Improve HACKING

Reword info about creating SSH key - it's not required to add it to
Github account. Mention adding created SSH key to Gerrit account -
without this step it's not possible to access Gerrit in further
steps.

Change-Id: Ibd81521fbe47d4b4beae0b77cdc9d939fd3ee20c
Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/1350
Tested-by: jenkins
Reviewed-by: Paul Fertser <fercerpav@gmail.com>

telnet: add telnet history support

adapted from Yoshinori Sato's patch:
https://github.com/ysat0/openocd/commit/2f07f4600a0da8206612d78c159bbe1171aa41c2

Change-Id: I084b86d316b0aa6e9593f007c024961dbda805e9
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1310
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

Support newer OSBDM firmware

OSBDM: add new VID:PID implemented in OSJTAG/OSBDM firmware somewhere
between versions 30.13 and 31.21.  PFLASH programming works with this
patch, tested on a Freescale Kinetis TWR-K20D72M using its onboard OSBDM
JTAG adapter.

Note: flash program testing required hacking kinetis_write() to force
longword programming, as the FTFL program section commands formulated by
kinetis_write() currently fail on this board's PK20DX256VLL7 processor.

Change-Id: Ib7b92ff2fe9ebf6158fb1489f554a19e96cd9651
Signed-off-by: R. Steve McKown <rsmckown@gmail.com>
Reviewed-on: http://openocd.zylin.com/1348
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

stm32f30x: Add boundary scan TAP ID to match silicon

Change-Id: I74ef3cfc437540aedd99da46ac3e0c6cd9c5cd8d
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/1354
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

mips: m4k alternate pracc code. Patch 4

Now all the functions with only fetch accesses are modified.
The same delay between scans has been added to mips32_pracc_fastdata_xfer(), it should work
at the same scan rates as the other pracc functions, but it needs higher scan_delays
to work.

Change-Id: Ifb31d8ea6de9d22674385782913d221a2494dbbf
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1196
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

mips: m4k alternate pracc code. Patch 3

Functions mips32_pracc_read_mem(), mips32_cp0_read() and mips32_pracc_read_regs() are now modified.
mips32_cp0_read() is very similar to mips32_read_u32() with one store access.
mips32_pracc_read_regs() is the only function that can not be executed from only one queue.
Now this function is modified to use reg8, it saves all the registers but does not restore reg8.
To remedy this, mips_ejtag_config_step() is called after mips32_save_context() in
mips_m4k_debug_entry(). Function mips_ejtag_config_step() is modified to use reg8 and
restore it from ejtag info instead of using DeSave for save/restore.

Change-Id: Icc224f6d7e41abdec94199483401cb512cc0b450
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1195
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

mips: m4k alternate pracc code. Patch 2

Each pracc function defines a variable ctx of type struct pracc_queue_info.
To simplify the code tree auxiliary functions are defined: pracc_queue_init(), pracc_add() and
pracc_queue_free().
The second parameter in pracc_add() is the store address if the instruction is a store at dmseg,
otherwise it should be 0.
The code is executed by mips32_pracc_queue_exec(). If ejtag_info->mode is 0 mips32_pracc_exec()
is called and it should work like with current code.
To generate the delay between scans the number of clock ticks are calculated with the help of
jtag_get_speed_khz(). Due to delays in the execution of each single ftdi instruction the number of ticks
are higher as it should be, specially at higher scan rates.
mips32_pracc_read_u32() should now work with the new code.

Change-Id: I471590a4fc89b56af10bd46c48767b4c64de154f
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1194
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

mips: m4k alternate pracc code. Patch 1

This patch and the following patches define another way of doing processor access without the need to read back
the pracc address as needed in current pracc code.
Current pracc code is executed linearly and unconditionally. The processor starts execution at 0xff200200
and the fetch address is ever incremented by 4, including the last instruction in the delay slot of the branch to start.
Most of the processor accesses are fetch and some are store accesses.
After a previous patch regarding the way of restoring registers (reg8 and reg9), there are no load processor accesses.
The pracc address for a store depends only on the store instruction given before.
m4k core has a 5 stage pipeline and the memory access is done in the 3rth stage. This means that the store access
will not arrive immediately after a store instruction, it appears after another instruction enters the pipeline.
For reference: MD00249 mips32 m4k manual.
A new struct pracc_queue_info is defined to help each function in generating the code. The field pracc_list holds in the
lower half the list of instructions and in the upper half the store addressess, if any. In this way the list can be used by
current code or by the new one to generate the sequence of pracc accesses.
For every pracc access only one scan to register "all" is used by calling the new function mips_ejtag_add_scan_96().
This function does not call jtag_execute_queue(), all the scans needed can be queued before calling for execution.
The pracc bit is not checked before execution, is checked after the queue has been executed.
Without calling the wait function the code works much faster, but the scan frequency must be limited. For pic32mx
with core clock at 4Mhz works  up to 600Khz and with 8Mhz up to 1200. To increase the scan frequency a delay
between scans is added by calling jtag_add_cloks().
A time delay in nano seconds is stored in scan_delay, a new field in ejtag_info, and a handler is provided for it.
A mode field is added to ejtag_info to hold the working mode. If a time delay of 2ms (2000000 ns) or higher is set,
current code is executed, if lower, new code is executed.
Initial default values are set in function mips32_init_arch_info. A reset does not change this settings.

