Mercurial > hg > sv-dependency-builds

comparison src/fftw-3.3.5/doc/FAQ/fftw-faq.html/section2.html @ 42:2cd0e3b3e1fd

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Current fftw source
author Chris Cannam
date Tue, 18 Oct 2016 13:40:26 +0100
parents
children
comparison
equal deleted inserted replaced
41:481f5f8c5634 42:2cd0e3b3e1fd
1 <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
2 <html>
3 <head><title>
4 FFTW FAQ - Section 2
5 </title>
6 <link rev="made" href="mailto:fftw@fftw.org">
7 <link rel="Contents" href="index.html">
8 <link rel="Start" href="index.html">
9 <link rel="Next" href="section3.html"><link rel="Previous" href="section1.html"><link rel="Bookmark" title="FFTW FAQ" href="index.html">
10 </head><body text="#000000" bgcolor="#FFFFFF"><h1>
11 FFTW FAQ - Section 2 <br>
12 Installing FFTW
13 </h1>
14
15 <ul>
16 <li><a href="#systems" rel=subdocument>Q2.1. Which systems does FFTW run on?</a>
17 <li><a href="#runOnWindows" rel=subdocument>Q2.2. Does FFTW run on Windows?</a>
18 <li><a href="#compilerCrashes" rel=subdocument>Q2.3. My compiler has trouble with FFTW.</a>
19 <li><a href="#solarisSucks" rel=subdocument>Q2.4. FFTW does not compile on Solaris, complaining about
20 <code>const</code>.</a>
21 <li><a href="#3dnow" rel=subdocument>Q2.5. What's the difference between <code>--enable-3dnow</code> and <code>--enable-k7</code>?</a>
22 <li><a href="#fma" rel=subdocument>Q2.6. What's the difference between the fma and the non-fma
23 versions?</a>
24 <li><a href="#languages" rel=subdocument>Q2.7. Which language is FFTW written in?</a>
25 <li><a href="#fortran" rel=subdocument>Q2.8. Can I call FFTW from Fortran?</a>
26 <li><a href="#cplusplus" rel=subdocument>Q2.9. Can I call FFTW from C++?</a>
27 <li><a href="#whynotfortran" rel=subdocument>Q2.10. Why isn't FFTW written in Fortran/C++?</a>
28 <li><a href="#singleprec" rel=subdocument>Q2.11. How do I compile FFTW to run in single precision?</a>
29 <li><a href="#64bitk7" rel=subdocument>Q2.12. --enable-k7 does not work on x86-64</a>
30 </ul><hr>
31
32 <h2><A name="systems">
33 Question 2.1. Which systems does FFTW run
34 on?
35 </A></h2>
36
37 FFTW is written in ANSI C, and should work on any system with a decent
38 C compiler. (See also <A href="#runOnWindows">Q2.2 `Does FFTW run on Windows?'</A>, <A href="#compilerCrashes">Q2.3 `My compiler has trouble with FFTW.'</A>.) FFTW can also take advantage of certain hardware-specific features,
39 such as cycle counters and SIMD instructions, but this is optional.
40
41 <h2><A name="runOnWindows">
42 Question 2.2. Does FFTW run on Windows?
43 </A></h2>
44
45 Yes, many people have reported successfully using FFTW on Windows with
46 various compilers. FFTW was not developed on Windows, but the source
47 code is essentially straight ANSI C. See also the
48 <A href="http://www.fftw.org/install/windows.html">FFTW Windows installation notes</A>, <A href="#compilerCrashes">Q2.3 `My compiler has trouble with FFTW.'</A>, and <A href="section3.html#vbetalia">Q3.18 `How do I call FFTW from the Microsoft language du
49 jour?'</A>.
50 <h2><A name="compilerCrashes">
51 Question 2.3. My compiler has trouble with
52 FFTW.
53 </A></h2>
54
55 Complain fiercely to the vendor of the compiler.
56
57 <p>
58 We have successfully used <code>gcc</code> 3.2.x on x86 and PPC, a recent Compaq C compiler for Alpha, version 6 of IBM's
59 <code>xlc</code> compiler for AIX, Intel's <code>icc</code> versions 5-7, and Sun WorkShop <code>cc</code> version 6.
