Message235151
>> When I compile and run the above (latest Cython from Git master), I
>> get:
>>
>> ()
>> ()
>
> With Cython version 0.20.1post0 I get:
>
> >>> foo.foo()
> (-1,)
> (-1,)
>
> If you get the correct output from the latest Cython, it looks like
> this issue has been fixed.
Oops, I was running a locally modified version of Cython that contained a patch meant to work around issue #23349.
When I run the *actual* upstream master I get the same behavior that you do.
But either way, I don't see why it's a problem that it prints (-1,) for the PyBUF_ND case. The consumer didn't request suboffsets information so I would expect it to be OK for the producer to put whatever it wants in the suboffsets field, including a junk pointer that would segfault upon dereference. |
|
Date |
User |
Action |
Args |
2015-02-01 09:21:18 | rhansen | set | recipients:
+ rhansen, pitrou, skrah, seberg |
2015-02-01 09:21:18 | rhansen | set | messageid: <1422782478.58.0.298625475587.issue23352@psf.upfronthosting.co.za> |
2015-02-01 09:21:18 | rhansen | link | issue23352 messages |
2015-02-01 09:21:18 | rhansen | create | |
|