1 \input texinfo @c -*-texinfo-*-
3 @setfilename openocd.info
4 @settitle OpenOCD User's Guide
5 @dircategory Development
7 * OpenOCD: (openocd). OpenOCD User's Guide
16 This User's Guide documents
17 release @value{VERSION},
18 dated @value{UPDATED},
19 of the Open On-Chip Debugger (OpenOCD).
22 @item Copyright @copyright{} 2008 The OpenOCD Project
23 @item Copyright @copyright{} 2007-2008 Spencer Oliver @email{spen@@spen-soft.co.uk}
24 @item Copyright @copyright{} 2008 Oyvind Harboe @email{oyvind.harboe@@zylin.com}
25 @item Copyright @copyright{} 2008 Duane Ellis @email{openocd@@duaneellis.com}
26 @item Copyright @copyright{} 2009 David Brownell
30 Permission is granted to copy, distribute and/or modify this document
31 under the terms of the GNU Free Documentation License, Version 1.2 or
32 any later version published by the Free Software Foundation; with no
33 Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
34 Texts. A copy of the license is included in the section entitled ``GNU
35 Free Documentation License''.
40 @titlefont{@emph{Open On-Chip Debugger:}}
42 @title OpenOCD User's Guide
43 @subtitle for release @value{VERSION}
44 @subtitle @value{UPDATED}
47 @vskip 0pt plus 1filll
56 @top OpenOCD User's Guide
62 * About:: About OpenOCD
63 * Developers:: OpenOCD Developers
64 * JTAG Hardware Dongles:: JTAG Hardware Dongles
65 * About JIM-Tcl:: About JIM-Tcl
66 * Running:: Running OpenOCD
67 * OpenOCD Project Setup:: OpenOCD Project Setup
68 * Config File Guidelines:: Config File Guidelines
69 * Daemon Configuration:: Daemon Configuration
70 * Interface - Dongle Configuration:: Interface - Dongle Configuration
71 * Reset Configuration:: Reset Configuration
72 * TAP Declaration:: TAP Declaration
73 * CPU Configuration:: CPU Configuration
74 * Flash Commands:: Flash Commands
75 * NAND Flash Commands:: NAND Flash Commands
76 * PLD/FPGA Commands:: PLD/FPGA Commands
77 * General Commands:: General Commands
78 * Architecture and Core Commands:: Architecture and Core Commands
79 * JTAG Commands:: JTAG Commands
80 * Boundary Scan Commands:: Boundary Scan Commands
82 * GDB and OpenOCD:: Using GDB and OpenOCD
83 * Tcl Scripting API:: Tcl Scripting API
84 * Upgrading:: Deprecated/Removed Commands
85 * Target Library:: Target Library
86 * FAQ:: Frequently Asked Questions
87 * Tcl Crash Course:: Tcl Crash Course
88 * License:: GNU Free Documentation License
90 @comment DO NOT use the plain word ``Index'', reason: CYGWIN filename
91 @comment case issue with ``Index.html'' and ``index.html''
92 @comment Occurs when creating ``--html --no-split'' output
93 @comment This fix is based on: http://sourceware.org/ml/binutils/2006-05/msg00215.html
94 * OpenOCD Concept Index:: Concept Index
95 * Command and Driver Index:: Command and Driver Index
102 OpenOCD was created by Dominic Rath as part of a diploma thesis written at the
103 University of Applied Sciences Augsburg (@uref{http://www.fh-augsburg.de}).
104 Since that time, the project has grown into an active open-source project,
105 supported by a diverse community of software and hardware developers from
108 @section What is OpenOCD?
112 The Open On-Chip Debugger (OpenOCD) aims to provide debugging,
113 in-system programming and boundary-scan testing for embedded target
116 @b{JTAG:} OpenOCD uses a ``hardware interface dongle'' to communicate
117 with the JTAG (IEEE 1149.1) compliant TAPs on your target board.
118 A @dfn{TAP} is a ``Test Access Port'', a module which processes
119 special instructions and data. TAPs are daisy-chained within and
120 between chips and boards.
122 @b{Dongles:} OpenOCD currently supports many types of hardware dongles: USB
123 based, parallel port based, and other standalone boxes that run
124 OpenOCD internally. @xref{JTAG Hardware Dongles}.
126 @b{GDB Debug:} It allows ARM7 (ARM7TDMI and ARM720t), ARM9 (ARM920T,
127 ARM922T, ARM926EJ--S, ARM966E--S), XScale (PXA25x, IXP42x) and
128 Cortex-M3 (Stellaris LM3 and ST STM32) based cores to be
129 debugged via the GDB protocol.
131 @b{Flash Programing:} Flash writing is supported for external CFI
132 compatible NOR flashes (Intel and AMD/Spansion command set) and several
133 internal flashes (LPC1700, LPC2000, AT91SAM7, AT91SAM3U, STR7x, STR9x, LM3, and
134 STM32x). Preliminary support for various NAND flash controllers
135 (LPC3180, Orion, S3C24xx, more) controller is included.
137 @section OpenOCD Web Site
139 The OpenOCD web site provides the latest public news from the community:
141 @uref{http://openocd.berlios.de/web/}
143 @section Latest User's Guide:
145 The user's guide you are now reading may not be the latest one
146 available. A version for more recent code may be available.
147 Its HTML form is published irregularly at:
149 @uref{http://openocd.berlios.de/doc/html/index.html}
151 PDF form is likewise published at:
153 @uref{http://openocd.berlios.de/doc/pdf/openocd.pdf}
155 @section OpenOCD User's Forum
157 There is an OpenOCD forum (phpBB) hosted by SparkFun:
159 @uref{http://forum.sparkfun.com/viewforum.php?f=18}
163 @chapter OpenOCD Developer Resources
166 If you are interested in improving the state of OpenOCD's debugging and
167 testing support, new contributions will be welcome. Motivated developers
168 can produce new target, flash or interface drivers, improve the
169 documentation, as well as more conventional bug fixes and enhancements.
171 The resources in this chapter are available for developers wishing to explore
172 or expand the OpenOCD source code.
174 @section OpenOCD Subversion Repository
176 You can download the current SVN version with an SVN client of your
177 choice from the following repositories:
179 @uref{svn://svn.berlios.de/openocd/trunk}
183 @uref{http://svn.berlios.de/svnroot/repos/openocd/trunk}
185 Using the SVN command line client, you can use the following command to
186 fetch the latest version (make sure there is no (non-svn) directory
187 called "openocd" in the current directory):
189 svn checkout svn://svn.berlios.de/openocd/trunk openocd
191 If you prefer GIT based tools, the @command{git-svn} package works too:
193 git svn clone -s svn://svn.berlios.de/openocd
195 The ``README'' file contains the instructions for building the project
198 Developers that want to contribute patches to the OpenOCD system are
199 @b{strongly} encouraged to base their work off of the most recent trunk
200 revision. Patches created against older versions may require additional
201 work from their submitter in order to be updated for newer releases.
203 @section Doxygen Developer Manual
205 During the development of the 0.2.0 release, the OpenOCD project began
206 providing a Doxygen reference manual. This document contains more
207 technical information about the software internals, development
208 processes, and similar documentation:
210 @uref{http://openocd.berlios.de/doc/doxygen/index.html}
212 This document is a work-in-progress, but contributions would be welcome
213 to fill in the gaps. All of the source files are provided in-tree,
214 listed in the Doxyfile configuration in the top of the repository trunk.
216 @section OpenOCD Developer Mailing List
218 The OpenOCD Developer Mailing List provides the primary means of
219 communication between developers:
221 @uref{https://lists.berlios.de/mailman/listinfo/openocd-development}
223 All drivers developers are enouraged to also subscribe to the list of
224 SVN commits to keep pace with the ongoing changes:
226 @uref{https://lists.berlios.de/mailman/listinfo/openocd-svn}
229 @node JTAG Hardware Dongles
230 @chapter JTAG Hardware Dongles
239 Defined: @b{dongle}: A small device that plugins into a computer and serves as
240 an adapter .... [snip]
242 In the OpenOCD case, this generally refers to @b{a small adapater} one
243 attaches to your computer via USB or the Parallel Printer Port. The
244 execption being the Zylin ZY1000 which is a small box you attach via
245 an ethernet cable. The Zylin ZY1000 has the advantage that it does not
246 require any drivers to be installed on the developer PC. It also has
247 a built in web interface. It supports RTCK/RCLK or adaptive clocking
248 and has a built in relay to power cycle targets remotely.
251 @section Choosing a Dongle
253 There are three things you should keep in mind when choosing a dongle.
256 @item @b{Voltage} What voltage is your target? 1.8, 2.8, 3.3, or 5V? Does your dongle support it?
257 @item @b{Connection} Printer Ports - Does your computer have one?
258 @item @b{Connection} Is that long printer bit-bang cable practical?
259 @item @b{RTCK} Do you require RTCK? Also known as ``adaptive clocking''
262 @section Stand alone Systems
264 @b{ZY1000} See: @url{http://www.zylin.com/zy1000.html} Technically, not a
265 dongle, but a standalone box. The ZY1000 has the advantage that it does
266 not require any drivers installed on the developer PC. It also has
267 a built in web interface. It supports RTCK/RCLK or adaptive clocking
268 and has a built in relay to power cycle targets remotely.
270 @section USB FT2232 Based
272 There are many USB JTAG dongles on the market, many of them are based
273 on a chip from ``Future Technology Devices International'' (FTDI)
274 known as the FTDI FT2232; this is a USB full speed (12 Mbps) chip.
275 See: @url{http://www.ftdichip.com} for more information.
276 In summer 2009, USB high speed (480 Mbps) versions of these FTDI
277 chips are starting to become available in JTAG adapters.
281 @* Link @url{http://www.hs-augsburg.de/~hhoegl/proj/usbjtag/usbjtag.html}
283 @* See: @url{http://www.amontec.com/jtagkey.shtml}
285 @* See: @url{http://www.amontec.com/jtagkey2.shtml}
287 @* See: @url{http://www.oocdlink.com} By Joern Kaipf
289 @* See: @url{http://www.signalyzer.com}
290 @item @b{evb_lm3s811}
291 @* See: @url{http://www.luminarymicro.com} - The Stellaris LM3S811 eval board has an FTD2232C chip built in.
292 @item @b{luminary_icdi}
293 @* See: @url{http://www.luminarymicro.com} - Luminary In-Circuit Debug Interface (ICDI) Board, included in the Stellaris LM3S9B90 and LM3S9B92 Evaluation Kits.
294 @item @b{olimex-jtag}
295 @* See: @url{http://www.olimex.com}
297 @* See: @url{http://www.tincantools.com}
298 @item @b{turtelizer2}
300 @uref{http://www.ethernut.de/en/hardware/turtelizer/index.html, Turtelizer 2}, or
301 @url{http://www.ethernut.de}
303 @* Link: @url{http://www.hitex.com/index.php?id=383}
305 @* Link @url{http://www.hitex.com/stm32-stick}
306 @item @b{axm0432_jtag}
307 @* Axiom AXM-0432 Link @url{http://www.axman.com}
309 @* Link @url{http://www.hitex.com/index.php?id=cortino}
312 @section USB JLINK based
313 There are several OEM versions of the Segger @b{JLINK} adapter. It is
314 an example of a micro controller based JTAG adapter, it uses an
315 AT91SAM764 internally.
318 @item @b{ATMEL SAMICE} Only works with ATMEL chips!
319 @* Link: @url{http://www.atmel.com/dyn/products/tools_card.asp?tool_id=3892}
320 @item @b{SEGGER JLINK}
321 @* Link: @url{http://www.segger.com/jlink.html}
323 @* Link: @url{http://www.iar.com/website1/1.0.1.0/369/1/index.php}
326 @section USB RLINK based
327 Raisonance has an adapter called @b{RLink}. It exists in a stripped-down form on the STM32 Primer, permanently attached to the JTAG lines. It also exists on the STM32 Primer2, but that is wired for SWD and not JTAG, thus not supported.
330 @item @b{Raisonance RLink}
331 @* Link: @url{http://www.raisonance.com/products/RLink.php}
332 @item @b{STM32 Primer}
333 @* Link: @url{http://www.stm32circle.com/resources/stm32primer.php}
334 @item @b{STM32 Primer2}
335 @* Link: @url{http://www.stm32circle.com/resources/stm32primer2.php}
341 @* Link: @url{http://www.embedded-projects.net/usbprog} - which uses an Atmel MEGA32 and a UBN9604
343 @item @b{USB - Presto}
344 @* Link: @url{http://tools.asix.net/prg_presto.htm}
346 @item @b{Versaloon-Link}
347 @* Link: @url{http://www.simonqian.com/en/Versaloon}
349 @item @b{ARM-JTAG-EW}
350 @* Link: @url{http://www.olimex.com/dev/arm-jtag-ew.html}
353 @section IBM PC Parallel Printer Port Based
355 The two well known ``JTAG Parallel Ports'' cables are the Xilnx DLC5
356 and the MacGraigor Wiggler. There are many clones and variations of
361 @item @b{Wiggler} - There are many clones of this.
362 @* Link: @url{http://www.macraigor.com/wiggler.htm}
364 @item @b{DLC5} - From XILINX - There are many clones of this
365 @* Link: Search the web for: ``XILINX DLC5'' - it is no longer
366 produced, PDF schematics are easily found and it is easy to make.
368 @item @b{Amontec - JTAG Accelerator}
369 @* Link: @url{http://www.amontec.com/jtag_accelerator.shtml}
372 @* Link: @url{http://www.gateworks.com/products/avila_accessories/gw16042.php}
375 @*@uref{http://www.ccac.rwth-aachen.de/@/~michaels/@/index.php/hardware/@/armjtag,
376 Improved parallel-port wiggler-style JTAG adapter}
378 @item @b{Wiggler_ntrst_inverted}
379 @* Yet another variation - See the source code, src/jtag/parport.c
381 @item @b{old_amt_wiggler}
382 @* Unknown - probably not on the market today
385 @* Link: Most likely @url{http://www.olimex.com/dev/arm-jtag.html} [another wiggler clone]
388 @* Link: @url{http://www.amontec.com/chameleon.shtml}
394 @* ispDownload from Lattice Semiconductor
395 @url{http://www.latticesemi.com/lit/docs/@/devtools/dlcable.pdf}
398 @* From ST Microsystems;
399 @uref{http://www.st.com/stonline/@/products/literature/um/7889.pdf,
400 FlashLINK JTAG programing cable for PSD and uPSD}
408 @* An EP93xx based Linux machine using the GPIO pins directly.
411 @* Like the EP93xx - but an ATMEL AT91RM9200 based solution using the GPIO pins on the chip.
416 @chapter About JIM-Tcl
420 OpenOCD includes a small ``Tcl Interpreter'' known as JIM-Tcl.
421 This programming language provides a simple and extensible
424 All commands presented in this Guide are extensions to JIM-Tcl.
425 You can use them as simple commands, without needing to learn
426 much of anything about Tcl.
427 Alternatively, can write Tcl programs with them.
429 You can learn more about JIM at its website, @url{http://jim.berlios.de}.
432 @item @b{JIM vs. Tcl}
433 @* JIM-TCL is a stripped down version of the well known Tcl language,
434 which can be found here: @url{http://www.tcl.tk}. JIM-Tcl has far
435 fewer features. JIM-Tcl is a single .C file and a single .H file and
436 implements the basic Tcl command set. In contrast: Tcl 8.6 is a
437 4.2 MB .zip file containing 1540 files.
439 @item @b{Missing Features}
440 @* Our practice has been: Add/clone the real Tcl feature if/when
441 needed. We welcome JIM Tcl improvements, not bloat.
444 @* OpenOCD configuration scripts are JIM Tcl Scripts. OpenOCD's
445 command interpreter today is a mixture of (newer)
446 JIM-Tcl commands, and (older) the orginal command interpreter.
449 @* At the OpenOCD telnet command line (or via the GDB mon command) one
450 can type a Tcl for() loop, set variables, etc.
451 Some of the commands documented in this guide are implemented
452 as Tcl scripts, from a @file{startup.tcl} file internal to the server.
454 @item @b{Historical Note}
455 @* JIM-Tcl was introduced to OpenOCD in spring 2008.
457 @item @b{Need a crash course in Tcl?}
458 @*@xref{Tcl Crash Course}.
463 @cindex command line options
465 @cindex directory search
467 The @option{--help} option shows:
471 --help | -h display this help
472 --version | -v display OpenOCD version
473 --file | -f use configuration file <name>
474 --search | -s dir to search for config files and scripts
475 --debug | -d set debug level <0-3>
476 --log_output | -l redirect log output to file <name>
477 --command | -c run <command>
478 --pipe | -p use pipes when talking to gdb
481 By default OpenOCD reads the file configuration file ``openocd.cfg''
482 in the current directory. To specify a different (or multiple)
483 configuration file, you can use the ``-f'' option. For example:
486 openocd -f config1.cfg -f config2.cfg -f config3.cfg
489 Once started, OpenOCD runs as a daemon, waiting for connections from
490 clients (Telnet, GDB, Other).
492 If you are having problems, you can enable internal debug messages via
495 Also it is possible to interleave JIM-Tcl commands w/config scripts using the
496 @option{-c} command line switch.
498 To enable debug output (when reporting problems or working on OpenOCD
499 itself), use the @option{-d} command line switch. This sets the
500 @option{debug_level} to "3", outputting the most information,
501 including debug messages. The default setting is "2", outputting only
502 informational messages, warnings and errors. You can also change this
503 setting from within a telnet or gdb session using @command{debug_level
504 <n>} (@pxref{debug_level}).
506 You can redirect all output from the daemon to a file using the
507 @option{-l <logfile>} switch.
509 Search paths for config/script files can be added to OpenOCD by using
510 the @option{-s <search>} switch. The current directory and the OpenOCD
511 target library is in the search path by default.
513 For details on the @option{-p} option. @xref{Connecting to GDB}.
515 Note! OpenOCD will launch the GDB & telnet server even if it can not
516 establish a connection with the target. In general, it is possible for
517 the JTAG controller to be unresponsive until the target is set up
518 correctly via e.g. GDB monitor commands in a GDB init script.
520 @node OpenOCD Project Setup
521 @chapter OpenOCD Project Setup
523 To use OpenOCD with your development projects, you need to do more than
524 just connecting the JTAG adapter hardware (dongle) to your development board
525 and then starting the OpenOCD server.
526 You also need to configure that server so that it knows
527 about that adapter and board, and helps your work.
529 @section Hooking up the JTAG Adapter
531 Today's most common case is a dongle with a JTAG cable on one side
532 (such as a ribbon cable with a 10-pin or 20-pin IDC connector)
533 and a USB cable on the other.
534 Instead of USB, some cables use Ethernet;
535 older ones may use a PC parallel port, or even a serial port.
538 @item @emph{Start with power to your target board turned off},
539 and nothing connected to your JTAG adapter.
540 If you're particularly paranoid, unplug power to the board.
541 It's important to have the ground signal properly set up,
542 unless you are using a JTAG adapter which provides
543 galvanic isolation between the target board and the
546 @item @emph{Be sure it's the right kind of JTAG connector.}
547 If your dongle has a 20-pin ARM connector, you need some kind
548 of adapter (or octopus, see below) to hook it up to
549 boards using 14-pin or 10-pin connectors ... or to 20-pin
550 connectors which don't use ARM's pinout.
