This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


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

Re: Are egcs optimizations safe for conservative GC?


> Date: Wed, 05 Jan 2000 00:57:15 -0500
> From: Kevin Atkinson <kevinatk@home.com>

> I will say so.  Won't this KILL performace.  With out being able to put
> things in registers is there much the optimizer can do?

That is easier for you to answer.  Just compile up your application
with and without it, and run it.  Differences are potential bugs, and
the slowdowns (if any) will tell you exactly what your performance
loss is.  Feel free to share your experience.

Also, bear in mind, you may be able to compile 95% of your app with
-fvolatile (for correctness) and the 5% that accounts for most of the
runtime you may be able to compile without and discover that
everything just works.

> But does jeff fell that way?

I engineeered my answer to be agreeable by all, or tried to anyway.

> I bet you he can right now list about 50 bugs by pointing out ways
> gcc will spoke a garbage detector when compiled with -O2.

Yes, I know, and if someone wanted to step forward to proactively fix
bugs, he might be willing to list some of them.  Also, I suspect there
are others that once work was started in this area, that might have
pointers and ideas about other parts of the code.

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