Opened 12 years ago
Closed 11 years ago
#6348 closed defect (duplicate)
gap4.4.10->gap4.4.12 revisited
Reported by: | wdj | Owned by: | joyner |
---|---|---|---|
Priority: | major | Milestone: | sage-duplicate/invalid/wontfix |
Component: | group theory | Keywords: | gap |
Cc: | awebb | Merged in: | |
Authors: | D. Joyner | Reviewers: | |
Report Upstream: | N/A | Work issues: | |
Branch: | Commit: | ||
Dependencies: | Stopgaps: |
Description
From an email:
I think we can upgrade gap to 4.12 now too, since I worked around the problem we had before with 4.12 on itanium. William
Background: Some months ago, we upgraded from gap 4.4.10 to gap 4.4.12 (gap 4.4.11 was skipped due to some technical problems). This caused some docstrings to break, since some of gap's print methods were changed.
The patch at http://sage.math.washington.edu/home/wdj/patches/gap-4.4.12.p0.spkg will also break some docstrings. However, it applies fine (using sage -f) to sage 4.0.2.rc1.
A patch fixing the breaks will be posted later.
Attachments (2)
Change History (36)
Changed 12 years ago by
comment:1 Changed 12 years ago by
- Summary changed from gap4.4.10->gap4.4.12 revisited to [with patch, needs review] gap4.4.10->gap4.4.12 revisited
First do sage -f http://sage.math.washington.edu/home/wdj/patches/gap-4.4.12.p0.spkg This will cause several (minor) docstring test failures. All are fixed in this patch, which passes sage -testall.
comment:2 Changed 12 years ago by
- Summary changed from [with patch, needs review] gap4.4.10->gap4.4.12 revisited to [with patch, positive review pending doctests passing] gap4.4.10->gap4.4.12 revisited
I have looked at the patch, and it indeed is just printing/ordering issues. Spkg looks fine, SPKG.txt updated. Tests are still running...
comment:3 Changed 12 years ago by
Seems to have broken my install... maybe it was something else, but now I've got to fix my Sage :(.
comment:4 Changed 12 years ago by
Just curious what happened. You installed the spkg using sage -f then maybe ran sage -b and then sage -testall? This failed somehow and rendered your Sage unusable?
comment:5 Changed 12 years ago by
Yep--segfaulted on startup (so you can imagine how many tests failed...). It was acting a bit weird beforehand, so I'm not blaming this spkg for sure.
comment:6 Changed 12 years ago by
- Summary changed from [with patch, positive review pending doctests passing] gap4.4.10->gap4.4.12 revisited to [with patch, needs work] gap4.4.10->gap4.4.12 revisited
This seems to be not quite a positive review. I'm changing this back to "needs work," at least for now.
comment:7 Changed 12 years ago by
- Summary changed from [with patch, needs work] gap4.4.10->gap4.4.12 revisited to [with patch, needs review] gap4.4.10->gap4.4.12 revisited
I think my sage was messed up from elsewhere, so the spkg probably doesn't need work, but it still needs to be reviewed.
comment:8 Changed 12 years ago by
- Summary changed from [with patch, needs review] gap4.4.10->gap4.4.12 revisited to [with patch, needs work] gap4.4.10->gap4.4.12 revisited
The file sage/graphs/graph_isom.pyx was removed in ticket #6112. That breaks this patch so I guess I will switch it back to needs work.
comment:9 Changed 12 years ago by
I'm not sure how to fix the patch. Can I take the patch, open it in an editor, and delete the section containing the fixes for graph_isom.pyx, save it and rename the patch?
comment:10 follow-up: ↓ 11 Changed 12 years ago by
I believe that you can. I tried cutting out that part and I was able to apply the patch but I don't know anything about Mercurial or if this is the "proper" way to do it. My first thought is to apply the "cut" patch and then export a new one to replace the one here.
