On Wednesday, March 23, 2016 at 4:50:10 PM UTC, Dima Pasechnik wrote:
>
>
>
> On Wednesday, March 23, 2016 at 3:33:11 PM UTC, Dima Pasechnik wrote:
>>
>>
>>
>> On Wednesday, March 23, 2016 at 3:06:46 PM UTC, Erik Bray wrote:
>>>
>>> On Wed, Mar 23, 2016 at 3:14 PM, Dima Pasechnik
>>> wrote:
>>
For the python-level leak, I think it's down to this:
x = polygen(ZZ)
v = x # or 1/3 or RR(3) but not, say 3
pre={id(c) for c in gc.get_objects()}
for _ in range(2): test =
sage.rings.polynomial.polynomial_element.Polynomial.__call__(x,v)
It ends up letting a lot of 1-element tuples of the
Strangely enough, it does not seem to be NTL related... I was able to
reproduce a somehow minimal Cython example without any use of Sage.
. See the report on Cython mailing list
https://groups.google.com/forum/#!topic/cython-users/g10b0911qq0
The reason why everything is fine with FLINT vers
On Wednesday, March 23, 2016 at 12:08:29 PM UTC-7, vdelecroix wrote:
>
>
> I am not sure about the Python analysis (though it might be some other
> problem).
>
> I observed the same thing. It looks like we're looking at two different
memory leaks in the same example:
1) A plain old malloc-type
On 23/03/16 15:39, Nils Bruin wrote:
On Wednesday, March 23, 2016 at 9:17:58 AM UTC-7, vdelecroix wrote:
Hello,
Some friend just sent an e-mail to me mentioning a memory leak. Here is
a minimal example
sage: x = polygen(ZZ)
sage: K = NumberField(x**3 - 2, 'cbrt2', embedding=RR(1.2599))
sage:
On Wednesday, March 23, 2016 at 9:17:58 AM UTC-7, vdelecroix wrote:
>
> Hello,
>
> Some friend just sent an e-mail to me mentioning a memory leak. Here is
> a minimal example
>
> sage: x = polygen(ZZ)
> sage: K = NumberField(x**3 - 2, 'cbrt2', embedding=RR(1.2599))
> sage: w = K.gen()
> sage:
Thanks Christian!
More specifically, it only concerns the NTL implementation backend. Here
is a more direct example
sage: R = PolynomialRing(ZZ, 'x', implementation='NTL')
sage: x = R.gen()
sage: p = x**2 - 3
sage: for _ in range(1): a = p(2)
sage: resource.getrusage(resource.RUSAGE_SELF).
I also found valgrind not very helpful here, but good old
code-dissection leads me to believe that the problem might originate in
the polynomial evaluation in the _richcmp_ routine in
src/sage/rings/number_field/number_field_element.pyx
That's because the following code shows the same leak
On Wednesday, March 23, 2016 at 3:33:11 PM UTC, Dima Pasechnik wrote:
>
>
>
> On Wednesday, March 23, 2016 at 3:06:46 PM UTC, Erik Bray wrote:
>>
>> On Wed, Mar 23, 2016 at 3:14 PM, Dima Pasechnik
>> wrote:
>> > Hi Erik,
>> >
>> > would this work on Windows running in a VM?
>> > Ideally, fo
Hello,
Some friend just sent an e-mail to me mentioning a memory leak. Here is
a minimal example
sage: x = polygen(ZZ)
sage: K = NumberField(x**3 - 2, 'cbrt2', embedding=RR(1.2599))
sage: w = K.gen()
sage: import resource
sage: resource.getrusage(resource.RUSAGE_SELF).ru_maxrss
180720
sage: fo
Hello,
I tried to use valgrind via sage-memcheck. The command "sage -advanced"
told me that the log should be in "/home/vincent/.sage". However I can
not found any file matching this...
In the script sage-valgrind the call to valgrind does not use the option
--log-file. I guess there is some
On Wednesday, March 23, 2016 at 3:06:46 PM UTC, Erik Bray wrote:
>
> On Wed, Mar 23, 2016 at 3:14 PM, Dima Pasechnik > wrote:
> > Hi Erik,
> >
> > would this work on Windows running in a VM?
> > Ideally, for testing purposes, that's what one would prefer over having
> > a dedicated box for
I'll bring a laptop with Windows 7 installed next week at Cernay.
--
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
On Wed, Mar 23, 2016 at 3:14 PM, Dima Pasechnik wrote:
> Hi Erik,
>
> would this work on Windows running in a VM?
> Ideally, for testing purposes, that's what one would prefer over having
> a dedicated box for testing.
It should work. I really ought to be doing that myself but I've been
using a
Hi Erik,
would this work on Windows running in a VM?
Ideally, for testing purposes, that's what one would prefer over having
a dedicated box for testing.
Dima
On Wednesday, March 23, 2016 at 11:15:58 AM UTC, Erik Bray wrote:
>
> Hi all,
>
> I've been working for a few weeks on an installer for
Hey Erik,
I will try it on my Windows 8 machine soon. Thank you for your work on
this. From my experience, Windows is very common in Asia, and this could be
another good approach to expanding Sage's influence in that region (in
addition to SMC).
Although I like the idea of supporting Sage vi
Hi all,
I've been working for a few weeks on an installer for Sage on Windows,
which takes advantage of Docker to accomplish this.* The goal of this
project is to make it possible to run Sage on Windows with as much
transparency as possible, such that the user isn't really aware that
there is any
unfortunately, it seems that setting max_behind_commits to 0 does not solve
the problem.
We will have to wait for the next beta. Hopefully soon ? Volker ?
I will try to solve this in the next patchbot release.
Frederic
Le mardi 22 mars 2016 08:37:12 UTC+1, Frédéric Chapoton a écrit :
>
> Hello
On Tuesday, March 22, 2016 at 10:49:44 PM UTC, John H Palmieri wrote:
>
>
>
> On Tuesday, March 22, 2016 at 3:24:13 PM UTC-7, David Roe wrote:
>>
>>
>>
>> On Tue, Mar 22, 2016 at 10:13 PM, Dima Pasechnik
>> wrote:
>>
>>> we now decompress tarfiles with python
>>>
>>> https://docs.python.org/2/l
On Wed, 23 Mar 2016, Jeroen Demeyer wrote:
Intentionally?
I don't think so.
OK. http://trac.sagemath.org/ticket/20255
I just reported this, hoping someone else to correct it. Also workaround
for 7.1 would be nice.
--
Jori Mäntysalo
On 2016-03-23 09:29, Jori Mäntysalo wrote:
Intentionally?
I don't think so.
--
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.c
It seems that on Sage 7.1 deprecated functions appears on html
documentation. This was not the case in 7.0.
Intentionally? If so, can I disable it?
--
Jori Mäntysalo
22 matches
Mail list logo