Change-Id: I266bdb386b24744435b6e29d8489a68c0c15ff65
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1193
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

Improve clone command in README

Without the explicit dir at the end the repository will be cloned to "code".

Change-Id: Icd8b55b4ba74f23b214c3844e2fb785377768119
Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/1349
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Tested-by: jenkins

stm32w: Added sample target configuration for STM32W108 with STLink-V2

As requested, here is the target configuration that I'm using for an
STLink-V2-attached STM32W108C8. For some reason, it only seems to work
with "reset_config trst_only".

Change-Id: Icbff4f83343e1f505d8afdfc53ff6f8b7496cac9
Signed-off-by: Ben Nahill <bnahill@gmail.com>
Reviewed-on: http://openocd.zylin.com/1347
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

topic: STM32W support added to em357 driver

The em357 driver only supported one page configuration (192k in 96 2048k)
pages. This is fine for em357 chips since that's the size they have, but
ST's STM32W chips (pretty much the same) have different flash
configurations available (64, 128, 192, 256k). I can't find anywhere
that would indicate the size of the chip anywhere in memory so the
selection must be manual, using the 'size' parameter. For backwards
compatibility, any size not known to be in use defaults to the 192k
configuration. I don't have any em357 devices to test, but I also found
that I had to re-assert the FPEC clock enable before performing an
erase. This is a single line and shouldn't break any configurations.

My testing so far has only been with a 64k device with 8k of RAM.

Change-Id: Ic0ac400a9696efaa09d1407dd4a4d456bc2c318b
Signed-off-by: Ben Nahill <bnahill@gmail.com>
Reviewed-on: http://openocd.zylin.com/1336
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

program: do not poll target after reset run

Disable polling the target before we issue a 'reset run'. This stops errors or
warnings if the target disables the SWD or JTAG interface as part of the
application code.

Change-Id: I5019dffdad41a8e210003ece1caf89069ee0f223
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1331
Tested-by: jenkins

stlink: fix connect under reset issues

We need to make sure that srst is asserted before we attempt to switch into
jtag or swd mode otherwise we receive a error (-9) - invalid device id.

Change-Id: I625166c751cfba8e8a5290f40122bb9afc9dbb39
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1315
Tested-by: jenkins

parport: fix parport_toggling_time regression

If parport_toggling_time is called before the adapter speed has been
configured then the call fails. Probably not the best fix, but does at least
enable parport_toggling_time to be used again.

This regression was added in commit 740b9e25b410c164e661d0334a9ea4168406726b

Change-Id: I90300916d6bda5ef053c557e5ac136c4f002bdd1
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1309
Tested-by: jenkins

ft2232: remove ft2232_large_scan memory leak

This is a very long outstanding issue see:
http://lists.berlios.de/pipermail/openocd-development/2011-June/019404.html

As this driver is deprecated the fix is added to purely to reduce the warnings
reported by clang.

Change-Id: I3a16a704e0e8db27efda50fdcfdd35abf5ebed0f
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1278
Tested-by: jenkins

libusb: disable debug messages by default

Change-Id: I15dec0f521502139b57adaff576516af7883a74b
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1314
Tested-by: jenkins

jimtcl: update embedded jimtcl

Update to latest jimtcl commit 2c1eba991e21a6f0b531fb0f83e21f9e6ee7c515.
This fixes issues when building on certain versions of Mac OSX.

Change-Id: I551477752d7913c84e6deb60b889d0c14bd200a0
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1311
Tested-by: jenkins
Reviewed-by: Xiaofan <xiaofanc@gmail.com>

cfg: Added cfg script for at91sam4sd32x targets

Change-Id: I3b8a54d89a180bfded3dae3f1fe3d940540e6e7d
Signed-off-by: Peter Dietzsch <peter.dietzsch@ib-dt.de>
Reviewed-on: http://openocd.zylin.com/1333
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

flash: Added support for at91sam4sd32c

Change-Id: I7223980602d7595a3dd7a3ceaac3f58d4f73f88d
Signed-off-by: Peter Dietzsch <peter.dietzsch@ib-dt.de>
Reviewed-on: http://openocd.zylin.com/1332
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

ft2232: fix input scan ending in drshift/irshift

The final bit was incorrectly added as output data, even if no data was
to be written. Changed it to match handling of other bits.

Change-Id: I91e5ba0c932876bfb579c22e6c7ef0300baa1534
Signed-off-by: Yann Vernier <yann.vernier@orsoc.se>
Reviewed-on: http://openocd.zylin.com/1049
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

arm_adi_v5: fix for csw nonsecure access.

