]> git.sur5r.net Git - openocd/log
openocd
6 years agoaarch64: fix crash on single-stepping
Matthias Welwarsky [Wed, 26 Jul 2017 09:23:19 +0000 (11:23 +0200)]
aarch64: fix crash on single-stepping

Make sure to skip un-examined targets (e.g. targets with deferred
examination) on single-stepping

fixes bug #158

Change-Id: Ib40c6fee03079fd078bd52a84615e6f8a081d128
Signed-off-by: Matthias Welwarsky <matthias.welwarsky@sysgo.com>
Reviewed-on: http://openocd.zylin.com/4184
Tested-by: jenkins
Reviewed-by: Yao Qi <qiyaoltc@gmail.com>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
6 years agoarm_adi_v5: reduce some CSW writes
Tomas Vanek [Thu, 15 Jun 2017 09:03:32 +0000 (11:03 +0200)]
arm_adi_v5: reduce some CSW writes

MEM-AP access through banked data registers MEM_AP_REG_BD0..3
does not increment TAR regardless of the current autoincrement mode.
mem_ap_read_u32() and mem_ap_write_u32() can keep the current
autoincrement mode instead of switching autoincrement off.

Change-Id: Ib7ec688d3e04f1da678363cd2819ce90e8910e58
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4163
Tested-by: jenkins
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-by: Christopher Head <chead@zaber.com>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agoarm_adi_v5: fix wrong addressing after change of CSW_ADDRINC
Tomas Vanek [Thu, 15 Jun 2017 06:59:01 +0000 (08:59 +0200)]
arm_adi_v5: fix wrong addressing after change of CSW_ADDRINC

Problem: If the same memory location is accessed alternatively
by MEM-AP banked data registers without autoincrement and by standard
autoincremented read/write, TAR register is not updated correctly.

How to replicate: On a Cortex-M issue
    mdw 0xe000edf0
multiple times. When poll is on (poll reads the same memory location)
only the first read is correct.
0xe000edf0: 01000000
0xe000edf0: 00000000
0xe000edf0: 20002640
0xe000edf0: 01000000
0xe000edf0: 00000000
0xe000edf0: 00000000

No problems with poll off.
0xe000edf0: 01000000
0xe000edf0: 01000000
0xe000edf0: 01000000

mem_ap_setup_tar() writes to MEM_AP_REG_TAR if requested TAR value
changed or CSW_ADDRINC_... is currently active.
However if an autoincremented access has been issued and autoinc
switched off in CSW afterwards, TAR does not get updated.

The change introduces mem_ap_update_tar_cache() which is called
after queuing of any access to MEM_AP_REG_DRW. It simulates
TAR increment to keep tar_value in sync with MEM_AP.
Crossing tar autoincrement block boundary invalidates cached value.

mem_ap_write() and mem_ap_read() do not check tar autoincrement
block boundary, mem_ap_setup_tar() is called before each transfer instead.

dap_invalidate_cache() is introduced to ensure invalidation
of all cached values during dap_dp_init() and swd_connect()

Change-Id: I815c2283d2989cffd6ea9a4100ce2f29dc3fb7b4
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4162
Tested-by: jenkins
Reviewed-by: Christopher Head <chead@zaber.com>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agotarget aarch64: rework memory read/write to use 8/16/32 bit operations
Bas Vermeulen [Sun, 26 Nov 2017 21:31:55 +0000 (22:31 +0100)]
target aarch64: rework memory read/write to use 8/16/32 bit operations

The existing code only used Memory Access mode to read memory,
which uses 32 bit operations only.
Rework the code to check the alignment/size of the read/write operation,
and use the Memory Access mode to read aligned 32 bit memory.
When using unaligned access, or 8 or 16 bit reads, use LDR{BHW} and STR{BHW}
instead.

The exception handling is still the same as it was before (meaning it breaks
when things go wrong), but I can now read an 8 bit register correctly.

Change-Id: I739a5ee825c0226ed4a89c32895cc2a047b8dc15
Signed-off-by: Bas Vermeulen <bas@daedalean.ai>
Reviewed-on: http://openocd.zylin.com/4301
Tested-by: jenkins
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
6 years agojtag/drivers/cmsis-dap: fix connect under reset
Tomas Vanek [Sun, 9 Apr 2017 08:59:57 +0000 (10:59 +0200)]
jtag/drivers/cmsis-dap: fix connect under reset

Commit ef02b69b14d133b061217a91add5a028a77e86bc included
a call to cmsis_dap_cmd_DAP_Connect() before calling
cmsis_dap_cmd_DAP_SWJ_Sequence(). According to comment
it is necessary for at least Keil ULINK-ME.

Commit 72c3464be42088dc75245cf2fcc8f5c6e6959b4b added
a cmsis_dap_cmd_DAP_Disconnect() before connect call to pair
connection/disconnection. It solves some problems on Atmel EDBG.

Unfortunately calling either of cmsis_dap_cmd_DAP_Connect()
or cmsis_dap_cmd_DAP_Disconnect() deasserts reset signal.
So these workarounds break ability to connect under reset.

Use cmsis_dap_cmd_DAP_Disconnect() and cmsis_dap_cmd_DAP_Connect()
pair only if both SRST and TRST are deasserted.

Change-Id: I0914dae0a1360b8c7fe48231ff3867caedfb2dbe
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reported-by: Leonardo Sabino dos Santos <leonardo.sabino@gmail.com>
Reviewed-on: http://openocd.zylin.com/4100
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
6 years agoconfig for ESPRESSObin from Globalscale Tech. Inc.
Jiri Kastner [Sat, 16 Dec 2017 21:21:14 +0000 (22:21 +0100)]
config for ESPRESSObin from Globalscale Tech. Inc.

Change-Id: I77f536a9d2e901ebcef0a7dd0f205e5332b1d382
Signed-off-by: Jiri Kastner <cz172638@gmail.com>
Reviewed-on: http://openocd.zylin.com/4303
Tested-by: jenkins
Reviewed-by: Forest Crossman <cyrozap@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
6 years agoconfigs for Marvell Armada 3700
Jiri Kastner [Sat, 16 Dec 2017 21:17:52 +0000 (22:17 +0100)]
configs for Marvell Armada 3700

Change-Id: I367f39c9bc9e58380d6d5b500d5368d5173d96bd
Signed-off-by: Jiri Kastner <cz172638@gmail.com>
Signed-off-by: Forest Crossman <cyrozap@gmail.com>
Reviewed-on: http://openocd.zylin.com/4302
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
6 years agotarget: remove unused event definitions
Tomas Vanek [Tue, 31 Oct 2017 11:32:22 +0000 (12:32 +0100)]
target: remove unused event definitions

Events reset-halt-pre, reset-halt-post, reset-wait-pre and
reset-wait-post are not used anywhere.

Change-Id: I9a0f94875b102d9b08f6c2fd9d73a9f05f8e8e79
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4285
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agoflash/nor/stm32f2x: fix erase on STM32F413/423
Tomas Vanek [Thu, 23 Nov 2017 13:47:37 +0000 (14:47 +0100)]
flash/nor/stm32f2x: fix erase on STM32F413/423

Theese devices do not have a gap in sector numbering.
The driver translates sectors numbers 12 13... to 16 17... as used on dual
bank flash devices. Therefore erase of sector 12 and above fails with error
'stm32x device protected'
on F413/423.

Drop sector number translation for devices without has_large_mem flag.

Change-Id: I65531c0dfe02e2fd0f3d68f0615e0926e9901391
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4299
Tested-by: jenkins
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
6 years agoflash/nor/stm32f2x: fix protection block size for F767 in dual bank mode
Tomas Vanek [Thu, 26 Oct 2017 16:00:33 +0000 (18:00 +0200)]
flash/nor/stm32f2x: fix protection block size for F767 in dual bank mode

A protection block comprises two adjacent sectors in dual bank mode.
As there are 64 and 128kB sectors joined in blocks 2 and 8, block size
should be computed as a sum of sector sizes.