552 In the same vein, make sure the voltage levels are compatible.
553 Not all JTAG adapters have the level shifters needed to work
554 with 1.2 Volt boards.
556 @item @emph{Be certain the cable is properly oriented} or you might
557 damage your board. In most cases there are only two possible
558 ways to connect the cable.
559 Connect the JTAG cable from your adapter to the board.
560 Be sure it's firmly connected.
562 In the best case, the connector is keyed to physically
563 prevent you from inserting it wrong.
564 This is most often done using a slot on the board's male connector
565 housing, which must match a key on the JTAG cable's female connector.
566 If there's no housing, then you must look carefully and
567 make sure pin 1 on the cable hooks up to pin 1 on the board.
568 Ribbon cables are frequently all grey except for a wire on one
569 edge, which is red. The red wire is pin 1.
571 Sometimes dongles provide cables where one end is an ``octopus'' of
572 color coded single-wire connectors, instead of a connector block.
573 These are great when converting from one JTAG pinout to another,
574 but are tedious to set up.
575 Use these with connector pinout diagrams to help you match up the
576 adapter signals to the right board pins.
578 @item @emph{Connect the adapter's other end} once the JTAG cable is connected.
579 A USB, parallel, or serial port connector will go to the host which
580 you are using to run OpenOCD.
581 For Ethernet, consult the documentation and your network administrator.
583 For USB based JTAG adapters you have an easy sanity check at this point:
584 does the host operating system see the JTAG adapter?
586 @item @emph{Connect the adapter's power supply, if needed.}
587 This step is primarily for non-USB adapters,
588 but sometimes USB adapters need extra power.
590 @item @emph{Power up the target board.}
591 Unless you just let the magic smoke escape,
592 you're now ready to set up the OpenOCD server
593 so you can use JTAG to work with that board.
597 Talk with the OpenOCD server using
598 telnet (@code{telnet localhost 4444} on many systems) or GDB.
599 @xref{GDB and OpenOCD}.
601 @section Project Directory
603 There are many ways you can configure OpenOCD and start it up.
605 A simple way to organize them all involves keeping a
606 single directory for your work with a given board.
607 When you start OpenOCD from that directory,
608 it searches there first for configuration files, scripts,
609 and for code you upload to the target board.
610 It is also the natural place to write files,
611 such as log files and data you download from the board.
613 @section Configuration Basics
615 There are two basic ways of configuring OpenOCD, and
616 a variety of ways you can mix them.
617 Think of the difference as just being how you start the server:
620 @item Many @option{-f file} or @option{-c command} options on the command line
621 @item No options, but a @dfn{user config file}
622 in the current directory named @file{openocd.cfg}
625 Here is an example @file{openocd.cfg} file for a setup
626 using a Signalyzer FT2232-based JTAG adapter to talk to
627 a board with an Atmel AT91SAM7X256 microcontroller:
630 source [find interface/signalyzer.cfg]
632 # GDB can also flash my flash!
633 gdb_memory_map enable
634 gdb_flash_program enable
636 source [find target/sam7x256.cfg]
639 Here is the command line equivalent of that configuration:
642 openocd -f interface/signalyzer.cfg \
643 -c "gdb_memory_map enable" \
644 -c "gdb_flash_program enable" \
645 -f target/sam7x256.cfg
648 You could wrap such long command lines in shell scripts,
649 each supporting a different development task.
650 One might re-flash the board with a specific firmware version.
651 Another might set up a particular debugging or run-time environment.
653 Here we will focus on the simpler solution: one user config
654 file, including basic configuration plus any TCL procedures
655 to simplify your work.
657 @section User Config Files
658 @cindex config file, user
659 @cindex user config file
660 @cindex config file, overview
662 A user configuration file ties together all the parts of a project
664 One of the following will match your situation best:
667 @item Ideally almost everything comes from configuration files
668 provided by someone else.
669 For example, OpenOCD distributes a @file{scripts} directory
670 (probably in @file{/usr/share/openocd/scripts} on Linux).
671 Board and tool vendors can provide these too, as can individual
672 user sites; the @option{-s} command line option lets you say
673 where to find these files. (@xref{Running}.)
674 The AT91SAM7X256 example above works this way.
676 Three main types of non-user configuration file each have their
677 own subdirectory in the @file{scripts} directory:
680 @item @b{interface} -- one for each kind of JTAG adapter/dongle
681 @item @b{board} -- one for each different board
682 @item @b{target} -- the chips which integrate CPUs and other JTAG TAPs
685 Best case: include just two files, and they handle everything else.
686 The first is an interface config file.
687 The second is board-specific, and it sets up the JTAG TAPs and
688 their GDB targets (by deferring to some @file{target.cfg} file),
689 declares all flash memory, and leaves you nothing to do except
693 source [find interface/olimex-jtag-tiny.cfg]
694 source [find board/csb337.cfg]
697 Boards with a single microcontroller often won't need more
698 than the target config file, as in the AT91SAM7X256 example.
699 That's because there is no external memory (flash, DDR RAM), and
700 the board differences are encapsulated by application code.
702 @item You can often reuse some standard config files but
703 need to write a few new ones, probably a @file{board.cfg} file.
704 You will be using commands described later in this User's Guide,
705 and working with the guidelines in the next chapter.
707 For example, there may be configuration files for your JTAG adapter
708 and target chip, but you need a new board-specific config file
709 giving access to your particular flash chips.
710 Or you might need to write another target chip configuration file
711 for a new chip built around the Cortex M3 core.
714 When you write new configuration files, please submit
715 them for inclusion in the next OpenOCD release.
716 For example, a @file{board/newboard.cfg} file will help the
717 next users of that board, and a @file{target/newcpu.cfg}
718 will help support users of any board using that chip.
722 You may may need to write some C code.
723 It may be as simple as a supporting a new ft2232 or parport
724 based dongle; a bit more involved, like a NAND or NOR flash
725 controller driver; or a big piece of work like supporting
726 a new chip architecture.
729 Reuse the existing config files when you can.
730 Look first in the @file{scripts/boards} area, then @file{scripts/targets}.
731 You may find a board configuration that's a good example to follow.
733 When you write config files, separate the reusable parts
734 (things every user of that interface, chip, or board needs)
735 from ones specific to your environment and debugging approach.
739 For example, a @code{gdb-attach} event handler that invokes
740 the @command{reset init} command will interfere with debugging
741 early boot code, which performs some of the same actions
742 that the @code{reset-init} event handler does.
745 Likewise, the @command{arm9tdmi vector_catch} command (or
747 its siblings @command{xscale vector_catch}
748 and @command{cortex_m3 vector_catch}) can be a timesaver
749 during some debug sessions, but don't make everyone use that either.
750 Keep those kinds of debugging aids in your user config file,
751 along with messaging and tracing setup.
752 (@xref{Software Debug Messages and Tracing}.)
755 You might need to override some defaults.
756 For example, you might need to move, shrink, or back up the target's
757 work area if your application needs much SRAM.
760 TCP/IP port configuration is another example of something which
761 is environment-specific, and should only appear in
762 a user config file. @xref{TCP/IP Ports}.
765 @section Project-Specific Utilities
767 A few project-specific utility
768 routines may well speed up your work.
769 Write them, and keep them in your project's user config file.
771 For example, if you are making a boot loader work on a
772 board, it's nice to be able to debug the ``after it's
773 loaded to RAM'' parts separately from the finicky early
774 code which sets up the DDR RAM controller and clocks.
775 A script like this one, or a more GDB-aware sibling,
779 proc ramboot @{ @} @{
780 # Reset, running the target's "reset-init" scripts
781 # to initialize clocks and the DDR RAM controller.
782 # Leave the CPU halted.
785 # Load CONFIG_SKIP_LOWLEVEL_INIT version into DDR RAM.
786 load_image u-boot.bin 0x20000000
793 Then once that code is working you will need to make it
794 boot from NOR flash; a different utility would help.
795 Alternatively, some developers write to flash using GDB.
796 (You might use a similar script if you're working with a flash
797 based microcontroller application instead of a boot loader.)
800 proc newboot @{ @} @{
801 # Reset, leaving the CPU halted. The "reset-init" event
802 # proc gives faster access to the CPU and to NOR flash;
803 # "reset halt" would be slower.
806 # Write standard version of U-Boot into the first two
807 # sectors of NOR flash ... the standard version should
808 # do the same lowlevel init as "reset-init".
809 flash protect 0 0 1 off
810 flash erase_sector 0 0 1
811 flash write_bank 0 u-boot.bin 0x0
812 flash protect 0 0 1 on
814 # Reboot from scratch using that new boot loader.
819 You may need more complicated utility procedures when booting
821 That often involves an extra bootloader stage,
822 running from on-chip SRAM to perform DDR RAM setup so it can load
823 the main bootloader code (which won't fit into that SRAM).
825 Other helper scripts might be used to write production system images,
826 involving considerably more than just a three stage bootloader.
829 @node Config File Guidelines
830 @chapter Config File Guidelines
832 This chapter is aimed at any user who needs to write a config file,
833 including developers and integrators of OpenOCD and any user who
834 needs to get a new board working smoothly.
835 It provides guidelines for creating those files.
837 You should find the following directories under @t{$(INSTALLDIR)/scripts}:
840 @item @file{interface} ...
841 think JTAG Dongle. Files that configure JTAG adapters go here.
842 @item @file{board} ...
843 think Circuit Board, PWA, PCB, they go by many names. Board files
844 contain initialization items that are specific to a board. For
845 example, the SDRAM initialization sequence for the board, or the type
846 of external flash and what address it uses. Any initialization
847 sequence to enable that external flash or SDRAM should be found in the
848 board file. Boards may also contain multiple targets: two CPUs; or
849 a CPU and an FPGA or CPLD.
850 @item @file{target} ...
851 think chip. The ``target'' directory represents the JTAG TAPs
853 which OpenOCD should control, not a board. Two common types of targets
854 are ARM chips and FPGA or CPLD chips.
855 When a chip has multiple TAPs (maybe it has both ARM and DSP cores),
856 the target config file defines all of them.
859 The @file{openocd.cfg} user config
860 file may override features in any of the above files by
861 setting variables before sourcing the target file, or by adding
862 commands specific to their situation.
864 @section Interface Config Files
867 should be able to source one of these files with a command like this:
870 source [find interface/FOOBAR.cfg]
873 A preconfigured interface file should exist for every interface in use
874 today, that said, perhaps some interfaces have only been used by the
875 sole developer who created it.
877 A separate chapter gives information about how to set these up.
878 @xref{Interface - Dongle Configuration}.
879 Read the OpenOCD source code if you have a new kind of hardware interface
880 and need to provide a driver for it.
882 @section Board Config Files
883 @cindex config file, board
884 @cindex board config file
887 should be able to source one of these files with a command like this:
890 source [find board/FOOBAR.cfg]
893 The point of a board config file is to package everything
894 about a given board that user config files need to know.
895 In summary the board files should contain (if present)
898 @item One or more @command{source [target/...cfg]} statements
899 @item NOR flash configuration (@pxref{NOR Configuration})
900 @item NAND flash configuration (@pxref{NAND Configuration})
901 @item Target @code{reset} handlers for SDRAM and I/O configuration
902 @item JTAG adapter reset configuration (@pxref{Reset Configuration})
903 @item All things that are not ``inside a chip''
906 Generic things inside target chips belong in target config files,
907 not board config files. So for example a @code{reset-init} event
908 handler should know board-specific oscillator and PLL parameters,
909 which it passes to target-specific utility code.
911 The most complex task of a board config file is creating such a
912 @code{reset-init} event handler.
913 Define those handlers last, after you verify the rest of the board
916 @subsection Communication Between Config files
918 In addition to target-specific utility code, another way that
919 board and target config files communicate is by following a
920 convention on how to use certain variables.
922 The full Tcl/Tk language supports ``namespaces'', but JIM-Tcl does not.
923 Thus the rule we follow in OpenOCD is this: Variables that begin with
924 a leading underscore are temporary in nature, and can be modified and
925 used at will within a target configuration file.
927 Complex board config files can do the things like this,
928 for a board with three chips:
931 # Chip #1: PXA270 for network side, big endian
934 source [find target/pxa270.cfg]
935 # on return: _TARGETNAME = network.cpu
936 # other commands can refer to the "network.cpu" target.
937 $_TARGETNAME configure .... events for this CPU..
939 # Chip #2: PXA270 for video side, little endian
942 source [find target/pxa270.cfg]
943 # on return: _TARGETNAME = video.cpu
944 # other commands can refer to the "video.cpu" target.
945 $_TARGETNAME configure .... events for this CPU..
947 # Chip #3: Xilinx FPGA for glue logic
950 source [find target/spartan3.cfg]
953 That example is oversimplified because it doesn't show any flash memory,
954 or the @code{reset-init} event handlers to initialize external DRAM
955 or (assuming it needs it) load a configuration into the FPGA.
956 Such features are usually needed for low-level work with many boards,
957 where ``low level'' implies that the board initialization software may
958 not be working. (That's a common reason to need JTAG tools. Another
959 is to enable working with microcontroller-based systems, which often
960 have no debugging support except a JTAG connector.)
962 Target config files may also export utility functions to board and user
963 config files. Such functions should use name prefixes, to help avoid
966 Board files could also accept input variables from user config files.
967 For example, there might be a @code{J4_JUMPER} setting used to identify
968 what kind of flash memory a development board is using, or how to set
969 up other clocks and peripherals.
971 @subsection Variable Naming Convention
972 @cindex variable names
974 Most boards have only one instance of a chip.
975 However, it should be easy to create a board with more than
976 one such chip (as shown above).
977 Accordingly, we encourage these conventions for naming
978 variables associated with different @file{target.cfg} files,
979 to promote consistency and
980 so that board files can override target defaults.
982 Inputs to target config files include:
985 @item @code{CHIPNAME} ...
986 This gives a name to the overall chip, and is used as part of
987 tap identifier dotted names.
988 While the default is normally provided by the chip manufacturer,
989 board files may need to distinguish between instances of a chip.
990 @item @code{ENDIAN} ...
991 By default @option{little} - although chips may hard-wire @option{big}.
992 Chips that can't change endianness don't need to use this variable.
993 @item @code{CPUTAPID} ...
994 When OpenOCD examines the JTAG chain, it can be told verify the
995 chips against the JTAG IDCODE register.
996 The target file will hold one or more defaults, but sometimes the
997 chip in a board will use a different ID (perhaps a newer revision).
1000 Outputs from target config files include:
1003 @item @code{_TARGETNAME} ...
1004 By convention, this variable is created by the target configuration
1005 script. The board configuration file may make use of this variable to
1006 configure things like a ``reset init'' script, or other things
1007 specific to that board and that target.
1008 If the chip has 2 targets, the names are @code{_TARGETNAME0},
1009 @code{_TARGETNAME1}, ... etc.
1012 @subsection The reset-init Event Handler
1013 @cindex event, reset-init
1014 @cindex reset-init handler
1016 Board config files run in the OpenOCD configuration stage;
1017 they can't use TAPs or targets, since they haven't been
1019 This means you can't write memory or access chip registers;
1020 you can't even verify that a flash chip is present.
1021 That's done later in event handlers, of which the target @code{reset-init}
1022 handler is one of the most important.
1024 Except on microcontrollers, the basic job of @code{reset-init} event
1025 handlers is setting up flash and DRAM, as normally handled by boot loaders.
1026 Microcontrollers rarely use boot loaders; they run right out of their
1027 on-chip flash and SRAM memory. But they may want to use one of these
1028 handlers too, if just for developer convenience.
1031 Because this is so very board-specific, and chip-specific, no examples
1033 Instead, look at the board config files distributed with OpenOCD.
1034 If you have a boot loader, its source code may also be useful.
1037 Some of this code could probably be shared between different boards.
1038 For example, setting up a DRAM controller often doesn't differ by
1039 much except the bus width (16 bits or 32?) and memory timings, so a
1040 reusable TCL procedure loaded by the @file{target.cfg} file might take
1041 those as parameters.
1042 Similarly with oscillator, PLL, and clock setup;
1043 and disabling the watchdog.
1044 Structure the code cleanly, and provide comments to help
1045 the next developer doing such work.
1046 (@emph{You might be that next person} trying to reuse init code!)
1048 The last thing normally done in a @code{reset-init} handler is probing
1049 whatever flash memory was configured. For most chips that needs to be
1050 done while the associated target is halted, either because JTAG memory
1051 access uses the CPU or to prevent conflicting CPU access.
1053 @subsection JTAG Clock Rate
1055 Before your @code{reset-init} handler has set up
1056 the PLLs and clocking, you may need to use
1057 a low JTAG clock rate; then you'd increase it later.
1058 For most ARM-based processors the fastest JTAG clock@footnote{A FAQ
1059 @uref{http://www.arm.com/support/faqdev/4170.html} gives details.}
1060 is one sixth of the CPU clock; or one eighth for ARM11 cores.
1061 Consult chip documentation to determine the peak JTAG clock rate,
1062 which might be less than that.
1065 On most ARMs, JTAG clock detection is coupled to the core clock, so
1066 software using a @option{wait for interrupt} operation blocks JTAG access.
1067 Adaptive clocking provides a partial workaround, but a more complete
1068 solution just avoids using that instruction with JTAG debuggers.
1071 If the board supports adaptive clocking, use the @command{jtag_rclk}
1072 command, in case your board is used with JTAG adapter which
1073 also supports it. Otherwise use @command{jtag_khz}.
1074 Set the slow rate at the beginning of the reset sequence,
1075 and the faster rate as soon as the clocks are at full speed.
1077 @section Target Config Files
1078 @cindex config file, target
1079 @cindex target config file
1081 Board config files communicate with target config files using
1082 naming conventions as described above, and may source one or
1083 more target config files like this:
1086 source [find target/FOOBAR.cfg]
1089 The point of a target config file is to package everything
1090 about a given chip that board config files need to know.
1091 In summary the target files should contain
1095 @item Add TAPs to the scan chain
1096 @item Add CPU targets (includes GDB support)
1097 @item CPU/Chip/CPU-Core specific features
1101 As a rule of thumb, a target file sets up only one chip.
1102 For a microcontroller, that will often include a single TAP,
1103 which is a CPU needing a GDB target, and its on-chip flash.
1105 More complex chips may include multiple TAPs, and the target
1106 config file may need to define them all before OpenOCD
1107 can talk to the chip.
1108 For example, some phone chips have JTAG scan chains that include
1109 an ARM core for operating system use, a DSP,
1110 another ARM core embedded in an image processing engine,
1111 and other processing engines.
1113 @subsection Default Value Boiler Plate Code
1115 All target configuration files should start with code like this,
1116 letting board config files express environment-specific
1117 differences in how things should be set up.
1120 # Boards may override chip names, perhaps based on role,
1121 # but the default should match what the vendor uses
1122 if @{ [info exists CHIPNAME] @} @{
1123 set _CHIPNAME $CHIPNAME
1125 set _CHIPNAME sam7x256
1128 # ONLY use ENDIAN with targets that can change it.
