Hi,
On Wednesday, October 18, 2006, at 11:29 AM, Steve Langasek wrote:
So given that you are concerned about the consequences that dropping
m68k
from the release will have on our roots, when can we expect that you
will be
doing your part to get m68k back into a releasable state by fixing the
kaffe
build failure there?
http://buildd.debian.org/fetch.cgi?pkg=kaffe;ver=2%3A1.1.7-
3;arch=m68k;stamp=1153832010
well, yes, testing kaffe on linux/68k again is one of my goals for
updating my 68k macs. I will test if current CVS head works, if not if
it can be fixed and if a patch for 1.1.7 can be generated.
Unfortunately when 1.1.7 was released, I was already not able to test
linux/68k, but IIRC correctly, netbsd/68k worked, so I suspect more a
build problem.
I notice from your buildlog that the classpath build is done using ecj.
I personally use jikes on any platform I test. Furthermore, on slow
architectures, I use a prebuild classpath jar, but I don't know if that
can be easily done in a package build system. Jikes requires for a
current classpath > 256MB of ram and plenty of swap space and it can
consume a big part of the build time, thus slowing down the testing
cycle considerably.
Gcj could be attemtped too and afaik it has means in the classpath
build to split up the compiation and thus require less resources, I
have never tried it either but now that gcc 4.1 is working on 68k, I
suppose gcj is usable too?
Cheers,
Riccardo
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
