ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/libecb/ecb.pod
Revision: 1.33
Committed: Fri Jun 17 18:46:19 2011 UTC (13 years ago) by root
Branch: MAIN
Changes since 1.32: +4 -0 lines
Log Message:
*** empty log message ***

File Contents

# User Rev Content
1 root 1.14 =head1 LIBECB - e-C-Builtins
2 root 1.3
3 root 1.14 =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 sf-exg 1.16 It's part of the e-suite of libraries, other members of which include
9 root 1.14 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 sf-exg 1.16 it provides a number of other lowlevel C utilities, such as endianness
18 root 1.14 detection, byte swapping or bit rotations.
19    
20 root 1.24 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 root 1.17
24 root 1.14 More might come.
25 root 1.3
26     =head2 ABOUT THE HEADER
27    
28 root 1.14 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 sf-exg 1.16 There are currently no object files to link to - future versions might
37 root 1.14 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 root 1.1
59     =head2 GCC ATTRIBUTES
60    
61 root 1.20 A major part of libecb deals with GCC attributes. These are additional
62 sf-exg 1.26 attributes that you can assign to functions, variables and sometimes even
63 root 1.20 types - much like C<const> or C<volatile> in C.
64    
65     While GCC allows declarations to show up in many surprising places,
66 sf-exg 1.26 but not in many expected places, the safest way is to put attribute
67 root 1.20 declarations before the whole declaration:
68    
69     ecb_const int mysqrt (int a);
70     ecb_unused int i;
71    
72     For variables, it is often nicer to put the attribute after the name, and
73     avoid multiple declarations using commas:
74    
75     int i ecb_unused;
76 root 1.3
77 root 1.1 =over 4
78    
79 root 1.2 =item ecb_attribute ((attrs...))
80 root 1.1
81 root 1.15 A simple wrapper that expands to C<__attribute__((attrs))> on GCC, and to
82     nothing on other compilers, so the effect is that only GCC sees these.
83    
84     Example: use the C<deprecated> attribute on a function.
85    
86     ecb_attribute((__deprecated__)) void
87     do_not_use_me_anymore (void);
88 root 1.2
89 root 1.3 =item ecb_unused
90    
91     Marks a function or a variable as "unused", which simply suppresses a
92     warning by GCC when it detects it as unused. This is useful when you e.g.
93     declare a variable but do not always use it:
94    
95 root 1.15 {
96     int var ecb_unused;
97 root 1.3
98 root 1.15 #ifdef SOMECONDITION
99     var = ...;
100     return var;
101     #else
102     return 0;
103     #endif
104     }
105 root 1.3
106 root 1.31 =item ecb_inline
107 root 1.29
108     This is not actually an attribute, but you use it like one. It expands
109     either to C<static inline> or to just C<static>, if inline isn't
110     supported. It should be used to declare functions that should be inlined,
111     for code size or speed reasons.
112    
113     Example: inline this function, it surely will reduce codesize.
114    
115 root 1.31 ecb_inline int
116 root 1.29 negmul (int a, int b)
117     {
118     return - (a * b);
119     }
120    
121 root 1.2 =item ecb_noinline
122    
123 root 1.9 Prevent a function from being inlined - it might be optimised away, but
124 root 1.3 not inlined into other functions. This is useful if you know your function
125     is rarely called and large enough for inlining not to be helpful.
126    
127 root 1.2 =item ecb_noreturn
128    
129 root 1.17 Marks a function as "not returning, ever". Some typical functions that
130     don't return are C<exit> or C<abort> (which really works hard to not
131     return), and now you can make your own:
132    
133     ecb_noreturn void
134     my_abort (const char *errline)
135     {
136     puts (errline);
137     abort ();
138     }
139    
140 sf-exg 1.19 In this case, the compiler would probably be smart enough to deduce it on
141     its own, so this is mainly useful for declarations.
142 root 1.17
143 root 1.2 =item ecb_const
144    
145 sf-exg 1.19 Declares that the function only depends on the values of its arguments,
146 root 1.17 much like a mathematical function. It specifically does not read or write
147     any memory any arguments might point to, global variables, or call any
148     non-const functions. It also must not have any side effects.
149    
150     Such a function can be optimised much more aggressively by the compiler -
151     for example, multiple calls with the same arguments can be optimised into
152     a single call, which wouldn't be possible if the compiler would have to
153     expect any side effects.
154    
155     It is best suited for functions in the sense of mathematical functions,
156 sf-exg 1.19 such as a function returning the square root of its input argument.
157 root 1.17
158     Not suited would be a function that calculates the hash of some memory
159     area you pass in, prints some messages or looks at a global variable to
160     decide on rounding.
