On Wed, Jul 7, 2010 at 8:22 AM, Chuck Blake <c...@pdos.csail.mit.edu> wrote: > The code generated by the following compiles without error but then > tries to INCREF a null pointer for x (all for closures branch tip): > > cdef f(x): # def here => works fine > def g(y): return y*x # cdef here => compile error > return g(x) # faults@ INCREF(.*cur_scope->.*v_x > print f(2) > > If the intent is not to support the above scenario then this is a > compiler-should-fail-but-doesn't bug, else a code generation bug. > > It may be a known issue, but I couldn't find a test case or a Trac > issue. So, I thought that I should mention it just in case.
Yep, looks like a bug to me: http://trac.cython.org/cython_trac/ticket/554 - Robert _______________________________________________ Cython-dev mailing list Cython-dev@codespeak.net http://codespeak.net/mailman/listinfo/cython-dev
