[openbeos] Re: gcc-2.95.3
- From: Oliver Tappe <openbeos@xxxxxxxxxxxxxxx>
- To: openbeos@xxxxxxxxxxxxx
- Date: Sun, 15 Aug 2004 00:18:14 +0200
Hi,
1) A new test-version of gcc-2.95.3 is available. You can find it here:
http://www.hirschkaefer.de/beos/gcc-2.95.3-zooey-test2.zip
I have tested it successfully with quite a set of different projects, but I
think I have reached a point where broader testing makes sense.
So it'd be great if people interested would test this version and tell me
where it fails (or improves upon gnupro >:o). Information about this
compiler's influence on performance and/or memory footprint of applications
is highly appreciated, too!
The impression I've got so far from my own testing is that gcc-2.95.3 is
quite a bit slower than the versions based on gnupro but that it crashes
less often and generates smaller code.
I haven't done any performance comparisons yet, as my first target is
correctness. Trying to compile the haiku-source exhibits a couple of new
warnings and four new errors, so there's still some more work to do...
Two new errors are particularly interesting:
- addons/kernel/file_systems/udp/udp.cpp fails to compile
- tests/kernel/core/util/OrderedMapTest.h triggers an internal compiler
error >:o(
Please tell me if you have any idea what might be the cause...
2) Additionally, there's a "fixed" version of Takashi's gcc-2.95.3
available from here:
http://www.hirschkaefer.de/beos/gcc-2.95.3-takashi.zip
The "fixing" just involved replacing the libstdc++ headers with the ones
from gnupro-000224. If you want to test this one, please copy stl_alloc.h
from the include/g++-2 folder into /boot/develop/headers/cpp (but keep a
backup of the one that lives there, you'll need it if you switch to another
compiler!).
cheers,
Oliver
- Follow-Ups:
- [openbeos] Re: gcc-2.95.3
- From: Ingo Weinhold
- [openbeos] Re: gcc-2.95.3
- From: Eric Petit
- References:
- [openbeos] Re: gcc-2.95.3
- From: Axel Dörfler
- [openbeos] Re: gcc-2.95.3
- From: Oliver Tappe
Other related posts:
- » [openbeos] gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- » [openbeos] Re: gcc-2.95.3
- [openbeos] Re: gcc-2.95.3
- From: Ingo Weinhold
- [openbeos] Re: gcc-2.95.3
- From: Eric Petit
- [openbeos] Re: gcc-2.95.3
- From: Axel Dörfler
- [openbeos] Re: gcc-2.95.3
- From: Oliver Tappe