Add support for the `--erase` option on `west flash`, for boards that
specify at least one `--cmd-erase` option.
Signed-off-by: Jordan Yates <jordan@embeint.com>
The Espressif branch of OpenOCD uses the version string "v0.12.0" instead
of "0.12.0" causing the version regex match to fail.
Add an optional 'v' prefix to avoid the failure.
Fixes: #83373
Signed-off-by: Eric Holmberg <eric.holmberg@northriversystems.co.nz>
Previously, rtt start would always fail because the target
(i.e. firmware) had not been started.
```
Info : rtt: Searching for control block 'SEGGER RTT'
Info : rtt: No control block found
```
When the command is 'rtt', run the binary via gdb before calling
'rtt start'. Firmware calls `SEGGER_RTT_Init()` shortly after
init which allows OpenOCD to find the RTT control block.
Similarly, only start the 'rtt server' after calling
'rtt start' to avoid any potential race conditions
internally within OpenOCD.
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
Previously, there was a warning that the to_num() method can
(and probably should) be unbound from OpenOcdBinaryRunner.
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
Moves the telnet client into runners/core.py as well, as this is now shared
between openocd and jlink.
Signed-off-by: Tobias Pisani <mail@topisani.dev>
This was non-trivial, as openocd is a bit weird to work with. Using only
commands passed with '-c' arguments, I couldn't get it to reliably resume
(or just not halt) the target when started. I tried using the 'sleep'
command, and various 'configure -event XX { resume }' events, but nothing
panned out, as it seems to always halt after all `-c` commands have been
run.
To avoid that, this waits for the TCL RPC port to be up, and sends a
resume command there. This works reliably.
Signed-off-by: Tobias Pisani <mail@topisani.dev>
Previous OpenOCD version finding would fail when additional tokens were
prepended to the 'openocd --version' output, as happens with some third-
party OpenOCD repackages (xPack for one).
Fixes: #71955
Signed-off-by: Nick Kraus <nick@nckraus.com>
The gdb-port defines the GDB port and the openocd runner passes
this value to openocd as well as gdb. However, the TI AM62x board
provides multiple ports for each subsystem. For example, systick
appears at 3333, A53 as 3334-3337, R5F as 3338 and M4F as 3339.
If we want to connect to the M4F, we need to add another port which
is different to the to the gdb-port value.
This patch adds an additional argument --gdb-client-port to define
the port which GDB should connect to. It defaults to 3333, identical
to gdb-port.
Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Add explicit capabilities to nios2, nsim and openocd runners to prevent
them from having potentially unwanted ones (for example when new
capabilities are added to Zephyr).
Signed-off-by: Michał Szprejda <mszprejda@antmicro.com>
when run package from another PC, the openocd path
may not the same, so try to use ZEPHYR_BASE when
not aligned
Signed-off-by: Hake Huang <hake.huang@oss.nxp.com>
mypy fails with:
Incompatible types in assignment
(expression has type "None", variable has type "Type[ELFFile]")
this happens because of the code:
try:
from elftools.elf.elffile import ELFFile
except ImportError:
ELFFile = None
ELFFile is set to None to allow later code to check if ELFFile was
imported by checking against None. Instead of setting ELFFile to None,
then update testing code to check if the class has been loaded, as:
if globals().get('ELFFile') is None:
Update the try-catch to `pass`.
Removed ELFFile cargo cult from intel_cyclonev.py and fix pylint
warnings.
Disable duplicate code check. The intel_cyclonev.py is already based
upon openocd.py, so although the duplication detection is correct then
this should not prevent other code changes / fixes to those files from
being applied.
Signed-off-by: Torsten Rasmussen <Torsten.Rasmussen@nordicsemi.no>
the options --verify and --verify-only where only used when flashing elf
files and were ignored for flashing hex files for example when using
openocd runner.
The command to verify the image is now dependant on the presence of
--verify or --verify-only, and the load command is not executed if
--verify-only is provided.
Signed-off-by: Nicolas VINCENT <nicolas.vincent@vossloh.com>
Instead of 'reset halt', default the command to reset and halt the target
to 'reset init'.
OpenOCD strictly requires 'reset init' to halt and prepare the device
before flash programming.
See https://openocd.org/doc/html/Flash-Commands.html chapter
"12.2 Preparing a Target before Flash Programming."
