This commit renames the NORETURN macro, indicating to the compiler
that a function does not return, into MP_NORETURN to maintain the same
naming convention of other similar macros.
To maintain compaitiblity with existing code NORETURN is aliased to
MP_NORETURN, but it is also deprecated for MicroPython v2.
This changeset was created using a similar process to
decf8e6a8b ("all: Remove the "STATIC"
macro and just use "static" instead."), with no documentation or python
scripts to change to reflect the new macro name.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
There's no specified behaviour for what should happen if both CPUs call
`lightsleep()` together, but the latest changes could cause a permanent
hang due to a race in the timer cleanup code. Add a flag to prevent hangs
if two threads accidentally lightsleep, at least.
This allows the new lightsleep test to pass on RPI_PICO and RPI_PICO2, and
even have much tighter time deltas. However, the test still fails on
wireless boards where the lwIP tick wakes them up too frequently.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
Stop using soft timer for `mp_wfe_or_timeout`. Now uses the alarm pool
again as issues with this code have been fixed. This resolves the "sev"
issue that stops the RP2350 going idle.
Also, change the lightsleep code to use the hardware timer library and
alarm 1, as alarm 2 is used by and soft timers and alarm 3 is used by the
alarm pool.
Signed-off-by: Peter Harper <peter.harper@raspberrypi.com>
Configure flash timings dynamically to match the system clock. Reconfigure
timings after flash writes.
Changes are:
- ports/rp2/main.c: Set default flash timings.
- ports/rp2/modmachine.c: Configure optimal flash timings on freq change.
- ports/rp2/rp2_flash.c: Reconfigure flash when leaving critical section.
Signed-off-by: Phil Howard <github@gadgetoid.com>
Performs a best-effort attempt to detect attached PSRAM, configure it and
*add* it to the MicroPython heap. If PSRAM is not present, should fall
back to use internal RAM.
Introduce two new port/board defines:
- MICROPY_HW_ENABLE_PSRAM to enable PSRAM.
- MICROPY_HW_PSRAM_CS_PIN to define the chip-select pin (required).
Changes are:
- ports/rp2/rp2_psram.[ch]: Add new PSRAM module.
- ports/rp2/main.c: Add optional PSRAM support.
- ports/rp2/CMakeLists.txt: Include rp2_psram.c.
- ports/rp2/mpconfigport.h: Add MICROPY_HW_ENABLE_PSRAM.
- ports/rp2/modmachine.c: Reconfigure PSRAM on freq change.
Co-authored-by: Kirk Benell <kirk.benell@sparkfun.com>
Co-authored-by: Mike Bell <mike@mercuna.com>
Signed-off-by: Phil Howard <phil@gadgetoid.com>
A better indication of whether a cyw43 event is pending is the actual flag
in the PendSV handler table. (If this fails, could also use the GPIO
interrupt enabled register bit).
This commit was needed of a previous version of the fix in the parent
commit, but it turned out not strictly necessary for the current version.
However, it's still a good clean up.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
The problem was introduced in d1423ef7a2,
calling `machine.lightsleep()` overwrites RP2xxx registers `SLEEP_EN0` and
`SLEEP_EN1` with their power on default values.
Prior to that commit the register values were saved on entry to lightsleep
and restored before returning.
These changes restores the earlier behavior.
Fixes issue #16502.
Signed-off-by: Carl Pottle <cpottle9@outlook.com>
Prior to this fix, `tests/extmod/vfs_lfs_mtime.py` would fail when run
after the `tests/ports/rp2/rp2_lightsleep.py` test, because
`time.time_ns()` would have a large and constant offset from `time.time()`.
Fix this by re-syncing the time-ns offset when coming out of lightsleep.
Signed-off-by: Damien George <damien@micropython.org>
This commit separates various build settings and include files that are
specific to RP2040 and RP2350, and uses the aon_timer interface instead of
rtc, to work across both MCU variants.
