jtag/drivers/cmsis_dap: move pending request FIFO

Move all pending request FIFO related variables to
the struct cmsis_dap

Pure refactoring, no functional changes.

Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Change-Id: I2cea9c1e265ac102ec0f314767b8b9afdcda9ee9
Reviewed-on: https://review.openocd.org/c/openocd/+/7360
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tools: add disassembler helper for files .inc

We are moving the binary helpers in files .inc in contrib/loaders/
but we have no support to disassemble them for checking their
content, nor documentation to give any hint.

Add a simple script that uses objdump to directly disassemble a
file .inc
Use Cortex-M settings as default, but provide the flexibility to
reuse the script for any other target CPU.

Change-Id: I12e79580f2936b1622fb7231d9a2484a763ba72a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7347
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

doc: usb_adapters: add other adapters

Mainly get from web or from OpenOCD users

Change-Id: Id4d6354bc9b6972622adfacf3549c7c3e0549bc2
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7298
Tested-by: jenkins

doc: usb_adapters: add lsusb dump of few adapters

Real dumps from adapters I have access to.
Serial numbers have been manually edited but are still consistent.

Change-Id: Iee2d16e09615746e29588f91fdfb77eb9fad4daa
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7297
Tested-by: jenkins

doc: usb_adapters: add container for lsusb dump of adapters

When extending an existing adapter driver, developer has to
carefully avoid to break backward compatibility.
But not all developers have access to many adapters for testing.

Prepare a folder aimed at containing 'lsusb -v' dumps of adapters
to be used as reference.
Add documentation and a simple dump script.

Change-Id: Ia3a9132d65ec8869734721f9605cebcebcf33608
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7296
Tested-by: jenkins

flash/nor, contrib/loaders: move numicro M0 write assembly to contrib/loaders

Simply move numicro M0 flash write assembly to contrib/loaders. Also,
modify corresponding Makefile and generated numicro_m0.inc.

To make the path more general, this patch rename the path
"contrib/loaders/flash/numicro_m4" to "contrib/loaders/flash/numicro" as
well.

Signed-off-by: Jian-Hong Pan <chienhung.pan@gmail.com>
Change-Id: I9adea24c2b3c97319a9b015cf50257f3d131dc26
Reviewed-on: https://review.openocd.org/c/openocd/+/7346
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash/nor, contrib/loaders: move numicro M4 write assembly to contrib/loaders

Simply move numicro M4 flash write assembly to contrib/loaders. Also,
add corresponding Makefile and generated numicro_m4.inc.

Signed-off-by: Jian-Hong Pan <chienhung.pan@gmail.com>
Change-Id: I22b8be0a245857335b11bd0b538db415fc34748d
Reviewed-on: https://review.openocd.org/c/openocd/+/7343
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl: add a configuration file for Nuvoton M541 & NUC442/472 series

This patch is picked from the tcl part of OpenOCD-Nuvoton's commit
("flash: supported Nuvoton M4 series. jtag: Used HW reset instead of
auto reset. tcl: added a configuration file for Nuvoton M4 series.") [1]
to support the communication with Nuvoton's Cortex-M4 chips: M541 &
NUC442/472 series.

This patch has been tested with Nuvoton's NuTiny-SDK-NUC472 development
board [2].

The code comes from the commit basically. Jian-Hong Pan tweaked for the
compatibility with current OpenOCD. So, leave the author as Zale Yu.

[1]: https://github.com/OpenNuvoton/OpenOCD-Nuvoton/commit/c2d5b8bfc705
[2]: https://www.nuvoton.com/export/resource-files/UM_NuTiny-SDK-
     NUC472_EN_Rev1.02.pdf

Signed-off-by: Zale Yu <cyyu@nuvoton.com>
Signed-off-by: Jian-Hong Pan <chienhung.pan@gmail.com>
Change-Id: I27ac58dd1c98a76e791a4f1117c31060cf5522e8
Reviewed-on: https://review.openocd.org/c/openocd/+/7330
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash: support Nuvoton M541 & NUC442/472 series

This patch is picked from the flash part of OpenOCD-Nuvoton's commit
("flash: supported Nuvoton M4 series. jtag: Used HW reset instead of
auto reset. tcl: added a configuration file for Nuvoton M4 series.") [1]
to support flashing Nuvoton's Cortex-M4 chips: M541 & NUC442/472 series.

The code comes from the commit basically. Jian-Hong Pan tweaked for the
compatibility with current OpenOCD. So, leave the author as Zale Yu.

[1]: https://github.com/OpenNuvoton/OpenOCD-Nuvoton/commit/c2d5b8bfc705

Signed-off-by: Zale Yu <cyyu@nuvoton.com>
Signed-off-by: Jian-Hong Pan <chienhung.pan@gmail.com>
Change-Id: I9dc69eccb851df14c1b0ce2f619d7b3da0aa92aa
Reviewed-on: https://review.openocd.org/c/openocd/+/7329
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash/nor/numicro: reorder the parts list

This patch is picked from the flash part of OpenOCD-Nuvoton's commit
("flash: supported Nuvoton M4 series. jtag: Used HW reset instead of
auto reset. tcl: added a configuration file for Nuvoton M4 series.") [1]
It reorders Nuvoton MCU part list and spreads out the way to define the
banks' base address & size of the parts.

The code comes from the commit basically. Jian-Hong Pan tweaked for the
compatibility with current OpenOCD. So, leave the author as Zale Yu.

[1]: https://github.com/OpenNuvoton/OpenOCD-Nuvoton/commit/c2d5b8bfc705

Signed-off-by: Zale Yu <cyyu@nuvoton.com>
Signed-off-by: Jian-Hong Pan <chienhung.pan@gmail.com>
Change-Id: I1f21f54dfdf53e5b8ea04d803347d1dbc8c321a0
Reviewed-on: https://review.openocd.org/c/openocd/+/7339
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/armv7m: Rename xPSR to xpsr

The org.gnu.gdb.arm.m-system GDB feature defines the name in lowercase
letters.[1] Not adhering to the definition can cause issues with tools
interacting with the GDB which expect the correct casing.

