[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Fwd: Help needed debugging on ARM

On Thu, Aug 30, 2012 at 07:13:29PM +0200, Michael Wild wrote:
> That would explain that particular failure. Would splitting of the build
> into -arch and -indep help in the sense that the -indep packages only
> get built on one architecture? If so, which one would that be?

Well it would partially solve it.  The package maintainer generally
uploads the -indep and one architecture as far as I can tell (I am not
a debian developer so I may have misunderstood this part).  I think the
buildd's only build the architecture specific part.

Of course I could also then file a release critical FTBFS bug against
the package since it would fail to compile completely from source on
one of the many architectures I have around.

So no it is NOT a solution.

It is however still a good idea to do the split since you don't want to
waste time generating documentation on all build machines, when it should
end up identical to what you could generate on one of the fast ones.

> That's my mistake. I simply forgot to take care of the non-Intel
> architectures when I ported the upstream project's (OpenFOAM) build
> system to CMake. I have those failures fixed already locally, but before
> filing an RFS I wanted to get to the bottom of the asymptote-failure...

Well fixing that is probably also necesary before arm will build properly.

> Yes, my co-maintainer probably just didn't know about that. I have that
> fixed also locally, but don't think it would get past the freeze, right?

I have no idea.  I don't think the package as is will get past any freeze.

> The doxygen-fixup is necessary because of the way that OpenFOAM
> organizes it's source tree. Doxygen just doesn't seem to like what I did
> in FreeFOAM, and thus the output needs patching. Possibly I could speed
> it up a lot, I just didn't have the time to look into that.

Well I can build a linux kernel in about 1 minute and 15 seconds on this
powerpc machine.  It takes over 10 minutes to get to the point where
openfoam fails (which cmake seems to claim is 3% in to the build).
The document cleanup python script took minutes to run, and this is not
a slow machine (It's a 3.7GHz power7 box).

I would hate to think what this package would do to the arm or mips
build machines.

Len Sorensen

Reply to: