… | |
… | |
68 | platform (currently C<4> or C<8>) and can be used in preprocessor |
68 | platform (currently C<4> or C<8>) and can be used in preprocessor |
69 | expressions. |
69 | expressions. |
70 | |
70 | |
71 | For C<ptrdiff_t> and C<size_t> use C<stddef.h>. |
71 | For C<ptrdiff_t> and C<size_t> use C<stddef.h>. |
72 | |
72 | |
73 | =head2 LANGUAGE/COMPILER VERSIONS |
73 | =head2 LANGUAGE/ENVIRONMENT/COMPILER VERSIONS |
74 | |
74 | |
75 | All the following symbols expand to an expression that can be tested in |
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 |
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). |
77 | ensure it's either C<0> or C<1> if you need that). |
78 | |
78 | |
… | |
… | |
104 | =item ECB_CPP11 |
104 | =item ECB_CPP11 |
105 | |
105 | |
106 | True if the implementation claims to be compliant to ISO/IEC 14882:2011 |
106 | True if the implementation claims to be compliant to ISO/IEC 14882:2011 |
107 | (C++11) or any later version. |
107 | (C++11) or any later version. |
108 | |
108 | |
109 | =item ECB_GCC_VERSION(major,minor) |
109 | =item ECB_GCC_VERSION (major, minor) |
110 | |
110 | |
111 | Expands to a true value (suitable for testing in by the preprocessor) |
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 |
112 | if the compiler used is GNU C and the version is the given version, or |
113 | higher. |
113 | higher. |
114 | |
114 | |
… | |
… | |
163 | C<__x86_64> stands for, well, the x86-64 ABI, making these macros |
163 | C<__x86_64> stands for, well, the x86-64 ABI, making these macros |
164 | necessary. |
164 | necessary. |
165 | |
165 | |
166 | =back |
166 | =back |
167 | |
167 | |
|
|
168 | =head2 MACRO TRICKERY |
|
|
169 | |
|
|
170 | =over 4 |
|
|
171 | |
|
|
172 | =item ECB_CONCAT (a, b) |
|
|
173 | |
|
|
174 | Expands any macros in C<a> and C<b>, then concatenates the result to form |
|
|
175 | a single token. This is mainly useful to form identifiers from components, |
|
|
176 | e.g.: |
|
|
177 | |
|
|
178 | #define S1 str |
|
|
179 | #define S2 cpy |
|
|
180 | |
|
|
181 | ECB_CONCAT (S1, S2)(dst, src); // == strcpy (dst, src); |
|
|
182 | |
|
|
183 | =item ECB_STRINGIFY (arg) |
|
|
184 | |
|
|
185 | Expands any macros in C<arg> and returns the stringified version of |
|
|
186 | it. This is mainly useful to get the contents of a macro in string form, |
|
|
187 | e.g.: |
|
|
188 | |
|
|
189 | #define SQL_LIMIT 100 |
|
|
190 | sql_exec ("select * from table limit " ECB_STRINGIFY (SQL_LIMIT)); |
|
|
191 | |
|
|
192 | =item ECB_STRINGIFY_EXPR (expr) |
|
|
193 | |
|
|
194 | Like C<ECB_STRINGIFY>, but additionally evaluates C<expr> to make sure it |
|
|
195 | is a valid expression. This is useful to catch typos or cases where the |
|
|
196 | macro isn't available: |
|
|
197 | |
|
|
198 | #include <errno.h> |
|
|
199 | |
|
|
200 | ECB_STRINGIFY (EDOM); // "33" (on my system at least) |
|
|
201 | ECB_STRINGIFY_EXPR (EDOM); // "33" |
|
|
202 | |
|
|
203 | // now imagine we had a typo: |
|
|
204 | |
|
|
205 | ECB_STRINGIFY (EDAM); // "EDAM" |
|
|
206 | ECB_STRINGIFY_EXPR (EDAM); // error: EDAM undefined |
|
|
207 | |
|
|
208 | =back |
|
|
209 | |
168 | =head2 GCC ATTRIBUTES |
210 | =head2 ATTRIBUTES |
169 | |
211 | |
170 | A major part of libecb deals with GCC attributes. These are additional |
212 | A major part of libecb deals with additional attributes that can be |
171 | attributes that you can assign to functions, variables and sometimes even |
213 | assigned to functions, variables and sometimes even types - much like |
172 | types - much like C<const> or C<volatile> in C. |
214 | C<const> or C<volatile> in C. They are implemented using either GCC |
173 | |
215 | attributes or other compiler/language specific features. Attributes |
174 | While GCC allows declarations to show up in many surprising places, |