[1]: https://sourceware.org/gdb/onlinedocs/gdb/ARM-Features.html

Change-Id: I0b6584a78f86b053947d79686baad5dac3ec4a00
Signed-off-by: Boris-Chengbiao Zhou <bobo1239@web.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/7292
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash/stm32l4x: avoid multiple assignments

Change-Id: I6d8e0fbfa7e05f26295fc22733c65c11f7460b51
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@st.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7282
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins

adapter: Improve three adapter-related prints

- Do not warn the user about adapter speed not being set
  if the selected adapter does not support configurable speed.
  This would be confusing to users.

- Minor reword in the "adapter speed is not selected" warning.

- Print the adapter name in the "adapter doesn't support
  configurable speed" log entry

- Another minor reword: autoselect --> autoselecting

Signed-off-by: Jan Matyas <matyas@codasip.com>
Change-Id: I4c198b547544fd8199921f092153562c5f159134
Reviewed-on: https://review.openocd.org/c/openocd/+/7269
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

riscv: drop deprecated command 'riscv test_sba_config_reg'

Change-Id: I51c1b1cb3de8cb86ee38280fa3f035f6f7a63dbc
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7272
Reviewed-by: Jan Matyas <matyas@codasip.com>
Reviewed-by: Tim Newsome <tim@sifive.com>
Tested-by: jenkins

riscv: drop deprecated command 'riscv set_prefer_sba'

Change-Id: I546efe4e1a6b673b26cfb4a74b5c3809fecda49c
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7271
Tested-by: jenkins
Reviewed-by: Jan Matyas <matyas@codasip.com>

tcl/target: remove rp2040-core0.cfg

rp2040-core0.cfg configuration file was intended for a special adapter
which selects a SWD multidrop target on its own. This means
that rp2040-core0.cfg is totally unusable with a standard SWD
adapter. The file was marked as deprecated in 0.12 release.

The reworked rp2040.cfg can be restricted to use just one core:
  openocd ... -c 'set USE_CORE 0' -f target/rp2040.cfg

Remove the obsoleted config.

Change-Id: Id886471622bb4a8cb83f5c4c3660657407aaaf74
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7326
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl/target: add SMP mode to rp2040.cfg

Add the variable selected configuration for SMP debug with rtos hwthread.

Use SMP by default.

Change-Id: I1c37d91688a3ab58d65c15686737892965711adc
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7242
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash/nor/rp2040: make SPI flash ID detection optional

Do not read ID from SPI flash and suppress autodetection
if non-zero flash bank size is configured.

Change-Id: Idcf9ee6ca17f9fa89964a60da7bf11e47b4af5e7
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7241
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/cortex_m: add SMP support for Cortex-M

Cortex-M support for SMP multicore targets.

This SMP implementation unlike older ones does not act
on the first halted target found. It polls targets
until a SMP group is finished and stores eventual
'halted' events instead of emitting them. As soon as polling
of a group is done, poll proceeds with SMP related tasks.
This approach improves detection of a reason why debug
stopped - a correct reason is detected for all targets,
not only for the first found.
Drawback: SMP target group should be defined in the same
order as the targets were defined.

Obsolete gdb 'J' packet/smp_gdb command core switching is not implemented,
use with rtos hwthread.

Only one core is resumed if debug_execution is requested.

Some ideas taken from Graham Sanderson's [4936]
and src/target/aarch64.c

Added error checking of armv7m_restore_context().

Change-Id: I60f5b79e74b624dc2b5835ff10e38ac2ccb23792
Link: [4936]: target/cortex_m: Add smp support for Cortex M | https://review.openocd.org/c/openocd/+/4936
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7239
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/armv7m: check error in armv7m_restore_context()

Return error if arm.write_core_reg() fails.

Change-Id: Ide8f5aa5958532b202dc9f5e13d3250a706d832d
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7238
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

armv8: Add support of pointer authentication

When pointer authentication is enabled, some upper bits of the link
register (LR[63:VA_SIZE]) are used to store a signature. Therefore, GDB
need to remove the signature to get backtraces.
GDB has support of pointer authentication. When pointer authenticaion is
enabled, GDB requests 8-bytes mask to the target to remove the
signature. mask[63:VA_SIZE] should be all set and mask[VA_SIZE-1:0]
should be all cleared. GDB removes the signature by addr&~mask or
addr|mask.
I added a feature to provide the mask for pointer authentication.

Signed-off-by: Koudai Iwahori <koudai@google.com>
Change-Id: I56fbbf9cc23619b6536ecd326f350c8bf137f322
Reviewed-on: https://review.openocd.org/c/openocd/+/7248
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

src/rtos/eCos : Support per-build configuration specific structure layouts

Extended feature. This is a large patch, but is self-contained to the
eCos support and does not affect any other openocd functionality. It
does NOT affect existing eCos RTOS plugin users where their
applications do not provide the extended symbolic helper
information. If the helper symbols are not available the rtos support
code will behave as before. This "dynamic" functionality is *required*
because eCos does NOT have a fixed/hardwired, known, layout for the
thread descriptor structure. The per-application build eCos
configuration can affect the shape of the thread descriptor structure
(field presence, and hence offsets of subsequent fields) such that
constant values cannot be used to consistently interpret all possible
eCos application configurations. For historical reasons, there is not
yet a consistent namespace for the helper symbols across eCos HALs
hence the support is currently limited to specific architectures
(Cortex-M and ARM/Cortex-A). No new Clang analyser warnings are raised
by this changeset.

Change-Id: Ib3a36877326eeb56595cbca55e21b9e59a59c98a
Signed-off-by: James G. Smith <jsmith@rallysmith.co.uk>
Reviewed-on: https://review.openocd.org/c/openocd/+/6275
Reviewed-by: Alex Schuilenburg <alex.schuilenburg@gmail.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins

flash/nor/spi: Add issi is25lq040b to device list

The HiFive Inventor uses this flash chip, so adding it will allow for
openocd to be used to program it. These values were taken from the
chip's documentation.

Signed-off-by: Fawaz Tirmizi <fawaz@rivosinc.com>
Change-Id: I15c9d35f99d4500f73134cdc2d1b9ab6279b491c
Reviewed-on: https://review.openocd.org/c/openocd/+/7135
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/xtensa: remove needless target_was_examined check

In any case flag will be set as examined.

Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: I5177ee656f726a807269e2f4725223f50e49e855
Reviewed-on: https://review.openocd.org/c/openocd/+/7231
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash/nor/avrf: add ATmega32U4 support

Add new chip info and tcl target

Change-Id: Ib9d33d1b145a8659857b7a6cc9c5acba047f41d1
Signed-off-by: Kyle Schwarz <zeranoe@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7081
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

openocd: revert workarounds for 'expr' syntax change

With OpenOCD v0.12.0 released, drop the workarounds for 'expr'
syntax change by reverting:
- commit 320043c054dc ("openocd: fix for polling during "expr"
  computation");
- commit c7eaaf620488 ("openocd: prepare for jimtcl 0.81 'expr'
  syntax change").

Replace the call to target_call_timer_callbacks_now() with call
to target_call_timer_callbacks().

Change-Id: Iae5afc50e3f688e11176a52648efc9a6577a9a11
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7405
Tested-by: jenkins

nds32: drop it, together with aice adapter driver

The target nds32 and its companion adapter aice have not received
any real improvement since 2013.
It has been hard to keep them aligned during the evolution of
OpenOCD code, with no way for maintainers to really check if they
are still working.
No real documentation is present for them in OpenOCD.
The nds32 code triggers ~50 errors/warnings with scan-build.

The arch nds32 has been dropped from Linux kernel v5.18-rc1.

For all the reasons above, this code has been deprecated with
commit 2e5df83de7f2 ("nds32: deprecate it, together with aice
adapter driver") and tagged to be dropped before v0.13.0.

Let it r.i.p. in OpenOCD git history.

While there, drop from checkpatch list the camelcase symbols that
where only used in this code.

Change-Id: Ide52a217f2228e9da2f1cc5036c48f3536f26952
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7382
Tested-by: jenkins

Deprecate libjaylink Git submodule

libjaylink was integrated into OpenOCD as a Git submodule, because at
that time there was no official release and no packages for it.
Today there are libjaylink packages for most popular distributions [1].

Removing libjaylink from OpenOCD reduces build complexity in both
projects and makes them more flexible with respect to the build system,
for example.

Disable the libjaylink submodule by default and announce it as
deprecated feature that will be removed in the future. This gives
package maintainers time to package libjaylink if not already done.

[1] https://repology.org/project/libjaylink/versions

Change-Id: I6166ba4757aee5c89a0506de867072f58fa5ec4b
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/7129
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

Restore normal development cycle

Signed-off-by: Paul Fertser <fercerpav@gmail.com>

The openocd-0.12.0 release

Signed-off-by: Paul Fertser <fercerpav@gmail.com>

configure.ac: fix check for jimtcl submodule

The file configure.ac checks if jimtcl submodule is present by
looking for the file jimtcl/configure.ac .
But jimtcl has switched to its own build system in 2011 and thus
dropped configure.ac . This cause a build issue on OpenOCD that
was incorrectly fixed by adding a dummy configure.ac in jimtcl.

Fix OpenOCD configure.ac to look for the correct file
jimtcl/configure .
After this fix, another fix would be proposed to jimtcl to drop
its useless dummy file configure.ac .

Change-Id: I705f72d83e374c8dd700baaa0c0bbe041f153605
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reported-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7437
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>

jtag: esp_usb_jtag: remove macro __packed

In FreeBSB 13.0 the build fails due to redefined macro __packed.

src/jtag/drivers/esp_usb_jtag.c:19:9: error: '__packed' macro
    redefined [-Werror,-Wmacro-redefined]
#define __packed __attribute__((packed))
        ^
/usr/include/sys/cdefs.h:223:9: note: previous definition is here
#define __packed        __attribute__((__packed__))
        ^
1 error generated.

Drop the macro and align the code with the other files in OpenOCD
project, where the attribute is directly applied without using a
macro.

Change-Id: I89ae943e77036206d40d4d54172cd4a73e76e5c5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reported-by: Wojciech Puchar <wojtek@puchar.net>
Reviewed-on: https://review.openocd.org/c/openocd/+/7435
Tested-by: jenkins
Reviewed-by: Erhan Kurubas <erhan.kurubas@espressif.com>

rtt: fix corner-cases of finding control block

This patch fixes two corner-cases of finding RTT control block.

The first one is when there was a partial match (even single byte) at
the end of loaded buffer (uint8_t buf[1024]), but this was not part of
full match. In that case `cb_offset` was not updated correctly and the
returned `*address` was lower by the legth of the partial match. In case
of searched 'SEGGER RTT' (the default control block ID) string, it was
enough to match `buf[1023] == 'S'`, which is quite likely to happen, and
the `*address` was offset by 1 (e.g. it was 0x20000fff instead of
0x20010000).

Updating (or even maintaining) `cb_offset` is not needed, as start
address of control block can be calculated based on memory address that
was loaded into `uint8_t buf[1024]`, the offset within this buffer and
the length of expected string.

The second issue is when control block is prepended with a byte that
matches first ID character, e.g. there is `SEGGER RTT` control block ID
is prepended by another `S`, making memory contents be `SSEGGER RTT`. In
that case there was no match found.

Fix that issue by making sure that tested byte is always compared with
first byte of expected control block ID.

While at it, change names of local variables to better describe their
meaning.

Signed-off-by: Marcin Niestroj <m.niestroj@emb.dev>
Change-Id: I12aa6e202bf12bedcbb888ab595751a2a2518a24
Reviewed-on: https://review.openocd.org/c/openocd/+/7429
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

Restore +dev suffix

Signed-off-by: Paul Fertser <fercerpav@gmail.com>

The openocd-0.12.0-rc3 release candidate

Change-Id: Id7ddf232593e1aa7cb36f2b30fe832ebf79c1535
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>

cortex_m: handle armv8m cores without security extension

Cores armv8m, e.g. Cortex-M33, can be instantiated without the
optional Security Extension.
In this case, the secure registers are not present and when GDB
try accessing them it triggers a set of errors.

For armv8m cores without security extension, don't provide to GDB
the description of the secure registers.

Change-Id: I254478a4cf883e85b786df3f62c726b2f40d88d9
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reported-by: Torbjörn SVENSSON <torbjorn.svensson@foss.st.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7402
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

jtag: fix build with configure --enable-verbose

With flag --enable-verbose, configure enables compiling some
conditional code that with new gcc triggers an error:
error: '%04x' directive output may be truncated writing
between 4 and 8 bytes into a region of size 5
[-Werror=format-truncation=]

Extend the buffer to contain the full 8 bytes of %04x on a 'int'
and change the limit in snprintf.
Skip the intermediate buffer 's[4]'.
Align the code to the coding style.

