tcl/interface/ftdi: Add imx8mp-evk internal JTAG interface

The imx8mp-evk board has a FT4232H chip connected to the chip's JTAG
pins.

Switching between this on-board interface and the external ARM-10
connector is controlled by an GPIO which is behind an i2c expander
connected to the same FTDI chip.

Switching can be performed using the NXP bcu tool:

https://github.com/NXPmicro/bcu

Change-Id: Ic910515f76eaf09ea6d0f755b026fb09cf09ccfc
Signed-off-by: Leonard Crestez <leonard.crestez@nxp.com>
Reviewed-on: http://openocd.zylin.com/5426
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/stm32h7x: use alignment infrastructure

Report the 32-byte alignemnt requirement via the bank structure rather
than enforcing it ad-hoc in the write routine. This allows people to do
non-32-byte-aligned writes if they want, with the infrastructure fixing
up the addresses passed to the low-level driver.

Change-Id: I2c4f532f2000435954a900224dbc9f2c30d1cc94
Signed-off-by: Christopher Head <chead@zaber.com>
Reviewed-on: http://openocd.zylin.com/5388
Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor/stm32f1x: Some small code cleanups

Change-Id: I1fc08b96b179a1376af233b713ae50d6ad7867a0
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: http://openocd.zylin.com/5404
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

flash/nor/stm32h7x: check OPTCHANGEERR

Without this, a failed attempt to change option bytes will silently
appear to succeed but without actually changing the option bytes
(confusingly, the option bytes will still read back as if they had been
changed until a reboot as well!).

Change-Id: Id529c6c384a8a16be75f5702310670d99d8fac79
Signed-off-by: Christopher Head <chead@zaber.com>
Reviewed-on: http://openocd.zylin.com/5418
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

flash/nor/stm32h7x: fix incorrect array indexing

Change-Id: Iec2246df284953d1442dfefdad8e70041690dfe2
Signed-off-by: Christopher Head <chead@zaber.com>
Reviewed-on: http://openocd.zylin.com/5417
Tested-by: jenkins
Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

stlink: add trace support in DAP direct mode

Reuse the existing tracing functionality of HLA mode to support
tracing in DAP direct mode.

Change-Id: I75a01e88ba5d3e45717e4108b99697ac3225db9e
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5409
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

armv8: check the core state to pass the correct arch to gdb

Commit 3799eded6761 ("target/aarch64: add support for
multi-architecture gdb") passes the constant string "aarch64" as
architecture to gdb. This is not working if the core is running
in 32 bits mode; gdb reports:
Truncated register 8 in remote 'g' packet
then closes the connection with OpenOCD.

Make the architecture string dependant from the current state of
the core.

Change-Id: I16e1614ea02ba29bf87f450b3dfe25c83c9a3612
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5234
Tested-by: jenkins
Reviewed-by: Muhammad Omair Javaid <omair.javaid@linaro.org>

coding style: use ARRAY_SIZE() when possible

We have the macro ARRAY_SIZE() already available. Use it!

Issue identified by checkpatch script from Linux kernel v5.1 using
the command

find src/ -type f -exec ./tools/scripts/checkpatch.pl \
-q --types ARRAY_SIZE -f {} \;

Change-Id: Ic7da9b710edf118eacb08f9e222f34208c580842
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5198
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

coding style: remove unnecessary parentheses

Identified by checkpatch script from Linux kernel v5.1 using the
command

find src/ -type f -exec ./tools/scripts/checkpatch.pl \
-q --types UNNECESSARY_PARENTHESES -f {} \;

then fixed manually.

Change-Id: Ia2d9a0953d9b89fc87dc1195aa05c7f63c068c48
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5196
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

jtag/startup.tcl: remove trailing whitespaces

Change-Id: I1d6f4f47ee6f8985c84ddb2647e029c5f4e6a55a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5194
Tested-by: jenkins
Reviewed-by: Marc Schink <dev@zapb.de>

target/nds32: fix type of magic number

The macro NDS32_COMMON_MAGIC was cast-ed to int to avoid compile
time error for comparison type mismatch while comparing it with
the field common_magic.
This is incorrect because the macro value is a 32 bit unsigned
value; better changing the type of the field common_magic to keep
the unsigned value.

Issue identified by checkpatch script from Linux kernel v5.1 using
the command

find src/ -type f -exec ./tools/scripts/checkpatch.pl \
-q --types TYPECAST_INT_CONSTANT -f {} \;

Change-Id: Ib5924b6cecdffe70ab5c78d3b30a9c8e4deb7c7b
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5193
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor/stm32lx: Minor code cleanups

Change-Id: I6440a4eb1f65a2f8ae2914b38f21a59955e85e0d
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: http://openocd.zylin.com/5438
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jtag_vpi: added an option to stop simulation on exit

Command CMD_STOP_SIMU had been defined in jtag_vpi for a long time
(since the beginning?) but has not been utilized until now.
Its purpose is to signal to the jtag_vpi server (i.e. the RTL
simulation software) that the simulation shall be stopped.

This commit adds a TCL configuration command that selects whether
CMD_STOP_SIMU will be sent to the jtag_vpi server when OpenOCD is
about to exit. This functionality is off by default to maintain
identical behavior as in previous OpenOCD versions, unless the user
enables it explicitly.

Change-Id: If3894af6efa61038ccf6c9191f664e2128f2ef11
Signed-off-by: Jan Matyas <matyas@codasip.com>
Reviewed-on: http://openocd.zylin.com/5407
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

Include start-of-region address

When $ADDRESS == $BASE that address used to be excluded.

After this commit $ADDRESS == $BASE is within the range.

Now the rule for "iswithin" is:
$BASE <= $ADDRESS < ($BASE + $LEN)

Thanks to Antonio Borneo <borneo.antonio@gmail.com> for noticing this.

