From: Richard Henderson <rth@twiddle.net>
To: "Joseph S. Myers" <joseph@codesourcery.com>
Cc: libc-alpha@sourceware.org, libc-ports@sourceware.org,
Marcus Shawcroft <marcus.shawcroft@linaro.org>,
Richard Henderson <rth@redhat.com>,
David Holsgrove <david.holsgrove@xilinx.com>,
Chris Metcalf <cmetcalf@tilera.com>,
"David S. Miller" <davem@davemloft.net>
Subject: Re: Implement fma in soft-fp
Date: Tue, 25 Jun 2013 16:31:00 -0000 [thread overview]
Message-ID: <51C9C5F5.1070809@twiddle.net> (raw)
In-Reply-To: <Pine.LNX.4.64.1306221707230.15367@digraph.polyomino.org.uk>
[-- Attachment #1: Type: text/plain, Size: 1321 bytes --]
With the following patch, I get
> testing long double (without inline functions)
> Failure: fma (0x1.fffffffffffffffffffffffffffcp-16382, 0x1.0000000000000000000000000001p-1, 0x1p-16494): Exception "Underflow" set
> Failure: fma (-0x1.fffffffffffffffffffffffffffcp-16382, 0x1.0000000000000000000000000001p-1, -0x1p-16494): Exception "Underflow" set
> Failure: fma (0x1p-16494, -0x1p-16494, 0x1p-16382): Exception "Underflow" set
> Failure: fma (0x1p-16494, 0x1p-16494, -0x1p-16382): Exception "Underflow" set
> Failure: fma_downward (-0x1.fffffffffffffffffffffffffffcp-16382, 0x1.0000000000000000000000000001p-1, -0x1p-16494): Exception "Underflow" set
> Failure: fma_downward (-0x1p-16494, 0x1.1p-1, -0x0.ffffffffffffffffffffffffffffp-16382): Exception "Underflow" set
> Failure: fma_downward (0x1p-16494, 0x1p-16494, -0x1p-16382): Exception "Underflow" set
> Failure: fma_upward (0x1.fffffffffffffffffffffffffffcp-16382, 0x1.0000000000000000000000000001p-1, 0x1p-16494): Exception "Underflow" set
> Failure: fma_upward (0x1p-16494, 0x1.1p-1, 0x0.ffffffffffffffffffffffffffffp-16382): Exception "Underflow" set
> Failure: fma_upward (0x1p-16494, -0x1p-16494, 0x1p-16382): Exception "Underflow" set
but I thought you'd already committed the patch to solve this
as commit 695c378f81263640618bdebf56eaa065f578251f ?
r~
[-- Attachment #2: z --]
[-- Type: text/plain, Size: 3765 bytes --]
diff --git a/ports/sysdeps/alpha/Implies b/ports/sysdeps/alpha/Implies
index d03783b..35ccba3 100644
--- a/ports/sysdeps/alpha/Implies
+++ b/ports/sysdeps/alpha/Implies
@@ -1,7 +1,7 @@
wordsize-64
+alpha/soft-fp
# Alpha uses IEEE 754 single, double and quad precision floating point.
ieee754/ldbl-128
ieee754/dbl-64/wordsize-64
ieee754/dbl-64
ieee754/flt-32
-alpha/soft-fp
diff --git a/ports/sysdeps/alpha/soft-fp/s_fmal.c b/ports/sysdeps/alpha/soft-fp/s_fmal.c
new file mode 100644
index 0000000..8eae7fb
--- /dev/null
+++ b/ports/sysdeps/alpha/soft-fp/s_fmal.c
@@ -0,0 +1,52 @@
+/* Implement fmal using soft-fp.
+ 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.
+
+ In addition to the permissions in the GNU Lesser General Public
+ License, the Free Software Foundation gives you unlimited
+ permission to link the compiled version of this file into
+ combinations with other programs, and to distribute those
+ combinations without any restriction coming from the use of this
+ file. (The Lesser General Public License restrictions do apply in
+ other respects; for example, they cover modification of the file,
+ and distribution when not linked into a combine executable.)
+
+ 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 <math.h>
+#include <soft-fp.h>
+#include <quad.h>
+#include <math_ldbl_opt.h>
+
+long double
+__fmal (long double a, long double b, long double c)
+{
+ FP_DECL_EX;
+ FP_DECL_Q(A); FP_DECL_Q(B); FP_DECL_Q(C); FP_DECL_Q(R);
+ long double r;
+ long _round = 4; /* dynamic rounding */
+
+ FP_INIT_ROUNDMODE;
+ FP_UNPACK_Q(A, a);
+ FP_UNPACK_Q(B, b);
+ FP_UNPACK_Q(C, c);
+ FP_FMA_Q(R, A, B, C);
+ FP_PACK_Q(r, R);
+ FP_HANDLE_EXCEPTIONS;
+
+ return r;
+}
+
+long_double_symbol (libm, __fmal, fmal);
diff --git a/ports/sysdeps/alpha/soft-fp/sfp-machine.h b/ports/sysdeps/alpha/soft-fp/sfp-machine.h
index be266fe..98e9850 100644
--- a/ports/sysdeps/alpha/soft-fp/sfp-machine.h
+++ b/ports/sysdeps/alpha/soft-fp/sfp-machine.h
@@ -34,6 +34,13 @@
#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(_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)
diff --git a/ports/sysdeps/unix/sysv/linux/alpha/Implies b/ports/sysdeps/unix/sysv/linux/alpha/Implies
index 1616efe..2a14f00 100644
--- a/ports/sysdeps/unix/sysv/linux/alpha/Implies
+++ b/ports/sysdeps/unix/sysv/linux/alpha/Implies
@@ -1,4 +1,6 @@
unix/sysv/linux/wordsize-64
+# Select the soft-fp versions of some long double implementations.
+alpha/soft-fp
# These supply the ABI compatibility for when long double was double.
ieee754/ldbl-64-128
ieee754/ldbl-opt
next prev parent reply other threads:[~2013-06-25 16:31 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-06-22 17:17 Joseph S. Myers
2013-06-25 15:30 ` Richard Henderson
2013-06-25 16:13 ` Joseph S. Myers
2013-06-25 16:23 ` Richard Henderson
2013-06-25 16:31 ` Richard Henderson [this message]
2013-06-25 16:45 ` Joseph S. Myers
2013-06-27 23:26 ` Joseph S. Myers
2013-07-02 12:57 ` Ping " Joseph S. Myers
2013-07-02 14:45 ` Richard Henderson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=51C9C5F5.1070809@twiddle.net \
--to=rth@twiddle.net \
--cc=cmetcalf@tilera.com \
--cc=davem@davemloft.net \
--cc=david.holsgrove@xilinx.com \
--cc=joseph@codesourcery.com \
--cc=libc-alpha@sourceware.org \
--cc=libc-ports@sourceware.org \
--cc=marcus.shawcroft@linaro.org \
--cc=rth@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).