1129 if @{ [info exists ENDIAN] @} @{
1135 # TAP identifiers may change as chips mature, for example with
1136 # new revision fields (the "3" here). Pick a good default; you
1137 # can pass several such identifiers to the "jtag newtap" command.
1138 if @{ [info exists CPUTAPID ] @} @{
1139 set _CPUTAPID $CPUTAPID
1141 set _CPUTAPID 0x3f0f0f0f
1144 @c but 0x3f0f0f0f is for an str73x part ...
1146 @emph{Remember:} Board config files may include multiple target
1147 config files, or the same target file multiple times
1148 (changing at least @code{CHIPNAME}).
1150 Likewise, the target configuration file should define
1151 @code{_TARGETNAME} (or @code{_TARGETNAME0} etc) and
1152 use it later on when defining debug targets:
1155 set _TARGETNAME $_CHIPNAME.cpu
1156 target create $_TARGETNAME arm7tdmi -chain-position $_TARGETNAME
1159 @subsection Adding TAPs to the Scan Chain
1160 After the ``defaults'' are set up,
1161 add the TAPs on each chip to the JTAG scan chain.
1162 @xref{TAP Declaration}, and the naming convention
1165 In the simplest case the chip has only one TAP,
1166 probably for a CPU or FPGA.
1167 The config file for the Atmel AT91SAM7X256
1168 looks (in part) like this:
1171 jtag newtap $_CHIPNAME cpu -irlen 4 -ircapture 0x1 -irmask 0xf \
1172 -expected-id $_CPUTAPID
1175 A board with two such at91sam7 chips would be able
1176 to source such a config file twice, with different
1177 values for @code{CHIPNAME}, so
1178 it adds a different TAP each time.
1180 If there are one or more nonzero @option{-expected-id} values,
1181 OpenOCD attempts to verify the actual tap id against those values.
1182 It will issue error messages if there is mismatch, which
1183 can help to pinpoint problems in OpenOCD configurations.
1186 JTAG tap: sam7x256.cpu tap/device found: 0x3f0f0f0f
1187 (Manufacturer: 0x787, Part: 0xf0f0, Version: 0x3)
1188 ERROR: Tap: sam7x256.cpu - Expected id: 0x12345678, Got: 0x3f0f0f0f
1189 ERROR: expected: mfg: 0x33c, part: 0x2345, ver: 0x1
1190 ERROR: got: mfg: 0x787, part: 0xf0f0, ver: 0x3
1193 There are more complex examples too, with chips that have
1194 multiple TAPs. Ones worth looking at include:
1197 @item @file{target/omap3530.cfg} -- with disabled ARM and DSP,
1198 plus a JRC to enable them
1199 @item @file{target/str912.cfg} -- with flash, CPU, and boundary scan
1200 @item @file{target/ti_dm355.cfg} -- with ETM, ARM, and JRC (this JRC
1201 is not currently used)
1204 @subsection Add CPU targets
1206 After adding a TAP for a CPU, you should set it up so that
1207 GDB and other commands can use it.
1208 @xref{CPU Configuration}.
1209 For the at91sam7 example above, the command can look like this;
1210 note that @code{$_ENDIAN} is not needed, since OpenOCD defaults
1211 to little endian, and this chip doesn't support changing that.
1214 set _TARGETNAME $_CHIPNAME.cpu
1215 target create $_TARGETNAME arm7tdmi -chain-position $_TARGETNAME
1218 Work areas are small RAM areas associated with CPU targets.
1219 They are used by OpenOCD to speed up downloads,
1220 and to download small snippets of code to program flash chips.
1221 If the chip includes a form of ``on-chip-ram'' - and many do - define
1222 a work area if you can.
1223 Again using the at91sam7 as an example, this can look like:
1226 $_TARGETNAME configure -work-area-phys 0x00200000 \
1227 -work-area-size 0x4000 -work-area-backup 0
1230 @subsection Chip Reset Setup
1232 As a rule, you should put the @command{reset_config} command
1233 into the board file. Most things you think you know about a
1234 chip can be tweaked by the board.
1236 Some chips have specific ways the TRST and SRST signals are
1237 managed. In the unusual case that these are @emph{chip specific}
1238 and can never be changed by board wiring, they could go here.
1240 Some chips need special attention during reset handling if
1241 they're going to be used with JTAG.
1242 An example might be needing to send some commands right
1243 after the target's TAP has been reset, providing a
1244 @code{reset-deassert-post} event handler that writes a chip
1245 register to report that JTAG debugging is being done.
1247 @subsection ARM Core Specific Hacks
1249 If the chip has a DCC, enable it. If the chip is an ARM9 with some
1250 special high speed download features - enable it.
1252 If present, the MMU, the MPU and the CACHE should be disabled.
1254 Some ARM cores are equipped with trace support, which permits
1255 examination of the instruction and data bus activity. Trace
1256 activity is controlled through an ``Embedded Trace Module'' (ETM)
1257 on one of the core's scan chains. The ETM emits voluminous data
1258 through a ``trace port''. (@xref{ARM Hardware Tracing}.)
1259 If you are using an external trace port,
1260 configure it in your board config file.
1261 If you are using an on-chip ``Embedded Trace Buffer'' (ETB),
1262 configure it in your target config file.
1265 etm config $_TARGETNAME 16 normal full etb
1266 etb config $_TARGETNAME $_CHIPNAME.etb
1269 @subsection Internal Flash Configuration
1271 This applies @b{ONLY TO MICROCONTROLLERS} that have flash built in.
1273 @b{Never ever} in the ``target configuration file'' define any type of
1274 flash that is external to the chip. (For example a BOOT flash on
1275 Chip Select 0.) Such flash information goes in a board file - not
1276 the TARGET (chip) file.
1280 @item at91sam7x256 - has 256K flash YES enable it.
1281 @item str912 - has flash internal YES enable it.
1282 @item imx27 - uses boot flash on CS0 - it goes in the board file.
1283 @item pxa270 - again - CS0 flash - it goes in the board file.
1286 @node Daemon Configuration
1287 @chapter Daemon Configuration
1288 @cindex initialization
1289 The commands here are commonly found in the openocd.cfg file and are
1290 used to specify what TCP/IP ports are used, and how GDB should be
1293 @section Configuration Stage
1294 @cindex configuration stage
1295 @cindex config command
1297 When the OpenOCD server process starts up, it enters a
1298 @emph{configuration stage} which is the only time that
1299 certain commands, @emph{configuration commands}, may be issued.
1300 In this manual, the definition of a configuration command is
1301 presented as a @emph{Config Command}, not as a @emph{Command}
1302 which may be issued interactively.
1304 Those configuration commands include declaration of TAPs,
1306 the interface used for JTAG communication,
1307 and other basic setup.
1308 The server must leave the configuration stage before it
1309 may access or activate TAPs.
1310 After it leaves this stage, configuration commands may no
1313 @deffn {Config Command} init
1314 This command terminates the configuration stage and
1315 enters the normal command mode. This can be useful to add commands to
1316 the startup scripts and commands such as resetting the target,
1317 programming flash, etc. To reset the CPU upon startup, add "init" and
1318 "reset" at the end of the config script or at the end of the OpenOCD
1319 command line using the @option{-c} command line switch.
1321 If this command does not appear in any startup/configuration file
1322 OpenOCD executes the command for you after processing all
1323 configuration files and/or command line options.
1325 @b{NOTE:} This command normally occurs at or near the end of your
1326 openocd.cfg file to force OpenOCD to ``initialize'' and make the
1327 targets ready. For example: If your openocd.cfg file needs to
1328 read/write memory on your target, @command{init} must occur before
1329 the memory read/write commands. This includes @command{nand probe}.
1332 @anchor{TCP/IP Ports}
1333 @section TCP/IP Ports
1338 The OpenOCD server accepts remote commands in several syntaxes.
1339 Each syntax uses a different TCP/IP port, which you may specify
1340 only during configuration (before those ports are opened).
1342 For reasons including security, you may wish to prevent remote
1343 access using one or more of these ports.
1344 In such cases, just specify the relevant port number as zero.
1345 If you disable all access through TCP/IP, you will need to
1346 use the command line @option{-pipe} option.
1348 @deffn {Command} gdb_port (number)
1350 Specify or query the first port used for incoming GDB connections.
1351 The GDB port for the
1352 first target will be gdb_port, the second target will listen on gdb_port + 1, and so on.
1353 When not specified during the configuration stage,
1354 the port @var{number} defaults to 3333.
1355 When specified as zero, this port is not activated.
1358 @deffn {Command} tcl_port (number)
1359 Specify or query the port used for a simplified RPC
1360 connection that can be used by clients to issue TCL commands and get the
1361 output from the Tcl engine.
1362 Intended as a machine interface.
1363 When not specified during the configuration stage,
1364 the port @var{number} defaults to 6666.
1365 When specified as zero, this port is not activated.
1368 @deffn {Command} telnet_port (number)
1369 Specify or query the
1370 port on which to listen for incoming telnet connections.
1371 This port is intended for interaction with one human through TCL commands.
1372 When not specified during the configuration stage,
1373 the port @var{number} defaults to 4444.
1374 When specified as zero, this port is not activated.
1377 @anchor{GDB Configuration}
1378 @section GDB Configuration
1380 @cindex GDB configuration
1381 You can reconfigure some GDB behaviors if needed.
1382 The ones listed here are static and global.
1383 @xref{Target Configuration}, about configuring individual targets.
1384 @xref{Target Events}, about configuring target-specific event handling.
1386 @anchor{gdb_breakpoint_override}
1387 @deffn {Command} gdb_breakpoint_override [@option{hard}|@option{soft}|@option{disable}]
1388 Force breakpoint type for gdb @command{break} commands.
1389 This option supports GDB GUIs which don't
1390 distinguish hard versus soft breakpoints, if the default OpenOCD and
1391 GDB behaviour is not sufficient. GDB normally uses hardware
1392 breakpoints if the memory map has been set up for flash regions.
1395 @deffn {Config Command} gdb_detach (@option{resume}|@option{reset}|@option{halt}|@option{nothing})
1396 Configures what OpenOCD will do when GDB detaches from the daemon.
1397 Default behaviour is @option{resume}.
1400 @anchor{gdb_flash_program}
1401 @deffn {Config Command} gdb_flash_program (@option{enable}|@option{disable})
1402 Set to @option{enable} to cause OpenOCD to program the flash memory when a
1403 vFlash packet is received.
1404 The default behaviour is @option{enable}.
1407 @deffn {Config Command} gdb_memory_map (@option{enable}|@option{disable})
1408 Set to @option{enable} to cause OpenOCD to send the memory configuration to GDB when
1409 requested. GDB will then know when to set hardware breakpoints, and program flash
1410 using the GDB load command. @command{gdb_flash_program enable} must also be enabled
1411 for flash programming to work.
1412 Default behaviour is @option{enable}.
1413 @xref{gdb_flash_program}.
1416 @deffn {Config Command} gdb_report_data_abort (@option{enable}|@option{disable})
1417 Specifies whether data aborts cause an error to be reported
1418 by GDB memory read packets.
1419 The default behaviour is @option{disable};
1420 use @option{enable} see these errors reported.
1423 @anchor{Event Polling}
1424 @section Event Polling
1426 Hardware debuggers are parts of asynchronous systems,
1427 where significant events can happen at any time.
1428 The OpenOCD server needs to detect some of these events,
1429 so it can report them to through TCL command line
1432 Examples of such events include:
1435 @item One of the targets can stop running ... maybe it triggers
1436 a code breakpoint or data watchpoint, or halts itself.
1437 @item Messages may be sent over ``debug message'' channels ... many
1438 targets support such messages sent over JTAG,
1439 for receipt by the person debugging or tools.
1440 @item Loss of power ... some adapters can detect these events.
1441 @item Resets not issued through JTAG ... such reset sources
1442 can include button presses or other system hardware, sometimes
1443 including the target itself (perhaps through a watchdog).
1444 @item Debug instrumentation sometimes supports event triggering
1445 such as ``trace buffer full'' (so it can quickly be emptied)
1446 or other signals (to correlate with code behavior).
1449 None of those events are signaled through standard JTAG signals.
1450 However, most conventions for JTAG connectors include voltage
1451 level and system reset (SRST) signal detection.
1452 Some connectors also include instrumentation signals, which
1453 can imply events when those signals are inputs.
1455 In general, OpenOCD needs to periodically check for those events,
1456 either by looking at the status of signals on the JTAG connector
1457 or by sending synchronous ``tell me your status'' JTAG requests
1458 to the various active targets.
1459 There is a command to manage and monitor that polling,
1460 which is normally done in the background.
1462 @deffn Command poll [@option{on}|@option{off}]
1463 Poll the current target for its current state.
1464 (Also, @pxref{target curstate}.)
1465 If that target is in debug mode, architecture
1466 specific information about the current state is printed.
1467 An optional parameter
1468 allows background polling to be enabled and disabled.
1470 You could use this from the TCL command shell, or
1471 from GDB using @command{monitor poll} command.
1474 background polling: on
1475 target state: halted
1476 target halted in ARM state due to debug-request, \
1477 current mode: Supervisor
1478 cpsr: 0x800000d3 pc: 0x11081bfc
1479 MMU: disabled, D-Cache: disabled, I-Cache: enabled
1484 @node Interface - Dongle Configuration
1485 @chapter Interface - Dongle Configuration
1486 @cindex config file, interface
1487 @cindex interface config file
1489 JTAG Adapters/Interfaces/Dongles are normally configured
1490 through commands in an interface configuration
1491 file which is sourced by your @file{openocd.cfg} file, or
1492 through a command line @option{-f interface/....cfg} option.
1495 source [find interface/olimex-jtag-tiny.cfg]
1499 OpenOCD what type of JTAG adapter you have, and how to talk to it.
1500 A few cases are so simple that you only need to say what driver to use:
1507 Most adapters need a bit more configuration than that.
1510 @section Interface Configuration
1512 The interface command tells OpenOCD what type of JTAG dongle you are
1513 using. Depending on the type of dongle, you may need to have one or
1514 more additional commands.
1516 @deffn {Config Command} {interface} name
1517 Use the interface driver @var{name} to connect to the
1521 @deffn Command {interface_list}
1522 List the interface drivers that have been built into
1523 the running copy of OpenOCD.
1526 @deffn Command {jtag interface}
1527 Returns the name of the interface driver being used.
1530 @section Interface Drivers
1532 Each of the interface drivers listed here must be explicitly
1533 enabled when OpenOCD is configured, in order to be made
1534 available at run time.
1536 @deffn {Interface Driver} {amt_jtagaccel}
1537 Amontec Chameleon in its JTAG Accelerator configuration,
1538 connected to a PC's EPP mode parallel port.
1539 This defines some driver-specific commands:
1541 @deffn {Config Command} {parport_port} number
1542 Specifies either the address of the I/O port (default: 0x378 for LPT1) or
1543 the number of the @file{/dev/parport} device.
1546 @deffn {Config Command} rtck [@option{enable}|@option{disable}]
1547 Displays status of RTCK option.
1548 Optionally sets that option first.
1552 @deffn {Interface Driver} {arm-jtag-ew}
1553 Olimex ARM-JTAG-EW USB adapter
1554 This has one driver-specific command:
1556 @deffn Command {armjtagew_info}
1561 @deffn {Interface Driver} {at91rm9200}
1562 Supports bitbanged JTAG from the local system,
1563 presuming that system is an Atmel AT91rm9200
1564 and a specific set of GPIOs is used.
1565 @c command: at91rm9200_device NAME
1566 @c chooses among list of bit configs ... only one option
1569 @deffn {Interface Driver} {dummy}
1570 A dummy software-only driver for debugging.
1573 @deffn {Interface Driver} {ep93xx}
1574 Cirrus Logic EP93xx based single-board computer bit-banging (in development)
1577 @deffn {Interface Driver} {ft2232}
1578 FTDI FT2232 (USB) based devices over one of the userspace libraries.
1579 These interfaces have several commands, used to configure the driver
1580 before initializing the JTAG scan chain:
1582 @deffn {Config Command} {ft2232_device_desc} description
1583 Provides the USB device description (the @emph{iProduct string})
1584 of the FTDI FT2232 device. If not
1585 specified, the FTDI default value is used. This setting is only valid
1586 if compiled with FTD2XX support.
1589 @deffn {Config Command} {ft2232_serial} serial-number
1590 Specifies the @var{serial-number} of the FTDI FT2232 device to use,
1591 in case the vendor provides unique IDs and more than one FT2232 device
1592 is connected to the host.
1593 If not specified, serial numbers are not considered.
1594 (Note that USB serial numbers can be arbitrary Unicode strings,
1595 and are not restricted to containing only decimal digits.)
1598 @deffn {Config Command} {ft2232_layout} name
1599 Each vendor's FT2232 device can use different GPIO signals
1600 to control output-enables, reset signals, and LEDs.
1601 Currently valid layout @var{name} values include:
1603 @item @b{axm0432_jtag} Axiom AXM-0432
1604 @item @b{comstick} Hitex STR9 comstick
1605 @item @b{cortino} Hitex Cortino JTAG interface
1606 @item @b{evb_lm3s811} Luminary Micro EVB_LM3S811 as a JTAG interface,
1607 either for the local Cortex-M3 (SRST only)
1608 or in a passthrough mode (neither SRST nor TRST)
1609 @item @b{luminary_icdi} Luminary In-Circuit Debug Interface (ICDI) Board
1610 @item @b{flyswatter} Tin Can Tools Flyswatter
1611 @item @b{icebear} ICEbear JTAG adapter from Section 5
1612 @item @b{jtagkey} Amontec JTAGkey and JTAGkey-Tiny (and compatibles)
1613 @item @b{jtagkey2} Amontec JTAGkey2 (and compatibles)
1614 @item @b{m5960} American Microsystems M5960
1615 @item @b{olimex-jtag} Olimex ARM-USB-OCD and ARM-USB-Tiny
1616 @item @b{oocdlink} OOCDLink
1617 @c oocdlink ~= jtagkey_prototype_v1
1618 @item @b{sheevaplug} Marvell Sheevaplug development kit
1619 @item @b{signalyzer} Xverve Signalyzer
1620 @item @b{stm32stick} Hitex STM32 Performance Stick
1621 @item @b{turtelizer2} egnite Software turtelizer2
1622 @item @b{usbjtag} "USBJTAG-1" layout described in the OpenOCD diploma thesis
1626 @deffn {Config Command} {ft2232_vid_pid} [vid pid]+
1627 The vendor ID and product ID of the FTDI FT2232 device. If not specified, the FTDI
1628 default values are used.
1629 Currently, up to eight [@var{vid}, @var{pid}] pairs may be given, e.g.
1631 ft2232_vid_pid 0x0403 0xcff8 0x15ba 0x0003
1635 @deffn {Config Command} {ft2232_latency} ms
1636 On some systems using FT2232 based JTAG interfaces the FT_Read function call in
1637 ft2232_read() fails to return the expected number of bytes. This can be caused by
1638 USB communication delays and has proved hard to reproduce and debug. Setting the
1639 FT2232 latency timer to a larger value increases delays for short USB packets but it
1640 also reduces the risk of timeouts before receiving the expected number of bytes.