Change-Id: I74373c6010e069372d886fa7ecd8120892616834
Signed-off-by: Piotr Kasprzyk <ciri@ciri.pl>
Reviewed-on: http://openocd.zylin.com/4799
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

flash/stm32lx: mention explicitly that this driver covers STM32 L0 and L1

this is to avoid confusion with STM32 L4, L4+ and L5 families

also:
 - a warning message is changed to error
 - stm32l0x and stm32l1x aliases has been created to permit
   the usage of either names

Change-Id: If3f16d2a3b7d1369959aa7407da37a9076ea91d7
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/5437
Reviewed-by: Marc Schink <dev@zapb.de>
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/startup.tcl: update stm32 flash driver aliases

This will enable us to use either name when calling flash driver commands.

For example the stm32wbx family use the same flash driver as the stm32l4x, so
the user has to use 'stm32l4x lock 0' which can be confusing.
Now the user can also use 'stm32wbx lock 0' with the same result.

Change-Id: Ic0d8da9afc202d7cc82d9b9949827e958a1cc824
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/5436
Tested-by: jenkins
Reviewed-by: Marc Schink <dev@zapb.de>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

jtag: Fix copy-paste error in 'irscan' help

Replace "Instruction Register (DR)" with "Instruction Register (IR)",
that is sed "s@DR@IR@", which was likely a copy-paste error.

Change-Id: I3e625872c855d655485b3efa5f50fe1c00ecbf52
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5446
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins

tcl: Remove executable bit

Change-Id: Ib452435b13c3cb8d14453d983151936238b9601d
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: http://openocd.zylin.com/5419
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

efm32: add EFR32ZG13P and EFR32ZG14P parts

This adds the EFR32 Zen Gecko Family parts. The device family values are
found in table 4.7.11 of
https://www.silabs.com/documents/public/reference-manuals/efr32xg14-rm.pdf.

Change-Id: I3858b7ba815784b1150e2214a2833e8ff7d249e1
Signed-off-by: Frank Hunleth <fhunleth@troodon-software.com>
Reviewed-on: http://openocd.zylin.com/5364
Tested-by: jenkins
Reviewed-by: Marc Schink <dev@zapb.de>

stlink: fix max packet size for 8 bit R/W on stlink-v3

While ST internal documentation for STLINK-V3 reports that 8 bits
read/write commands handle 512 bytes of data, a firmware bug makes
it crashing on high data size.
This is fixed with firmware V3J6 (shipped together with V2J36).

Check for firmware version to use the proper data size.

Change-Id: Iaba6cd26bbe130097c1c19de610680e0e8b69bfc
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Fixes: https://sourceforge.net/p/openocd/tickets/259/
Reviewed-on: http://openocd.zylin.com/5408
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

jtag: flush jtag queue after jtag_add_tlr()

If the TLR sequence is sent as result of the command "adapter
assert trst" while polling is off, the TLR sequence is not sent out
until a following jtag operation.

Flush the jtag queue before return.

Change-Id: I20efd7137cb7b1d1c4f73c1362cbe4e57aeaae49
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5405
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

Switch to HTTPS for submodules

repo.or.cz already redirects HTTP requests to HTTPS. There is therefore
no possible benefit to keeping the submodule URLs using HTTP—anyone who
can’t access via HTTPS will fail anyway, immediately after the redirect.
Changing the submodule URLs eliminates one unnecessary HTTP request and,
more importantly, eliminates SSLStrip-style attacks.

Change-Id: I9faf1ec8aa87bcfd1acafe2c445a0baf2abfbd09
Signed-off-by: Christopher Head <chead@zaber.com>
Reviewed-on: http://openocd.zylin.com/5406
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

tcl: fix remaining scripts after rework adapter commands

Some script have been added or modified after the patches for
reworking the adapter commands were pushed in gerrit.
Such scripts use the old command syntax and trigger a "deprecated"
warning at runtime.

Fix them with the same sed commands used for the other scripts:
sed -i 's/^interface /adapter driver /' $(find tcl/ -type f)
sed -i 's/adapter_khz/adapter speed/g' $(find tcl/ -type f)
sed -i 's/adapter_nsrst_delay/adapter srst delay/g' $(find tcl/ -type f)
sed -i 's/adapter_nsrst_assert_width/adapter srst pulse_width/g' $(find tcl/ -type f)

Change-Id: I0824d6c506a9af3eb9129b74c02a92b4eb1b100d
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5424
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Tested-by: jenkins

tcl: Add interface config for isodebug isolated JTAG/SWD+UART

Currently only used internally at Unjo.

Change-Id: I7786e95e1bd755a73156ecad5b6d7f6273d8ddef
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/5251
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jtag: drivers: xlnx-pcie-xvc: fix build on Linux pre v4.10

The macro PCI_CFG_SPACE_EXP_SIZE is exposed to userspace from
Linux kernel v4.10, with commit cc10385b6fde ("PCI: Move config
space size macros to pci_regs.h")
http://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=cc10385b6fde

Define the macro in the driver code, if not already defined.

Change-Id: I610219a2587eff2c142102b9f7830e3da882af78
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5435
Reviewed-by: Moritz Fischer <moritzf@google.com>
Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Tested-by: jenkins

flash/nor: Add Renesas RPC HF driver

Add driver for the RPC block in HF mode on Renesas R-Car Gen3 SoCs.
This driver allows operating the on-SIP HF memory.

Note that HF is CFI compliant flash, but it is not memory mapped,
hence the need to replace all the memory accessors and read/write
functions. The write function is entirely replaced to increase
performance and is Spansion/AMD specific, since there is no known
SIP with other HF from another vendor.

Add the following two lines to board TCL file to bind the driver on
R-Car Gen3 SoC using HyperFlash:

  set _FLASHNAME $_CHIPNAME.flash
  flash bank $_FLASHNAME rpchf 0x08000000 0x4000000 2 2 $_CHIPNAME.a57.0

Change-Id: Ie18729d017eeb46e1363333ffe002d010dfc5ead
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5149
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

flash/nor: Export various functions from the CFI core

Export various functions needed by other driver, specifically the
upcoming Renesas RPC HF driver. No functional change.

Change-Id: I551258979a7221288fb4f4382f857db5cfe0b0de
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5148
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

jtag: Fix jtag_reset fallback

The fallback provided for the jtag_reset command always fails with a
strange message: 'Error: invalid command name "de"'

This is caused by incorrect quoting inside the warning message.