Change-Id: Ie915df8cf7ca232c4565d7e0c514c8933e71fdfe
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4271
Tested-by: jenkins
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
6 years agodoc: improve stm32 flash driver documentation
Spencer Oliver [Thu, 7 Dec 2017 12:30:35 +0000 (12:30 +0000)]
doc: improve stm32 flash driver documentation

also remove legacy footnote as it adds no value.

Change-Id: I3892acf244bd8fba6f844a5d82a66004e193a395
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/4309
Tested-by: jenkins
Reviewed-by: Karl Palsson <karlp@tweak.net.au>
6 years agojtag: drivers: stlink: handle all versions with single config
Paul Fertser [Thu, 26 Jan 2017 20:41:40 +0000 (23:41 +0300)]
jtag: drivers: stlink: handle all versions with single config

Extend HLA interface to allow multiple VID/PID pairs and use it to
autodetect the connected stlink version.

Change-Id: I35cd895b2260e23cf0e8fcb1fc11a78c2b99c69b
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3961
Tested-by: jenkins
Reviewed-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
6 years agoconfig: stm32l01x and stm32l02x chips support
elmot [Thu, 7 Dec 2017 08:08:28 +0000 (10:08 +0200)]
config: stm32l01x and stm32l02x chips support

New low-end chips have only 2k of RAM, workarea size adjusted

Change-Id: Ibfccd73fef9e6dabffc87d901736c5626ce411fe
Signed-off-by: Ilia Motornyi <elijah.mot@gmail.com>
Reviewed-on: http://openocd.zylin.com/4308
Tested-by: jenkins
Reviewed-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agodocs: add missing stm32 flash driver documentation
Spencer Oliver [Wed, 6 Dec 2017 22:04:49 +0000 (22:04 +0000)]
docs: add missing stm32 flash driver documentation

Change-Id: I433780646e6fdfd0c2527b4a68025946ccb79d8b
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/4307
Tested-by: jenkins
Reviewed-by: <alexandre.torgue@st.com>
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
6 years agostm8 : new target
Ake Rehnman [Mon, 6 Nov 2017 18:56:28 +0000 (19:56 +0100)]
stm8 : new target

New STM8 target based mostly on mips4k. Target communication
through STLINK/SWIM. No flash driver yet but it is still possible
to program flash through load_image command. The usual target debug
methods are implemented.

Change-Id: I7216f231d3ac7c70cae20f1cd8463c2ed864a329
Signed-off-by: Ake Rehnman <ake.rehnman@gmail.com>
Reviewed-on: http://openocd.zylin.com/3953
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
6 years agoAdd STM32H7 config files
Alexandre Torgue [Mon, 13 Nov 2017 16:00:58 +0000 (17:00 +0100)]
Add STM32H7 config files

Add 2 target files:
-stm32h7x.cfg
-stm32h7x_dual_bank.cfg

Add 2 config files for:
-STM32H743zi-nucleo bord
-STM32H743i and STM32H753i eval boards.

Change-Id: I2aae2c5acff4f3ff8e1bf232fda5a11a87f71703
Signed-off-by: Alexandre TORGUE <alexandre.torgue@st.com>
Reviewed-on: http://openocd.zylin.com/4182
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
6 years agoflash: Add new stm32h7x driver support
Alexandre Torgue [Mon, 13 Nov 2017 13:05:28 +0000 (14:05 +0100)]
flash: Add new stm32h7x driver support

Add basic support for:
     -STM32H7x (Embedded flash 2M)

Erase and write tested on stm32h743.

Change-Id: Ie8d8786227cdeee39fcf5663167a053ad8dcef4c
Signed-off-by: Rémi Prud'homme <remi.prudhomme@st.com>
Signed-off-by: Alexandre TORGUE <alexandre.torgue@st.com>
Reviewed-on: http://openocd.zylin.com/4181
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
6 years agoOnly call cmsis_dap_cmd_DAP_SWD_Configure when swd_mode is enabled
Bas Vermeulen [Tue, 21 Nov 2017 16:12:24 +0000 (17:12 +0100)]
Only call cmsis_dap_cmd_DAP_SWD_Configure when swd_mode is enabled

The CMSIS-DAP used by NXP's LS1012ARDB board only supports JTAG,
and not SWD. Calling cmsis_dap_cmd_DAP_SWD_Configure returns with an
error (and doesn't actually do anything in the debugger).

Wrap the call to cmsis_dap_cmd_DAP_SWD_Configure in a check for
swd_mode, to make sure initialisation doesn't fail needlessly.

Change-Id: Id7e568cb6e36886bd7c5b3699d198a77a51c28c9
Signed-off-by: Bas Vermeulen <bas@daedalean.ai>
Reviewed-on: http://openocd.zylin.com/4294
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Matthias Welwarsky <matthias@welwarsky.de>
6 years agospi: add n25q256 flash
Robert Jordens [Fri, 19 Aug 2016 21:49:49 +0000 (05:49 +0800)]
spi: add n25q256 flash

* 256 MBit SPI flash
* https://www.micron.com/~/media/documents/products/data-sheet/nor-flash/serial-nor/mt25q/die-rev-a/mt25q_qljs_l_256_aba_0.pdf spells out the entire zoo of IDs
* used e.g. on Xilinx KCU105

Change-Id: I18b19292b4869627adb9071266271962fec68fb4
Signed-off-by: Robert Jordens <jordens@gmail.com>
Reviewed-on: http://openocd.zylin.com/4186
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
6 years agoserver/gdb: Use 'bool' instead of 'int' for boolean values
Marc Schink [Sun, 29 Oct 2017 18:54:00 +0000 (19:54 +0100)]
server/gdb: Use 'bool' instead of 'int' for boolean values

Change-Id: I71c2f2553a29e9ef167ff3313cc06c7b31c64190
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4278
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agoserver/gdb: Use get_target_from_connection()
Marc Schink [Thu, 22 Sep 2016 20:36:28 +0000 (22:36 +0200)]
server/gdb: Use get_target_from_connection()

Change-Id: I2c66bf6da734a3b71e358553943e9fc3c6578c39
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4277
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agotarget: Constify parameter of is_armv7m()
Marc Schink [Sun, 29 Oct 2017 14:58:41 +0000 (15:58 +0100)]
target: Constify parameter of is_armv7m()

Change-Id: Ieea1b0dec88818e9e8d5c8c5d54aa8959556d77b
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4275
Tested-by: jenkins
Reviewed-by: Christopher Head <chead@zaber.com>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agortos: Use 'bool' as return type for detect_rtos()
Marc Schink [Sun, 29 Oct 2017 10:39:51 +0000 (11:39 +0100)]
rtos: Use 'bool' as return type for detect_rtos()

Change-Id: I91ad0431d44ed94f48d20c4690f8642d66f52a9b
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4274
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
6 years agoadi_v5_swd: Add error message when SWD fails to connect
Jonas Norling [Wed, 25 Oct 2017 08:38:49 +0000 (10:38 +0200)]
adi_v5_swd: Add error message when SWD fails to connect

Error message instead of failing silently.

Change-Id: Ie54a5bf68459d3c0e96cc38080ffad8de0a4b5ce
Signed-off-by: Jonas Norling <jonas.norling@cyanconnode.com>
Reviewed-on: http://openocd.zylin.com/4269
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoftdi: Enable SWDIO output before sending data on it
Jonas Norling [Wed, 25 Oct 2017 09:33:06 +0000 (11:33 +0200)]
ftdi: Enable SWDIO output before sending data on it

The SWDIO buffer has to be enabled, by setting SWDIO_OE, for data on
SWDIO to reach the target. Explicitly do this before sending the
switch sequences for JTAG-to-SWD, etc.