I see that graph_isom.pxd is still there although it is not needed. Should it be deleted? On this or a new ticket?
comment:11 in reply to: ↑ 10 Changed 12 years ago by
comment:12 follow-up: ↓ 13 Changed 12 years ago by
I installed the spkg and applied a modified patch (removed the graph_isom.pyx part). I ran 'sage -ptest and everything passed. I could not change the patch but otherwise it looks positive.
comment:13 in reply to: ↑ 12 Changed 12 years ago by
- Summary changed from [with patch, needs work] gap4.4.10->gap4.4.12 revisited to [with patch, needs review] gap4.4.10->gap4.4.12 revisited
Replying to awebb:
I installed the spkg and applied a modified patch (removed the graph_isom.pyx part). I ran 'sage -ptest and everything passed. I could not change the patch but otherwise it looks positive.
New patch (removed the graph_isom.pyx part) seems to solve the problem, passes sage -testall.
comment:14 Changed 12 years ago by
By the way, the best way to fix a patch is to use Mercurial queues. First, you use hg qimport
to import the patch, then hg qpush
to get it in the queue. Rebuild, edit, do whatever you like, and then you can use hg qrefresh
to incorporate the changes into the patch, and hg export tip > filename.patch
to get the new patch, and finally hg qpop
if you want the patch back off the queue again.
comment:15 follow-up: ↓ 17 Changed 12 years ago by
- Summary changed from [with patch, needs review] gap4.4.10->gap4.4.12 revisited to [with patch, needs work] gap4.4.10->gap4.4.12 revisited
After installing this spkg and applying the patch, I get the following:
---------------------------------------------------------------------- | Sage Version 4.1, Release Date: 2009-07-09 | | Type notebook() for the GUI, and license() for information. | ---------------------------------------------------------------------- sage: gap('2+2') A workspace appears to have been corrupted... automatically rebuilding (this is harmless). --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /home/rlmill/.sage/temp/geom/22741/_home_rlmill__sage_init_sage_0.py in <module>() /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in __call__(self, x, name) 1019 1020 if isinstance(x, basestring): -> 1021 return cls(self, x, name=name) 1022 try: 1023 return self._coerce_from_special_method(x) /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in __init__(self, parent, value, is_name, name) 1420 else: 1421 try: -> 1422 self._name = parent._create(value, name=name) 1423 except (TypeError, KeyboardInterrupt, RuntimeError, ValueError), x: 1424 self._session_number = -1 /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in _create(self, value, name) 1188 def _create(self, value, name=None): 1189 name = self._next_var_name() if name is None else name -> 1190 self.set(name, value) 1191 return name 1192 /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/gap.pyc in set(self, var, value) 529 """ 530 cmd = ('%s:=%s;;'%(var,value)).replace('\n','') --> 531 out = self._eval_line(cmd, allow_use_file=True) 532 533 def get(self, var, use_file=False): /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/gap.pyc in _eval_line(self, line, allow_use_file, wait_for_prompt) 690 try: 691 (normal, error) = self._execute_line(line, wait_for_prompt=wait_for_prompt, --> 692 expect_eof= (self._quit_string() in line)) 693 694 if len(error)> 0: /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/gap.pyc in _execute_line(self, line, wait_for_prompt, expect_eof) 590 current_outputs = normal_outputs 591 while True: --> 592 x = E.expect_list(self._compiled_full_pattern) 593 current_outputs.append(E.before) 594 if x == 0: # @p /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in __getattr__(self, attrname) 1309 def __getattr__(self, attrname): 1310 if attrname[:1] == "_": -> 1311 raise AttributeError 1312 return self._function_class()(self, attrname) 1313 AttributeError:
comment:16 Changed 12 years ago by
Also, after quitting the session above, I get a seemingly infinite cascade of the following:
Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored <bound method GapElement.__del__ of (invalid object -- defined in terms of closed session)> ignored Exception RuntimeError: 'maximum recursion depth exceeded in __subclasscheck__' in <type 'exceptions.AttributeError'> ignored Exception RuntimeError: 'maximum recursion depth exceeded in __subclasscheck__' in <type 'exceptions.RuntimeError'> ignored Exception RuntimeError: RuntimeError('maximum recursion depth exceeded',) in Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored <bound method GapElement.__del__ of (invalid object -- defined in terms of closed session)> ignored Exception RuntimeError: 'maximum recursion depth exceeded in __subclasscheck__' in <type 'exceptions.AttributeError'> ignored Exception RuntimeError: 'maximum recursion depth exceeded in __subclasscheck__' in <type 'exceptions.RuntimeError'> ignored Exception RuntimeError: RuntimeError('maximum recursion depth exceeded',) in Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored
comment:17 in reply to: ↑ 15 Changed 12 years ago by
Replying to rlm:
After installing this spkg and applying the patch, I get the following:
---------------------------------------------------------------------- | Sage Version 4.1, Release Date: 2009-07-09 | | Type notebook() for the GUI, and license() for information. | ---------------------------------------------------------------------- sage: gap('2+2') A workspace appears to have been corrupted... automatically rebuilding (this is harmless). --------------------------------------------------------------------------- AttributeError Traceback (most recent call last) /home/rlmill/.sage/temp/geom/22741/_home_rlmill__sage_init_sage_0.py in <module>() /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in __call__(self, x, name) 1019 1020 if isinstance(x, basestring): -> 1021 return cls(self, x, name=name) 1022 try: 1023 return self._coerce_from_special_method(x) /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in __init__(self, parent, value, is_name, name) 1420 else: 1421 try: -> 1422 self._name = parent._create(value, name=name) 1423 except (TypeError, KeyboardInterrupt, RuntimeError, ValueError), x: 1424 self._session_number = -1 /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in _create(self, value, name) 1188 def _create(self, value, name=None): 1189 name = self._next_var_name() if name is None else name -> 1190 self.set(name, value) 1191 return name 1192 /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/gap.pyc in set(self, var, value) 529 """ 530 cmd = ('%s:=%s;;'%(var,value)).replace('\n','') --> 531 out = self._eval_line(cmd, allow_use_file=True) 532 533 def get(self, var, use_file=False): /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/gap.pyc in _eval_line(self, line, allow_use_file, wait_for_prompt) 690 try: 691 (normal, error) = self._execute_line(line, wait_for_prompt=wait_for_prompt, --> 692 expect_eof= (self._quit_string() in line)) 693 694 if len(error)> 0: /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/gap.pyc in _execute_line(self, line, wait_for_prompt, expect_eof) 590 current_outputs = normal_outputs 591 while True: --> 592 x = E.expect_list(self._compiled_full_pattern) 593 current_outputs.append(E.before) 594 if x == 0: # @p /space/rlm/sage-4.1/local/lib/python2.6/site-packages/sage/interfaces/expect.pyc in __getattr__(self, attrname) 1309 def __getattr__(self, attrname): 1310 if attrname[:1] == "_": -> 1311 raise AttributeError 1312 return self._function_class()(self, attrname) 1313 AttributeError:
Details???
I get:
wdj@hera:~/sagefiles/sage-4.1.rc0$ ./sage ---------------------------------------------------------------------- | Sage Version 4.1.rc0, Release Date: 2009-07-04 | | Type notebook() for the GUI, and license() for information. | ---------------------------------------------------------------------- Loading Sage library. Current Mercurial branch is: gap sage: gap("2+2") 4
This is in amd64 ubuntu 9.04, phemon chip.
comment:18 follow-up: ↓ 19 Changed 12 years ago by
On sage.math.
comment:19 in reply to: ↑ 18 Changed 12 years ago by
Replying to rlm:
On sage.math.
I had no problem on an amd64 ubuntu 9.04 machine and likewise for an intel macbook running 10.4.11.
I tried to build sage 4.1 from source on sage.math and ran into some compiling errors. Can you tell me what version of sage you used on sage.math?
comment:20 Changed 12 years ago by
This was a fresh build of Sage-4.1. What errors did you run into when building sage-4.1?
comment:21 Changed 12 years ago by
I ran into an error building atlas. I don't know how to replicate your error so don't know how to try to fix this ticket. As I said, it works on my 2 machines. Do you have any suggestions?
comment:22 Changed 12 years ago by
Strangely, doing the exact same thing again gave different results.