Add command to fix CSW_SPROT in register AP_CSW.
This solves dap apmem access in non secure access.

Change-Id: I7cfcb6434d75f5cfd4a2630a059901cdeea010ce
Signed-off-by: Michel JAOUEN <michel.jaouen@stericsson.com>
Reviewed-on: http://openocd.zylin.com/1276
Tested-by: jenkins
Reviewed-by: mike brown
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

mips: code cleanup in cp0 command handlers

After calling mips32_cp0_read() nothing has been queued, the call to jtag_exec_queue() is unnecessary.

Change-Id: Ie25438045a8e9b6b1b170df7b52609d45f284b5a
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1190
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

mips: change in restoring debug working register

In current devel code there are 3 functions (related to m4k code) that need to restore register 8 from pracc stack:
mips32_pracc_read_u32()
mips32_cp0_read()
mips32_pracc_write_mem_generic()

And mips32_pracc_read_mem() needs to restore regs 8 and 9 from pracc stack.

Values in this registers should be the same as read by mips32_pracc_read_regs() when entering debug
mode and can be modified by mips32_pracc_write_regs() when leaving debug mode.
There is no need to read their values from the processor registers every time.

The fields reg8 and reg9 are added to struct mips_ejtag to store these register values
and the call to mips32_save_context() is shifted in mips_m4k_debug_entry() in order
to store them before any other function needs to restore these registers.
For the same reason in function mips_m4k_step() the call to mips_m4k_set_breakpoint(), if needed,
should be made after calling mips_m4k_debug_entry().
For single word write the number of pracc accesses are now 9 or 8, from 13 or 12 in current code,
single word read takes now 10 instead of 12.

This patch is really the first in a set of patches for an alternate m4k pracc code
much faster that current code. At least for me with pic32mx works fine.

Change-Id: Ibd9df5e8b9f78ce05a180949ba6a561c761b61d6
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1146
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

mips: mips32_pracc_fastdata_xfer() little modification

In this function after loading the handler code and the jump code there is a call
to wait_for_pracc_rw() to verify that a pracc access is pending.
Next the address is read to verify that the handler is running, the address should be at
fastdata area.
Next, another call is made to wait_for_pracc_rw(). This call is not needed, we now already
that a pracc access is pending.
Better we call this function before loading the end address to be sure it is loaded correctly.

Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Change-Id: If311450ea634786fc28cf1a8e18ed24ce5257d20
Reviewed-on: http://openocd.zylin.com/1142
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

pic32mx: 0 wait state option

By default pic32mx starts after any reset with 1 wait state for RAM access/exec.
It can be changed to 0 wait states by clearing the BMXWSDRM bit (bit 6) in BMXCON register.
With 0 wait states near doubles the execution speed. CRC check sum can be done much faster
increasing verify_image speed. Fast data transfer also works with a bit higher scan rate, up to 1500 Khz.
This option can be set at any time with
mww 0xbf882004 0x40
or cleared with
mww 0xbf882008 0x40.
Some numbers for FTDI/HS with current devel code and a elf file:

Core clock / wait states              verify_image speed
------------------------------------|------------------------------

        4 Mhz    /    1                            21 KiB/s
        4 Mhz    /    0                            36 KiB/s
        8 Mhz    /    1                            37 KiB/s
        8 Mhz    /    0                            57 KiB/s

Change-Id: I4092ad0f3753f72f77108718d0ed3a3ab84e3b23
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1141
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-by: Xiaofan <xiaofanc@gmail.com>

pic32mx: false pending at low core clock

To show up the fail try to step with the core clock set to 31.25Khz
and with a ftdi/hs adapter or with a wiggler, -not with ft2232-.
The scan frequency should be set to 300Khz or higher, at lower frequency probably will not fail.

The code exits with error because the pracc address is at 0x0.

It also fails when using the "all" register, but in this case the code works without any message because the
pracc address is at 0xff202004 when it fails.

I never saw this fail with the core clock set to 500Khz or higher, but ...

The workaround simply puts a 1 ms delay after the execution of the DERET instruction.

Change-Id: I38e8c01a9c39aedd3282140543b83a0844d8ad29
Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/1139
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

Added support for ARMv7-M in arm io.

Added support for ARMv7-M targets in arm_nandwrite and
arm_nandread.

Change-Id: Iab1d78d401f735e191c6a8519f3619035a300fae
Signed-off-by: Henrik Nilsson <henrik.nilsson@bytequest.se>
Reviewed-on: http://openocd.zylin.com/1188
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

Add abort when JTAG-DP transaction times out.

Fixes system hang for devices that don't ignore
transactions to bad addresses.

Change-Id: Ia98344d7efc12951ef79dbc82b8f792b70a22cee
Signed-off-by: Evan Hunter <ehunter@broadcom.com>
Reviewed-on: http://openocd.zylin.com/1115
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

arm_adi_v5: fix mem_ap_read_buf_u32() JTAG nastiness..