This makes the code insensitive to the state of SWDIO_OE specified in
ftdi_layout_init. It used to work only on adapters with a non-inverted
SWDIO_OE inited to 1, or inverted SWDIO_OE inited to 0.

Change-Id: I4b9e520ac1c7ce2a437251a05fc036bc68de718e
Signed-off-by: Jonas Norling <jonas.norling@cyanconnode.com>
Reviewed-on: http://openocd.zylin.com/4270
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
7 years agotcl/interface/ftdi/openrd: Fix FTDI channel + device description
Jonathan McDowell [Fri, 13 Oct 2017 06:53:32 +0000 (07:53 +0100)]
tcl/interface/ftdi/openrd: Fix FTDI channel + device description

Similar to the Sheevaplug fix inf95f8b70fbd0f7e9c91a2d9006b1abb2dd07ebf2
the OpenRD device has its JTAG interface on the first channel of the
ft2232, which is 0 for the new driver but was 1 for the old one. Correct
the config file appropriately. Also the device description was missing
the trailing " B" and thus not picking up the device correctly. Finally
add an adapter_khz setting in the OpenRD board configuration file - set
to 2MHz to match the Sheeva variant.

Confirmed as working thanks to Phil Hands providing me access to his
hardware to test on.

See also Debian Bug#793214; https://bugs.debian.org/793214

Change-Id: Ifacf53124eaa330bbbdf36dfa79e3256bf2a5201
Signed-off-by: Jonathan McDowell <noodles@earth.li>
Reviewed-on: http://openocd.zylin.com/4254
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoCortex-M: Delete an unnecessary local variable
Christopher Head [Tue, 3 Oct 2017 20:23:57 +0000 (13:23 -0700)]
Cortex-M: Delete an unnecessary local variable

The dhcsr_save variable was used to save the value of
cortex_m->dcb_dhcsr so it could be restored later. However, all writes
in between the save and the restore use mem_ap_write_atomic_u32, not
cortex_m_write_debug_halt_mask, which means cortex_m->dcb_dhcsr isn’t
changed anyway. Delete the unnecessary local.

Change-Id: I064a3134e21398e1ecfc9f1fa7efd7b020b52341
Signed-off-by: Christopher Head <chead@zaber.com>
Reviewed-on: http://openocd.zylin.com/4240
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoCortex-M: fix stale DHCSR cache values
Christopher Head [Tue, 3 Oct 2017 19:59:06 +0000 (12:59 -0700)]
Cortex-M: fix stale DHCSR cache values

In cortex_m_assert_reset, in two locations, DHCSR is written directly
using mem_ap_write_u32. This means that the cached version,
target_to_cm(target)->dcb_dhcsr, is not updated when these writes are
performed, so subsequent writes to DHCSR that use
cortex_m_write_debug_halt_mask will change those bits back to their old
values which, unless modified in that particular invocation, come from
the cache. This causes an actual, observable bug on an STM32F7 in which
running “reset run” immediately after “program” can in some cases result
in execution proceeding with C_MASKINTS set (it is cleared on line 1021
but is then set immediately afterward in cortex_m_clear_halt), causing
failure of the application. Replace these mem_ap_write_u32 calls with
cortex_m_write_debug_halt_mask calls to do the same jobs.

Change-Id: Id35ca7f6057c2df2ba9cd67c53a73b50816d0b71
Signed-off-by: Christopher Head <chead@zaber.com>
Reviewed-on: http://openocd.zylin.com/4239
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoFix GCC7 warnings about string truncation
Freddie Chopin [Thu, 29 Jun 2017 21:49:03 +0000 (23:49 +0200)]
Fix GCC7 warnings about string truncation

GCC7 with -Wall warns about possible string truncation with
snprint()-type functions with "directive output may be truncated writing
1 byte into a region of size between 0 and 9
[-Werror=format-truncation=]" + "note: ‘snprintf’ output between 5 and
14 bytes into a destination of size 12" (or similar). Fix this by
increasing sizes of buffers.

See https://gcc.gnu.org/gcc-7/changes.html

Change-Id: Ib848f2a56dd658783534158947ae1be7c0e99d45
Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/4175
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Tested-by: jenkins
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>
7 years agoFix GCC7 warnings about switch-case fallthroughs
Freddie Chopin [Thu, 29 Jun 2017 21:48:19 +0000 (23:48 +0200)]
Fix GCC7 warnings about switch-case fallthroughs

GCC7 with -Wextra warns about switch-case blocks which fallthrough with
"this statement may fall through [-Werror=implicit-fallthrough=]". This
can be fixed by adding "special" comments: "/* fallthrough */".

See https://gcc.gnu.org/gcc-7/changes.html

Change-Id: Iba0be791dbdd86984489b2d9a0592bb59828da1e
Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/4174
Tested-by: jenkins
7 years agoarm: semihosting: set command line arguments
Andreas Fritiofson [Wed, 28 Dec 2016 22:43:53 +0000 (23:43 +0100)]
arm: semihosting: set command line arguments

Add "arm semihosting_cmdline [argv0 argv1 ...]" for setting the
command line arguments for the debuggee.

[andreas.fritiofson@gmail.com]: Dynamic allocation, empty default

Change-Id: I831ddd161d602f251940e29608a154e9590fdee1
Signed-off-by: Christian Groessler <chris@groessler.org>
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3106
Tested-by: jenkins
7 years agoDifferentiate "target not halted" messages.
Tim Newsome [Fri, 6 Oct 2017 21:36:34 +0000 (14:36 -0700)]
Differentiate "target not halted" messages.

Change-Id: I8728fa007289d7af5c6791142e76eb5777c83ca4
Signed-off-by: Tim Newsome <tim@sifive.com>
Reviewed-on: http://openocd.zylin.com/4244
Tested-by: jenkins
Reviewed-by: Liviu Ionescu <ilg@livius.net>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
7 years agoDocument `struct reg` fields.
Tim Newsome [Sun, 8 Oct 2017 19:03:25 +0000 (12:03 -0700)]
Document `struct reg` fields.

Change-Id: I286316079e2e4d4f09427a4ffbecadb48c5dc9d9
Signed-off-by: Tim Newsome <tim@sifive.com>
Reviewed-on: http://openocd.zylin.com/4250
Tested-by: jenkins
Reviewed-by: Liviu Ionescu <ilg@livius.net>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
7 years agotcl: add hi3798 target and Tocoding Poplar board config
Peter Griffin [Mon, 12 Jun 2017 15:28:03 +0000 (16:28 +0100)]
tcl: add hi3798 target and Tocoding Poplar board config

This config covers the 4x Cortex A53 CPUs. A custom connector
is required from J14 to standard ARM JTAG on v1 boards. However
v2 hardware should have a standard FTSH-105-01-L-DV connector.

Pinmuxing code to enable JTAG pins is included in l-loader-poplar
repository, so board is flashed with open source code, JTAG
is available at very early boot. Alternatively the following
pokes can be issued from U-Boot to enable JTAG (e.g. to debug
hisilicon SDK).

mw 0xf8a210ec 0x130;
mw 0xf8a210f0 0x130;
mw 0xf8a210f4 0x130;
mw 0xf8a210f8 0x130;
mw 0xf8a210fc 0x130;
mw 0xf8a21100 0x130;

Change-Id: I2b83dfcb3dc5461c1620f94dd99aa7b31fdda59b
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Reviewed-on: http://openocd.zylin.com/4161
Tested-by: jenkins
Reviewed-by: Jiri Kastner <cz172638@gmail.com>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
7 years agonrf5: Add nRF52832-QFAA support
Slowcoder [Thu, 31 Aug 2017 20:34:16 +0000 (22:34 +0200)]
nrf5: Add nRF52832-QFAA support

Change-Id: Ica9e34e873cac182662b1e32a9b3164dbc0c935f
Signed-off-by: Slowcoder <slowcoder@gmail.com>
Reviewed-on: http://openocd.zylin.com/4210
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agonrf51: Rename to nrf5
Slowcoder [Thu, 31 Aug 2017 19:09:42 +0000 (21:09 +0200)]
nrf51: Rename to nrf5

Renaming of all nrf51 NOR flash code to nrf5, as to prepare the
code for being able to flash nrf51 and nrf52 chips.

