Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Hpcgap merge nov25 #375

Merged
merged 116 commits into from
Nov 27, 2015
Merged

Hpcgap merge nov25 #375

merged 116 commits into from
Nov 27, 2015

Conversation

ChrisJefferson
Copy link
Contributor

New mega-merge for HPC-GAP.

EDIT: Now passing all tests (hopefully)

The biggest issue is that the new Timeout functionality doesn't work. I haven't investigated at all how hard it would be to fix it, but by "luck", as HPC-GAP doesn't use configure, it compiles as disabled so users can't call it anyway.

Alexander Konovalov and others added 30 commits September 22, 2015 16:11
Draft overview of changes in GAP 4.8 beta.
bae58af
@ChrisJefferson
Add optimisation to syFgetsNoEdit to get string straight from buffer
29666dd
@ChrisJefferson
Add C implementation of InstallHiddenTrueMethod, and methods involvin…
a1207fe 
…g HIDDEN_IMP
@ChrisJefferson
Add UncheckedIS_SUBSET_FLAGS, which avoids checking the types of it's…
a175a03 
… input
Incorporated suggestions by Frank and Max.
ff01b89
@ChrisJefferson
Fix warnings when HAVE_POPEN is undefined
7eaf127
@ChrisJefferson
Merge pull request gap-system#286 from ChrisJefferson/new_quick_gets
3edb1ed 
Add optimisation to syFgetsNoEdit to get string straight from buffer
@ChrisJefferson
Merge pull request gap-system#288 from ChrisJefferson/hiddentrue
ab43c98 
Port HiddenTrueMethod related code to C
@ChrisJefferson
Fix cygwin warnings
0ee7da9
@ChrisJefferson
Remove XSET_ type-caching optimisations
61a1ba4
@hulpke
GQuotients from fp group is slightly more clever with kernel computat…
88f9f91 
…ion.
@alex-konovalov
Merge pull request gap-system#250 from alex-konovalov/changes-4.8
ae048a4 
Draft overview of changes in GAP 4.8 beta.
Renormalised GAP4Stones.
79206f6 
Run tests on Ubuntu 8-core machine. No tweaking of number afterwards.
@markuspf
Merge pull request gap-system#297 from ChrisJefferson/XSET_removal
06eda0a 
Remove XSET_ type-caching optimisations
@hulpke
Method for multiplication of automorphisms should not apply to iso*auto.
a08b278
@hulpke
Further additions. AH
04fadad
@hulpke
Next
9eb561d
@hulpke
Next
ac1af47
@ChrisJefferson
Improve profile coverage of if and for-range statements
0897a72
@hulpke
Next step of Magma translator: Process folder with files and deal with
4b42040 
dependencies, namespaces.
@ChrisJefferson
Merge pull request gap-system#299 from ChrisJefferson/profilecoverage
753c0eb 
Improve profile coverage of else and for-range statements
@ChrisJefferson
Clean up profiling and add some code documentation
60969a5
@ChrisJefferson
Merge branch 'profilecoverage'
4ab3d35
@ChrisJefferson
Add some more documentation for profiling
2768776
@alex-konovalov
Merge pull request gap-system#301 from ChrisJefferson/profiledoc
edddca1 
Add some more documentation for profiling
@markuspf
Initialise GAPInfo.LoadPackageLevel
2b487c3 
* This is not properly initialised if one does not have any
  needed packages
@ChrisJefferson
Make IsDirectoryPath take Directory objects, and add some simple dire…
9c9be72 
…ctory tests
@markuspf
Replace NormedVectors by NormedRowVectors
ea33adf 
* NormedVectors is declared obsolete in `lib/obsolete.gd`
@markuspf
Remove use of MutableIdentityMat
f50af2e 
This is obsolete according to `lib/obsolete.gd`, and identical with
IdentityMat.
@markuspf
Remove 2 uses of Tuple() from norad.gi
79a6994 
Tuple is marked as obsolete.
markuspf and others added 14 commits November 24, 2015 12:47
@markuspf
Merge pull request gap-system#317 from james-d-mitchell/semigroups
cf7c8d9 
Semigroups updates.
@markuspf
Cleanup commented out code
cfbd63c
@markuspf
Move hashfunctions into separate file
73e8e20 
This removes a dependency of dict.gi on the rest of the library.
@markuspf
Move HashKeyBag around, don't load hash.g{d,i}
3406937
@markuspf
Delete hash.g{d,i}
a8c1323
@markuspf
Merge pull request gap-system#320 from markuspf/dict-cleanup
2f9c069 
Cleanup dict.g{i,d}, remove hash.g{i,d}
Updated Tutorial examples changed after recent commits by AH
4f2fd99
@fingolfin
Restore compatibility with semigroups versions < 2.7
cc799f0 
This was broken by PR gap-system#317
@ChrisJefferson
Disable timeout tests in HPCGAP
179a352
@ChrisJefferson
Do not edit original home path
f8a4b6d
@ChrisJefferson
Remove const which does not match header
ce95841
@ChrisJefferson
Do not edit original home path
7773e49
@ChrisJefferson
imerge 'master': automatic merge 9-59
a156a72
@ChrisJefferson
Make the constness of SyHaveAlarms consistent
904b71d
@rbehrends
Copy link
Contributor

