This commit lets the Thumb native emitter generate a proper opcode
sequence when calculating an indexed register offset for load/store
operations with said offset beight both greater than 65535 and not
able to be represented as a shifted 8-bit bitmask.
The original code would assume the scaled index would always fit in 16
bits and silently discard upper bits of the offset. Now an optimised
constant loading sequence is emitted instead, and the final offset is
also stored in the correct register in all cases.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
This commit adds a series of test cases to exercise the Viper code
generator load/store emitting capabilities on certain boundary
conditions.
The new test cases check whether the emitted load/store code performs
correctly when dealing with specific memory offsets, which trigger
specific code generation sequences on different architectures.
Right now the cases are for unsigned offsets whose bitmasks span up to
5, 8, and 12 bits (respectively Arm/Thumb, Xtensa, RV32).
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
This commit removes the ASM_LOAD_REG_REG and ASM_STORE_REG_REG generic
ASM API opcodes from all backends, as they are not used anymore in the
native emitter framework.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
This commit cleans up the Viper code generation blocks for
register-indexed load and store operations.
An attempt is made to simplify the code in the common code generator
code block, by moving architecture-specific code to the appropriate
native generation backends whenever possible. This should make that
specific bit of code in the Viper generator clearer and easier to
maintain in the long term.
To achieve this, six generic assembler meta-opcodes have been
introduced, named `ASM_{LOAD,STORE}{8,16,32}_REG_REG_REG`. A
platform-independent implementation for those operations is provided, so
backends that cannot emit a shorter sequence for the requested operation
or are fine with the platform-independent implementation can just not
provide said meta-opcodes.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
This commit lets the Viper code generator use optimised code sequence
for register-indexed load and store operations when generating Thumb
code.
Register-indexed load and store operations for Thumb now can take at
most two machine opcodes for halfword and word values, and just a single
machine opcode for byte values. The original implementation could
generate up to four opcodes in the worst case (dealing with word
values).
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
This commit lets the Viper code generator use optimised code sequences
for register-indexed load and store operations when generating Arm code.
The existing code defaulted to generic multi-operations code sequences
for Arm code on most cases. Now optimised implementations are provided
for register-indexed loads and stores of all data sizes, taking at most
two machine opcodes for each operation.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
esp32 port will disconnect if active(0) is called on a STA
interface, but rp2 port stays associated without this change.
This work was funded through GitHub Sponsors.
Signed-off-by: Angus Gratton <angus@redyak.com.au>
This commit makes the JSON parser raise an exception when handling
objects or arrays whose declaration is incomplete, as in missing the
closing marker (brace or bracket) and if the missing marker would have
been the last non-whitespace character in the incoming string.
Since CPython's JSON parser would raise an exception in such a case,
unlike MicroPython's, this commit aligns MicroPython's behaviour with
CPython.
This commit fixes issue #17141.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
The output `_result.json` file generated by `run-tests.py` currently
contains a list of failed tests. This commit adds to the output a list of
passed and skipped tests, and so now provides full information about which
tests were run and what their results were.
Signed-off-by: Damien George <damien@micropython.org>
Add `mpremote fs tree` command to show a tree of the device's files. It:
- Shows a treeview from current path or specified path.
- Uses the graph chars ("├── ", "└── ") (not configurable).
- Has the options:
-v/--verbose adds the serial device name to the top of the tree
-s/--size add a size to the files
-h/--human add a human readable size to the files
Signed-off-by: Jos Verlinde <jos_verlinde@hotmail.com>
Support the new PHY_GENERIC device type, added in ESP-IDF v5.4.0 [1].
This PHY driver was added to ESP-IDF to support "generic"/oddball PHY
LAN chips like the JL1101, which offer no features beyond the bare
802.3 PHY standard and don't actually need a chip-specific driver (see
discussion at [2]).
[1] 0738314308
[2] https://github.com/espressif/esp-eth-drivers/pull/28
Signed-off-by: Elvis Pfutzenreuter <epxx@epxx.co>
The default I2C init does not require setting SCL or SDA but the default
I2C0 pins for C3, S3 conflict with the espressif GPIO usage.
For the C3, pins 18/19 are for USB/JTAG. If used for I2C() they will cause
the REPL to hang on initialization of the I2C.
For the S3 pin 19 is allocated for USB/JTAG also but the defaults do not
seem to affect the REPL.
See related #16956.
Fixes issue #17103.
Signed-off-by: Rick Sorensen <rick.sorensen@gmail.com>
Signed-off-by: Angus Gratton <angus@redyak.com.au>
This adds a new function, `esp32.idf_task_info()`, that can be used to
retrieve task statistics which is useful for diagnosing issues where some
tasks are using up a lot of CPU time.
