NAND/CORE: fix clang warning

The fix is inline with the Linux coding style that forbids
assignment in if condition

Change-Id: I42a371d6adfdf3b3fb867705211c47d89776ee2a
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/85
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

FLASH/CFI: fix clang warnings

Total of 5 warnings:
3x "Dead store": removed dead assignment to variable;
1x "Dereference of null pointer": this is not an error, but a
   limited visibility of clang, since pointer erase_region_info
   is initialized inside cfi_fixup_non_cfi();
1x "Branch condition evaluates to a garbage value":
   this is a real coding bug that could issue SIGSEGV, since
   "goto cleanup" can be executed before initialization
   of "source".

Change-Id: Id3c323c82bb15cbd3bb8fc04b23541f11145f109
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/84
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

mx2: add error propagation and remove warnings

Change-Id: Idd4fb452790e5d7921a749679dbd865586e5a4a9
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/48
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

fm3: fix warning for superfluous assignment

Change-Id: I4f8e8c2e676a2728ddc6227daf9ca6a7ceb3d505
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/46
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

kinetis: fix warning about malloc(0) w/assert

Change-Id: Ib40204675bfc5429c744f9ed7e2f7098384b753d
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/47
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

tms470: remove dead assignment warning

Change-Id: I21e7ac47f80d93c9c0d7b169f8848b929c664b20
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/45
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

add Freescale Kinetis K40 devices and Kwikstik eval board

Change-Id: I4817921d09ab915c50f42651bc073690033450fe
Signed-off-by: Mathias K <kesmtp@freenet.de>
Reviewed-on: http://openocd.zylin.com/51
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

clang: fix warning w/assert so that clang knows that there will be no division by zero

Change-Id: I98ac62a22f21043bb535a667a4f1f1537ccde2a4
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/42
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

clang: fix malloc() warning with assert

Change-Id: I989d2655622a9f11f4a0a2994014e42822587ecd
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/41
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

tms470: removed unnecessary operations

This should silence a warning.

Change-Id: Id91a9ebacae836083b1db2654a8e7bf24b2300e9
Signed-off-by: Edgar Grimberg <edgar.grimberg@gmail.com>
Reviewed-on: http://openocd.zylin.com/52
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

FLASH/STMSMI: fix clang "dead store" warning

Change-Id: Icfdefdc48432db2057d3fea19dc424571d2385eb
Signed-off-by: Antonio Borneo <borneo.antonio@gmail.com>
Reviewed-on: http://openocd.zylin.com/50
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

buspirate: ignore return value and fix warning

Perhaps we could do one better and propagate the error?

Change-Id: Idc45f516c26f09de4ee01fe05e8d3475f4b80db3
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/43
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

warning: fix false positive

may be used uninitialized in this function [-Werror=uninitialized]

Change-Id: Ida2cf8efe4e7da6fd9f669b806a20894563ac3d4
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/49
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

server: remove warning due to dead assignment

Change-Id: Idb08aef0c11b3fae92286e8fcea61ab09ab09e74
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/44
Reviewed-by: Peter Stuge <peter@stuge.se>
Tested-by: jenkins

jim: add missing jim license

Change-Id: Ib8e34739d92cd54655b9b47d07b856a82ff25f3c
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/39
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

rtos: remove broken code for handling the deprecated qP packet

Change-Id: Icca522c1e2a2877abb20665b171c61079b1d8f48
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/37
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

rtos: return the correct value if the T or H packets are handled

Change-Id: Iea31e20ee4e35c1a9cb7b93424c92b3f38081067
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/38
Tested-by: jenkins
Reviewed-by: Evan Hunter <evan@ozhiker.com>
Reviewed-by: Peter Stuge <peter@stuge.se>

Unused variables

Fix a few errors with set and unused variables detected by GCC 4.7.0

Change-Id: I59b748e18e514ee9f0cde7883b4ed5116198bd4a
Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/36
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

Updated fm3.c, added Flash type 2 support, error handling improved

Change-Id: I684aca11c4554290d0e57c6d3318d8082980c1ef
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Signed-off-by: Spencer Oliver <ntfreak@users.sourceforge.net>
Reviewed-on: http://openocd.zylin.com/10
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

TMPA900/910 MCUs are always little endian.

Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Change-Id: I8839f2cf0faf1b5ba9f99901c5ee028b199fabd2
Reviewed-on: http://openocd.zylin.com/35
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

Toshiba TMPA900 config: Fix incorrect working area.

The Toshiba TMPA900 series (TMPA900/901) only has internal RAM regions
RAM-0 (16kB) and RAM-1 (8kB) which we can use as working area.

This is probably a copy-paste error from tmpa910.cfg, which has the
correct values and sizes for the TMPA910 series (TMPA910/911/912/913):
there are RAM-0, RAM-1, and RAM-2 (each 16kB).

Also, change "built-in RAM" to "internal RAM" to match what the
datasheet uses.

Change-Id: I993cd6b7fadc28cf34e5cc18426bb2bb42597670
Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Reviewed-on: http://openocd.zylin.com/34
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

xscale: fix bug in xscale_receive()

The code in xscale_receive() that tries to skip invalid reads (i.e.
reads that don't have the DBG_SR[0] 'valid' bit set) seems to be
wrong, as it only looks at the first word's valid flag rather than
each word's own valid flag. Am I reading the code correctly? If so,
the attached patch should fix it.

If this looks correct, I'll generate a proper patch and commit message.

Matt

Change-Id: I74ebe2ad7a36d340a9dd3b8487578b6ea7f3cf1e
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/32
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

luminary: add new targets

update target support from latest SW-DRL 8049

Change-Id: I40aba4d30fe2b79fd955f466c64d99a1dfd63ecf
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/31
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>

luminary: add peripheral reset script

some luminary device classes require a reset script
to emulate a hardware reset.

Change-Id: Id505c92451244b48b0238c2130aebab2df8d208b
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/30
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>

AM/DM37x: Use ICEPick warm reset and include halt when gdb connects.

Using the ICEPick reset seems to allow the processor to be halted sooner
and the halt on gdb connection makes the connect process more robust.

Change-Id: I0586f6e6becc60a729030509ef58907a19d545ec
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/23
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

ICEPick-C: Add support for warm reset through JTAG controller and provide finer detail functions.

This sets up simple functions that can later be used to provide additional
ICEPick Operations.

Change-Id: I313b8679267696fad87d23f3692963e513f2fe21
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/22
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

Add the SAM3N familly to the chip_details table

Change-Id: Ic122d324eacf6e667ed6008ebb84708be944222c
Signed-off-by: Attila Kinali <attila@kinali.ch>
Reviewed-on: http://openocd.zylin.com/29
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>

target_request: add target_got_message() that can be used to improve DCC performance

API change to allow implementing a back-off algorithm for
polling hardware.

Change-Id: I6cbe8b4534c8dfeb8442305171ea96b5481c1f17
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/26
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>

target: DCC / target message backoff algorithm

by immediately polling again when we have received a message from
the target instead of waiting 100ms, we can hope for much better
performance. More than 100x? :-)

Change-Id: Ieaf0c6c8b6e5addc482895670ffbf9a743e07a29
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/27
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>

Add some more detail for setting up Gerrit account.

1) Add a couple more steps when setting up the Gerrit account.

Change-Id: I5d81feac4650d4d28653d14cfc0baf14270424c1
Signed-off-by: Jim Norris <u17263@att.net>
Reviewed-on: http://openocd.zylin.com/28
Reviewed-by: Peter Stuge <peter@stuge.se>
Tested-by: Peter Stuge <peter@stuge.se>

arm-jtag-ew: whitespace cleanup

Change-Id: I8861e825f9c84525e0c09c3adaa3fe300640770d
Signed-off-by: Spencer Oliver <ntfreak@users.sourceforge.net>
Reviewed-on: http://openocd.zylin.com/21
Tested-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

flash: fix lpc2000 driver typo

Change-Id: I3a759ed98a27fd186c12355b846d5e97dba86c5b
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>

Add an interface file for DLP Design DLP-USB1232H.