1641 The OpenOCD default value is 2 and for some systems a value of 10 has proved useful.
1644 For example, the interface config file for a
1645 Turtelizer JTAG Adapter looks something like this:
1649 ft2232_device_desc "Turtelizer JTAG/RS232 Adapter"
1650 ft2232_layout turtelizer2
1651 ft2232_vid_pid 0x0403 0xbdc8
1655 @deffn {Interface Driver} {gw16012}
1656 Gateworks GW16012 JTAG programmer.
1657 This has one driver-specific command:
1659 @deffn {Config Command} {parport_port} number
1660 Specifies either the address of the I/O port (default: 0x378 for LPT1) or
1661 the number of the @file{/dev/parport} device.
1665 @deffn {Interface Driver} {jlink}
1666 Segger jlink USB adapter
1667 @c command: jlink_info
1669 @c command: jlink_hw_jtag (2|3)
1670 @c sets version 2 or 3
1673 @deffn {Interface Driver} {parport}
1674 Supports PC parallel port bit-banging cables:
1675 Wigglers, PLD download cable, and more.
1676 These interfaces have several commands, used to configure the driver
1677 before initializing the JTAG scan chain:
1679 @deffn {Config Command} {parport_cable} name
1680 The layout of the parallel port cable used to connect to the target.
1681 Currently valid cable @var{name} values include:
1684 @item @b{altium} Altium Universal JTAG cable.
1685 @item @b{arm-jtag} Same as original wiggler except SRST and
1686 TRST connections reversed and TRST is also inverted.
1687 @item @b{chameleon} The Amontec Chameleon's CPLD when operated
1688 in configuration mode. This is only used to
1689 program the Chameleon itself, not a connected target.
1690 @item @b{dlc5} The Xilinx Parallel cable III.
1691 @item @b{flashlink} The ST Parallel cable.
1692 @item @b{lattice} Lattice ispDOWNLOAD Cable
1693 @item @b{old_amt_wiggler} The Wiggler configuration that comes with
1695 Amontec's Chameleon Programmer. The new version available from
1696 the website uses the original Wiggler layout ('@var{wiggler}')
1697 @item @b{triton} The parallel port adapter found on the
1698 ``Karo Triton 1 Development Board''.
1699 This is also the layout used by the HollyGates design
1700 (see @uref{http://www.lartmaker.nl/projects/jtag/}).
1701 @item @b{wiggler} The original Wiggler layout, also supported by
1702 several clones, such as the Olimex ARM-JTAG
1703 @item @b{wiggler2} Same as original wiggler except an led is fitted on D5.
1704 @item @b{wiggler_ntrst_inverted} Same as original wiggler except TRST is inverted.
1708 @deffn {Config Command} {parport_port} number
1709 Either the address of the I/O port (default: 0x378 for LPT1) or the number of
1710 the @file{/dev/parport} device
1712 When using PPDEV to access the parallel port, use the number of the parallel port:
1713 @option{parport_port 0} (the default). If @option{parport_port 0x378} is specified
1714 you may encounter a problem.
1717 @deffn {Config Command} {parport_write_on_exit} (on|off)
1718 This will configure the parallel driver to write a known
1719 cable-specific value to the parallel interface on exiting OpenOCD
1722 For example, the interface configuration file for a
1723 classic ``Wiggler'' cable might look something like this:
1728 parport_cable wiggler
1732 @deffn {Interface Driver} {presto}
1733 ASIX PRESTO USB JTAG programmer.
1734 @c command: presto_serial str
1735 @c sets serial number
1738 @deffn {Interface Driver} {rlink}
1739 Raisonance RLink USB adapter
1742 @deffn {Interface Driver} {usbprog}
1743 usbprog is a freely programmable USB adapter.
1746 @deffn {Interface Driver} {vsllink}
1747 vsllink is part of Versaloon which is a versatile USB programmer.
1750 This defines quite a few driver-specific commands,
1751 which are not currently documented here.
1755 @deffn {Interface Driver} {ZY1000}
1756 This is the Zylin ZY1000 JTAG debugger.
1759 This defines some driver-specific commands,
1760 which are not currently documented here.
1763 @deffn Command power [@option{on}|@option{off}]
1764 Turn power switch to target on/off.
1765 No arguments: print status.
1772 JTAG clock setup is part of system setup.
1773 It @emph{does not belong with interface setup} since any interface
1774 only knows a few of the constraints for the JTAG clock speed.
1775 Sometimes the JTAG speed is
1776 changed during the target initialization process: (1) slow at
1777 reset, (2) program the CPU clocks, (3) run fast.
1778 Both the "slow" and "fast" clock rates are functions of the
1779 oscillators used, the chip, the board design, and sometimes
1780 power management software that may be active.
1782 The speed used during reset can be adjusted using pre_reset
1783 and post_reset event handlers.
1784 @xref{Target Events}.
1786 If your system supports adaptive clocking (RTCK), configuring
1787 JTAG to use that is probably the most robust approach.
1788 However, it introduces delays to synchronize clocks; so it
1789 may not be the fastest solution.
1791 @b{NOTE:} Script writers should consider using @command{jtag_rclk}
1792 instead of @command{jtag_khz}.
1794 @deffn {Command} jtag_khz max_speed_kHz
1795 A non-zero speed is in KHZ. Hence: 3000 is 3mhz.
1796 JTAG interfaces usually support a limited number of
1797 speeds. The speed actually used won't be faster
1798 than the speed specified.
1800 Chip data sheets generally include a top JTAG clock rate.
1801 The actual rate is often a function of a CPU core clock,
1802 and is normally less than that peak rate.
1803 For example, most ARM cores accept at most one sixth of the CPU clock.
1805 Speed 0 (khz) selects RTCK method.
1807 If your system uses RTCK, you won't need to change the
1808 JTAG clocking after setup.
1809 Not all interfaces, boards, or targets support ``rtck''.
1810 If the interface device can not
1811 support it, an error is returned when you try to use RTCK.
1814 @defun jtag_rclk fallback_speed_kHz
1815 @cindex adaptive clocking
1817 This Tcl proc (defined in @file{startup.tcl}) attempts to enable RTCK/RCLK.
1818 If that fails (maybe the interface, board, or target doesn't
1819 support it), falls back to the specified frequency.
1821 # Fall back to 3mhz if RTCK is not supported
1826 @node Reset Configuration
1827 @chapter Reset Configuration
1828 @cindex Reset Configuration
1830 Every system configuration may require a different reset
1831 configuration. This can also be quite confusing.
1832 Resets also interact with @var{reset-init} event handlers,
1833 which do things like setting up clocks and DRAM, and
1834 JTAG clock rates. (@xref{JTAG Speed}.)
1835 They can also interact with JTAG routers.
1836 Please see the various board files for examples.
1839 To maintainers and integrators:
1840 Reset configuration touches several things at once.
1841 Normally the board configuration file
1842 should define it and assume that the JTAG adapter supports
1843 everything that's wired up to the board's JTAG connector.
1845 However, the target configuration file could also make note
1846 of something the silicon vendor has done inside the chip,
1847 which will be true for most (or all) boards using that chip.
1848 And when the JTAG adapter doesn't support everything, the
1849 user configuration file will need to override parts of
1850 the reset configuration provided by other files.
1853 @section Types of Reset
1855 There are many kinds of reset possible through JTAG, but
1856 they may not all work with a given board and adapter.
1857 That's part of why reset configuration can be error prone.
1861 @emph{System Reset} ... the @emph{SRST} hardware signal
1862 resets all chips connected to the JTAG adapter, such as processors,
1863 power management chips, and I/O controllers. Normally resets triggered
1864 with this signal behave exactly like pressing a RESET button.
1866 @emph{JTAG TAP Reset} ... the @emph{TRST} hardware signal resets
1867 just the TAP controllers connected to the JTAG adapter.
1868 Such resets should not be visible to the rest of the system; resetting a
1869 device's the TAP controller just puts that controller into a known state.
1871 @emph{Emulation Reset} ... many devices can be reset through JTAG
1872 commands. These resets are often distinguishable from system
1873 resets, either explicitly (a "reset reason" register says so)
1874 or implicitly (not all parts of the chip get reset).
1876 @emph{Other Resets} ... system-on-chip devices often support
1877 several other types of reset.
1878 You may need to arrange that a watchdog timer stops
1879 while debugging, preventing a watchdog reset.
1880 There may be individual module resets.
1883 In the best case, OpenOCD can hold SRST, then reset
1884 the TAPs via TRST and send commands through JTAG to halt the
1885 CPU at the reset vector before the 1st instruction is executed.
1886 Then when it finally releases the SRST signal, the system is
1887 halted under debugger control before any code has executed.
1888 This is the behavior required to support the @command{reset halt}
1889 and @command{reset init} commands; after @command{reset init} a
1890 board-specific script might do things like setting up DRAM.
1891 (@xref{Reset Command}.)
1893 @anchor{SRST and TRST Issues}
1894 @section SRST and TRST Issues
1896 Because SRST and TRST are hardware signals, they can have a
1897 variety of system-specific constraints. Some of the most
1902 @item @emph{Signal not available} ... Some boards don't wire
1903 SRST or TRST to the JTAG connector. Some JTAG adapters don't
1904 support such signals even if they are wired up.
1905 Use the @command{reset_config} @var{signals} options to say
1906 when either of those signals is not connected.
1907 When SRST is not available, your code might not be able to rely
1908 on controllers having been fully reset during code startup.
1909 Missing TRST is not a problem, since JTAG level resets can
1910 be triggered using with TMS signaling.
1912 @item @emph{Signals shorted} ... Sometimes a chip, board, or
1913 adapter will connect SRST to TRST, instead of keeping them separate.
1914 Use the @command{reset_config} @var{combination} options to say
1915 when those signals aren't properly independent.
1917 @item @emph{Timing} ... Reset circuitry like a resistor/capacitor
1918 delay circuit, reset supervisor, or on-chip features can extend
1919 the effect of a JTAG adapter's reset for some time after the adapter
1920 stops issuing the reset. For example, there may be chip or board
1921 requirements that all reset pulses last for at least a
1922 certain amount of time; and reset buttons commonly have
1923 hardware debouncing.
1924 Use the @command{jtag_nsrst_delay} and @command{jtag_ntrst_delay}
1925 commands to say when extra delays are needed.
1927 @item @emph{Drive type} ... Reset lines often have a pullup
1928 resistor, letting the JTAG interface treat them as open-drain
1929 signals. But that's not a requirement, so the adapter may need
1930 to use push/pull output drivers.
1931 Also, with weak pullups it may be advisable to drive
1932 signals to both levels (push/pull) to minimize rise times.
1933 Use the @command{reset_config} @var{trst_type} and
1934 @var{srst_type} parameters to say how to drive reset signals.
1936 @item @emph{Special initialization} ... Targets sometimes need
1937 special JTAG initialization sequences to handle chip-specific
1938 issues (not limited to errata).
1939 For example, certain JTAG commands might need to be issued while
1940 the system as a whole is in a reset state (SRST active)
1941 but the JTAG scan chain is usable (TRST inactive).
1942 (@xref{JTAG Commands}, where the @command{jtag_reset}
1943 command is presented.)
1946 There can also be other issues.
1947 Some devices don't fully conform to the JTAG specifications.
1948 Trivial system-specific differences are common, such as
1949 SRST and TRST using slightly different names.
1950 There are also vendors who distribute key JTAG documentation for
1951 their chips only to developers who have signed a Non-Disclosure
1954 Sometimes there are chip-specific extensions like a requirement to use
1955 the normally-optional TRST signal (precluding use of JTAG adapters which
1956 don't pass TRST through), or needing extra steps to complete a TAP reset.
1958 In short, SRST and especially TRST handling may be very finicky,
1959 needing to cope with both architecture and board specific constraints.
1961 @section Commands for Handling Resets
1963 @deffn {Command} jtag_nsrst_delay milliseconds
1964 How long (in milliseconds) OpenOCD should wait after deasserting
1965 nSRST (active-low system reset) before starting new JTAG operations.
1966 When a board has a reset button connected to SRST line it will
1967 probably have hardware debouncing, implying you should use this.
1970 @deffn {Command} jtag_ntrst_delay milliseconds
1971 How long (in milliseconds) OpenOCD should wait after deasserting
1972 nTRST (active-low JTAG TAP reset) before starting new JTAG operations.
1975 @deffn {Command} reset_config mode_flag ...
1976 This command tells OpenOCD the reset configuration
1977 of your combination of JTAG board and target in target
1978 configuration scripts.
1980 Information earlier in this section describes the kind of problems
1981 the command is intended to address (@pxref{SRST and TRST Issues}).
1982 As a rule this command belongs only in board config files,
1983 describing issues like @emph{board doesn't connect TRST};
1984 or in user config files, addressing limitations derived
1985 from a particular combination of interface and board.
1986 (An unlikely example would be using a TRST-only adapter
1987 with a board that only wires up SRST.)
1989 The @var{mode_flag} options can be specified in any order, but only one
1990 of each type -- @var{signals}, @var{combination}, @var{trst_type},
1991 and @var{srst_type} -- may be specified at a time.
1992 If you don't provide a new value for a given type, its previous
1993 value (perhaps the default) is unchanged.
1994 For example, this means that you don't need to say anything at all about
1995 TRST just to declare that if the JTAG adapter should want to drive SRST,
1996 it must explicitly be driven high (@option{srst_push_pull}).
1998 @var{signals} can specify which of the reset signals are connected.
1999 For example, If the JTAG interface provides SRST, but the board doesn't
2000 connect that signal properly, then OpenOCD can't use it.
2001 Possible values are @option{none} (the default), @option{trst_only},
2002 @option{srst_only} and @option{trst_and_srst}.
2005 If your board provides SRST or TRST through the JTAG connector,
2006 you must declare that or else those signals will not be used.
2009 The @var{combination} is an optional value specifying broken reset
2010 signal implementations.
2011 The default behaviour if no option given is @option{separate},
2012 indicating everything behaves normally.
2013 @option{srst_pulls_trst} states that the
2014 test logic is reset together with the reset of the system (e.g. Philips
2015 LPC2000, "broken" board layout), @option{trst_pulls_srst} says that
2016 the system is reset together with the test logic (only hypothetical, I
2017 haven't seen hardware with such a bug, and can be worked around).
2018 @option{combined} implies both @option{srst_pulls_trst} and
2019 @option{trst_pulls_srst}.
2021 The optional @var{trst_type} and @var{srst_type} parameters allow the
2022 driver mode of each reset line to be specified. These values only affect
2023 JTAG interfaces with support for different driver modes, like the Amontec
2024 JTAGkey and JTAGAccelerator. Also, they are necessarily ignored if the
2025 relevant signal (TRST or SRST) is not connected.
2027 Possible @var{trst_type} driver modes for the test reset signal (TRST)
2028 are @option{trst_push_pull} (default) and @option{trst_open_drain}.
2029 Most boards connect this signal to a pulldown, so the JTAG TAPs
2030 never leave reset unless they are hooked up to a JTAG adapter.
2032 Possible @var{srst_type} driver modes for the system reset signal (SRST)
2033 are the default @option{srst_open_drain}, and @option{srst_push_pull}.
2034 Most boards connect this signal to a pullup, and allow the
2035 signal to be pulled low by various events including system
2036 powerup and pressing a reset button.
2040 @node TAP Declaration
2041 @chapter TAP Declaration
2042 @cindex TAP declaration
2043 @cindex TAP configuration
2045 @emph{Test Access Ports} (TAPs) are the core of JTAG.
2046 TAPs serve many roles, including:
2049 @item @b{Debug Target} A CPU TAP can be used as a GDB debug target
2050 @item @b{Flash Programing} Some chips program the flash directly via JTAG.
2051 Others do it indirectly, making a CPU do it.
2052 @item @b{Program Download} Using the same CPU support GDB uses,
2053 you can initialize a DRAM controller, download code to DRAM, and then
2054 start running that code.
2055 @item @b{Boundary Scan} Most chips support boundary scan, which
2056 helps test for board assembly problems like solder bridges
2057 and missing connections
2060 OpenOCD must know about the active TAPs on your board(s).
2061 Setting up the TAPs is the core task of your configuration files.
2062 Once those TAPs are set up, you can pass their names to code
2063 which sets up CPUs and exports them as GDB targets,
2064 probes flash memory, performs low-level JTAG operations, and more.
2066 @section Scan Chains
2069 TAPs are part of a hardware @dfn{scan chain},
2070 which is daisy chain of TAPs.
2071 They also need to be added to
2072 OpenOCD's software mirror of that hardware list,
2073 giving each member a name and associating other data with it.
2074 Simple scan chains, with a single TAP, are common in
2075 systems with a single microcontroller or microprocessor.
2076 More complex chips may have several TAPs internally.
2077 Very complex scan chains might have a dozen or more TAPs:
2078 several in one chip, more in the next, and connecting
2079 to other boards with their own chips and TAPs.
2081 You can display the list with the @command{scan_chain} command.
2082 (Don't confuse this with the list displayed by the @command{targets}
2083 command, presented in the next chapter.
2084 That only displays TAPs for CPUs which are configured as
2086 Here's what the scan chain might look like for a chip more than one TAP:
2089 TapName Enabled IdCode Expected IrLen IrCap IrMask Instr
2090 -- ------------------ ------- ---------- ---------- ----- ----- ------ -----
2091 0 omap5912.dsp Y 0x03df1d81 0x03df1d81 38 0 0 0x...
2092 1 omap5912.arm Y 0x0692602f 0x0692602f 4 0x1 0 0xc
2093 2 omap5912.unknown Y 0x00000000 0x00000000 8 0 0 0xff
2096 Unfortunately those TAPs can't always be autoconfigured,
2097 because not all devices provide good support for that.
2098 JTAG doesn't require supporting IDCODE instructions, and
2099 chips with JTAG routers may not link TAPs into the chain
2100 until they are told to do so.
2102 The configuration mechanism currently supported by OpenOCD
2103 requires explicit configuration of all TAP devices using
2104 @command{jtag newtap} commands, as detailed later in this chapter.
2105 A command like this would declare one tap and name it @code{chip1.cpu}:
2108 jtag newtap chip1 cpu -irlen 7 -ircapture 0x01 -irmask 0x55
2111 Each target configuration file lists the TAPs provided
2113 Board configuration files combine all the targets on a board,
2115 Note that @emph{the order in which TAPs are declared is very important.}
2116 It must match the order in the JTAG scan chain, both inside
2117 a single chip and between them.
2118 @xref{FAQ TAP Order}.
2120 For example, the ST Microsystems STR912 chip has
2121 three separate TAPs@footnote{See the ST
2122 document titled: @emph{STR91xFAxxx, Section 3.15 Jtag Interface, Page:
2123 28/102, Figure 3: JTAG chaining inside the STR91xFA}.
2124 @url{http://eu.st.com/stonline/products/literature/ds/13495.pdf}}.
2125 To configure those taps, @file{target/str912.cfg}
2126 includes commands something like this:
2129 jtag newtap str912 flash ... params ...
