Fix endianness for cortex_r4 and cortex_r5 when inserting software
breakpoints. Because the cortex_a target is used by the cortex_r
architecture and some chips start in BE by default (e.g. TMS570)
Change-Id: I68b7fe7c4604de67fee2e64fff0fad2691659a58
Signed-off-by: Lucien Dufour <lucien.buchmann@dufour.aero>
Reviewed-on: https://review.openocd.org/c/openocd/+/8909
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
When SMP is enabled, gdb will always use the first target in the SMP
group. That doesn't work when that first target is unavailable, but
others in the SMP group are still available.
For cases where gdb expects an operation to affect the entire group (run
control, memory access), find the first available target in an SMP group
and use that.
Imported from
https://github.com/riscv-collab/riscv-openocd/pull/767
Change-Id: I4bed600da3ac0fdfe4287d8fdd090a58452db501
Signed-off-by: Tim Newsome <tim@sifive.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8912
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
The numeric value '3' for the ASCII character CTRL-C is not
immediately readable, even if the lines that follow explicitly
mention CTRL-C.
Use the same macro present in `telnet_server.c` to replace the
numeric value.
Change-Id: Iaf4296b1f0e384f8122d8a4875cad17e8ddaf66a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8922
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins
GDB server debug logging eat Ctrl-C when gdb user issues interrupt
in time of communication between OpenOCD and gdb.
E.g. Ctrl-C after `next` gdb command taking many
gdb remote protocol $vCont;s (steps)
Change-Id: I4a65446a9bb25a28e50566607b3dec116fa7d2cd
Suggested-by: Tim Newsome <tim@sifive.com>
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8920
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Since the initial commit f444c57bf2 ("arm_cti: add cti command
group") the helper handle_cti_dump() return JIM error codes.
Fix it by returning standard OpenOCD error codes.
Change-Id: Ia36b82083d213aff90fe22fcfe7fbe26172806a3
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8904
Tested-by: jenkins
With all OpenOCD commands converted to COMMAND_HANDLER, we can
drop the management of jim_handler commands.
Drop also from documentation the subsection on Jim Command
Registration.
Change-Id: I4d13abc7e384e64ecb155cb40bbbd52bb79ec672
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8901
Tested-by: jenkins
Rewrite only the command, but still use the old jimtcl specific
code in arm_tpiu_swo_configure(), shared with command 'create'.
Change-Id: If2258f048403f54faf229e602d9b395b71894f97
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8900
Tested-by: jenkins
Rewrite only the command, but still use the old jimtcl specific
code in arm_tpiu_swo_configure(), shared with commands 'configure'
and 'cget'.
Change-Id: I39c69b1cdc23f7b5f875df3e15be987c715b0bcf
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8899
Tested-by: jenkins
Rewrite only the command, but still use the old jimtcl specific
code in cti_configure().
Change-Id: I29fb952a7c8148416b301cbf78b6e342979af7d3
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8898
Tested-by: jenkins
Rewrite only the command, but still use the old jimtcl specific
code in dap_configure().
Change-Id: I3360884616367aae52f5b32247d9864000c53fdc
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8897
Tested-by: jenkins
On ARMv7a/r the single-step is implemented through a HW breakpoint
that hits instructions at any address except the address of the
current instruction.
The method above fails in case of an infinite loop coded by a
single instruction that jumps on itself; in such case, the same
instruction (at the same address) is executed over and over and
the breakpoint never hits. In current code this case is wrongly
considered as an error.
Reduce the timeout while waiting for the HW breakpoint being hit,
then halt.
The jump on itself would be executed several times before the
timeout and the halt, but this is not an issue. There are few
"pathological" instructions in ARMv7a/r that jumps on itself and
that can have side effects if executed more than once. They are
listed in the code. We do not consider these as real use cases
generated by a compiler.
Document the method in the code.
Report that the single-step function is not properly managing the
HW breakpoints if it exits on error. To be fixed in the future.
