1 |
=head1 LIBECB - e-C-Builtins |
2 |
|
3 |
=head2 ABOUT LIBECB |
4 |
|
5 |
Libecb is currently a simple header file that doesn't require any |
6 |
configuration to use or include in your project. |
7 |
|
8 |
It's part of the e-suite of libraries, other members of which include |
9 |
libev and libeio. |
10 |
|
11 |
Its homepage can be found here: |
12 |
|
13 |
http://software.schmorp.de/pkg/libecb |
14 |
|
15 |
It mainly provides a number of wrappers around GCC built-ins, together |
16 |
with replacement functions for other compilers. In addition to this, |
17 |
it provides a number of other lowlevel C utilities, such as endianness |
18 |
detection, byte swapping or bit rotations. |
19 |
|
20 |
Or in other words, things that should be built into any standard C system, |
21 |
but aren't, implemented as efficient as possible with GCC, and still |
22 |
correct with other compilers. |
23 |
|
24 |
More might come. |
25 |
|
26 |
=head2 ABOUT THE HEADER |
27 |
|
28 |
At the moment, all you have to do is copy F<ecb.h> somewhere where your |
29 |
compiler can find it and include it: |
30 |
|
31 |
#include <ecb.h> |
32 |
|
33 |
The header should work fine for both C and C++ compilation, and gives you |
34 |
all of F<inttypes.h> in addition to the ECB symbols. |
35 |
|
36 |
There are currently no object files to link to - future versions might |
37 |
come with an (optional) object code library to link against, to reduce |
38 |
code size or gain access to additional features. |
39 |
|
40 |
It also currently includes everything from F<inttypes.h>. |
41 |
|
42 |
=head2 ABOUT THIS MANUAL / CONVENTIONS |
43 |
|
44 |
This manual mainly describes each (public) function available after |
45 |
including the F<ecb.h> header. The header might define other symbols than |
46 |
these, but these are not part of the public API, and not supported in any |
47 |
way. |
48 |
|
49 |
When the manual mentions a "function" then this could be defined either as |
50 |
as inline function, a macro, or an external symbol. |
51 |
|
52 |
When functions use a concrete standard type, such as C<int> or |
53 |
C<uint32_t>, then the corresponding function works only with that type. If |
54 |
only a generic name is used (C<expr>, C<cond>, C<value> and so on), then |
55 |
the corresponding function relies on C to implement the correct types, and |
56 |
is usually implemented as a macro. Specifically, a "bool" in this manual |
57 |
refers to any kind of boolean value, not a specific type. |
58 |
|
59 |
=head2 TYPES / TYPE SUPPORT |
60 |
|
61 |
ecb.h makes sure that the following types are defined (in the expected way): |
62 |
|
63 |
int8_t uint8_t int16_t uint16_t |
64 |
int32_t uint32_t int64_t uint64_t |
65 |
intptr_t uintptr_t |
66 |
|
67 |
The macro C<ECB_PTRSIZE> is defined to the size of a pointer on this |
68 |
platform (currently C<4> or C<8>) and can be used in preprocessor |
69 |
expressions. |
70 |
|
71 |
For C<ptrdiff_t> and C<size_t> use C<stddef.h>. |
72 |
|
73 |
=head2 LANGUAGE/COMPILER VERSIONS |
74 |
|
75 |
All the following symbols expand to an expression that can be tested in |
76 |
preprocessor instructions as well as treated as a boolean (use C<!!> to |
77 |
ensure it's either C<0> or C<1> if you need that). |
78 |
|
79 |
=over 4 |
80 |
|
81 |
=item ECB_C |
82 |
|
83 |
True if the implementation defines the C<__STDC__> macro to a true value, |
84 |
while not claiming to be C++. |
85 |
|
86 |
=item ECB_C99 |
87 |
|
88 |
True if the implementation claims to be compliant to C99 (ISO/IEC |
89 |
9899:1999) or any later version, while not claiming to be C++. |
90 |
|
91 |
Note that later versions (ECB_C11) remove core features again (for |
92 |
example, variable length arrays). |
93 |
|
94 |
=item ECB_C11 |
95 |
|
96 |
True if the implementation claims to be compliant to C11 (ISO/IEC |
97 |
