boards/pba-d-01-kw2x: fix flashing from invalid state

[cladmi]

Contribution description

When in an invalid state, the board could not flash anymore.
Like for example after flashing tests/driver_adt7310 this makes it
flash again.

I used reset_config srst_only as the other kinetis boards are also
configured that way.

RIOT/boards/common/frdm/dist/old-openocd-kx.cfg

Lines 38 to 39 in c439346

	cortex_m reset_config sysresetreq
	#reset_config srst_only srst_nogate connect_assert_srst

Testing procedure

Thy flashing two or more times tests/driver_adt7310 compile first and run flash-only.

With this pull request it works without issues.

RIOT_CI_BUILD=1 BOARD=pba-d-01-kw2x make -C tests/driver_adt7310/ flash-only

RIOT_CI_BUILD=1 BOARD=pba-d-01-kw2x make -C tests/driver_adt7310/ flash-only
make: Entering directory '/home/harter/work/git/worktree/riot_master/tests/driver_adt7310'
/home/harter/work/git/worktree/riot_master/dist/tools/openocd/openocd.sh flash /home/harter/work/git/worktree/riot_master/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf
### Flashing Target ###
/home/harter/work/git/worktree/riot_master/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf is fine.
GNU MCU Eclipse 64-bit Open On-Chip Debugger 0.10.0+dev-00462-gdd1d90111 (2019-01-18-11:37)
Licensed under GNU GPL v2
For bug reports, read
        http://openocd.org/doc/doxygen/bugs.html
Info : auto-selecting first available session transport "swd". To override use 'transport select <transport>'.
Info : add flash_bank kinetis kx.pflash
adapter speed: 1000 kHz
none separate
cortex_m reset_config sysresetreq
srst_only separate srst_nogate srst_open_drain connect_assert_srst
Info : CMSIS-DAP: SWD  Supported
Info : CMSIS-DAP: FW Version = 1.0
Info : CMSIS-DAP: Interface Initialised (SWD)
Info : SWCLK/TCK = 0 SWDIO/TMS = 1 TDI = 0 TDO = 0 nTRST = 0 nRESET = 1
Info : Connecting under reset
Info : CMSIS-DAP: Interface ready
Info : clock speed 1000 kHz
Info : SWD DPIDR 0x2ba01477
Info : MDM: Chip is unsecured. Continuing.
Info : kx.cpu: hardware has 6 breakpoints, 4 watchpoints
Info : Listening on port 38973 for gdb connections
    TargetName         Type       Endian TapName            State
--  ------------------ ---------- ------ ------------------ ------------
 0* kx.cpu             cortex_m   little kx.cpu             reset
Info : MDM: Chip is unsecured. Continuing.
target halted due to debug-request, current mode: Thread
xPSR: 0x01000000 pc: 0x00000740 msp: 0x1fffc200
auto erase enabled
Info : Kinetis MK21DN512xxx5 detected: 2 flash blocks
Info : 2 PFlash banks: 512k total
wrote 16384 bytes from file /home/harter/work/git/worktree/riot_master/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf in 0.605979s (26.404 KiB/s)
34 bytes written at address 0x20000000
downloaded 34 bytes in 0.004076s (8.146 KiB/s)
target halted due to breakpoint, current mode: Thread
xPSR: 0x01000000 pc: 0x20000020 msp: 0x1fffc200
verified 14388 bytes in 0.437898s (32.087 KiB/s)
Info : MDM: Chip is unsecured. Continuing.
Error: Failed to write memory and, additionally, failed to find out where
Error: Failed to enable the FPB
Polling target kx.cpu failed, trying to reexamine
Error: MDM: failed to read ID register
Error: Could not find MEM-AP to control the core
Examination failed, GDB will be halted. Polling again in 100ms
shutdown command invoked
Done flashing
make: Leaving directory '/home/harter/work/git/worktree/riot_master/tests/driver_adt7310'

With master after the second time it fails:

Flash fails without error message

RIOT_CI_BUILD=1 BOARD=pba-d-01-kw2x make -C tests/driver_adt7310/ flash-only
make: Entering directory '/home/harter/work/git/worktree/riot_master/tests/driver_adt7310'
/home/harter/work/git/worktree/riot_master/dist/tools/openocd/openocd.sh flash /home/harter/work/git/worktree/riot_master/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf
### Flashing Target ###
/home/harter/work/git/worktree/riot_master/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf is fine.
GNU MCU Eclipse 64-bit Open On-Chip Debugger 0.10.0+dev-00462-gdd1d90111 (2019-01-18-11:37)
Licensed under GNU GPL v2
For bug reports, read
        http://openocd.org/doc/doxygen/bugs.html
Info : auto-selecting first available session transport "swd". To override use 'transport select <transport>'.
Info : add flash_bank kinetis kx.pflash
adapter speed: 1000 kHz
none separate
cortex_m reset_config sysresetreq
Info : CMSIS-DAP: SWD  Supported
Info : CMSIS-DAP: FW Version = 1.0
Info : CMSIS-DAP: Interface Initialised (SWD)
Info : SWCLK/TCK = 0 SWDIO/TMS = 1 TDI = 0 TDO = 0 nTRST = 0 nRESET = 1
Info : CMSIS-DAP: Interface ready
Info : clock speed 1000 kHz


/home/harter/work/git/worktree/riot_master/tests/driver_adt7310/../../Makefile.include:541: recipe for target 'flash-only' failed
make: *** [flash-only] Error 1
make: Leaving directory '/home/harter/work/git/worktree/riot_master/tests/driver_adt7310'

Issues/PRs references

Found while running the test suite on the board in #10870

[cladmi]

I think I should also test debugging. I put this as a reminder.

[fjmolinas]

This PR fixes the issue, but DEBUG is not working properly anymore on this branch, while it is on master. I'll take a closer look tomorrow.

[cladmi]

It still works for me for debugging:

BOARD=pba-d-01-kw2x make -C examples/hello-world/ flash debug
...
reset_handler_default () at /home/harter/work/git/RIOT/cpu/cortexm_common/vectors_cortexm.c:79
79      {
(gdb) b main
Breakpoint 1 at 0x10fe: file /home/harter/work/git/RIOT/examples/hello-world/main.c, line 26.
(gdb) monitor reset halt
MDM: Chip is unsecured. Continuing.
target halted due to debug-request, current mode: Thread
xPSR: 0x01000000 pc: 0x00000740 msp: 0x1fffc200
(gdb) monitor reset init
MDM: Chip is unsecured. Continuing.
target halted due to debug-request, current mode: Thread
xPSR: 0x01000000 pc: 0x00000740 msp: 0x1fffc200
Disabling Kinetis watchdog (initial WDOG_STCTRLH = 0x01d3)
WDOG_STCTRLH = 0x01d2
(gdb) c
Continuing.
Note: automatically using hardware breakpoints for read-only addresses.

Breakpoint 1, main () at /home/harter/work/git/RIOT/examples/hello-world/main.c:26
26          puts("Hello World!");

I could also debug examples/default.

[fjmolinas]

I executed this:

BOARD=pba-d-01-kw2x make -C examples/hello-world/ flash debug

on master:

### Starting Debugging ###
Open On-Chip Debugger 0.10.0+dev-00703-g92bb76a4-dirty (2019-07-19-14:27)
Licensed under GNU GPL v2
For bug reports, read
	http://openocd.org/doc/doxygen/bugs.html
Reading symbols from /home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf...
Remote debugging using :3333
pm_set_lowest () at /home/francisco/workspace/RIOT/cpu/cortexm_common/include/cpu.h:177
177	    irq_restore(state);

with this PR:

### Starting Debugging ###
Open On-Chip Debugger 0.10.0+dev-00703-g92bb76a4-dirty (2019-07-19-14:27)
Licensed under GNU GPL v2
For bug reports, read
	http://openocd.org/doc/doxygen/bugs.html
Reading symbols from /home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf...
Remote debugging using :3333
0x2000002e in ?? ()

[cladmi]

This one is now based on #11976

@fjmolinas would this work for your board right now? and if I replace the reset_config to reset_config trst_and_srst ?

With trst_and_srst it works on mine.

[cladmi]

Rebased now that #11976 is merged.

[fjmolinas]

@cladmi I'm still having the same issue, but as discussed IRL this could be an issue with my board. Maybe @smlng could test? I think he has some pba-d-01-kw2x. If not i'll try to find some time to see If I can find a working config for my board.

[cladmi]

@cladmi I'm still having the same issue, but as discussed IRL this could be an issue with my board. Maybe @smlng could test? I think he has some pba-d-01-kw2x. If not i'll try to find some time to see If I can find a working config for my board.