Moved JTAG code out of transport-neutral file (arm_adi_v5.c) into
transport specific file (adi_v5_jtag.c).
Added ap_block_read to dap_ops interface (arm_adi_v5.h) to support
the move.

Change-Id: I796d3984f138aad052b97c77ac9c12ffd1158f74
Signed-off-by: mike brown <mike@theshedworks.org.uk>
Reviewed-on: http://openocd.zylin.com/1277
Tested-by: jenkins
Reviewed-by: Michel JAOUEN <michel.jaouen@stericsson.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

gdb server: Fix buffer overrun - sprintf appends a terminating null to the data which was overrunning the supplied buffer.

Fixes regression introduced in commit 07dcd5648d146d38f9ffa619f0737587e592d0b6

Signed-off-by: Evan Hunter <ehunter@broadcom.com>
Change-Id: Iec64233c0da5a044fb984c4b1803309cb636efe9
Reviewed-on: http://openocd.zylin.com/1312
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

ti_icdi: add icdi_usb_query result check

Change-Id: I0b40586677a77ee6ae46fe120a677616bde22d1e
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1279
Tested-by: jenkins
Reviewed-by: Xiaofan <xiaofanc@gmail.com>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

docs: update incorrect urls

These were missed when git was moved to the new SF platform during Nov 2012.

Change-Id: I7b4ae9dea010d95f9bf4c26841b5b724f41768be
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1248
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Tested-by: jenkins

docs: remove unnecessary whitespace

Change-Id: I11bad3de145d941b61e9bd4920bc3281ece91ab3
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1245
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

docs: fix html anchor xref links

makeinfo has a long outstanding bug that means @anchors are not correctly
formatted for split html, see:
http://lists.gnu.org/archive/html/bug-texinfo/2012-06/msg00000.html

The issue relates to using spaces or hyphens in the @anchor name.
Issue also reported via Trac #44

Change-Id: Id72e23375dd167674b2ae5b314e8242b90a72a5f
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1244
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

at91sam3: Wrong PLLA frequency calculations

The command 'at91sam3 info' ignores PLLA DIV values >1. This patch fixes it.
Tested on a SAM3S4C chip.

Change-Id: I051f41bb3dcefe1ac785fbcb48477a807daa16a2
Signed-off-by: Thomas Schmid <thomas.schmid@gmail.com>
Reviewed-on: http://openocd.zylin.com/1307
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

rtos: fixed handling of qThreadExtraInfo packets

The commit "gdbserver: use common hexify/unhexify routines" [3d62c3d]
mis-replaced a call to "str_to_hex" with a call to "unhexify". "hexify"
should have been used instead.

Change-Id: I5f5904b1b422f819a6308e2c0740ea43d22c7d0b
Signed-off-by: Christian Gudrian <christian.gudrian@gmx.de>
Reviewed-on: http://openocd.zylin.com/1308
Tested-by: jenkins
Reviewed-by: Peter Stuge <peter@stuge.se>

target: fix broken Cortex-R4 support

This regression was caused due to the recent addition of R4 support and
the removal of the bulk_write_memory handler.

Change-Id: Ide692737f235c0e9906becb6f3502ba52c5907aa
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1246
Tested-by: jenkins

target: Retire target_bulk_write_memory()

The only caller was arm_nandwrite(). Replace that call with
target_write_buffer() instead, which in turn may end up calling the same
bulk_write_memory target API function.

Change-Id: If34c7474df5cf14af3b732fb4774816818f28e79
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/1214
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

target: Add default implementation of bulk_write_memory

Remove dummy implementations from all targets except arm7_9 and mips, which
are the only ones with real implementations. Replace with a single default
implementation simply calling target_write_memory().

Change-Id: I9228104240bc0b50661be20bc7909713ccda2164
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/1213
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

arch: Added ARMv7R and Cortex-R4 support

Rewrite to merge Cortex-A and Cortex-R code

Change-Id: I4541557980d43d1bba6e8d1bfeb04f536ed25a00
Signed-off-by: Evan Hunter <ehunter@broadcom.com>
Reviewed-on: http://openocd.zylin.com/358
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

adi_v5: search for Debug and Memory AP support

Adds dap_find_ap() function.

Change-Id: I6643025624009b12d4936de67a605da52c07be49
Signed-off-by: Evan Hunter <ehunter@broadcom.com>
Reviewed-on: http://openocd.zylin.com/909
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

cortex_a : optimize apb read/write access.

Rewrite: Adheres more closely to 'fast read/write' examples in TRM.
up to 50x faster

Change-Id: Ieb4da57d8367628f3e7306827a5b1f0ab550e641
Signed-off-by: Evan Hunter <ehunter@broadcom.com>
Reviewed-on: http://openocd.zylin.com/903
Tested-by: jenkins
Reviewed-by: Michel JAOUEN <michel.jaouen@stericsson.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

jtag: add support for some probes that are mostly compatible with opendous

This patch adds support for usbprog-jtag and usbvlab that are mostly compatible
to opendous except for IN and OUT endpoints and usb transfer mode.