It's best used in conjunction with the `utop` module from micropython-lib.
Signed-off-by: Daniël van de Giessen <daniel@dvdgiessen.nl>
This adds support for LAN8670 to the esp32 port. Enabled conditionally for
the esp32 target, if ESP-IDF version is new enough (v5.3 or newer).
Fixes issue #15731.
Signed-off-by: Damien George <damien@micropython.org>
Signed-off-by: Angus Gratton <angus@redyak.com.au>
This version of the IDF uses about 1KB more IRAM and 1KB more DRAM on most
boards, but 6.5KB more DRAM usage on the S3. It seems that's due to a lot
of small increases in many components.
Signed-off-by: Ihor Nehrutsa <Ihor.Nehrutsa@gmail.com>
Python threads (created via the `_thread` module) are backed by a FreeRTOS
task. Managing the deletion of the task can be tricky, and there are
currently some bugs with this in the esp32 port.
The actual crash seen was in FreeRTOS' `uxListRemove()`, and that's because
of two calls to `vTaskDelete()` for the same task: one in
`freertos_entry()` when the task ran to completion, and the other in
`mp_thread_deinit()`. The latter tried to delete the task a second time
because it was still in the linked list, because `vTaskPreDeletionHook()`
had not yet been called. And the reason `vTaskPreDeletionHook()` was yet
to be called is because the FreeRTOS idle task was starved.
This commit fixes that.
There are three things done by this commit:
- remove the `vTaskPreDeletionHook`, it's not needed anymore because task
stack memory is allocated by the IDF, not on the MicroPython heap
- when a task finishes it now removes itself from the linked list, just
before it deletes itself
- on soft reset, all tasks are deleted and removed from the linked list in
one swoop (while the mutex is held)
Signed-off-by: Damien George <damien@micropython.org>
This reduce inconsistencies between esp32 PWM and other ports:
1. duty_u16() high value is 2**16-1 == 65535
2. Invert PWM wave with invert=1 parameter
3. Enable PWM in light sleep mode
4. Allow PWM output and read pulse input simultaneously on the same Pin()
5. Code refactoring
Co-Authored-By: Angus Gratton <angus@redyak.com.au>
Co-Authored-By: robert-hh <robert@hammelrath.com>
Co-Authored-By: Andrew Leech <andrew.leech@planetinnovation.com.au>
Co-Authored-By: Yoann Darche <yoannd@hotmail.com>
Signed-off-by: Ihor Nehrutsa <Ihor.Nehrutsa@gmail.com>
This reduces inconsitencies between esp32 and other ports.
According to the discussion in #10817.
Signed-off-by: Ihor Nehrutsa <Ihor.Nehrutsa@gmail.com>
Previously, the navigation ended up messy when the (long) description of
the item became used as a 2nd level header, meaning that it was placed in
the navigation. Check for this when generating cpydiff so that new cases
don't sneak in unnoticed.
Signed-off-by: Jeff Epler <jepler@gmail.com>
The heading character for the difference title was always "~", but items
had been added which had just a single heading level. This made the
generated table of contents confused about heading levels, because heading
levels are not fixed in rst, but are inferred from the order they appear in
the document.
Signed-off-by: Jeff Epler <jepler@gmail.com>
In the syntax_space cpydiff, all the warnings were shown after the other
output. This is because the output always showed all of stdout first and
all of stdout second.
By running Python in unbuffered mode and using `stderr=STDOUT`, the two
streams are interleaved in exactly the order they're printed, so the
SyntaxWarnings are interleaved with the other output.
By using the `encoding=` argument of Popen, the need to explicitly convert
to utf-8 is avoided. The encoding of the input also becomes utf-8 in this
case, which all the test cases are (well, they're all ASCII, I think). As
in `run-tests.py`, setting PYTHONIOENCODING ensures the Python
interpreter's input and output are in utf-8, which is not always the case,
especially on Windows systems.
I spot-checked the generated doc pages and they all seemed to make sense
still.
Signed-off-by: Jeff Epler <jepler@gmail.com>
The ESP32 PPP implementation predates the generic
implementation in extmod. The new extmod
implementation has a few advantages such as a
better deinitialisation procedure (the ESP32
implemementation would not clean up properly and
cause crashes if recreated) and using the UART IRQ
functionality instead of running a task to read
data from the UART.