2130 jtag newtap str912 cpu ... params ...
2131 jtag newtap str912 bs ... params ...
2134 Actual config files use a variable instead of literals like
2135 @option{str912}, to support more than one chip of each type.
2136 @xref{Config File Guidelines}.
2138 @deffn Command {jtag names}
2139 Returns the names of all current TAPs in the scan chain.
2140 Use @command{jtag cget} or @command{jtag tapisenabled}
2141 to examine attributes and state of each TAP.
2143 foreach t [jtag names] @{
2144 puts [format "TAP: %s\n" $t]
2149 @deffn Command {scan_chain}
2150 Displays the TAPs in the scan chain configuration,
2152 The set of TAPs listed by this command is fixed by
2153 exiting the OpenOCD configuration stage,
2154 but systems with a JTAG router can
2155 enable or disable TAPs dynamically.
2156 In addition to the enable/disable status, the contents of
2157 each TAP's instruction register can also change.
2160 @c FIXME! "jtag cget" should be able to return all TAP
2161 @c attributes, like "$target_name cget" does for targets.
2163 @c Probably want "jtag eventlist", and a "tap-reset" event
2164 @c (on entry to RESET state).
2169 When TAP objects are declared with @command{jtag newtap},
2170 a @dfn{dotted.name} is created for the TAP, combining the
2171 name of a module (usually a chip) and a label for the TAP.
2172 For example: @code{xilinx.tap}, @code{str912.flash},
2173 @code{omap3530.jrc}, @code{dm6446.dsp}, or @code{stm32.cpu}.
2174 Many other commands use that dotted.name to manipulate or
2175 refer to the TAP. For example, CPU configuration uses the
2176 name, as does declaration of NAND or NOR flash banks.
2178 The components of a dotted name should follow ``C'' symbol
2179 name rules: start with an alphabetic character, then numbers
2180 and underscores are OK; while others (including dots!) are not.
2183 In older code, JTAG TAPs were numbered from 0..N.
2184 This feature is still present.
2185 However its use is highly discouraged, and
2186 should not be relied on; it will be removed by mid-2010.
2187 Update all of your scripts to use TAP names rather than numbers,
2188 by paying attention to the runtime warnings they trigger.
2189 Using TAP numbers in target configuration scripts prevents
2190 reusing those scripts on boards with multiple targets.
2193 @section TAP Declaration Commands
2195 @c shouldn't this be(come) a {Config Command}?
2196 @anchor{jtag newtap}
2197 @deffn Command {jtag newtap} chipname tapname configparams...
2198 Declares a new TAP with the dotted name @var{chipname}.@var{tapname},
2199 and configured according to the various @var{configparams}.
2201 The @var{chipname} is a symbolic name for the chip.
2202 Conventionally target config files use @code{$_CHIPNAME},
2203 defaulting to the model name given by the chip vendor but
2206 @cindex TAP naming convention
2207 The @var{tapname} reflects the role of that TAP,
2208 and should follow this convention:
2211 @item @code{bs} -- For boundary scan if this is a seperate TAP;
2212 @item @code{cpu} -- The main CPU of the chip, alternatively
2213 @code{arm} and @code{dsp} on chips with both ARM and DSP CPUs,
2214 @code{arm1} and @code{arm2} on chips two ARMs, and so forth;
2215 @item @code{etb} -- For an embedded trace buffer (example: an ARM ETB11);
2216 @item @code{flash} -- If the chip has a flash TAP, like the str912;
2217 @item @code{jrc} -- For JTAG route controller (example: the ICEpick modules
2218 on many Texas Instruments chips, like the OMAP3530 on Beagleboards);
2219 @item @code{tap} -- Should be used only FPGA or CPLD like devices
2221 @item @code{unknownN} -- If you have no idea what the TAP is for (N is a number);
2222 @item @emph{when in doubt} -- Use the chip maker's name in their data sheet.
2223 For example, the Freescale IMX31 has a SDMA (Smart DMA) with
2224 a JTAG TAP; that TAP should be named @code{sdma}.
2227 Every TAP requires at least the following @var{configparams}:
2230 @item @code{-ircapture} @var{NUMBER}
2231 @*The bit pattern loaded by the TAP into the JTAG shift register
2232 on entry to the @sc{ircapture} state, such as 0x01.
2233 JTAG requires the two LSBs of this value to be 01.
2234 The value is used to verify that instruction scans work correctly.
2235 @item @code{-irlen} @var{NUMBER}
2236 @*The length in bits of the
2237 instruction register, such as 4 or 5 bits.
2238 @item @code{-irmask} @var{NUMBER}
2239 @*A mask for the IR register.
2240 For some devices, there are bits in the IR that aren't used.
2241 This lets OpenOCD mask them off when doing IDCODE comparisons.
2242 In general, this should just be all ones for the size of the IR.
2245 A TAP may also provide optional @var{configparams}:
2248 @item @code{-disable} (or @code{-enable})
2249 @*Use the @code{-disable} parameter to flag a TAP which is not
2250 linked in to the scan chain after a reset using either TRST
2251 or the JTAG state machine's @sc{reset} state.
2252 You may use @code{-enable} to highlight the default state
2253 (the TAP is linked in).
2254 @xref{Enabling and Disabling TAPs}.
2255 @item @code{-expected-id} @var{number}
2256 @*A non-zero value represents the expected 32-bit IDCODE
2257 found when the JTAG chain is examined.
2258 These codes are not required by all JTAG devices.
2259 @emph{Repeat the option} as many times as required if more than one
2260 ID code could appear (for example, multiple versions).
2264 @c @deffn Command {jtag arp_init-reset}
2265 @c ... more or less "init" ?
2267 @anchor{Enabling and Disabling TAPs}
2268 @section Enabling and Disabling TAPs
2270 @cindex JTAG Route Controller
2273 In some systems, a @dfn{JTAG Route Controller} (JRC)
2274 is used to enable and/or disable specific JTAG TAPs.
2275 Many ARM based chips from Texas Instruments include
2276 an ``ICEpick'' module, which is a JRC.
2277 Such chips include DaVinci and OMAP3 processors.
2279 A given TAP may not be visible until the JRC has been
2280 told to link it into the scan chain; and if the JRC
2281 has been told to unlink that TAP, it will no longer
2283 Such routers address problems that JTAG ``bypass mode''
2287 @item The scan chain can only go as fast as its slowest TAP.
2288 @item Having many TAPs slows instruction scans, since all
2289 TAPs receive new instructions.
2290 @item TAPs in the scan chain must be powered up, which wastes
2291 power and prevents debugging some power management mechanisms.
2294 The IEEE 1149.1 JTAG standard has no concept of a ``disabled'' tap,
2295 as implied by the existence of JTAG routers.
2296 However, the upcoming IEEE 1149.7 framework (layered on top of JTAG)
2297 does include a kind of JTAG router functionality.
2299 @c (a) currently the event handlers don't seem to be able to
2300 @c fail in a way that could lead to no-change-of-state.
2301 @c (b) eventually non-event configuration should be possible,
2302 @c in which case some this documentation must move.
2304 @deffn Command {jtag cget} dotted.name @option{-event} name
2305 @deffnx Command {jtag configure} dotted.name @option{-event} name string
2306 At this writing this mechanism is used only for event handling.
2307 Three events are available. Two events relate to TAP enabling
2308 and disabling, one to post reset handling.
2310 The @code{configure} subcommand assigns an event handler,
2311 a TCL string which is evaluated when the event is triggered.
2312 The @code{cget} subcommand returns that handler.
2313 The three possible values for an event @var{name} are @option{tap-disable}, @option{tap-enable} and @option{post-reset}.
2315 So for example, when defining a TAP for a CPU connected to
2316 a JTAG router, you should define TAP event handlers using
2317 code that looks something like this:
2320 jtag configure CHIP.cpu -event tap-enable @{
2321 echo "Enabling CPU TAP"
2322 ... jtag operations using CHIP.jrc
2324 jtag configure CHIP.cpu -event tap-disable @{
2325 echo "Disabling CPU TAP"
2326 ... jtag operations using CHIP.jrc
2330 If you need some post reset action, you can do:
2333 jtag configure CHIP.cpu -event post-reset @{
2335 ... jtag operations to be done after reset
2340 @deffn Command {jtag tapdisable} dotted.name
2341 @deffnx Command {jtag tapenable} dotted.name
2342 @deffnx Command {jtag tapisenabled} dotted.name
2343 These three commands all return the string "1" if the tap
2344 specified by @var{dotted.name} is enabled,
2345 and "0" if it is disbabled.
2346 The @command{tapenable} variant first enables the tap
2347 by sending it a @option{tap-enable} event.
2348 The @command{tapdisable} variant first disables the tap
2349 by sending it a @option{tap-disable} event.
2352 Humans will find the @command{scan_chain} command more helpful
2353 than the script-oriented @command{tapisenabled}
2354 for querying the state of the JTAG taps.
2358 @node CPU Configuration
2359 @chapter CPU Configuration
2362 This chapter discusses how to set up GDB debug targets for CPUs.
2363 You can also access these targets without GDB
2364 (@pxref{Architecture and Core Commands},
2365 and @ref{Target State handling}) and
2366 through various kinds of NAND and NOR flash commands.
2367 If you have multiple CPUs you can have multiple such targets.
2369 We'll start by looking at how to examine the targets you have,
2370 then look at how to add one more target and how to configure it.
2372 @section Target List
2373 @cindex target, current
2374 @cindex target, list
2376 All targets that have been set up are part of a list,
2377 where each member has a name.
2378 That name should normally be the same as the TAP name.
2379 You can display the list with the @command{targets}
2381 This display often has only one CPU; here's what it might
2382 look like with more than one:
2384 TargetName Type Endian TapName State
2385 -- ------------------ ---------- ------ ------------------ ------------
2386 0* at91rm9200.cpu arm920t little at91rm9200.cpu running
2387 1 MyTarget cortex_m3 little mychip.foo tap-disabled
2390 One member of that list is the @dfn{current target}, which
2391 is implicitly referenced by many commands.
2392 It's the one marked with a @code{*} near the target name.
2393 In particular, memory addresses often refer to the address
2394 space seen by that current target.
2395 Commands like @command{mdw} (memory display words)
2396 and @command{flash erase_address} (erase NOR flash blocks)
2397 are examples; and there are many more.
2399 Several commands let you examine the list of targets:
2401 @deffn Command {target count}
2402 @emph{Note: target numbers are deprecated; don't use them.
2403 They will be removed shortly after August 2010, including this command.
2404 Iterate target using @command{target names}, not by counting.}
2406 Returns the number of targets, @math{N}.
2407 The highest numbered target is @math{N - 1}.
2409 set c [target count]
2410 for @{ set x 0 @} @{ $x < $c @} @{ incr x @} @{
2411 # Assuming you have created this function
2412 print_target_details $x
2417 @deffn Command {target current}
2418 Returns the name of the current target.
2421 @deffn Command {target names}
2422 Lists the names of all current targets in the list.
2424 foreach t [target names] @{
2425 puts [format "Target: %s\n" $t]
2430 @deffn Command {target number} number
2431 @emph{Note: target numbers are deprecated; don't use them.
2432 They will be removed shortly after August 2010, including this command.}
2434 The list of targets is numbered starting at zero.
2435 This command returns the name of the target at index @var{number}.
2437 set thename [target number $x]
2438 puts [format "Target %d is: %s\n" $x $thename]
2442 @c yep, "target list" would have been better.
2443 @c plus maybe "target setdefault".
2445 @deffn Command targets [name]
2446 @emph{Note: the name of this command is plural. Other target
2447 command names are singular.}
2449 With no parameter, this command displays a table of all known
2450 targets in a user friendly form.
2452 With a parameter, this command sets the current target to
2453 the given target with the given @var{name}; this is
2454 only relevant on boards which have more than one target.
2457 @section Target CPU Types and Variants
2462 Each target has a @dfn{CPU type}, as shown in the output of
2463 the @command{targets} command. You need to specify that type
2464 when calling @command{target create}.
2465 The CPU type indicates more than just the instruction set.
2466 It also indicates how that instruction set is implemented,
2467 what kind of debug support it integrates,
2468 whether it has an MMU (and if so, what kind),
2469 what core-specific commands may be available
2470 (@pxref{Architecture and Core Commands}),
2473 For some CPU types, OpenOCD also defines @dfn{variants} which
2474 indicate differences that affect their handling.
2475 For example, a particular implementation bug might need to be
2476 worked around in some chip versions.
2478 It's easy to see what target types are supported,
2479 since there's a command to list them.
2480 However, there is currently no way to list what target variants
2481 are supported (other than by reading the OpenOCD source code).
2483 @anchor{target types}
2484 @deffn Command {target types}
2485 Lists all supported target types.
2486 At this writing, the supported CPU types and variants are:
2489 @item @code{arm11} -- this is a generation of ARMv6 cores
2490 @item @code{arm720t} -- this is an ARMv4 core
2491 @item @code{arm7tdmi} -- this is an ARMv4 core
2492 @item @code{arm920t} -- this is an ARMv5 core
2493 @item @code{arm926ejs} -- this is an ARMv5 core
2494 @item @code{arm966e} -- this is an ARMv5 core
2495 @item @code{arm9tdmi} -- this is an ARMv4 core
2496 @item @code{avr} -- implements Atmel's 8-bit AVR instruction set.
2497 (Support for this is preliminary and incomplete.)
2498 @item @code{cortex_a8} -- this is an ARMv7 core
2499 @item @code{cortex_m3} -- this is an ARMv7 core, supporting only the
2500 compact Thumb2 instruction set. It supports one variant:
2502 @item @code{lm3s} ... Use this when debugging older Stellaris LM3S targets.
2503 This will cause OpenOCD to use a software reset rather than asserting
2504 SRST, to avoid a issue with clearing the debug registers.
2505 This is fixed in Fury Rev B, DustDevil Rev B, Tempest; these revisions will
2506 be detected and the normal reset behaviour used.
2508 @item @code{fa526} -- resembles arm920 (w/o Thumb)
2509 @item @code{feroceon} -- resembles arm926
2510 @item @code{mips_m4k} -- a MIPS core. This supports one variant:
2512 @item @code{ejtag_srst} ... Use this when debugging targets that do not
2513 provide a functional SRST line on the EJTAG connector. This causes
2514 OpenOCD to instead use an EJTAG software reset command to reset the
2516 You still need to enable @option{srst} on the @command{reset_config}
2517 command to enable OpenOCD hardware reset functionality.
2519 @item @code{xscale} -- this is actually an architecture,
2520 not a CPU type. It is based on the ARMv5 architecture.
2521 There are several variants defined:
2523 @item @code{ixp42x}, @code{ixp45x}, @code{ixp46x},
2524 @code{pxa27x} ... instruction register length is 7 bits
2525 @item @code{pxa250}, @code{pxa255},
2526 @code{pxa26x} ... instruction register length is 5 bits
2531 To avoid being confused by the variety of ARM based cores, remember
2532 this key point: @emph{ARM is a technology licencing company}.
2533 (See: @url{http://www.arm.com}.)
2534 The CPU name used by OpenOCD will reflect the CPU design that was
2535 licenced, not a vendor brand which incorporates that design.
2536 Name prefixes like arm7, arm9, arm11, and cortex
2537 reflect design generations;
2538 while names like ARMv4, ARMv5, ARMv6, and ARMv7
2539 reflect an architecture version implemented by a CPU design.
2541 @anchor{Target Configuration}
2542 @section Target Configuration
2544 Before creating a ``target'', you must have added its TAP to the scan chain.
2545 When you've added that TAP, you will have a @code{dotted.name}
2546 which is used to set up the CPU support.
2547 The chip-specific configuration file will normally configure its CPU(s)
2548 right after it adds all of the chip's TAPs to the scan chain.
2550 Although you can set up a target in one step, it's often clearer if you
2551 use shorter commands and do it in two steps: create it, then configure
2553 All operations on the target after it's created will use a new
2554 command, created as part of target creation.
2556 The two main things to configure after target creation are
2557 a work area, which usually has target-specific defaults even
2558 if the board setup code overrides them later;
2559 and event handlers (@pxref{Target Events}), which tend
2560 to be much more board-specific.
2561 The key steps you use might look something like this
2564 target create MyTarget cortex_m3 -chain-position mychip.cpu
2565 $MyTarget configure -work-area-phys 0x08000 -work-area-size 8096
2566 $MyTarget configure -event reset-deassert-pre @{ jtag_rclk 5 @}
2567 $MyTarget configure -event reset-init @{ myboard_reinit @}
2570 You should specify a working area if you can; typically it uses some
2572 Such a working area can speed up many things, including bulk
2573 writes to target memory;
2574 flash operations like checking to see if memory needs to be erased;
2575 GDB memory checksumming;
2579 On more complex chips, the work area can become
2580 inaccessible when application code
2581 (such as an operating system)
2582 enables or disables the MMU.
2583 For example, the particular MMU context used to acess the virtual
2584 address will probably matter ... and that context might not have
2585 easy access to other addresses needed.
2586 At this writing, OpenOCD doesn't have much MMU intelligence.
2589 It's often very useful to define a @code{reset-init} event handler.
2590 For systems that are normally used with a boot loader,
2591 common tasks include updating clocks and initializing memory
2593 That may be needed to let you write the boot loader into flash,
2594 in order to ``de-brick'' your board; or to load programs into
2595 external DDR memory without having run the boot loader.
2597 @deffn Command {target create} target_name type configparams...
2598 This command creates a GDB debug target that refers to a specific JTAG tap.
2599 It enters that target into a list, and creates a new
2600 command (@command{@var{target_name}}) which is used for various
2601 purposes including additional configuration.
2604 @item @var{target_name} ... is the name of the debug target.
2605 By convention this should be the same as the @emph{dotted.name}
2606 of the TAP associated with this target, which must be specified here
2607 using the @code{-chain-position @var{dotted.name}} configparam.
2609 This name is also used to create the target object command,
2610 referred to here as @command{$target_name},
2611 and in other places the target needs to be identified.
2612 @item @var{type} ... specifies the target type. @xref{target types}.
2613 @item @var{configparams} ... all parameters accepted by
2614 @command{$target_name configure} are permitted.
2615 If the target is big-endian, set it here with @code{-endian big}.
2616 If the variant matters, set it here with @code{-variant}.
2618 You @emph{must} set the @code{-chain-position @var{dotted.name}} here.
2622 @deffn Command {$target_name configure} configparams...
2623 The options accepted by this command may also be
2624 specified as parameters to @command{target create}.
2625 Their values can later be queried one at a time by
2626 using the @command{$target_name cget} command.
2628 @emph{Warning:} changing some of these after setup is dangerous.
2629 For example, moving a target from one TAP to another;
2630 and changing its endianness or variant.
2634 @item @code{-chain-position} @var{dotted.name} -- names the TAP
2635 used to access this target.
2637 @item @code{-endian} (@option{big}|@option{little}) -- specifies
2638 whether the CPU uses big or little endian conventions
2640 @item @code{-event} @var{event_name} @var{event_body} --
2641 @xref{Target Events}.