The nrf51 command is retained for backwards compatability.

"nRF5" is also the name Nordic Semiconductor uses to describe
both the nrf51 and nrf52 chips.

Change-Id: I5f4e3f1ec780184b28ad44f735a746e68908c502
Signed-off-by: Slowcoder <slowcoder@gmail.com>
Reviewed-on: http://openocd.zylin.com/4209
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agonrf51: Refactor device-list
Slowcoder [Thu, 31 Aug 2017 18:53:57 +0000 (20:53 +0200)]
nrf51: Refactor device-list

This cleans up the list of supported nrf51 chips considerably.

Change-Id: Ic74685657bb72a8703c0a49df4c48c54604ec2a7
Signed-off-by: Slowcoder <slowcoder@gmail.com>
Reviewed-on: http://openocd.zylin.com/4208
Tested-by: jenkins
Reviewed-by: Fredrik Hederstierna <fredrik@hederstierna.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agostm32lx.c: Read IDcode at appropriate address.
Uwe Bonnes [Mon, 14 Nov 2016 18:12:38 +0000 (19:12 +0100)]
stm32lx.c: Read IDcode at appropriate address.

Trying to read the L0 idcode at the L1 idcode address 0xE0042000 often
resulted in an uncatched error. Reading at the right L0 address 0x40015800
afterwards results in reading 0. So access to the device is denied..

Change-Id: I6de92cf99a5d5d46c72f9ba055613cbc5753a951
Signed-off-by: Uwe Bonnes <bon@elektron.ikp.physik.tu-darmstadt.de>
Reviewed-on: http://openocd.zylin.com/3883
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agotarget: atheros_ar9344: add simple uart0 test
Oleksij Rempel [Sun, 6 Aug 2017 07:50:05 +0000 (09:50 +0200)]
target: atheros_ar9344: add simple uart0 test

in some cases we need something to test if uart is actually
properly connected.

Change-Id: I5a16b053164b34bb30ae8370753be12887a85c51
Signed-off-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-on: http://openocd.zylin.com/4194
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoboard: add TP-Link WDR4300 config
Oleksij Rempel [Sat, 15 Jul 2017 15:53:37 +0000 (17:53 +0200)]
board: add TP-Link WDR4300 config

tested on TP-Link WDR4300 v1.7

Change-Id: If2b456afac835ab3a3987d434d20824c7ba75b93
Signed-off-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-on: http://openocd.zylin.com/4192
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agotarget: add atheros_ar9344.cfg
Oleksij Rempel [Sat, 15 Jul 2017 15:52:01 +0000 (17:52 +0200)]
target: add atheros_ar9344.cfg

Change-Id: I005b4c78ccb0fec8d38a25430cb49c580dcd8df5
Signed-off-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-on: http://openocd.zylin.com/4191
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoboard: tp-link_tl-mr3020: add ath79 support
Oleksij Rempel [Sun, 6 Aug 2017 05:58:31 +0000 (07:58 +0200)]
board: tp-link_tl-mr3020: add ath79 support

Finally we can use this driver by default!

Change-Id: I09d215d1bd1dc16873a7379637e6869af65ad8f1
Signed-off-by: Oleksij Rempel <linux@rempel-privat.de>
Reviewed-on: http://openocd.zylin.com/4193
Tested-by: jenkins
Reviewed-by: Dmytro <dioptimizer@hotmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoserver: Add port number to socket bind error
Robert Foss [Wed, 7 Jun 2017 20:49:29 +0000 (16:49 -0400)]
server: Add port number to socket bind error

Make this error message more useful by providing the port number
that we tried to bind to.

Change-Id: Ieb18adf0725a6ae99c77ebfaadc49d64ed407bbe
Signed-off-by: Robert Foss <robert.foss@collabora.com>
Reviewed-on: http://openocd.zylin.com/4157
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agotcl/interface/ftdi: improve minimodule config
Diego Herranz [Thu, 3 Aug 2017 06:37:41 +0000 (07:37 +0100)]
tcl/interface/ftdi: improve minimodule config

- Tested on a real FT2232H MiniModule, so warning removed.
- Every pin initially set to high impedance except TCK, TDI,
  TDO and TMS: Safest values given it's an evaluation board
  and the rest of pins might be connected to something else.
- Reset is now initially de-asserted (it was asserted
  which is not recommended).
- nRST pin choice is arbitrary so comment added (wondering
  if it should be an "echo").
- "-oe" option added to NRST signal so it can be set as
  high impedance (tri-stated).

Change-Id: I967ab0c7bbccf72dbf6d6d78b3180b74e016e0d6
Signed-off-by: Diego Herranz <diegoherranz@diegoherranz.com>
Reviewed-on: http://openocd.zylin.com/4185
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoflash: efm32: add support for EFR-familty (e.g. bluegecko)
Andrea Merello [Wed, 1 Mar 2017 10:32:07 +0000 (11:32 +0100)]
flash: efm32: add support for EFR-familty (e.g. bluegecko)

This patch adds support for Blue Gecko and Mighty Gecko chips from
Silabs.

They have different EFM32_MSC_REGBASE and LOCK register offset.

Based on the original patch from Andreas Kemnade.

Change-Id: I166c14960ced7c880b68083badd1b31372fefabe
Cc: Andreas Kemnade <andreas@kemnade.info>
Signed-off-by: Andrea Merello <andrea.merello@gmail.com>
Reviewed-on: http://openocd.zylin.com/4034
Reviewed-by: Jonas Norling <jonas.norling@cyanconnode.com>
Tested-by: jenkins
Reviewed-by: Fredrik Hederstierna <fredrik@hederstierna.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: chrysn <chrysn@fsfe.org>
7 years agotcl/interface/ftdi/sheevaplug: Fix FTDI channel configuration
Jonathan McDowell [Sun, 27 Aug 2017 18:12:21 +0000 (19:12 +0100)]
tcl/interface/ftdi/sheevaplug: Fix FTDI channel configuration

The migration from the old ft2232 driver to the new generic ftdi driver
ended up breaking support for the SheevaPlug device. The old driver
defaulted to channel 1, but numbered the channels 1 to 4. The new driver
starts at 0. The SheevaPlug JTAG is on interface A (interface B is the
serial console), so it should be using channel 0. Fix this. Confirmed
as working; serial console remains available and a new u-boot image can
be transferred across using the JTAG link.

See also Debian Bug#837989; https://bugs.debian.org/837989

Change-Id: I4ac2bfeb0d1e7e99d70fa47dc55f186e6af2c542
Signed-off-by: Jonathan McDowell <noodles@earth.li>
Reviewed-on: http://openocd.zylin.com/4206
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Tested-by: jenkins
7 years agoAdd missing break
Fabio Utzig [Fri, 11 Aug 2017 16:17:03 +0000 (13:17 -0300)]
Add missing break

Change-Id: Ie1de679fe6ab5ace05fc3e156c71f34b296b3d3b
Signed-off-by: Fabio Utzig <utzig@apache.org>
Reviewed-on: http://openocd.zylin.com/4200
Tested-by: jenkins
Reviewed-by: Andreas Färber <afaerber@suse.de>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: Esben Haabendal <esbenhaabendal@gmail.com>
7 years agostm32f2x: Fix left shift of negative value
Marc Schink [Thu, 24 Aug 2017 15:42:28 +0000 (17:42 +0200)]
stm32f2x: Fix left shift of negative value

Use unsigned constant for left shift operation in order to avoid the
following error with GCC >= 6.0:

../src/flash/nor/stm32f2x.c: In function ‘stm32x_handle_unlock_command’:
../src/flash/nor/stm32f2x.c:1324:67: error: left shift of negative value [-Werror=shift-negative-value]
   stm32x_info->option_bytes.optcr2_pcrop = OPTCR2_PCROP_RDP | (~1 << bank->num_sectors);

Change-Id: I0ac082bd0dbb8dc2f61ffff8fdf486ab7962d2e0
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4207
Tested-by: jenkins
Reviewed-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: Anton Fosselius <anton.fosselius@gmail.com>
Reviewed-by: Esben Haabendal <esbenhaabendal@gmail.com>
7 years agoFix typo in comment.
Tim Newsome [Thu, 9 Jun 2016 16:23:10 +0000 (09:23 -0700)]
Fix typo in comment.

Change-Id: I6567f85f399315e1dac98881765dfaa6eab5facb
Signed-off-by: Tim Newsome <tim@sifive.com>
Reviewed-on: http://openocd.zylin.com/4238
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agojlink: Disable TCP/IP discovery
Marc Schink [Sat, 2 Sep 2017 08:16:41 +0000 (10:16 +0200)]
jlink: Disable TCP/IP discovery

If no serial number is specified, disable TCP/IP device discovery to
ensure that a user does not unintentionally operate on a remote device.

Change-Id: I6a7e913b8b679fae003825468cd86d2014849b29
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4214
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agojlink: Disable automatic device selection
Marc Schink [Sat, 2 Sep 2017 08:10:00 +0000 (10:10 +0200)]
jlink: Disable automatic device selection

If multiple devices are attached, do not automatically use the first
device found. Otherwise, a user may unintentionally operate on the
wrong device.

Change-Id: I08c4110b82e911e9e3e744d41830ffc6c56c44bf
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4213
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agojlink: Use error description in log messages
Marc Schink [Sat, 2 Sep 2017 07:50:44 +0000 (09:50 +0200)]
jlink: Use error description in log messages

Use a human-readable error description rather than just the error name
in log messages.

Change-Id: Iab4ff7a7e4d9993983a07eab9f462820d4ee8190
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4212
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agojlink: Make libusb optional
Marc Schink [Sat, 20 May 2017 12:34:21 +0000 (14:34 +0200)]
jlink: Make libusb optional

libusb is now optional for libjaylink because support for TCP/IP
devices is always available.

Change-Id: I03f2566f8e1703276671ac0f353f72394d21f2f0
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4204
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agojlink: Make use of debug level for I/O messages
Marc Schink [Wed, 26 Jul 2017 21:51:22 +0000 (23:51 +0200)]
jlink: Make use of debug level for I/O messages

Change-Id: Iba08c119a80041f9c1b4c9bd7e83bb4f9bbb7199
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4203
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agolibjaylink: Update to latest Git version
Marc Schink [Fri, 28 Jul 2017 15:34:07 +0000 (17:34 +0200)]
libjaylink: Update to latest Git version

Update to latest Git version and bump required libjaylink package
version to 0.2.0.

This version introduces support for devices with TCP/IP interface (e.g.
SEGGER Flasher ARM) and an additional debug level for I/O messages.

Change-Id: I030236aa704a91d1bb1843dd30010865947747e0
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4202
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agohelper/options: Add missing #include for MinGW and MSYS2
Marc Schink [Mon, 21 Aug 2017 18:56:43 +0000 (20:56 +0200)]
helper/options: Add missing #include for MinGW and MSYS2

Change-Id: I3bb295f52706b641661241e3e047306811ca915e
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/4201
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoconfigure: bring back all default JimTcl extensions
Paul Fertser [Sun, 11 Jun 2017 13:18:27 +0000 (16:18 +0300)]
configure: bring back all default JimTcl extensions

This partially reverts commit 56d163ce79510c7756567df00ae54155757eaa0f.

That change certainly caused more pain than gain.

Change-Id: Ifb126abd1e6b89d29db8bf6a7b8af5dfc815c163
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-on: http://openocd.zylin.com/4159
Tested-by: jenkins
Reviewed-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agotelnet_server: increase buffer sizes to allow longer commands.
Karl Palsson [Wed, 10 May 2017 14:37:51 +0000 (14:37 +0000)]
telnet_server: increase buffer sizes to allow longer commands.

A common use case seen in the wild is echoing a string of commands to an
existing openocd instance via netcat.  The sequence of ; separated
commands can easily run over the line limit of only 256 chars.

Increasing this dramatically reduces surprises, at the expense of a tiny
amount of extra ram usage.

Change-Id: I2389d99d316a96b5fa03f0894b43c412308e12c4
Signed-off-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-on: http://openocd.zylin.com/4132
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agotelnet_server: drop unused options
Karl Palsson [Wed, 10 May 2017 14:31:18 +0000 (14:31 +0000)]
telnet_server: drop unused options

They're never used, so just drop them.

Change-Id: Ie137deed3e7258f9d6af7e0cb508e73df0f53ee0
Signed-off-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-on: http://openocd.zylin.com/4131
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoSupport for STM32F722, F723, F413 and F423
Andreas Bolsch [Sun, 5 Mar 2017 18:01:06 +0000 (19:01 +0100)]
Support for STM32F722, F723, F413 and F423

IDs for STM32F722, F723, F413 and F423 added, handling of PCROP
for F722/723 and additional nWPRT bits for F413/423 implemented.
The additional protection bit positions for F413/423 conflict
with other options bits for the F7xx variants, additionally the
last two sectors share a common bit.

Protection for F413 and F767/777 now use protection blocks
rather sectors for dealing with protections bits.

Checking for halted state in 'lock' and 'unlock' removed: When
PCROP is activated in F723, halted state is not detected properly,
but lock/unlock sequence is required to disable PCROP.

Tested with STM32F723E-Disco, STM32F413ZH-Nucleo.

Change-Id: Ie6ddab47a9ae8461087d369b4f289b7f9d1e031c
Signed-off-by: Andreas Bolsch <hyphen0break@gmail.com>
Reviewed-on: http://openocd.zylin.com/4045
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
7 years agolog: Add a new debug level (4) for verbose I/O debug
Andreas Fritiofson [Sun, 2 Apr 2017 10:54:43 +0000 (12:54 +0200)]
log: Add a new debug level (4) for verbose I/O debug

Change ftdi SWD driver and CMSIS-DAP to use it instead of LOG_DEBUG().

Change-Id: I17ba3de2086c7159209db61fba3faf067dfc5023
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/3805
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agostm32l1: Devid 0x429 only has 8bit flash size register
Karl Palsson [Wed, 9 Aug 2017 17:11:29 +0000 (17:11 +0000)]
stm32l1: Devid 0x429 only has 8bit flash size register

A footnote in RM0038r14.

Change-Id: Ic31894d846fbbe917a7290b2b7ff8fb582bb65da
Signed-off-by: Karl Palsson <karlp@tweak.net.au>
Reviewed-on: http://openocd.zylin.com/4198
Tested-by: jenkins
Reviewed-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
7 years agoflash Kinetis: fix probe for FlexNVM partitioned as EEPROM backup
Tomas Vanek [Thu, 13 Jul 2017 19:35:22 +0000 (21:35 +0200)]
flash Kinetis: fix probe for FlexNVM partitioned as EEPROM backup

If a MCU has FlexNVM partitioned as EEPROM backup only
(no data flash), kinetis_probe_chip() detects zero fcfg2_maxaddr1
and adjusts flash banks count to 1, what is obviously wrong.

The change limits the test to devices without FlexNVM.

Computation of program flash/FlexNVM blocks is now more robust.

