Is there any difference when compilers implements C inline function in different platforms -
I was told that implementation (such as MIP vs. X86) for C inline functions in different platforms is a bit different. For example, inline function for one of them (amps or x86) still allocates a heap and thus is not so efficient, is this true? What are all the differences? Thanks
It is not guaranteed that inline inline (this is just a suggestion for a compiler) Em> any functions which are not underlined for architecture will be considered as normal tasks, and therefore Push & amp; When you call them (like any normal non-inline function) pops up from the stack
and how your compiler is configured, it is very much dependent. For example, there are such flags:
- Finnish functions -finline-limit = n - the ultimate name = value
Consider all the functions of inlineing, even if the inline has not been declared, the compiler predicts that In this way, which functions are being consolidated in this way.
By default, the GCC limits the size of tasks that can be inline. This flag allows thick control of this range is the size of the N function which can be inline in the number of pseudo instructions.
In some places, the GCC uses to control various constants that the amount of optimization, for example GCC functions In which there are more than certain instructions you can control some of these constants using the command --param option on the command line.
Comments
Post a Comment