2642 Note that this updates a list of named event handlers.
2643 Calling this twice with two different event names assigns
2644 two different handlers, but calling it twice with the
2645 same event name assigns only one handler.
2647 @item @code{-variant} @var{name} -- specifies a variant of the target,
2648 which OpenOCD needs to know about.
2650 @item @code{-work-area-backup} (@option{0}|@option{1}) -- says
2651 whether the work area gets backed up; by default,
2652 @emph{it is not backed up.}
2653 When possible, use a working_area that doesn't need to be backed up,
2654 since performing a backup slows down operations.
2655 For example, the beginning of an SRAM block is likely to
2656 be used by most build systems, but the end is often unused.
2658 @item @code{-work-area-size} @var{size} -- specify/set the work area
2660 @item @code{-work-area-phys} @var{address} -- set the work area
2661 base @var{address} to be used when no MMU is active.
2663 @item @code{-work-area-virt} @var{address} -- set the work area
2664 base @var{address} to be used when an MMU is active.
2669 @section Other $target_name Commands
2670 @cindex object command
2672 The Tcl/Tk language has the concept of object commands,
2673 and OpenOCD adopts that same model for targets.
2675 A good Tk example is a on screen button.
2676 Once a button is created a button
2677 has a name (a path in Tk terms) and that name is useable as a first
2678 class command. For example in Tk, one can create a button and later
2679 configure it like this:
2683 button .foobar -background red -command @{ foo @}
2685 .foobar configure -foreground blue
2687 set x [.foobar cget -background]
2689 puts [format "The button is %s" $x]
2692 In OpenOCD's terms, the ``target'' is an object just like a Tcl/Tk
2693 button, and its object commands are invoked the same way.
2696 str912.cpu mww 0x1234 0x42
2697 omap3530.cpu mww 0x5555 123
2700 The commands supported by OpenOCD target objects are:
2702 @deffn Command {$target_name arp_examine}
2703 @deffnx Command {$target_name arp_halt}
2704 @deffnx Command {$target_name arp_poll}
2705 @deffnx Command {$target_name arp_reset}
2706 @deffnx Command {$target_name arp_waitstate}
2707 Internal OpenOCD scripts (most notably @file{startup.tcl})
2708 use these to deal with specific reset cases.
2709 They are not otherwise documented here.
2712 @deffn Command {$target_name array2mem} arrayname width address count
2713 @deffnx Command {$target_name mem2array} arrayname width address count
2714 These provide an efficient script-oriented interface to memory.
2715 The @code{array2mem} primitive writes bytes, halfwords, or words;
2716 while @code{mem2array} reads them.
2717 In both cases, the TCL side uses an array, and
2718 the target side uses raw memory.
2720 The efficiency comes from enabling the use of
2721 bulk JTAG data transfer operations.
2722 The script orientation comes from working with data
2723 values that are packaged for use by TCL scripts;
2724 @command{mdw} type primitives only print data they retrieve,
2725 and neither store nor return those values.
2728 @item @var{arrayname} ... is the name of an array variable
2729 @item @var{width} ... is 8/16/32 - indicating the memory access size
2730 @item @var{address} ... is the target memory address
2731 @item @var{count} ... is the number of elements to process
2735 @deffn Command {$target_name cget} queryparm
2736 Each configuration parameter accepted by
2737 @command{$target_name configure}
2738 can be individually queried, to return its current value.
2739 The @var{queryparm} is a parameter name
2740 accepted by that command, such as @code{-work-area-phys}.
2741 There are a few special cases:
2744 @item @code{-event} @var{event_name} -- returns the handler for the
2745 event named @var{event_name}.
2746 This is a special case because setting a handler requires
2748 @item @code{-type} -- returns the target type.
2749 This is a special case because this is set using
2750 @command{target create} and can't be changed
2751 using @command{$target_name configure}.
2754 For example, if you wanted to summarize information about
2755 all the targets you might use something like this:
2758 foreach name [target names] @{
2759 set y [$name cget -endian]
2760 set z [$name cget -type]
2761 puts [format "Chip %d is %s, Endian: %s, type: %s" \
2767 @anchor{target curstate}
2768 @deffn Command {$target_name curstate}
2769 Displays the current target state:
2770 @code{debug-running},
2773 @code{running}, or @code{unknown}.
2774 (Also, @pxref{Event Polling}.)
2777 @deffn Command {$target_name eventlist}
2778 Displays a table listing all event handlers
2779 currently associated with this target.
2780 @xref{Target Events}.
2783 @deffn Command {$target_name invoke-event} event_name
2784 Invokes the handler for the event named @var{event_name}.
2785 (This is primarily intended for use by OpenOCD framework
2786 code, for example by the reset code in @file{startup.tcl}.)
2789 @deffn Command {$target_name mdw} addr [count]
2790 @deffnx Command {$target_name mdh} addr [count]
2791 @deffnx Command {$target_name mdb} addr [count]
2792 Display contents of address @var{addr}, as
2793 32-bit words (@command{mdw}), 16-bit halfwords (@command{mdh}),
2794 or 8-bit bytes (@command{mdb}).
2795 If @var{count} is specified, displays that many units.
2796 (If you want to manipulate the data instead of displaying it,
2797 see the @code{mem2array} primitives.)
2800 @deffn Command {$target_name mww} addr word
2801 @deffnx Command {$target_name mwh} addr halfword
2802 @deffnx Command {$target_name mwb} addr byte
2803 Writes the specified @var{word} (32 bits),
2804 @var{halfword} (16 bits), or @var{byte} (8-bit) pattern,
2805 at the specified address @var{addr}.
2808 @anchor{Target Events}
2809 @section Target Events
2811 At various times, certain things can happen, or you want them to happen.
2814 @item What should happen when GDB connects? Should your target reset?
2815 @item When GDB tries to flash the target, do you need to enable the flash via a special command?
2816 @item During reset, do you need to write to certain memory locations
2817 to set up system clocks or
2818 to reconfigure the SDRAM?
2821 All of the above items can be addressed by target event handlers.
2822 These are set up by @command{$target_name configure -event} or
2823 @command{target create ... -event}.
2825 The programmer's model matches the @code{-command} option used in Tcl/Tk
2826 buttons and events. The two examples below act the same, but one creates
2827 and invokes a small procedure while the other inlines it.
2830 proc my_attach_proc @{ @} @{
2834 mychip.cpu configure -event gdb-attach my_attach_proc
2835 mychip.cpu configure -event gdb-attach @{
2841 The following target events are defined:
2844 @item @b{debug-halted}
2845 @* The target has halted for debug reasons (i.e.: breakpoint)
2846 @item @b{debug-resumed}
2847 @* The target has resumed (i.e.: gdb said run)
2848 @item @b{early-halted}
2849 @* Occurs early in the halt process
2851 @item @b{examine-end}
2852 @* Currently not used (goal: when JTAG examine completes)
2853 @item @b{examine-start}
2854 @* Currently not used (goal: when JTAG examine starts)
2856 @item @b{gdb-attach}
2857 @* When GDB connects
2858 @item @b{gdb-detach}
2859 @* When GDB disconnects
2861 @* When the target has halted and GDB is not doing anything (see early halt)
2862 @item @b{gdb-flash-erase-start}
2863 @* Before the GDB flash process tries to erase the flash
2864 @item @b{gdb-flash-erase-end}
2865 @* After the GDB flash process has finished erasing the flash
2866 @item @b{gdb-flash-write-start}
2867 @* Before GDB writes to the flash
2868 @item @b{gdb-flash-write-end}
2869 @* After GDB writes to the flash
2871 @* Before the target steps, gdb is trying to start/resume the target
2873 @* The target has halted
2875 @item @b{old-gdb_program_config}
2876 @* DO NOT USE THIS: Used internally
2877 @item @b{old-pre_resume}
2878 @* DO NOT USE THIS: Used internally
2880 @item @b{reset-assert-pre}
2881 @* Issued as part of @command{reset} processing
2882 after SRST and/or TRST were activated and deactivated,
2883 but before reset is asserted on the tap.
2884 @item @b{reset-assert-post}
2885 @* Issued as part of @command{reset} processing
2886 when reset is asserted on the tap.
2887 @item @b{reset-deassert-pre}
2888 @* Issued as part of @command{reset} processing
2889 when reset is about to be released on the tap.
2891 For some chips, this may be a good place to make sure
2892 the JTAG clock is slow enough to work before the PLL
2893 has been set up to allow faster JTAG speeds.
2894 @item @b{reset-deassert-post}
2895 @* Issued as part of @command{reset} processing
2896 when reset has been released on the tap.
2898 @* Issued as the final step in @command{reset} processing.
2900 @item @b{reset-halt-post}
2901 @* Currently not used
2902 @item @b{reset-halt-pre}
2903 @* Currently not used
2905 @item @b{reset-init}
2906 @* Used by @b{reset init} command for board-specific initialization.
2907 This event fires after @emph{reset-deassert-post}.
2909 This is where you would configure PLLs and clocking, set up DRAM so
2910 you can download programs that don't fit in on-chip SRAM, set up pin
2911 multiplexing, and so on.
2912 @item @b{reset-start}
2913 @* Issued as part of @command{reset} processing
2914 before either SRST or TRST are activated.
2916 @item @b{reset-wait-pos}
2917 @* Currently not used
2918 @item @b{reset-wait-pre}
2919 @* Currently not used
2921 @item @b{resume-start}
2922 @* Before any target is resumed
2923 @item @b{resume-end}
2924 @* After all targets have resumed
2928 @* Target has resumed
2932 @node Flash Commands
2933 @chapter Flash Commands
2935 OpenOCD has different commands for NOR and NAND flash;
2936 the ``flash'' command works with NOR flash, while
2937 the ``nand'' command works with NAND flash.
2938 This partially reflects different hardware technologies:
2939 NOR flash usually supports direct CPU instruction and data bus access,
2940 while data from a NAND flash must be copied to memory before it can be
2941 used. (SPI flash must also be copied to memory before use.)
2942 However, the documentation also uses ``flash'' as a generic term;
2943 for example, ``Put flash configuration in board-specific files''.
2947 @item Configure via the command @command{flash bank}
2948 @* Do this in a board-specific configuration file,
2949 passing parameters as needed by the driver.
2950 @item Operate on the flash via @command{flash subcommand}
2951 @* Often commands to manipulate the flash are typed by a human, or run
2952 via a script in some automated way. Common tasks include writing a
2953 boot loader, operating system, or other data.
2955 @* Flashing via GDB requires the flash be configured via ``flash
2956 bank'', and the GDB flash features be enabled.
2957 @xref{GDB Configuration}.
2960 Many CPUs have the ablity to ``boot'' from the first flash bank.
2961 This means that misprogramming that bank can ``brick'' a system,
2962 so that it can't boot.
2963 JTAG tools, like OpenOCD, are often then used to ``de-brick'' the
2964 board by (re)installing working boot firmware.
2966 @anchor{NOR Configuration}
2967 @section Flash Configuration Commands
2968 @cindex flash configuration
2970 @deffn {Config Command} {flash bank} driver base size chip_width bus_width target [driver_options]
2971 Configures a flash bank which provides persistent storage
2972 for addresses from @math{base} to @math{base + size - 1}.
2973 These banks will often be visible to GDB through the target's memory map.
2974 In some cases, configuring a flash bank will activate extra commands;
2975 see the driver-specific documentation.
2978 @item @var{driver} ... identifies the controller driver
2979 associated with the flash bank being declared.
2980 This is usually @code{cfi} for external flash, or else
2981 the name of a microcontroller with embedded flash memory.
2982 @xref{Flash Driver List}.
2983 @item @var{base} ... Base address of the flash chip.
2984 @item @var{size} ... Size of the chip, in bytes.
2985 For some drivers, this value is detected from the hardware.
2986 @item @var{chip_width} ... Width of the flash chip, in bytes;
2987 ignored for most microcontroller drivers.
2988 @item @var{bus_width} ... Width of the data bus used to access the
2989 chip, in bytes; ignored for most microcontroller drivers.
2990 @item @var{target} ... Names the target used to issue
2991 commands to the flash controller.
2992 @comment Actually, it's currently a controller-specific parameter...
2993 @item @var{driver_options} ... drivers may support, or require,
2994 additional parameters. See the driver-specific documentation
2995 for more information.
2998 This command is not available after OpenOCD initialization has completed.
2999 Use it in board specific configuration files, not interactively.
3003 @comment the REAL name for this command is "ocd_flash_banks"
3004 @comment less confusing would be: "flash list" (like "nand list")
3005 @deffn Command {flash banks}
3006 Prints a one-line summary of each device declared
3007 using @command{flash bank}, numbered from zero.
3008 Note that this is the @emph{plural} form;
3009 the @emph{singular} form is a very different command.
3012 @deffn Command {flash probe} num
3013 Identify the flash, or validate the parameters of the configured flash. Operation
3014 depends on the flash type.
3015 The @var{num} parameter is a value shown by @command{flash banks}.
3016 Most flash commands will implicitly @emph{autoprobe} the bank;
3017 flash drivers can distinguish between probing and autoprobing,
3018 but most don't bother.
3021 @section Erasing, Reading, Writing to Flash
3022 @cindex flash erasing
3023 @cindex flash reading
3024 @cindex flash writing
3025 @cindex flash programming
3027 One feature distinguishing NOR flash from NAND or serial flash technologies
3028 is that for read access, it acts exactly like any other addressible memory.
3029 This means you can use normal memory read commands like @command{mdw} or
3030 @command{dump_image} with it, with no special @command{flash} subcommands.
3031 @xref{Memory access}, and @ref{Image access}.
3033 Write access works differently. Flash memory normally needs to be erased
3034 before it's written. Erasing a sector turns all of its bits to ones, and
3035 writing can turn ones into zeroes. This is why there are special commands
3036 for interactive erasing and writing, and why GDB needs to know which parts
3037 of the address space hold NOR flash memory.
3040 Most of these erase and write commands leverage the fact that NOR flash
3041 chips consume target address space. They implicitly refer to the current
3042 JTAG target, and map from an address in that target's address space
3043 back to a flash bank.
3044 @comment In May 2009, those mappings may fail if any bank associated
3045 @comment with that target doesn't succesfuly autoprobe ... bug worth fixing?
3046 A few commands use abstract addressing based on bank and sector numbers,
3047 and don't depend on searching the current target and its address space.
3048 Avoid confusing the two command models.
3051 Some flash chips implement software protection against accidental writes,
3052 since such buggy writes could in some cases ``brick'' a system.
3053 For such systems, erasing and writing may require sector protection to be
3055 Examples include CFI flash such as ``Intel Advanced Bootblock flash'',
3056 and AT91SAM7 on-chip flash.
3057 @xref{flash protect}.
3059 @anchor{flash erase_sector}
3060 @deffn Command {flash erase_sector} num first last
3061 Erase sectors in bank @var{num}, starting at sector @var{first} up to and including
3062 @var{last}. Sector numbering starts at 0.
3063 The @var{num} parameter is a value shown by @command{flash banks}.
3066 @deffn Command {flash erase_address} address length
3067 Erase sectors starting at @var{address} for @var{length} bytes.
3068 The flash bank to use is inferred from the @var{address}, and
3069 the specified length must stay within that bank.
3070 As a special case, when @var{length} is zero and @var{address} is
3071 the start of the bank, the whole flash is erased.
3074 @deffn Command {flash fillw} address word length
3075 @deffnx Command {flash fillh} address halfword length
3076 @deffnx Command {flash fillb} address byte length
3077 Fills flash memory with the specified @var{word} (32 bits),
3078 @var{halfword} (16 bits), or @var{byte} (8-bit) pattern,
3079 starting at @var{address} and continuing
3080 for @var{length} units (word/halfword/byte).
3081 No erasure is done before writing; when needed, that must be done
3082 before issuing this command.
3083 Writes are done in blocks of up to 1024 bytes, and each write is
3084 verified by reading back the data and comparing it to what was written.
3085 The flash bank to use is inferred from the @var{address} of
3086 each block, and the specified length must stay within that bank.
3088 @comment no current checks for errors if fill blocks touch multiple banks!
3090 @anchor{flash write_bank}
3091 @deffn Command {flash write_bank} num filename offset
3092 Write the binary @file{filename} to flash bank @var{num},
3093 starting at @var{offset} bytes from the beginning of the bank.
3094 The @var{num} parameter is a value shown by @command{flash banks}.
3097 @anchor{flash write_image}
3098 @deffn Command {flash write_image} [erase] filename [offset] [type]
3099 Write the image @file{filename} to the current target's flash bank(s).
3100 A relocation @var{offset} may be specified, in which case it is added
3101 to the base address for each section in the image.
3102 The file [@var{type}] can be specified
3103 explicitly as @option{bin} (binary), @option{ihex} (Intel hex),
3104 @option{elf} (ELF file), @option{s19} (Motorola s19).
3105 @option{mem}, or @option{builder}.
3106 The relevant flash sectors will be erased prior to programming
3107 if the @option{erase} parameter is given.
3108 The flash bank to use is inferred from the @var{address} of
3112 @section Other Flash commands
3113 @cindex flash protection
3115 @deffn Command {flash erase_check} num
3116 Check erase state of sectors in flash bank @var{num},
3117 and display that status.
3118 The @var{num} parameter is a value shown by @command{flash banks}.
3119 This is the only operation that
3120 updates the erase state information displayed by @option{flash info}. That means you have
3121 to issue an @command{flash erase_check} command after erasing or programming the device
3122 to get updated information.
3123 (Code execution may have invalidated any state records kept by OpenOCD.)
3126 @deffn Command {flash info} num
3127 Print info about flash bank @var{num}
3128 The @var{num} parameter is a value shown by @command{flash banks}.
3129 The information includes per-sector protect status.
3132 @anchor{flash protect}
3133 @deffn Command {flash protect} num first last (on|off)
3134 Enable (@var{on}) or disable (@var{off}) protection of flash sectors
3135 @var{first} to @var{last} of flash bank @var{num}.
3136 The @var{num} parameter is a value shown by @command{flash banks}.
3139 @deffn Command {flash protect_check} num
3140 Check protection state of sectors in flash bank @var{num}.
3141 The @var{num} parameter is a value shown by @command{flash banks}.
3142 @comment @option{flash erase_sector} using the same syntax.
3145 @anchor{Flash Driver List}
3146 @section Flash Drivers, Options, and Commands
3147 As noted above, the @command{flash bank} command requires a driver name,
3148 and allows driver-specific options and behaviors.
3149 Some drivers also activate driver-specific commands.
3151 @subsection External Flash
3153 @deffn {Flash Driver} cfi
3154 @cindex Common Flash Interface
3156 The ``Common Flash Interface'' (CFI) is the main standard for
3157 external NOR flash chips, each of which connects to a
3158 specific external chip select on the CPU.
3159 Frequently the first such chip is used to boot the system.
3160 Your board's @code{reset-init} handler might need to
3161 configure additional chip selects using other commands (like: @command{mww} to
3162 configure a bus and its timings) , or
3163 perhaps configure a GPIO pin that controls the ``write protect'' pin
3165 The CFI driver can use a target-specific working area to significantly
3168 The CFI driver can accept the following optional parameters, in any order:
3171 @item @var{jedec_probe} ... is used to detect certain non-CFI flash ROMs,
3172 like AM29LV010 and similar types.