9899:2011) or any later version, while not claiming to be C++. |
98 |
|
99 |
=item ECB_CPP |
100 |
|
101 |
True if the implementation defines the C<__cplusplus__> macro to a true |
102 |
value, which is typically true for C++ compilers. |
103 |
|
104 |
=item ECB_CPP11 |
105 |
|
106 |
True if the implementation claims to be compliant to ISO/IEC 14882:2011 |
107 |
(C++11) or any later version. |
108 |
|
109 |
=item ECB_GCC_VERSION (major, minor) |
110 |
|
111 |
Expands to a true value (suitable for testing in by the preprocessor) |
112 |
if the compiler used is GNU C and the version is the given version, or |
113 |
higher. |
114 |
|
115 |
This macro tries to return false on compilers that claim to be GCC |
116 |
compatible but aren't. |
117 |
|
118 |
=item ECB_EXTERN_C |
119 |
|
120 |
Expands to C<extern "C"> in C++, and a simple C<extern> in C. |
121 |
|
122 |
This can be used to declare a single external C function: |
123 |
|
124 |
ECB_EXTERN_C int printf (const char *format, ...); |
125 |
|
126 |
=item ECB_EXTERN_C_BEG / ECB_EXTERN_C_END |
127 |
|
128 |
These two macros can be used to wrap multiple C<extern "C"> definitions - |
129 |
they expand to nothing in C. |
130 |
|
131 |
They are most useful in header files: |
132 |
|
133 |
ECB_EXTERN_C_BEG |
134 |
|
135 |
int mycfun1 (int x); |
136 |
int mycfun2 (int x); |
137 |
|
138 |
ECB_EXTERN_C_END |
139 |
|
140 |
=item ECB_STDFP |
141 |
|
142 |
If this evaluates to a true value (suitable for testing in by the |
143 |
preprocessor), then C<float> and C<double> use IEEE 754 single/binary32 |
144 |
and double/binary64 representations internally I<and> the endianness of |
145 |
both types match the endianness of C<uint32_t> and C<uint64_t>. |
146 |
|
147 |
This means you can just copy the bits of a C<float> (or C<double>) to an |
148 |
C<uint32_t> (or C<uint64_t>) and get the raw IEEE 754 bit representation |
149 |
without having to think about format or endianness. |
150 |
|
151 |
This is true for basically all modern platforms, although F<ecb.h> might |
152 |
not be able to deduce this correctly everywhere and might err on the safe |
153 |
side. |
154 |
|
155 |
=item ECB_AMD64, ECB_AMD64_X32 |
156 |
|
157 |
These two macros are defined to C<1> on the x86_64/amd64 ABI and the X32 |
158 |
ABI, respectively, and undefined elsewhere. |
159 |
|
160 |
The designers of the new X32 ABI for some inexplicable reason decided to |
161 |
make it look exactly like amd64, even though it's completely incompatible |
162 |
to that ABI, breaking about every piece of software that assumed that |
163 |
C<__x86_64> stands for, well, the x86-64 ABI, making these macros |
164 |
necessary. |
165 |
|
166 |
=back |
167 |
|
168 |
=head2 ATTRIBUTES |
169 |
|
170 |
A major part of libecb deals with additional attributes that can be |
171 |
assigned to functions, variables and sometimes even types - much like |
172 |
C<const> or C<volatile> in C. They are implemented using either GCC |
173 |
attributes or other compiler/language specific features. Attributes |
174 |
declarations must be put before the whole declaration: |
175 |
|
176 |
ecb_const int mysqrt (int a); |
177 |
ecb_unused int i; |
178 |
|
179 |
=over 4 |
180 |
|
181 |
=item ecb_unused |
182 |
|
183 |
Marks a function or a variable as "unused", which simply suppresses a |
184 |
warning by GCC when it detects it as unused. This is useful when you e.g. |
185 |
declare a variable but do not always use it: |
186 |
|
187 |
{ |
188 |
ecb_unused int var; |
189 |
|
190 |
#ifdef SOMECONDITION |
191 |
var = ...; |
192 |
return var; |
193 |
#else |
194 |
return 0; |
195 |
#endif |
196 |
} |
197 |
|
198 |
=item ecb_deprecated |
199 |
|
200 |
Similar to C<ecb_unused>, but marks a function, variable or type as |
201 |
deprecated. This makes some compilers warn when the type is used. |
202 |