|
|
175 | but not in many expected places, the safest way is to put attribute |
|
|
176 | declarations before the whole declaration: |
216 | declarations must be put before the whole declaration: |
177 | |
217 | |
178 | ecb_const int mysqrt (int a); |
218 | ecb_const int mysqrt (int a); |
179 | ecb_unused int i; |
219 | ecb_unused int i; |
180 | |
220 | |
181 | For variables, it is often nicer to put the attribute after the name, and |
|
|
182 | avoid multiple declarations using commas: |
|
|
183 | |
|
|
184 | int i ecb_unused; |
|
|
185 | |
|
|
186 | =over 4 |
221 | =over 4 |
187 | |
|
|
188 | =item ecb_attribute ((attrs...)) |
|
|
189 | |
|
|
190 | A simple wrapper that expands to C<__attribute__((attrs))> on GCC, and to |
|
|
191 | nothing on other compilers, so the effect is that only GCC sees these. |
|
|
192 | |
|
|
193 | Example: use the C<deprecated> attribute on a function. |
|
|
194 | |
|
|
195 | ecb_attribute((__deprecated__)) void |
|
|
196 | do_not_use_me_anymore (void); |
|
|
197 | |
222 | |
198 | =item ecb_unused |
223 | =item ecb_unused |
199 | |
224 | |
200 | Marks a function or a variable as "unused", which simply suppresses a |
225 | Marks a function or a variable as "unused", which simply suppresses a |
201 | warning by GCC when it detects it as unused. This is useful when you e.g. |
226 | warning by GCC when it detects it as unused. This is useful when you e.g. |
202 | declare a variable but do not always use it: |
227 | declare a variable but do not always use it: |
203 | |
228 | |
204 | { |
229 | { |
205 | int var ecb_unused; |
230 | ecb_unused int var; |
206 | |
231 | |
207 | #ifdef SOMECONDITION |
232 | #ifdef SOMECONDITION |
208 | var = ...; |
233 | var = ...; |
209 | return var; |
234 | return var; |
210 | #else |
235 | #else |
211 | return 0; |
236 | return 0; |
212 | #endif |
237 | #endif |
213 | } |
238 | } |
214 | |
239 | |
|
|
240 | =item ecb_deprecated |
|
|
241 | |
|
|
242 | Similar to C<ecb_unused>, but marks a function, variable or type as |
|
|
243 | deprecated. This makes some compilers warn when the type is used. |
|
|
244 | |
|
|
245 | =item ecb_deprecated_message (message) |
|
|
246 | |
|
|
247 | Same as C<ecb_deprecated>, but if possible, the specified diagnostic is |
|
|
248 | used instead of a generic depreciation message when the object is being |
|
|
249 | used. |
|
|
250 | |
215 | =item ecb_inline |
251 | =item ecb_inline |
216 | |
252 | |
217 | This is not actually an attribute, but you use it like one. It expands |
|
|
218 | either to C<static inline> or to just C<static>, if inline isn't |
253 | Expands either to C<static inline> or to just C<static>, if inline |
219 | supported. It should be used to declare functions that should be inlined, |
254 | isn't supported. It should be used to declare functions that should be |
220 | for code size or speed reasons. |
255 | inlined, for code size or speed reasons. |
221 | |
256 | |
222 | Example: inline this function, it surely will reduce codesize. |
257 | Example: inline this function, it surely will reduce codesize. |
223 | |
258 | |
224 | ecb_inline int |
259 | ecb_inline int |
225 | negmul (int a, int b) |
260 | negmul (int a, int b) |
… | |
… | |
227 | return - (a * b); |
262 | return - (a * b); |
228 | } |
263 | } |
229 | |
264 | |
230 | =item ecb_noinline |
265 | =item ecb_noinline |
231 | |
266 | |
232 | Prevent a function from being inlined - it might be optimised away, but |
267 | Prevents a function from being inlined - it might be optimised away, but |
233 | not inlined into other functions. This is useful if you know your function |
268 | not inlined into other functions. This is useful if you know your function |
234 | is rarely called and large enough for inlining not to be helpful. |