Fixes: c07b774e8f49 ("jtag: replace command "jtag_reset" with "adapter [de]assert"")
Change-Id: Icd47fca2b5a7b33474bfb0040e88193a0968f301
Signed-off-by: Leonard Crestez <leonard.crestez@nxp.com>
Reviewed-on: http://openocd.zylin.com/5416
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

tcl: replace the deprecated commands with "adapter ..."

Avoid annoying "deprecated" messages while running the scripts
distributed with OpenOCD code.
Change automatically created with commands
sed -i 's/adapter_khz/adapter speed/g' $(find tcl/ -type f)
sed -i 's/adapter_nsrst_delay/adapter srst delay/g' $(find tcl/ -type f)
sed -i 's/adapter_nsrst_assert_width/adapter srst pulse_width/g' $(find tcl/ -type f)

Minor indentation issue fixed manually in
tcl/board/at91sam9g20-ek.cfg
tcl/target/at91sam9260_ext_RAM_ext_flash.cfg

Change-Id: I425fd56c0c88cd6b06124621306eeb89166dfe71
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5284
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

doc: replace example command "interface" with "adapter driver"

Keep documentation consisted after commands renaming.

Change-Id: I97b43887cae9d7c224b07e4ba0b7d04915a19fc4
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5285
Tested-by: jenkins
Reviewed-by: Marc Schink <dev@zapb.de>
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

tcl: replace command "interface" with "adapter driver"

Avoid annoying "deprecated" messages while running the scripts
distributed with OpenOCD code.
Change automatically created with command
sed -i 's/^interface /adapter driver /' $(find tcl/ -type f)

Change-Id: I2291dfb96e164beecbeb3366ce83f9df2ad6c197
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5283
Reviewed-by: Marc Schink <dev@zapb.de>
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

openocd: fix minor inconsistencies after renaming "adapter" command

Replace in the code any reference to the deprecated commands.

Change-Id: I75d28064017d664990b4024967900f32e196230a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5282
Tested-by: jenkins
Reviewed-by: Marc Schink <dev@zapb.de>
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

move ftdi_location deprecation helper to proper place

Change-Id: I927d4e918acbf321aea1dd7a8de95fbaa8fbbbf0
Signed-off-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-on: http://openocd.zylin.com/5278
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jtag: adapter: rework adapter related commands

currently we have different types of same command group:
- starting with adapter_*
- starting with interface*
- without adapter or interface prefix.

Since interface name is already used, we can only use "adapter" command
group by keeping old commands as well.

Change-Id: Id0a1cb63a2ea6860c67ae1e7a3a06a37ddf464f4
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-on: http://openocd.zylin.com/4774
Reviewed-by: Marc Schink <dev@zapb.de>
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

flash/nor/stm32l4x: fix minor errors in flash write/async algo

Fix comment of tested errors in asm src.

List all relevant errors in FLASH_ERROR mask: FLASH_PROGERR was missing
and any trial to re-program already programmed double word ended up
in the error bit held uncleared and flash write permanetly repeating
the error message until reset.

Lock the bank also after unsuccesfull write_block run.

Set async target algo block size to size of double word.

Remove warning in case of write_block success. In case of error
use LOG_ERROR instead of warning.

Change-Id: Ibf6d5e306a4c2eaa43de67d636b4902c737f02f3
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5360
Tested-by: jenkins
Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>

flash/nor/stm32l4x: use flash infrastructure to align write

The original code paded the write chunk with random bytes by overrunning
the buffer. An user can easily regard the random bytes to
be a programming error.

Change-Id: Ib0f47b5bc406bc6a7c32f3d929bf324a17c7c1e1
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5359
Tested-by: jenkins
Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>

tcl/board: Add config for STM32WB Nucleo board

Change-Id: Ic29802306c706bcf3e261c60facd01d101c9e1ce
Signed-off-by: Marc Schink <dev@zapb.de>
Reviewed-on: http://openocd.zylin.com/5358
Tested-by: jenkins
Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Ilya Kharin <akscram@gmail.com>

target/arm_cti: fix regression from Tcl_return_values series

Since commit 7f260f5009a774f2d66b5f3037f8f595c6881d4d native OpenOCD
command handlers should not directly use Jim_SetResult functions.
The Tcl result of a native command is built as concatenation of
command_print() strings and Jim_SetResult() is called after return
of the command handler.

Replace "wrong number of args" error messages (now not delivered to user)
by simply return ERROR_COMMAND_SYNTAX_ERROR

Change-Id: I40c1374a13859cefbdef68e0f1c13ab93538bd50
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5363
Tested-by: jenkins
Reviewed-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>

flash/nor/nrf5: Fix build error on OSX

The chip->hwid is uint32_t , fix the print format.
This was detected by TravisCI on OSX, where this triggers a build error.

Change-Id: I776a7bb50e396c8fccc24500dec4750190da7982
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5401
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Ilya Kharin <akscram@gmail.com>

jtag: drivers: xlnx-pcie-xvc: fix build after merge

Commit [1] was submitted in gerrit well before the conflicting
commit [2] get merged in master branch. While it was fine
committing in master branch [1] alone, it should not be
committed "as is" after [2].
Unfortunately gerrit did not complained committing [1] after [2].
The result is that master branch does not build anymore when the
driver xlnx-pcie-xvc is enabled at configure time by the optional
flag --enable-xlnx-pcie-xvc.

Apply to the driver the required changes as in [2].
While there, remove the duplicated struct xlnx_pcie_xvc_transports
and the struct field already implicitly initialized to zero.

[1] ff6d0704ecd6 ("jtag: drivers: xlnx-pcie-xvc: Add support for
    Xilinx XVC/PCIe")
[2] efd1d642220a ("adapter: switch from struct jtag_interface to
    adapter_driver")

Change-Id: I5498479b802a231afbee1b845ae9775e1da7c728
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5402
Reviewed-by: Moritz Fischer <moritzf@google.com>
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

flash/nor/sh_qspi: Add SH QSPI driver

Add driver for the SH QSPI controller. This SPI controller is often
connected to the boot SPI NOR flash on R-Car Gen2 platforms.

Add the following two lines to board TCL file to bind the driver on
R-Car Gen2 SoC and make SRAM work area available:

  flash bank flash0 sh_qspi 0xe6b10000 0 0 0 ${_TARGETNAME}0 cs0
  ${_TARGETNAME}0 configure -work-area-phys 0xe6300000 -work-area-virt 0xe6300000 -work-area-size 0x10000 -work-area-backup 0

To install mainline U-Boot on the board, use the following procedure:

  proc update_uboot {} {
    # SPL
    flash erase_sector 0 0x0 0x0
    flash write_bank 0 /u-boot/spl/u-boot-spl.bin 0x0
    # U-Boot
    flash erase_sector 0 0x5 0x6
    flash write_bank 0 /u-boot/u-boot.img 0x140000
  }

Change-Id: Ief22f61e93bcabae37f6e371156dece6c4be3459
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
---
V2: - Add Makefile and linker script for the SH QSPI IO algorithm
    - Include the algorithm code instead of hard-coding it
Reviewed-on: http://openocd.zylin.com/5143
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

jtag_libusb_bulk_read|write: return error code instead of size

A USB bulk write/read operation may fail with different errors:
 LIBUSB_ERROR_TIMEOUT if the transfer timed out (and populates transferred)
 LIBUSB_ERROR_PIPE if the endpoint halted
 LIBUSB_ERROR_OVERFLOW if the device offered more data, see Packets and overflows
 LIBUSB_ERROR_NO_DEVICE if the device has been disconnected
 another LIBUSB_ERROR code on other failures

Current OpenOCD code is using the transfer size based error detection.
Which may not always work. For example for LIBUSB_ERROR_OVERFLOW as libusb
documentation says:
"Problems may occur if the device attempts to send more data than can fit in
the buffer. libusb reports LIBUSB_TRANSFER_OVERFLOW for this condition but
other behaviour is largely undefined: actual_length may or may not be accurate,
the chunk of data that can fit in the buffer (before overflow) may or may not
have been transferred."

This patch is refactoring code to use actual error return value for
error detection instead of size.

Change-Id: Iec0798438ca7b5c76e2e2912af21d9aa76ee0217
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-on: http://openocd.zylin.com/4590
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

mips_ejtag: there is no DCR.MIPS64 bit

available revisions (2.60, 3.10, 5.06 and 6.10) of
MD00047 (EJTAG specification), have only in IMPCODE
MIPS32/MIPS64 bit/flag.

Change-Id: If9191b6ced760c59bb7551bb041cd72b0a060bb1
Signed-off-by: Jiri Kastner <cz172638@gmail.com>
Reviewed-on: http://openocd.zylin.com/4628
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>
Tested-by: jenkins

jtag: drivers: xlnx-pcie-xvc: Add support for Xilinx XVC/PCIe

Add support for Xilinx Virtual Cable over PCIe JTAG controller.
It is commonly used in Xilinx based PCI Express designs with JTAG IP
in the FPGA fabric.

Access to the JTAG registers happens via the PCI Express extended
configuration space.

This can be used to debug soft-cores instantiated in the FPGA fabric.

The clang static checker doesn't find any new problems with this change.

Change-Id: Ib12ede0d1f26dacfda808d5e05b947b640c5bde7
Signed-off-by: Moritz Fischer <moritzf@google.com>
Reviewed-on: http://openocd.zylin.com/5314
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Marex
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

flash/nor: Rename flash_address() to cfi_flash_address()

This is a preparatory change, align the function name with the rest
of the API, no functional change.

Change-Id: I6a810d2a54edcd13ad9a87d24a7334802c41623b
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5391
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor: Rename get_cfi_info() to cfi_get_info()

This is a preparatory change, align the function name with the rest
of the API, no functional change.

Change-Id: Ib967520f027b03eb1792b36ede52335df8e23941
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5390
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor: Allow CFI memory read/write functions be overriden

Add possibility to supply custom CFI memory accessors via cfi_info
and override the default memory-mapped ones.

Change-Id: I1b6bc1db69fc33e8cdef96c41742c40e6d8917e9
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5147
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor: Drop size argument of cfi_target_{read,write}_memory()

The size argument is always set to bank->bus_width and bank pointer
is now passed into cfi_target_{read,write}_memory(), so the size
can be accessed through the bank pointer inside the function instead
of being explicitly passed in.

Change-Id: I0abc1cc3bf513281c10cb5de7a21cb0e75cb7676
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5389
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor: Pass flash_bank to memory accessors

Replace passing in struct target with passing in struct flash_bank,
so that the later can contain function pointers to custom per-driver
memory accessor functions.

Change-Id: Id2573a6d5f1a73ed9c4f73c53592a9a335a11c99
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5146
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

target/xscale: remove duplicated command

Reported by clang static analyzer.

Change-Id: I893af10852af4885507ed62d024008159a80dd56
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5382
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

target/etm: add check for calloc error

and fix one more clang static analyzer warning.

Change-Id: I17f03e318e1cf7617e7f753e7ca960552be547e5
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5381
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

target/semihosting_common: fix minor memory leak

Reported by clang static analyzer.

Change-Id: Ie663f49d92588c0d8b502cfdd8fc34004b308066
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5380
Reviewed-by: Liviu Ionescu <ilg@livius.net>
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

target/target: fix clang static analyzer warnings

Change-Id: I23e6586be60915f21a7179a994a1ec93fb9b2c36
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5379
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

target/avr32_jtag: fix error returns

Fixed only 2 error returns discovered by clang static analyzer.
There are obviously many more missing error tests in avr32_jtag.c
These was not fixed to keep this change minimal.
Not tested with hw.

Change-Id: I6c79f6248db774990ddb42c0dacdb621651ed69e
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5378
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

target/arm_dpm: add missing error returns

Discoverd by clang static analyzer.

Change-Id: I93d5de0a36216e62b170fe8cc870431226a7777f
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5377
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

target/arm946e: add missing error detection

Discoverd by clang static analyzer.
While on it remove useless type casts from arm946e_read_cp15() parameter.

Change-Id: I549e19685b431400243800ee0f7d1bbe6cdb14b4
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5376
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

server/server: fix clang static analyzer warning

Change-Id: I317e189b62540e3688a20d88a95f551280317f14
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5373
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

xsvf: fix memory leak

Discovered by clang static analyzer.

