summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
* | | | | | | | | | | | | | Merge pull request #16604 from Iamsoto/bugfix_16492_segfault_on_pyfragmentsSebastian Berg2023-02-221-0/+2
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | BUG: SWIG overloaded long functions on pyfragments.swg will seg fault
| * \ \ \ \ \ \ \ \ \ \ \ \ \ Merge branch 'main' into bugfix_16492_segfault_on_pyfragmentsCharles Harris2022-06-091635-69760/+224782
| |\ \ \ \ \ \ \ \ \ \ \ \ \ \
| * | | | | | | | | | | | | | | BUG: default/overloadedfunctions with long will seg faultiamsoto2020-06-141-0/+2
| | | | | | | | | | | | | | | |
* | | | | | | | | | | | | | | | TST: Test (weird) empty value put/putmaskSebastian Berg2023-02-221-0/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Mainly to cover early return path which previously evaded initialization and thus crashed during cleanup
* | | | | | | | | | | | | | | | BUG: Fix initialization and cleanup of cast_info in put/putmaskSebastian Berg2023-02-221-9/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This slipped a bit, I am surprised the compilers didn't warn on CI that the goto requires the variables... Closes gh-23258
* | | | | | | | | | | | | | | | Merge pull request #23257 from rgommers/skip-gha-cleanupSebastian Berg2023-02-218-10/+9
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | CI: clean up GHA boilerplate for `[skip github]`
| * | | | | | | | | | | | | | | | CI: clean up GHA boilerplate for `[skip github]`Ralf Gommers2023-02-218-10/+9
|/ / / / / / / / / / / / / / / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | GitHub has had builtin support for this for quite a while now, via `skip actions`. So we no longer need this. It's left only in the one Meson job; that line is already taken care of in gh-23165. [skip cirrus] [skip circle] [skip azp] [skip travis]
* | | | | | | | | | | | | | | | Merge pull request #23256 from FrancescElies/cesc-set-generic-parameters-to-anySebastian Berg2023-02-216-43/+43
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | TYP,MAINT: Add a missing explicit Any parameter
| * | | | | | | | | | | | | | | | TYP,MAINT: Add a missing explicit Any parameterFrancesc Elies2023-02-216-43/+43
| | |_|_|_|_|/ / / / / / / / / / | |/| | | | | | | | | | | | | |
* | | | | | | | | | | | | | | | Merge pull request #23251 from seberg/further-removalsMatti Picus2023-02-213-38/+172
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | |/ / / / / / / / / / / / / / / |/| | | | | | | | | | | | | | | MAINT: Further removal of PyArray_Item_INCREF use
| * | | | | | | | | | | | | | | TST: Add at least some coverage for put and putmaskSebastian Berg2023-02-201-0/+58
| | | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | | BUG: Pass threadstte to check_and_adjust_indexSebastian Berg2023-02-201-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | (This is not actually currently reachable, but in theory wrong; the reason is that all "reference" dtypes we have also need the API anyway)
| * | | | | | | | | | | | | | | BENCH: Add benchmark for put and types for putmaskSebastian Berg2023-02-201-1/+1
| | | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | | MAINT: Avoid PyArray_Item_INCREF in putmaskSebastian Berg2023-02-201-12/+29
| | | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | | MAINT: Further removal of PyArray_Item_INCREF useSebastian Berg2023-02-202-25/+84
|/ / / / / / / / / / / / / / /
* | | | | | | | | | | | | | | Merge pull request #23248 from seberg/fast-take-no-refcntMatti Picus2023-02-202-27/+56
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | |_|_|/ / / / / / / / / / / / |/| | | | | | | | | | | | | | ENH: Avoid use of item XINCREF and DECREF in fasttake
| * | | | | | | | | | | | | | MAINT: Move/simplify fast-take dest incrememntSebastian Berg2023-02-201-6/+3
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | ENH: Avoid use of item XINCREF and DECREF in fasttakeSebastian Berg2023-02-202-24/+56
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Rather, use the cast function directly when the copy is not trivial (which we know based on it not passing REFCHK, if that passes we assume memcpy is fine). Also uses memcpy, since no overlap is possible here.