269 | is rarely called and large enough for inlining not to be helpful. |
235 | |
270 | |
236 | =item ecb_noreturn |
271 | =item ecb_noreturn |
237 | |
272 | |
… | |
… | |
258 | |
293 | |
259 | Example: multiply a vector, and allow the compiler to parallelise the |
294 | Example: multiply a vector, and allow the compiler to parallelise the |
260 | loop, because it knows it doesn't overwrite input values. |
295 | loop, because it knows it doesn't overwrite input values. |
261 | |
296 | |
262 | void |
297 | void |
263 | multiply (float *ecb_restrict src, |
298 | multiply (ecb_restrict float *src, |
264 | float *ecb_restrict dst, |
299 | ecb_restrict float *dst, |
265 | int len, float factor) |
300 | int len, float factor) |
266 | { |
301 | { |
267 | int i; |
302 | int i; |
268 | |
303 | |
269 | for (i = 0; i < len; ++i) |
304 | for (i = 0; i < len; ++i) |
… | |
… | |
367 | |
402 | |
368 | =head2 OPTIMISATION HINTS |
403 | =head2 OPTIMISATION HINTS |
369 | |
404 | |
370 | =over 4 |
405 | =over 4 |
371 | |
406 | |
372 | =item bool ecb_is_constant(expr) |
407 | =item bool ecb_is_constant (expr) |
373 | |
408 | |
374 | Returns true iff the expression can be deduced to be a compile-time |
409 | Returns true iff the expression can be deduced to be a compile-time |
375 | constant, and false otherwise. |
410 | constant, and false otherwise. |
376 | |
411 | |
377 | For example, when you have a C<rndm16> function that returns a 16 bit |
412 | For example, when you have a C<rndm16> function that returns a 16 bit |
… | |
… | |
395 | return is_constant (n) && !(n & (n - 1)) |
430 | return is_constant (n) && !(n & (n - 1)) |
396 | ? rndm16 () & (num - 1) |
431 | ? rndm16 () & (num - 1) |
397 | : (n * (uint32_t)rndm16 ()) >> 16; |
432 | : (n * (uint32_t)rndm16 ()) >> 16; |
398 | } |
433 | } |
399 | |
434 | |
400 | =item bool ecb_expect (expr, value) |
435 | =item ecb_expect (expr, value) |
401 | |
436 | |
402 | Evaluates C<expr> and returns it. In addition, it tells the compiler that |
437 | Evaluates C<expr> and returns it. In addition, it tells the compiler that |
403 | the C<expr> evaluates to C<value> a lot, which can be used for static |
438 | the C<expr> evaluates to C<value> a lot, which can be used for static |
404 | branch optimisations. |
439 | branch optimisations. |
405 | |
440 | |
… | |
… | |
452 | { |
487 | { |
453 | if (ecb_expect_false (current + size > end)) |
488 | if (ecb_expect_false (current + size > end)) |
454 | real_reserve_method (size); /* presumably noinline */ |
489 | real_reserve_method (size); /* presumably noinline */ |
455 | } |
490 | } |
456 | |
491 | |
457 | =item bool ecb_assume (cond) |
492 | =item ecb_assume (cond) |
458 | |
493 | |
459 | Try to tell the compiler that some condition is true, even if it's not |
494 | Tries to tell the compiler that some condition is true, even if it's not |
460 | obvious. |
495 | obvious. This is not a function, but a statement: it cannot be used in |
|
|
496 | another expression. |
461 | |
497 | |
462 | This can be used to teach the compiler about invariants or other |
498 | This can be used to teach the compiler about invariants or other |
463 | conditions that might improve code generation, but which are impossible to |
499 | conditions that might improve code generation, but which are impossible to |
464 | deduce form the code itself. |
500 | deduce form the code itself. |
465 | |
501 | |
… | |
… | |
482 | |
518 | |
483 | Then the compiler I<might> be able to optimise out the second call |
519 | Then the compiler I<might> be able to optimise out the second call |
484 | completely, as it knows that C<< current + 1 > end >> is false and the |
520 | completely, as it knows that C<< current + 1 > end >> is false and the |
485 | call will never be executed. |