60 <p>
61 FFTW is likely to push compilers to their limits, however, and several
62 compiler bugs have been exposed by FFTW. A partial list follows.
63
64 <p>
65 <code>gcc</code> 2.95.x for Solaris/SPARC produces incorrect code for
66 the test program (workaround: recompile the
67 <code>libbench2</code> directory with <code>-O2</code>).
68 <p>
69 NetBSD/macppc 1.6 comes with a <code>gcc</code> version that also miscompiles the test program. (Please report a workaround if you know
70 one.)
71 <p>
72 <code>gcc</code> 3.2.3 for ARM reportedly crashes during compilation.
73 This bug is reportedly fixed in later versions of
74 <code>gcc</code>.
75 <p>
76 Versions 8.0 and 8.1 of Intel's <code>icc</code> falsely claim to be <code>gcc</code>, so you should specify <code>CC=&quot;icc -no-gcc&quot;</code>; this is automatic in FFTW 3.1. <code>icc-8.0.066</code> reportely produces incorrect code for FFTW 2.1.5, but is fixed in version 8.1.
77 <code>icc-7.1</code> compiler build 20030402Z appears to produce
78 incorrect dependencies, causing the compilation to fail.
79 <code>icc-7.1</code> build 20030307Z appears to work fine. (Use
80 <code>icc -V</code> to check which build you have.) As of 2003/04/18,
81 build 20030402Z appears not to be available any longer on Intel's
82 website, whereas the older build 20030307Z is available.
83
84 <p>
85 <code>ranlib</code> of GNU <code>binutils</code> 2.9.1 on Irix has been observed to corrupt the FFTW libraries, causing a link failure when
86 FFTW is compiled. Since <code>ranlib</code> is completely superfluous on Irix, we suggest deleting it from your system and replacing it with
87 a symbolic link to <code>/bin/echo</code>.
88 <p>
89 If support for SIMD instructions is enabled in FFTW, further compiler
90 problems may appear:
91 <p>
92 <code>gcc</code> 3.4.[0123] for x86 produces incorrect SSE2 code for
93 FFTW when <code>-O2</code> (the best choice for FFTW) is used, causing
94 FFTW to crash (<code>make check</code> crashes). This bug is fixed in <code>gcc</code> 3.4.4. On x86_64 (amd64/em64t), <code>gcc</code> 3.4.4 reportedly still has a similar problem, but this is fixed as of
95 <code>gcc</code> 3.4.6.
96 <p>
97 <code>gcc-3.2</code> for x86 produces incorrect SIMD code if
98 <code>-O3</code> is used. The same compiler produces incorrect SIMD
99 code if no optimization is used, too. When using
100 <code>gcc-3.2</code>, it is a good idea not to change the default
101 <code>CFLAGS</code> selected by the <code>configure</code> script.
102 <p>
103 Some 3.0.x and 3.1.x versions of <code>gcc</code> on <code>x86</code> may crash. <code>gcc</code> so-called 2.96 shipping with RedHat 7.3 crashes
104 when compiling SIMD code. In both cases, please upgrade to
105 <code>gcc-3.2</code> or later.
106 <p>
107 Intel's <code>icc</code> 6.0 misaligns SSE constants, but FFTW has a
108 workaround. <code>icc</code> 8.x fails to compile FFTW 3.0.x because it
109 falsely claims to be <code>gcc</code>; we believe this to be a bug in <code>icc</code>, but FFTW 3.1 has a workaround.
110 <p>
111 Visual C++ 2003 reportedly produces incorrect code for SSE/SSE2 when
112 compiling FFTW. This bug was reportedly fixed in VC++ 2005;
113 alternatively, you could switch to the Intel compiler. VC++ 6.0 also
114 reportedly produces incorrect code for the file
115 <code>reodft11e-r2hc-odd.c</code> unless optimizations are disabled for that file.
116 <p>
117 <code>gcc</code> 2.95 on MacOS X miscompiles AltiVec code (fixed in
118 later versions). <code>gcc</code> 3.2.x miscompiles AltiVec permutations, but FFTW has a workaround.