The DLP Design DLP-USB1232H UART/SPI/JTAG module is based on an FTDI FT2232H
chip. Among other things, it can used as JTAG programmer if connected to
the JTAG target properly. I have successfully wired the module to an
Olimex STM32-H103 eval board and flashed a firmware onto that using OpenOCD.

The setup details and schematics are documented at:
http://randomprojects.org/wiki/DLP-USB1232H_and_OpenOCD_based_JTAG_adapter

Change-Id: I5eb9255a61eeece233009bee77d7dc3b5d1afb8b
Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Reviewed-on: http://openocd.zylin.com/20
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Tested-by: Spencer Oliver <spen@spen-soft.co.uk>

Add a board file for the Glyn Tonga2.

This is a Toshiba TMPA900CMXBG (ARM9) based SO-DIMM CPU module with 64MB
DDR SDRAM, 256MB NAND flash, and on-board Ethernet.

The board file provides a tonga2_init function which sets up the
PLL/clocks and memory (SDRAM and SRAM), which allows writing a boot-loader
into RAM via JTAG.

Change-Id: I60522b97997bdf50e1f25aebab910d93a98522fb
Signed-off-by: Uwe Hermann <uwe@hermann-uwe.de>
Reviewed-on: http://openocd.zylin.com/19
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Tested-by: Spencer Oliver <spen@spen-soft.co.uk>

target: whitespace cleanup

Change-Id: I1453f4f3dc0add529da20577e38b8b82d7d00366
Signed-off-by: Spencer Oliver <ntfreak@users.sourceforge.net>
Reviewed-on: http://openocd.zylin.com/18
Reviewed-by: Alex Austin <alex.austin@spectrumdsi.com>
Tested-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

breakpoint : smp software breakpoint issue

Change-Id: Iefe78bad71d4fdb38ae412ab8fe2f6282836c22e
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/14
Tested-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

Merge "docs: update HACKING to point to Gerrit"

docs: update HACKING to point to Gerrit

Change-Id: If79e86c731ac06aaefca1aebde40e7cb3de68e4d
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>

docs: update more url's

Change-Id: I476078f32910579fed55777c3b0e6da3ef3363b7
Signed-off-by: Spencer Oliver <ntfreak@users.sourceforge.net>

docs: update project url's

Change-Id: I54fc3aff722ed25143aad85e58d19b72fcecbba0
Signed-off-by: Spencer Oliver <ntfreak@users.sourceforge.net>

breakpoint : indentation

Change-Id: Icdb8f72dbb516cd0dfc612c3d61b6801f6382be6
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>

replace berlios url's with sourceforge url's

Change-Id: I1c9957bb64df87cee7c5e832f21453eb8934a5fb
Signed-off-by: Spencer Oliver <ntfreak@users.sourceforge.net>

contrib: remove extra lf

Change-Id: I6e16010e13ad2ea0cdff99b2e8805c74bcd0eb56
Signed-off-by: Spencer Oliver <ntfreak@users.sourceforge.net>

stm32f1x: use async algorithm in flash programming routine

Let the target algorithm be running in the background and buffer data
continuously through a FIFO. This reduces or removes the effect of latency
because only a very small number of queue executions needs to be done per
buffer fill. Previously, the many repeated target state changes, register
accesses (really inefficient) and algorithm uploads caused the flash
programming to be latency bound in many cases. Now it should scale better
with increased throughput.

Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

stm32f1x: use register base instead of register offset

Access the different flash banks' registers using a bank specific register
base and a register specific offset. This is equivalent but feels more
natural.

Some accesses were discovered that maybe should not be hard coded to bank0
registers. Add a note about that.

Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

cortex_m3: use armv7m's async algorithm implementation

Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

armv7m: implement async algorithm functions

Split armv7m_run_algorithm into two pieces and use them to reimplement it.
The arch_info parameter is used to keep context between the two calls, so
both calls must refer to the same armv7m_algorithm struct. Ugly but works
for a proof-of-concept.

Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

target: add async algorithm entries to the target type

On supported targets, this may be used to start a long running algorithm in
the background so the target may be interacted with during execution and
later wait for its completion.

The most obvious use case is a double buffered flash algorithm that can
upload the next block of data while the algorithm is flashing the current.

Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

arm-jtag-ew: Send GDB keep_alive() messages when logging USB communication

- Ticket: #35

arm-jtag-ew: Formatting

arm-jtag-ew: In armjtagew_init(), set initial JTAG speed to 32 kHz (before TAP initialization).

This prevents rare communication errors during startup.

arm-jtag-ew: Emit a warning if interface firmware version != 1.6

arm-jtag-ew: Declare interface as `jtag_only'

arm-jtag-ew: Provide armjtagew_speed_div() in order to fix interactive use of `adapter_khz'

arm-jtag-ew: Fix setting interface speed (2/2)

Interface expects speed in Hz, not kHz

- Ticket #34

arm-jtag-ew: Fix setting interface speed (1/2)

CMD_SET_TCK_FREQUENCY message length is 5, not 4

- Ticket: #34

Add udev rules for openmoko neo1973 debug board

STM32L: Added flash driver and target

Added the flash driver for the STM32L family, which highly differ from the STM32F family.
Added the TCL target file for JTAG access.

Verdex: Add support for Gumstix Verdex boards.

Gumstix Verdex is a PXA270-based series of computer-on-modules. This
configuration file is based off the voipac.cfg configuration with
a different flash memory configuration. This has been tested flyswatter
adapter to reflash a Gumstix Verdex XL6P board.

u8500 : config for L2 cache

armv7a ,cortex a : add L1, L2 cache support, va to pa support

jim-nvp is moving from jimtcl to openocd

The jim-nvp code is specific to openocd, so it belongs in openocd,
not in the core jimtcl.

Signed-off-by: Steve Bennett <steveb@workware.net.au>

AM/DM37x: add ES1.2 silicon type into account

The missing value for ES1.2 silicon revision is mentioned in
sprugn4m.pdf, and the recent TI Beagleboard XM is powered by it,
so let support the revision.

add target events, run algorithm and default r/w buffer api

Target events are added to get better gdb support. The run
algorithm functionality are implemented to support feature
fast flash write functionality. The new r/w buffer api is now
used to support the special memory address handling. The output
of the md command was fixed.

kinetis : fix deadlock on device having hasidcode false.

kinetis cpu flash driver

Initial release of the freescale kinetis cpu flash driver.

contrib: fix udev rules for tty based adaptors

Most serial adaptors are identified by udev with SUBSYSTEM=tty and
without DEVTYPE. This patch fix udev rules to work with any listed
tty-based adaptor. It has been tested with a FTDI-based Bus Pirate.

Signed-off-by: Luca Bruno <lucab@debian.org>

kinetis auto mass erase on secured devices

This is a proof of concept to get access to the debug port of a
secured kinetis cpu. On full flash erase the cpu is automatically
secured and the debug port is not accessible.
To get this to work the srst line is needed and the necessary
configuration should be added to the configuration file.

Fixes and spellchecks for various Buspirate output messages

Signed-off-by: Luca Bruno <lucab@debian.org>

Add definition for the STEVAL-PCC010 board with the STM32F207

Implementation of a new jtag remote_bitbang driver.

The driver sends ascii encoded bitbang commands over unix sockets or TCP to
another process. This driver is useful for debugging software running on
processors which are being simulated.

ULINK driver: Remove typedefs in ulink.c

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Remove typedefs in OpenULINK firmware: Use typedefs from stdint.h (uint8_t, uint16_t) instead of custom typedefs in shorttypes.h (u8, u16, ...)

