summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorMichael W. Hudson <mwh@python.net>2005-06-20 16:37:03 (GMT)
committerMichael W. Hudson <mwh@python.net>2005-06-20 16:37:03 (GMT)
commitfb662972e001aa051e5085862cf5eac323e1756f (patch)
treeb6ba8e4136a4fdddcdd20ab78d8cc1c822dfa8f4
parent7390942aa16be9b894c31563ec9f2b502ed28338 (diff)
downloadcpython-fb662972e001aa051e5085862cf5eac323e1756f.zip
cpython-fb662972e001aa051e5085862cf5eac323e1756f.tar.gz
cpython-fb662972e001aa051e5085862cf5eac323e1756f.tar.bz2
Add a(nother) warning about mixing Py_NewInterpreter and the
PyGILState_* functions.
-rw-r--r--Doc/api/init.tex6
1 files changed, 6 insertions, 0 deletions
diff --git a/Doc/api/init.tex b/Doc/api/init.tex
index 940aab9..51b925f 100644
--- a/Doc/api/init.tex
+++ b/Doc/api/init.tex
@@ -131,6 +131,12 @@
objects may affect the wrong (sub-)interpreter's dictionary of
loaded modules. (XXX This is a hard-to-fix bug that will be
addressed in a future release.)
+
+ Also note that the use of this functionality is incompatible with
+ extension modules such as PyObjC and ctypes that use the
+ \cfunction{PyGILState_*} APIs (and this is inherent in the way the
+ \cfunction{PyGILState_*} functions work). Simple things may work,
+ but confusing behavior will always be near.
\end{cfuncdesc}
\begin{cfuncdesc}{void}{Py_EndInterpreter}{PyThreadState *tstate}