161    
162     See C<ecb_pure> for a slightly less restrictive class of functions.
163    
164 root 1.2 =item ecb_pure
165    
166 root 1.17 Similar to C<ecb_const>, declares a function that has no side
167     effects. Unlike C<ecb_const>, the function is allowed to examine global
168     variables and any other memory areas (such as the ones passed to it via
169     pointers).
170    
171     While these functions cannot be optimised as aggressively as C<ecb_const>
172     functions, they can still be optimised away in many occasions, and the
173     compiler has more freedom in moving calls to them around.
174    
175     Typical examples for such functions would be C<strlen> or C<memcmp>. A
176     function that calculates the MD5 sum of some input and updates some MD5
177     state passed as argument would I<NOT> be pure, however, as it would modify
178     some memory area that is not the return value.
179    
180 root 1.2 =item ecb_hot
181    
182 root 1.17 This declares a function as "hot" with regards to the cache - the function
183     is used so often, that it is very beneficial to keep it in the cache if
184     possible.
185    
186     The compiler reacts by trying to place hot functions near to each other in
187     memory.
188    
189 sf-exg 1.19 Whether a function is hot or not often depends on the whole program,
190 root 1.17 and less on the function itself. C<ecb_cold> is likely more useful in
191     practise.
192    
193 root 1.2 =item ecb_cold
194    
195 root 1.17 The opposite of C<ecb_hot> - declares a function as "cold" with regards to
196     the cache, or in other words, this function is not called often, or not at
197     speed-critical times, and keeping it in the cache might be a waste of said
198     cache.
199    
200     In addition to placing cold functions together (or at least away from hot
201     functions), this knowledge can be used in other ways, for example, the
202     function will be optimised for size, as opposed to speed, and codepaths
203     leading to calls to those functions can automatically be marked as if
204 root 1.27 C<ecb_expect_false> had been used to reach them.
205 root 1.17
206     Good examples for such functions would be error reporting functions, or
207     functions only called in exceptional or rare cases.
208    
209 root 1.2 =item ecb_artificial
210    
211 root 1.17 Declares the function as "artificial", in this case meaning that this
212     function is not really mean to be a function, but more like an accessor
213     - many methods in C++ classes are mere accessor functions, and having a
214     crash reported in such a method, or single-stepping through them, is not
215     usually so helpful, especially when it's inlined to just a few instructions.
216    
217     Marking them as artificial will instruct the debugger about just this,
218     leading to happier debugging and thus happier lives.
219    
220     Example: in some kind of smart-pointer class, mark the pointer accessor as
221     artificial, so that the whole class acts more like a pointer and less like
222     some C++ abstraction monster.
223    
224     template<typename T>
225     struct my_smart_ptr
226     {
227     T *value;
228    
229     ecb_artificial
230     operator T *()
231     {
232     return value;
233     }
234     };
235    
236 root 1.2 =back
237 root 1.1
238     =head2 OPTIMISATION HINTS
239    
240     =over 4
241    
242 root 1.14 =item bool ecb_is_constant(expr)
243 root 1.1
244 root 1.3 Returns true iff the expression can be deduced to be a compile-time
245     constant, and false otherwise.
246    
247     For example, when you have a C<rndm16> function that returns a 16 bit
248     random number, and you have a function that maps this to a range from
249 root 1.5 0..n-1, then you could use this inline function in a header file:
250 root 1.3
251     ecb_inline uint32_t
252     rndm (uint32_t n)
253     {
254 root 1.6 return (n * (uint32_t)rndm16 ()) >> 16;
255 root 1.3 }
256    
257     However, for powers of two, you could use a normal mask, but that is only
258     worth it if, at compile time, you can detect this case. This is the case
259     when the passed number is a constant and also a power of two (C<n & (n -
260     1) == 0>):
261    
262     ecb_inline uint32_t
263     rndm (uint32_t n)
264     {
265     return is_constant (n) && !(n & (n - 1))
266     ? rndm16 () & (num - 1)
267 root 1.6 : (n * (uint32_t)rndm16 ()) >> 16;
268 root 1.3 }
269    
270 root 1.14 =item bool ecb_expect (expr, value)
271 root 1.1
272 root 1.7 Evaluates C<expr> and returns it. In addition, it tells the compiler that
273     the C<expr> evaluates to C<value> a lot, which can be used for static
274     branch optimisations.
