[Bug middle-end/64621] New: MIssed tail call oppurtunity

[Bug middle-end/64621] New: MIssed tail call oppurtunity

[hubicka at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64621

            Bug ID: 64621
           Summary: MIssed tail call oppurtunity
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: middle-end
          Assignee: unassigned at gcc dot gnu.org
          Reporter: hubicka at gcc dot gnu.org

Compiling the following at x86_64:
extern long int strtol (__const char *__restrict __nptr,                        
   char **__restrict __endptr, int __base)                                      
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1))) ;          
extern __inline int                                                             
__attribute__ ((__nothrow__)) atoi (__const char *__nptr)                       
{                                                                               
  return (int) strtol (__nptr, (char **) ((void *)0), 10);                      
}                                                                               

will not produce tailcall even though it is perfectly possible because of the
extra (int) conversion that is noop and could be ignored.

[Bug middle-end/64621] MIssed tail call oppurtunity

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64621

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |INVALID

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Well, this really really asks for the function signatures to be updated to the
call ABI at the GIMPLE level...

OTOH, (int) strtol (...) possibly truncates the value so it's not a noop
(it's a noop on i?86 only).

[Bug middle-end/64621] MIssed tail call oppurtunity

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=64621

--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #1)
> Well, this really really asks for the function signatures to be updated to
> the
> call ABI at the GIMPLE level...
> 
> OTOH, (int) strtol (...) possibly truncates the value so it's not a noop
> (it's a noop on i?86 only).

Actually it is a NOP on targets which define the upper bits of the register
being undefined when returning.  That includes x86_64 and aarch64.

Thanks,
Andrew Pinski