On Friday, 8 May 2015 09:27:57 UTC+1, Nathann Cohen wrote:
>
> Small update on the 'modular decomposition story'.
>
> Today I ran valgrind on the code, and ended up finding where the error
> comes from. Around line 972 of dm.c, one can find:
>
> for(v = n-1; v>=0; v--)
> if(ds[v-1] != -1){
> L2[v]=v;
> while( pile[sommet] < ds[v-1])
> L2[pile[sommet--]]=v;
> }
>
> Now, because v can be equal to 0 in the loop, ds[v-1] is actually
> ds[-1] and leads, unsurprisingly, to a wrong area of the memory.
> Valgrind reports it like that:
>
> ==23980== 1 errors in context 3 of 4:
> ==23980== Invalid read of size 4
> ==23980== at 0x40269D: algo2 (dm.c:972)
>
> Thus it is rather obvious where the error comes from (there are some
> other occurrences of the same problem). I was about to write an email
> to the authors, when I noticed that..... I had already sent an email
> with the very same information, i.e. line number+explanation+short
> tutorial on valgrind, and that was... one year ago. On the 6th of
> April 2014, to be specific.
>
Was it about the same version of their code?
Maybe we should tell the code authors that we will have to remove it from
Sage if they will not fix the bug?
(and not have certainly wrong code in Sage?)
>
> So that is not a problem of having to debug under mac OS X.
>
> Nathann
>
--
You received this message because you are subscribed to the Google Groups
"sage-support" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sage-support.
For more options, visit https://groups.google.com/d/optout.