For targets implementing the init hook, it will allows to perform some
target specific init improvements (such as speed flash improvements).
Targets not implementing this hook will end up executing the usual and
previous 'halt'.
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
Provide a way to specify the openocd target script handle used to
describe the core target.
In most target scripts '_TARGETNAME' is used, but it can happen that
'_TARGETNAME.foo' or '_CHIPNAME.bar' is used, specially on SoCs subject
to multicore.
Today, this option is required to enable rtos thread awareness,
using '$_TARGETNAME configure -rtos Zephyr' command, which is generated
by this runner.
It could be useful for similar needs in future.
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
Whenever build/zephyr/runners.yaml does not specify the location of the
openocd executable we seemd to be leaving the executable variable as
None. Instead, use its executable name.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
Add yet another option, --verify-only, to west flash. This option
runs openocd with verify_image but without load_image.
Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
OpenOCD has verify_image command. Add `--verify` option to west flash
so that openocd can do vierfy_image right after load_image.
Signed-off-by: Yasushi SHOJI <yashi@spacecubics.com>
Always add the boards/<arch>/<board>/support directory to the OpenOCD
runner search path if the directory exists.
This simplifies using custom --config <partial-board.cfg> runner
arguments without having to use the full path to the cfg file.
Signed-off-by: Henrik Brix Andersen <hebad@vestas.com>
#38557 Reordered the argument for openocd, but we should not try
to call the init command before giving the board configuration
file stored in 'self.cfg_cmd'.
Move back this variable to it's original position.
Signed-off-by: Julien Massot <julien.massot@iot.bzh>
Commits 49bcc08033 introduced
a possible None object 'gdb_init', do not try to iterate over
this variable if it is a None object.
Signed-off-by: Julien Massot <julien.massot@iot.bzh>
Debugging ESP32 SoCs is not possible if the commands
halt, target and load are enabled by default.
This MR allows this configs to be used as input parameters.
Signed-off-by: Sylvio Alves <sylvio.alves@espressif.com>
When enable thread awareness feature for OpenOCD the search path was
converted to a list. In some environments OPENOCD_DEFAULT_PATH may
not be defined. That create an empty search path list system fails.
This add a test to skips fill search_args with openocd_search values
when list is empty.
Fixes#38272.
Signed-off-by: Gerson Fernando Budke <nandojve@gmail.com>
Add support for specifying a custom OpenOCD command to be used for
resetting and halting a target.
As noted in the OpenOCD documentation, not all targets support the
"reset halt" command for halt-on-reset.
Some targets support a software emulation via the "soft_reset_halt"
command. Other targets may require a custom command (e.g. a command
defined in the target configuration file).
Signed-off-by: Henrik Brix Andersen <henrik@brixandersen.dk>
Zephyr thread awareness is available for openocd but boards don't
have debuggers configuration. This configure OpenOCD runner
automatically to complete configuration.
User still require enable CONFIG_DEBUG_THREAD_INFO=y to visualize
thread debug information.
Signed-off-by: Gerson Fernando Budke <nandojve@gmail.com>
Some targets require no 'halt' to be issued i the gdb server command.
Add a --no-halt option to make it possible.
Keep use of halt as the default case.
Signed-off-by: Erwan Gouriou <erwan.gouriou@linaro.org>
Add support for passing the --config argument to the openocd west runner
multiple times.
This allows for using modular openocd configuration files (e.g. CPU core
configuration in one file, independent of the selected JTAG interface
type).
Signed-off-by: Henrik Brix Andersen <henrik@brixandersen.dk>
There are use cases where --use-elf is given and the hex file does not
exist. Handle them.
Fixes: #31944
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
The RunnerConfig class stores the locations of the Zephyr output files
in various formats (elf, hex, bin). A longstanding issue with the
representation is that these might not exist if the corresponding
Kconfig options are not set. For example, if
CONFIG_BUILD_OUTPUT_BIN=n, there is no .bin file.
Change this so the type system knows these are Optional[str], not str.
Fix the runners that use non-ELF outputs so they check for the
existence of the relevant file before using it, mostly using a new
ZephyrBinaryRunner.ensure_output helper.