521 | call will never be executed. |
486 | |
522 | |
487 | =item bool ecb_unreachable () |
523 | =item ecb_unreachable () |
488 | |
524 | |
489 | This function does nothing itself, except tell the compiler that it will |
525 | This function does nothing itself, except tell the compiler that it will |
490 | never be executed. Apart from suppressing a warning in some cases, this |
526 | never be executed. Apart from suppressing a warning in some cases, this |
491 | function can be used to implement C<ecb_assume> or similar functions. |
527 | function can be used to implement C<ecb_assume> or similar functionality. |
492 | |
528 | |
493 | =item bool ecb_prefetch (addr, rw, locality) |
529 | =item ecb_prefetch (addr, rw, locality) |
494 | |
530 | |
495 | Tells the compiler to try to prefetch memory at the given C<addr>ess |
531 | Tells the compiler to try to prefetch memory at the given C<addr>ess |
496 | for either reading (C<rw> = 0) or writing (C<rw> = 1). A C<locality> of |
532 | for either reading (C<rw> = 0) or writing (C<rw> = 1). A C<locality> of |
497 | C<0> means that there will only be one access later, C<3> means that |
533 | C<0> means that there will only be one access later, C<3> means that |
498 | the data will likely be accessed very often, and values in between mean |
534 | the data will likely be accessed very often, and values in between mean |
499 | something... in between. The memory pointed to by the address does not |
535 | something... in between. The memory pointed to by the address does not |
500 | need to be accessible (it could be a null pointer for example), but C<rw> |
536 | need to be accessible (it could be a null pointer for example), but C<rw> |
501 | and C<locality> must be compile-time constants. |
537 | and C<locality> must be compile-time constants. |
502 | |
538 | |
|
|
539 | This is a statement, not a function: you cannot use it as part of an |
|
|
540 | expression. |
|
|
541 | |
503 | An obvious way to use this is to prefetch some data far away, in a big |
542 | An obvious way to use this is to prefetch some data far away, in a big |
504 | array you loop over. This prefetches memory some 128 array elements later, |
543 | array you loop over. This prefetches memory some 128 array elements later, |
505 | in the hope that it will be ready when the CPU arrives at that location. |
544 | in the hope that it will be ready when the CPU arrives at that location. |
506 | |
545 | |
507 | int sum = 0; |
546 | int sum = 0; |
… | |
… | |
559 | |
598 | |
560 | =item bool ecb_is_pot32 (uint32_t x) |
599 | =item bool ecb_is_pot32 (uint32_t x) |
561 | |
600 | |
562 | =item bool ecb_is_pot64 (uint32_t x) |
601 | =item bool ecb_is_pot64 (uint32_t x) |
563 | |
602 | |
564 | Return true iff C<x> is a power of two or C<x == 0>. |
603 | Returns true iff C<x> is a power of two or C<x == 0>. |
565 | |
604 | |
566 | For smaller types then C<uint32_t> you can safely use C<ecb_is_pot32>. |
605 | For smaller types than C<uint32_t> you can safely use C<ecb_is_pot32>. |
567 | |
606 | |
568 | =item int ecb_ld32 (uint32_t x) |
607 | =item int ecb_ld32 (uint32_t x) |
569 | |
608 | |
570 | =item int ecb_ld64 (uint64_t x) |
609 | =item int ecb_ld64 (uint64_t x) |
571 | |
610 | |
… | |
… | |
647 | |
686 | |
648 | =head2 FLOATING POINT FIDDLING |
687 | =head2 FLOATING POINT FIDDLING |
649 | |
688 | |
650 | =over 4 |
689 | =over 4 |
651 | |
690 | |
|
|
691 | =item ECB_INFINITY |
|
|
692 | |
|
|
693 | Evaluates to positive infinity if supported by the platform, otherwise to |
|
|
694 | a truly huge number. |
|
|
695 | |
|
|
696 | =item ECB_NAN |
|
|
697 | |
|
|
698 | Evaluates to a quiet NAN if supported by the platform, otherwise to |
|
|
699 | C<ECB_INFINITY>. |
|
|
700 | |
|
|
701 | =item float ecb_ldexpf (float x, int exp) |
|
|
702 | |