Change-Id: I2980586aea5ee43226adb1f4cf72e7ba1dfddd83
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5372
Tested-by: jenkins
Reviewed-by: Oleksij Rempel <linux@rempel-privat.de>

flash/nor/stm32l4x: add support of STM32L4P5/L4Q5x devices

STM32L4P/Q devices have:
 - similar flash layout as STM32L4R/S devices
 - 1024K of flash memory (some parts have 512K only)

tested on NUCLEO-L4P5ZG using board/st_nucleo_l4.cfg

Change-Id: I77047351bc7dcd7c76d0f31a77be73005104a06f
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/5392
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor: add support of STM32WB on top STM32L4 flash driver

Change-Id: I9fb6700085d817d35a691f6484193f67939a4e0f
Signed-off-by: Laurent LEMELE <laurent.lemele@st.com>
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/4933
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor: add support of STM32L41/L42xx

tested using STM32L412KB

Change-Id: I1e2ae93d8c740db219f0fb579940de7f2fffac15
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/4934
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor/stm32l4x : add structure containers to hold devices' information

This rework is inspired from the 'flash/nor/stm32h7x.c'
This rework will ease the support of new devices on top of this driver:
  for example: STM32WB have different flash base and size addresses

Notes:
 - stm32l4_probe modified in order to charge the correct part_info from
   the defined stm32l4_parts according to the device id
 - stm32l4_flash_bank.bank2_start is replaced by .part_info->bank1_sectors
 - STM32_FLASH_BASE is removed , part_info->flash_regs_base will be used instead
   based on that flash register addresses are changed to offsets,
   >> stm32l4_get_flash_reg was modified accordingly
 - stm32l4_read_option and stm32l4_write_option was modified to accept an
   offset instead of an absolute address, luckily this is the commands'
   argument by default
 - stm32l4_mass_erase modifications :
     - use MER2 only on top of dual bank devices
     - wait for BUSY bit before starting the mass erase

Change-Id: Ib35bfc3cbadc76bbeaaaba9005b82077b9e1e744
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/4932
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>

target/target: parse value as proper type

The `value` variable is passed into `target_fill_mem` as its
second-to-last parameter. That parameter is of type `uint64_t`. It is
appropriate to parse the value as that type, since otherwise a target
with a 32-bit address space but 64-bit data write capabilities would not
be able to exercise those capabilities.

Change-Id: Ib336d47d42c27cd2b5ba1206b04e8f740f167dba
Signed-off-by: Christopher Head <chead@zaber.com>
Reviewed-on: http://openocd.zylin.com/5219
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins

stlink: fix handling of DPv1 and DPv2 banked registers

Arm DPv1 and DPv2 support banked registers by setting the bank in
field DPBANKSEL of register DP_SELECT.
Old ST-Link firmware don't support banked registers and setting a
bank other than bank zero on DPv1 or DPv2 cause issues in the
firmware because it cannot set back bank zero to read CTRL/STAT.
New ST-Link firmware mask away DPBANKSEL bits while writing in
DP_SELECT but support banked register using the same packed method
used by OpenOCD:
#define BANK_REG(bank, reg) (((bank) << 4) | (reg))

Add a new macro STLINK_F_HAS_DPBANKSEL for firmware that support
arm DPv1 and DPv2, plus trigger an error if banked registers are
requested on old firmware.
Prevent changing DPBANKSEL on old firmware.
Log a debug message when changing DPBANKSEL will be ignored.

Change-Id: Iaa592517831d63f8da2290db54f6b32504e3081b
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4978
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

stlink: add DAP direct driver

STLINK-V3 and the new firmware V2J24 for ST-LINK/V2 provide API
to directly access the DAP registers.
This mode permits to use the native target in cortex_m.c, with no
need to override it with the target in hla_target.c.
Other advantages wrt HLA are: support for Cortex-A cores, support
for SoC multi-core and/or multi AP, support for OpenOCD commands
"dap" thus including control of CSW.
This obsoletes the existing HLA driver for ST-Link, that should
anyway be kept for those cases where it's not possible to update
the ST-Link firmware.

This commit introduces the minimal implementation for direct DAP
access. The implementation is much slower than the HLA because
every memory transfer requires several USB packets. Further
commits will close the performance gap.
The whole ST-Link driver is compiled under BUILD_HLADAPTER, to
remove the need to split the driver between the two modes. This
has to be reworked, but it's quite invasive!
A new interface file stlink-dap.cfg is added and should be used
in place of stlink.cfg to enable the DAP mode.
Documentation is updated and reports limitation on the maximum AP
number that can be accessed by ST-Link for some firmware already
tested.

Change-Id: I932ffe16bc81d00b1fe489e2944fda13470cce9b
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4904
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

adi_v5_dapdirect: add support for adapter drivers that provide DAP API

Some high level adapters, like STLINK-V3 and new firmware for
ST-Link/V2, provide API to directly access the DAP registers
hiding the details of the physical transport JTAG or SWD.
OpenOCD has already the intermediate API in struct dap_ops that
are suitable for such adapters, but are not exposed to the
adapter drivers.

Add in struct adapter_driver two independent struct dap_ops for
the cases of physical JTAG and SWD transport.
Add new transport names "dapdirect_jtag" and "dapdirect_swd", to
be used by the drivers that provide one or both DAP API.
Add the necessarily glue in target/adi_v5_dapdirect.c

Change-Id: I2bb8e3a80fba750f2c218d877cfa5888428e3c28
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4903
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

TODO: add restructuring of JTAG/adapter layer

SWD is already implemented, so remove the item.
Rename the section JTAG as Adapter, including the subsections.
Add an initial list of pending activities after the restructure
of the JTAG layer.

Change-Id: I540777344c62a746df8347538fe8b29e4d72e1c7
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4901
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

adapter: switch from struct jtag_interface to adapter_driver

To reorganize the adapters code, introduce an adapter_driver
struct that contains all the adapter generic part, while
keeping in two separate struct the specific API jtag_ops and
swd_ops.
Move the allocation of *adapter_driver from the JTAG-specific
file core.c to the more adapter-specific file adapter.c
While splitting the old jtag_interface for every driver, put
the fields in the same order as in the struct declaration so
we keep a consistent code across all the drivers.