Signed-off-by: Damien George <damien@micropython.org>
Signed-off-by: Phil Howard <phil@gadgetoid.com>
This isn't fully working, the CPU often wakes up early. That will be fixed
when a newer version of pico-sdk is released.
Signed-off-by: Damien George <damien@micropython.org>
By default, the peripheral clock for UART and SPI is set to 48 MHz and will
not be affected by the MCU clock change. This can be changed by a second
argument to `machine.freq(freq, peripheral_freq)`. The second argument
must be either 48 MHz or identical with the first argument.
Note that UART and SPI baud rates may have to be re-configured after
changing the MCU clock.
Signed-off-by: robert-hh <robert@hammelrath.com>
Updates rp2 port to always resume from idle within 1ms max.
When rp2 port went tickless the behaviour of machine.idle() changed as
there is no longer a tick interrupt to wake it up every millisecond. On a
quiet system it would now block indefinitely. No other port does this.
See parent commit for justification of why this change is useful.
Also adds a test case that fails without this change.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
Follow-up to a84c7a0ed9, this commit works most of the time but has an
intermittent bug where USB doesn't resume as expected after waking from
light sleep.
Turns out waking calls clocks_init() which will re-initialise the USB PLL.
Most of the time this is OK but occasionally it seems like the clock
glitches the USB peripheral and it stops working until the next hard reset.
Adds a machine.lightsleep() test that consistently hangs in the first
two dozen iterations on rp2 without this fix. Passed over 100 times in a
row with this fix.
The test is currently rp2-only as it seems similar lightsleep USB issues
exist on other ports (both pyboard and ESP32-S3 native USB don't send any
data to the host after waking, until they receive something from the host
first.)
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
Without this change going to lightsleep stops the USB peripheral clock, and
can lead to either the device going into a weird state or the host deciding
to issue a bus reset.
This change only keeps the USB peripheral clocks enabled if the USB device
is currently active and a host has configured the device. This means the
USB device continues to respond to host transfers and (presumably) will
even complete pending endpoint transfers. All other requests are NAKed
while still asleep, but the interaction with the host seems to resume
correctly on wake
Otherwise, if USB is not active or configured by a host, USB clocks are
disabled, the same as before.
With the change, one can issue a `machine.lightsleep(...)` with USB CDC
connected and the USB CDC remains connected during the sleep and resumes
when the lightsleep finishes.
Tested on a RPi Pico, the power consumption is:
- During normal idle at the REPL, about 15.3mA.
- During lightsleep, prior to this change, about 1.35mA.
- During lightsleep, with this change and USB CDC connected, about 3.7mA.
If power consumption should be as low as possible when USB is connected,
one can use `machine.USBDevice` to disable the USB before entering
lightsleep.
As discussed at https://github.com/orgs/micropython/discussions/14401
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
To avoid undefined references to `mp_thread_begin_atomic_section()` /
`mp_thread_end_atomic_section()`, replace them with the
`MICROPY_BEGIN_ATOMIC_SECTION` / `MICROPY_END_ATOMIC_SECTION`
macros. That way, it's possible to build again with `MICROPY_PY_THREAD`
disabled (made possible by efa54c27b9).
Fixes commit 19844b4983.
Signed-off-by: Matthias Blankertz <matthias@blankertz.org>
When `lightsleep()` is called from within a thread the interrupts may not
be enabled on current core, and thus the call to `lightsleep()` never
completes.
Fixes issue #14092.
Signed-off-by: Simon Wood <simon@mungewell.org>
The STATIC macro was introduced a very long time ago in commit
d5df6cd44a. The original reason for this was
to have the option to define it to nothing so that all static functions
become global functions and therefore visible to certain debug tools, so
one could do function size comparison and other things.
This STATIC feature is rarely (if ever) used. And with the use of LTO and
heavy inline optimisation, analysing the size of individual functions when
they are not static is not a good representation of the size of code when
fully optimised.
