The file `ports/unix/moduos.c` uses `errno` so it needs to include
`errno.h`, otherwise a compiler error can occur.
Signed-off-by: David Lechner <david@pybricks.com>
Ensure that nimble and cyw43-driver are initialised when the board requires
it. Also make these work with `make submodules`.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
Commit 9670a156da missed one renaming of
MICROPY_PY_WIZNET5K to MICROPY_PY_NETWORK_WIZNET5K which prevented the
Wiznet interface from being enabled.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
This includes:
- Configuration file for the cyw43-driver.
- Integration of cyw43-driver into the build, using lwIP.
- Enhancements to machine.Pin to support extension IO pins provided by the
CYW43xx.
- More mp-hal pin helper functions.
- mp_hal_get_mac_ascii MAC address helper function.
- Addition of rp2.country() function to set the country code.
A board can enable this driver by setting MICROPY_PY_NETWORK_CYW43 in their
cmake snippet.
Work done in collaboration with Graham Sanderson and Peter Harper.
Signed-off-by: Damien George <damien@micropython.org>
It is more reliable and scales better when more components need it.
Work done in collaboration with Graham Sanderson and Peter Harper.
Signed-off-by: Damien George <damien@micropython.org>
Prior to this commit the following code would lock up the device when
Ctrl-D is entered at the REPL:
import gc, _thread
def collect_thread():
while True:
gc.collect()
_thread.start_new_thread(collect_thread, [])
Fixes part of #8494.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
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>
This ROM level is not yet fully defined, but it at least enables
MICROPY_PY_SYS_TRACEBACKLIMIT. The coverage build should have everything
enabled, so it makes sense to use this ROM level for it.
Signed-off-by: Damien George <damien@micropython.org>
So that the default configuration for the dev and coverage variants
includes all options set by MICROPY_CONFIG_ROM_LEVEL_EXTRA_FEATURES.
Note that enabling MICROPY_PY_SYS_STDIO_BUFFER on unix doesn't do anything
because unix doesn't use shared/runtime/sys_stdio_mphal.c.
Signed-off-by: Damien George <damien@micropython.org>
The default is the same as before: MICROPY_PY_USELECT=0 and
MICROPY_PY_USELECT_POSIX=1. But now this can be easily overridden at the
make command-line using, eg:
make VARIANT=dev CFLAGS_EXTRA=-DMICROPY_PY_USELECT=1
Signed-off-by: Damien George <damien@micropython.org>
Prior to this commit, running scan() without any APs available would give:
>>> wl.scan()
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
RuntimeError: Wifi Unknown Error 0x0102
Signed-off-by: Damien George <damien@micropython.org>
When tested, this reduces default MP binary sizes by approx 2-2.5%, and
very marginally increases performance in benchmarks. Build times seem very
similar to non-LTO when using gcc 12.
See #8733 for further discussion.
Signed-off-by: Angus Gratton <gus@projectgus.com>
Prerequisite for enabling Link Time Optimisation.
The _bl_state address is the same as _estack, but _estack is referred to as
a uint32_t elsewhere in the code. LTO doesn't like it when the same symbol
has two different types.
Signed-off-by: Angus Gratton <gus@projectgus.com>
Replaces preprocessor macro for SDRAM option from #ifdef to #if in order to
allow always setting the define `MICROPY_HW_SDRAM_AVAIL` just with the
appropriate value 0/1. This eliminates one `if` in the Makefile.
Add esp32.wake_on_ulp() to give access to esp_sleep_enable_ulp_wakeup(),
which is needed to allow the ULP co-processor to wake the main CPU from
deep sleep.
Allow esp32.ULP.load_binary() to use the maximum amount of memory available
again, which is 2040 bytes unless MICROPY_HW_RTC_USER_MEM_MAX is
customized.
This value regressed in 3d49b157b8
Using it for the rx-timeout. The value is given as ms, which is then
converted to character times. A value of less than a character time will
cause the rx call to return immediately after 1 character, which may be
inefficient at high transmission rates.
Addresses #8778.
This separates extmod source files from `py.mk`. Previously, `py.mk`
assumed that every consumer of the py/ directory also wanted to include
extmod/. However, this is not the case. For example, building mpy-cross
uses py/ but doesn't need extmod/.
This commit moves all extmod-specific items from `py.mk` to `extmod.mk` and
explicitly includes `extmod.mk` in ports that use it.
Signed-off-by: David Lechner <david@pybricks.com>
The following changes are made:
- Guard entire file with MICROPY_PY_LWIP, so it can be included in the
build while still being disabled (for consistency with other extmod
modules).
- Add modlwip.c to list of all extmod source in py/py.mk and
extmod/extmod.cmake so all ports can easily use it.
- Move generic modlwip GIT_SUBMODULES build configuration code from
ports/rp2/CMakeLists.txt to extmod/extmod.cmake, so it can be reused by
other ports.
- Remove now unnecessary inclusion of modlwip.c in EXTMOD_SRC_C in esp8266
port, and in SRC_QSTR in mimxrt port.
Signed-off-by: Damien George <damien@micropython.org>
The WLAN.config() method now supports "ssid", "security" and "key" as
aliases to the existing "essid", "authmode" and "password", which are now
deprecated. The help text and setup helper are also updated.
Addresses issue #8083.
The WLAN.config() method now supports "ssid", "security" and "key" as
aliases to the existing "essid", "authmode" and "password", which are now
deprecated.
Addresses issue #8083.
Updates the Zephyr port build instructions and CI to use the latest Zephyr
release tag.
Tested on frdm_k64f.
Signed-off-by: Maureen Helm <maureen.helm@intel.com>
When MICROPY_PY_MACHINE_I2C_TRANSFER_WRITE1 is enabled the port's hardware
I2C transfer functions should support the MP_MACHINE_I2C_FLAG_WRITE1
option, but software I2C will not. So add a flag to the I2C protocol
struct so each individual protocol can indicate whether it supports this
option or not.
Fixes issue #8765.
Signed-off-by: Damien George <damien@micropython.org>
Both led_init and led_state are configurable via MBOOT_BOARD_LED_INIT and
MBOOT_BOARD_LED_STATE respectively, so don't need to be MP_WEAK.
Furthermore, led_state and led0_state are private to ui.c so can be made
static.
Signed-off-by: Damien George <damien@micropython.org>
This IO was enabled in IDF commit 68f8b999bb69563f2f3d1d897bc073968f41f3bf,
which is available in IDF release v4.3.2 and above.
Signed-off-by: Damien George <damien@micropython.org>
Otherwise include directories are added unconditionally to the build
variables if the component (submodule) is checked out. This can lead to,
eg, the esp32 build using lib/lwip header files, instead of lwip header
files from the IDF.
Fixes issue #8727.
Signed-off-by: Damien George <damien@micropython.org>
System config block contains hardware unrelated to USB. So calling
`__SYSCFG_CLK_DISABLE()` during `HAL_PCD_MspDeInit()` has an adverse effect
on other system functionality.
Removing call to `__SYSCFG_CLK_DISABLE()` to rectify this issue.
This call was there since the beginning of the USB CDC code, added in
b30c02afa0.
According to ST Errata ES0206 Rev 18, Section 2.2.1, on STM32F427x,
STM32F437x, STM32F429x and STM32F439x.
If the system tick interrupt is enabled during stop mode while certain
bits are set in the DBGMCU_CR, then the system will immediately wake
from stop mode.
Suggested workaround is to disable system tick timer interrupt when
entering stop mode.
According to ST Errate ES0394 Rev 11, Section 2.2.17, on STM32WB55Cx and
STM32WB35Cx.
If the system tick interrupt is enabled during stop 0, stop 1 or stop 2
while certain bits are set in DBGMCU_CR, then system will immediately
wake from stop mode but the system remains in low power state. The CPU
therefore fetches incorrect data from inactive Flash, which can cause a
hard fault.
Suggested workaround is to disable system tick timer interrupt when
entering stop mode.
This was added by mistake in 8f68e26f79 when
adding support for G4 MCUs, which does not using this get_bank() function.
FLASH_OPTR_DBANK is only defined on G4 and L4 MCUs, so on H7 this
FLASH_BANK_2 code was being wrongly excluded.
Signed-off-by: Damien George <damien@micropython.org>
Also remove redundant modusocket.c and modnetwork.c sources, they are
already added by extmod/extmod.cmake.
Signed-off-by: Damien George <damien@micropython.org>
The callback passed to add_alarm_in_ms must return microseconds, even
though the initial delay is in milliseconds. Fix this use, and to avoid
further confusion use the add_alarm_in_us function instead.
Signed-off-by: Damien George <damien@micropython.org>
The nxp_driver v2.10 allows for/requires some changes to the code:
- Remove some part of pwm_backlog.*, which is provided by the lib now.
- Change eth.c: the newer versions have additional parameters of the
library versions.
- Change sdcard.c: use TransferBlocking instead of TransferNonblocking.
- Add some support for the MIMXRT1176 device.
- Set the clocks for UART, I2C, Timer.
- Integrate the I2S module and fix a rebase error.
- Use blocking transfer only for SPI. It's faster and interferes less with
other modules.
- Use the clock_config.c files of library v2.8.5. The mimxrt files keeps
the clock_config.c files from Verson 2.8.5. With clock_config.c from
v2.10, the boards do not work. Refactoring of the clock set-up is on the
to-do list.
- Enable expiry timers for UART, I2C and SPI, avoiding a stall in library
code.
- The clock_config.* files are moved from the board-specific directories to
the boards directory and given a MCU related name.
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 cleans up the parsing of MP_REGISTER_MODULE() and generation of
genhdr/moduledefs.h so that it uses the same process as compressed error
string messages, using the output of qstr extraction.
This makes sure all MP_REGISTER_MODULE()'s that are part of the build are
correctly picked up. Previously the extraction would miss some (eg if you
had a mod.c file in the board directory for an stm32 board).
Build speed is more or less unchanged.
Thanks to @stinos for the ports/windows/msvc/genhdr.targets changes.
Signed-off-by: Damien George <damien@micropython.org>
The following changes are made:
- Use software SPI for external SPI flash access when building mboot.
- Enable the mboot filesystem-loading feature, with FAT FS support.
- Increase the frequency of the CPU when in mboot to 96MHz, to increase the
speed of SPI flash accesses and programming.
Signed-off-by: Damien George <damien@micropython.org>
This allows a board to modify initial_r0 if needed.
Also make default board behaviour functions always available, named as
mboot_get_reset_mode_default and mboot_state_change_default.
Signed-off-by: Damien George <damien@micropython.org>
I2C transfers are much more efficient if they are combined, instead of
doing separate writes and reads.
Fixes issue #7134.
Signed-off-by: Damien George <damien@micropython.org>
This follows on from a5324a1074 and allows
mpy-cross to dynamically select whether ARMv7-M instructions are supported
in @micropython.asm_thumb functions.
The config option MICROPY_EMIT_INLINE_THUMB_ARMV7M is no longer needed, it
is now controlled by MICROPY_EMIT_THUMB_ARMV7M.
Signed-off-by: Damien George <damien@micropython.org>
For ports with MICROPY_VFS and MICROPY_PY_IO enabled their configuration
can now be simplified to use the defaults for mp_import_stat and
mp_builtin_open.
This commit makes no functional change, except for the following minor
points:
- the built-in "open" is removed from the minimal port (it previously did
nothing)
- the duplicate built-in "input" is removed from the esp32 port
- qemu-arm now delegates to VFS import/open
Signed-off-by: Damien George <damien@micropython.org>
The following changes are made:
- If MICROPY_VFS is enabled then mp_vfs_import_stat and mp_vfs_open are
automatically used for mp_import_stat and mp_builtin_open respectively.
- If MICROPY_PY_IO is enabled then "open" is automatically included in the
set of builtins, and points to mp_builtin_open_obj.
This helps to clean up and simplify the most common port configuration.
Signed-off-by: Damien George <damien@micropython.org>
This reverts commit 2668337f36.
The issue with potential breaking of the BLE RX path in the radio is fixed
since WS v1.12.0.
Signed-off-by: Damien George <damien@micropython.org>
This adds a new command line option to the unix port `-X realtime` to
enable realtime priority on threads. This enables high precision timers
for applications that need more accurate timers.
Related docs:
https://developer.apple.com/library/archive/technotes/tn2169/_index.html
Fixes issue #8621.
Signed-off-by: David Lechner <david@pybricks.com>
This adds the `mp_` prefix to the `thread_t` type. The name `thread_t`
conflicts with the same in `mach/mach_types.h` on macOS.
Signed-off-by: David Lechner <david@lechnology.com>
Reasons for removal:
- It did not work properly because it stopped the hardware watchdog
timer while keeping the software watchdog running (issue #8597).
- There isn't a deinit method for the WDT in any other port.
- "The watchdog is not intended to be stopped. That is a feature."
(See #8600.)
Replace the timer-based sleep with the standard win32 call since the former
has no benefits: even though it allows specifying the time in 100uSec
chunks, the actual resolution is still limited by the OS and is never
better than 1mSec.
For clarity move all of this next to the mp_hal_delay_ms definition so all
related functions are in one place.
This replaces occurences of
foo_t *foo = m_new_obj(foo_t);
foo->base.type = &foo_type;
with
foo_t *foo = mp_obj_malloc(foo_t, &foo_type);
Excludes any places where base is a sub-field or when new0/memset is used.
Signed-off-by: Jim Mussared <jim.mussared@gmail.com>
* Change device name table to list style to show properly.
* Change the link of cable connection information to the latest.
Signed-off-by: Takeo Takahashi <takeo.takahashi.xv@renesas.com>
This contains a string useful for identifying the underlying machine. This
string is kept consistent with the second part of the REPL banner via the
new config option MICROPY_BANNER_MACHINE.
This makes os.uname() more or less redundant, as all the information in
os.uname() is now available in the sys module.
Signed-off-by: Damien George <damien@micropython.org>
This commit adds the git hash and build date to sys.version. This is
allowed according to CPython docs, and is what PyPy does. The docs state:
A string containing the version number of the Python interpreter plus
additional information on the build number and compiler used.
Eg on CPython:
Python 3.10.4 (main, Mar 23 2022, 23:05:40) [GCC 11.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import sys
>>> sys.version
'3.10.4 (main, Mar 23 2022, 23:05:40) [GCC 11.2.0]'
and PyPy:
Python 2.7.12 (5.6.0+dfsg-4, Nov 20 2016, 10:43:30)
[PyPy 5.6.0 with GCC 6.2.0 20161109] on linux2
Type "help", "copyright", "credits" or "license" for more information.
>>>> import sys
>>>> sys.version
'2.7.12 (5.6.0+dfsg-4, Nov 20 2016, 10:43:30)\n[PyPy 5.6.0 with GCC ...
With this commit on MicroPython we now have:
MicroPython v1.18-371-g9d08eb024 on 2022-04-28; linux [GCC 11.2.0] v...
Use Ctrl-D to exit, Ctrl-E for paste mode
>>> import sys
>>> sys.version
'3.4.0; MicroPython v1.18-371-g9d08eb024 on 2022-04-28'
Note that the start of the banner is the same as the end of sys.version.
This helps to keep code size under control because the string can be reused
by the compiler.
Signed-off-by: Damien George <damien@micropython.org>
This commit adds a board definition for NUCLEO_G0B1RE. This board has the
REPL on UART2 which is connected to the on-board ST-link USB-UART.
Signed-off-by: Asensio Lorenzo Sempere <asensio.aerospace@gmail.com>
This implements self-triggering of the Flash NVIC interrupt on Cortex-M0
devices, which allows enabling internal storage on those MCUs.
Signed-off-by: Asensio Lorenzo Sempere <asensio.aerospace@gmail.com>
These files that are reformatted only now fall under the list of files to
apply uncrustify/black formatting to.
Signed-off-by: Damien George <damien@micropython.org>
To avoid any I/O glitches in mp_hal_pin_config, make sure a valid alternate
function is set in AFR first before switching the pin mode. When switching
from AF to INPUT or OUTPUT, the AF in AFR will remain valid up until the
pin mode is switched.
On ESP32 S2/S3 variants, GPIO0 through GPIO21 are valid RTC pins. This
commit defines the valid RTC_VALID_EXT_PINS for the S2/S3 variants,
otherwise, it keeps functionality the same.
For ESP32-S3 configurations, CONFIG_SPIRAM_MODE_OCT requires pins 33-37 for
PSRAM. So exclude them from the machine_pin_type and machine_pin_irq_type
object tables.
These boards do not build with IDF v4.4 because the section .iram0.text
does not fit in region iram0_0_seg. Enabling SPIRAM increases the code
size so use -Os instead of -O2 to build these boards.
Fixes issue #8260.
Some S2/S3 modules don't use the native USB interface but instead have an
external USB-UART. To make the GENERIC_S3/S3 firmware work on these boards
the UART REPL is enabled in addition to the native USB CDC REPL.
Fixes issues #8418 and #8524.
Signed-off-by: Damien George <damien@micropython.org>
Duplication of characters is caused by re-entrant calls from separate cores
of uart_fill_tx_fifo(). This patch uses a mutex to ensure that a
re-entrant execution of the function returns without affecting the UART
FIFO.
Fixes issues #8344 and #8360.
- Add board-level configuration option to set the SMPS supply mode.
- Wait for valid voltage levels after configuring the SMPS mode.
- Wait for external supply ready flag if SMPS supplies external circuitry.
