Spencer Oliver [Mon, 24 Sep 2012 11:29:47 +0000 (12:29 +0100)]
ftdi: fix adapter_init rclk fallback
adapter_init expects jtag_get_speed (via ftdi_khz) to return a valid
fallback speed if the adapter does not support rclk. The call was failing
and so was the rest of the adapter init.
The makes the new ftdi driver emulate the old ftdi driver.
George Harris [Fri, 17 Aug 2012 00:53:32 +0000 (17:53 -0700)]
Added SPIFI flash driver, algorithms, and docs
Added a flash driver designed to allow program/erase of
memory-mapped SPI flash chips for LPC43xx/LPC18xx family
micros. This driver includes three algorithms - erase,
write, and SPIFI peripheral initialization (to allow
memory-mapped access after a reset). The driver has been
added to the flash driver table (drivers.c), and the
OpenOCD documentation has been updated to include the flash
driver configuration command.
George Harris [Fri, 17 Aug 2012 00:35:39 +0000 (17:35 -0700)]
SPI nor drivers refactor
Moved common SPI flash driver code (device table,
common commands) into flash/nor/spi.c and spi.h.
Updated flash/nor/stmsmi.c to reflect this refactor.
In at91sam3.c for Atmel SAM3 flash support, there are arithmetic errors in the functions sam3_page_read() and sam3_page_write().
Address locations are computed incorrectly due to an extra addition operation. This leads to memory locations being skipped during
flash writes and reads.
Smaller programs are written successfully into flash, with memory gaps, while larger programs of legitimate size fail because the
skipped memory is not utilized and therefore unavailable.
The changes address this condition, and have been tested with an Atmel SAM3X-EK evaluation board.
Spencer Oliver [Thu, 30 Aug 2012 11:53:34 +0000 (12:53 +0100)]
cortex_m: suggest using hardware srst if VECTRESET used
If the target does not support SYSRESETREQ we fall back to using VECTRESET.
This however does not reset the peripherals and we issue a warning to the user
to suggest using a reset-init script.
Also suggest that using hardware srst will give them the same functionality
as using SYSRESETREQ.
Change-Id: Ie1781c4b849fed66c52222e6539735537c879fb3 Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/802 Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com> Tested-by: jenkins Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Salvador Arroyo [Sat, 25 Aug 2012 14:22:58 +0000 (16:22 +0200)]
Pic32mx: make row programming work with any offset
In function pic32mx_write_block() if the parameter
offset is not a multiple of row size the row offset
(offset % row_size) will be ignored by the flash
controller, shifting the code to the beginning of
the row.
Word programming gets it right.
Change-Id: I134913e3d533688f791bbcb0c6e8983524197f3c Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/796 Tested-by: jenkins Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk> Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Spencer Oliver [Tue, 21 Aug 2012 10:10:18 +0000 (11:10 +0100)]
server: warn if user changes server port after init
So the user can view the current port number these cmds were changed to COMMAND_ANY.
However this means that the user can also attempt to change the port number after init,
even though this is not supported. Issue a warning that this is not supported.
ftdi: fix overflow if last field of a scan is empty
The last bit of a scan is clocked during TAP movement so it's necessary
for the last field to have at least one bit. Strip trailing empty fields
and make sure the TAP is not affected if there's nothing to scan. Clients
probably shouldn't add empty fields so add a debug message to be able to
track and fix them.
Change-Id: I27552568bc11146570b9b99ed8a1ae81b5fb2c50 Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/794 Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk> Tested-by: jenkins Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
mpsse: check available buffer space even for discarded data scans
When there's no data to scan in or out, we still use the clock data out
command and fill the buffer with zeroes, so make sure the buffer is
checked for available space.
Change-Id: Ia6005c40c81f7fdb89379f1b5023fe383184d210 Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/793 Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk> Tested-by: jenkins Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Spencer Oliver [Tue, 28 Aug 2012 11:19:26 +0000 (12:19 +0100)]
adapter: add 'adapter speed:' prefix to output
Currently only the adapter speed is printed, which can be rather misleading
when DEBUG_INFO is disabled, all the user sees is
6000 kHz
instead lets print
adapter speed: 6000 kHz
Freddie Chopin [Fri, 17 Aug 2012 09:54:24 +0000 (11:54 +0200)]
Add another scripts search path for Windows builds
Add single "scripts" folder to search path for Windows OpenOCD builds
that don't use cygwin
bin/openocd.exe
scripts/interface/dummy.cfg
scripts/target/at91eb40a.cfg
Do some refactoring of current code (thx to Andreas).
The number if longwords or "sections" (Freescale term) written for a Kinetis
flash write (4, 8, or 16 bytes depending on the part density/granularity) are
now rounded up to ensure there are no truncations when the desired write is
not a multiple of the minimum write size.
Always scan out all bits, but make sure only the allowed number of bytes
end up in the caller-provided buffer. Discard the rest by adding another
scan field when size < 4.
Rewrite the endianness callback to avoid reading outside allocated memory.
Make it directly usable as a callback without the need for a wrapper. Move
the shared callback to a more suitable home in arm7_9_common.