So the macro does not have much use and it's simpler to just remove it.
Then you know exactly what it's doing. For example, newcomers don't have
to learn what the STATIC macro is and why it exists. Reading the code is
also less "loud" with a lowercase static.
One other minor point in favour of removing it, is that it stops bugs with
`STATIC inline`, which should always be `static inline`.
Methodology for this commit was:
1) git ls-files | egrep '\.[ch]$' | \
xargs sed -Ei "s/(^| )STATIC($| )/\1static\2/"
2) Do some manual cleanup in the diff by searching for the word STATIC in
comments and changing those back.
3) "git-grep STATIC docs/", manually fixed those cases.
4) "rg -t python STATIC", manually fixed codegen lines that used STATIC.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
The ports esp32, mimxrt, rp2 and samd all shared exactly the same
implementation of machine.disable_irq() and machine.enable_irq(),
implemented in terms of MICROPY_{BEGIN,END}_ATOMIC_SECTION. This commit
factors these implementations into extmod/modmachine.c.
The cc3200, esp8266, nrf, renesas-ra and stm32 ports do not yet use this
common implementation.
Signed-off-by: Damien George <damien@micropython.org>
Minor changes for consistency are:
- nrf gains: unique_id(), freq() [they do nothing]
- samd: deepsleep() now resets after calling lightsleep()
- esp32: lightsleep()/deepsleep() no longer take kw arg "sleep", instead
it's positional to match others. also, passing 0 here will now do a 0ms
sleep instead of acting like nothing was passed.
reset_cause() no longer takes any args (before it would just ignore them)
- mimxrt: freq() with an argument and lightsleep() both raise
NotImplementedError
Signed-off-by: Damien George <damien@micropython.org>
And use it in all ports. The ports are unchanged, except esp8266 which now
just returns None from this function instead of the time elapsed (to match
other ports), and qemu-arm which gains this function.
Signed-off-by: Damien George <damien@micropython.org>
This is a code factoring to have the dict for the machine module in one
location, and all the ports use that same dict. The machine.soft_reset()
function implementation is also factored because it's the same for all
ports that did already implement it. Eventually more functions/bindings
can be factored.
All ports remain functionally the same, except:
- cc3200 port: gains soft_reset, mem8, mem16, mem32, Signal; loses POWER_ON
(which was a legacy constant, replaced long ago by PWRON_RESET)
- nrf port: gains Signal
- qemu-arm port: gains soft_reset
- unix port: gains soft_reset
- zephyr port: gains soft_reset, mem8, mem16, mem32
Signed-off-by: Damien George <damien@micropython.org>
The main motivation for doing this was to reduce the latency when the
system is woken by a USB interrupt. The best_effort_wfe_or_timeout()
function calls into the pico-sdk dynamic timer framework which sets up a
new dynamic timer instance each time, and then has to tear it down before
continuing after a WFE.
Testing Python interrupt latency, it seems to be improved by about 12us
(from average of 46us to 34us running a Pin IRQ). C-based "scheduled
nodes" should see even lower latency.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
The contents of machine_mem.h, machine_i2c.h and machine_spi.h have been
moved into extmod/modmachine.h.
Signed-off-by: Damien George <damien@micropython.org>
The contents of machine_bitstream.h, machine_pinbase.h, machine_pulse.h and
machine_signal.h have been moved into extmod/modmachine.h.
Signed-off-by: Damien George <damien@micropython.org>
This is a code factoring to have the Python bindings in one location, and
all the ports use those same bindings. For all ports except the two listed
below there is no functional change.
The nrf port has UART.sendbreak() removed, but this method previously did
nothing.
The zephyr port has the following methods added:
- UART.init(): supports setting timeout and timeout_char.
- UART.deinit(): does nothing, just returns None.
- UART.flush(): raises OSError(EINVAL) because it's not implemented.
- UART.any() and UART.txdone(): raise NotImplementedError.