Change-Id: I9641a4a3e2f68b83897ccf3a12d3c34e98a7805c
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8871
Tested-by: jenkins
This patch ports FIELD_{GET,PREP,FIT} macros and related macro
from FreeBSD, referenced file:
- `src/tree/sys/compat/linuxkpi/common/include/linux/bitfield.h`
Checkpatch-ignore: MACRO_ARG_REUSE
Change-Id: I6fdf4514d3f95d62fadf7654409a4878d470a600
Signed-off-by: Walter Ji <walter.ji@oss.cipunited.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8171
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
`target create` calls can get quite long and an indication what is the
option that caused the error can be helpful.
Also, there can be multiple `-event` options for different events,
therefore indicating which one is it is also helpful.
Change-Id: I5ea61437ca9705e790ed8343183883a3fdfebc80
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8861
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
There are a couple of issues with the usage string for `target create`,
namely:
* `-chain-position` is allowed to be not the first option.
* `-chain-position` should be ommited alltogether on ARM targets when
DAP is specified.
Before the patch:
```
> openocd -c 'target create name testee'
...
target create name type '-chain-position' name [options ...]
```
After the patch:
```
> openocd -c 'target create name testee'
...
-chain-position ?name? required when creating target
> openocd -c 'target create'
...
target create name type [options ...]
```
Change-Id: Ia21a99ce6a4086e2e0676f5ef4685da3514a4f69
Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8860
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins
Function target_configure() when processing a "cget" command needs
to print the result to the console. Currently the result is only printed
when an error occurs. To fix this, move the command print statement from
the error handling section to the common code section.
The code was tested by executing a "$target_name cget -dap" command
and reviewing the result.
Change-Id: Iff1999de8c8e9a837055ba95714137aa03e68d4b
Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com>
Fixes: 61890e3dc3 ("target: rewrite function target_configure() as COMMAND_HELPER")
Reviewed-on: https://review.openocd.org/c/openocd/+/8870
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Commit 236208a5ff ("transport: use a bitmask for the transport")
has an incorrect C statement in place of a return.
The code is working thanks to the previous condition never true.
The issue has been detected by clang scan-build in OpenOCD ACI
since the missing return can make the following statement
dereferencing a NULL pointer.
Fix it!
Change-Id: I3bbe04d99ad9b1288f55ba3c45e2e487aef9ae40
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: 236208a5ff ("transport: use a bitmask for the transport")
Reviewed-on: https://review.openocd.org/c/openocd/+/8868
Tested-by: jenkins
The field 'outer_cache' is always initialized and used as a
pointer to 'struct armv7a_l2x_cache'.
There is no reason for using type 'void *' for it.
Change the type of 'outer_cache'.
Drop the useless cast while reading 'outer_cache'.
Change-Id: Iaea9d02e247da26e230f887c85fbf8e9d7be34d5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8858
Tested-by: jenkins
The command was already tagged as deprecated in 2015 with commit
0df5577282 ("armv7a: remove l1 flush all data handler") but has
never been removed.
An equivalent command 'cache l2x conf' was introduced at the same
time in commit cd440bd32a ("add armv7a_cache handlers").
Drop it and deprecate it.
Replace the old command in the Tcl script.
Change-Id: Ie24eccc99a78786903704d10ee1d9f6c924529b5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8857
Tested-by: jenkins
Currently, we are using offset of xListEnd.pxPrevious in List_t for
list_next_offset and offset of pxPrevious in ListItem_t for
list_elem_next_offset. This is confusing. Fix this.
As the related lists are doubly linked lists, only iteration order is
changed without breaking functionality.
Also document those offsets.
Change-Id: I8beacc235ee781ab4e3b415fccad7b72ec55b098
Signed-off-by: Chien Wong <m@xv97.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8833
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Drop the NULL sentinel at the end of the array and use ARRAY_SIZE()
to bound the loops.
Adapt rtos_try_next() to use ARRAY_SIZE().
While there:
- change to bool the return type of rtos_try_next();
- move rtos_try_next() to avoid the forward declaration.
Change-Id: I1bee11db943b670789e62f1bebe2509bbef451a0
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8855
Tested-by: jenkins
To simplify the caller of rtos_create(), convert the code from
jimtcl oriented to OpenOCD commands.
While there, fix inconsistencies in almost every rtos create()
method and reset rtos_auto_detect to better cooperate on run-time
rtos configuration.