Salvador Arroyo [Sun, 5 Aug 2012 09:18:27 +0000 (11:18 +0200)]
Severe bug in Pracc code
The function wait_for_pracc_rw() fails if Pracc bit is 0.
The variable ejtag_ctrl is loaded with the content of the
control register in the first scan.
In the second scan Pracc bit is scanned out as 0, letting
the proccesor go. The result is unpredictable.
All the strange data corruption when scanning at certain
frequencies, or the strange delays needed when entering
or leaving fasdata area are retated to this bug.
Now the code works at any scan frequency, tested up to 15000Khz
and indepently of processor speed, tested at 31.25Khz and 4/8Mhz.
cfi: fix type-punning warnings in cfi_spansion_write_block
Retest the condition when needed, instead of abusing the common_magic
field as a flag. There are only two options here. Either it's an armv7m or
it's another arm. is_arm(...) will return true even for armv7m, so it's
imperative to check in the right order.
Change-Id: Ic227f19f7babf1b0b0fe075f9a3abc4eabc7d5f1 Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/779 Tested-by: jenkins Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk> Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Freddie Chopin [Thu, 2 Aug 2012 12:14:47 +0000 (14:14 +0200)]
Add missing files (header and .txt) for release.
make distcheck is used to make packages with OpenOCD release, this
command uses information from Makefile.am files to know which files
should be included in the package and which can be left only in
repository. This patch makes a few headers from recent JTAG drivers
and one txt file with info about target tcl config files included in
released packages.
Cache type register C0.C is read-only, and display
hard core configuration information.
This information is unlikely be changed in runtime.
- removed C0.C access when result is not used in
arm946e_invalidate_dcache()
- access C0.C only once per target, store result
in cp15_cache_info field of target structure
- fix cache index count calculation
Global variables 'dc' 'ic' had been used in the code
to keep target's state of D-cache and I-cache
on debug entry.
This may lead to incorrect operation in configurations
with multiple cores and unequal cache states.
Fix: move cache state to the appropriate bits of the
'cp15_control_reg' field (already present but unused).
Vaule of cp15 control register stored here on
arm946e_post_debug_entry(), and analyzed later
in arm946e_write_memory().
flash: stm32f1x: Pad odd byte writes early to avoid 16-bit writes
For odd byte counts, stm32x_write() pads the last byte and writes it using
a discrete 16-bit access. The stlink debugger can't issue 16-bit writes so
it fails for odd byte writes.
This patch changes stm32x_write() to pad odd byte writes into a new buffer
and use the normal code path with a single block write. The fallback path,
when working area cannot be allocated, has to use 16-bit writes though
which means that sufficient working area is required for stlink and odd
byte writes.
Change-Id: I4c5dc456300b6e1056f76b0095be8aceee3e954f Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/756 Tested-by: jenkins Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk> Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Salvador Arroyo [Thu, 7 Jun 2012 14:44:34 +0000 (16:44 +0200)]
Bugfixes in mips32_pracc.c
When testing a pic32mx220f032b with different values
for adapter_khz and cpu clocks i was getting a lot of
corrupted data from the chip. From time to time
openocd fails with segmentation faults or is aborted
due to memory corruption.
Change-Id: I134743f75c477b3d55dc74ae4474598e153b4a4a Signed-off-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Reviewed-on: http://openocd.zylin.com/690 Tested-by: jenkins Reviewed-by: Andreas Fritiofson <andreas.fritiofson@gmail.com> Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
flash: reduce code duplication in stm32 flash probe
Remove a lot of the repetitive code in stm32f1x flash probe by converting
the large if-selector to a switch, moving the common checks outside it and
concentrating the failure handling to a single point.
Do the same with stm32f2x and stm32lx for consistency.
Change-Id: Ic0ecfb1533c49f5d2108cda5fd20c8372d7c71ef Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/746 Tested-by: jenkins Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk> Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Vandra Akos [Sun, 27 May 2012 10:36:55 +0000 (12:36 +0200)]
lpc1768.cfg abstracted and moved to lpc17xx.cfg
- Moved variant-independent code to lpc17xx.cfg, which will be included from
lpc17??.cfg files automatically.
- lpc1768.cfg filled with variant-dependent code.
FT232H chips are new highspeed devices from FTDI. Basically these are a half of FT2232H (or a quarter of FT4232H), so only one channel which can be used as OpenOCD
interface. The chips are supported by libftdi 0.20 or later and by ftd2xx 2.08.12 or later.
Change-Id: Ic9a2c279167c3419a24f0d6befacbb83c4ffeb25 Signed-off-by: Freddie Chopin <freddie.chopin@gmail.com>
Reviewed-on: http://openocd.zylin.com/736 Tested-by: jenkins Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk> Reviewed-by: Salvador Arroyo <sarroyofdez@yahoo.es>
Alan Bowman [Wed, 25 Jul 2012 18:58:11 +0000 (19:58 +0100)]
Test the incoming debug_level, rather than the current.