3173 @item @var{x16_as_x8} ... when a 16-bit flash is hooked up to an 8-bit bus.
3176 To configure two adjacent banks of 16 MBytes each, both sixteen bits (two bytes)
3177 wide on a sixteen bit bus:
3180 flash bank cfi 0x00000000 0x01000000 2 2 $_TARGETNAME
3181 flash bank cfi 0x01000000 0x01000000 2 2 $_TARGETNAME
3183 @c "cfi part_id" disabled
3186 @subsection Internal Flash (Microcontrollers)
3188 @deffn {Flash Driver} aduc702x
3189 The ADUC702x analog microcontrollers from Analog Devices
3190 include internal flash and use ARM7TDMI cores.
3191 The aduc702x flash driver works with models ADUC7019 through ADUC7028.
3192 The setup command only requires the @var{target} argument
3193 since all devices in this family have the same memory layout.
3196 flash bank aduc702x 0 0 0 0 $_TARGETNAME
3200 @deffn {Flash Driver} at91sam3
3202 All members of the AT91SAM3 microcontroller family from
3203 Atmel include internal flash and use ARM's Cortex-M3 core. The driver
3204 currently (6/22/09) recognizes the AT91SAM3U[1/2/4][C/E] chips. Note
3205 that the driver was orginaly developed and tested using the
3206 AT91SAM3U4E, using a SAM3U-EK eval board. Support for other chips in
3207 the family was cribbed from the data sheet. @emph{Note to future
3208 readers/updaters: Please remove this worrysome comment after other
3209 chips are confirmed.}
3211 The AT91SAM3U4[E/C] (256K) chips have two flash banks; most other chips
3212 have one flash bank. In all cases the flash banks are at
3213 the following fixed locations:
3216 # Flash bank 0 - all chips
3217 flash bank at91sam3 0x00080000 0 1 1 $_TARGETNAME
3218 # Flash bank 1 - only 256K chips
3219 flash bank at91sam3 0x00100000 0 1 1 $_TARGETNAME
3222 Internally, the AT91SAM3 flash memory is organized as follows.
3223 Unlike the AT91SAM7 chips, these are not used as parameters
3224 to the @command{flash bank} command:
3227 @item @emph{N-Banks:} 256K chips have 2 banks, others have 1 bank.
3228 @item @emph{Bank Size:} 128K/64K Per flash bank
3229 @item @emph{Sectors:} 16 or 8 per bank
3230 @item @emph{SectorSize:} 8K Per Sector
3231 @item @emph{PageSize:} 256 bytes per page. Note that OpenOCD operates on 'sector' sizes, not page sizes.
3234 The AT91SAM3 driver adds some additional commands:
3236 @deffn Command {at91sam3 gpnvm}
3237 @deffnx Command {at91sam3 gpnvm clear} number
3238 @deffnx Command {at91sam3 gpnvm set} number
3239 @deffnx Command {at91sam3 gpnvm show} [@option{all}|number]
3240 With no parameters, @command{show} or @command{show all},
3241 shows the status of all GPNVM bits.
3242 With @command{show} @var{number}, displays that bit.
3244 With @command{set} @var{number} or @command{clear} @var{number},
3245 modifies that GPNVM bit.
3248 @deffn Command {at91sam3 info}
3249 This command attempts to display information about the AT91SAM3
3250 chip. @emph{First} it read the @code{CHIPID_CIDR} [address 0x400e0740, see
3251 Section 28.2.1, page 505 of the AT91SAM3U 29/may/2009 datasheet,
3252 document id: doc6430A] and decodes the values. @emph{Second} it reads the
3253 various clock configuration registers and attempts to display how it
3254 believes the chip is configured. By default, the SLOWCLK is assumed to
3255 be 32768 Hz, see the command @command{at91sam3 slowclk}.
3258 @deffn Command {at91sam3 slowclk} [value]
3259 This command shows/sets the slow clock frequency used in the
3260 @command{at91sam3 info} command calculations above.
3264 @deffn {Flash Driver} at91sam7
3265 All members of the AT91SAM7 microcontroller family from Atmel include
3266 internal flash and use ARM7TDMI cores. The driver automatically
3267 recognizes a number of these chips using the chip identification
3268 register, and autoconfigures itself.
3271 flash bank at91sam7 0 0 0 0 $_TARGETNAME
3274 For chips which are not recognized by the controller driver, you must
3275 provide additional parameters in the following order:
3278 @item @var{chip_model} ... label used with @command{flash info}
3280 @item @var{sectors_per_bank}
3281 @item @var{pages_per_sector}
3282 @item @var{pages_size}
3283 @item @var{num_nvm_bits}
3284 @item @var{freq_khz} ... required if an external clock is provided,
3285 optional (but recommended) when the oscillator frequency is known
3288 It is recommended that you provide zeroes for all of those values
3289 except the clock frequency, so that everything except that frequency
3290 will be autoconfigured.
3291 Knowing the frequency helps ensure correct timings for flash access.
3293 The flash controller handles erases automatically on a page (128/256 byte)
3294 basis, so explicit erase commands are not necessary for flash programming.
3295 However, there is an ``EraseAll`` command that can erase an entire flash
3296 plane (of up to 256KB), and it will be used automatically when you issue
3297 @command{flash erase_sector} or @command{flash erase_address} commands.
3299 @deffn Command {at91sam7 gpnvm} bitnum (@option{set}|@option{clear})
3300 Set or clear a ``General Purpose Non-Volatle Memory'' (GPNVM)
3301 bit for the processor. Each processor has a number of such bits,
3302 used for controlling features such as brownout detection (so they
3303 are not truly general purpose).
3305 This assumes that the first flash bank (number 0) is associated with
3306 the appropriate at91sam7 target.
3311 @deffn {Flash Driver} avr
3312 The AVR 8-bit microcontrollers from Atmel integrate flash memory.
3313 @emph{The current implementation is incomplete.}
3314 @comment - defines mass_erase ... pointless given flash_erase_address
3317 @deffn {Flash Driver} ecosflash
3318 @emph{No idea what this is...}
3319 The @var{ecosflash} driver defines one mandatory parameter,
3320 the name of a modules of target code which is downloaded
3324 @deffn {Flash Driver} lpc2000
3325 Most members of the LPC1700 and LPC2000 microcontroller families from NXP
3326 include internal flash and use Cortex-M3 (LPC1700) or ARM7TDMI (LPC2000) cores.
3329 There are LPC2000 devices which are not supported by the @var{lpc2000}
3331 The LPC2888 is supported by the @var{lpc288x} driver.
3332 The LPC29xx family is supported by the @var{lpc2900} driver.
3335 The @var{lpc2000} driver defines two mandatory and one optional parameters,
3336 which must appear in the following order:
3339 @item @var{variant} ... required, may be
3340 @var{lpc2000_v1} (older LPC21xx and LPC22xx)
3341 @var{lpc2000_v2} (LPC213x, LPC214x, LPC210[123], LPC23xx and LPC24xx)
3342 or @var{lpc1700} (LPC175x and LPC176x)
3343 @item @var{clock_kHz} ... the frequency, in kiloHertz,
3344 at which the core is running
3345 @item @var{calc_checksum} ... optional (but you probably want to provide this!),
3346 telling the driver to calculate a valid checksum for the exception vector table.
3349 LPC flashes don't require the chip and bus width to be specified.
3352 flash bank lpc2000 0x0 0x7d000 0 0 $_TARGETNAME \
3353 lpc2000_v2 14765 calc_checksum
3356 @deffn {Command} {lpc2000 part_id} bank
3357 Displays the four byte part identifier associated with
3358 the specified flash @var{bank}.
3362 @deffn {Flash Driver} lpc288x
3363 The LPC2888 microcontroller from NXP needs slightly different flash
3364 support from its lpc2000 siblings.
3365 The @var{lpc288x} driver defines one mandatory parameter,
3366 the programming clock rate in Hz.
3367 LPC flashes don't require the chip and bus width to be specified.
3370 flash bank lpc288x 0 0 0 0 $_TARGETNAME 12000000
3374 @deffn {Flash Driver} lpc2900
3375 This driver supports the LPC29xx ARM968E based microcontroller family
3378 The predefined parameters @var{base}, @var{size}, @var{chip_width} and
3379 @var{bus_width} of the @code{flash bank} command are ignored. Flash size and
3380 sector layout are auto-configured by the driver.
3381 The driver has one additional mandatory parameter: The CPU clock rate
3382 (in kHz) at the time the flash operations will take place. Most of the time this
3383 will not be the crystal frequency, but a higher PLL frequency. The
3384 @code{reset-init} event handler in the board script is usually the place where
3387 The driver rejects flashless devices (currently the LPC2930).
3389 The EEPROM in LPC2900 devices is not mapped directly into the address space.
3390 It must be handled much more like NAND flash memory, and will therefore be
3391 handled by a separate @code{lpc2900_eeprom} driver (not yet available).
3393 Sector protection in terms of the LPC2900 is handled transparently. Every time a
3394 sector needs to be erased or programmed, it is automatically unprotected.
3395 What is shown as protection status in the @code{flash info} command, is
3396 actually the LPC2900 @emph{sector security}. This is a mechanism to prevent a
3397 sector from ever being erased or programmed again. As this is an irreversible
3398 mechanism, it is handled by a special command (@code{lpc2900 secure_sector}),
3399 and not by the standard @code{flash protect} command.
3401 Example for a 125 MHz clock frequency:
3403 flash bank lpc2900 0 0 0 0 $_TARGETNAME 125000
3406 Some @code{lpc2900}-specific commands are defined. In the following command list,
3407 the @var{bank} parameter is the bank number as obtained by the
3408 @code{flash banks} command.
3410 @deffn Command {lpc2900 signature} bank
3411 Calculates a 128-bit hash value, the @emph{signature}, from the whole flash
3412 content. This is a hardware feature of the flash block, hence the calculation is
3413 very fast. You may use this to verify the content of a programmed device against
3418 signature: 0x5f40cdc8:0xc64e592e:0x10490f89:0x32a0f317
3422 @deffn Command {lpc2900 read_custom} bank filename
3423 Reads the 912 bytes of customer information from the flash index sector, and
3424 saves it to a file in binary format.
3427 lpc2900 read_custom 0 /path_to/customer_info.bin
3431 The index sector of the flash is a @emph{write-only} sector. It cannot be
3432 erased! In order to guard against unintentional write access, all following
3433 commands need to be preceeded by a successful call to the @code{password}
3436 @deffn Command {lpc2900 password} bank password
3437 You need to use this command right before each of the following commands:
3438 @code{lpc2900 write_custom}, @code{lpc2900 secure_sector},
3439 @code{lpc2900 secure_jtag}.
3441 The password string is fixed to "I_know_what_I_am_doing".
3444 lpc2900 password 0 I_know_what_I_am_doing
3445 Potentially dangerous operation allowed in next command!
3449 @deffn Command {lpc2900 write_custom} bank filename type
3450 Writes the content of the file into the customer info space of the flash index
3451 sector. The filetype can be specified with the @var{type} field. Possible values
3452 for @var{type} are: @var{bin} (binary), @var{ihex} (Intel hex format),
3453 @var{elf} (ELF binary) or @var{s19} (Motorola S-records). The file must
3454 contain a single section, and the contained data length must be exactly
3456 @quotation Attention
3457 This cannot be reverted! Be careful!
3461 lpc2900 write_custom 0 /path_to/customer_info.bin bin
3465 @deffn Command {lpc2900 secure_sector} bank first last
3466 Secures the sector range from @var{first} to @var{last} (including) against
3467 further program and erase operations. The sector security will be effective
3468 after the next power cycle.
3469 @quotation Attention
3470 This cannot be reverted! Be careful!
3472 Secured sectors appear as @emph{protected} in the @code{flash info} command.
3475 lpc2900 secure_sector 0 1 1
3477 #0 : lpc2900 at 0x20000000, size 0x000c0000, (...)
3478 # 0: 0x00000000 (0x2000 8kB) not protected
3479 # 1: 0x00002000 (0x2000 8kB) protected
3480 # 2: 0x00004000 (0x2000 8kB) not protected
3484 @deffn Command {lpc2900 secure_jtag} bank
3485 Irreversibly disable the JTAG port. The new JTAG security setting will be
3486 effective after the next power cycle.
3487 @quotation Attention
3488 This cannot be reverted! Be careful!
3492 lpc2900 secure_jtag 0
3497 @deffn {Flash Driver} ocl
3498 @emph{No idea what this is, other than using some arm7/arm9 core.}
3501 flash bank ocl 0 0 0 0 $_TARGETNAME
3505 @deffn {Flash Driver} pic32mx
3506 The PIC32MX microcontrollers are based on the MIPS 4K cores,
3507 and integrate flash memory.
3508 @emph{The current implementation is incomplete.}
3511 flash bank pix32mx 0 0 0 0 $_TARGETNAME
3514 @comment numerous *disabled* commands are defined:
3515 @comment - chip_erase ... pointless given flash_erase_address
3516 @comment - lock, unlock ... pointless given protect on/off (yes?)
3517 @comment - pgm_word ... shouldn't bank be deduced from address??
3518 Some pic32mx-specific commands are defined:
3519 @deffn Command {pic32mx pgm_word} address value bank
3520 Programs the specified 32-bit @var{value} at the given @var{address}
3521 in the specified chip @var{bank}.
3525 @deffn {Flash Driver} stellaris
3526 All members of the Stellaris LM3Sxxx microcontroller family from
3528 include internal flash and use ARM Cortex M3 cores.
3529 The driver automatically recognizes a number of these chips using
3530 the chip identification register, and autoconfigures itself.
3531 @footnote{Currently there is a @command{stellaris mass_erase} command.
3532 That seems pointless since the same effect can be had using the
3533 standard @command{flash erase_address} command.}
3536 flash bank stellaris 0 0 0 0 $_TARGETNAME
3540 @deffn {Flash Driver} stm32x
3541 All members of the STM32 microcontroller family from ST Microelectronics
3542 include internal flash and use ARM Cortex M3 cores.
3543 The driver automatically recognizes a number of these chips using
3544 the chip identification register, and autoconfigures itself.
3547 flash bank stm32x 0 0 0 0 $_TARGETNAME
3550 Some stm32x-specific commands
3551 @footnote{Currently there is a @command{stm32x mass_erase} command.
3552 That seems pointless since the same effect can be had using the
3553 standard @command{flash erase_address} command.}
3556 @deffn Command {stm32x lock} num
3557 Locks the entire stm32 device.
3558 The @var{num} parameter is a value shown by @command{flash banks}.
3561 @deffn Command {stm32x unlock} num
3562 Unlocks the entire stm32 device.
3563 The @var{num} parameter is a value shown by @command{flash banks}.
3566 @deffn Command {stm32x options_read} num
3567 Read and display the stm32 option bytes written by
3568 the @command{stm32x options_write} command.
3569 The @var{num} parameter is a value shown by @command{flash banks}.
3572 @deffn Command {stm32x options_write} num (@option{SWWDG}|@option{HWWDG}) (@option{RSTSTNDBY}|@option{NORSTSTNDBY}) (@option{RSTSTOP}|@option{NORSTSTOP})
3573 Writes the stm32 option byte with the specified values.
3574 The @var{num} parameter is a value shown by @command{flash banks}.
3578 @deffn {Flash Driver} str7x
3579 All members of the STR7 microcontroller family from ST Microelectronics
3580 include internal flash and use ARM7TDMI cores.
3581 The @var{str7x} driver defines one mandatory parameter, @var{variant},
3582 which is either @code{STR71x}, @code{STR73x} or @code{STR75x}.
3585 flash bank str7x 0x40000000 0x00040000 0 0 $_TARGETNAME STR71x
3588 @deffn Command {str7x disable_jtag} bank
3589 Activate the Debug/Readout protection mechanism
3590 for the specified flash bank.
3594 @deffn {Flash Driver} str9x
3595 Most members of the STR9 microcontroller family from ST Microelectronics
3596 include internal flash and use ARM966E cores.
3597 The str9 needs the flash controller to be configured using
3598 the @command{str9x flash_config} command prior to Flash programming.
3601 flash bank str9x 0x40000000 0x00040000 0 0 $_TARGETNAME
3602 str9x flash_config 0 4 2 0 0x80000
3605 @deffn Command {str9x flash_config} num bbsr nbbsr bbadr nbbadr
3606 Configures the str9 flash controller.
3607 The @var{num} parameter is a value shown by @command{flash banks}.
3610 @item @var{bbsr} - Boot Bank Size register
3611 @item @var{nbbsr} - Non Boot Bank Size register
3612 @item @var{bbadr} - Boot Bank Start Address register
3613 @item @var{nbbadr} - Boot Bank Start Address register
3619 @deffn {Flash Driver} tms470
3620 Most members of the TMS470 microcontroller family from Texas Instruments
3621 include internal flash and use ARM7TDMI cores.
3622 This driver doesn't require the chip and bus width to be specified.
3624 Some tms470-specific commands are defined:
3626 @deffn Command {tms470 flash_keyset} key0 key1 key2 key3
3627 Saves programming keys in a register, to enable flash erase and write commands.
3630 @deffn Command {tms470 osc_mhz} clock_mhz
3631 Reports the clock speed, which is used to calculate timings.
3634 @deffn Command {tms470 plldis} (0|1)
3635 Disables (@var{1}) or enables (@var{0}) use of the PLL to speed up
3640 @subsection str9xpec driver
3643 Here is some background info to help
3644 you better understand how this driver works. OpenOCD has two flash drivers for
3648 Standard driver @option{str9x} programmed via the str9 core. Normally used for
3649 flash programming as it is faster than the @option{str9xpec} driver.
3651 Direct programming @option{str9xpec} using the flash controller. This is an
3652 ISC compilant (IEEE 1532) tap connected in series with the str9 core. The str9
3653 core does not need to be running to program using this flash driver. Typical use
3654 for this driver is locking/unlocking the target and programming the option bytes.
3657 Before we run any commands using the @option{str9xpec} driver we must first disable
3658 the str9 core. This example assumes the @option{str9xpec} driver has been
3659 configured for flash bank 0.
3661 # assert srst, we do not want core running
3662 # while accessing str9xpec flash driver
3664 # turn off target polling
3667 str9xpec enable_turbo 0
3669 str9xpec options_read 0
3670 # re-enable str9 core
3671 str9xpec disable_turbo 0
3675 The above example will read the str9 option bytes.
3676 When performing a unlock remember that you will not be able to halt the str9 - it
3677 has been locked. Halting the core is not required for the @option{str9xpec} driver
3678 as mentioned above, just issue the commands above manually or from a telnet prompt.
3680 @deffn {Flash Driver} str9xpec
3681 Only use this driver for locking/unlocking the device or configuring the option bytes.
3682 Use the standard str9 driver for programming.
3683 Before using the flash commands the turbo mode must be enabled using the
3684 @command{str9xpec enable_turbo} command.
3686 Several str9xpec-specific commands are defined:
3688 @deffn Command {str9xpec disable_turbo} num
3689 Restore the str9 into JTAG chain.
3692 @deffn Command {str9xpec enable_turbo} num
3693 Enable turbo mode, will simply remove the str9 from the chain and talk
3694 directly to the embedded flash controller.