Change-Id: I59c443aaed77a48174facdfc86db75d6b28c8480
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8830
Tested-by: jenkins
Commit 34ec5536c0 ("stlink: deprecate HLA support") makes hard
to use the still functional HLA transport with the stlink listed
in board config files.
Now that the prefixes 'hla_' and 'dapdirect_' has been dropped
from the transport name, allow overriding the transport by using
the 'stlink-hla' script in front of the board file, e.g.:
openocd -f interface/stlink-hla.cfg -f board/st_nucleo_f4.cfg
Revert the documentation changes of the change above.
Improve the documentation to explain how to use the compatibility
HLA mode.
Improve the error message in stlink driver to guide the user to
update the stlink firmware and to use the compatibility HLA mode.
Change-Id: I5d0bc7954511692cebe024bda2aaa72767b97681
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8679
Tested-by: jenkins
The transport used on adapter's wires is either 'jtag' or 'swd'
but, depending on the adapter, in the command 'transport select'
we have to use either 'jtag' or 'swd' or the similar 'hla_jtag',
'hla_swd', 'dapdirect_jtag' or 'dapdirect_swd'.
This becomes cumbersome when we just want to change adapter and we
get forced to modify the 'transport select' command too.
There is no reason for an adapter driver to support two of the
similar transports. In fact 'dapdirect' one is a superset of the
'hla', and the native 'jtag' or 'swd' is a superset of the
'dapdirect' one.
While the adapter could support more than one similar transports,
its adapter driver should only support the most complete of these
similar transports.
Modify the 'transport select' code to accept 'jtag' or 'swd' for
the 'dapdirect' and the 'hla' adapters too.
Issue a deprecated message for the old 'dapdirect' and 'hla'
transport names.
In command 'transport list', print only the transport names that
can be selected through 'transport select' skipping information
about 'dapdirect' and 'hla' versions and avoid duplicated entries.
This improvement was listed in the TODO file. Update it!
Change-Id: I626b50e7a94c141c042eab388cd1ffe77eb864c2
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8677
Tested-by: jenkins
While this operation has no real interest so far, it will be used
later to avoid listing twice protocols with the same name.
Change-Id: I59f3634830f94dc992d28863cf29d5d869726918
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8685
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
Verify that it contains only valid transports.
While JTAG and SWD are the more permissive transports, the
respective 'dapdirect' versions are slightly limited, and the
respective 'hla' versions are even more limited.
A driver should not provide two version of the same transport.
Verify that only one JTAG and only one SWD transport is present.
Verify that the preferred transport is valid too.
Change-Id: Iace2f881dd65fc763e81b33e6a7113961a7008af
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8676
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
Reviewed-by: Jan Matyas <jan.matyas@codasip.com>
In every driver, replace the array of strings with a bitmask that
lists the supported transports.
Add an extra field to carry the former first listed transport as a
"preferred" transport. It would be used as default when no command
'transport select' is used. This keeps backward compatibility with
scripts that do not define the transport, relying on such default.
Change-Id: I4976583f1a38fdcc1f85045023dc7c629001f743
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8675
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
Move the transport's names in a local array in the transport
framework.
Replace the string struct transport::name, that identifies the
transport, with a bitmask where each bit corresponds to one of the
available transports.
Change-Id: I6bdf7264d5979c355299f63fcf80bf54dcd95cee
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8674
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
Historically, if the user does not specify a transport, OpenOCD
select automatically the first transport listed in the adapter
driver.
This auto-selection can behave differently by changing adapter,
so the transport should be enforced in the configuration file.
Deprecate the auto-selection and print a warning message when a
transport gets auto-selected.
There are two cases:
- adapter offers one transport only. The code early auto-selects
the transport but does not print anything. If later the user
selects the transport then no deprecation will be printed during
'transport init';
- user runs 'transport select', e.g. in 'swj-dp' script, and this
triggers the auto-selection and the deprecated message.
Change-Id: I2e55b9dcc6da77ca937978fbfb36bc365b803f0d
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8692
Reviewed-by: Jan Matyas <jan.matyas@codasip.com>
Reviewed-by: zapb <dev@zapb.de>
Tested-by: jenkins
Modify the command 'adapter list' to output the list of transports
supported by each adapter driver.
Drop the line number, as there is no real interest on it.
