"kolloid" <tom### [at] sobotanet> wrote:
> "kolloid" <tom### [at] sobotanet> wrote:
> > Le_Forgeron <jgr### [at] freefr> wrote:
> > > Le 09/02/2015 21:25, kolloid a écrit :
> > > (or where did you get it from ?)
> > >
> > > The 3.7-stable is one year old, and it worked at that time. But in the
> > > meantime, they changed automake and its default (someone never heard
> > > about backward compatibility, on the same line as someone at boost never
> > > heard of it either some years ago).
> > >
> > > The master branch have that part fixed (and more).
> >
> > Well, the default download is the 3.7-stable release, which is what I installed.
> > Now I downloaded master and will compile that, to see what happens.
> >
> > Agreed about automake. You end up having several incompatible versions
> > installed. Crazy ...
> >
> > Tom
>
> Ok, I built master. As you said, it's been fixed and now it compiles without
> problems. That's the good news.
>
> The bad news is that it bombed on the first script that I tested it with. On the
> second script also. A malloc corrupting memory somewhere. Tomorrow I will
> compile it with debug flags to see what is going on.
>
> So I reinstalled 3.7-stable and this runs perfectly...
>
> Tom

Ok, I rebuilt 3.7-master in debug mode.
The program is dying during parsing, in source/backend/math/splines.cpp,
function mkfree, at line 440. The code is:

se.resize(se.size()+1);

I notice that this code has changed from 3.7-stable (where it works, or at least
it doesn't bomb).

This happens when attempting to run the spectral rendering macros in
http://www.lilysoft.org/CGI/SR/Spectral%20Render.htm . I cannot see anything
related to splines in these macros, however.

My machine is 64 bits, running Linux Mint, based on Ubuntu 14.04.

I'm not an expert in STL, so I'm somewhat at a loss trying to find the error...

Tom

Post a reply to this message