Change-Id: I44557c2449fe7473295038efa6ae4fc8d80ec7bf
Signed-off-by: Stefan Mahr <stefan.mahr@sphairon.com>
Reviewed-on: http://openocd.zylin.com/687
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

jtag: usb_blaster: fix allocation of usb_blaster_device_desc

usb_blaster_device_desc was allocated, but never freed.

Change-Id: I764bd092c71b8c260b98aab0e7a1710fd7bfa9fd
Signed-off-by: Stefan Mahr <stefan.mahr@sphairon.com>
Reviewed-on: http://openocd.zylin.com/1224
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Tested-by: jenkins

jtag: parport: avoid freeing read-only memory section

If command parport_cable is not executed, parport_cable points to
const char array in read-only memory as default. On exit free()
will try to free this read-only memory. This patch uses strdup to
allocate memory when defining default setting.

Change-Id: I290e707ac6a37e9dc1b45c85ca51d8bd6aac6761
Signed-off-by: Stefan Mahr <stefan.mahr@sphairon.com>
Reviewed-on: http://openocd.zylin.com/1223
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

jtag: opendous: fix tap buffer overflow

Appending bits to TAP buffer doesn't check if there's enough space left.
This patch adds this check to fix TAP overflow error.

Change-Id: If80d5ab4a24983ad24f3cab31f9676d1590ebf5d
Signed-off-by: Stefan Mahr <stefan.mahr@sphairon.com>
Reviewed-on: http://openocd.zylin.com/1216
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

target: Remove read_memory_imp

Change-Id: Idc6ef3b075ccbb5945df8fea746011cb17175d8f
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/1219
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

target: Remove write_memory_imp

Change-Id: I5d933bc19443bba8a0193c90471fdd0614324a92
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/1218
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

target: Remove soft_reset_halt_imp

Change-Id: I12c907584ef73de570eba2dcfeb8477cabc6098f
Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/1217
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

doc: opendous interface based on ft2232H

It was listed in the ft2232 based cables supported. Moreover, the
ft2232_channel option, which was added to support this cable, was explained.

Change-Id: I82ebc7bc10d6472f96ab150e78d623a617edccd2
Signed-off-by: Rodrigo Melo <rmelo@inti.gob.ar>
Reviewed-on: http://openocd.zylin.com/1098
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

interface: opendous_ftdi config file added

This config file add support to the opendous cable based on the chip ft2232H,
using the ft2232 interface driver.

Change-Id: I8171a0c475af8d61e081844ee86466a392138fb0
Signed-off-by: Rodrigo Melo <rmelo@inti.gob.ar>
Reviewed-on: http://openocd.zylin.com/1096
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

ft2232: ft2232_channel option added

With this option a different channel of the ft2232 chip can be selected using
a previously existing layout. It was made for a partner called Salvador
Tropea.

Change-Id: Ia0dedb2f50e232d089e73788735edc8f47ee23e6
Signed-off-by: Rodrigo Melo <rmelo@inti.gob.ar>
Reviewed-on: http://openocd.zylin.com/1095
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

doc: Added MIPS target document

This document explains MIPS EJTAG operations
implementations within OpenOCD.

Change-Id: Ieaf01f8bc5a97d7b0f2c847bac5fbb2ccf8ba088
Signed-off-by: Drasko DRASKOVIC <drasko.draskovic@gmail.com>
Reviewed-on: http://openocd.zylin.com/904
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

Kinetis: Flash command function matches datasheet

The kinetis datasheets specify the flash registers as bytes rather
than as words, as the previous implementation did.  This also makes
a few code sections slightly less endian-magical.

Change-Id: If8f4adfc7f4341085ae5b6eacbf7d74bbd74cf08
Signed-off-by: Alex Austin <alex.austin@spectrumdsi.com>
Reviewed-on: http://openocd.zylin.com/1192
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

Kinetis: Symbolic names for Addresses and Commands

Change-Id: I2165b66c37bd1608139b5dd00f48124161e13ef0
Signed-off-by: Alex Austin <alex.austin@spectrumdsi.com>
Reviewed-on: http://openocd.zylin.com/1191
Tested-by: jenkins
Reviewed-by: Christopher Kilgour <techie@whiterocker.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

docs: add gerrit http password url

Add url to show where the http password is configured.
Also add note that password can also be saved to url git config.

Change-Id: I3c1a022580e5f73372b0c50e8d1d2f0b1498966f
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1207
Tested-by: jenkins

tcl: add flash programming helper

This adds a program proc that simplifies using OpenOCD as a standalone
programmer.