Missing case 0x07 is added to switch (fcfg1_depart)

Change-Id: I0bd6030a0fe1ab62aeb0223bbdf2aee1505bf6a0
Reported-by: simon.haines@scalardata.com
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4180
Tested-by: jenkins
Reviewed-by: Simon Haines <simon.haines@scalardata.com>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
7 years agoflash Kinetis: fix devices with smallest program flash (8 and 16 kB)
Tomas Vanek [Thu, 29 Jun 2017 21:06:55 +0000 (23:06 +0200)]
flash Kinetis: fix devices with smallest program flash (8 and 16 kB)

Change-Id: I2692b9877a7f877104528f279a69e8cc1cfbcdbf
Reported-by: David Miller Lowe <milhead@gmail.com>
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4173
Tested-by: jenkins
Reviewed-by: Miller Lowe <miller.lowe@trailtech.net>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
7 years agomips32: inline functions in headers must be static
Andreas Fritiofson [Thu, 6 Jul 2017 07:51:40 +0000 (09:51 +0200)]
mips32: inline functions in headers must be static

Change-Id: If1d0fc6766cadc2db33408ae5c0968de6b7a1b94
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/4178
Tested-by: jenkins
Reviewed-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-by: Stian Skjelstad <stian@nixia.no>
7 years agortos: better sanity checking for uCOS-III
Steven Stallion [Thu, 22 Jun 2017 05:14:08 +0000 (22:14 -0700)]
rtos: better sanity checking for uCOS-III

This patch improves the OSRunning check. If the rtos_running check
fails, update_threads will return an error rather than attempt to update
the thread list using bad values.

Change-Id: I8614c325504d3a9ab19aebb6862b1fe445a0c8e7
Signed-off-by: Steven Stallion <stallion@squareup.com>
Reviewed-on: http://openocd.zylin.com/4166
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
7 years agortos: style corrections for uCOS-III
Steven Stallion [Wed, 3 May 2017 17:17:45 +0000 (12:17 -0500)]
rtos: style corrections for uCOS-III

This patch corrects a number of style infractions in RTOS support for
uC/OS-III. These were missed during initial review last year prior to
the 0.10.0 release.

Change-Id: Ia2139f6ca381d4087fd8ee989f7a03ac474d7440
Signed-off-by: Steven Stallion <stallion@squareup.com>
Reviewed-on: http://openocd.zylin.com/4120
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
7 years agozynq_7000: Add expected id for Zynq 7z100 devices
Moritz Fischer [Fri, 30 Jun 2017 18:54:38 +0000 (11:54 -0700)]
zynq_7000: Add expected id for Zynq 7z100 devices

As found on the NI Project Sulfur SDR board.

Change-Id: I47bdd38ae85cf45cedad8797ea03bf3105153320
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
Reviewed-on: http://openocd.zylin.com/4176
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash/nor/efm32: Support EZR32HG devices.
Richard Watts [Mon, 2 Jan 2017 11:56:36 +0000 (11:56 +0000)]
flash/nor/efm32: Support EZR32HG devices.

Recognise the family number for Silicon Labs EZR32HG devices and
select the correct flash page size.

Change-Id: I876e930f3a9f679557fa0d0acac33e9bbfb28c46
Signed-off-by: Richard Watts <rrw@kynesim.co.uk>
Reviewed-on: http://openocd.zylin.com/3934
Tested-by: jenkins
Reviewed-by: Fredrik Hederstierna <fredrik@hederstierna.com>
Reviewed-by: Jonas Norling <jonas.norling@cyanconnode.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agotarget: Fix snprintf format string and argument mismatch in md output
Andreas Fritiofson [Thu, 22 Jun 2017 12:23:58 +0000 (14:23 +0200)]
target: Fix snprintf format string and argument mismatch in md output

Commit 47b8cf842 changed the fixed type of the value argument to snprint
but didn't change the format string to match for sizes != 64 bit.

Change-Id: I908b06f49ab69d04224282949190a0de883048e0
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/4167
Tested-by: jenkins
Reviewed-by: Stian Skjelstad <stian@nixia.no>
Reviewed-by: Philipp Guehring <pg@futureware.at>
7 years agoflash/nor/tcl: Make read_bank parameters optional
Marc Schink [Fri, 4 Nov 2016 08:19:21 +0000 (09:19 +0100)]
flash/nor/tcl: Make read_bank parameters optional

Make 'offset' and 'length' parameters optional, if both are omitted
simply read the whole flash bank.

Additionally, check if the 'offset' and 'length' arguments are out of
bounds of the flash bank.

Change-Id: Ib9c1b0538a2c78ebcf702e2da11468dff407f8ff
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/3862
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoflash/nor/tcl: Respect flash bank boundary in write_bank
Marc Schink [Fri, 4 Nov 2016 08:02:11 +0000 (09:02 +0100)]
flash/nor/tcl: Respect flash bank boundary in write_bank

Respect the flash bank boundary and write only to the remaining part of
the bank even if the file content is larger.

Change-Id: I8f4c1b161c103a77bdb30c6bf052293b5ed48c41
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/3861
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoflash/nor/tcl: Make write_bank parameter optional
Marc Schink [Fri, 4 Nov 2016 07:39:03 +0000 (08:39 +0100)]
flash/nor/tcl: Make write_bank parameter optional

Make the 'offset' parameter optional, if omitted simply start at the
beginning of the flash bank.

Additionally, check if the argument is out of bounds of the flash bank.

Change-Id: I8e9632b539ad9e83211e1ac6a06da4c8109cbc60
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/3860
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoflash/nor/tcl: Respect flash bank boundary in verify_bank
Marc Schink [Fri, 4 Nov 2016 07:18:45 +0000 (08:18 +0100)]
flash/nor/tcl: Respect flash bank boundary in verify_bank

Respect the flash bank boundary and compare only the remaining content
of the bank even if the file content is larger.

Change-Id: I4d75979c7893fdd4d18372fa6b0321a0486b4fa9
Signed-off-by: Marc Schink <openocd-dev@marcschink.de>
Reviewed-on: http://openocd.zylin.com/3859
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoflash Kinetis: Add support for newer KW series
Joakim Nohlgård [Tue, 11 Apr 2017 07:55:41 +0000 (09:55 +0200)]
flash Kinetis: Add support for newer KW series

Add support for flashing newer members of the NXP Kinetis KW family

Supported devices:
 - KW20Z
 - KW30Z
 - KW40Z
 - KW21Z
 - KW31Z
 - KW41Z

The earlier KW2xD and KW01Z devices are already supported by the code
for the older K-series.

Verified working on the FRDM-KW41Z development board.
Tested flashing both via GDB `load` and directly via OpenOCD flash
write commands.

Change-Id: I73eae477127a8b54a33005b3b526b5439450a808
Signed-off-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-on: http://openocd.zylin.com/4104
Tested-by: jenkins
Reviewed-by: Johann Fischer <johann_fischer@posteo.de>
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agoflash Kinetis: add KL28 device
Tomas Vanek [Fri, 10 Mar 2017 20:43:46 +0000 (21:43 +0100)]
flash Kinetis: add KL28 device

This device differs a lot from others in KL series.

Unfortunately the System Integration Module, where device
identification resides, moved to a new address so probe now have
to try both addresses of SIM_SDID.

Introduce a new bank creation option: -sim-base to ensure error free probe.

WDOG32 is slightly different from KE1x and on different address.
System Mode Controler changed layout to word aligned.

Change-Id: I2c9dca0c4ad4228fcc941d6078d15f5e394833ff
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/4059
Tested-by: jenkins
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash Kinetis: handle all types of watchdog, disable in reset-init
Tomas Vanek [Mon, 26 Dec 2016 18:59:10 +0000 (19:59 +0100)]
flash Kinetis: handle all types of watchdog, disable in reset-init

Active watchdog forces reset during armv7m_checksum_memory()
in verify_image command if run just after reset init.

COP watchdog in KL series and WDOG32 in KE1 series
have longer timeout however they need to be disabled too.

The change extends 'kinetis disable_wdog' command to optionally
probe the chip and use appropriate algorithm to disable watchdog.

Setting of cache type is also split from flash_support flags.

Tcl command 'kinetis disable_wdog' is called in reset-init event.

