Skip to content

cortexm/pyswd

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

89 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

PYSWD

Is a python module for debugging microcontrollers with SWD using ST-Link/V2 (/V2-1) or V3 debugger.

This package also contain small command line tool.

Goal

Is to create python module for access debugging interface on MCU with SWD interface.

Main purpose of python module is to create automated functional and hardware tests from simple python scripts and without special firmware for microcontroller.

Compatibility

OS

PYSWD will work on Linux, Mac and Windows.

Python

Python 3.7+

Dependencies

  • pyusb - prefer latest version from github, especially on Windows platgorm
  • libusb

Installation:

from downloaded sources

pip3 install .

reinstall (upgrade):

pip3 install --upgrade .

install for editing

pip3 install --editable .

uninstall:

pip3 uninstall pyswd

using make

make test
make install
make editable
make uninstall

Python SWD module documentation

swd.Swd:

swd.Swd(swd_frequency=4000000, logger=None, serial_no='')

Arguments:

  • swd_frequency: SWD communication frequency
  • logger: logging interface (optional)
  • serial_no: serial number of connected USB ST-Link debugger (optional). Serial number can be also part from begin or end, if more devices are detected then it stops with error
>>> import swd
>>> dev = swd.Swd()

ST-Link version

property with ST-Link version

Return:

instance of StlinkVersion

>>> dev.get_version().str
'ST-Link/V2 V2J27S6'

Target voltage

Get target voltage measured by ST-Link

Return:

float target voltage in volts

>>> dev.get_target_voltage()
3.21

ID code

Get MCU ID code

Return:

32bit unsigned with ID code

>>> hex(dev.get_idcode())
'0xbb11477'

Get memory register

get_mem32(address)

Arguments:

  • address: address in memory, must be aligned to 32bits

Return:

32bit unsigned data from memory

>>> hex(dev.get_mem32(0x08000000))
'0x20001000'

Set memory register

set_mem32(address, data)

Arguments:

  • address: address in memory, must be aligned to 32bits
  • data: 32bit unsigned data
>>> dev.set_mem32(0x20000200, 0x12345678)
>>> hex(dev.get_mem32(0x20000200))
'0x12345678'

Read memory

  • read_mem(address, size) - automatically select read access
  • read_mem8(address, size) - read using 8 bit access
  • read_mem16(address, size) - read using 16 bit access
  • read_mem32(address, size) - read using 32 bit access

Arguments:

  • address: address in memory
  • size: number of bytes to read from memory

Return:

iterable of read data

>>> data = dev.read_mem(0x08000000, 16)
>>> ' '.join(['%02x' % d for d in data])
'00 10 00 20 45 00 00 08 41 00 00 08 41 00 00 08'

Write memory

  • write_mem(address, data) - automatically select write access
  • write_mem8(address, data) - write using 8 bit access
  • write_mem16(address, data) - write using 16 bit access
  • write_mem32(address, data) - write using 32 bit access

Arguments:

  • address: address in memory
  • data: list or iterable of bytes whic will be stored into memory
>>> dev.write_mem(0x20000100, [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15])
>>> data = dev.read_mem(0x20000100, 15)
>>> ' '.join(['%02x' % d for d in data])
'01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f'

Fill memory

  • fill_mem(address, pattern, size) - automatically select fill access
  • fill_mem8(address, pattern, size) - fill using 8 bit access
  • fill_mem16(address, pattern, size) - fill using 16 bit access
  • fill_mem32(address, pattern, size) - fill using 32 bit access

Arguments:

  • address: address in memory
  • pattern: list or iterable of bytes whic will be stored into memory
  • size: number of bytes to fill memory
>>> dev.fill_mem(0x20000300, [5, 6, 7], 20)
>>> data = dev.read_mem(0x20000300, 20)
>>> ' '.join(['%02x' % d for d in data])
'05 06 07 05 06 07 05 06 07 05 06 07 05 06 07 05 06 07 05 06'

Read core register

get_reg(register) On CortexM platform this will work only if program is halted

Arguments:

  • register: is numeric coded register (e.g. 0: R0, 1: R1, ...)