All user interface (LED, button) code has been moved to ui.c, and the
interface to this code with the rest of the system now goes through calls
to mboot_state_change(). This state-change function can be overridden by a
board to fully customise the user interface behaviour.
Signed-off-by: Damien George <damien@micropython.org>
This follows the CPython change: https://bugs.python.org/issue21455
Socket listen backlog defaults to 2 if not given, based on most bare metal
targets not having many resources for a large backlog. On UNIX it defaults
to SOMAXCONN or 128, whichever is less.
Changes in this commit:
- Fix USB CDC RX handling to not block when unprocessed. The fix follows
5873390226.
- Fix dupterm rx.
- Remove some obsolete lines.
This commit changes the method of waiting for SPI being not busy. Instead
of the FIFO size, the TransferBusyFlag is probed.
Also, raise an error if the transfer failed.
Changes in this commit:
- Start the RTC Timer at system boot. Otherwise time.time() will advance
only if an RTC() object was created.
- Set the time to a more recent date than Jan 1, 1970, if not set. That is
2013/10/14, 19:53:11, MicroPython's first commit.
- Compensate an underflow in in timeutils_seconds_since_2000(), called by
time.time(), if the time is set to a pre-2000 date.
This is enabled at MICROPY_CONFIG_ROM_LEVEL_EXTRA_FEATURES, which is the
default for stm32. Not setting the value in mpconfigboard.h allows boards
to optionally configure it.
Signed-off-by: Damien George <damien@micropython.org>
Prior to this commit, the USB CDC OUT endpoint got NACK'd if a character
was received but not consumed by the application, e.g. via
sys.stdin.read(). This meant that USB CDC was blocked and no additional
characters could be sent from the host. In particular a ctrl-C could not
interrupt the application if another character was pending.
To fix the issue, the approach in this commit uses a callback tud_cdc_rx_cb
which is called by the TinyUSB stack on reception of new CDC data. By
consuming the data immediately, the endpoint does not stall anymore. The
previous handler tud_cdc_rx_wanted_cb was made obsolete and removed.
In addition some cleanup was done along the way: by adding interrupt_char.c
and removing the existing code mp_hal_set_interrupt_char(). Also, there is
now only one (stdin) ringbuffer.
Fixes issue #7996.
The CAN.initfilterbanks() class method is removed, and its functionality is
replaced with the "num_filter_banks" keyword argument to the CAN
constructor and CAN.init(). This configures the filter bank split.
This new approach provides more flexibility configuring the resources used
by a given CAN instance, allowing other MCUs like H7 to fit the API. It
also brings CAN closer to how other machine peripherals are configured,
where everything is done in the constructor/init method.
This is a breaking change to the CAN API.
CAN.recv() now returns a 5-tuple, with the new element in the second
position being a boolean, True if the ID is extended.
This is a breaking change of the API for CAN.recv().
A CAN bus can have mixed classic/FD nodes. Prior to this patch the CAN API
could be configured for either standard or extended ID, but not both/mixed
operation.
This patch allows extended IDs to be filtered and enabled on a per-message
basis, in send(), setfilter() and clearfilter().
This is a breaking change to the API: init() no longer accepts the extframe
keyword argument.
- Enable CAN FD frame support and BRS.
- Optimize the message RAM usage per FDCAN instance.
- Document the usage and different sections of the Message RAM.
This commit adds support for machine.I2S on the mimxrt port. The I2S API
is consistent with the existing stm32, esp32, and rp2 implementations.
I2S features:
- controller transmit and controller receive
- 16-bit and 32-bit sample sizes
- mono and stereo formats
- sampling frequencies from 8kHz to 48kHz
- 3 modes of operation:
- blocking
- non-blocking with callback
- uasyncio
- configurable internal buffer
- optional MCK
Tested with the following development boards:
- MIMXRT1010_EVK, MIMXRT1015_EVK, MIMXRT1020_EVK, MIMXRT1050_EVK
- Teensy 4.0, Teensy 4.1
- Olimex RT1010
- Seeed ARCH MIX
Tested with the following I2S hardware peripherals:
- UDA1334
- GY-SPH0645LM4H
- WM8960 codec on board the MIMXRT boards and separate breakout board
- INMP441
- PCM5102
- SGTL5000 on the Teensy audio shield
Signed-off-by: Mike Teachman <mike.teachman@gmail.com>
In particular, it is called by the constructor if the instance already
exists. So if the previous instance was deinit'd then it will be deinit'd
a second time.
