abc/GmSSL
1
0
Fork 0
mirror of https://github.com/guanzhi/GmSSL.git synced 2026-08-14 01:53:41 +08:00
Code Issues Packages Projects Releases Wiki Activity

quantum init

Browse Source
This commit is contained in:
zhaoxiaomeng
2018-01-04 13:38:57 +08:00
committed by Simon
parent d11f845fde
commit 53af3b51ae
2361 changed files with 387455 additions and 144458 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
crypto/bn/asm/ia64.S
View File

@@ -3,13 +3,6 @@
.ident "ia64.S, Version 2.1"
.ident "IA-64 ISA artwork by Andy Polyakov <appro@fy.chalmers.se>"
// Copyright 2001-2016 The OpenSSL Project Authors. All Rights Reserved.
//
// Licensed under the OpenSSL license (the "License"). You may not use
// this file except in compliance with the License. You can obtain a copy
// in the file LICENSE in the source distribution or at
// https://www.openssl.org/source/license.html
//
// ====================================================================
// Written by Andy Polyakov <appro@fy.chalmers.se> for the OpenSSL
@@ -29,7 +22,7 @@
// ports is the same, i.e. 2, while I need 4. In other words, to this
// module Itanium2 remains effectively as "wide" as Itanium. Yet it's
// essentially different in respect to this module, and a re-tune was
// required. Well, because some instruction latencies has changed. Most
// required. Well, because some intruction latencies has changed. Most
// noticeably those intensively used:
//
// Itanium Itanium2
@@ -370,7 +363,7 @@ bn_mul_words:
// The loop therefore spins at the latency of xma minus 1, or in other
// words at 6*(n+4) ticks:-( Compare to the "production" loop above
// that runs in 2*(n+11) where the low latency problem is worked around
// by moving the dependency to one-tick latent integer ALU. Note that
// by moving the dependency to one-tick latent interger ALU. Note that
// "distance" between ldf8 and xma is not latency of ldf8, but the
// *difference* between xma and ldf8 latencies.
.L_bn_mul_words_ctop:
@@ -432,7 +425,7 @@ bn_mul_add_words:
// version was performing *all* additions in IALU and was starving
// for those even on Itanium 2. In this version one addition is
// moved to FPU and is folded with multiplication. This is at cost
// of propagating the result from previous call to this subroutine
// of propogating the result from previous call to this subroutine
// to L2 cache... In other words negligible even for shorter keys.
// *Overall* performance improvement [over previous version] varies
// from 11 to 22 percent depending on key length.
@@ -502,7 +495,7 @@ bn_sqr_words:
// scalability. The decision will very likely be reconsidered after the
// benchmark program is profiled. I.e. if perfomance gain on Itanium
// will appear larger than loss on "wider" IA-64, then the loop should
// be explicitly split and the epilogue compressed.
// be explicitely split and the epilogue compressed.
.L_bn_sqr_words_ctop:
{ .mfi; (p16) ldf8 f32=[r33],8
(p25) xmpy.lu f42=f41,f41