While other transport specific API could/would be added as
separate ops, nothing is done here for HLA.

Change-Id: I2d60f97ac514c0dd2d93a6ec9be66fd9d388dad5
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4900
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

drivers/kitprog: remove unused JTAG execute_queue method

kitprog is SWD only and we do not rely on JTAG queue anymore.
Remove the remaining JTAG heritage.

Change-Id: Ic586278368301eb669bc6e4e641f683a81cb171d
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4899
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

hla: remove empty JTAG execute_queue method

We do not rely on JTAG queue anymore.
Remove the remaining JTAG heritage.

Change-Id: I6c87d9ffebaa383c998cf273188b3e7f28b3fe95
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4898
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

jtag: print an errmsg on using jtag API for non jtag transport

After the cleanup of swd and hla, there should be no more calls
to jtag_execute_queue() or to queue jtag commands if current
transport is not jtag. Thus we can start removing the jtag
specific code from adapters that do not support jtag.

To prevent some remaining call to jtag_execute_queue() to crash
openocd, verify the transport, print an error message if the
transport is not jtag, call the adapter's jtag_execute_queue()
only if it exist.
To identify code that still add commands in the jtag queue even
if transport is not jtag, print an error message in the function
jtag_queue_command(). For the moment, still queue the message,
even if will cause a memory leak if there is no following call
to jtag_execute_queue(); the target is to identify the issue and
cleanup the code, thus solving also the leak.

Change-Id: I8fc85f754aa057aad1df05ff0448c8619897da23
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4897
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

hla: use the new system_reset API

HLA uses its own internal driver's API to control the adapter's
system reset, but at the same time it calls jtag_add_reset() to
avoid breaking the internal logic of OpenOCD. This implicitly
forces HLA to rely on jtag queue mechanism, even if HLA has no
link with JTAG state machine. It requires HLA to implement an
empty execute_queue() to comply with the JTAG queue.

Modify the HLA framework and the HLA targets to use the new
adapter API for system_reset and decouple HLA from JTAG queue.
Rename the HLA static functions adapter_assert_reset() and
adapter_deassert_reset() to avoid overlap with the global
functions with same name.
While there, fix a minor typo in a comment s/incase/in case/.

Do not remove from HLA the JTAG specific API execute_queue(),
even if not required anymore, because OpenOCD code still has
calls to jtag_execute_queue() in case of non JTAG transport.

Change-Id: I0e65e3e557bd665bd3d3aeaa84ea609b55a05e48
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4896
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

bitbang: jtag-only drivers: switch to new reset API

Remove the JTAG_RESET command from the bitbang execute queue now
that all bitbang drivers have moved away from old reset method.
Remove also the internal reset API in struct bitbang_interface.

Tested parport only.

Change-Id: I12b157ef442f4c9912406b19b7a4d32ba6ec0b53
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5300
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

swd: get rid of jtag queue to assert/deassert srst

The transport SWD uses the JTAG queue to assert/deassert the
system reset srst. This is the major inconsistency that has to be
removed to properly split JTAG and SWD.

Introduce a new driver API, reset(), to controls both the signals
trst and srst in the driver, skipping the JTAG queue. Put the new
API in struct jtag_interface, even if in this patch it's used for
SWD only; the goal is to get it reused by the other transports.

Add the implementation of the API in all the drivers that
implement SWD. Such implementation is almost the same of the old
code in JTAG queue.

Create a wrapper adapter_system_reset() to use the new API and
remove the SWD specific swd_add_reset(). In the wrapper replace
jtag_add_sleep() with jtag_sleep(), because the former uses the
JTAG queue too.
Rename the old jtag_add_reset() as legacy_jtag_add_reset() with
the target to remove it when all drivers would be ported to the
new reset API. Create a new jtag_add_reset() that calls the
legacy function for drivers still on the old reset API.

Use the new API also on JTAG transport for the drivers that can
support both SWD and JTAG.

For the moment, do not modify the implementation of JTAG-only
drivers, which will continue using the usual method. This should
be cleaned-up in future commits.

Change-Id: I32331c88313f6059b25e12c6bb0156aebc1c074f
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4895
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor: Factor out CFI memory read/write functions

Create separate memory read/write functions which facilitate access
to the CFI NOR, so that they can be replaced by controller-specific
functions if necessary. This would become necessary when implementing
support for e.g. HyperFlash controllers, which do not directly map
the HyperFlash into the address space.

Change-Id: I1bba1edfd397cb37bfedb43efe2dd03feb26a375
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5145
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>

flash/nor: Factor out cfi_spansion_unlock_seq()

Factor out the spansion unlock sequence to deduplicate the code.

Change-Id: Id78522e9a2f0e701870ef816772289d08257476a
Signed-off-by: Marek Vasut <marek.vasut@gmail.com>
Reviewed-on: http://openocd.zylin.com/5144
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>

flash/nor/stm32f2x: Support value line chips with trimmed flash

The current code assumes an STM32's flash bank is laid-out in either of
two configurations:
- 4 x 16kB + 1 x 64kB + n x 128kB
- 4 x 32kB + 1 x 128kB + n x 256kB

This is quite ad-hoc but works fine in practice, as long as there are at
least 5 sectors (if n=0). Unfortunately, some newer STM32s are shipping
with only 64 kB of flash (4 x 16kB sectors).

This patch still assumes the same sector layout, but only keeps adding
sectors to the bank if the bank's capacity has not been reached. This
prevents openocd from crashing on some newer STM32s.

Change-Id: If00e5d7a328d11b399babc0bb2111e3ad8a3217e
Signed-off-by: Romain Goyet <romain.goyet@numworks.com>
Signed-off-by: Keir Fraser <keir.xen@gmail.com>
Reviewed-on: http://openocd.zylin.com/4926
Tested-by: jenkins
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

arm_adi_v5: add API send_sequence() and use it

The method to send an arbitrary sequence to DAP depends on the
transport and is thus different on JTAG and SWD. This is already
coded in dap_to_jtag() and dap_to_swd().