Change-Id: I6ece492cd878c170b734e8bb2e09fe8c4557d5a6
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1199
Tested-by: jenkins
Reviewed-by: Jörg Fischer <turboj@gmx.de>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

rtos: fixes warning

Change-Id: I45db15b16b52c71009d8830985f42ac88eabe160
Signed-off-by: Øyvind Harboe <oyvind.harboe@zylin.com>
Reviewed-on: http://openocd.zylin.com/1209
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

gdb: use stdio functions to convert parameters

This reduces the number of gdb conversion routines we have to maintain.

Change-Id: Ia43d6cac86cbe4f76fe0875b9d9c16ac340296db
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1128
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

interface: opendous_ftdi config file added

This config file add support to the opendous cable based on the chip ft2232H,
using the ftdi interface driver.

Change-Id: I4491f99d7b14f7078a04583ef0c4acd8692c4349
Signed-off-by: Rodrigo Melo <rmelo@inti.gob.ar>
Reviewed-on: http://openocd.zylin.com/1097
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

docs: use doxygen keywords for formatting

Rather than use bold fonts etc, use the correct keywords so doxygen formats
better.

Change-Id: Id9d63f0fc3465665376d7a536c4d6da71998f40e
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1210
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

cfg: add Netgear DG834v3 configuration

Change-Id: I3f4880d8b07b9623544b94d316b37e6d0ae97020
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1189
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

flash: fix stm32 failed probe using incorrect flash size

This fixes an issue if the device is manually probed after the initial probe
fails due to being unable to read flash size register. In this situation the
driver assumes the user has overridden the flash size when infact this may
not be the case.

It also seems on the older stm32f1 devices the flash register is not readable
when locked, this does not seem to apply to the newer parts - f0, f3, f4.

Change-Id: I125f872fcb2d962ca6705f97b62d957e2b31303b
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1187
Tested-by: jenkins
Reviewed-by: Johan Almquist <johan.almquist@assaabloy.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

stm32: update stm32f1x driver for f0x and f3x option bytes

The stm32f0 and stm32f3 share the same option byte location, but the format
differs.

Adding an option_offset fixes the broken options_read cmd and incorrectly
setting Hardware Watchdog when unlocking a f3x device.

Change-Id: I82d66b6198294ea9eedb44ca8b2fb368c0cb15e8
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1184
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

ti-icdi: catch failed icdi_send_cmd

warnings detected by clang.

Change-Id: I1532bcc12a8ab7446646dfb2a7afa8894ff03679
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1180
Tested-by: jenkins
Reviewed-by: Xiaofan <xiaofanc@gmail.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

cfg: add basic support of Freescale i.MX6 series targets

This change adds a simple target configuration for Freescale
single/dual/quad core i.MX6 SoCs, only one core is configured by default.

Change-Id: I853dd27f4c6765b7f731be2ddea82e85d496c6a4
Signed-off-by: Vladimir Zapolskiy <vz@mleia.com>
Reviewed-on: http://openocd.zylin.com/1135
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

target: use common target_name to access target::cmd_name member

Change-Id: I203b89ef25a072c3b00b504483d5f2a83477fad6
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1182
Tested-by: jenkins
Reviewed-by: Mathias Küster <kesmtp@freenet.de>
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

speed up ftdi by reorder to out-in

When the ftdi driver calls finally the mpsse_flush function, it first
initiate the USB in and finally the corresponding USB out transaction.
Because data in is requested too early the USB device will always answer
the first USB in by a NAK. That can prevented by a simple reordering of
the out and then the in transfer and can improve the Jtag performance for
high JTAG clock rates.

Change-Id: I17abf1487c914c92e2e447ee6d30562ef629f327
Signed-off-by: Peter Henn <Peter.Henn@web.de>
Reviewed-on: http://openocd.zylin.com/942
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-by: Xiaofan <xiaofanc@gmail.com>

Change reset configuration.

This patch change the default reset config from SYSRESETREQ to the working
VECTRESET.

Change-Id: I21a9a74b9c0c68cfa3a6e6dac9b123acc98a93cb
Signed-off-by: Mathias K <kesmtp@freenet.de>
Reviewed-on: http://openocd.zylin.com/1186
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

Move back off timer to target struct

Move the global target back off timer to the target struct. This will
fix the wrong error handling with multi target devices like smp systems.

Change-Id: Ia327182ed5d13ca87323700017a8c40ecc6b25a3
Signed-off-by: Mathias K <kesmtp@freenet.de>
Reviewed-on: http://openocd.zylin.com/1179
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

Add the target name to debug output for better understanding and error identification.

Change-Id: I1054debea6cd3a6548aadeae2d84000a0039814e
Signed-off-by: Mathias K <kesmtp@freenet.de>
Reviewed-on: http://openocd.zylin.com/1178
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

Update HACKING info

Inform about possibility of discarding negative review in Gerrit when specific conditions are met.