Signed-off-by: Damien George <damien@micropython.org>
No functional change, just code factoring to have the Python bindings in
one location, and all the ports use those same bindings.
Signed-off-by: Damien George <damien@micropython.org>
With public declarations moved to extmod/modmachine.h. It's now mandatory
for a port to define MICROPY_PY_MACHINE_PWM_INCLUDEFILE if it enables
MICROPY_PY_MACHINE_PWM. This follows how extmod/machine_wdt.c works.
All ports have been updated to work with this modified scheme.
Signed-off-by: Damien George <damien@micropython.org>
There are currently 7 ports that implement machine.WDT and a lot of code is
duplicated across these implementations. This commit factors the common
parts of all these implementations to a single location in
extmod/machine_wdt.c. This common code provides the top-level Python
bindings (class and method wrappers), and then each port implements the
back end specific to that port.
With this refactor the ports remain functionally the same except for:
- The esp8266 WDT constructor now takes keyword arguments, and accepts the
"timeout" argument but raises an exception if it's not the default value
(this port doesn't support changing the timeout).
- The mimxrt and samd ports now interpret the argument to WDT.timeout_ms()
as signed and if it's negative truncate it to the minimum timeout (rather
than it being unsigned and a negative value truncating to the maximum
timeout).
Signed-off-by: Damien George <damien@micropython.org>
While cyw43 is deinitialized, an interrupt occurs. That is handled with
these lines: ports/rp2/mpnetworkport.c#L59-L61 and as pendsv is disabled
while in network code, the poll function then just waits there.
When deinit has finished, the poll func is executed, but skipped:
src/cyw43_ctrl.c#L222-L225 this skips the `CYW43_POST_POLL_HOOK` which
would re-enable interrupts, but also reset `cyw43_has_pending`.
And in that state, the lightsleep code, will skip sleeping as it thinks
there is a network packet pending to be handled.
With this change applied, lightsleep works as expected when the wifi chip
is enabled, and when it's powered off.
This renames the builtin-modules, such that help('modules') and printing
the module object will show "module" rather than "umodule".
This work was funded through GitHub Sponsors.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
This commit prevents the device from "hanging" when using lightsleep while
the WiFi chip is active.
Whenever the WiFi chip wants to interrupt the microcontroller to notify it
for a new package, it sets the CYW43_PIN_WL_HOST_WAKE pin to high,
triggering an IRQ. However, as polling the chip cannot happen in an
interrupt handler, it subsequently notifies the pendsv-service to do a poll
as soon as the interrupt handler ended. In order to prevent a new
interrupt from happening immediately afterwards, even before the poll has
run, the IRQ handler disables interrupts from the pin.
The first problem occurs, when a WiFi package arrives while the main loop
is in cyw43-code. In order to prevent concurrent access of the hardware,
the network code blocks pendsv from running again while entering lwIP code.
The same holds for direct cyw43 code (like changing the cyw43-gpios, i.e.
the LED on the Pico W). While the pendsv is disabled, interrupts can still
occur to schedule a poll (and disable further interrupts), but it will not
run. This can happen while the microcontroller is anywhere in rp2040 code.
In order to preserve power while waiting for cyw43 responses,
cyw43_configport.h defines CYW43_DO_IOCTL_WAIT and
CYW43_SDPCM_SEND_COMMON_WAIT to __WFI(). While this might work in most
cases, there are 2 edge cases where it fails:
- When an interrupt has already been received by the cyw43 stack, for
example due to an incoming ethernet packet.
- When the interrupt from the cyw43 response comes before the
microcontroller entered the __WFI() instruction.
When that happens, wfi will just block forever as no further interrupts are
received. The only way to safely use wfi to wake up from an interrupt is
inside a critical section, as this delays interrupts until the wfi is
entered, possibly resuming immediately until interrupts are reenabled and
the interrupt handler is run. Additionally this critical section needs to
check whether the interrupt has already been disabled and pendsv was
triggered, as in such a case, wfi can never be woken up, and needs to be
skipped, because there is already a package from the network chip waiting.
