aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorChris Metcalf <cmetcalf@tilera.com>2013-07-03 11:23:01 -0400
committerChris Metcalf <cmetcalf@tilera.com>2013-07-07 11:33:24 -0400
commit36002f0786631a0f923ee2c60fb09a4fdba85ead (patch)
tree643d61f49535b976fe629f9895d877fe54db20e4
parent72c90ed01fb5ba17bbcfbe0a67ec044f4b868618 (diff)
downloadglibc-36002f0786631a0f923ee2c60fb09a4fdba85ead.tar
glibc-36002f0786631a0f923ee2c60fb09a4fdba85ead.tar.gz
glibc-36002f0786631a0f923ee2c60fb09a4fdba85ead.tar.bz2
glibc-36002f0786631a0f923ee2c60fb09a4fdba85ead.zip
tile: use soft-fp for fma() and fmaf()
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.
-rw-r--r--ports/ChangeLog.tile6
-rw-r--r--ports/sysdeps/tile/s_fma.c5
-rw-r--r--ports/sysdeps/tile/s_fmaf.c5
-rw-r--r--ports/sysdeps/tile/sfp-machine.h97
4 files changed, 105 insertions, 8 deletions
diff --git a/ports/ChangeLog.tile b/ports/ChangeLog.tile
index e5a92fc426..079b35c1db 100644
--- a/ports/ChangeLog.tile
+++ b/ports/ChangeLog.tile
@@ -1,3 +1,9 @@
+2013-07-03 Chris Metcalf <cmetcalf@tilera.com>
+
+ * 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.
+
2013-07-04 Andreas Jaeger <aj@suse.de>
* sysdeps/unix/sysv/linux/tile/sys/ptrace.h (PTRACE_PEEKSIGINFO):
diff --git a/ports/sysdeps/tile/s_fma.c b/ports/sysdeps/tile/s_fma.c
index c84a9bb2df..d9613fa67c 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 <sysdeps/ieee754/dbl-64/s_fma.c>
+#include <soft-fp/fmadf4.c>
diff --git a/ports/sysdeps/tile/s_fmaf.c b/ports/sysdeps/tile/s_fmaf.c
index 84e9adc41a..aa5c9b2d91 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 <sysdeps/ieee754/dbl-64/s_fmaf.c>
+#include <soft-fp/fmasf4.c>
diff --git a/ports/sysdeps/tile/sfp-machine.h b/ports/sysdeps/tile/sfp-machine.h
new file mode 100644
index 0000000000..3ddc4b4c83
--- /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
+ <http://www.gnu.org/licenses/>. */
+
+#include <bits/wordsize.h>
+
+#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)