Change-Id: I432b6c93cefc368fa22ce1096bea4cd174e03816
Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/747
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>
Tested-by: Øyvind Harboe <oyvindharboe@gmail.com>

gdbserver: use common hexify/unhexify routines

Change-Id: I9989b625666e9c60ec9867cf6f4d94f41c998c3f
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1105
Tested-by: jenkins
Reviewed-by: Mathias Küster <kesmtp@freenet.de>
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

Fix buffer overflow in versaloon interface

The USB buffer will need space for both TMS and TDI buffers.
Each holds tap_buffer_size bytes maximum, so tap_buffer_size must be
smaller than half of usb buf_size.

Change-Id: Id8f39936a894cbd98deb89eec5a859aef1e2b783
Signed-off-by: Joerg Fischer <turboj@gmx.de>
Reviewed-on: http://openocd.zylin.com/1136
Tested-by: jenkins
Reviewed-by: simon qian <simonqian.openocd@gmail.com>
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

target: hla correctly use target events

Because we were always running using target state TARGET_RUNNING target
algorithm's were a bit verbose compared to other targets.

This brings the hla target inline with the other targets.

Change-Id: I3a257fdc878b87660fac8b5eca22b421eee5b349
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1134
Tested-by: jenkins

armv7m: update to use correct register core_cache

The was missed when the armv7m was moved over to using the std arm
core_cache, probably because it is disabled by default.

Change-Id: I2f5a18ef6dd783b36e8c29f4c52379104bda4583
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1138
Tested-by: jenkins

stm32: add support for stm32l1x 256k high density single bank devices

Added support for new ST devices in the stm32lx portfolio, with device
id 0x427. These have 256k flash, but in a single bank compared to
device id 0x436 which is a dual bank flash.

Change-Id: Iafdfe990f24bd04b0d6e00385ee70690f3bf8d5f
Signed-off-by: Johan Almquist <johan.almquist@assaabloy.com>
Reviewed-on: http://openocd.zylin.com/1140
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

stm32: add support for the STM32Lx 384kb dual bank flash

This update adds support for the STM32Lx 384kb dual bank flash. Previously there was a problem when writing an
image that was larger than 192Kb. That lead to openocd printing out two error messages like
"Error: access denied / write protected" and "Error: invalid program address". The reason was that the stm32lx
driver tried to write half pages which overlapped into the next flash bank.
A new configuration file stm32lx_dual_bank.cfg can be used for stm32lx chips with dual bank flash (256kb or 384kb devices).
A sanity check was added for probed flash size values to fix the issue seen on some ST samples that answered incorrectly.

Change-Id: I69e25131983d88613be8606b438f98870c5f1e52
Signed-off-by: Johan Almquist <johan.almquist@assaabloy.com>
Reviewed-on: http://openocd.zylin.com/1125
Tested-by: jenkins
Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>

stm32: enable flash bank size override

It has been seen on some stm32 targets that the flash size register that
is probed by the driver may contain an invalid size.

This change enables the user to override the probed value.

Change-Id: I09359e59a96f9133d3d939670957d32a830a944e
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1132
Tested-by: jenkins
Reviewed-by: Johan Almquist <johan.almquist@assaabloy.com>

jtag_interface: .speed can be NULL  when not needed

adapter_init (core.c) won't check speed configuration
of the selected interface if it's not needed (.speed = NULL).

When it's not needed, we can now omit adapter_khz in
init scripts and we don't have to implement dummy handlers
for speed_div and khz functions.

It also removes calls to adapter_khz in interface configuration
files when not used anymore.

Change-Id: I6eb1894385503fede542a368f297cec6565eed44
Signed-off-by: Franck Jullien <franck.jullien@gmail.com>
Reviewed-on: http://openocd.zylin.com/1131
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

armv7m: use ARM_MODE_THREAD core mode for algoorithm's

This makes sure we are using privileged mode when executing any loaders.

Change-Id: I18bf32ec92e1c76a66ab25e3712652bc3650b332
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1108
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

armv7m: restore core mode after executing algorithm

Make sure we restore the core mode after executing any algorithm.

We also now check that we actually need to swap the core mode, we may
already be in the correct mode.

Change-Id: Ia48af2c108e0f9868aae241bf25f60323503f092
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1107
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

armv7m: use generic arm read/write_core_reg

Change-Id: I0c15acc1278d2972269d294078495e6b069c830b
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/969
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

armv7m: use generic register core_cache

This removes the armv7m::core_cache and uses the generic arm::core_cache.

Change-Id: If854281b31486cea8be005008f6a71a691b4c208
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/968
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

armv7m: remove unused armv7m_regtype

This simplifies the armv7m_core_reg structure ready for the move to using
the generic struct arm_reg.

Change-Id: I8edb9d77cc54965d49cd2e754568ebcea4cf6964
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/967
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

armv7m: use generic arm::core_mode

To simplify things change over to using the generic core_mode struct rather
than maintaining a armv7m specific one.

Change-Id: Ibf32b785d896fef4f33307fabe0d8eb266f7086f
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/966
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

helper: hexify correctly handle signed chars

The current implementation of hexify was not correctly handling signed chars.

This function is currently used by the ti-icdi driver and as such was causing
random write issues.