I was able to install gap-4.4.12.p0
on top of sage-4.1
on sage.math, and after applying the patch, I'm running long doctests everywhere. Will post the results shortly.
comment:23 follow-up: ↓ 24 Changed 12 years ago by
All tests pass!
comment:24 in reply to: ↑ 23 Changed 12 years ago by
Replying to rlm:
All tests pass!
Whew (with hand wiping sweating brow:-)!
Does this mean it gets a positive review?
comment:25 Changed 12 years ago by
The Mercurial repo is corrupted, but I am posting an spkg that will fix this. Tell me, what is the purpose of the patch
file that was added? I can't read it in plain text, and I don't know what format it is...
comment:26 Changed 12 years ago by
Hmm, I've deleted that patch
file, reinstalled (this time on 32-bit OS X), and I'm running long doctests on this. Next I'll test 64-bit OS X. I don't know what patch
was, but it doesn't seem necessary.
comment:27 Changed 12 years ago by
Oops, the AttributeError
rears its ugly head again: this time on OS X.
My modified spkg can be found here: http://sage.math.washington.edu/home/rlmill/gap-4.4.12.p1.spkg
Here is one example of what is happening:
sage -t -long sage/combinat/partition.py *** *** Error: TIMED OUT! PROCESS KILLED! *** *** *** *** Error: TIMED OUT! *** *** ********************************************************************** File "/Users/rlmill/sage-4.1.32bit/devel/sage-main/sage/combinat/partition.py", line 523: sage: Partition([5,3]).sign() Exception raised: Traceback (most recent call last): File "/Users/rlmill/sage-4.1.32bit/local/bin/ncadoctest.py", line 1231, in run_one_test self.run_one_example(test, example, filename, compileflags) File "/Users/rlmill/sage-4.1.32bit/local/bin/sagedoctest.py", line 38, in run_one_example OrigDocTestRunner.run_one_example(self, test, example, filename, compileflags) File "/Users/rlmill/sage-4.1.32bit/local/bin/ncadoctest.py", line 1172, in run_one_example compileflags, 1) in test.globs File "<doctest __main__.example_10[2]>", line 1, in <module> Partition([Integer(5),Integer(3)]).sign()###line 523: sage: Partition([5,3]).sign() File "/Users/rlmill/sage-4.1.32bit/local/lib/python/site-packages/sage/combinat/partition.py", line 580, in sign ans=gap.eval("SignPartition(%s)"%(self)) File "/Users/rlmill/sage-4.1.32bit/local/lib/python/site-packages/sage/interfaces/gap.py", line 480, in eval result = Expect.eval(self, input_line, **kwds) File "/Users/rlmill/sage-4.1.32bit/local/lib/python/site-packages/sage/interfaces/expect.py", line 972, in eval return '\n'.join([self._eval_line(L, **kwds) for L in code.split('\n') if L != '']) File "/Users/rlmill/sage-4.1.32bit/local/lib/python/site-packages/sage/interfaces/gap.py", line 692, in _eval_line expect_eof= (self._quit_string() in line)) File "/Users/rlmill/sage-4.1.32bit/local/lib/python/site-packages/sage/interfaces/gap.py", line 592, in _execute_line x = E.expect_list(self._compiled_full_pattern) File "/Users/rlmill/sage-4.1.32bit/local/lib/python/site-packages/sage/interfaces/expect.py", line 1311, in __getattr__ raise AttributeError AttributeError **********************************************************************
comment:28 Changed 12 years ago by
These tests also ended with the seemingly infinite barrage of
Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored
comment:29 Changed 12 years ago by
These tests also pass the second time around. Could this have something to do with temp files or workspaces or something?
comment:30 Changed 12 years ago by
I don't see how that problem can arise. There were reasons why 4.4.12 was downgraded to 4.4.10. Maybe these problems persist?