With regard to CallWithTimeout(), I discussed the details here. I'm currently a bit busy with trying to get generational GC to work with HPC-GAP, so I don't have the time to dig into the CallWithTimeout() issues myself at the moment.

@olexandr-konovalov
Copy link
Member

@ChrisJefferson thanks - seems to me that CallWithTimeout is the only thing that fails Travis tests. I am inclined to merge this anyway while it's fresh, and maybe fix or exclude the test temporarily from HPC-GAP until timeouts will be enabled.

@ChrisJefferson
Copy link
Contributor Author

The timeout stuff now hopefully fixed (by being siabled, so the functions always fail)

@rbehrends
Copy link
Contributor

To clarify something else: HPC-GAP does use configure, but the one in hpc/configure, which doesn't generate a Makefile. This is not a happy solution (because hpc/configure is manually generated by simply removing the ln -sf ... Makefile line from the original configure), but HPC-GAP still depends on and uses the configure script. The generation of hpc/configure could in principle be automated via:

egrep -v '^ln -sf .* Makefile$' configure > hpc/configure

or something similar.

@ChrisJefferson
Copy link
Contributor Author

Actually, please just wait a while before merging -- I used some new software and the history looks horrible.

@fingolfin
Copy link
Member

I'm currently a bit busy with trying to get generational GC to work with HPC-GAP [...]

That's pretty cool, actually :). Will this be usable together with malloc allocation?

@rbehrends
Copy link
Contributor

That's pretty cool, actually :). Will this be usable together with malloc allocation?

It is still using the Boehm GC (which does have a generational API), so, yes. I have no idea yet what this will give us in terms of performance, though. I'm nowhere near the point where it is ready for testing.

@olexandr-konovalov
Copy link
Member

@ChrisJefferson is this ready to merge now? Seems like you've managed to squeeze all similarly looking "imerge" commits.

@ChrisJefferson ChrisJefferson mentioned this pull request Nov 27, 2015
Closed
@ChrisJefferson
Copy link
Contributor Author

Yep, this is good (I hope!). I have made issue #377 to represent the unimplemented timeout functionality.

olexandr-konovalov pushed a commit that referenced this pull request Nov 27, 2015
@alex-konovalov
Merge pull request #375 from ChrisJefferson/hpcgap-merge-nov25
1b9c2a6 
Merge of the master branch (as on Nov 25th) into hpcgap-default.
@olexandr-konovalov olexandr-konovalov merged commit 1b9c2a6 into gap-system:hpcgap-default Nov 27, 2015
@ChrisJefferson ChrisJefferson deleted the hpcgap-merge-nov25 branch November 27, 2015 17:31
@olexandr-konovalov olexandr-konovalov added the topic: HPC-GAP Issues and PRs related to HPC-GAP label Nov 30, 2015
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
topic: HPC-GAP Issues and PRs related to HPC-GAP
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
10 participants
@ChrisJefferson @rbehrends @olexandr-konovalov @fingolfin @markuspf @hulpke @alex-konovalov @stevelinton @frankluebeck @dimpase