As a note perhaps a better long term fix would be to change to using uint8_t
buffers rather than char. This will require changes to the ti-icdi driver
aswell.

Change-Id: I572e69ff2b99227a7d412de056458c0393794b03
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1124
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

rtos: fix error message

Probably a copy&paste error or remainings of some older version.

Change-Id: Ifb81a9a1fe8242f3b114cd0686dd264fbaad4920
Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/1123
Tested-by: jenkins
Reviewed-by: Øyvind Harboe <oyvindharboe@gmail.com>

rtos: do not use LOG_OUTPUT

LOG_OUTPUT is not intended for general output so use the correct LOG_*
functions instead.

Change-Id: I48d0fe765637024dbafc68f2ea08219d3ff42754
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1104
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

rtos: Fix regression preventing use of first RTOS & clean up rtos_qsymbol()

ThreadX support was not working due to it being first in the list of RTOS - regression.
Auto-detect off, an RTOS was always be marked as successfully detected, even if symbols are not found.
Lines 223-227 were unnecessary as they are done in rtos_try_next()
Added lots of comments
Improved readability by separating: GDB not finding a symbol vs no more symbols being available

Regression caused by patch which was allowed only 52 minutes for review : http://openocd.zylin.com/895

Change-Id: Ib4decb01db595ddb3796837c6d8338ce6b9a91ca
Signed-off-by: Evan Hunter <ehunter@broadcom.com>
Reviewed-on: http://openocd.zylin.com/986
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

jtag: only change state if necessary

All the other drivers will only change the state if required.
This brings all the other drivers inline with this behaviour.

The original issue relates to problems on xscale commit 7989000e0969c1ccf69acbc3ce649a020bc1ee66

Change-Id: Ifc90ec2eef68a70a14f37c00931a07982bfa200c
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1114
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

libusb: idProduct of USB device may be zero

There is no constraint about idProduct in USB spec.  So, pids[i] may be 0 for USB devices.

Change-Id: I19d8974f4e7082e8b7e1f2d33c019ac4e61bc1e2
Signed-off-by: Hsiangkai Wang <hsiangkai@gmail.com>
Reviewed-on: http://openocd.zylin.com/1091
Tested-by: jenkins
Reviewed-by: Xiaofan <xiaofanc@gmail.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

hla: enable DWT component and fix watchpoints

The makes sure the DWT component is always enabled so that watchpoints
work as expected.

This does need merging into the existing cortex_m logic, however at the
moment this is non trivial.

Change-Id: Ic6cccd1badb51f70a2ca8ea9ab6923788a94c1bf
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1122
Tested-by: jenkins
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

flash: add stm32lx loader Hard Fault workaround

An issue has been seen with the stm32lx flash driver that if a
power cycle/reset is applied after a erase, any ram loader will Hard Fault
on execution.

A similar issue is mentioned in the errata for the device.
Two solution's seem to workaround this issue:
1, Handle the exception, this means adding exception vectors to the loader
   and changing the exception address using nvic vtor register.
2. falling back to using slower direct page writes - approx 50% slower.

Using solution 1 would mean restrictions are placed on the loader location.
Solution 2 was chosen mainly as it was simpler too implement.

Change-Id: I429f06b5a3e3b1d8de90071a88a7df11fc9b46a7
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1010
Tested-by: jenkins

flash: reduce stm32lx loader timeout

Waiting 20secs is a bit much excessive, we could probably reduce to 5.

Change-Id: Iffb97adb99c2541a075fe78dbc88a53ddf340214
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1009
Tested-by: jenkins

flash: cleanup stm32lx driver

Handle any leading bytes upto the next 128 byte page, enabling us to safely
use the faster page write.

Rather than use a separate word/byte write to program any trailing bytes
we use a combined write function.

Use memcpy for byte writes and change loader to using bytes.

Change-Id: Ie0164a30388f018dd00e752cf5ff87d4f96ced97
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1008
Tested-by: jenkins

docs: update stm32f1x/stm32f2x driver info

As we use the two ST flash drivers for multiple stm32 variants update the
docs as to which targets use which driver.

Change-Id: I84943ff45482a22b3d3dd8491bb4242d79415939
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/990
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

flash: stm32f2x support write protection

Change-Id: I42662681104bb06e28148229464ae144c4a54538
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/989
Tested-by: jenkins

flash: add stm32f2x flash lock/unlock cmds

Change-Id: I35344cc47fa4f0a49c034455c5abf479faa0344a
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/988
Tested-by: jenkins
Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>

build: replace deprecated AM_CONFIG_HEADER with AC_CONFIG_HEADERS

automake-1.13 has now deprecated AM_CONFIG_HEADER, use the correct
AC_CONFIG_HEADERS instead.

Change-Id: I8adaec64cbad7f7318ff69091176c30b707cbb0b
Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/1117
Tested-by: jenkins
Reviewed-by: Mikko Viitamäki <mikko.viitamaki@gmail.com>
Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>

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