The "patch" was (as described in one of the comments above) simply the original patch but with the section on graph_isom.pyx removed (I opened up the old patch, deleted the lines in the graph_isom.pyx section in emacs, and saved that as a new file). The problem was that the original patch would not apply since graph_isom.pyx has since been removed. Fortunately the edited patch worked (at least for me).
comment:31 Changed 12 years ago by
Patches to the Sage library don't belong in spkgs-- but this is gone in gap-4.4.12.p1
.
My guess is that you can reproduce this behavior on sage.math as follows:
Install a fresh sage-4.1
, and when it is finished, run the full doctest suite on long. This will guarantee that gap-4.4.10
gets used nontrivially. Then upgrade to gap-4.4.12.p1
. Run the doctests again.
If I copy over a sage-4.1 install and run long doctests, after having successfully used gap-4.4.12, the gap-4.4.10 in there has problems which look just like the problems I had going the other way:
Traceback (most recent call last): File "/space/rlm/sage-4.1.gap/tmp/all.py", line 2, in <module> from sage.all_cmdline import *; File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/all_cmdline.py", line 14, in <module> from sage.all import * File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/all.py", line 64, in <module> from sage.misc.all import * # takes a while File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/misc/all.py", line 70, in <module> from sage_input import sage_input File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/misc/sage_input.py", line 163, in <module> from sage.misc.functional import parent File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/misc/functional.py", line 37, in <module> from sage.rings.complex_double import CDF File "complex_double.pyx", line 88, in sage.rings.complex_double (sage/rings/complex_double.c:13486) File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/rings/complex_field.py", line 86, in ComplexField C = ComplexField_class(prec) File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/rings/complex_field.py", line 177, in __init__ self._populate_coercion_lists_(coerce_list=[complex_number.RRtoCC(self._real_field(), self)]) File "complex_number.pyx", line 1937, in sage.rings.complex_number.RRtoCC.__init__ (sage/rings/complex_number.c:12660) File "map.pyx", line 41, in sage.categories.map.Map.__init__ (sage/categories/map.c:1909) File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/categories/homset.py", line 149, in Hom from sage.rings.homset import RingHomset File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/rings/homset.py", line 17, in <module> import quotient_ring File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/rings/quotient_ring.py", line 32, in <module> import sage.rings.polynomial.multi_polynomial_ideal File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/rings/polynomial/multi_polynomial_ideal.py", line 229, in <module> from sage.interfaces.all import (singular as singular_default, File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/interfaces/all.py", line 8, in <module> from gap import gap, gap_reset_workspace, gap_console, gap_version, is_GapElement, Gap File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/interfaces/gap.py", line 960, in <module> gap_reset_workspace(verbose=False) File "/space/rlm/sage-4.1.gap/local/lib/python/site-packages/sage/interfaces/gap.py", line 938, in gap_reset_workspace os.unlink(WORKSPACE) OSError: [Errno 2] No such file or directory: '/home/rlmill/.sage//gap/workspace-3758812723191186113'
It looks like this happens whenever the version of GAP used is changed. This is why I was hinting at temp files and workspaces before. Something is ringing a bell, about GAP failing to reset its own workspace on upgrade.
comment:32 Changed 12 years ago by
Hmm, after doing the above, upgrading went just fine. I just can't reproduce the AssertionError
, even though I've seen it on Linux and OS X. Maybe it's because of the state of my DOT_SAGE
directories.
I'm almost ready to give this a positive review, but it makes me uneasy, since it seems each user will experience this once on upgrade, possibly even if they've just installed a new version from scratch...
comment:33 Changed 12 years ago by
- Cc awebb added
comment:34 Changed 11 years ago by
- Milestone changed from sage-4.3.2 to sage-duplicate/invalid/wontfix
- Report Upstream set to N/A
- Resolution set to duplicate
- Status changed from needs_work to closed
- Summary changed from [with patch, needs work] gap4.4.10->gap4.4.12 revisited to gap4.4.10->gap4.4.12 revisited
Closing this as a duplicate of #8076.
applies to 4.0.2.rc1