Add a new API send_sequence() in struct dap_ops.
Add the implementations of send_sequence() in adi_v5_jtag.c and
adi_v5_swd.c
Rewrite dap_to_jtag() and dap_to_swd() using the new API.
Move the enum swd_special_seq in arm_adi_v5.h to solve a circular
dependencies among swd.h and arm_adi_v5.h

Change-Id: I9db13a00f129761eab283783c094cfff2dd92610
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/4902
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

tcl: update scripts after "jtag_reset" got deprecated

Avoid annoying "deprecated" messages in the scripts
distributed with OpenOCD code.

Change-Id: I82d27cd420db30f0653efbd286a627ef56a8c1fd
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5287
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

jtag: replace command "jtag_reset" with "adapter [de]assert"

Replace the JTAG transport specific command with a more generic
one. Deprecate "jtag_reset" and update the documentation.
While there, fix an error in the documentation, where the command
"jtag_reset" was used in place of command "reset_config".

Change-Id: I41a988d37ce69f7b35a960cbaf5306aab0299b99
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5286
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

adapter: add command "adapter [de]assert srst|trst [[de]assert srst|trst]"

Inspired from http://openocd.zylin.com/#/c/3720/1

Add commands to control the adapter's signals srst and trst.
Add macros for the flag's values assert/deassert to make clear what
they mean and to propose a uniform set of values across the code.

Change-Id: Ia8b13f4ded892942916cad7bda49540a896e7218
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5277
Tested-by: jenkins

jtag_vpi: multiple improvements

- Fix: Proper handling of read_socket() and write_socket()
in case of "partial" read/write.

- Added low-level JTAG IO debug capability (_DEBUG_JTAG_IO_)

- Zero-fill packet buffers, avoid sending pieces of uninitialized
memory over the network (memset struct vpi_cmd)

- Use close_socket() instead of close() - needed for Win32

- Fixed usage messages of jtag_vpi_command_handlers

Change-Id: I8bd19bc5c9512fe8e798600212e8a95213f50f5b
Signed-off-by: Jan Matyas <matyas@codasip.com>
Reviewed-on: http://openocd.zylin.com/5177
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor/stm32f1x: fix options reading from locked chip

In commit cea40152f82f52bfc718c7bda9fa4a9d70bf9cfd option bytes
reading was changed to direct access to option bytes area.
While there are no problems with stm32f0xx and stm32f3xx chips,
option block (0x1ffff800..0x1ffff80F) is unreadable from locked
stm32f10x chips.
As a result, stm32f1x unlock command writes dirty values to user
options, user data and write protection bits.

Option bytes reading reverted from direct access to option bytes area
to reading currently loaded bytes from  FLASH_OBR/FLASH_WRPR registers.

Tested on stm32f100, stm32f103, stm32f107 as well as on stm32f030 and
stm32f303.

Change-Id: Iad476351ffdaca5ace12e02272dacea7f3d08f52
Signed-off-by: Oleksandr Redchuk <real@real.kiev.ua>
Reviewed-on: http://openocd.zylin.com/4940
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

contrib/rpc_examples: Add (dis)connect methods

Add these methods such that the OpenOcd class can also be used outside
of a 'with' statement.

Change-Id: I927c93fff2dc05cc74daa56360a7262e736a639f
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/5189
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

rtos/FreeRTOS: Fix FreeRTOS thread list reading

Change-Id: I749ae94ec7279907b1905c02ecc1e9661f43ef70
Signed-off-by: Rahul Masurkar <rahulgm@marvell.com>
Reviewed-on: http://openocd.zylin.com/5273
Tested-by: jenkins
Reviewed-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor/kinetis: fix clang static analyzer warnings

Use assert to remove "Dereference of null pointer" warnings.

Change-Id: Ie204c234a71758e6470351e1d9f22da3dd887f56
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5357
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

target/arm_cti: add new 'ack' and 'channel' commands

these commands have been introduced to ease the manipulation of CTI trough
script files, these commands are:
 - $cti_name ack $event : to acknowledge a CTI event
 - $cti_name channel $channel_number $operation: to perform an operation
    on a specific channel, the possible operations are:
    gate, ungate, set, clear and pulse

Change-Id: I35463867a3c85072f3776c3aeb1e5788953ec435
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/5315
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins

flash/nor/stm32h7: (minor) use proper variable names and enhance logging

-fsize_base should be fsize_addr as it is the address of FLASH_SIZE register
-flash_base should be flash_regs_base to avoid confusion with flash block start
-add LOG_ERROR to functions stm32x_[read|write]_flash_reg(...)

Change-Id: I86f035314bcd616fc0bdf445692d945a85c15481
Signed-off-by: Tarek BOCHKATI <tarek.bouchkati@gmail.com>
Reviewed-on: http://openocd.zylin.com/5362
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

tcl/target: swm050: fix to allow to use with ST-Link

Currently the code assumes the adapter uses raw SWD, and the expected ID
code of the CPU is even wrong. An adapter speed is also not specified.
All these prevents the config file to be used with ST-Link.

Fix the config file, to allow it to be used with ST-Link.

Change-Id: I1244320fabfe8ee23da5a56a592dbeddc72cc8d5
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
Reviewed-on: http://openocd.zylin.com/5297
Tested-by: jenkins
Reviewed-by: Caleb Szalacinski <contact@skiboy.net>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

target/armv7m_trace: Improve SWO frequency auto-detection

The SWO frequency auto-detection with J-Link adapters does not work
properly in the current implementation. This is because the trace layer
has only information about the highest possible SWO frequency supported
by the adapter. With that the trace layer calculates the SWO prescaler
which usually leads to a frequency deviation greater than what is
permitted by J-Link adapters.

Move the calculation of the SWO prescaler from the trace layer into the
trace configuration of the adapter to overcome this problem.
The adapter has the necessary information to choose a suitable SWO
frequency and calculate the corresponding prescaler that complies with
the maximum allowed frequency deviation.

Tested with:
  - STM32L152RC Discovery Kit (ST-Link)
  - EFM32GG-STK3700 (J-Link)

Change-Id: I38ff2b89d32f0a92c597989b590afe5c75cf4902
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/3903
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>

helper/command: remove unused field in struct command_registration

The field jim_handler_data in struct command_registration is never
assigned, thus has always value NULL. It was added in commit
17a9dea53a71 ("add jim_handler to command_registration") on Nov 23
2009, together with the homonym field jim_handler_data in struct
command, but never used since then.
Only the field jim_handler_data in struct command is used.