Does it also fail with trst_and_srst? because both configurations work on my board (it is not in a 'non flashable' state though).

diff --git a/boards/pba-d-01-kw2x/dist/openocd.cfg b/boards/pba-d-01-kw2x/dist/openocd.cfg
index 94a704212..1a5dbc6a2 100644
--- a/boards/pba-d-01-kw2x/dist/openocd.cfg
+++ b/boards/pba-d-01-kw2x/dist/openocd.cfg
@@ -1,3 +1,3 @@
 source [find target/kx.cfg]
-reset_config srst_only
+reset_config trst_and_srst
 $_TARGETNAME configure -rtos auto

[cladmi]

@fjmolinas indeed, there was no change as you mentioned in #12261 (comment) I had the diff to try in the previous comment. I pushed it to simplify testing :)

[fjmolinas]

@cladmi I'm trying to get it to work on my board ass wll but I still have the same issues as stated before. I tried to go back to the board documentation but it seems @PHYTEC-Messtechnik-GmbH has dropped there wiki entries for this board...

[fjmolinas]

@cladmi as I had told you IRL my board had some issue that could have explained why it wasn't working for me. I managed to fix them and know I have the following behaviour:

• Flashing tests/driver_adt7310/ twice suceeds:

RIOT_CI_BUILD=1 BOARD=pba-d-01-kw2x make -C tests/driver_adt7310/ flash

make: Entering directory '/home/francisco/workspace/RIOT/tests/driver_adt7310'
Building application "tests_driver_adt7310" for "pba-d-01-kw2x" with MCU "kinetis".

   text	   data	    bss	    dec	    hex	filename
  14516	    116	   2632	  17264	   4370	/home/francisco/workspace/RIOT/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf
/home/francisco/workspace/RIOT/dist/tools/openocd/openocd.sh flash /home/francisco/workspace/RIOT/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf
### Flashing Target ###
/home/francisco/workspace/RIOT/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf is not locked.
Open On-Chip Debugger 0.10.0+dev-00703-g92bb76a4-dirty (2019-07-19-14:27)
Licensed under GNU GPL v2
For bug reports, read
	http://openocd.org/doc/doxygen/bugs.html
Info : auto-selecting first available session transport "swd". To override use 'transport select <transport>'.
Info : add flash_bank kinetis kx.pflash
adapter speed: 1000 kHz
none separate
cortex_m reset_config sysresetreq
srst_only separate srst_nogate srst_open_drain connect_deassert_srst
srst_only separate srst_nogate srst_open_drain connect_assert_srst
Info : CMSIS-DAP: SWD  Supported
Info : CMSIS-DAP: FW Version = 1.0
Info : CMSIS-DAP: Interface Initialised (SWD)
Info : SWCLK/TCK = 0 SWDIO/TMS = 1 TDI = 0 TDO = 0 nTRST = 0 nRESET = 1
Info : Connecting under reset
Info : CMSIS-DAP: Interface ready
Info : clock speed 1000 kHz
Info : SWD DPIDR 0x2ba01477
Info : MDM: Chip is unsecured. Continuing.
Info : kx.cpu: hardware has 6 breakpoints, 4 watchpoints
Info : kx.cpu: external reset detected
Info : Listening on port 35111 for gdb connections
    TargetName         Type       Endian TapName            State       
--  ------------------ ---------- ------ ------------------ ------------
 0* kx.cpu             cortex_m   little kx.cpu             reset
Info : MDM: Chip is unsecured. Continuing.
target halted due to debug-request, current mode: Thread 
xPSR: 0x01000000 pc: 0x0000073c msp: 0x1fffc200
auto erase enabled
Info : Kinetis MK21DN512xxx5 detected: 2 flash blocks
Info : 2 PFlash banks: 512k total
wrote 16384 bytes from file /home/francisco/workspace/RIOT/tests/driver_adt7310/bin/pba-d-01-kw2x/tests_driver_adt7310.elf in 0.599877s (26.672 KiB/s)
34 bytes written at address 0x20000000
downloaded 34 bytes in 0.004033s (8.233 KiB/s)
target halted due to breakpoint, current mode: Thread 
xPSR: 0x01000000 pc: 0x20000020 msp: 0x1fffc200
verified 14632 bytes in 0.403877s (35.380 KiB/s)
Info : MDM: Chip is unsecured. Continuing.
Error: Failed to write memory and, additionally, failed to find out where
Error: Failed to enable the FPB
Polling target kx.cpu failed, trying to reexamine
Error: MDM: failed to read ID register
Error: Could not find MEM-AP to control the core
Examination failed, GDB will be halted. Polling again in 100ms
shutdown command invoked
Done flashing
make: Leaving directory '/home/francisco/workspace/RIOT/tests/driver_adt7310'

• flash debug works correctly:

.RIOT_CI_BUILD=1 BOARD=pba-d-01-kw2x make -C examples/hello-world/ flash debug

make: Entering directory '/home/francisco/workspace/RIOT/examples/hello-world'
Building application "hello-world" for "pba-d-01-kw2x" with MCU "kinetis".

   text	   data	    bss	    dec	    hex	filename
   9552	    116	   2564	  12232	   2fc8	/home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf
/home/francisco/workspace/RIOT/dist/tools/openocd/openocd.sh flash /home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf
### Flashing Target ###
/home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf is not locked.
Open On-Chip Debugger 0.10.0+dev-00703-g92bb76a4-dirty (2019-07-19-14:27)
Licensed under GNU GPL v2
For bug reports, read
	http://openocd.org/doc/doxygen/bugs.html
Info : auto-selecting first available session transport "swd". To override use 'transport select <transport>'.
Info : add flash_bank kinetis kx.pflash
adapter speed: 1000 kHz
none separate
cortex_m reset_config sysresetreq
srst_only separate srst_nogate srst_open_drain connect_deassert_srst
srst_only separate srst_nogate srst_open_drain connect_assert_srst
Info : CMSIS-DAP: SWD  Supported
Info : CMSIS-DAP: FW Version = 1.0
Info : CMSIS-DAP: Interface Initialised (SWD)
Info : SWCLK/TCK = 0 SWDIO/TMS = 1 TDI = 0 TDO = 0 nTRST = 0 nRESET = 1
Info : Connecting under reset
Info : CMSIS-DAP: Interface ready
Info : clock speed 1000 kHz
Info : SWD DPIDR 0x2ba01477
Info : MDM: Chip is unsecured. Continuing.
Info : kx.cpu: hardware has 6 breakpoints, 4 watchpoints
Info : kx.cpu: external reset detected
Info : Listening on port 40109 for gdb connections
    TargetName         Type       Endian TapName            State       
--  ------------------ ---------- ------ ------------------ ------------
 0* kx.cpu             cortex_m   little kx.cpu             reset
Info : MDM: Chip is unsecured. Continuing.
target halted due to debug-request, current mode: Thread 
xPSR: 0x01000000 pc: 0x0000073c msp: 0x1fffc200
auto erase enabled
Info : Kinetis MK21DN512xxx5 detected: 2 flash blocks
Info : 2 PFlash banks: 512k total
wrote 10240 bytes from file /home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf in 0.385966s (25.909 KiB/s)
34 bytes written at address 0x20000000
downloaded 34 bytes in 0.004036s (8.227 KiB/s)
target halted due to breakpoint, current mode: Thread 
xPSR: 0x01000000 pc: 0x20000020 msp: 0x1fffc200
verified 9668 bytes in 0.378957s (24.914 KiB/s)
Info : MDM: Chip is unsecured. Continuing.
shutdown command invoked
Done flashing
/home/francisco/workspace/RIOT/dist/tools/openocd/openocd.sh debug /home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf
### Starting Debugging ###
Open On-Chip Debugger 0.10.0+dev-00703-g92bb76a4-dirty (2019-07-19-14:27)
Licensed under GNU GPL v2
For bug reports, read
	http://openocd.org/doc/doxygen/bugs.html
Reading symbols from /home/francisco/workspace/RIOT/examples/hello-world/bin/pba-d-01-kw2x/hello-world.elf...
Remote debugging using :3333
pm_set_lowest () at /home/francisco/workspace/RIOT/cpu/cortexm_common/include/cpu.h:177
177	    irq_restore(state);
(gdb) 

This is independent of using srst_only and tsrst_and_srst. So please squash keeping the initial srst_only config. I'm sorry that the issue I had with my board stalled this PR so much.

[fjmolinas]

ACK!

[cladmi]

@cladmi as I had told you IRL my board had some issue that could have explained why it wasn't working for me. I managed to fix them and know I have the following behaviour:

Do you know in which state you were before as a reference? or what did you do to become again in the "normal" state?

IIRC @MichelRottleuthner had some boards that could not flash and it may help him too.

This is independent of using srst_only and tsrst_and_srst. So please squash keeping the initial srst_only config. I'm sorry that the issue I had with my board stalled this PR so much.

I agree, as srst_only works, no reason to add the other one.
I tried to see if it would trigger your board reset also but was not the case.

No worry for the delay, these things are still black magic for me and any additional input forces to re-consider what I know.
Also waiting allowed basing on your PR which made it way better.

[cladmi]

I will remove the last commit and push force.

[cladmi]

Thank you for the review and testing.