|
203 |
=item ecb_inline |
204 |
|
205 |
Expands either to C<static inline> or to just C<static>, if inline |
206 |
isn't supported. It should be used to declare functions that should be |
207 |
inlined, for code size or speed reasons. |
208 |
|
209 |
Example: inline this function, it surely will reduce codesize. |
210 |
|
211 |
ecb_inline int |
212 |
negmul (int a, int b) |
213 |
{ |
214 |
return - (a * b); |
215 |
} |
216 |
|
217 |
=item ecb_noinline |
218 |
|
219 |
Prevent a function from being inlined - it might be optimised away, but |
220 |
not inlined into other functions. This is useful if you know your function |
221 |
is rarely called and large enough for inlining not to be helpful. |
222 |
|
223 |
=item ecb_noreturn |
224 |
|
225 |
Marks a function as "not returning, ever". Some typical functions that |
226 |
don't return are C<exit> or C<abort> (which really works hard to not |
227 |
return), and now you can make your own: |
228 |
|
229 |
ecb_noreturn void |
230 |
my_abort (const char *errline) |
231 |
{ |
232 |
puts (errline); |
233 |
abort (); |
234 |
} |
235 |
|
236 |
In this case, the compiler would probably be smart enough to deduce it on |
237 |
its own, so this is mainly useful for declarations. |
238 |
|
239 |
=item ecb_restrict |
240 |
|
241 |
Expands to the C<restrict> keyword or equivalent on compilers that support |
242 |
them, and to nothing on others. Must be specified on a pointer type or |
243 |
an array index to indicate that the memory doesn't alias with any other |
244 |
restricted pointer in the same scope. |
245 |
|
246 |
Example: multiply a vector, and allow the compiler to parallelise the |
247 |
loop, because it knows it doesn't overwrite input values. |
248 |
|
249 |
void |
250 |
multiply (ecb_restrict float *src, |
251 |
ecb_restrict float *dst, |
252 |
int len, float factor) |
253 |
{ |
254 |
int i; |
255 |
|
256 |
for (i = 0; i < len; ++i) |
257 |
dst [i] = src [i] * factor; |
258 |
} |
259 |
|
260 |
=item ecb_const |
261 |
|
262 |
Declares that the function only depends on the values of its arguments, |
263 |
much like a mathematical function. It specifically does not read or write |
264 |
any memory any arguments might point to, global variables, or call any |
265 |
non-const functions. It also must not have any side effects. |
266 |
|
267 |
Such a function can be optimised much more aggressively by the compiler - |
268 |
for example, multiple calls with the same arguments can be optimised into |
269 |
a single call, which wouldn't be possible if the compiler would have to |
270 |
expect any side effects. |
271 |
|
272 |
It is best suited for functions in the sense of mathematical functions, |
273 |
such as a function returning the square root of its input argument. |
274 |
|
275 |
Not suited would be a function that calculates the hash of some memory |
276 |
area you pass in, prints some messages or looks at a global variable to |
277 |
decide on rounding. |
278 |
|
279 |
See C<ecb_pure> for a slightly less restrictive class of functions. |
280 |
|
281 |
=item ecb_pure |
282 |
|
283 |
Similar to C<ecb_const>, declares a function that has no side |
284 |
effects. Unlike C<ecb_const>, the function is allowed to examine global |
285 |
variables and any other memory areas (such as the ones passed to it via |
286 |
pointers). |
287 |
|
288 |
While these functions cannot be optimised as aggressively as C<ecb_const> |
289 |
functions, they can still be optimised away in many occasions, and the |
290 |
compiler has more freedom in moving calls to them around. |
291 |
|
292 |
Typical examples for such functions would be C<strlen> or C<memcmp>. A |
293 |
function that calculates the MD5 sum of some input and updates some MD5 |
294 |
state passed as argument would I<NOT> be pure, however, as it would modify |
295 |