275 root 1.1
276 root 1.27 Usually, you want to use the more intuitive C<ecb_expect_true> and
277     C<ecb_expect_false> functions instead.
278 root 1.1
279 root 1.27 =item bool ecb_expect_true (cond)
280 root 1.1
281 root 1.27 =item bool ecb_expect_false (cond)
282 root 1.1
283 root 1.7 These two functions expect a expression that is true or false and return
284     C<1> or C<0>, respectively, so when used in the condition of an C<if> or
285     other conditional statement, it will not change the program:
286    
287     /* these two do the same thing */
288     if (some_condition) ...;
289 root 1.27 if (ecb_expect_true (some_condition)) ...;
290 root 1.7
291 root 1.27 However, by using C<ecb_expect_true>, you tell the compiler that the
292     condition is likely to be true (and for C<ecb_expect_false>, that it is
293     unlikely to be true).
294 root 1.7
295 root 1.9 For example, when you check for a null pointer and expect this to be a
296 root 1.27 rare, exceptional, case, then use C<ecb_expect_false>:
297 root 1.7
298     void my_free (void *ptr)
299     {
300 root 1.27 if (ecb_expect_false (ptr == 0))
301 root 1.7 return;
302     }
303    
304     Consequent use of these functions to mark away exceptional cases or to
305     tell the compiler what the hot path through a function is can increase
306     performance considerably.
307    
308 root 1.27 You might know these functions under the name C<likely> and C<unlikely>
309     - while these are common aliases, we find that the expect name is easier
310     to understand when quickly skimming code. If you wish, you can use
311     C<ecb_likely> instead of C<ecb_expect_true> and C<ecb_unlikely> instead of
312     C<ecb_expect_false> - these are simply aliases.
313    
314 root 1.7 A very good example is in a function that reserves more space for some
315     memory block (for example, inside an implementation of a string stream) -
316 root 1.9 each time something is added, you have to check for a buffer overrun, but
317 root 1.7 you expect that most checks will turn out to be false:
318    
319     /* make sure we have "size" extra room in our buffer */
320     ecb_inline void
321     reserve (int size)
322     {
323 root 1.27 if (ecb_expect_false (current + size > end))
324 root 1.7 real_reserve_method (size); /* presumably noinline */
325     }
326    
327 root 1.14 =item bool ecb_assume (cond)
328 root 1.7
329     Try to tell the compiler that some condition is true, even if it's not
330     obvious.
331    
332     This can be used to teach the compiler about invariants or other
333     conditions that might improve code generation, but which are impossible to
334     deduce form the code itself.
335    
336 root 1.27 For example, the example reservation function from the C<ecb_expect_false>
337 root 1.7 description could be written thus (only C<ecb_assume> was added):
338    
339     ecb_inline void
340     reserve (int size)
341     {
342 root 1.27 if (ecb_expect_false (current + size > end))
343 root 1.7 real_reserve_method (size); /* presumably noinline */
344    
345     ecb_assume (current + size <= end);
346     }
347    
348     If you then call this function twice, like this:
349    
350     reserve (10);
351     reserve (1);
352    
353     Then the compiler I<might> be able to optimise out the second call
354     completely, as it knows that C<< current + 1 > end >> is false and the
355     call will never be executed.
356    
357     =item bool ecb_unreachable ()
358    
359     This function does nothing itself, except tell the compiler that it will
360 root 1.9 never be executed. Apart from suppressing a warning in some cases, this
361 root 1.7 function can be used to implement C<ecb_assume> or similar functions.
362    
363 root 1.14 =item bool ecb_prefetch (addr, rw, locality)
364 root 1.7
365     Tells the compiler to try to prefetch memory at the given C<addr>ess
366 root 1.10 for either reading (C<rw> = 0) or writing (C<rw> = 1). A C<locality> of
367 root 1.7 C<0> means that there will only be one access later, C<3> means that
368     the data will likely be accessed very often, and values in between mean
369     something... in between. The memory pointed to by the address does not
370     need to be accessible (it could be a null pointer for example), but C<rw>
371     and C<locality> must be compile-time constants.
372    
373     An obvious way to use this is to prefetch some data far away, in a big
374 root 1.9 array you loop over. This prefetches memory some 128 array elements later,
375 root 1.7 in the hope that it will be ready when the CPU arrives at that location.
376    
377     int sum = 0;
378    
379     for (i = 0; i < N; ++i)
380     {
381     sum += arr [i]
382     ecb_prefetch (arr + i + 128, 0, 0);
383     }
384    
385     It's hard to predict how far to prefetch, and most CPUs that can prefetch
386     are often good enough to predict this kind of behaviour themselves. It
387     gets more interesting with linked lists, especially when you do some fair
388     processing on each list element:
389    
390     for (node *n = start; n; n = n->next)
391     {
392     ecb_prefetch (n->next, 0, 0);
393     ... do medium amount of work with *n
394     }
395    
396     After processing the node, (part of) the next node might already be in
397     cache.
