On Aug 30, 7:28 pm, Martin Albrecht <[EMAIL PROTECTED]>
wrote:
> On Thursday 30 August 2007 19:11, mabshoff wrote:
>
> > > On a suse 101. amd64: SAGE compiled file. The sage -testall
> > > The above crash looks pretty serious -- it's evidently a 64-bit only
> > > issue, which was probably introduced by our work to fix other bugs (??)
> > > and/or memory leaks. This will be a blocker that has to be fixed before
> > > sage-2.8.3 can be released.
>
> I cannot reproduce this on my 64-bit Core2Duo system, i.e. 'make test' passes.
> So it is not a generic 64-bit problem.
>
> > The only thing we touched was the de-initialization of the tables in
> > m4ri, and I saw some potential issue in there with int* pointers:
>
> Could you elaborate why you see an issue with int* there? I don't see it.
>
Disclaumer: I am a little rusty regarding memory management in C, I
prefer C++ and size_t for pointers and new & delete. And I don't like
casts without overflow checks.
Okay, I might be totally wrong here, but: From the calloc man-page:
void *calloc(size_t nmemb, size_t size);
DESCRIPTION
calloc() allocates memory for an array of nmemb elements of
size bytes each and returns a pointer to the allocated memory. The
memory is set to zero.
So far so good. With
codebook[k]->ord =(int *)calloc(TWOPOW(k),sizeof(int));
you cast the void* to int* and assign it. Can we agree on that part?
Now where I see the problem: On a box with more than 2GB of RAM calloc
might return a pointer that lies above the range of signed int. Then
the cast to (int*) returns a pointer to the allocated memory module
2^31. That can be a reason for the code to go boom. I am not saying
(under the assumption that my analysis is correct) that it goes boom
in this place, but there might be similar issues elsewhere in the
code.
My apologies if that is total crap ;)
> Martin
>
Cheers,
Michael
> --
> name: Martin Albrecht
> _pgp:http://pgp.mit.edu:11371/pks/lookup?op=get&search=0x8EF0DC99
> _www:http://www.informatik.uni-bremen.de/~malb
> _jab: [EMAIL PROTECTED]
--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to sage-devel@googlegroups.com
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at http://groups.google.com/group/sage-devel
URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/
-~----------~----~----~----~------~----~------~--~---