This change restructures the ESP implementation to
be much closer to the new extmod version, while
also bringing a few tiny improvements from the
ESP32 version to the extmod version. The diff
between extmod/network_ppp_lwip.c and
ports/esp32/network_ppp.c is now a small set of
easy to review ESP32 port-specific changes.
Signed-off-by: Daniël van de Giessen <daniel@dvdgiessen.nl>
This commit changes the error handler for WiFi operations to recognise
out of memory conditions reported by ESP-IDF functions, and report them
as more descriptive exceptions rather than a generic "error 0x101".
The error handler only provided a human-readable error description for
WiFi-specific error codes (codes in the ESP_ERR_WIFI_BASE range), but
WiFi functions are known to return other codes. Now ESP_ERR_NO_MEM is
covered with a specific error message, making it easier to debug issues
related to running out of ESP-IDF heap.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
If the driver was reinitialised while there was
already an event task running the queue that task
is trying to receive from would be deleted,
causing it to try to take a lock that no longer
existed and deadlocking the CPU.
This change ensures the task is always shut down
before recreating the queue and recreates the task
afterwards.
It also allows setting an IRQ handler before the
UART is initialized (like other ports allow),
removes the task when the UART is deinitialized
(which was previously missing), adds a check that
no event task can be started when no queue exists,
and adds a check to prevent reinitialising the
UART driver unnecessarily.
Signed-off-by: Daniël van de Giessen <daniel@dvdgiessen.nl>
This commit introduces a new port configuration entry allowing the entry
point function name to be changed, from "app_main" to a custom name.
This is needed when MicroPython is embedded as an ESP-IDF component,
since the "app_main" symbol is already provided elsewhere, making
compilation not possible. Marking MicroPython's symbol as weak would
make it compile and make it possible to create and start the MicroPython
task anyway with the right FreeRTOS task creation incantation, but it is
probably easier to just rename the initialisation function into
something else that can be accessed from outside.
When MicroPython is embedded as an ESP-IDF component, the
MICROPY_ESP_IDF_ENTRY definition can be set to indicate the new entry
point function name. The new function name prototype should still be
defined in external code to let linking succeed.
Also, the NLR failure callback is marked as weak to give the chance of
handling such error in a more controlled fashion rather than trigger an
unconditional board restart.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
This commit introduces two extra CMake variables, MICROPY_DEF_COMPONENT
and MICROPY_COMPILE_COMPONENT, that make it easier to integrate
MicroPython as a custom ESP-IDF component.
Whilst there is no official MicroPython component available for ESP-IDF,
integration can be achieved with some minor CMake scripting outside the
MicroPython tree - except for customisation of compilation defines and
build flags, which is what this commit tries to provide.
Compilation defines customisation is especially important for
MicroPython configuration, as it is not possible to inject a value for
MP_CONFIGFILE otherwise. This means that unless MicroPython itself is
forked first to edit ports/esp32/mpconfigport.h, it is not possible to
perform any meaningful configuration of the interpreter/runtime when
included as a component.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
Raises a value error in that case, which happens after a timer was created
but not initialized, or after calling `timer.deinit()`.
Fixes issue #17033.
Signed-off-by: robert-hh <robert@hammelrath.com>
This commit removes the explicit dependency on the vendored tinyusb
version for the ESP32S2 and ESP32S3 boards.
Tinyusb is still available to MicroPython through a dependency on the
`espressif/esp_tinyusb` ESP-IDF component, which in turn depends on
the `espressif/tinyusb` component itself.
Signed-off-by: Alessandro Gatti <a.gatti@frob.it>
This groups non-decimal values by fours, such as bbb_bbbb_bbbb. It also
supports `{:_d}` to use underscore for decimal numbers (grouped in threes).
Use of incorrect ":,b" is not diagnosed.
Thanks to @dpgeorge for the suggestion to reduce code size.
Signed-off-by: Jeff Epler <jepler@gmail.com>
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>
This commit adds a new network multi-test which sends a burst of UDP
packets from the client, and the server doesn't recv them until they have
all been sent.
Signed-off-by: Damien George <damien@micropython.org>
The bare-metal lwIP socket interface is currently quite limited when used
for UDP streams, because it only allows one outstanding incoming UDP
packet. If one UDP packet is waiting to be socket.recv'd and another one
comes along, then the second one is simply dropped.
This commit implements a queue for incoming UDP and raw packets. The queue
depth is fixed at compile time, and is currently 4.
This allows better use of UDP connections, eg more efficient. It also
makes DTLS work better which sometimes has a queue of UDP packets (eg
during the connection phase).
Signed-off-by: Damien George <damien@micropython.org>
