On Jul 28, 2009, at 6:18 AM, Simon King wrote:

>
> Hi Robert,
>
> On Jul 28, 1:47 pm, Robert Bradshaw <[email protected]>
> wrote:
> ...
>> I would start up Sage without importing your code to get a baseline.
>> That's very surprising--I don't know much about valgrind, but it only
>> shows where the memory in question was allocated, not where it was
>> leaked.
>
> Yes, you are right: The memory leak also occurs when simply starting
> sage -valgrind and immediately quitting sage. See
> http://sage.math.washington.edu/home/SimonKing/logs/memcheck.basic
>
> But when you say that valgrind is not so informative, what else would
> you recommend?

Wish I knew. It's better than nothing though.

> And speaking about profiling: I also found that the performance of my
> programs is "leaking": The computation time for one cohomology ring
> depends on how many other rings have been computed before. Say, some
> example usually takes 10 seconds, but if the example is computed after
> 100 other examples, it takes 2 minutes.
>
> What tools are there for analysing such things? prun? What else?

That might help narrow things down, but it's impossible to say  
without just thinking about the code a lot.

- Robert

--~--~---------~--~----~------------~-------~--~----~
To post to this group, send email to [email protected]
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://www.sagemath.org
-~----------~----~----~----~------~----~------~--~---

Reply via email to