To have the flexibilty in application; to use any of the section
(data/bss/heap) without updating linker script in every use case,
following decisions are made:
1. Fixed size and small sections moved to SRAM2 (32K)
Vectors
Crash data
Remaining section - RW / ZI
2. Large memory space should be used for variable sections
RW/ZI
Heap - (Minimum - 0x12000)
Stack - At bottom
Please note the heap address of the both the banks must not be contigious else
GCC considers it to be single memory bank and does allocation across the banks,
which might result into hard-fault
To have the flexibilty in application; to use any of the section
(data/bss/heap) without updating linker script in every use case,
following decisions are made:
1. Fixed size and small sections moved to SRAM2 (32K)
Vectors
Crash data
Stack
Remaining section - Heap memory
2. Large memory space should be used for variable sections
Data
BSS
Heap - Remaining section
Heap is moved to the end of both sections as GCC allocates till 4K boundary,
if end of heap is not aligned to 4K, that chunk of memory will go unutilized
This test case uses `hidapi` -- a cross-platform Python module.
To keep the initial Mbed setup as simple as possible, the `hidapi`
module is skipped on Linux hosts because of its external dependancies
for this platform.
The module can be easily installed following instructions from the
README file.
The test case is skipped if the host machine lacks `hidapi` module.
Wait for the host driver to finish setup before sending any HID reports
from the device.
USBHID::wait_ready() blocks until the device reaches 'configured' state,
but the state of the host HID driver remains unknown to the device.
To successfully use pyusb on Windows hosts, a Zadig configuration has to
be performed. Since config for basic tests has already been provided,
use it again.
Assert failure took a critical section before calling `mbed_error`.
There's no need to take a critical section on assert failure -
mbed_error does not do this, and is designed to operate from normal
contexts.
Avoiding the critical section will improve the chances of console
initialisation due to assert failure working nicely.
The length calculation in UARTSerial::write_unbuffered was wrong,
meaning it would truncate output data to half length.
This would show up if `platform.stdio-buffered-serial` was configured to
true, `platform.stdio-convert-newlines` was still false - `mbed_error`
crashes would be garbled.
This wasn't usually spotted because applications generally have both
settings false or both true, and if newline conversion is on, then
`mbed_error_puts` writes 1 character at a time to FileHandle::write,
avoiding the length error.
Using malloc will require us to add stdlib.h somewhere in the path for
the application. Maybe the CI apps are adding stdlib.h and that's why
the code would have worked. In a custom app, it can happen that the
header is not included. Using new avoids the need to add stdlib.h
anywhere and it is more in line with C++.
For default construction of the driver, it would be necessary to define
the pins. However as we don't know the correct pins, until/unless user
sets them up for us. So it makes sense to mark the pins as NC. It would
be easier to debug if the pins are set to NC if user forgot to set them
up.
Cruz Monrreal
Deepika
Shrikant Tudavekar
Marcus Chang
Brian Daniels
Filip Jagodzinski
Kevin Bracey
Jimmy Brisson
Teppo Järvelin
Martin Kojtal
Hasnain Virk
Mirela Chirica
Russ Butler