Signed-off-by: Damien George <damien@micropython.org>
Tested on PYBV10 and PYBD_SF6, with MBOOT_FSLOAD enabled and programming
new firmware from a .dfu.gz file stored on the SD card.
Signed-off-by: Damien George <damien@micropython.org>
If enabled via MBOOT_ADDRESS_SPACE_64BIT (it's disabled by default) then
read addresses will be 64-bit.
Signed-off-by: Damien George <damien@micropython.org>
Even if MBOOT_FSLOAD is disabled, mboot should still check for 0x70ad0080
so it can immediately return to the application if this feature is not
enabled. Otherwise mboot will get stuck in DFU mode.
Signed-off-by: Damien George <damien@micropython.org>
The main functionality of this info function is available via the existing
micropython.mem_info() and micropython.qstr_info() functions. The printing
of the address space layout doesn't add much and removing esp.info() saves
about 600 bytes.
Signed-off-by: Damien George <damien@micropython.org>
Add a new function to control whether held pins will retain their function
through deep-sleep.
Also document this function and explain how to use this in quickref to
retain pin configuration during deep-sleep.
The current pull=Pin.PULL_HOLD argument doesn't make a lot of sense in the
context of what it actually does vs what the ESP32 quickref document says
it does.
This commit removes PULL_HOLD and adds a new hold=True|False keyword
argument to Pin()/Pin.init(). Setting this to True will cause the ESP32 to
lock the configuration of the pin – including direction, output value,
drive strength, pull-up/-down – such that it can't be accidentally changed
and will be retained through a watchdog or internal reset.
Fixes issue #8283, and see also #8284.
According to the C standard the free(void *ptr) function: if ptr is a null
pointer, no action occurs.
Signed-off-by: Peter Züger <zueger.peter@icloud.com>
All variants now use extmod/moduos.c as their uos module implementation.
In particular this means they all have MICROPY_VFS enabled and use VfsPosix
for their filesystem.
As part of this, the available functions in uos become more consistent with
other ports:
- coverage variant gets uos.urandom
- minimal and standard variant get: unlink, chdir, getcwd, listdir
Signed-off-by: Damien George <damien@micropython.org>
Ensure the symmetry of PWM: the duty rate of X and Q channels was not 50%,
when it should have been. That is evident at high frequencies, like 15Mhz
or 37.5 MHz. At low frequencies the deviation mattered less. The A/B
channels were fine.
Also round up or down non-integer division factors. Before, always the
floor value was used.
That caused Ethernet to lock up at high data rates after ~200MByte data
average in a row. Tested now with data bursts up to 10 GByte and overall
data rates of ~8MByte/s at the Eth100 port.
Sometimes frames could not be sent immediately because the controller was
still busy with previous frames. Then, an error was returned to lwip.
This fix adds a limited number of retries for this busy state, waiting
100µs before the next attempt. Typically the transmit succeeds now at the
second attempt.
Second change: Reset the controller for a clean state after soft reset.
OCOTP_Init() has been removed from mphalport.c. The library files are
missing for the MIMXRT1015, and for just reading the OCOTP the Init is not
required.
The disk_access header was moved to a different path in Zephyr v2.6.0.
The old path was deprecated for two releases (v2.6.0 and v2.7.0) and
will no longer be supported after Zephyr v2.7.0.
Signed-off-by: Maureen Helm <maureen.helm@intel.com>
If setting the frequency to a value used already by an existing timer, this
timer will be used. But still, the duty cycle for that channel may have to
be changed.
Fixes issues #8306 and #8345.
If MicroPython threads are enabled, loops waiting for an incoming event
should release the GIL and suspend, allowing other tasks to run while they
wait.
Prior to this commit, the problem can easily be observed by running a
thread that is both busy and regularly releases the GIL (for example a loop
doing something then sleeping a few ms after each iteration). When the
main task is at the REPL, the thread is significantly stalled. If the main
task is manually made to release the GIL (for example, by calling
utime.sleep_ms(500)) the other thread can be seen immediately working at
the expected speed again.
Additionally, there are various instances in where blocking functions run
MICROPY_EVENT_POLL_HOOK in a loop while they wait for a certain event/
condition. For example the uselect methods poll objects to determine
whether data is available, but uses 100% of CPU while it does, constantly
calling MICROPY_EVENT_POLL_HOOK in the process.
The MICROPY_EVENT_POLL_HOOK macro is only ever used in waiting loops, where
(if threads are enabled) it makes sense to yield for a single tick so that
these loops do not consume all CPU cycles but instead other threads may
execute. (In fact, the thing these loops wait for may even indirectly or
directly depend on another task being able to run.)
