[Gmp-commit] /var/hg/gmp: 3 new changesets

[mercurial at gmplib.org]

details:   /var/hg/gmp/rev/233a090d640f
changeset: 15733:233a090d640f
user:      Torbjorn Granlund <tege at gmplib.org>
date:      Wed Apr 17 17:04:54 2013 +0200
description:
(arm64 count_trailing_zeros): New.

details:   /var/hg/gmp/rev/feca99cc7d06
changeset: 15734:feca99cc7d06
user:      Torbjorn Granlund <tege at gmplib.org>
date:      Wed Apr 17 17:05:39 2013 +0200
description:
ARM64 invert_limb.

details:   /var/hg/gmp/rev/3d08c3752df9
changeset: 15735:3d08c3752df9
user:      Torbjorn Granlund <tege at gmplib.org>
date:      Wed Apr 17 17:05:56 2013 +0200
description:
ChangeLog

diffstat:

 ChangeLog                 |   4 ++
 longlong.h                |   2 +
 mpn/arm64/invert_limb.asm |  72 +++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 78 insertions(+), 0 deletions(-)

diffs (101 lines):

diff -r 6a00316342da -r 3d08c3752df9 ChangeLog
--- a/ChangeLog	Wed Apr 17 14:43:51 2013 +0200
+++ b/ChangeLog	Wed Apr 17 17:05:56 2013 +0200
@@ -1,5 +1,9 @@
 2013-04-17  Torbjorn Granlund  <tege at gmplib.org>
 
+	* longlong.h (arm64 count_trailing_zeros): New.
+
+	* mpn/arm64/invert_limb.asm: New file.
+
 	* mpn/generic/dive_1.c: Rewrite to use Hensel division also for
 	size = 1.
 
diff -r 6a00316342da -r 3d08c3752df9 longlong.h
--- a/longlong.h	Wed Apr 17 14:43:51 2013 +0200
+++ b/longlong.h	Wed Apr 17 17:05:56 2013 +0200
@@ -548,6 +548,8 @@
   } while (0)
 #define count_leading_zeros(count, x) \
   __asm__ ("clz\t%0, %1" : "=r" (count) : "r" (x))
+#define count_trailing_zeros(count, x) \
+  __asm__ ("rbit\t%0, %1\n\tclz\t%0, %0" : "=r" (count) : "r" (x))
 #define COUNT_LEADING_ZEROS_0 64
 #endif /* __aarch64__ */
 
diff -r 6a00316342da -r 3d08c3752df9 mpn/arm64/invert_limb.asm
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/mpn/arm64/invert_limb.asm	Wed Apr 17 17:05:56 2013 +0200
@@ -0,0 +1,72 @@
+dnl  ARM64 mpn_invert_limb -- Invert a normalized limb.
+
+dnl  Contributed to the GNU project by Torbjörn Granlund.
+
+dnl  Copyright 2013 Free Software Foundation, Inc.
+
+dnl  This file is part of the GNU MP Library.
+
+dnl  The GNU MP Library is free software; you can redistribute it and/or modify
+dnl  it under the terms of the GNU Lesser General Public License as published
+dnl  by the Free Software Foundation; either version 3 of the License, or (at
+dnl  your option) any later version.
+
+dnl  The GNU MP Library is distributed in the hope that it will be useful, but
+dnl  WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
+dnl  or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public
+dnl  License for more details.
+
+dnl  You should have received a copy of the GNU Lesser General Public License
+dnl  along with the GNU MP Library.  If not, see http://www.gnu.org/licenses/.
+
+include(`../config.m4')
+
+C            cycles/limb
+C Cortex-A53     ?
+C Cortex-A57     ?
+
+C Compiler generated, mildly edited.  Could surely be further optimised.
+
+ASM_START()
+PROLOGUE(mpn_invert_limb)
+	lsr	x2, x0, 54
+	adrp	x1, approx_tab
+	and	x2, x2, #0x1fe
+	add	x1, x1, :lo12:approx_tab
+	ldrh	w3, [x1,x2]
+	lsr	x4, x0, 24
+	add	x4, x4, 1
+	ubfiz	x2, x3, 11, 16
+	umull	x3, w3, w3
+	mul	x3, x3, x4
+	sub	x2, x2, #1
+	sub	x2, x2, x3, lsr 40
+	lsl	x3, x2, 60
+	mul	x1, x2, x2
+	msub	x1, x1, x4, x3
+	lsl	x2, x2, 13
+	add	x1, x2, x1, lsr 47
+	and	x2, x0, 1
+	neg	x3, x2
+	and	x3, x3, x1, lsr 1
+	add	x2, x2, x0, lsr 1
+	msub	x2, x1, x2, x3
+	umulh	x2, x2, x1
+	lsl	x1, x1, 31
+	add	x1, x1, x2, lsr 1
+	mul	x3, x1, x0
+	umulh	x2, x1, x0
+	adds	x4, x3, x0
+	adc	x0, x2, x0
+	sub	x0, x1, x0
+	ret
+EPILOGUE()
+
+	RODATA
+	ALIGN(2)
+	TYPE(   approx_tab, object)
+	SIZE(   approx_tab, 512)
+approx_tab:
+forloop(i,256,512-1,dnl
+`	.hword	eval(0x7fd00/i)
+')dnl