Note that this turns cyw43_yield into a nop (and thereby the cyw43-loops
into busy waits) from the second time onwards, as after the first call, a
pendsv request will definitely be pending. More logic could be added, to
explicitly enable the interrupt in this case.
Regarding lightsleep, this code has a similar problem. When an interrupt
occurs during lightsleep, the IRQ and pendsv handler and thereby poll are
run immediately, with the clocks still disabled, causing the SPI transfers
to fail. If we don't want to add complex logic inside the IRQ handler we
need to protect the whole lightsleep procedure form interrupts with a
critical section, exiting out early if an interrupt is pending for whatever
reason. Only then we can start to shut down clocks and only enable
interrupts when the system is ready again. Other interrupt handlers might
also be happy, that they are only run when the system is fully operational.
Tested on a Pico W, calling machine.lightsleep() within an endless loop and
pinging from the outside.
This gets basic machine.lightsleep([n]) behaviour working on the rp2 port.
It supports:
- Calling lightsleep without a specified period, in which case it uses xosc
dormant mode. There's currently no way to wake it up from this state,
unless you write to raw registers to enable a GPIO wake up source.
- Calling lightsleep with a period n in milliseconds. This period must be
less than about 72 minutes and uses timer alarm3 to wake it up.
The RTC continues to run during lightsleep, but other peripherals have
their clock turned off during the sleep.
It doesn't yet support longer periods than 72 minutes, or waking up from
GPIO IRQ.
Measured current consumption from the USB port on a PICO board is about
1.5mA when doing machine.lightsleep(5000), and about 0.9mA when doing
machine.lightsleep().
Addresses issue #8770.
Signed-off-by: Damien George <damien@micropython.org>
It's no longer needed because this macro is now processed after
preprocessing the source code via cpp (in the qstr extraction stage), which
means unused MP_REGISTER_MODULE's are filtered out by the preprocessor.
Signed-off-by: Damien George <damien@micropython.org>
This commit adds I2S protocol support for the rp2 port:
- I2S API is consistent with STM32 and ESP32 ports
- I2S configurations supported:
- master transmit and master receive
- 16-bit and 32-bit sample sizes
- mono and stereo formats
- sampling frequency
- 3 modes of operation:
- blocking
- non-blocking with callback
- uasyncio
- internal ring buffer size can be tuned
- DMA IRQs are managed on an I2S object basis, allowing other
RP2 entities to use DMA IRQs when I2S is not being used
- MicroPython documentation
- tested on Raspberry Pi Pico development board
- build metric changes for this commit: text(+4552), data(0), bss(+8)
Signed-off-by: Mike Teachman <mike.teachman@gmail.com>
This commit refactors machine.PWM and creates extmod/machine_pwm.c. The
esp8266, esp32 and rp2 ports all use this and provide implementations of
the required PWM functionality. This helps to reduce code duplication and
keep the same Python API across ports.
This commit does not make any functional changes.
Signed-off-by: Damien George <damien@micropython.org>
Initial support for machine.RTC on rp2 port. It only supports datetime()
method and nothing else. The method gets/returns a tuple of 8 items, just
like esp32 port, for example, but the usec parameter is ignored as the RP2
RTC only works up to seconds precision.
The Pico RTC isn't very useful as the time is lost during reset and there
seems to be no way to easily power up just the RTC clock with a low current
voltage, but still there seems to be use-cases for that, see issues #6831,
and a Thonny issue #1592. It was also requested for inclusion on v1.15
roadmap on #6832.
Signed-off-by: Krzysztof Adamski <k@japko.eu>
When UART is used for REPL and the MCU frequency is changed, the UART
has to be re-initialised. Besides that the UART may have to be recreated
after a frequency change, but with USB REPL this is not a problem.
Thanks to @HermannSW for spotting and providing the change.