|
|
703 | Same as C<ldexpf>, but always available. |
|
|
704 | |
652 | =item uint32_t ecb_float_to_binary32 (float x) [-UECB_NO_LIBM] |
705 | =item uint32_t ecb_float_to_binary32 (float x) [-UECB_NO_LIBM] |
653 | |
706 | |
654 | =item uint64_t ecb_double_to_binary64 (double x) [-UECB_NO_LIBM] |
707 | =item uint64_t ecb_double_to_binary64 (double x) [-UECB_NO_LIBM] |
655 | |
708 | |
656 | These functions each take an argument in the native C<float> or C<double> |
709 | These functions each take an argument in the native C<float> or C<double> |
… | |
… | |
679 | /* this results in a single add instruction to toggle the bit, and 4 extra */ |
732 | /* this results in a single add instruction to toggle the bit, and 4 extra */ |
680 | /* instructions to move the float value to an integer register and back. */ |
733 | /* instructions to move the float value to an integer register and back. */ |
681 | |
734 | |
682 | x = ecb_binary32_to_float (ecb_float_to_binary32 (x) ^ 0x80000000U) |
735 | x = ecb_binary32_to_float (ecb_float_to_binary32 (x) ^ 0x80000000U) |
683 | |
736 | |
|
|
737 | =item float ecb_binary16_to_float (uint16_t x) [-UECB_NO_LIBM] |
|
|
738 | |
684 | =item float ecb_binary32_to_float (uint32_t x) [-UECB_NO_LIBM] |
739 | =item float ecb_binary32_to_float (uint32_t x) [-UECB_NO_LIBM] |
685 | |
740 | |
686 | =item double ecb_binary32_to_double (uint64_t x) [-UECB_NO_LIBM] |
741 | =item double ecb_binary64_to_double (uint64_t x) [-UECB_NO_LIBM] |
687 | |
742 | |
688 | The reverse operation of the previos function - takes the bit representation |
743 | The reverse operation of the previous function - takes the bit |
689 | of an IEEE binary32 or binary64 number and converts it to the native C<float> |
744 | representation of an IEEE binary16, binary32 or binary64 number and |
690 | or C<double> format. |
745 | converts it to the native C<float> or C<double> format. |
691 | |
746 | |
692 | This function should work even when the native floating point format isn't |
747 | This function should work even when the native floating point format isn't |
693 | IEEE compliant, of course at a speed and code size penalty, and of course |
748 | IEEE compliant, of course at a speed and code size penalty, and of course |
694 | also within reasonable limits (it tries to convert normals and denormals, |
749 | also within reasonable limits (it tries to convert normals and denormals, |
695 | and might be lucky for infinities, and with extraordinary luck, also for |
750 | and might be lucky for infinities, and with extraordinary luck, also for |
… | |
… | |
785 | dependencies on the math library (usually called F<-lm>) - these are |
840 | dependencies on the math library (usually called F<-lm>) - these are |
786 | marked with [-UECB_NO_LIBM]. |
841 | marked with [-UECB_NO_LIBM]. |
787 | |
842 | |
788 | =back |
843 | =back |
789 | |
844 | |
|
|
845 | =head1 UNDOCUMENTED FUNCTIONALITY |
790 | |
846 | |
|
|
847 | F<ecb.h> is full of undocumented functionality as well, some of which is |
|
|
848 | intended to be internal-use only, some of which we forgot to document, and |
|
|
849 | some of which we hide because we are not sure we will keep the interface |
|
|
850 | stable. |
|
|
851 | |
|
|
852 | While you are welcome to rummage around and use whatever you find useful |
|
|
853 | (we can't stop you), keep in mind that we will change undocumented |
|
|
854 | functionality in incompatible ways without thinking twice, while we are |
|
|
855 | considerably more conservative with documented things. |
|
|
856 | |
|
|
857 | =head1 AUTHORS |
|
|
858 | |
|
|
859 | C<libecb> is designed and maintained by: |
|
|
860 | |
|
|
861 | Emanuele Giaquinta <e.giaquinta@glauco.it> |
|
|
862 | Marc Alexander Lehmann <schmorp@schmorp.de> |
|
|
863 | |
|
|
864 | |