Change-Id: Ifc8a6e4686555578a7355a1f6049471fd5e31913
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reported-by: Karl Hammar <karl@aspodata.se>
Reported-by: Tommy Murphy <tommy_murphy@hotmail.com>
Fixes: https://sourceforge.net/p/openocd/tickets/376/
Reviewed-on: https://review.openocd.org/c/openocd/+/7403
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

at91samd: wait for nvm ready

Flashing a SAMD21J17D was failing during NVM erase.  The samd21
datasheet specifies that one cause of error conditions is executing an
NVM command while the previous command is still running.  The solution
is to wait for INTFLAG.READY after a command is issued.

SAMD21J17A was not exhibiting this problem.  Perhaps the later silicon
revision has slower NVM erase times.

Signed-off-by: Dan Stahlke <dan@stahlke.org>
Change-Id: I19745dae4d3fc6e3a7611dcac628e067cb41e0f0
Reviewed-on: https://review.openocd.org/c/openocd/+/7391
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

driver: vdebug: fix mode of cmd 'vdebug mem_path'

The command 'vdebug mem_path' is reported in the documentation as
'{Config Command}', but the code sets mode = COMMAND_ANY.
The code of the commands sets some value that is only used during
the init phase, so the documentation is correct.

Change mode of command 'vdebug mem_path' to COMMAND_CONFIG.

Change-Id: Icb940fe382cbc75015273b35dcc8a88fc2a7d0ac
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7395
Tested-by: jenkins
Reviewed-by: Jacek Wuwer <jacekmw8@gmail.com>

target: fix assert in 'monitor profile' on constant PC

When target is stopped in WFI/WFE or is in an infinite loop, the
sampled PC will always return the same value.
Command 'profile' requires that distance between min and max PC
should be at least 2, which is not the case for constant PC, and
incorrectly enforces the check through as assert().

Move the code that reads the optional parameters 'start' and 'end'
and check the gap 'end - start' before running the profile.
For self-computed min and max, increase max (or decrease min) to
match the required constraint.
Drop the assert().

Change-Id: I2be8df8568ce8c889923888c492e4f7ce354b16b
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: https://sourceforge.net/p/openocd/tickets/370/
Reviewed-on: https://review.openocd.org/c/openocd/+/7400
Tested-by: jenkins

target: fix unsigned computation in 'monitor profile'

The implementation of command 'monitor profile' has few
issues:
- the address_space is a signed int, so cannot wrap-around on
  space over INT_MAX;
- max address is incremented without check for overflow;
- assert() used on errors instead of returning error codes;
- only handles 32 bits PC;
- output file created and left empty on error.

This patch fixes the first two issues, as a wider fix would be too
invasive and should be postponed in a following series.

Change-Id: Id8ead3f6db0fd5730682a0d1638f11836d06a632
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: https://sourceforge.net/p/openocd/tickets/370/
Reviewed-on: https://review.openocd.org/c/openocd/+/7394
Tested-by: jenkins

target: cortex_a: fix clang error core.CallAndMessage

Clang complains about the variable 'orig_dfsr' that can be used
uninitialized both in cortex_a_read_cpu_memory() and in
cortex_a_write_cpu_memory().

The issue is caused by an incorrect error path that used to jump
through 'goto out'. The code after the label 'out' is specific to
handle the case of an error during memory R/W; it is incorrect to
jump there to handle an error during the initialization that
precedes the memory R/W.

Replace the 'goto out' with 'return retval'.
Remove the label 'out' that is now unused.

Change-Id: Ib4b140221d1c1b63419de109579bde8b63fc2e8c
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7393
Tested-by: jenkins

arm_adi_v5: fix SIGSEGV due to failing re-examine

Commit 35a503b08d14 ("arm_adi_v5: add ap refcount and add get/put
around ap use") modifies the examine functions of mem_ap, cortex_m,
cortex_a and aarch64 by calling dap_put_ap() and then looking again
for the mem-ap and calling dap_get_ap().
This causes an issue if the system is irresponsive and the examine
fails and left the AP pointer to NULL. If the system was already
examined the NULL pointer will cause a SIGSEGV.

Commit b6dad912b85d ("target/cortex_m: prevent segmentation fault
in cortex_m_poll()") proposes a fix for one specific case and only
on cortex_m.

Modify all the examine functions by skipping look-up for the AP if
it was already set in a previous examine; the target's AP is not
supposed to change during runtime.

Remove the partial fix for cortex_m as it is not needed anymore.

Change-Id: I806ec3b1b02fcc76e141c8dd3a65044febbf0a8c
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: 35a503b08d14 ("arm_adi_v5: add ap refcount and add get/put around ap use")
Reviewed-on: https://review.openocd.org/c/openocd/+/7392
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

src/target/mips_m4k.c: set missing flag in set_watchpoint

Without the fix, will see "Can not find free FP Comparator" error log

Change-Id: Id0d91cc02b7055e44d27507f9c05ccd48ff49838
Signed-off-by: Nima Palizban <n.palizban@gmail.com>
Fixes: fb43f1ff4e2f (target: Rework 'set' variable of break-/watchpoints)
Reviewed-on: https://review.openocd.org/c/openocd/+/7389
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

doc: describe tcl port consistently.

One place described the tcl port as 5555, which was changed in 163bd86071a
Reported on IRC.

Change-Id: If740a29443793d6a4d4f8c9db54f0fc8344a6c1c
Signed-off-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-on: https://review.openocd.org/c/openocd/+/7385
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

Fix jim_target_smp for smp rtos target

If multiple targets are specified as -rtos <rtos_type>, the
rtos_update_threads was called only if the last target was specified as
rtos, which is inconsistent with other checks of whether or not smp target
is an rtos one.

Signed-off-by: Evgeniy Naydanov <evgeniy.naydanov@syntacore.com>
Change-Id: Ie52bc6b6c8f841d31b9590fcbc44e985d3cba0eb
Reviewed-on: https://review.openocd.org/c/openocd/+/7244
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

hwthread: Restore current_threadid in hwthread_update_threads

When OpenOCD receives a step-execution command from GDB and the target
is configured as rtos=hwthread, OpenOCD reconstructs the thread-info.
However, OpenOCD does not restore the thread id which is currently
selected by GDB. Due to this issue, OpenOCD sends the information of
wrong thread to GDB after the step execution.
This commit fixes the above issue by adding a code to save/restore the
thread id selected by GDB.

