Shriramana Sharma wrote:
> In man:gcc under -fkeep-inline-functions I first came across the usage
> of "extern inline". For normal functions, extern declares that the
> function exists elsewhere. But this does not seem to work for inline
> functions:
>
> lib.cpp:
Note that way in which "inline" is handled differs between C and C++.
> ---------------------------------------
> inline int foo ( void ) { return 2 ; }
> ---------------------------------------
>
> main.cpp:
> ---------------------------------------
> # include <cstdio>
> extern inline int foo ( void ) ;
> int main ( void ) {
> printf ( "%d\n", foo () ) ;
> }
> ---------------------------------------
>
> $ g++ -c main.cpp lib.cpp
> main.cpp:3: warning: inline function -Fʽint foo()ʼ used but never defined-A
> $ g++ -o main main.o lib.o
> main.o: In function `main':
> main.cpp:(.text+0x12): undefined reference to `foo()'
> collect2: ld returned 1 exit status
> $
>
> nm on lib.o returned absolutely nothing! Apparently code for an inline
> function is not produced if it is never called.
True for C++, not for C.
> If it is called, and it
> is not labeled static inline, then it is both inlined and compiled
> separately. Strange, but ok.
>
> So I tried adding a dummy function to lib.cpp calling foo() just to make
> it get compiled. Then compiling main.cpp gave the same warning as above,
> but linking and execution went on ok.
>
> But I discovered that I don't need to have the inline keyword in the
> declaration of foo() in main.cpp. In fact, removing the inline keyword
> allows compilation without warning.
>
> So what is the *unique* use of extern inline? I mean, where we cannot do
> without it?
"extern inline" is meaningless. You can only inline functions which
occur in the current translation unit.
--
Glynn Clements <glynn@xxxxxxxxxxxxxxxxxx>
-
To unsubscribe from this list: send the line "unsubscribe linux-c-programming" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