Change-Id: I3191e230f38b679ed74f2a97fe323ef8fb3fe22e
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3901
Tested-by: jenkins
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash Kinetis: implement automatic bank creation based on device probe
Tomas Vanek [Mon, 26 Dec 2016 14:20:33 +0000 (15:20 +0100)]
flash Kinetis: implement automatic bank creation based on device probe

Kinetis flash driver services huge number of MCU types. They have
one, two or four flash banks with option of FlexNVM. It would
require ~36 config files just for Kx series, more for KLx, KVx and KE1x.

The change implements alternative approach:
- configuration file creates just one pflash bank (common for all devices)
- when a device is probed, additional pflash or flexnvm banks are created
based on flash layout of the connected MCU
- created banks have names with optional numbering e.g. kx.pflash0 kx.pflash1
kx.flexnvm0 kx.flexnvm1
- the first bank gets renamed if numbering is used

Automatic bank creation is enabled by tcl command 'kinetis create_banks'.

Used solution has a drawback: other banks than pflash0 are not accessible
until pflash0 is probed. Fortunately gdb attach and standard programming
accesses banks in right sequence.

Change-Id: I5b9037cbefdb8a4176b7715fbcc3af4da4c1ab60
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3925
Tested-by: jenkins
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash Kinetis: split kinetis_chip from kinetis_flash_bank
Tomas Vanek [Mon, 26 Dec 2016 09:15:06 +0000 (10:15 +0100)]
flash Kinetis: split kinetis_chip from kinetis_flash_bank

Kinetis flash driver probed and decoded chip repeatedly for each flash
bank. Bank ordering used global bank number so multi-target
configuration was broken.

The change introduces kinetis_probe_chip() which reads SIM SDID
and SIM FCFG registers, decodes Kinetis series and family
and fills struct kinetis_chip. This probe runs once for all banks.

struct kinetis_chip contains pointers to all flash banks embeded
in the MCU. It simplifies iteration over all or specific MCU banks.

kinetis_probe_chip() generates MCU name and some informational messages
are improved.

Change-Id: I990db5c63ba490667eec0e5459086d83936662fb
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3924
Tested-by: jenkins
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash Kinetis: add KV5x family
Tomas Vanek [Sat, 3 Dec 2016 11:14:10 +0000 (12:14 +0100)]
flash Kinetis: add KV5x family

Real time control MCU has a Cortex-M7 and numerous changes in flash layout.
Introduced a new ID of MDM-AP.

While on it a LOG_DEBUG format error fixed.

Change-Id: I1018660ce0c3dd63ac5e2563408fabff3c3daef7
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3900
Tested-by: jenkins
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash Kinetis: add KL8x family, fix erase check
Tomas Vanek [Fri, 2 Dec 2016 14:47:01 +0000 (15:47 +0100)]
flash Kinetis: add KL8x family, fix erase check

Secure devices KL81Z7 and KL82Z7 have no SERIESID field in ID register
so they have to be decoded in Kx branch (not KLx).

The flash controller in KL8x and also in K8x devices does not implement
FTFx_CMD_BLOCKSTAT command. Fix kinetis_blank_check() to work properly
using FTFx_CMD_SECTSTAT command only.
Introduce a new flag FS_NO_CMD_BLOCKSTAT to avoid use of FTFx_CMD_BLOCKSTAT
on these devices.

Change-Id: I3ff58718480acd8cce69f618f71667b6b1d9c4f3
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3898
Tested-by: jenkins
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agotcl/target: make sure kex.cfg is not used for Kinetis KE1x families
Tomas Vanek [Fri, 2 Dec 2016 17:09:27 +0000 (18:09 +0100)]
tcl/target: make sure kex.cfg is not used for Kinetis KE1x families

Config file renamed to ke0x.cfg and a notice added to texi.

While on ke0x.cfg CPUTAPID setting fixed: device has SWD port only, no JTAG.

Removed per device configs as they set CHIPNAME and nothing more.
Let's use reasonably universal chip name 'ke' set in family config.

Change-Id: I313db87a59f25f968eb3c27df155780b67becee8
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3897
Tested-by: jenkins
Reviewed-by: Ivan Meleca <ivan@artekit.eu>
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash Kinetis: add KE1xZ and KE1xF families
Tomas Vanek [Wed, 30 Nov 2016 20:48:59 +0000 (21:48 +0100)]
flash Kinetis: add KE1xZ and KE1xF families

The new Kinetis KE1x families use FTFE flash controller unlike KE0x.
Also SDID coding corresponds to new K, KL and KV families.
That's why KE1x is handled by kinetis driver instead of kinetis_ke

Change-Id: Ibb73e28e41dfbb086e761e1f006b089825dab854
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3896
Tested-by: jenkins
Reviewed-by: Joakim Nohlgård <joakim.nohlgard@eistec.se>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash/nor: at91samd modified to use real erase sector size
Tomas Vanek [Thu, 14 Jul 2016 18:52:13 +0000 (20:52 +0200)]
flash/nor: at91samd modified to use real erase sector size

Before this change SAMD driver defined "sector" equal to a flash
protection block. Oversize sectors (16kB for the biggest flash size)
made problems for flashing firmware split to two or more parts.

Removed superfluous test of sector protection before erase.

Change-Id: I8e6a6bda6ccd91eda2df67ec48270c69faa1bdd1
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3546
Tested-by: jenkins
Reviewed-by: Stian Skjelstad <stian@nixia.no>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash/nor/at91sam4: remove FWS=6, rename at91samg to atsamg
Tomas Vanek [Mon, 26 Dec 2016 21:53:44 +0000 (22:53 +0100)]
flash/nor/at91sam4: remove FWS=6, rename at91samg to atsamg

FWS=6 workaround removed, as this appears to be a copy-paste error
from the SAM3X family. Originally addressed in http://openocd.zylin.com/3837
but not all occurences were removed.

Atmel changed chip naming and removed 91 prefix for atsamg, samd...

Change-Id: Ia2b43da82b2ff9b1c85fdb456a0a198ab095243d
Signed-off-by: Tomas Vanek <vanekt@fbl.cz>
Reviewed-on: http://openocd.zylin.com/3926
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agoflash: nor: ath79: fix build failure due to recent MIPS changes
Paul Fertser [Wed, 31 May 2017 13:57:34 +0000 (16:57 +0300)]
flash: nor: ath79: fix build failure due to recent MIPS changes

Change-Id: I7139b0658f048afea2d16216c93e8946356a630d
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
Reviewed-on: http://openocd.zylin.com/4151
Tested-by: jenkins
Reviewed-by: Salvador Arroyo <sarroyofdez@yahoo.es>
7 years agoflash: Add support for Atheros (ath79) SPI interface
Tobias Diedrich [Mon, 1 Aug 2016 13:57:20 +0000 (15:57 +0200)]
flash: Add support for Atheros (ath79) SPI interface

Supported SoCs: AR71xx, AR724x, AR91xx, AR93xx, QCA9558

Extended and revised version of my original patch submitted by Dmytro
here: http://openocd.zylin.com/#/c/3390

This driver is using pure SPI mode, so the flash base address is not
used except some flash commands (e.g. "flash program") need it to
distinguish the banks.