I'm not going to bother with checking for the ELF file itself; that's
always there as far as I can tell.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
Commit 3124c02 ("cmake: flash/debug: refactor runner configuration")
changed the way the hex_file and elf_file inputs in the RunnerConfig
object are created. In particular, they are now host-style paths.
This breaks flashing with openocd on Windows, which doesn't handle that
properly. Fix that by "casting" the internal hex_file and elf_file
attributes to POSIX paths.
Fixes: #28138
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
Require all implementations to provide a do_create(), a new
ZephyrBinaryRunner abstract class method, and make create() itself
concrete.
This allows us to enforce common conventions related to individual
runner capabilities as each runner provides to the core via
RunnerCaps.
For now, just enforce that:
- common options related to capabilities are always added, so runners
can't reuse them for different ends
- common options provided for runners which don't support them emit
sensible error messages that should be easy to diagnose and support
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
We can just call super() instead of super(MyClassName, self). The
original extra verbosity is likely due to old habits of mine from
Python 2 which are no longer necessary, but got copy/pasted around.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
Catch ImportError whenever a non-standard module import fails from any
runners that do one. Complain at runtime about it if the user actually
needs the runner.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
In case of flashing (which is typically used with OpenOCD)
we do reset of the target after programming application binary
in the non-volatile memory.
In case of Elf execution we need to reset the target before
loading Elf sections so that we might be sure our target
is in sane & expected state before we start execution.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
In some cases especially for on-going development & debugging of real
application it might be useful to load and run not from flash but
from RAM in that case there's one catch: we cannot reset the board
after loading memory with our app.
That's because:
a) RAM we use might be either cleared on reset or might enter
unpredictable state with portion of previously loaded data
being corrupted.
b) Reset vector most probably still point to ROM/flash and so
our current application won't be executed on reset.
So instead of "run reset" command of OpenOCD we'll use
"resume 0x12345678". Where 0x12345678 is our application's
entry-point (which BTW may very well not match beginning of
the .text section or link base).
Now to extract the entry-point we need our application's zephyr.elf
and since we already have a requirement for Elf we may use it for
loading because OpenOCD does it perfectly fine moreover automatically
detecting loaded image type (binary, hex, Elf etc).
And to use that nice feature just add "--use-elf" to west's
command-line for boards that use "openocd" runner. Like that:
----------->8--------------
west flash --use-elf
----------->8--------------
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
To be used in setups with multiple boards attached to the same one
host we need to have an ability to specify precisely which JTAG probe
to use for a particular board.
This is done by passing "ftdi_serial XXX" command to OpenOCD.
And the serial ("XXX") is supposed to be passed from higher level,
typically via west's options. And exactly for that we add another
"openocd" runner option "--serial=XXX" which sets
a Tcl's "_ZEPHYR_BOARD_SERIAL" variable that later gets passed
to OpenOCD's "ftdi_serial" command.
See more discussions on the matter here:
https://github.com/zephyrproject-rtos/zephyr/issues/22543
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
In some cases the config is already provided for us e.g. when setting
OPENOCD_NRF5_SUBFAMILY. Also it looks like config was always supposed to
be optional (there are checks for it on do_run()).
Signed-off-by: Rihards Skuja <rihardssk@mikrotik.com>
Use the folder the config file is in as an additional search directory
for openocd. This way additional files, like custom debug interface
files can be put in the support folder of a board and the openocd.cfg
can use them.
Signed-off-by: Christian Taedcke <christian.taedcke@lemonbeat.com>
commit 0df4a53107 changed the behavior of
how openocd commands are passed to openocd. We used to add -c to each
command, now the commands are being added without -c causing an error.
This adds "-c" to all commands instead of just passing a list.
Also fixes#20449.
Signed-off-by: Anas Nashif <anas.nashif@intel.com>
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
Let the user specify these arguments as many times as they want:
--cmd-pre-init
--cmd-pre-load
--cmd-post-verify
This makes it a bit easier to handle scripts that need to do a few
things in a row depending on conditions.
Handle --cmd-pre-init and the port arguments properly in the debug
related targets.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
I'd like to support nRF boards in a more generic way, so make it
optional. We can clean up the STM32 files later if there's a lot of
overlap in their .cfg files.
Signed-off-by: Martí Bolívar <marti.bolivar@nordicsemi.no>