Return:

32bit unsigned data

>>> hex(dev.get_reg(1))
'0x0800012e'

Read all core registers

get_reg_all() On CortexM platform this will work only if program is halted

Return:

list of 32bit unsigned data for all registers

>>> dev.get_reg_all()
[0,  0,  16942,  10,  100,  0,  0,  0,  0,  0,  0,  0,  10,  604502776,  134288075,  134284002,  1627389952,  604502776,  0,  0,  67125248]

Write core register

get_reg(register) On CortexM platform this will work only if program is halted

Arguments:

  • register: is numeric coded register (e.g. 0: R0, 1: R1, ...)
  • data: 32bit unsigned data
>>> dev.set_reg(1, 0x12345678)

swd.CortexM:

swd.CortexM(swd)

Arguments:

  • swd: instance of Swd
>>> import swd
>>> dev = swd.Swd()
>>> cm = swd.CortexM(dev)

Read core register

get_reg(register) On CortexM platform this will work only if program is halted

Arguments:

  • register: name of register (e.g.: 'R0', 'R1', 'SP', 'PC', ...)

Return:

32bit unsigned data

>>> hex(cm.get_reg('PC'))
'0x0800012e'

Write core register

set_reg(register) On CortexM platform this will work only if program is halted

Arguments:

  • register: name of register (e.g.: 'R0', 'R1', 'SP', 'PC', ...)
  • data: 32bit unsigned data
>>> cm.set_reg('R2', 0x12345678)

Read all core registers

get_reg_all() On CortexM platform this will work only if program is halted

Return:

dictionary with register name as key and as value 32bit unsigned data for each register

>>> cm.get_reg_all()
{'LR': 134288075,
 'MSP': 604502776,
 'PC': 134284002,
 'PSP': 0,
 'PSR': 1627389952,
 'R0': 0,
 'R1': 0,
 'R10': 0,
 'R11': 0,
 'R12': 10,
 'R2': 16942,
 'R3': 10,
 'R4': 100,
 'R5': 0,
 'R6': 0,
 'R7': 0,
 'R8': 0,
 'R9': 0,
 'SP': 604502776}

Reset

reset()

>>> cm.reset()

Reset and halt

reset_halt()

>>> cm.reset_halt()

Halt core

halt()

>>> cm.halt()

step program

step()

>>> cm.step()

Run in debug mode

run()

>>> cm.run()

Disable debug mode and run

nodebug()

>>> cm.nodebug()

Check if MCU is halted

is_halted()

Return:

True if MCU is halted, or False if is running

>>> cm.is_halted()
True

Python application

Simple tool for access MCU debugging features from command line. Is installed together with python module.

$ pyswd --help

Usage:

pyswd [-h] [-V] [-q] [-d] [-i] [-v] [-f FREQ] [action [action ...]]

positional arguments:

action                actions will be processed sequentially

Optional arguments:

-h, --help            show this help message and exit
-V, --version         show program's version number and exit
-q, --quite           quite output
-d, --debug           increase debug output
-i, --info            increase info output
-v, --verbose         increase verbose output
-f FREQ, --freq FREQ  set SWD frequency
-s SERIAL, --serial SERIAL
                        select ST-Link by serial number (enough is part of serial number: begin or end

List of available actions:

  dump8:{addr}[:{size}]     print content of memory 8 bit register or dump
  dump16:{addr}[:{size}]    print content of memory 16 bit register or dump
  dump32:{addr}[:{size}]    print content of memory 32 bit register or dump
  dump:{addr}[:{size}]      print content of memory 32 bit register or 8 bit dump

  set8:{addr}:{data}[:{data}..]     set 8 bit memory
  set16:{addr}:{data}[:{data}..]    set 16 bit memory
  set32:{addr}:{data}[:{data}..]    set 32 bit memory
  set:{addr}:{data}[:{data}..]      set 32 bit memory register or 8 bit memory area

  fill8:{addr}:{size}:{pattern}     fill memory with 8 bit pattern

  reg:all                   print all core register
  reg:{reg}                 print content of core register
  reg:{reg}:{data}          set core register

  reset[:halt]              reset core or halt after reset
  run[:nodebug]             run core
  step[:{n}]                step core (n-times)
  halt                      halt core

  sleep:{seconds}           sleep (float) - insert delay between commands

(numerical values can be in different formats, like: 42, 0x2a, 0o52, 0b101010, 32K, 1M, ..)

License

Whole project is under MIT license