Some work on the NumPy test #783
Conversation
rudi-c
force-pushed
the
numpy_fix
branch
6 times, most recently
from
5c536e9 to
5bf904b
Compare
|
@kmod This got rebased on the ctypes changes and should be ready for review now. |
There was a problem hiding this comment.
hmm this is a bit odd to me -- why do we have to do this when cpython doesn't?
There was a problem hiding this comment.
It is in PyType_Ready, typeobject.cpp:4123
There was a problem hiding this comment.
ah ok, right (not sure why I didn't see that).
|
Sorry, we really should get around to adding an exception-style linter :/ btw, how hard would it be to get it to the point that we can actually enable the numpy test? |
|
There's almost nothing to do besides adding a line to the script that applies a patch to NumPy. |
|
That being said the NumPy test currently only tests if NumPy crashes, not if it produces the same output. |
rudi-c
force-pushed
the
numpy_fix
branch
2 times, most recently
from
f52dd00 to
f3636f2
Compare
src/codegen/parser.cpp
Outdated
There was a problem hiding this comment.
Is there actually anything to do here? I think we should avoid putting this in if it will silently handle ellipses the wrong way, but it looks like this might actually be all we need and we can remove the TODO? Same for the serializer.
|
Ok I merged the testsuite change, could you update this PR to include the new submodule commit? |
rudi-c
force-pushed
the
numpy_fix
branch
2 times, most recently
from
417a31d to
a52252e
Compare
NumPy is huge, bigger than our previous (arbitrary) number by an order of magnitude.
C extensions (NumPy) might inherit classes in C code and expect to find tp_number. This is just copied from CPython's PyType_Ready. This requires assigning some of the runtime functions to thesq_ and mp_ slots otherwise there are infinite loops from Pyston attributes.
Those should never exist because all Python objects should be created through the CPython API except for type objects. Unfortunately, some places like NumPy do that so we need a mean of patching it for now.
|
Cool this is passing again. |
|
awesome! |
2 participants
Depends on #767
These are just a few small changes that help run NumPy which should be useful independently of the rest of the NumPy work.