119 <code>gcc</code> 4.0.1 on MacOS for Intel crashes when compiling FFTW; a workaround is to
120 compile one file without optimization: <code>cd kernel; make CFLAGS=&quot; &quot; trig.lo</code>.
121 <p>
122 <code>gcc</code> 4.1.1 reportedly crashes when compiling FFTW for MIPS;
123 the workaround is to compile the file it crashes on
124 (<code>t2_64.c</code>) with a lower optimization level.
125 <p>
126 <code>gcc</code> versions 4.1.2 to 4.2.0 for x86 reportedly miscompile
127 FFTW 3.1's test program, causing <code>make check</code> to crash (<code>gcc</code> bug #26528). The bug was reportedly fixed in
128 <code>gcc</code> version 4.2.1 and later. A workaround is to compile
129 <code>libbench2/verify-lib.c</code> without optimization.
130 <h2><A name="solarisSucks">
131 Question 2.4. FFTW does not compile on Solaris, complaining about
132 <code>const</code>.
133 </A></h2>
134
135 We know that at least on Solaris 2.5.x with Sun's compilers 4.2 you
136 might get error messages from <code>make</code> such as
137 <p>
138 <code>&quot;./fftw.h&quot;, line 88: warning: const is a keyword in ANSI
139 C</code>
140 <p>
141 This is the case when the <code>configure</code> script reports that <code>const</code> does not work:
142 <p>
143 <code>checking for working const... (cached) no</code>
144 <p>
145 You should be aware that Solaris comes with two compilers, namely,
146 <code>/opt/SUNWspro/SC4.2/bin/cc</code> and <code>/usr/ucb/cc</code>. The latter compiler is non-ANSI. Indeed, it is a perverse shell script
147 that calls the real compiler in non-ANSI mode. In order
148 to compile FFTW, change your path so that the right
149 <code>cc</code> is used.
150 <p>
151 To know whether your compiler is the right one, type
152 <code>cc -V</code>. If the compiler prints ``<code>ucbcc</code>'', as in
153 <p>
154 <code>ucbcc: WorkShop Compilers 4.2 30 Oct 1996 C
155 4.2</code>
156 <p>
157 then the compiler is wrong. The right message is something like
158
159 <p>
160 <code>cc: WorkShop Compilers 4.2 30 Oct 1996 C
161 4.2</code>
162 <h2><A name="3dnow">
163 Question 2.5. What's the difference between
164 <code>--enable-3dnow</code> and <code>--enable-k7</code>?
165 </A></h2>
166
167 <code>--enable-k7</code> enables 3DNow! instructions on K7 processors
168 (AMD Athlon and its variants). K7 support is provided by assembly
169 routines generated by a special purpose compiler.
170 As of fftw-3.2, --enable-k7 is no longer supported.
171
172 <p>
173 <code>--enable-3dnow</code> enables generic 3DNow! support using <code>gcc</code> builtin functions. This works on earlier AMD
174 processors, but it is not as fast as our special assembly routines.
175 As of fftw-3.1, --enable-3dnow is no longer supported.
176
177 <h2><A name="fma">
178 Question 2.6. What's the difference between the fma and the non-fma
179 versions?
180 </A></h2>
181
182 The fma version tries to exploit the fused multiply-add instructions
183 implemented in many processors such as PowerPC, ia-64, and MIPS. The
184 two FFTW packages are otherwise identical. In FFTW 3.1, the fma and
185 non-fma versions were merged together into a single package, and the
186 <code>configure</code> script attempts to automatically guess which
187 version to use.
188 <p>
189 The FFTW 3.1 <code>configure</code> script enables fma by default on PowerPC, Itanium, and PA-RISC, and disables it otherwise. You can
190 force one or the other by using the <code>--enable-fma</code> or <code>--disable-fma</code> flag for <code>configure</code>.
191 <p>
192 Definitely use fma if you have a PowerPC-based system with
193 <code>gcc</code> (or IBM <code>xlc</code>). This includes all GNU/Linux systems for PowerPC and the older PowerPC-based MacOS systems. Also
194 use it on PA-RISC and Itanium with the HP/UX compiler.
195
196 <p>
197 Definitely do not use the fma version if you have an ia-32 processor
198 (Intel, AMD, MacOS on Intel, etcetera).