This change moves the sleep that was inside the REPL input function to
inside the MICROPY_EVENT_POLL_HOOK macro, where the GIL is already being
released, solving both the blocking REPL issue and the 100% CPU use issue
at the same time.
Signed-off-by: Daniël van de Giessen <daniel@dvdgiessen.nl>
The stack (and arg) of core1 is itself a root pointer, not just the entries
in it. Without this fix the GC could reclaim the entire stack (and
argument object).
Fixes issues #7124 and #7981.
.py files are valid source files and shouldn't be ignored. This line was
from the early days when .py files in the unix directory were used for
testing.
Signed-off-by: Damien George <damien@micropython.org>
The xmlns attribute is required for older msbuild version (e.g. for
VS2015). Add it where needed, and reorder the attributes so all
files look the same.
Background: .mpy files are precompiled .py files, built using mpy-cross,
that contain compiled bytecode functions (and can also contain machine
code). The benefit of using an .mpy file over a .py file is that they are
faster to import and take less memory when importing. They are also
smaller on disk.
But the real benefit of .mpy files comes when they are frozen into the
firmware. This is done by loading the .mpy file during compilation of the
firmware and turning it into a set of big C data structures (the job of
mpy-tool.py), which are then compiled and downloaded into the ROM of a
device. These C data structures can be executed in-place, ie directly from
ROM. This makes importing even faster because there is very little to do,
and also means such frozen modules take up much less RAM (because their
bytecode stays in ROM).
The downside of frozen code is that it requires recompiling and reflashing
the entire firmware. This can be a big barrier to entry, slows down
development time, and makes it harder to do OTA updates of frozen code
(because the whole firmware must be updated).
This commit attempts to solve this problem by providing a solution that
sits between loading .mpy files into RAM and freezing them into the
firmware. The .mpy file format has been reworked so that it consists of
data and bytecode which is mostly static and ready to run in-place. If
these new .mpy files are located in flash/ROM which is memory addressable,
the .mpy file can be executed (mostly) in-place.
With this approach there is still a small amount of unpacking and linking
of the .mpy file that needs to be done when it's imported, but it's still
much better than loading an .mpy from disk into RAM (although not as good
as freezing .mpy files into the firmware).
The main trick to make static .mpy files is to adjust the bytecode so any
qstrs that it references now go through a lookup table to convert from
local qstr number in the module to global qstr number in the firmware.
That means the bytecode does not need linking/rewriting of qstrs when it's
loaded. Instead only a small qstr table needs to be built (and put in RAM)
at import time. This means the bytecode itself is static/constant and can
be used directly if it's in addressable memory. Also the qstr string data
in the .mpy file, and some constant object data, can be used directly.
Note that the qstr table is global to the module (ie not per function).
In more detail, in the VM what used to be (schematically):
qst = DECODE_QSTR_VALUE;
is now (schematically):
idx = DECODE_QSTR_INDEX;
qst = qstr_table[idx];
That allows the bytecode to be fixed at compile time and not need
relinking/rewriting of the qstr values. Only qstr_table needs to be linked
when the .mpy is loaded.
Incidentally, this helps to reduce the size of bytecode because what used
to be 2-byte qstr values in the bytecode are now (mostly) 1-byte indices.
If the module uses the same qstr more than two times then the bytecode is
smaller than before.