Signed-off-by: Koudai Iwahori <koudai@google.com>
Change-Id: I761a1141c04d48f1290e4f09baa7c7024f86f36a
Reviewed-on: https://review.openocd.org/c/openocd/+/7358
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

hwthread: Add register validity check in get_thread_reg_list

When OpenOCD receives 'g' packet (read general registers) from GDB and
target is configured as rtos=hwthread, hwthread_get_thread_reg_list is
called. However, it does not check if the register valid or not. Due to
this issue, OpenOCD returns invalid register values to GDB.
This commit adds a validity check to hwthread_get_thread_reg_list. If
the register is not valid, it tries to read the register from the
target.

Signed-off-by: Koudai Iwahori <koudai@google.com>
Change-Id: Iad6424b62124271ec411b1dfc044b57dfc460280
Reviewed-on: https://review.openocd.org/c/openocd/+/7357
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jtag: xds110: fix clang error core.StackAddressEscape

Clang ignores that xds110_swd_write_reg() is always called with
bit SWD_CMD_RNW in 'cmd' set to zero.
It then complains that the local variable 'value' gets passed by
address to xds110_swd_queue_cmd() and in case of 'read request'
such stack address get stored for later use:
src/jtag/drivers/xds110.c:1363:1: warning: Address of
 stack memory associated with local variable 'value' is
 still referred to by the global variable 'xds110' upon
 returning to the caller. This will be a dangling
 reference [core.StackAddressEscape]

To both xds110_swd_write_reg() and xds110_swd_read_reg(), add an
assert() to inform Clang about the state of bit SWD_CMD_RNW.

Change-Id: I7687c055ec71424b642e152f478723a930966e3a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7380
Tested-by: jenkins

flash: lpc2900: fix clang error 'dead assignment'

The variable retval is assigned a value that is never used.
Scan-build reports:
Although the value stored to 'retval' is used in the
enclosing expression, the value is never actually read
from 'retval'.

Drop the dead assignment.

Change-Id: I11588dee748a55d52aa7f35bc1967b7df55af7fc
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7379
Tested-by: jenkins

flash/nor/stm32lx: Add revision '1, X' for Cat.2 devices

Change-Id: I0ff1e2102175ee952b066b325c9acbcb598b3af7
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: https://review.openocd.org/c/openocd/+/7378
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl: max326xx: fix target scripts for latest version of OpenOCD

Change-Id: Iec5aba3a082f2e25f21d7ca173ed710894b370a4

Signed-off-by: Rocco Marco Guglielmi <roccomarco.guglielmi@gmail.com>
Change-Id: Ia83850e326661c8acb0712a280fdf961258322a4
Reviewed-on: https://review.openocd.org/c/openocd/+/7373
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jtag/drivers/cmsis_dap.c: Fix Length of SWO Baudrate Command

The command should now send the full 5 byte command length, which
includes the command tag (0x19) and the 4-byte baudrate word, instead
of only the last 3 bytes of the baudrate.

Signed-off-by: Nick Kraus <nick@nckraus.com>
Change-Id: Idd6e084efd7492489aa900cdbf08f540944041cb
Reviewed-on: https://review.openocd.org/c/openocd/+/7370
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/armv7m: fix feature name of ARMv8M security extension regs

gdb requires this feature to enable stack unwinding of secure/nonsecure
interstate calls and exceptions on an ARMv8M target with
the security extension.

Tested on STM32L5 (Cortex-M33).

Change-Id: Ib09780c011afbc095b352074068597559ad14fcd
Link: https://sourceware.org/git/?p=binutils-gdb.git;a=commit;h=ae7e2f45aa4798be449f282bbf75ad41e73f055e
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7265
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

rtos/ThreadX: added check for NULL-named tasks

Thread name loading was not correctly handled if a ThreadX task has a NULL
name.

Signed-off-by: Giulio Fieramosca <giulio@glgprograms.it>
Change-Id: I03071930182bc2585b61ce5d8c67491710883dd6
Reviewed-on: https://review.openocd.org/c/openocd/+/7328
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

Fix for segfault and some clang reported problems in stmqspi

Change-Id: Id003adb574085cdd603cc13aeb6f2efec73593f1
Signed-off-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7345
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jep106: update to revision JEP106BF.01 Oct 2022

Change-Id: Ia1f19dcce48da997c036ccffa65e76e179de2eb9
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7341
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor/stm32f1x: adjust size of the flash loader buffer

target_run_flash_async_algorithm expects the source_buffer to have
at least 2 words reserved for read and write pointers in addition to the
FIFO buffer. If the size of the data to be flashed is <= 8 bytes then
the flash function will fail with "corrupted fifo read pointer" error.

Ensure the allocated buffer is big enough to hold both FIFO buffer and
read/write pointers.

Change-Id: I09c22eaac517b8cfea8e0b463f5deb6b98afd267
Signed-off-by: Simon Smiganovski <simon.smiganovski@fruitcore.de>
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7342
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/armv7m: prevent saving and restoring non existent regs

armv7m_start_algorithm() saves register values to arch_info->context.
armv7m_wait_algorithm() restores register values from arch_info->context.
Exclude registers with flag exist = false from both loops.

While on it refactor the register restore: introduce 'struct reg' pointer
and dereference it instead of numerous accesses by a full path
from armv7m pointer.

Change-Id: I1600084db84809ee13bcf8e7828b79f8c9ff9077
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7276
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl/interface: fix raspberrypi2-native.cfg speed coefficient

The speed coefficient for Raspberry Pi 2 was probably calibrated
for a scaled down clock frequency.

To prevent JTAG/SWD overclocking, use the value corresponding
to the 'official' maximum CPU clock.

