-
-
Notifications
You must be signed in to change notification settings - Fork 481
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Refcounting for Singular rings #11339
Comments
comment:2
The documentation is fairly clear, you are only allowed to touch C data structures in Really, this shows the difference between the C++ and the Python object model. In Python, executing constructors/destructors in derived classes is optional and leaves a lot of freedom to the programmer. Because of the garbage collection you can allocate resources wherever you want. By comparison, C++ is very strict and essentially forces you to use RAII. This is why there is no try/finally in C++. Which is why there is necessarily some tension in Cython between C++ and Python object instantiation/finalization. Having said that, it seems like the cleanest solution would be if Cython would also call the Python destructor
|
comment:3
Thanks for your input Volker. I think we should definitely put some work in libs/singular to have some |
comment:4
Just to clarify, right now Cython extension classes do not call the Python destructor |
comment:5
OK, well something as to be done about this if we want to move to python-2.7. I think this problem is probably the biggest issue in the way of #9958 there are a few other things to look at but that's the only one producing a segfault. |
comment:6
I notice that PyTypeObject has a member |
Workaround patch |
Attachment: bug11339a.patch.gz Workaround hg bundle |
comment:8
Attachment: bug11339a.bundle.gz For the record this works with sage-on-gentoo with python-2.7.1 so as far as I am concerned it works as advertised so far. I can do a review against a vanilla sage using python-2.6 later this week provided that I get the all clear to go back to work (we got a couple more earthquakes down here - the magnitude 6 one was scary but very few people got hurt). What is the "workaround hg bundle" Martin? I can reformat things if needed. |
comment:9
Replying to @kiwifb:
It's what I perceived to be the hg equivalent of a "bzr send": a file which describes a number of hg commits inclusing metadata, so that other hg users can pull from it. |
comment:10
Hi Martin, Steve sent me a note about it earlier this morning. The patch is the only thing necessary here, am I right? |
comment:11
Replying to @kiwifb:
This patch works fine for me too on sage-on-gentoo with python-2.7.1. I have applied this patch to the vanilla sage-4.7.1.alpha2 (python-2.6.4.p10) spkg and all long tests pass. I have also applied the patch along with the patches and required spkgs in ticket #9958 to the sage-4.7.1.alpha2 (python-2.7.1) spkg and all the tests that failed without this patch because of segfaults, i.e. sage -t -force_lib "devel/sage/sage/schemes/generic/scheme.py" sage -t -force_lib "devel/sage/sage/rings/morphism.pyx" sage -t -force_lib "devel/sage/sage/rings/homset.py" now pass. |
This comment has been minimized.
This comment has been minimized.
Reviewer: François Bissey, Steven Trogdon |
comment:12
I am giving this a positive review on my and Steve's observations. I unfortunately won't be able to run some extra tests before Monday it seems. |
comment:13
As it is, the patch is spectacularly ugly. I agree that it works, but we should investigate alternatives. It is not a particularly pressing issue since Sage is not going to switch to Python-2.7 this week. I'll have a closer look at it with Burcin duing this week at SD31. |
comment:14
I am ok with this. I don't plan to push python-2.7.1 in 4.7.1. I would like most of the other pending bits and pieces that will make testing easier to be in 4.7.1 however. It would be nice if at least some testing could be done in 4.7.2 and really land it by the next release. |
comment:15
Replying to @vbraun:
I agree, that's why I titled the patch "workaround" instead of "solution". I can't think of a way to address this in sage alone, though, as in my opinion a proper solution would entail modifications to cython as discussed above. |
This comment has been minimized.
This comment has been minimized.
comment:16
In the case at hand we don't really need the Cython class The patch should fix the crash, but I don't have a Sage-on-Python-2.7 install to test it. Can somebody give it a try? |
Author: Volker Braun, Martin von Gagern |
Changed work issues from Conflict with #11856 to none |
comment:84
You should run the tests with your patches applied on top of sage-4.7.2.alpha3. At least with the alpha3-prerelease, I get some doctest errors, for example So, it seems it would have been better to change this patch, not #11856 and #11068. Too bad. |
comment:85
The doctest failure in sage/misc/sageinspect (would have been nice if it would have used its own module as a doctest example btw) is fixed in #10903. There are no doctest failures if you use sage-4.7.2.alpha3 + #11339 + #10903. Really the two tickets should be one but thats not how it panned out historically. |
comment:86
Replying to @vbraun:
So we have a cyclic dependency now... Somehow matches the ticket's title. |
comment:87
Question to the release manager: How are the odds that both #11339/#10903 and #11856 (which both fix critical bugs) will be merged into sage-4.7.2? If #11339/#10903 are ready to be merged then I guess one can easily modify #11856 so that it matches (it is just a different line break in one hunk). And #11068 was shifted to sage-4.7.3 anyway, so, it can wait. |
comment:88
Replying to @vbraun:
The point was to have a cdef class, but sageinspect is pure Python, so it wouldn't have been a good example. |
comment:90
Replying to @vbraun:
Exactly. Just make sure you clearly state the order in which patches have to be applied. The "cyclic" dependency is not really a problem as these tickets will be merged together anyway. |
comment:91
Replying to @jdemeyer:
I'll implement merging strongly connected components... ;-) |
Dependencies: #10903 (for doctests) |
Merged: sage-4.7.2.alpha4 |
Ref counting for Singular rings
Python makes no guarantees that destructors are called if circular references are involved. This patch implements an extra Python proxy layer that correctly refcounts Singular rings. In contrast to our earlier code, it will release the singular rings once they are no longer in use. This triggers various issues in Sage/libSingular that are dealt with in the follow-up ticket #10903
Historic discussion
As described for the Sage on Gentoo project, there is a problem in how parts of sage use
__deallocate__
in their Cython code. I've actually traced an instance of groebner_strategy.pyx causing segmentation faults under Python 2.7.What happens is that the garbage collector invokes the tp_clear function for the object. Its implementation is provided by Cython, and one of its effects is that every python reference will be set to
None
. A bit later on, tp_dealloc is called and invokes the__deallocate__
method. By that time,_parent
isNone
, so accessing_parent._ring
is a bad idea, and in this case it turns out to be null.Others have had similar problems before. There are crude workarounds floating around. I guess a proper solution would be twaking cython to allow custom code in the tp_clear function. In other words, have a "magic" mehod
__clear__
similar to the magic__deallocate__
. But I'll wait for comments here first, before taking this to cython upstream. Perhaps people with more experience have better solutions to offer. And I won't mind if you decide to take this to Cython devs yourselves.Apply
Apply:
Depends on #10903
Dependencies: #10903 (for doctests)
CC: @kiwifb @robertwb @burcin @malb @simon-king-jena @nexttime
Component: algebra
Keywords: sd31
Author: Volker Braun, Martin von Gagern
Reviewer: François Bissey, Steven Trogdon, Martin Albrecht
Merged: sage-4.7.2.alpha4
Issue created by migration from https://trac.sagemath.org/ticket/11339
The text was updated successfully, but these errors were encountered: