/usr/share/doc/libntl-dev/NTL/tour-tips.html is in libntl-dev 5.5.2-2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 | <html>
<head>
<title>
A Tour of NTL: Tips for Getting the Best Performance out of NTL
</title>
</head>
<body bgcolor="#fff9e6">
<center>
<a href="tour-win.html"><img src="arrow1.gif" alt="[Previous]" align=bottom></a>
<a href="tour.html"><img src="arrow2.gif" alt="[Up]" align=bottom></a>
<a href="tour-impl.html"> <img src="arrow3.gif" alt="[Next]" align=bottom></a>
</center>
<h1>
<p align=center>
A Tour of NTL: Tips for Getting the Best Performance out of NTL
</p>
</h1>
<p> <hr> <p>
<ol>
<li>
Build NTL using GMP as the primary long integer package.
This is extremely important, as the GMP implementation
of long integer arithmetic is <i>much</i> faster
than the default implementation.
Go <a href="tour-gmp.html">here</a> for details.
<p>
<li>
On many machines that optionally offer 64-bit integer arithmetic
(recent Mac OSX machines, for instance),
you should
compile using <tt>gcc</tt> with the option <tt>-m64</tt>
to get the full benefit.
To do this,
pass <tt>"CFLAGS=-O2 -m64"</tt>
to the <tt>configure</tt> script (note the use of quotes).
If you are using NTL with GMP on such a machine,
you <i>must</i> do this to get compatible code.
<p>
<li>
On Sparcs,
pass the argument <tt>"CFLAGS=-O2 -mcpu=v8"</tt>
to the <tt>configure</tt> script.
On more recent, 64-bit sparcs, pass <tt>"CFLAGS=-O2 -mcpu=v9 -m64"</tt>
to get the full instruction set and 64-bit code.
<p>
<li>
Make sure you run the configuration wizard when you install NTL.
This is the default behaviour in the makefile
in the Unix distribution, so don't change this;
in the Windows distribution, there is unfortunately no
easy way to run the wizard.
<p>
<li>
In time-critical code, avoid creating unnecessary temporary
objects.
For example, instead of
<pre>
ZZ InnerProduct(const ZZ *a, const ZZ *b, long n)
{
long i;
ZZ res;
for (i = 0; i < n; i++)
res += a[i] * b[i];
return res;
}
</pre>
write this as
<pre>
ZZ InnerProduct(const ZZ *a, const ZZ *b, long n)
{
long i;
ZZ res, t;
for (i = 0; i < n; i++) {
mul(t, a[i], b[i]);
add(res, res, t);
}
return res;
}
</pre>
The first version of <tt>InnerProduct</tt>
creates and destroys a temporary object, holding the value
<tt>a[i]*b[i]</tt>, in every loop iteration.
The second does not.
<p>
<li>
If you use the class <tt>ZZ_p</tt>, try to avoid switching the modulus
too often, as this can be a rather expensive operation.
If you <i>must</i> switch the modulus often,
use the class <tt>ZZ_pContext</tt> to save the information
associated with the modulus (see <a href="ZZ.txt">ZZ.txt</a>).
</ol>
<p>
<center>
<a href="tour-win.html"><img src="arrow1.gif" alt="[Previous]" align=bottom></a>
<a href="tour.html"><img src="arrow2.gif" alt="[Up]" align=bottom></a>
<a href="tour-impl.html"> <img src="arrow3.gif" alt="[Next]" align=bottom></a>
</center>
</body>
</html>
|