Change-Id: Iaff58b092198dce6d6552c9d31d6a3ba4aaaa2d5
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7305
Tested-by: jenkins
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>

tcl/cpld/xilinx-xcu: fix typo

Fix typo in comments

Signed-off-by: George Voicu <razvanvg@hotmail.com>
Change-Id: Icc2d770e73f896e20dd347de324328030544bdb9
Reviewed-on: https://review.openocd.org/c/openocd/+/7333
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

pld/virtex2: small doc extension

Change-Id: I174cd702388be04268b38178fbfacb90db452f72
Signed-off-by: Daniel Anselmi <danselmi@gmx.ch>
Reviewed-on: https://review.openocd.org/c/openocd/+/7303
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

rtos: hwthread: fix clang error core.NullDereference

Clang spots a potential NULL pointer dereferencing that is instead
an incorrect use of an array of pointers:

src/rtos/hwthread.c:254:32: warning: Dereference of null pointer
  [core.NullDereference]
(*rtos_reg_list)[j].number = (*reg_list)[i].number;
                             ^~~~~~~~~~~~~~~~~~~~~
The error has not been spotted before because:
- this function is not called for the first core of the SMP node,
- for the other cores on Cortex-A it still returns valid register
  value for the first 12 ARM registers, then it diverges.

Also Valgrind does not spot any issue at runtime.

Address the array correctly.

While there, use DIV_ROUND_UP() macro for the computation.

Change-Id: Ib87e60e0edfd9671091f5dcfa9aedaf1aed800d1
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7337
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

Revert "Remove duplicate of a counter in hwthread_update_threads"

Commit 0cedf10f8fd6 ("Remove duplicate of a counter in
hwthread_update_threads") introduced a code bug.

In the second foreach_smp_target() loop, variable "threads_found"
gets passed to routine hwthread_fill_thread(). By removing the
counting of threads_found from the second loop, the
incorrect thread counter value gets passed to hwthread_fill_thread().

Change-Id: Ie89e53ccd28bb72b6838ef2f12106a1fe8d00994
Suggested-by: Daniel Goehring <dgoehrin@os.amperecomputing.com>
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7307
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jtag/drivers/cmsis_dap: prevent CDC missdetect as CMSIS-DAP bulk

The autodetection of CMSIS-DAP v2 bulk interface is tricky
as not all adapters conform CMSIS-DAP specs.

If an interface has a string descriptor containing CMSIS-DAP,
then OpenOCD did not insisted on the correct interface class
LIBUSB_CLASS_VENDOR_SPEC.

However the relaxed test caused false autodetection of v2 bulk
interface on some CMSIS-DAP v1 adapters with an additional serial
interface with the string descriptor stupidly containing
CMSIS-DAP text.

Make the test less relaxed, refuse autodetection of the interfaces
with the class number of well known functions including CDC and MSC.

Link: https://sourceforge.net/p/openocd/tickets/368/
Change-Id: I917cb257eb42aab93560cc39c61ec35a60ce52e3
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7279
Tested-by: jenkins
Reviewed-by: SilverFox <yyjdelete@126.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash: stmqspi: fix clang error 'dead assignment'

The variable retval is assigned a value that is never used, as it
is reassigned few lines below.

Drop the dead assignment.

Change-Id: Id4e9134408fab3e04936d36e95724bf8d3ab55aa
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7304
Tested-by: jenkins

dsp5680xx: fix clang error core.UndefinedBinaryOperatorResult

Clang get confused by initializing the array uint16_t lock_word[],
casting it to (uint8_t *), then accessing the second element of
the uint8_t pointer.

  src/target/dsp5680xx.c:2046:41: warning: The left operand of '<<'
    is a garbage value [core.UndefinedBinaryOperatorResult]
        uint16_t tmp = (buffer[0] | (buffer[1] << 8));
                                     ~~~~~~~~~ ^
Fix it by replacing the array with a single uint16_t.

The code is still depending on host endianness; no fix for this is
proposed.

Change-Id: I16dfd60cab117dd145aeecf10d9593574ff233a2
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7302
Tested-by: jenkins

esirisc_jtag: fix clang error core.VLASize

The function esirisc_jtag_recv() can be called with argument
num_in_fields = 0, for example as consequence of calling
esirisc_jtag_continue().
In this case, num_in_bytes is zero and the allocation of the
variable-length array 'r' requires size zero.

  src/target/esirisc_jtag.c:133:2: warning: Declared variable-length
    array (VLA) has zero size [core.VLASize]
        uint8_t r[num_in_bytes * 2];
        ^~~~~~~~~ ~~~~~~~~~~~~~~~~

Fix it by forcing size one when num_in_bytes is zero.

Change-Id: Id764c7b5ec4f5b3c18c7da650bbff39fc98ed049
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7301
Tested-by: jenkins

openrisc: fix clang error core.CallAndMessage

Clang assumes that size could assume a value that is not 1 nor 2
nor 4. In such condition the buffer in t is allocated (size != 1)
and not initialized. This triggers an error:
  src/target/openrisc/or1k_du_adv.c:655:14: warning: 2nd function
    call argument is an uninitialized value [core.CallAndMessage]
                crc_calc = adbg_compute_crc(crc_calc, data[i], 8);
                           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Add the default case to cover other values of size.

After this fix, clang still complains on the same line, this time
misunderstanding the limits of the loop and considering that
buf_bswap16() only swaps the first 16 bits, thus passing not
initialized value data[2] to adbg_compute_crc()

Replace malloc() with calloc() to silent it.

Change-Id: I358d7fb2ebefd69255670641bd435b770762a301
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7300
Tested-by: jenkins

helper/types: use unsigned type for all h_u64_to_le() and similar

All the converters functions:
h_u64_to_le()
h_u64_to_be()
h_u32_to_le()
h_u32_to_be()
h_u24_to_le()
h_u24_to_be()
h_u16_to_le()
h_u16_to_be()
have signed type in their prototype, while the function name and
all the current use cases pass an unsigned value.

Change the prototypes to use unsigned types.

Change-Id: I76dcfdd7912b81f60902184712b2907eae9843f7
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7299
Tested-by: jenkins

ThreadX: set current_thread for kernel execution

If we just invented thread 1 to represent the current execution, we
need to make sure the RTOS object also claims it's the current thread
so that threadx_get_thread_reg_list() doesn't attempt to read a
thread control block at 0x00000001.