The existing code tests the current debug_level for validity, allowing
the user to set it too high and never change it after that. The new
behaviour is to test the debug_level that the user has requested.
Based on ft2232.c but uses the MPSSE layer for low-level access, greatly
simplifying the JTAG logic. Remove all libftdi/FTD2XX code and all layout
specific code. Layout specifications are instead handled in Tcl.
Use a signal abstraction to enable Tcl configuration files to define
outputs for one or several FTDI GPIO. These outputs can then be
controlled using the ftdi_set_signal command. Special signal names are
reserved for nTRST, nSRST and LED (for blink) so that they, if defined,
will be used for their customary purpose.
Depending on the type of buffer attached to the FTDI GPIO, the outputs
have to be controlled differently. In order to support tristateable
signals such as nSRST, both a data GPIO and an output-enable GPIO can be
specified for each signal. The following output buffer configurations are
supported:
* Push-pull with one FTDI output as (non-)inverted data line
* Open drain with one FTDI output as (non-)inverted output-enable
* Tristate with one FTDI output as (non-)inverted data line and another
FTDI output as (non-)inverted output-enable
* Unbuffered, using the FTDI GPIO as a tristate output directly by
switching data and direction as necessary
The data and output-enables are specified as 16-bit bitmasks,
corresponding to the concatenation of the high and low FTDI GPIO
registers. To specify an unbuffered output, use the same bitmask
for both data and output-enable.
The adapter configuration file must also specify default values for the
FTDI data and direction GPIO registers, and the channel being used (if
different from 0).
Change-Id: I287a41d4c696cf5fc74eb10d5e63578b0dc7f826 Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/452 Tested-by: jenkins Reviewed-by: Peter Stuge <peter@stuge.se>
This is a higher-level libftdi replacement for use when implementing
protocol drivers for FT2232, FT2232H or FT4232H. It takes care of device
open/close and, unlike libftdi, also MPSSE command abstraction, command
queueing, buffer handling and return data parsing.
The FTDI device is accessed through libusb-1.0 in asynchronous mode.
Change-Id: I051adb574dcc39f8ca9cd7f6dbe6ae4aeea5f4c8 Signed-off-by: Andreas Fritiofson <andreas.fritiofson@gmail.com>
Reviewed-on: http://openocd.zylin.com/451 Tested-by: jenkins Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com> Reviewed-by: Peter Stuge <peter@stuge.se>
Spencer Oliver [Wed, 27 Jun 2012 22:25:07 +0000 (23:25 +0100)]
target: detect correct Cortex-M tar auto increment size
The ADIv5 spec guarentees that tar_autoincr_block will be 10bits.
Make this the default for Cortex-M family until we detect a Cortex-M3/M4,
we then change autoincrement to 12bits.
Change-Id: Ie8c89134aa036879bdd8a3c312cee9715dbc6913 Signed-off-by: Spencer Oliver <spen@spen-soft.co.uk>
Reviewed-on: http://openocd.zylin.com/730 Tested-by: jenkins Reviewed-by: simon qian <simonqian.openocd@gmail.com> Reviewed-by: Freddie Chopin <freddie.chopin@gmail.com>
Vandra Akos [Fri, 25 May 2012 01:02:55 +0000 (03:02 +0200)]
target.c, jim_target_md using command_print_sameline
jim_target_md is supposed to print out results with command_print in
hexdump format. It was using command_print which appends a newline character
aftre every invocation. Using command_print_sameline instead
Vandra Akos [Thu, 24 May 2012 13:17:34 +0000 (15:17 +0200)]
target/target.c, jim_target_md refractored
- Added a few lines of comment before the function explaining the usage and the
output generated by the command
- Added a few lines of comment in the body explaining what is happening to improve
code readability
- Renamed a few variables to improve readability:
* a -> addr
* b -> dwidth
* c -> count
- Added a new variable, named byte to contain the number of bytes to read, instead
of overwriting the count parameter, to avoid confusion between the two values.
Interface drivers regularly need to keep track of where each part of a long
read buffer should be copied, once that data arrives. Both source and
destination are often at an arbitrary bit offset.
This queued bit-level copy can help with that, by allowing the driver to
perform postponed reads from the receive buffer already when building the
transmit buffer, and have those reads executed at a later time when data is
available.
For simplicity, it uses the linked list implementation list.h imported from
the Linux kernel.
Mike Crowe [Tue, 29 May 2012 20:23:27 +0000 (15:23 -0500)]
topic: flash: description/id added for ATSAM3SD8C
New flash description for ATSAM3SD8C used on SAM3S-EK2 development boards. Name used
is "at91sam3sd8c" and chipid is 0x29ab0a60. Mirrors description of other similar parts.
Change-Id: I7fc4b82e7969451645ab067223663f08b76d866b Signed-off-by: Mike Crowe <mpcrowe@gmail.com>
Reviewed-on: http://openocd.zylin.com/684 Tested-by: jenkins Reviewed-by: Spencer Oliver <spen@spen-soft.co.uk>
In the previous iteration, the page counter for erases would not be updated with
the erase size. This patch keeps the page counter synced with the sector counter.