some memory area that is not the return value. |
296 |
|
297 |
=item ecb_hot |
298 |
|
299 |
This declares a function as "hot" with regards to the cache - the function |
300 |
is used so often, that it is very beneficial to keep it in the cache if |
301 |
possible. |
302 |
|
303 |
The compiler reacts by trying to place hot functions near to each other in |
304 |
memory. |
305 |
|
306 |
Whether a function is hot or not often depends on the whole program, |
307 |
and less on the function itself. C<ecb_cold> is likely more useful in |
308 |
practise. |
309 |
|
310 |
=item ecb_cold |
311 |
|
312 |
The opposite of C<ecb_hot> - declares a function as "cold" with regards to |
313 |
the cache, or in other words, this function is not called often, or not at |
314 |
speed-critical times, and keeping it in the cache might be a waste of said |
315 |
cache. |
316 |
|
317 |
In addition to placing cold functions together (or at least away from hot |
318 |
functions), this knowledge can be used in other ways, for example, the |
319 |
function will be optimised for size, as opposed to speed, and codepaths |
320 |
leading to calls to those functions can automatically be marked as if |
321 |
C<ecb_expect_false> had been used to reach them. |
322 |
|
323 |
Good examples for such functions would be error reporting functions, or |
324 |
functions only called in exceptional or rare cases. |
325 |
|
326 |
=item ecb_artificial |
327 |
|
328 |
Declares the function as "artificial", in this case meaning that this |
329 |
function is not really meant to be a function, but more like an accessor |
330 |
- many methods in C++ classes are mere accessor functions, and having a |
331 |
crash reported in such a method, or single-stepping through them, is not |
332 |
usually so helpful, especially when it's inlined to just a few instructions. |
333 |
|
334 |
Marking them as artificial will instruct the debugger about just this, |
335 |
leading to happier debugging and thus happier lives. |
336 |
|
337 |
Example: in some kind of smart-pointer class, mark the pointer accessor as |
338 |
artificial, so that the whole class acts more like a pointer and less like |
339 |
some C++ abstraction monster. |
340 |
|
341 |
template<typename T> |
342 |
struct my_smart_ptr |
343 |
{ |
344 |
T *value; |
345 |
|
346 |
ecb_artificial |
347 |
operator T *() |
348 |
{ |
349 |
return value; |
350 |
} |
351 |
}; |
352 |
|
353 |
=back |
354 |
|
355 |
=head2 OPTIMISATION HINTS |
356 |
|
357 |
=over 4 |
358 |
|
359 |
=item bool ecb_is_constant (expr) |
360 |
|
361 |
Returns true iff the expression can be deduced to be a compile-time |
362 |
constant, and false otherwise. |
363 |
|
364 |
For example, when you have a C<rndm16> function that returns a 16 bit |
365 |
random number, and you have a function that maps this to a range from |
366 |
0..n-1, then you could use this inline function in a header file: |
367 |
|
368 |
ecb_inline uint32_t |
369 |
rndm (uint32_t n) |
370 |
{ |
371 |
return (n * (uint32_t)rndm16 ()) >> 16; |
372 |
} |
373 |
|
374 |
However, for powers of two, you could use a normal mask, but that is only |
375 |
worth it if, at compile time, you can detect this case. This is the case |
376 |
when the passed number is a constant and also a power of two (C<n & (n - |
377 |
1) == 0>): |
378 |
|
379 |
ecb_inline uint32_t |
380 |
rndm (uint32_t n) |
381 |
{ |
382 |
return is_constant (n) && !(n & (n - 1)) |
383 |
? rndm16 () & (num - 1) |
384 |
: (n * (uint32_t)rndm16 ()) >> 16; |
385 |
} |
386 |
|
387 |
=item bool ecb_expect (expr, value) |
388 |
|
389 |
Evaluates C<expr> and returns it. In addition, it tells the compiler that |
390 |
the C<expr> evaluates to C<value> a lot, which can be used for static |
391 |
branch optimisations. |
392 |
|
393 |
Usually, you want to use the more intuitive C<ecb_expect_true> and |