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Remove typedefs in OpenULINK firmware USB descriptor structures

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

usleep to jtag_sleep

this will help avoid platform specific timing issues

static for some functions

made two functions into static, since they are not required by anything external

speed up, relocate function

added an attempt to use the non-reseting halting sequence. if it fails, then the full sequence will be attempted. this makes things a bit faster most of the time.
changed the location of a function, avoiding a forward def

Merge branch 'dsp5680xx_cherry' of git://repo.or.cz/openocd/dsp568013 into fix

ULINK driver: Update firmware image to reflect the latest changes in source code.

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Add '-lm' linker flag when building this driver (required for correct calculation of JTAG TCK speed setting)

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Implement variable TCK frequency in OpenOCD driver

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Implement variable TCK frequency in OpenULINK firmware

Also, speed up jtag_clock_tck() significantly (150 kHz -> 375 kHz)

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Fix whitespace in OpenULINK firmware usb.c module (trivial)

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Implement JTAG_PATHMOVE command

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Implement JTAG_STABLECLOCKS command

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Implement command to manually force downloading firmware image from arbitrary location

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Re-order queue functions to reflect the order in commands.h

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Properly propagate return values in ulink_execute_queue()

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

ULINK driver: Update some comments, fix some coding mistakes

Signed-off-by: Martin Schmölzer <martin.schmoelzer@student.tuwien.ac.at>

vsllink driver compile fails with'vsllink_debug_buffer' defined but not used

USB communication is handled by code under versaloon directory.
So _DEBUG_USB_COMMS_  should not be used in vsllink.c.
Attachment is the patch.

removed trailing whitespaces

emacs is awesome.
replace-regexp RET [ ]+$ RET RET
and it's done

fix enter debug mode for locking

added an alternative way to enter debug mode, which does not require restarting the chip.
this will not always work, but in general it will (failure 0.3%), and failure is not a dramatic issue, simply have to use the full sequence.
the user can only access "halt", which uses the full sequence, so the user should not have any problems.
restarting the chip requires reconfiguring the flash module. the doc is very poor, so i'd rather have the two methods, and live with the 0.3%.

fix irlen handling

sometimes the master tap will be enabled, since tap switching is required during halt/lock/unlocking procedures.
now irscan handles this, avoiding unnecessary warnings and preventing errors.

fix debug mode,lock,unlock

got new info regarding setting the chip to debug mode, and locking/unlocking flash memory.
the newer implementation is a bit slower, but always works.
the previous implementation would randomly (as once every 25k-70k times) get the chip into a state where the freescale tool would be necessary. this is fixed now.

added functions to play around with the jtag state machine. they are not the happiest, but are necessary to be able to execute the halting/locking/unlocking sequences.

Conflicts:

src/target/dsp5680xx.c

fix error handling during halt

the user can execute halt, but no enter_debug_mode. modified the error handling to suite this.
the new implementation of unlocking will use enter_debug_mode, and should not get the same errors as the user would, because not being able to enter debug mode is actually success when checking for locked flash.

optional crc for flash writing

crc check was always performed on newly flashed data, now it is optional
flash mem can be locked by writing a specific word to a specific address in flash.
to verify flash, target must be halted, and this will (when the new halt sequence is implemented) require reseting the chip. if the target is reset after writing the lock words, then it will lock, hence the CRC will fail because it is not possible to read stuff from the target.

also added a function that resets the jtag state machine.
this is not used yet, but will be soon.
it is implemented to allow strict control over JTAG state machine, necessary to implement to halt and unlocking sequences.

def syntax to match tap irlen

the master tap has a 4 bit irlen
changed the instructions to be 4 bit, ie, removed the zeros.
it makes it clearer to interpret.

dsp568013 disable polling by default

cortex_a hybrid & context breakpoints

Fix off by one bug in FreeRTOS