199
200 <p>
201 For other architectures/compilers, the situation is not so clear. For
202 example, ia-64 has the fma instruction, but
203 <code>gcc-3.2</code> appears not to exploit it correctly. Other compilers may do the right thing,
204 but we have not tried them. Please send us your feedback so that we
205 can update this FAQ entry.
206 <h2><A name="languages">
207 Question 2.7. Which language is FFTW written
208 in?
209 </A></h2>
210
211 FFTW is written in ANSI C. Most of the code, however, was
212 automatically generated by a program called
213 <code>genfft</code>, written in the Objective Caml dialect of ML. You do not need to know ML or to
214 have an Objective Caml compiler in order to use FFTW.
215
216 <p>
217 <code>genfft</code> is provided with the FFTW sources, which means that
218 you can play with the code generator if you want. In this case, you
219 need a working Objective Caml system. Objective Caml is available
220 from <A href="http://caml.inria.fr">the Caml web page</A>.
221 <h2><A name="fortran">
222 Question 2.8. Can I call FFTW from Fortran?
223 </A></h2>
224
225 Yes, FFTW (versions 1.3 and higher) contains a Fortran-callable
226 interface, documented in the FFTW manual.
227
228 <p>
229 By default, FFTW configures its Fortran interface to work with the
230 first compiler it finds, e.g. <code>g77</code>. To configure for a different, incompatible Fortran compiler
231 <code>foobar</code>, use <code>./configure F77=foobar</code> when installing FFTW. (In the case of <code>g77</code>, however, FFTW 3.x also includes an extra set of
232 Fortran-callable routines with one less underscore at the end of
233 identifiers, which should cover most other Fortran compilers on Linux
234 at least.)
235 <h2><A name="cplusplus">
236 Question 2.9. Can I call FFTW from C++?
237 </A></h2>
238
239 Most definitely. FFTW should compile and/or link under any C++
240 compiler. Moreover, it is likely that the C++
241 <code>&lt;complex&gt;</code> template class is bit-compatible with FFTW's complex-number format
242 (see the FFTW manual for more details).
243
244 <h2><A name="whynotfortran">
245 Question 2.10. Why isn't FFTW written in
246 Fortran/C++?
247 </A></h2>
248
249 Because we don't like those languages, and neither approaches the
250 portability of C.
251 <h2><A name="singleprec">
252 Question 2.11. How do I compile FFTW to run in single
253 precision?
254 </A></h2>
255
256 On a Unix system: <code>configure --enable-float</code>. On a non-Unix system: edit <code>config.h</code> to <code>#define</code> the symbol <code>FFTW_SINGLE</code> (for FFTW 3.x). In both cases, you must then
257 recompile FFTW. In FFTW 3, all FFTW identifiers will then begin with
258 <code>fftwf_</code> instead of <code>fftw_</code>.
259 <h2><A name="64bitk7">
260 Question 2.12. --enable-k7 does not work on
261 x86-64
262 </A></h2>
263
264 Support for --enable-k7 was discontinued in fftw-3.2.
265
266 <p>
267 The fftw-3.1 release supports --enable-k7. This option only works on
268 32-bit x86 machines that implement 3DNow!, including the AMD Athlon
269 and the AMD Opteron in 32-bit mode. --enable-k7 does not work on AMD
270 Opteron in 64-bit mode. Use --enable-sse for x86-64 machines.
271
272 <p>
273 FFTW supports 3DNow! by means of assembly code generated by a
274 special-purpose compiler. It is hard to produce assembly code that
275 works in both 32-bit and 64-bit mode. <hr>
276 Next: <a href="section3.html" rel=precedes>Using FFTW</a>.<br>
277 Back: <a href="section1.html" rev=precedes>Introduction and General Information</a>.<br>
278 <a href="index.html" rev=subdocument>Return to contents</a>.<p>
279 <address>
280 <A href="http://www.fftw.org">Matteo Frigo and Steven G. Johnson</A> / <A href="mailto:fftw@fftw.org">fftw@fftw.org</A>
281 - 30 July 2016
282 </address><br>
283 Extracted from FFTW Frequently Asked Questions with Answers,
284 Copyright &copy; 2016 Matteo Frigo and Massachusetts Institute of Technology.
285 </body></html>