* | | | | | | | | | | | | | | Merge pull request #23154 from seberg/astype-refCharles Harris2023-02-2012-134/+243
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | |/ / / / / / / / / / / / / / |/| | | | | | | | | | | | | | ENH: Allow using dtype classes in `arr.astype()`
| * | | | | | | | | | | | | | DOC: Fix and expand docs based on Nathan's reviewSebastian Berg2023-02-032-1/+22
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: And fix final cast warning for missing DTypeMeta cast.Sebastian Berg2023-02-031-1/+1
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: Remove now (currently) unnecessary castSebastian Berg2023-02-031-1/+1
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: Add missing includes to descriptor.h (for the extractor API)Sebastian Berg2023-02-032-0/+2
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | TST: Test new paths for `arr.astype()` to check it accepts DType classesSebastian Berg2023-02-032-0/+12
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | ENH: Use new converter API for `arr.astype()` allowing DType classesSebastian Berg2023-02-031-4/+11
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: Make AdaptDescrToArray work with legacy descr OR dtype+descrSebastian Berg2023-02-036-24/+44
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: Use dtype+descr converter for private array params discoverySebastian Berg2023-02-033-105/+15
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | ENH: Create new DType class aware converters for `dtype=`Sebastian Berg2023-02-032-0/+137
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This commit also moves the check for legacy dtypes (but does not yet delete it).
* | | | | | | | | | | | | | | Merge pull request #22924 from seberg/refactor-decref-meth-onlyMatti Picus2023-02-2019-431/+938
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | |_|_|_|_|/ / / / / / / / / / |/| | | | | | | | | | | | | | MAINT: Refactor clearing up of array data
| * | | | | | | | | | | | | | MAINT: "comment out" unreachable subarray clear data cloneSebastian Berg2023-02-201-2/+10
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | TST: Add some more test coverage (but probably not the big clone funcs)Sebastian Berg2023-02-201-2/+4
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | DOC: Modify comments about untested functionsSebastian Berg2023-02-201-9/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We only ever use the "clone" mechanism for `nditer.copy()` but even there, it is only used for buffer transfers (not final cleanups). It appears that structured dtypes are always a single cast step which doesn't require the clearing normally (so this is hard to hit).
| * | | | | | | | | | | | | | TST: Add test and comment about path that seems uncoverableSebastian Berg2023-02-192-1/+36
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | Apply suggestions from code reviewSebastian Berg2023-02-193-8/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Co-authored-by: Matti Picus <matti.picus@gmail.com>
| * | | | | | | | | | | | | | DOC: Expand comment for `get_clear_loop` and sprinkle dealloc+nullingSebastian Berg2023-02-193-6/+16
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | TST: Do not use `dtypemeta.h` in tests, they should use "external" APISebastian Berg2023-02-191-3/+2
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | DOC: Add code comment about why flags are not used in refcount.cSebastian Berg2023-02-191-4/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We simply don't bother releasing the GIL right now. This could make sense in principle for some user dtypes, though. (NumPy only ever uses clearing with Python objects so the GIL is always needed). I see no plausible use of checking floating point errors, so do not bother to add such cruft, if anyone ever needs it, they should add it.
| * | | | | | | | | | | | | | MAINT: Address review commentsSebastian Berg2023-02-195-27/+37
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Co-authored-by: Nathan Goldbaum <nathan.goldbaum@gmail.com>
| * | | | | | | | | | | | | | BUG: Readd accidentally remove function return typeSebastian Berg2023-02-192-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Also remove accidentally auto-inserted include...
| * | | | | | | | | | | | | | MAINT: Make buffer clearing explicitly stored in the iteratorSebastian Berg2023-02-196-35/+57
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This removes re-using `write` for the same purpose, which seems surprisingly clean, considering that the "clearing" path was already special in any case.
| * | | | | | | | | | | | | | Fixups, but npyiter is blocking me as it actually uses the "clear" logic onceSebastian Berg2023-02-1911-77/+160
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | WIP: Further fixups and full implementation for structured dtypesSebastian Berg2023-02-194-177/+299
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | WIP: Move traversal and simplify signatureSebastian Berg2023-02-197-259/+357
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: Reorganize array dealloc and nditer buffer cleanupSebastian Berg2023-02-196-35/+72
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: Refactor clear function to live on the DTypeSebastian Berg2023-02-196-136/+220
|/ / / / / / / / / / / / / /
* | | | | | | | | | | | | | Merge pull request #23186 from seberg/random-ints-boundscheckingCharles Harris2023-02-192-31/+59
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | MAINT: Fixup random bounds checking code
| * | | | | | | | | | | | | | TST: Add a random integers bounds checking test that previously failedSebastian Berg2023-02-131-0/+2
| | | | | | | | | | | | | | |
| * | | | | | | | | | | | | | MAINT: Fixup random bounds checking codeSebastian Berg2023-02-131-31/+57
| | | | | | | | | | | | | | |
* | | | | | | | | | | | | | | Merge pull request #22051 from r-devulap/spr-supportMatti Picus2023-02-197-50/+99
|\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | BLD: Add compile and runtime checks for AVX512_SPR
| * | | | | | | | | | | | | | | Update cpu_features.inc and compilers-diff.incRaghuveer Devulapalli2023-02-082-46/+49
| | | | | | | | | | | | | | | |
View Lorry Controller Status