search for: pyeval_threadsinitialized

...const uint64_t *array, size_t array_len) { - PyGILState_STATE py_save = PyGILState_UNLOCKED; + PyGILState_STATE py_save; PyObject *py_callback = callback; PyObject *py_array; PyObject *args; PyObject *a; size_t i; PyObject *py_r; + int threads_initialized = PyEval_ThreadsInitialized (); + + if (threads_initialized) + py_save = PyGILState_Ensure (); py_array = PyList_New (array_len); for (i = 0; i < array_len; ++i) { @@ -132,14 +136,8 @@ guestfs_int_py_event_callback_wrapper (guestfs_h *g, buf, buf_len, py_array); Py_INCREF (args); -...

...;, py_%s" n + | CBInt n | CBInt64 n + | CBString n + | CBUInt n | CBUInt64 n -> pr ", %s" n + | CBArrayAndLen _ | CBMutable _ -> assert false + ) cbargs; + pr ");\n"; + pr " Py_INCREF (py_args);\n"; + pr "\n"; + pr " if (PyEval_ThreadsInitialized ())\n"; + pr " py_save = PyGILState_Ensure ();\n"; + pr "\n"; + pr " py_ret = PyObject_CallObject ((PyObject *)user_data, py_args);\n"; + pr "\n"; + pr " if (PyEval_ThreadsInitialized ())\n"; + pr " PyGILState_Release...

A couple of fixes for Python 3 to the Python binding. Unfortunately a behaviour change is needed, although it fixes broken types used. Changes from v1: - handle also FBuffer in structs Pino Toscano (2): python: fix call of Python handlers of events python: change types for RBufferOut/FBuffer with Python 3 (RHBZ#1661871) generator/python.ml | 9 +++++++++ python/handle.c | 3 ++- 2

A couple of fixes for Python 3 to the Python binding. Unfortunately a behaviour change is needed, although it fixes broken types used. Pino Toscano (2): python: fix call of Python handlers of events python: change return type for RBufferOut with Python 3 (RHBZ#1661871) generator/python.ml | 4 ++++ python/handle.c | 3 ++- 2 files changed, 6 insertions(+), 1 deletion(-) -- 2.20.1

Patch 2 certainly allows you to set the thread model. However patch 3 shows that if you set it to nbdkit.THREAD_MODEL_PARALLEL it will crash. If you look closely at the stack trace (attached below) you can see that ignoring threads which are in parts of nbdkit unrelated to Python: Thread 4: In pread, waiting in time.sleep(). This thread has released the GIL. Thread 2: Started to

Hi, some of the bindings may try to malloc with 0 bytes as size when closing an handle, because there were no event handlers registered. Since this can have different behaviours in POSIX, avoid that situation altogether by just skipping allocating anything when there were no event handlers. Thanks, Pino Toscano (4): ocaml: do not try to malloc 0 elements in get_all_event_callbacks python:

...Mutable _ + | OpaqueAndCallbacks _ | OpaqueAndCallbacksPersist _ + | Path _ | SockAddrAndLen _ | StringList _ + | UInt _ | UInt32 _ -> assert false + ) cb.cbargs; + pr ");\n"; + pr " Py_INCREF (py_args);\n"; + pr "\n"; + pr " if (PyEval_ThreadsInitialized ())\n"; + pr " py_save = PyGILState_Ensure ();\n"; + pr "\n"; + pr " py_ret = PyObject_CallObject (data->%s, py_args);\n" cb.name; + pr "\n"; + pr " if (PyEval_ThreadsInitialized ())\n"; + pr " PyGILState_Relea...

...{ args; optargs; ret; may_set_error } = | CBArrayAndLen _ | CBMutable _ -> assert false ) cbargs; pr ");\n"; - pr " Py_INCREF (py_args);\n"; + pr " Py_INCREF (py_args);\n"; pr "\n"; - pr " if (PyEval_ThreadsInitialized ())\n"; - pr " py_save = PyGILState_Ensure ();\n"; + pr " if (PyEval_ThreadsInitialized ())\n"; + pr " py_save = PyGILState_Ensure ();\n"; pr "\n"; - pr " py_ret = PyObject_CallObject ((PyObject *)user_dat...