398 root 1.1
399 root 1.2 =back
400 root 1.1
401     =head2 BIT FIDDLING / BITSTUFFS
402    
403 root 1.4 =over 4
404    
405 root 1.3 =item bool ecb_big_endian ()
406    
407     =item bool ecb_little_endian ()
408    
409 sf-exg 1.11 These two functions return true if the byte order is big endian
410     (most-significant byte first) or little endian (least-significant byte
411     first) respectively.
412    
413 root 1.24 On systems that are neither, their return values are unspecified.
414    
415 root 1.3 =item int ecb_ctz32 (uint32_t x)
416    
417 sf-exg 1.11 Returns the index of the least significant bit set in C<x> (or
418 root 1.24 equivalently the number of bits set to 0 before the least significant bit
419 sf-exg 1.32 set), starting from 0. If C<x> is 0 the result is undefined. For example:
420 sf-exg 1.11
421 root 1.15 ecb_ctz32 (3) = 0
422     ecb_ctz32 (6) = 1
423 sf-exg 1.11
424 root 1.3 =item int ecb_popcount32 (uint32_t x)
425    
426 sf-exg 1.11 Returns the number of bits set to 1 in C<x>. For example:
427    
428 root 1.15 ecb_popcount32 (7) = 3
429     ecb_popcount32 (255) = 8
430 sf-exg 1.11
431 root 1.8 =item uint32_t ecb_bswap16 (uint32_t x)
432    
433 root 1.3 =item uint32_t ecb_bswap32 (uint32_t x)
434    
435 root 1.21 These two functions return the value of the 16-bit (32-bit) value C<x>
436     after reversing the order of bytes (0x11223344 becomes 0x44332211).
437 sf-exg 1.13
438 root 1.3 =item uint32_t ecb_rotr32 (uint32_t x, unsigned int count)
439    
440     =item uint32_t ecb_rotl32 (uint32_t x, unsigned int count)
441    
442 root 1.33 =item uint64_t ecb_rotr64 (uint64_t x, unsigned int count)
443    
444     =item uint64_t ecb_rotl64 (uint64_t x, unsigned int count)
445    
446 root 1.22 These two functions return the value of C<x> after rotating all the bits
447 sf-exg 1.11 by C<count> positions to the right or left respectively.
448    
449 root 1.20 Current GCC versions understand these functions and usually compile them
450     to "optimal" code (e.g. a single C<roll> on x86).
451    
452 root 1.3 =back
453 root 1.1
454     =head2 ARITHMETIC
455    
456 root 1.3 =over 4
457    
458 root 1.14 =item x = ecb_mod (m, n)
459 root 1.3
460 root 1.25 Returns C<m> modulo C<n>, which is the same as the positive remainder
461     of the division operation between C<m> and C<n>, using floored
462     division. Unlike the C remainder operator C<%>, this function ensures that
463     the return value is always positive and that the two numbers I<m> and
464     I<m' = m + i * n> result in the same value modulo I<n> - in other words,
465     C<ecb_mod> implements the mathematical modulo operation, which is missing
466     in the language.
467 root 1.14
468 sf-exg 1.23 C<n> must be strictly positive (i.e. C<< >= 1 >>), while C<m> must be
469 root 1.14 negatable, that is, both C<m> and C<-m> must be representable in its
470 root 1.30 type (this typically excludes the minimum signed integer value, the same
471 root 1.25 limitation as for C</> and C<%> in C).
472 sf-exg 1.11
473 root 1.24 Current GCC versions compile this into an efficient branchless sequence on
474 root 1.28 almost all CPUs.
475 root 1.24
476     For example, when you want to rotate forward through the members of an
477     array for increasing C<m> (which might be negative), then you should use
478     C<ecb_mod>, as the C<%> operator might give either negative results, or
479     change direction for negative values:
480    
481     for (m = -100; m <= 100; ++m)
482     int elem = myarray [ecb_mod (m, ecb_array_length (myarray))];
483    
484 root 1.3 =back
485 root 1.1
486     =head2 UTILITY
487    
488 root 1.3 =over 4
489    
490 sf-exg 1.23 =item element_count = ecb_array_length (name)
491 root 1.3
492 sf-exg 1.13 Returns the number of elements in the array C<name>. For example:
493    
494     int primes[] = { 2, 3, 5, 7, 11 };
495     int sum = 0;
496    
497     for (i = 0; i < ecb_array_length (primes); i++)
498     sum += primes [i];
499    
500 root 1.3 =back
501 root 1.1
502