The following changes are measured for this commit compared to the
previous (the baseline):
- average 7%-9% reduction in size of .mpy files
- frozen code size is reduced by about 5%-7%
- importing .py files uses about 5% less RAM in total
- importing .mpy files uses about 4% less RAM in total
- importing .py and .mpy files takes about the same time as before
The qstr indirection in the bytecode has only a small impact on VM
performance. For stm32 on PYBv1.0 the performance change of this commit
is:
diff of scores (higher is better)
N=100 M=100 baseline -> this-commit diff diff% (error%)
bm_chaos.py 371.07 -> 357.39 : -13.68 = -3.687% (+/-0.02%)
bm_fannkuch.py 78.72 -> 77.49 : -1.23 = -1.563% (+/-0.01%)
bm_fft.py 2591.73 -> 2539.28 : -52.45 = -2.024% (+/-0.00%)
bm_float.py 6034.93 -> 5908.30 : -126.63 = -2.098% (+/-0.01%)
bm_hexiom.py 48.96 -> 47.93 : -1.03 = -2.104% (+/-0.00%)
bm_nqueens.py 4510.63 -> 4459.94 : -50.69 = -1.124% (+/-0.00%)
bm_pidigits.py 650.28 -> 644.96 : -5.32 = -0.818% (+/-0.23%)
core_import_mpy_multi.py 564.77 -> 581.49 : +16.72 = +2.960% (+/-0.01%)
core_import_mpy_single.py 68.67 -> 67.16 : -1.51 = -2.199% (+/-0.01%)
core_qstr.py 64.16 -> 64.12 : -0.04 = -0.062% (+/-0.00%)
core_yield_from.py 362.58 -> 354.50 : -8.08 = -2.228% (+/-0.00%)
misc_aes.py 429.69 -> 405.59 : -24.10 = -5.609% (+/-0.01%)
misc_mandel.py 3485.13 -> 3416.51 : -68.62 = -1.969% (+/-0.00%)
misc_pystone.py 2496.53 -> 2405.56 : -90.97 = -3.644% (+/-0.01%)
misc_raytrace.py 381.47 -> 374.01 : -7.46 = -1.956% (+/-0.01%)
viper_call0.py 576.73 -> 572.49 : -4.24 = -0.735% (+/-0.04%)
viper_call1a.py 550.37 -> 546.21 : -4.16 = -0.756% (+/-0.09%)
viper_call1b.py 438.23 -> 435.68 : -2.55 = -0.582% (+/-0.06%)
viper_call1c.py 442.84 -> 440.04 : -2.80 = -0.632% (+/-0.08%)
viper_call2a.py 536.31 -> 532.35 : -3.96 = -0.738% (+/-0.06%)
viper_call2b.py 382.34 -> 377.07 : -5.27 = -1.378% (+/-0.03%)
And for unix on x64:
diff of scores (higher is better)
N=2000 M=2000 baseline -> this-commit diff diff% (error%)
bm_chaos.py 13594.20 -> 13073.84 : -520.36 = -3.828% (+/-5.44%)
bm_fannkuch.py 60.63 -> 59.58 : -1.05 = -1.732% (+/-3.01%)
bm_fft.py 112009.15 -> 111603.32 : -405.83 = -0.362% (+/-4.03%)
bm_float.py 246202.55 -> 247923.81 : +1721.26 = +0.699% (+/-2.79%)
bm_hexiom.py 615.65 -> 617.21 : +1.56 = +0.253% (+/-1.64%)
bm_nqueens.py 215807.95 -> 215600.96 : -206.99 = -0.096% (+/-3.52%)
bm_pidigits.py 8246.74 -> 8422.82 : +176.08 = +2.135% (+/-3.64%)
misc_aes.py 16133.00 -> 16452.74 : +319.74 = +1.982% (+/-1.50%)
misc_mandel.py 128146.69 -> 130796.43 : +2649.74 = +2.068% (+/-3.18%)
misc_pystone.py 83811.49 -> 83124.85 : -686.64 = -0.819% (+/-1.03%)
misc_raytrace.py 21688.02 -> 21385.10 : -302.92 = -1.397% (+/-3.20%)
The code size change is (firmware with a lot of frozen code benefits the
most):
bare-arm: +396 +0.697%
minimal x86: +1595 +0.979% [incl +32(data)]
unix x64: +2408 +0.470% [incl +800(data)]
unix nanbox: +1396 +0.309% [incl -96(data)]
stm32: -1256 -0.318% PYBV10
cc3200: +288 +0.157%
esp8266: -260 -0.037% GENERIC
esp32: -216 -0.014% GENERIC[incl -1072(data)]
nrf: +116 +0.067% pca10040
rp2: -664 -0.135% PICO
samd: +844 +0.607% ADAFRUIT_ITSYBITSY_M4_EXPRESS
As part of this change the .mpy file format version is bumped to version 6.
And mpy-tool.py has been improved to provide a good visualisation of the
contents of .mpy files.
In summary: this commit changes the bytecode to use qstr indirection, and
reworks the .mpy file format to be simpler and allow .mpy files to be
executed in-place. Performance is not impacted too much. Eventually it
will be possible to store such .mpy files in a linear, read-only, memory-
mappable filesystem so they can be executed from flash/ROM. This will
essentially be able to replace frozen code for most applications.
Signed-off-by: Damien George <damien@micropython.org>
The unix port's main.c gets used by unix and windows ports, and with a
variety of compilers, so it's convenient to see which version is actually
being used immediately when starting micropython. This is similar to what
CPython does.