Signed-off-by: Ben McMorran <bemcmorr@microsoft.com>
Change-Id: I7f71e730d047858898297e4cb31db8e47e0c371c
Reviewed-on: https://review.openocd.org/c/openocd/+/7280
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

Restore +dev suffix

Signed-off-by: Paul Fertser <fercerpav@gmail.com>

The openocd-0.12.0-rc2 release candidate

Signed-off-by: Paul Fertser <fercerpav@gmail.com>

tcl/target: add basic RP2040 target config

The existing rp2040-core0.cfg configuration file was intended
for a special adapter which selects a SWD multidrop target on its own.
This means that rp2040-core0.cfg is totally unusable with a standard SWD
adapter.

To fix the problem, mark rp2040-core0.cfg as deprecated and
add rp2040.cfg, a basic config file with multidrop target selection.

Change-Id: I5194e42f529a2d9645481424b7c66ab61efa44ee
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7275
Tested-by: jenkins
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

doc: fix semihosting_redirect command documentation

Change-Id: I78c82a21e4160851a5c0b58394ac7897479808ff
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@st.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7278
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl/stm32l5x|u5x: support HLA adapters in non-secure mode only

instrument "target/stm32x5x_common.cfg" used by both STM32L5x/U5x
to support HLA adapters like "interface/stlink.cfg" in non-secure mode

if the device switches to secure mode, the debug session will be
stopped immediately (with an explanatory message).

Change-Id: I645fdd55e3448ef82d0ddcc396f42fd7b2f39ac3
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reported-by: Patrik Bachan <diggit@users.sourceforge.net>
Fixes: https://sourceforge.net/p/openocd/tickets/317/
Reviewed-on: https://review.openocd.org/c/openocd/+/6546
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins

target/xtensa: remove redundant call for `TARGET_EVENT_HALTED`

`xtensa_do_step` is invoked from `xtensa_prepare_resume` to silently
step over BP/WP before resuming.
For example; in the case of WPs (DEBUGCAUSE_DB), in the current
implementation `xtensa_do_step` will generate one more
`TARGET_EVENT_HALTED` after the original one caused by WP itself.

This patch moves the halted event cb call after
the step is done successfully.

Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: I9048e14fb316dc124847a42cfaefb1f76b5ce53e
Reviewed-on: https://review.openocd.org/c/openocd/+/7274
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

doc: fix riscv commands

- Fix the declaration of riscv command 'set_mem_access'.
- Remove non existing riscv command 'set_scratch_ram'.
- Add riscv commands 'info', 'reset_delays'; copy the description
  from the 'help' text.
- Don't add riscv commands 'set_prefer_sba' and 'test_sba_config_reg'
  as they are marked as deprecated.
- Ensure that 'test_sba_config_reg' prints a deprecation warning
  when used.

Change-Id: I39dc3aec4e7f13b69ac19685f1b593790acdde83
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Signed-off-by: Jan Matyas <matyas@codasip.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7268
Reviewed-by: Tim Newsome <tim@sifive.com>
Tested-by: jenkins

doc: remove unreferenced anchors

Remove the @anchor{} tags that are not referenced in the
documentation.

Change-Id: Ia8e9f75afb08e08ef99d0c8fd82115d689e4a267
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7270
Tested-by: jenkins

doc: fix espusbjtag commands type

Exec commands should be reported as {Command}.

Change-Id: Iacb50d77b354617ecd24b0f1c2ec24e240179698
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7267
Tested-by: jenkins
Reviewed-by: Erhan Kurubas <erhan.kurubas@espressif.com>

doc: fix xtensa commands type

Config commands should be reported as {Config Command}

Change-Id: Ic778df31bb1dc9aefdbe3d8006b06bb370d25e6f
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7266
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Erhan Kurubas <erhan.kurubas@espressif.com>

jtag/drivers: bcm2835gpio: implement memory barriers when bitbashing

This GPIO driver is common to SoCs that have in-order ARM cores
(BCM2835) as well as superscalar (BCM2836-7) and speculative
out-of-order cores (BCM2711).

For BCM2837 and BCM2711, the processor can dual-issue stores and
is free to merge writes to peripheral memory for pages mapped
MT_NORMAL_NC, which is the default provided by /dev/[gpio]mem.

This can cause glitches (or missing edges) on GPIO pins when
toggled with no delay, as pipelined writes to the same address
can get arbitrarily squelched.

To prevent this happening, make sure the preceding write ops are
flushed outside the shareable domain by using a memory barrier.

Signed-off-by: Jonathan Bell <jonathan@raspberrypi.com>
Change-Id: I8805cc0911667bcb9b7f4ca340d7f4f1cb25d096
Reviewed-on: https://review.openocd.org/c/openocd/+/7258
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

README: update build dependency list

Add info on what is optional,

Change-Id: Iedfa969243d95736aaf1b236caa2c2b33f563fe7
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7256
Tested-by: jenkins

README: cleanup requirements for pkg-config

FreeBSD fully supports pkg-config; the .pc files for the internal
libusb has been added with
https://cgit.freebsd.org/src/commit/?id=041d3f3f09b8
and became part of FreeBSD 10.0 in 2014-01-16.

Remove the obsoleted requirements for adding .pc files.

While there, add pkgconf as an alternative to pkg-config.

Change-Id: I16aea735c44107cb71945f225a979682c8c92d0a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7255
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: jenkins

target/xtensa: fill register number field in the cache

Currently 'number' field is zero in the register cache and
this causes an issue on `rtos get_thread_reg_list` calls.

Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: Iaef11e01f55d012969bbc1933f82847d5e02fec5
Reviewed-on: https://review.openocd.org/c/openocd/+/7246
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/esp32s2: check xtensa_poll return value

Although scan build couldn't catch, return value overwritten
without checking.

Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: I02b10002b03640604315047e8a8a639824724c16
Reviewed-on: https://review.openocd.org/c/openocd/+/7247
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl/xtensa: some fixes at xtensa-core-esp32s3.cfg

Some config changes required to run ESP32-S3 with full feature set

Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: I38022bb5ff5830e1cf9d11d6fe795ea99d91e9db
Reviewed-on: https://review.openocd.org/c/openocd/+/7254
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl/xtensa: some fixes at xtensa-core-esp32s2.cfg

Some config changes required to run ESP32-S2 with full feature set

Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: Ie0a742442254ec6e95d4e05be40213b079a94dab
Reviewed-on: https://review.openocd.org/c/openocd/+/7253
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl/xtensa: some fixes at xtensa-core-esp32.cfg

Some config changes required to run ESP32 with full feature set

Signed-off-by: Erhan Kurubas <erhan.kurubas@espressif.com>
Change-Id: I484324f8497ec7934bb73164c638fc5f6460fcc4
Reviewed-on: https://review.openocd.org/c/openocd/+/7252
Tested-by: jenkins
Reviewed-by: Ian Thompson <ianst@cadence.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

tcl/target: fix rp2040-core0.cfg work area backup.

