This is the mail archive of the fortran@gcc.gnu.org mailing list for the GNU Fortran project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: Inlining?


Hi!

Yes, it's very general, I'm more talking about tiny routines (say some
20 lines), which I'm under the impression that do not get inlined, but
I'll give it a go with the additional flag you mentioned, and try to
read the manual a bit more carefully...

Thanks a lot!

Philippe

Tim Prince wrote:
> 
> Philippe Schaffnit wrote:
> > Hi!
> >
> > Thanks a lot for your reply!
> >
> > Sorry for not supplying an example! I am more like asking in a general
> > way... I have set "-finline-functions" and
> > "-finline-functions-called-once" but I have to admit that I overlooked
> > "-fkeep-inline-functions": I'll try to add it!
> >
> In a general way, the compiler must observe some target-dependent limits
> on the complexity of a function which is feasible for inlining.
> gfortran doesn't fall into the trap of certain commercial compilers,
> where the experts who work for the compiler vendor recommend splitting
> source so as to prevent the compiler getting stuck on inter-procedural
> analysis.  I trust you aren't running one of those scripts which use
> fsplit, or a more modern alternative, then put the pieces back together
> with a pre-link step such as ld -o bigobject -r littleobjects.  If you
> are speaking in a general way, such common practices aren't excluded.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]