394 |
C<ecb_expect_false> functions instead. |
395 |
|
396 |
=item bool ecb_expect_true (cond) |
397 |
|
398 |
=item bool ecb_expect_false (cond) |
399 |
|
400 |
These two functions expect a expression that is true or false and return |
401 |
C<1> or C<0>, respectively, so when used in the condition of an C<if> or |
402 |
other conditional statement, it will not change the program: |
403 |
|
404 |
/* these two do the same thing */ |
405 |
if (some_condition) ...; |
406 |
if (ecb_expect_true (some_condition)) ...; |
407 |
|
408 |
However, by using C<ecb_expect_true>, you tell the compiler that the |
409 |
condition is likely to be true (and for C<ecb_expect_false>, that it is |
410 |
unlikely to be true). |
411 |
|
412 |
For example, when you check for a null pointer and expect this to be a |
413 |
rare, exceptional, case, then use C<ecb_expect_false>: |
414 |
|
415 |
void my_free (void *ptr) |
416 |
{ |
417 |
if (ecb_expect_false (ptr == 0)) |
418 |
return; |
419 |
} |
420 |
|
421 |
Consequent use of these functions to mark away exceptional cases or to |
422 |
tell the compiler what the hot path through a function is can increase |
423 |
performance considerably. |
424 |
|
425 |
You might know these functions under the name C<likely> and C<unlikely> |
426 |
- while these are common aliases, we find that the expect name is easier |
427 |
to understand when quickly skimming code. If you wish, you can use |
428 |
C<ecb_likely> instead of C<ecb_expect_true> and C<ecb_unlikely> instead of |
429 |
C<ecb_expect_false> - these are simply aliases. |
430 |
|
431 |
A very good example is in a function that reserves more space for some |
432 |
memory block (for example, inside an implementation of a string stream) - |
433 |
each time something is added, you have to check for a buffer overrun, but |
434 |
you expect that most checks will turn out to be false: |
435 |
|
436 |
/* make sure we have "size" extra room in our buffer */ |
437 |
ecb_inline void |
438 |
reserve (int size) |
439 |
{ |
440 |
if (ecb_expect_false (current + size > end)) |
441 |
real_reserve_method (size); /* presumably noinline */ |
442 |
} |
443 |
|
444 |
=item bool ecb_assume (cond) |
445 |
|
446 |
Try to tell the compiler that some condition is true, even if it's not |
447 |
obvious. |
448 |
|
449 |
This can be used to teach the compiler about invariants or other |
450 |
conditions that might improve code generation, but which are impossible to |
451 |
deduce form the code itself. |
452 |
|
453 |
For example, the example reservation function from the C<ecb_expect_false> |
454 |
description could be written thus (only C<ecb_assume> was added): |
455 |
|
456 |
ecb_inline void |
457 |
reserve (int size) |
458 |
{ |
459 |
if (ecb_expect_false (current + size > end)) |
460 |
real_reserve_method (size); /* presumably noinline */ |
461 |
|
462 |
ecb_assume (current + size <= end); |
463 |
} |
464 |
|
465 |
If you then call this function twice, like this: |
466 |
|
467 |
reserve (10); |
468 |
reserve (1); |
469 |
|
470 |
Then the compiler I<might> be able to optimise out the second call |
471 |
completely, as it knows that C<< current + 1 > end >> is false and the |
472 |
call will never be executed. |
473 |
|
474 |
=item bool ecb_unreachable () |
475 |
|
476 |
This function does nothing itself, except tell the compiler that it will |
477 |
never be executed. Apart from suppressing a warning in some cases, this |
478 |
function can be used to implement C<ecb_assume> or similar functions. |
479 |
|
480 |
=item bool ecb_prefetch (addr, rw, locality) |
481 |
|
482 |
Tells the compiler to try to prefetch memory at the given C<addr>ess |
483 |
for either reading (C<rw> = 0) or writing (C<rw> = 1). A C<locality> of |
484 |
C<0> means that there will only be one access later, C<3> means that |
485 |