Example config with all 3 chip selects:
flash bank flash0 ath79 0 0 0 0 $_TARGETNAME cs0
flash bank flash1 ath79 0x10000000 0 0 0 $_TARGETNAME cs1
flash bank flash2 ath79 0x20000000 0 0 0 $_TARGETNAME cs2

Example usage:
> flash probe flash0
Found flash device 'win w25q128fv' (ID 0x001840ef)
flash 'ath79' found at 0x00000000
> flash probe flash1
No SPI flash found
> flash probe flash2
No SPI flash found
> flash banks
> flash read_bank flash0 /tmp/test.bin 0x00000000 0x1000
reading 4096 bytes from flash @0x00000000
wrote 4096 bytes to file /tmp/test.bin from flash bank 0 at offset
0x00000000 in 28.688066s (0.139 KiB/s)

Change-Id: I5feb697722c07e83a9c1b361a9db7b06bc699aa8
Signed-off-by: Tobias Diedrich <ranma+openocd@tdiedrich.de>
Reviewed-on: http://openocd.zylin.com/3612
Tested-by: jenkins
Reviewed-by: Dmytro <dioptimizer@hotmail.com>
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
7 years agoimx_gpio: add mmap based jtag interface for IMX processors
Grzegorz Kostka [Tue, 18 Apr 2017 14:12:58 +0000 (16:12 +0200)]
imx_gpio: add mmap based jtag interface for IMX processors

For some targets (like nrf51) sysfs driver is too slow. This
patch implements memory maped driver for IMX processors.
Mostly based on bcm2835gpio. Tested on imx6ul CPU. However, it should
work on any NXP IMX CPU.

Change-Id: Idace4c98181c6e9c64dd158bfa52631204b5c4a7
Signed-off-by: Grzegorz Kostka <kostka.grzegorz@gmail.com>
Reviewed-on: http://openocd.zylin.com/4106
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
7 years agojtag/drivers: Add Cypress KitProg driver
Forest Crossman [Sat, 30 Jan 2016 05:23:49 +0000 (00:23 -0500)]
jtag/drivers: Add Cypress KitProg driver

This patch adds a driver for the SWD-only Cypress KitProg
programmer/debugger.

Change-Id: I3a9a8011a762781d560ebb305597e782a4f9a8e5
Signed-off-by: Forest Crossman <cyrozap@gmail.com>
Reviewed-on: http://openocd.zylin.com/3221
Tested-by: jenkins
Reviewed-by: Tomas Vanek <vanekt@fbl.cz>
7 years agotcl/cpld: add config file for Altera 5M570Z CPLD (MAXV family)
Moritz Fischer [Wed, 26 Apr 2017 05:18:49 +0000 (22:18 -0700)]
tcl/cpld: add config file for Altera 5M570Z CPLD (MAXV family)

Change-Id: I229c746be27b7c4fa01f48a6ed54ab2679b50ab1
Signed-off-by: Moritz Fischer <moritz.fischer@ettus.com>
Reviewed-on: http://openocd.zylin.com/4109
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>
7 years agostm32l4: support flashing L45x/46x devices
Juha Niskanen [Thu, 4 May 2017 10:43:08 +0000 (13:43 +0300)]
stm32l4: support flashing L45x/46x devices

Also fixes incorrect comment about MSI range.

Change-Id: If1339a00e50db44195dfcd5c767ba3f5d9035451
Signed-off-by: Juha Niskanen <juha.niskanen@haltian.com>
Reviewed-on: http://openocd.zylin.com/4122
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32, use scan32 function for reading impcode/idcode.
Salvador Arroyo [Mon, 20 Feb 2017 19:45:37 +0000 (20:45 +0100)]
mips32, use scan32 function for reading impcode/idcode.

There is no need to implement scan code in functions
mips_ejtag_get_idcode/impcode(), use mips_ejtag_drscan_32().
Impcode/idcode saved in ejtag.info.
Reorder the code in the callers of this functions.

Change-Id: Ia829c783a0b24c6a65cade736113fa6f67b0a170
Signed-off-by: Salvador Arroyo <salvador@telecable.es>
Reviewed-on: http://openocd.zylin.com/4003
Tested-by: jenkins
Reviewed-by: Peter Mamonov <pmamonov@gmail.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32: add micromips breakpoints support
Salvador Arroyo [Mon, 8 May 2017 07:44:50 +0000 (09:44 +0200)]
mips32: add micromips breakpoints support

Breakpoint setting based on length (kind) only.
Added 16bit aligned 32bit software breakpoints
support and same filtering before setting
breakpoint.
Set the required isa bit in hardware breakpoints.
Drop the isa bit in software breakpoints.

Change-Id: I7020f27be16015194b76f385d9b8e5af496d0dfc
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4124
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32: add micromips isa handling
Salvador Arroyo [Sun, 7 May 2017 16:39:17 +0000 (18:39 +0200)]
mips32: add micromips isa handling

Read and save configuration registers, up to 4.
Config3 holds the micromips implementation info.
Added isa implementation info to mips32_common.
Added isa filter to avoid common mistakes, but only
if one isa mode is implemented.
When resuming the isa requested is set if more than
one isa mode is implemented.

Change-Id: I1d6526c5525bffac8d75e031b842b2edc6310e28
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4123
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32, convert miniprograms with code definition
Salvador Arroyo [Sun, 7 May 2017 11:36:12 +0000 (13:36 +0200)]
mips32, convert miniprograms with code definition

Needed to run in micromips mode. Seems that if an isa
is supported in debug mode it also supported in kernel
mode. The contrary is not true.

Change-Id: I1feb8e2c376f4db97089f05c20bc0cd177208fb3
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4033
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32, add support for micromips in debug mode
Salvador Arroyo [Sun, 7 May 2017 09:58:25 +0000 (11:58 +0200)]
mips32, add support for micromips in debug mode

Micromips is 16bit oriented, branch and jumps are
16 bit based. The upper half 16bits of a 32bit instruction
with the major opcode, must go first in the instruction
stream, hence the SWAP16 macro and swap16 array function,
needed if the code is written as 32 bit word in little endian
cores. Endianess info added to ejtag_iinfo. Pointer to
ejtag_info and isa field added to pracc context.
MIPS32 code are renamed to MIPS32_ISA_...
To select the isa, the new code has an additional isa parameter
(1 for micromips, 0 for mips32).
In JR instruction the isa bit must be set to execute
micromips code.
The suffix u is added to the OP codes to avoid signed/unsigned
comparison errors and to make sure the right shift is
performed logically.
The isa in debug mode is updated in the poll function.
Code for miniprograms, in kernel mode, need to be converted.
CFI code only for mips32.

Change-Id: I79a8b637d49b0e2d92b6dd5eb5aa8aa0520bf938
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4032
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32, add microMips instruction subset
Salvador Arroyo [Mon, 27 Feb 2017 12:06:46 +0000 (13:06 +0100)]
mips32, add microMips instruction subset

Only reencoded MIPS32 instructions. Added some instructions
for crc code. Micromips isa in debug mode is only needed for
pic32mm cores. Pic32mz seems that only works with MIPS32
isa when in debug mode.

Change-Id: I07059e153a7000ea9204f20b6b37edf6a7623455
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4022
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32, add option to avoid check in last instruction
Salvador Arroyo [Mon, 27 Feb 2017 11:25:49 +0000 (12:25 +0100)]
mips32, add option to avoid check in last instruction

This option is needed, for example, when exiting
debug mode in bmips targets. The last instruction
is a NOP, not a DERET. When working in async mode
this check is not done, mips32_pracc_queue_exec() pass
the parameter to mips32_pracc_exec() and never use it.

Change-Id: I4c7ed4feb1588b62e2645b955b501b6671113b36
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4021
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32, add realloc code
Salvador Arroyo [Sun, 26 Feb 2017 19:35:01 +0000 (20:35 +0100)]
mips32, add realloc code

If max_code is reached realloc memory. If fails to realloc
the error is propagated and every call to pracc_add() returns
immediately. The exec function logs the error.

Change-Id: Idd4ed9d9b8b19b7d6842d0bc5ebb05f943726705
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4020
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
7 years agomips32, change in pracc_list for dynamic allocation
Salvador Arroyo [Sun, 26 Feb 2017 08:11:02 +0000 (09:11 +0100)]
mips32, change in pracc_list for dynamic allocation

pracc_list points to an array with code in the lower half
and addr in the upper half. Change it to a struct with
an instruction field and an address field.
Requiered to make reallocation easier.
As a side effect the code is less quirky.

Change-Id: Ibf904a33a2f35a7f69284d2a2114f4b4ae79219f
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/4019
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>