_MEAT_S(R,X,Y) _FP_DIV_MEAT_1_imm(S,R,X,Y,_FP_DIV_HELP_imm) +#define _FP_DIV_MEAT_D(R,X,Y) _FP_DIV_MEAT_1_udiv_norm(D,R,X,Y) +#define _FP_DIV_MEAT_Q(R,X,Y) _FP_DIV_MEAT_2_udiv(Q,R,X,Y) + +#define _FP_NANFRAC_S _FP_QNANBIT_S +#define _FP_NANFRAC_D _FP_QNANBIT_D +#define _FP_NANFRAC_Q _FP_QNANBIT_Q, 0 + +#else /* _FP_W_TYPE_SIZE == 32 */ + +#define _FP_MUL_MEAT_S(R,X,Y) \ + _FP_MUL_MEAT_1_wide(_FP_WFRACBITS_S,R,X,Y,umul_ppmm) +#define _FP_MUL_MEAT_D(R,X,Y) \ + _FP_MUL_MEAT_2_wide(_FP_WFRACBITS_D,R,X,Y,umul_ppmm) +#define _FP_MUL_MEAT_Q(R,X,Y) \ + _FP_MUL_MEAT_4_wide(_FP_WFRACBITS_Q,R,X,Y,umul_ppmm) + +#define _FP_MUL_MEAT_DW_S(R,X,Y) \ + _FP_MUL_MEAT_DW_1_wide(_FP_WFRACBITS_S,R,X,Y,umul_ppmm) +#define _FP_MUL_MEAT_DW_D(R,X,Y) \ + _FP_MUL_MEAT_DW_2_wide(_FP_WFRACBITS_D,R,X,Y,umul_ppmm) +#define _FP_MUL_MEAT_DW_Q(R,X,Y) \ + _FP_MUL_MEAT_DW_4_wide(_FP_WFRACBITS_Q,R,X,Y,umul_ppmm) + +#define _FP_DIV_MEAT_S(R,X,Y) _FP_DIV_MEAT_1_loop(S,R,X,Y) +#define _FP_DIV_MEAT_D(R,X,Y) _FP_DIV_MEAT_2_udiv(D,R,X,Y) +#define _FP_DIV_MEAT_Q(R,X,Y) _FP_DIV_MEAT_4_udiv(Q,R,X,Y) + +#define _FP_NANFRAC_S _FP_QNANBIT_S +#define _FP_NANFRAC_D _FP_QNANBIT_D, 0 +#define _FP_NANFRAC_Q _FP_QNANBIT_Q, 0, 0, 0 + +#endif + +#define _FP_NANSIGN_S 1 +#define _FP_NANSIGN_D 1 +#define _FP_NANSIGN_Q 1 + +#define _FP_KEEPNANFRACP 1 +#define _FP_QNANNEGATEDP 0 + +#define _FP_CHOOSENAN(fs, wc, R, X, Y, OP) \ + do { \ + if ((_FP_FRAC_HIGH_RAW_##fs(X) & _FP_QNANBIT_##fs) \ + && !(_FP_FRAC_HIGH_RAW_##fs(Y) & _FP_QNANBIT_##fs)) \ + { \ + R##_s = Y##_s; \ + _FP_FRAC_COPY_##wc(R,Y); \ + } \ + else \ + { \ + R##_s = X##_s; \ + _FP_FRAC_COPY_##wc(R,X); \ + } \ + R##_c = FP_CLS_NAN; \ + } while (0) -- 1.7.1