the data will likely be accessed very often, and values in between mean |
486 |
something... in between. The memory pointed to by the address does not |
487 |
need to be accessible (it could be a null pointer for example), but C<rw> |
488 |
and C<locality> must be compile-time constants. |
489 |
|
490 |
An obvious way to use this is to prefetch some data far away, in a big |
491 |
array you loop over. This prefetches memory some 128 array elements later, |
492 |
in the hope that it will be ready when the CPU arrives at that location. |
493 |
|
494 |
int sum = 0; |
495 |
|
496 |
for (i = 0; i < N; ++i) |
497 |
{ |
498 |
sum += arr [i] |
499 |
ecb_prefetch (arr + i + 128, 0, 0); |
500 |
} |
501 |
|
502 |
It's hard to predict how far to prefetch, and most CPUs that can prefetch |
503 |
are often good enough to predict this kind of behaviour themselves. It |
504 |
gets more interesting with linked lists, especially when you do some fair |
505 |
processing on each list element: |
506 |
|
507 |
for (node *n = start; n; n = n->next) |
508 |
{ |
509 |
ecb_prefetch (n->next, 0, 0); |
510 |
... do medium amount of work with *n |
511 |
} |
512 |
|
513 |
After processing the node, (part of) the next node might already be in |
514 |
cache. |
515 |
|
516 |
=back |
517 |
|
518 |
=head2 BIT FIDDLING / BIT WIZARDRY |
519 |
|
520 |
=over 4 |
521 |
|
522 |
=item bool ecb_big_endian () |
523 |
|
524 |
=item bool ecb_little_endian () |
525 |
|
526 |
These two functions return true if the byte order is big endian |
527 |
(most-significant byte first) or little endian (least-significant byte |
528 |
first) respectively. |
529 |
|
530 |
On systems that are neither, their return values are unspecified. |
531 |
|
532 |
=item int ecb_ctz32 (uint32_t x) |
533 |
|
534 |
=item int ecb_ctz64 (uint64_t x) |
535 |
|
536 |
Returns the index of the least significant bit set in C<x> (or |
537 |
equivalently the number of bits set to 0 before the least significant bit |
538 |
set), starting from 0. If C<x> is 0 the result is undefined. |
539 |
|
540 |
For smaller types than C<uint32_t> you can safely use C<ecb_ctz32>. |
541 |
|
542 |
For example: |
543 |
|
544 |
ecb_ctz32 (3) = 0 |
545 |
ecb_ctz32 (6) = 1 |
546 |
|
547 |
=item bool ecb_is_pot32 (uint32_t x) |
548 |
|
549 |
=item bool ecb_is_pot64 (uint32_t x) |
550 |
|
551 |
Return true iff C<x> is a power of two or C<x == 0>. |
552 |
|
553 |
For smaller types then C<uint32_t> you can safely use C<ecb_is_pot32>. |
554 |
|
555 |
=item int ecb_ld32 (uint32_t x) |
556 |
|
557 |
=item int ecb_ld64 (uint64_t x) |
558 |
|
559 |
Returns the index of the most significant bit set in C<x>, or the number |
560 |
of digits the number requires in binary (so that C<< 2**ld <= x < |
561 |
2**(ld+1) >>). If C<x> is 0 the result is undefined. A common use case is |
562 |
to compute the integer binary logarithm, i.e. C<floor (log2 (n))>, for |
563 |
example to see how many bits a certain number requires to be encoded. |
564 |
|
565 |
This function is similar to the "count leading zero bits" function, except |
566 |
that that one returns how many zero bits are "in front" of the number (in |
567 |
the given data type), while C<ecb_ld> returns how many bits the number |
568 |
itself requires. |
569 |
|
570 |
For smaller types than C<uint32_t> you can safely use C<ecb_ld32>. |
571 |
|
572 |
=item int ecb_popcount32 (uint32_t x) |
573 |
|
574 |
=item int ecb_popcount64 (uint64_t x) |
575 |
|
576 |
Returns the number of bits set to 1 in C<x>. |
577 |
|
578 |
For smaller types than C<uint32_t> you can safely use C<ecb_popcount32>. |
579 |
|
580 |
For example: |
581 |
|
582 |
ecb_popcount32 (7) = 3 |
583 |
ecb_popcount32 (255) = 8 |
584 |
|
585 |
=item uint8_t ecb_bitrev8 (uint8_t x) |
586 |
|
587 |
=item uint16_t ecb_bitrev16 (uint16_t x) |
588 |
|
589 |
