This is the mail archive of the java-discuss@sources.redhat.com mailing list for the Java project.


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

Inlining (was: Re: GCJ macro in libjava/Makefile)


Alexandre Petit-Bianco wrote:
> > Inlining does work for me now with -O and above, provided that the
> > classes are in the same source.
> 
> And, I believe, provided that you use this gcc patch which got never
> approved:
> 
>   http://gcc.gnu.org/ml/gcc-patches/1999-11/msg00628.html

Interesting... my test code _did_ inline, even without the calls.c patch.  I
double-checked it to be sure.

Too bad the optimizing compiler is currently broken :-(

Anybody else seeing stack corruption when using gcj -O2?  I guess I should try
to assemble a test case, though there are several in libgcj.

> But anyways, further inlining work should emphasize inlining at the
> tree level.

Yes.  At least for Java source.  Bytecode is another matter...

--
Jeff Sturm
jeff.sturm@commerceone.com

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