Relies on arbitrary precision math, so won't run on a port which
has threads & limited bigint support.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
If GIL is disabled then there's threat of a race condition if some other
code specifically requests USB processing (i.e. to unblock stdio), while
a scheduled TinyUSB callback is already running on another thread.
Relies on the change in the parent commit, where scheduler is restricted
to main thread if GIL is disabled.
Fixes#15390 - "TinyUSB callback can't recurse" exceptions on rp2 when
using _thread module and USB serial I/O.
Adds a unit test for stdin functioning correctly in threads (fails on rp2
port without this fix).
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
Explicitly yield each time a thread mutex is unlocked.
Key to understanding this bug is that Python threads run at equal RTOS
priority, and although ESP-IDF FreeRTOS (and I think vanilla FreeRTOS)
scheduler will round-robin equal priority tasks in the ready state it does
not make a similar guarantee for tasks moving between ready and waiting.
The pathological case of this bug is when one Python thread task is busy
(i.e. never blocks) it will hog the CPU more than expected, sometimes for
an unbounded amount of time. This happens even though it periodically
unlocks the GIL to allow another task to run.
Assume T1 is busy and T2 is blocked waiting for the GIL. T1 is executing
and hits a condition to yield execution:
1. T1 calls MP_THREAD_GIL_EXIT
2. FreeRTOS sees T2 is waiting for the GIL and moves it to the Ready list
(but does not preempt, as T2 is same priority, so T1 keeps running).
3. T1 immediately calls MP_THREAD_GIL_ENTER and re-takes the GIL.
4. Pre-emptive context switch happens, T2 wakes up, sees GIL is not
available, and goes on the waiting list for the GIL again.
To break this cycle step 4 must happen before step 3, but this may be a
very narrow window of time so it may not happen regularly - and
quantisation of the timing of the tick interrupt to trigger a context
switch may mean it never happens.
Yielding at the end of step 2 maximises the chance for another task to run.
Adds a test that fails on esp32 before this fix and passes afterwards.
Fixes issue #15423.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
Because the main thread executes `thread_entry()` it means there's an
additional one added to `count`, so the test must wait for the count to
reach `n_thread + 1`.
Signed-off-by: Damien George <damien@micropython.org>
mp_thread_begin_atomic_section() is expected to be recursive (i.e. for
nested machine.disable_irq() calls, or if Python code calls disable_irq()
and then the Python runtime calls mp_handle_pending() which also enters an
atomic section to check the scheduler state).
On rp2 when not using core1 the atomic sections are recursive.
However when core1 was active (i.e. _thread) then there was a bug that
caused the core to live-lock if an atomic section recursed.
Adds a test case specifically for mutual exclusion and recursive atomic
sections when using two threads. Without this fix the test immediately
hangs on rp2.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
Otherwise GC stays disabled (not re-enabled by soft reset) and later test
runs fail with MemoryError.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
The existing thread_sleep1.py test only tests execution, not accuracy, of
time.sleep. Also the existing test only tests sleep(0) on targets like rp2
that can only create a single thread.
The new test in this commit checks for timing accuracy on the main thread
and one other thread when they run at the same time.
Signed-off-by: Damien George <damien@micropython.org>
The aim of this commit is to make it so that the existing thread tests can
be used to test the _thread module on the rp2 port. The rp2 port only
allows up to one thread to be created at a time, and does not have the GIL
enabled.
The following changes have been made:
- run-tests.py skips mutation tests on rp2, because there's no GIL.
- run-tests.py skips other tests on rp2 that require more than one thread.
- The tests stop trying to start a new thread after there is an OSError,
which indicates that the system cannot create more threads.
- Some of these tests also now run the test function on the main thread,
not just the spawned threads.
- In some tests the output printing is adjusted so it's the same regardless
of how many threads were spawned.
- Some time.sleep(1) are replaced with time.sleep(0) to make the tests run
a little faster (finish sooner when the work is done).
For the most part the tests are unchanged for existing platforms like esp32
and unix.
Signed-off-by: Damien George <damien@micropython.org>
In CPython, `_thread.start_new_thread()` returns an ID that is the same ID
that is returned by `_thread.get_ident()`. The current MicroPython
implementation of `_thread.start_new_thread()` always returns `None`.
This modifies the required functions to return a value. The native thread
id is returned since this can be used for interop with other functions, for
example, `pthread_kill()` on *nix. `_thread.get_ident()` is also modified
to return the native thread id so that the values match and avoids the need
for a separate `native_id` attribute.
Fixes issue #12153.
Signed-off-by: David Lechner <david@pybricks.com>
The RP2040 has 2 cores and supports running at most 2 Python threads (the
main one plus another), and will raise OSError if a thread cannot be
created because core1 is already in use. This commit adjusts some thread
tests to be robust against such OSError's. These tests now pass on rp2
boards.
Signed-off-by: Damien George <damien@micropython.org>
This commit makes gc_lock_depth have one counter per thread, instead of one
global counter. This makes threads properly independent with respect to
the GC, in particular threads can now independently lock the GC for
themselves without locking it for other threads. It also means a given
thread can run a hard IRQ without temporarily locking the GC for all other
threads and potentially making them have MemoryError exceptions at random
locations (this really only occurs on MCUs with multiple cores and no GIL,
eg on the rp2 port).
The commit also removes protection of the GC lock/unlock functions, which
is no longer needed when the counter is per thread (and this also fixes the
cas where a hard IRQ calling gc_lock() may stall waiting for the mutex).
It also puts the check for `gc_lock_depth > 0` outside the GC mutex in
gc_alloc, gc_realloc and gc_free, to potentially prevent a hard IRQ from
waiting on a mutex if it does attempt to allocate heap memory (and putting
the check outside the GC mutex is now safe now that there is a
gc_lock_depth per thread).
Signed-off-by: Damien George <damien@micropython.org>
So a lock can be acquired on one Python thread and then released on
another. A test for this is added.
Signed-off-by: Damien George <damien@micropython.org>
When threading is enabled without the GIL then there can be races between
the threads accessing the globals dict. Avoid this issue by making sure
all globals variables are allocated before starting the threads.
Signed-off-by: Damien George <damien@micropython.org>
This is consistent with the other 'micro' modules and allows implementing
additional features in Python via e.g. micropython-lib's sys.
Note this is a breaking change (not backwards compatible) for ports which
do not enable weak links, as "import sys" must now be replaced with
"import usys".
On arm64 with CPython:
>>> _thread.stack_size(32*1024)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
ValueError: size not valid: 32768 bytes
So increase the CPython value in the test to 512k so it runs on more
systems (on modern Linux the default stack size is usually 8MB).
This adds the Python files in the tests/ directory to be formatted with
./tools/codeformat.py. The basics/ subdirectory is excluded for now so we
aren't changing too much at once.
In a few places `# fmt: off`/`# fmt: on` was used where the code had
special formatting for readability or where the test was actually testing
the specific formatting.
Depending on the thread scheduler, a busy-wait loop can hog the CPU and
make the tests very slow. So convert such loops to loops that have an
explicit sleep, allowing the worker threads to do their job.