=item uint32_t ecb_bitrev32 (uint32_t x) |
590 |
|
591 |
Reverses the bits in x, i.e. the MSB becomes the LSB, MSB-1 becomes LSB+1 |
592 |
and so on. |
593 |
|
594 |
Example: |
595 |
|
596 |
ecb_bitrev8 (0xa7) = 0xea |
597 |
ecb_bitrev32 (0xffcc4411) = 0x882233ff |
598 |
|
599 |
=item uint32_t ecb_bswap16 (uint32_t x) |
600 |
|
601 |
=item uint32_t ecb_bswap32 (uint32_t x) |
602 |
|
603 |
=item uint64_t ecb_bswap64 (uint64_t x) |
604 |
|
605 |
These functions return the value of the 16-bit (32-bit, 64-bit) value |
606 |
C<x> after reversing the order of bytes (0x11223344 becomes 0x44332211 in |
607 |
C<ecb_bswap32>). |
608 |
|
609 |
=item uint8_t ecb_rotl8 (uint8_t x, unsigned int count) |
610 |
|
611 |
=item uint16_t ecb_rotl16 (uint16_t x, unsigned int count) |
612 |
|
613 |
=item uint32_t ecb_rotl32 (uint32_t x, unsigned int count) |
614 |
|
615 |
=item uint64_t ecb_rotl64 (uint64_t x, unsigned int count) |
616 |
|
617 |
=item uint8_t ecb_rotr8 (uint8_t x, unsigned int count) |
618 |
|
619 |
=item uint16_t ecb_rotr16 (uint16_t x, unsigned int count) |
620 |
|
621 |
=item uint32_t ecb_rotr32 (uint32_t x, unsigned int count) |
622 |
|
623 |
=item uint64_t ecb_rotr64 (uint64_t x, unsigned int count) |
624 |
|
625 |
These two families of functions return the value of C<x> after rotating |
626 |
all the bits by C<count> positions to the right (C<ecb_rotr>) or left |
627 |
(C<ecb_rotl>). |
628 |
|
629 |
Current GCC versions understand these functions and usually compile them |
630 |
to "optimal" code (e.g. a single C<rol> or a combination of C<shld> on |
631 |
x86). |
632 |
|
633 |
=back |
634 |
|
635 |
=head2 FLOATING POINT FIDDLING |
636 |
|
637 |
=over 4 |
638 |
|
639 |
=item uint32_t ecb_float_to_binary32 (float x) [-UECB_NO_LIBM] |
640 |
|
641 |
=item uint64_t ecb_double_to_binary64 (double x) [-UECB_NO_LIBM] |
642 |
|
643 |
These functions each take an argument in the native C<float> or C<double> |
644 |
type and return the IEEE 754 bit representation of it. |
645 |
|
646 |
The bit representation is just as IEEE 754 defines it, i.e. the sign bit |
647 |
will be the most significant bit, followed by exponent and mantissa. |
648 |
|
649 |
This function should work even when the native floating point format isn't |
650 |
IEEE compliant, of course at a speed and code size penalty, and of course |
651 |
also within reasonable limits (it tries to convert NaNs, infinities and |
652 |
denormals, but will likely convert negative zero to positive zero). |
653 |
|
654 |
On all modern platforms (where C<ECB_STDFP> is true), the compiler should |
655 |
be able to optimise away this function completely. |
656 |
|
657 |
These functions can be helpful when serialising floats to the network - you |
658 |
can serialise the return value like a normal uint32_t/uint64_t. |
659 |
|
660 |
Another use for these functions is to manipulate floating point values |
661 |
directly. |
662 |
|
663 |
Silly example: toggle the sign bit of a float. |
664 |
|
665 |
/* On gcc-4.7 on amd64, */ |
666 |
/* this results in a single add instruction to toggle the bit, and 4 extra */ |
667 |
/* instructions to move the float value to an integer register and back. */ |
668 |
|
669 |
x = ecb_binary32_to_float (ecb_float_to_binary32 (x) ^ 0x80000000U) |
670 |
|
671 |
=item float ecb_binary16_to_float (uint16_t x) [-UECB_NO_LIBM] |
672 |
|
673 |
=item float ecb_binary32_to_float (uint32_t x) [-UECB_NO_LIBM] |
674 |
|
675 |
=item double ecb_binary32_to_double (uint64_t x) [-UECB_NO_LIBM] |
676 |
|
677 |
The reverse operation of the previous function - takes the bit |
678 |
representation of an IEEE binary16, binary32 or binary64 number and |
679 |
converts it to the native C<float> or C<double> format. |
680 |
|
681 |
This function should work even when the native floating point format isn't |
682 |