The work area should be backed up.
The flash probe runs an algorithm on the target CPU.
The flash is probed during gdb connect if gdb_memory_map is enabled
(is enabled by default).
Without backup the target memory gets corrupted on gdb connect.

Change-Id: I3344b9dc6cbf904d49f3b05ab104b541d1d63422
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7257
Tested-by: jenkins
Reviewed-by: Jonathan Bell <jonathan@raspberrypi.com>

target: re-examine before arp_waitstate in ocd_process_reset_inner

arp_waitstate will not work on not-examined state

Change-Id: I56c3e1c7e63af108e4ed1dbacebb567f9bf46264
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7230
Tested-by: jenkins
Reviewed-by: Erwan Gouriou
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/hla_target: try to re-examine under reset in hl_assert_reset()

An application often idling in real sleep mode may make a Cortex-M target
hard to access as CPU clock are gated and debug requests are responded
by WAIT ack.

Try to examine the target under reset as the last resort.

Change-Id: I7c3de39fb1e6c23b76e2a0a85ab75f23aac94c4d
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7229
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/cortex_m: try to re-examine under reset in cortex_m_assert_reset()

An application often idling in real sleep mode may make a Cortex-M target
hard to access as CPU clock are gated and debug requests are responded
by WAIT ack.

Try to examine the target under reset as the last resort.

Change-Id: Ife875a966a838c37dde987bc584ad0a1f4d020d6
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/7228
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/cortex_m: make reset robust again

After merging [1] 'reset halt' does not work on not responding Cortex-M.

Relax the examined tests and try to set vector catch VC_CORERESET
if debug_ap is available.

While on it add an info about examination state to debug logs.

Fixes: [1] commit 98d9f1168cbd ("target: reset target examined flag if target::examine() fails")
Change-Id: Ie2e018610026180af5997d70231061a275f05c76
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: https://review.openocd.org/c/openocd/+/6745
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

openocd: fix build with 'configure --without-capstone'

When configure option --without-capstone is used, the macro
HAVE_CAPSTONE is not defined in config.h, and the following lines
are instead present:
/* 1 if you have Capstone disassembly framework. */
/* #undef HAVE_CAPSTONE */

This cause compile error with message:
arm_disassembler.h:190:5: error: "HAVE_CAPSTONE" is not
defined, evaluates to 0 [-Werror=undef]
  190 | #if HAVE_CAPSTONE
      |     ^~~~~~~~~~~~~

This is caused by configure.ac that does not call AC_DEFINE when
--without-capstone option is present.

Fix configure.ac to always provide the autoconf macro
HAVE_CAPSTONE, with either value 0 or 1.

Change-Id: Ie5ac98b2c25746dd721812c91baaac61ec877ecd
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: https://review.openocd.org/c/openocd/+/7224
Tested-by: jenkins

target/adiv5: 64-bit TAR setup bugfix

For 64-bit TAR setup, if 'tar_valid == false' perform the upper 32-bit
write even if the cached copy matches the upper TAR value to be written.

Signed-off-by: Daniel Goehring <dgoehrin@os.amperecomputing.com>
Change-Id: I320377dc90a9d1d7b64cbb281b2527e56c7621ee
Reviewed-on: https://review.openocd.org/c/openocd/+/7245
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins

tcl/target/ti_k3: Handle swd vs jtag

Since all the device definition when accessing device from jtag is also
valid when accessing from swd, lets make sure the configuration can
handle the same.

Signed-off-by: Nishanth Menon <nm@ti.com>
Change-Id: I5af071137fd8c3b52cc4ef72401f8eba952f9cad
Reviewed-on: https://review.openocd.org/c/openocd/+/7090
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

fix leaky file-handle in virtex2 driver

Change-Id: I2784a66c42be71f2982dff7746f9fb2eb1dc8ca6
Signed-off-by: Daniel Anselmi <danselmi@gmx.ch>
Reviewed-on: https://review.openocd.org/c/openocd/+/7243
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

don't return ERROR_OK in error cases

Change-Id: I7e046df85838692c9044fe9c9d67e8b2c821eb0f
Signed-off-by: Daniel Anselmi <danselmi@gmx.ch>
Reviewed-on: https://review.openocd.org/c/openocd/+/7236
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins

fix memory leak in virtex2 driver

Change-Id: Ia08f7aaad25631132885acd5898477c1106f0ec4
Signed-off-by: Daniel Anselmi <danselmi@gmx.ch>
Reviewed-on: https://review.openocd.org/c/openocd/+/7235
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

1024 SHA256:YKx8b7u5ZWdcbp7/4AeXNaqElP49m6QrwfXaqQGJAOk gerrit-code-review@openocd.zylin.com (DSA)
384 SHA256:jHIbSQa4REvwCFG4cq5LBlBLxmxSqelQPem/EXIrxjk gerrit-code-review@openocd.org (ECDSA)
521 SHA256:UAOPYkU9Fjtcao0Ul/Rrlnj/OsQvt+pgdYSZ4jOYdgs gerrit-code-review@openocd.org (ECDSA)
256 SHA256:A13M5QlnozFOvTllybRZH6vm7iSt0XLxbA48yfc2yfY gerrit-code-review@openocd.org (ECDSA)
256 SHA256:spYMBqEYoAOtK7yZBrcwE8ZpYt6b68Cfh9yEVetvbXg gerrit-code-review@openocd.org (ED25519)
+--[ED25519 256]--+
|=..              |
|+o..   .         |
|*.o   . .        |
|+B . . .         |
|Bo. = o S        |
|Oo.+ + =         |
|oB=.* = . o      |
| =+=.+   + E     |
|. .=o   . o      |
+----[SHA256]-----+
2048 SHA256:0Onrb7/PHjpo6iVZ7xQX2riKN83FJ3KGU0TvI0TaFG4 gerrit-code-review@openocd.zylin.com (RSA)