Format the output as a TCL dictionary indexed by the adapter name
and containing the transports in a TCL list. E.g:
dummy { jtag }
ftdi { jtag swd }
This format is easily handled by TCL scripts, e.g.:
dict get [adapter list] ftdi
Document the command output.
Change-Id: I69f73b71da2f1756866a63bc2c0ba33459a29063
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8691
Tested-by: jenkins
Comments and output strings still reference the term 'interface',
while 'adapter driver' should be used.
While there, drop the useless test if CMD_ARGV[0] is an empty
string, as this is not possible when CMD_ARGC > 0.
Change-Id: I7b46b5dd3cec53d8b5b7559d941ee9ae3bd1d89b
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8808
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
The file stm8.h is only included by stm8.c and provides some basic
declaration that can be simply part of the C file.
Drop the file stm8.h and move its content in stm8.c
Replace the macro 'STM8_NUM_CORE_REGS' with the existing macro
'STM8_NUM_REGS'.
While there:
- drop the useless include of "hello.h".
Change-Id: Iecd1a27f0630cdbbfd51033d34aa3d468aa63464
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8856
Tested-by: jenkins
Reviewed-by: zapb <dev@zapb.de>
Use proper format specifiers for uint16_t and uint32_t arguments.
Use LOG_TARGET_DEBUG instead of target->cmd_name as a parameter.
Use command_print() in command handler.
Drop dots and new lines at end of messages.
Change-Id: I37c7d3680a352210b1d7e69f2c9b4ba0efe6ec15
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reported-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8809
Tested-by: jenkins
Calling ROM API set_bootrom_stack() function allows ROM API functionality
after OpenOCD halt or reset halt in RISC-V bootloder (emulated ARM code)
Change-Id: I3b255738d61876e876a94207804d9cbe1a7593c2
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Signed-off-by: Luke Wren <luke@raspberrypi.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/8729
Tested-by: jenkins
struct rp2xxx_rom_call_batch_record consists of uint32_t in the host
endianness. Therefore it should be converted to the target endianness
not just simply copied by target_write_buffer().
Concatenate algo code, converted batch records and terminator
to the host resident buffer and copy it at once to the target and
save some adapter turnaround times.
While on it remove typedef rp2xxx_rom_call_batch_record_t
Change-Id: I0e698396003869bee5dde4141d48ddd7d62b3cbc
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8460
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Tested-by: jenkins
The compiler knows what variable remains constant during its lifetime
and there is no need to emphasise constantness.
Change-Id: Ib515f96a3c77afea87274f33b8ccac7a71bfb932
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/8459
Tested-by: jenkins
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
While on it use calloc() instead of malloc()/memset()
Drop useless implementation of rp2040_flash_free_driver_priv()
- exactly same as default_flash_free_driver_priv()
Code style fixes forced by checkpatch
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I5c56c4a7d586c0dcab164a45e8f6200ea9a3bd1d
Reviewed-on: https://review.openocd.org/c/openocd/+/8455
Tested-by: jenkins
Invalidate cache and restore flash XIP mode after erase and also
in error cleanup after write/erase.
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: If7e0c2d75f50f923e6bcbf0aa7bab53fe91b6cc8
Reviewed-on: https://review.openocd.org/c/openocd/+/8454
Tested-by: jenkins
While on it:
Define the names for the fixed flash page/sector sizes and use them
instead of magic values.
Fix memory leak on error return.
Partially backported from former upstream rp2040.c
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: If51c912f4d381ee47756a70f616ecdbee1ac0da7
Reviewed-on: https://review.openocd.org/c/openocd/+/8453
Tested-by: jenkins
Flash erase and write require this guard, unfortunately it is also
partially needed in the flash probe.
Partially backported from former upstream rp2040.c
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: Ie8a240e66c3ed68e08f872cbbfdd90a6d80e1f1e
Reviewed-on: https://review.openocd.org/c/openocd/+/8452
Tested-by: jenkins
Also keep size override by FLASHSIZE Tcl variable possible.
Partially backported from former upstream rp2040.c
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I224c3644450e8b46e35714bfc5436219ffdee563
Reviewed-on: https://review.openocd.org/c/openocd/+/8451
Tested-by: jenkins