Remove the field from struct command_registration and use NULL
where it was referenced (or remove the assignment if the recipient
is already zero, e.g. allocated with calloc()).

Removing the field decreases the total size of OpenOCD binary by
only 4944 byte on a 64 bit x86. Not a significant improvement from
this point of view.

Change-Id: I9f1d281e3de6b2eb398e2d883c5e9ff92628aecd
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5225
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

target/armv4_5: use c99 array designator to init arm_core_regs[]

During code analysis and development, counting again and again the
lines to find the index of the register is a boring error-prone
brain-damaging activity.
Use the c99 syntax and add once forever the array designators to
specify the index values.
The code behavior is not changed.

Change-Id: I2c70f70794475679efb91a8dfadc00f50715bd3f
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5256
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

sysfsgpio: give time to udev to change gpio permission

When a gpio is exported by writing in /sys/class/gpio/export, the
corresponding gpio control files appear immediately in sysfs but
with default access permission for root user only. The daemon udev
requires some time to get notified of the new files before it can
change the permissions to allow access to unprivileged users.
Due to this race condition, sysfsgpio can fail with EACCES error
if OpenOCD is executed by any unprivileged user.

Give 0.5 seconds to udev to identify the new files and change the
permission.

Tested with udev rules:
SUBSYSTEM=="gpio*", PROGRAM="/bin/sh -c 'find -L /sys/class/gpio/ -maxdepth 2 -exec chown root:uucp {} \; -exec chmod g=u {} \; || true'"

Change-Id: I1316c66ff103ffe23e5e4720f33372dc272a3766
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5302
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

gdb_server: fix extended_protocol for multi-target

The flag extended_protocol is currently a single static variable
thus, in case of multiple targets, it is shared among all the gdb
connections. This is an issue if the gdb connections are not all
using extended protocol, but also when one connection get closed
because the code sets the flag to zero impacting the other
connections still open.

Move the flag extended_protocol in the per-connection struct
gdb_connection.

Change-Id: I19d565f925df6a31767fd8d392242f60867109f2
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5310
Tested-by: jenkins
Reviewed-by: Moritz Fischer <moritzf@google.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

gdb_server: fix string length with semihosting_fileio

The GDB file-I/O remote protocol extension, used for implementing
the semihosting file I/O, requires the length of strings to
include the trailing zero character, as explicitly stated inside a
comment in GDB source code [1]:
/* 1. Parameter: Ptr to pathname / length incl. trailing zero.  */

ARM specification for semihosting [2] requires the string length
to not include the trailing zero character, e.g. in SYS_OPEN
specifications:
"field 3: An integer that gives the length of the string
 pointed to by field 1. The length does not include the
 terminating null character that must be present."

The mismatch above requires OpenOCD to add "one" to the string
length before passing it to GDB. Such conversion is missing
either in the generic semihosting provider of the data, the
function semihosting_common(), and in the consumer of the data,
the gdb_server function gdb_fileio_reply().
The conversion is already implemented in the target specific
function nds32_get_gdb_fileio_info(), but it's not the preferred
place for such GDB specific requirement.

This issue affects the semihosting calls "open", "unlink",
"rename" and "system".

Remove the "+1" conversion from nds32_get_gdb_fileio_info().
Add the "+1" conversion in gdb_fileio_reply().

[1] http://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;f=gdb/remote-fileio.c;h=11c141e42c4d#l381
[2] "Semihosting for AArch32 and AArch64, Release 2.0"
    https://static.docs.arm.com/100863/0200/semihosting.pdf

Change-Id: I35461bcb30f734fe2d51f7f0d418e3d04b4af506
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/5322
Tested-by: jenkins
Reviewed-by: Steven Stallion <sstallion@gmail.com>
Reviewed-by: Muhammad Omair Javaid <omair.javaid@linaro.org>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

flash/nor/stm32f2x: fix clang static analyzer warning

Using a signed variable as a parameter of FLASH_SNB() macro
generated "warning: The result of the left shift is undefined
because the left operand is negative"

Change-Id: I8b3fe840f9308962460906097df6ddd848c07b25
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5356
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>

jtag_vpi: ensured constant packet size & endianness

Made sure that size and endianness of jtag_vpi structures sent
over the network is the same, regardless of the platform.

Little endian chosen to maintain as much compatibility
with existing OpenOCD builds as possible.

Matching change in the original jtag_vpi server:
https://github.com/fjullien/jtag_vpi/pull/4

Change-Id: Ib839fea9bb2d5190b5643c970b89333b286dce71
Signed-off-by: Jan Matyas <matyas@codasip.com>
Reviewed-on: http://openocd.zylin.com/5152
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Tested-by: jenkins

flash/nor/nrf5: remove check for protected sector

The new "Access control list" flash protection scheme used in nRF52840
is not yet supported. Do not prevent sector erase if protection
state is unknown.

Change-Id: Iae9a869a54ffbdc888fb3ec478dafb5c942d9ea0
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/5348
Tested-by: jenkins

server/tcl_server.c: Fix buffer overrun

The input buffer size is checked only after writing past its end.

Change-Id: I6a9651c5b7d82efe338468d67bf6caca41004b01
Signed-off-by: Jimmy <nhminus@gmail.com>
Reviewed-on: http://openocd.zylin.com/5352
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>

server: gdb_server: fix memory map generation on a 32-bit BE host

Due to lack of printf format check wrong specifier was used and it
actually broke operation on a 32-bit BE host.

So fix this and add the necessary function attributes so that the bugs
like that can be uncovered automaticaly.

Reported and pinpointed by Karl Palsson on IRC.

Change-Id: I254ec28fcd9bb30594d607f74a6dba5456c2c7a1
Tested-by: Karl Palsson <karlp@tweak.net.au>
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-on: http://openocd.zylin.com/5342
Tested-by: jenkins
Reviewed-by: Karl Palsson <karlp@tweak.net.au>
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)