Hi Nadim, On 2019-02-12, Nadim Rustom <restom.na...@gmail.com> wrote: > The following shows a small memory leak: > R.<X,Y> = ZZ[] > > > P = (X+Y)^120*Y^100 > > > mem1 = get_memory_usage() > for i in range(100): > Q = P(X,Y) > mem2 = get_memory_usage() > > > print mem2 - mem1 > > > whereas the following gives a much bigger memory leak: > > R.<X,Y> = ZZ[] > > > P = (X+Y)^120*Y^100 > > > mem1 = get_memory_usage() > for i in range(100): > Q = P(X,Y) > mem2 = get_memory_usage()
The two examples are identical. Copy-and-paste error? In any case, I can confirm that it leaks (and I am surprised that it does), even if python's cyclic garbage collection is invoked explicitly. > Additionally, I noticed that operations with multivariate polynomials are > way faster on Sage 7.4 than on Sage 8.*. Is there any work around for this > problem? I tried to search in the open tickets, the closest thing I could > find is this: > > https://trac.sagemath.org/ticket/13447 > > Could this be what is causing these issues? Certainly not, since #13447 is not merged yet. Leaks could be in at least two places here: (1) An unsolicited reference chain that prevents a Sage polynomial from being garbage collected. (2) The underlying libsingular polynomial object is not freed when the Sage polynomial becomes collected. #13447 tries to sanitize the refcounting of libsingular objects. So, if anything, #13477 could have the potential to fix that memory leak. And in addition, the examples given on the ticket show that it results in a speed-up for creation and deletion of polynomials. Before opening a new ticket for this issue, I'd like to test if #13477 helps with this issue. Best regards, Simon -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To post to this group, send email to sage-devel@googlegroups.com. Visit this group at https://groups.google.com/group/sage-devel. For more options, visit https://groups.google.com/d/optout.