IEEE compliant, of course at a speed and code size penalty, and of course |
683 |
also within reasonable limits (it tries to convert normals and denormals, |
684 |
and might be lucky for infinities, and with extraordinary luck, also for |
685 |
negative zero). |
686 |
|
687 |
On all modern platforms (where C<ECB_STDFP> is true), the compiler should |
688 |
be able to optimise away this function completely. |
689 |
|
690 |
=back |
691 |
|
692 |
=head2 ARITHMETIC |
693 |
|
694 |
=over 4 |
695 |
|
696 |
=item x = ecb_mod (m, n) |
697 |
|
698 |
Returns C<m> modulo C<n>, which is the same as the positive remainder |
699 |
of the division operation between C<m> and C<n>, using floored |
700 |
division. Unlike the C remainder operator C<%>, this function ensures that |
701 |
the return value is always positive and that the two numbers I<m> and |
702 |
I<m' = m + i * n> result in the same value modulo I<n> - in other words, |
703 |
C<ecb_mod> implements the mathematical modulo operation, which is missing |
704 |
in the language. |
705 |
|
706 |
C<n> must be strictly positive (i.e. C<< >= 1 >>), while C<m> must be |
707 |
negatable, that is, both C<m> and C<-m> must be representable in its |
708 |
type (this typically excludes the minimum signed integer value, the same |
709 |
limitation as for C</> and C<%> in C). |
710 |
|
711 |
Current GCC versions compile this into an efficient branchless sequence on |
712 |
almost all CPUs. |
713 |
|
714 |
For example, when you want to rotate forward through the members of an |
715 |
array for increasing C<m> (which might be negative), then you should use |
716 |
C<ecb_mod>, as the C<%> operator might give either negative results, or |
717 |
change direction for negative values: |
718 |
|
719 |
for (m = -100; m <= 100; ++m) |
720 |
int elem = myarray [ecb_mod (m, ecb_array_length (myarray))]; |
721 |
|
722 |
=item x = ecb_div_rd (val, div) |
723 |
|
724 |
=item x = ecb_div_ru (val, div) |
725 |
|
726 |
Returns C<val> divided by C<div> rounded down or up, respectively. |
727 |
C<val> and C<div> must have integer types and C<div> must be strictly |
728 |
positive. Note that these functions are implemented with macros in C |
729 |
and with function templates in C++. |
730 |
|
731 |
=back |
732 |
|
733 |
=head2 UTILITY |
734 |
|
735 |
=over 4 |
736 |
|
737 |
=item element_count = ecb_array_length (name) |
738 |
|
739 |
Returns the number of elements in the array C<name>. For example: |
740 |
|
741 |
int primes[] = { 2, 3, 5, 7, 11 }; |
742 |
int sum = 0; |
743 |
|
744 |
for (i = 0; i < ecb_array_length (primes); i++) |
745 |
sum += primes [i]; |
746 |
|
747 |
=back |
748 |
|
749 |
=head2 SYMBOLS GOVERNING COMPILATION OF ECB.H ITSELF |
750 |
|
751 |
These symbols need to be defined before including F<ecb.h> the first time. |
752 |
|
753 |
=over 4 |
754 |
|
755 |
=item ECB_NO_THREADS |
756 |
|
757 |
If F<ecb.h> is never used from multiple threads, then this symbol can |
758 |
be defined, in which case memory fences (and similar constructs) are |
759 |
completely removed, leading to more efficient code and fewer dependencies. |
760 |
|
761 |
Setting this symbol to a true value implies C<ECB_NO_SMP>. |
762 |
|
763 |
=item ECB_NO_SMP |
764 |
|
765 |
The weaker version of C<ECB_NO_THREADS> - if F<ecb.h> is used from |
766 |
multiple threads, but never concurrently (e.g. if the system the program |
767 |
runs on has only a single CPU with a single core, no hyperthreading and so |
768 |
on), then this symbol can be defined, leading to more efficient code and |
769 |
fewer dependencies. |
770 |
|
771 |
=item ECB_NO_LIBM |
772 |
|
773 |
When defined to C<1>, do not export any functions that might introduce |
774 |
dependencies on the math library (usually called F<-lm>) - these are |
775 |
marked with [-UECB_NO_LIBM]. |
776 |
|
777 |
=back |
778 |
|
779 |
|