Posting now that I got something to compile (at the expense of breaking OCaml bindings), but I'm open to ideas on how to improve it. Eric Blake (2): generator: Tweak print_c_arg_list to take alternate first arg RFC: generator: Handle shared callbacks in Python generator/generator | 556 ++++++++++++++++++++++---------------------- 1 file changed, 280 insertions(+), 276 deletions(-) --

...-4084,53 +4076,55 @@ let print_python_closure_wrapper { cbname; cbargs } = | CBArrayAndLen _ | CBMutable _ -> assert false ) cbargs; pr ");\n"; - pr " Py_INCREF (py_args);\n"; + pr " Py_INCREF (py_args);\n"; pr "\n"; - pr " if (PyEval_ThreadsInitialized ())\n"; - pr " py_save = PyGILState_Ensure ();\n"; + pr " if (PyEval_ThreadsInitialized ())\n"; + pr " py_save = PyGILState_Ensure ();\n"; pr "\n"; - pr " py_ret = PyObject_CallObject ((PyObject *)user_data, py_args);\n"; +...

Basically the same as this patch series, but for Python: https://www.redhat.com/archives/libguestfs/2019-August/msg00235.html plus adding the 590 asynch test at the end. Rich.

...@@ let print_python_binding name { args; ret } = | UInt _ | UInt32 _ -> assert false ) cbargs; pr ");\n"; - pr " Py_INCREF (py_args);\n"; + pr " Py_INCREF (py_args);\n"; pr "\n"; - pr " if (PyEval_ThreadsInitialized ())\n"; - pr " py_save = PyGILState_Ensure ();\n"; + pr " if (PyEval_ThreadsInitialized ())\n"; + pr " py_save = PyGILState_Ensure ();\n"; pr "\n"; - pr " py_ret = PyObject_CallObject ((PyObject *)user_dat...

...| Flags _ | Mutable _ - | Path _ | SockAddrAndLen _ | StringList _ - | UInt _ | UInt32 _ -> assert false - ) cbargs; - pr ");\n"; - pr " Py_INCREF (py_args);\n"; - pr "\n"; - pr " if (PyEval_ThreadsInitialized ())\n"; - pr " py_save = PyGILState_Ensure ();\n"; - pr "\n"; - pr " py_ret = PyObject_CallObject (((struct %s_user_data *)user_data)->%s, py_args);\n" name cbname; - pr "\n"; - pr " if (PyEv...

v1 was here: https://www.redhat.com/archives/libguestfs/2019-July/thread.html#00231 The changes address everything that Eric picked up in his review of the first two patches. I have also added two more patches (4 and 5) which respectively fix docs and change int status -> unsigned status, as discussed. Passes make, check, check-valgrind. Rich.

Patches 1-4 are basically uncontroversial, straightforward refactoring and IMHO we should just push them. Possibly 1-3 should be squashed together, but I posted them separately so they are easier to review. Patches 5 and 6 together implement OClosure. Patch 5 adds the feature and is simple to understand. Patch 6 changes the Closure completion callbacks into OClosure, but because it doesn't

..._ | CallbackPersist _ - | Flags _ | Mutable _ - | Path _ | SockAddrAndLen _ | StringList _ - | UInt _ | UInt32 _ -> assert false - ) args; - pr ");\n"; - pr " Py_INCREF (py_args);\n"; - pr "\n"; - pr " if (PyEval_ThreadsInitialized ())\n"; - pr " py_save = PyGILState_Ensure ();\n"; - pr "\n"; - pr " py_ret = PyObject_CallObject (_data->fn, py_args);\n"; - pr "\n"; - pr " if (PyEval_ThreadsInitialized ())\n"; - pr " PyGIL...

As before, but this one has working Python bindings. OCaml still TBD. Rich.

As proposed here: https://www.redhat.com/archives/libguestfs/2019-August/msg00130.html I didn't actually read Eric's replies to that yet because I've been concentrating on writing these patches all day. Anyway here they are and I'll look at what Eric said about the proposal next. Rich.

This implements most of what I wrote here: https://www.redhat.com/archives/libguestfs/2019-July/msg00213.html

The safe_malloc (etc) functions call g->abort_fn on failure. That's not appropriate for language bindings, and we never intended that these internal functions be used from language bindings, that was just a historical accident. This patch series removes any external use of the safe_* functions. Rich.