3697 @deffn Command {str9xpec lock} num
3698 Lock str9 device. The str9 will only respond to an unlock command that will
3702 @deffn Command {str9xpec part_id} num
3703 Prints the part identifier for bank @var{num}.
3706 @deffn Command {str9xpec options_cmap} num (@option{bank0}|@option{bank1})
3707 Configure str9 boot bank.
3710 @deffn Command {str9xpec options_lvdsel} num (@option{vdd}|@option{vdd_vddq})
3711 Configure str9 lvd source.
3714 @deffn Command {str9xpec options_lvdthd} num (@option{2.4v}|@option{2.7v})
3715 Configure str9 lvd threshold.
3718 @deffn Command {str9xpec options_lvdwarn} bank (@option{vdd}|@option{vdd_vddq})
3719 Configure str9 lvd reset warning source.
3722 @deffn Command {str9xpec options_read} num
3723 Read str9 option bytes.
3726 @deffn Command {str9xpec options_write} num
3727 Write str9 option bytes.
3730 @deffn Command {str9xpec unlock} num
3739 @subsection mFlash Configuration
3740 @cindex mFlash Configuration
3742 @deffn {Config Command} {mflash bank} soc base RST_pin target
3743 Configures a mflash for @var{soc} host bank at
3745 The pin number format depends on the host GPIO naming convention.
3746 Currently, the mflash driver supports s3c2440 and pxa270.
3748 Example for s3c2440 mflash where @var{RST pin} is GPIO B1:
3751 mflash bank s3c2440 0x10000000 1b 0
3754 Example for pxa270 mflash where @var{RST pin} is GPIO 43:
3757 mflash bank pxa270 0x08000000 43 0
3761 @subsection mFlash commands
3762 @cindex mFlash commands
3764 @deffn Command {mflash config pll} frequency
3765 Configure mflash PLL.
3766 The @var{frequency} is the mflash input frequency, in Hz.
3767 Issuing this command will erase mflash's whole internal nand and write new pll.
3768 After this command, mflash needs power-on-reset for normal operation.
3769 If pll was newly configured, storage and boot(optional) info also need to be update.
3772 @deffn Command {mflash config boot}
3773 Configure bootable option.
3774 If bootable option is set, mflash offer the first 8 sectors
3778 @deffn Command {mflash config storage}
3779 Configure storage information.
3780 For the normal storage operation, this information must be
3784 @deffn Command {mflash dump} num filename offset size
3785 Dump @var{size} bytes, starting at @var{offset} bytes from the
3786 beginning of the bank @var{num}, to the file named @var{filename}.
3789 @deffn Command {mflash probe}
3793 @deffn Command {mflash write} num filename offset
3794 Write the binary file @var{filename} to mflash bank @var{num}, starting at
3795 @var{offset} bytes from the beginning of the bank.
3798 @node NAND Flash Commands
3799 @chapter NAND Flash Commands
3802 Compared to NOR or SPI flash, NAND devices are inexpensive
3803 and high density. Today's NAND chips, and multi-chip modules,
3804 commonly hold multiple GigaBytes of data.
3806 NAND chips consist of a number of ``erase blocks'' of a given
3807 size (such as 128 KBytes), each of which is divided into a
3808 number of pages (of perhaps 512 or 2048 bytes each). Each
3809 page of a NAND flash has an ``out of band'' (OOB) area to hold
3810 Error Correcting Code (ECC) and other metadata, usually 16 bytes
3811 of OOB for every 512 bytes of page data.
3813 One key characteristic of NAND flash is that its error rate
3814 is higher than that of NOR flash. In normal operation, that
3815 ECC is used to correct and detect errors. However, NAND
3816 blocks can also wear out and become unusable; those blocks
3817 are then marked "bad". NAND chips are even shipped from the
3818 manufacturer with a few bad blocks. The highest density chips
3819 use a technology (MLC) that wears out more quickly, so ECC
3820 support is increasingly important as a way to detect blocks
3821 that have begun to fail, and help to preserve data integrity
3822 with techniques such as wear leveling.
3824 Software is used to manage the ECC. Some controllers don't
3825 support ECC directly; in those cases, software ECC is used.
3826 Other controllers speed up the ECC calculations with hardware.
3827 Single-bit error correction hardware is routine. Controllers
3828 geared for newer MLC chips may correct 4 or more errors for
3829 every 512 bytes of data.
3831 You will need to make sure that any data you write using
3832 OpenOCD includes the apppropriate kind of ECC. For example,
3833 that may mean passing the @code{oob_softecc} flag when
3834 writing NAND data, or ensuring that the correct hardware
3837 The basic steps for using NAND devices include:
3839 @item Declare via the command @command{nand device}
3840 @* Do this in a board-specific configuration file,
3841 passing parameters as needed by the controller.
3842 @item Configure each device using @command{nand probe}.
3843 @* Do this only after the associated target is set up,
3844 such as in its reset-init script or in procures defined
3845 to access that device.
3846 @item Operate on the flash via @command{nand subcommand}
3847 @* Often commands to manipulate the flash are typed by a human, or run
3848 via a script in some automated way. Common task include writing a
3849 boot loader, operating system, or other data needed to initialize or
3853 @b{NOTE:} At the time this text was written, the largest NAND
3854 flash fully supported by OpenOCD is 2 GiBytes (16 GiBits).
3855 This is because the variables used to hold offsets and lengths
3856 are only 32 bits wide.
3857 (Larger chips may work in some cases, unless an offset or length
3858 is larger than 0xffffffff, the largest 32-bit unsigned integer.)
3859 Some larger devices will work, since they are actually multi-chip
3860 modules with two smaller chips and individual chipselect lines.
3862 @anchor{NAND Configuration}
3863 @section NAND Configuration Commands
3864 @cindex NAND configuration
3866 NAND chips must be declared in configuration scripts,
3867 plus some additional configuration that's done after
3868 OpenOCD has initialized.
3870 @deffn {Config Command} {nand device} controller target [configparams...]
3871 Declares a NAND device, which can be read and written to
3872 after it has been configured through @command{nand probe}.
3873 In OpenOCD, devices are single chips; this is unlike some
3874 operating systems, which may manage multiple chips as if
3875 they were a single (larger) device.
3876 In some cases, configuring a device will activate extra
3877 commands; see the controller-specific documentation.
3879 @b{NOTE:} This command is not available after OpenOCD
3880 initialization has completed. Use it in board specific
3881 configuration files, not interactively.
3884 @item @var{controller} ... identifies the controller driver
3885 associated with the NAND device being declared.
3886 @xref{NAND Driver List}.
3887 @item @var{target} ... names the target used when issuing
3888 commands to the NAND controller.
3889 @comment Actually, it's currently a controller-specific parameter...
3890 @item @var{configparams} ... controllers may support, or require,
3891 additional parameters. See the controller-specific documentation
3892 for more information.
3896 @deffn Command {nand list}
3897 Prints a one-line summary of each device declared
3898 using @command{nand device}, numbered from zero.
3899 Note that un-probed devices show no details.
3902 @deffn Command {nand probe} num
3903 Probes the specified device to determine key characteristics
3904 like its page and block sizes, and how many blocks it has.
3905 The @var{num} parameter is the value shown by @command{nand list}.
3906 You must (successfully) probe a device before you can use
3907 it with most other NAND commands.
3910 @section Erasing, Reading, Writing to NAND Flash
3912 @deffn Command {nand dump} num filename offset length [oob_option]
3913 @cindex NAND reading
3914 Reads binary data from the NAND device and writes it to the file,
3915 starting at the specified offset.
3916 The @var{num} parameter is the value shown by @command{nand list}.
3918 Use a complete path name for @var{filename}, so you don't depend
3919 on the directory used to start the OpenOCD server.
3921 The @var{offset} and @var{length} must be exact multiples of the
3922 device's page size. They describe a data region; the OOB data
3923 associated with each such page may also be accessed.
3925 @b{NOTE:} At the time this text was written, no error correction
3926 was done on the data that's read, unless raw access was disabled
3927 and the underlying NAND controller driver had a @code{read_page}
3928 method which handled that error correction.
3930 By default, only page data is saved to the specified file.
3931 Use an @var{oob_option} parameter to save OOB data:
3933 @item no oob_* parameter
3934 @*Output file holds only page data; OOB is discarded.
3935 @item @code{oob_raw}
3936 @*Output file interleaves page data and OOB data;
3937 the file will be longer than "length" by the size of the
3938 spare areas associated with each data page.
3939 Note that this kind of "raw" access is different from
3940 what's implied by @command{nand raw_access}, which just
3941 controls whether a hardware-aware access method is used.
3942 @item @code{oob_only}
3943 @*Output file has only raw OOB data, and will
3944 be smaller than "length" since it will contain only the
3945 spare areas associated with each data page.
3949 @deffn Command {nand erase} num offset length
3950 @cindex NAND erasing
3951 @cindex NAND programming
3952 Erases blocks on the specified NAND device, starting at the
3953 specified @var{offset} and continuing for @var{length} bytes.
3954 Both of those values must be exact multiples of the device's
3955 block size, and the region they specify must fit entirely in the chip.
3956 The @var{num} parameter is the value shown by @command{nand list}.
3958 @b{NOTE:} This command will try to erase bad blocks, when told
3959 to do so, which will probably invalidate the manufacturer's bad
3961 For the remainder of the current server session, @command{nand info}
3962 will still report that the block ``is'' bad.
3965 @deffn Command {nand write} num filename offset [option...]
3966 @cindex NAND writing
3967 @cindex NAND programming
3968 Writes binary data from the file into the specified NAND device,
3969 starting at the specified offset. Those pages should already
3970 have been erased; you can't change zero bits to one bits.
3971 The @var{num} parameter is the value shown by @command{nand list}.
3973 Use a complete path name for @var{filename}, so you don't depend
3974 on the directory used to start the OpenOCD server.
3976 The @var{offset} must be an exact multiple of the device's page size.
3977 All data in the file will be written, assuming it doesn't run
3978 past the end of the device.
3979 Only full pages are written, and any extra space in the last
3980 page will be filled with 0xff bytes. (That includes OOB data,
3981 if that's being written.)
3983 @b{NOTE:} At the time this text was written, bad blocks are
3984 ignored. That is, this routine will not skip bad blocks,
3985 but will instead try to write them. This can cause problems.
3987 Provide at most one @var{option} parameter. With some
3988 NAND drivers, the meanings of these parameters may change
3989 if @command{nand raw_access} was used to disable hardware ECC.
3991 @item no oob_* parameter
3992 @*File has only page data, which is written.
3993 If raw acccess is in use, the OOB area will not be written.
3994 Otherwise, if the underlying NAND controller driver has
3995 a @code{write_page} routine, that routine may write the OOB
3996 with hardware-computed ECC data.
3997 @item @code{oob_only}
3998 @*File has only raw OOB data, which is written to the OOB area.
3999 Each page's data area stays untouched. @i{This can be a dangerous
4000 option}, since it can invalidate the ECC data.
4001 You may need to force raw access to use this mode.
4002 @item @code{oob_raw}
4003 @*File interleaves data and OOB data, both of which are written
4004 If raw access is enabled, the data is written first, then the
4006 Otherwise, if the underlying NAND controller driver has
4007 a @code{write_page} routine, that routine may modify the OOB
4008 before it's written, to include hardware-computed ECC data.
4009 @item @code{oob_softecc}
4010 @*File has only page data, which is written.
4011 The OOB area is filled with 0xff, except for a standard 1-bit
4012 software ECC code stored in conventional locations.
4013 You might need to force raw access to use this mode, to prevent
4014 the underlying driver from applying hardware ECC.
4015 @item @code{oob_softecc_kw}
4016 @*File has only page data, which is written.
4017 The OOB area is filled with 0xff, except for a 4-bit software ECC
4018 specific to the boot ROM in Marvell Kirkwood SoCs.
4019 You might need to force raw access to use this mode, to prevent
4020 the underlying driver from applying hardware ECC.
4024 @section Other NAND commands
4025 @cindex NAND other commands
4027 @deffn Command {nand check_bad_blocks} [offset length]
4028 Checks for manufacturer bad block markers on the specified NAND
4029 device. If no parameters are provided, checks the whole
4030 device; otherwise, starts at the specified @var{offset} and
4031 continues for @var{length} bytes.
4032 Both of those values must be exact multiples of the device's
4033 block size, and the region they specify must fit entirely in the chip.
4034 The @var{num} parameter is the value shown by @command{nand list}.
4036 @b{NOTE:} Before using this command you should force raw access
4037 with @command{nand raw_access enable} to ensure that the underlying
4038 driver will not try to apply hardware ECC.
4041 @deffn Command {nand info} num
4042 The @var{num} parameter is the value shown by @command{nand list}.
4043 This prints the one-line summary from "nand list", plus for
4044 devices which have been probed this also prints any known
4045 status for each block.
4048 @deffn Command {nand raw_access} num (@option{enable}|@option{disable})
4049 Sets or clears an flag affecting how page I/O is done.
4050 The @var{num} parameter is the value shown by @command{nand list}.
4052 This flag is cleared (disabled) by default, but changing that
4053 value won't affect all NAND devices. The key factor is whether
4054 the underlying driver provides @code{read_page} or @code{write_page}
4055 methods. If it doesn't provide those methods, the setting of
4056 this flag is irrelevant; all access is effectively ``raw''.
4058 When those methods exist, they are normally used when reading
4059 data (@command{nand dump} or reading bad block markers) or
4060 writing it (@command{nand write}). However, enabling
4061 raw access (setting the flag) prevents use of those methods,
4062 bypassing hardware ECC logic.
4063 @i{This can be a dangerous option}, since writing blocks
4064 with the wrong ECC data can cause them to be marked as bad.
4067 @anchor{NAND Driver List}
4068 @section NAND Drivers, Options, and Commands
4069 As noted above, the @command{nand device} command allows
4070 driver-specific options and behaviors.
4071 Some controllers also activate controller-specific commands.
4073 @deffn {NAND Driver} davinci
4074 This driver handles the NAND controllers found on DaVinci family
4075 chips from Texas Instruments.
4076 It takes three extra parameters:
4077 address of the NAND chip;
4078 hardware ECC mode to use (hwecc1, hwecc4, hwecc4_infix);
4079 address of the AEMIF controller on this processor.
4081 nand device davinci dm355.arm 0x02000000 hwecc4 0x01e10000
4083 All DaVinci processors support the single-bit ECC hardware,
4084 and newer ones also support the four-bit ECC hardware.
4085 The @code{write_page} and @code{read_page} methods are used
4086 to implement those ECC modes, unless they are disabled using
4087 the @command{nand raw_access} command.
4090 @deffn {NAND Driver} lpc3180
4091 These controllers require an extra @command{nand device}
4092 parameter: the clock rate used by the controller.
4093 @deffn Command {lpc3180 select} num [mlc|slc]
4094 Configures use of the MLC or SLC controller mode.
4095 MLC implies use of hardware ECC.
4096 The @var{num} parameter is the value shown by @command{nand list}.
4099 At this writing, this driver includes @code{write_page}
4100 and @code{read_page} methods. Using @command{nand raw_access}
4101 to disable those methods will prevent use of hardware ECC
4102 in the MLC controller mode, but won't change SLC behavior.
4104 @comment current lpc3180 code won't issue 5-byte address cycles
4106 @deffn {NAND Driver} orion
4107 These controllers require an extra @command{nand device}
4108 parameter: the address of the controller.
4110 nand device orion 0xd8000000
4112 These controllers don't define any specialized commands.
4113 At this writing, their drivers don't include @code{write_page}
4114 or @code{read_page} methods, so @command{nand raw_access} won't
4115 change any behavior.
4118 @deffn {NAND Driver} s3c2410
4119 @deffnx {NAND Driver} s3c2412
4120 @deffnx {NAND Driver} s3c2440
4121 @deffnx {NAND Driver} s3c2443
4122 These S3C24xx family controllers don't have any special
4123 @command{nand device} options, and don't define any
4124 specialized commands.
4125 At this writing, their drivers don't include @code{write_page}
4126 or @code{read_page} methods, so @command{nand raw_access} won't
4127 change any behavior.
4130 @node PLD/FPGA Commands
4131 @chapter PLD/FPGA Commands
4135 Programmable Logic Devices (PLDs) and the more flexible
4136 Field Programmable Gate Arrays (FPGAs) are both types of programmable hardware.
4137 OpenOCD can support programming them.
4138 Although PLDs are generally restrictive (cells are less functional, and
4139 there are no special purpose cells for memory or computational tasks),
4140 they share the same OpenOCD infrastructure.
4141 Accordingly, both are called PLDs here.
4143 @section PLD/FPGA Configuration and Commands
4145 As it does for JTAG TAPs, debug targets, and flash chips (both NOR and NAND),
4146 OpenOCD maintains a list of PLDs available for use in various commands.
4147 Also, each such PLD requires a driver.
4149 They are referenced by the number shown by the @command{pld devices} command,
4150 and new PLDs are defined by @command{pld device driver_name}.
4152 @deffn {Config Command} {pld device} driver_name tap_name [driver_options]
4153 Defines a new PLD device, supported by driver @var{driver_name},
4154 using the TAP named @var{tap_name}.
4155 The driver may make use of any @var{driver_options} to configure its
4159 @deffn {Command} {pld devices}
4160 Lists the PLDs and their numbers.
4163 @deffn {Command} {pld load} num filename
4164 Loads the file @file{filename} into the PLD identified by @var{num}.
4165 The file format must be inferred by the driver.
4168 @section PLD/FPGA Drivers, Options, and Commands
4170 Drivers may support PLD-specific options to the @command{pld device}
4171 definition command, and may also define commands usable only with
4172 that particular type of PLD.
4174 @deffn {FPGA Driver} virtex2
4175 Virtex-II is a family of FPGAs sold by Xilinx.
4176 It supports the IEEE 1532 standard for In-System Configuration (ISC).
4177 No driver-specific PLD definition options are used,
4178 and one driver-specific command is defined.
4180 @deffn {Command} {virtex2 read_stat} num
4181 Reads and displays the Virtex-II status register (STAT)
4186 @node General Commands
4187 @chapter General Commands
4190 The commands documented in this chapter here are common commands that
4191 you, as a human, may want to type and see the output of. Configuration type
4192 commands are documented elsewhere.
4196 @item @b{Source Of Commands}
4197 @* OpenOCD commands can occur in a configuration script (discussed
4198 elsewhere) or typed manually by a human or supplied programatically,
4199 or via one of several TCP/IP Ports.
4201 @item @b{From the human}
4202 @* A human should interact with the telnet interface (default port: 4444)
4203 or via GDB (default port 3333).
4205 To issue commands from within a GDB session, use the @option{monitor}
4206 command, e.g. use @option{monitor poll} to issue the @option{poll}
4207 command. All output is relayed through the GDB session.
4209 @item @b{Machine Interface}
4210 The Tcl interface's intent is to be a machine interface. The default Tcl
4215 @section Daemon Commands
4217 @deffn {Command} exit
4218 Exits the current telnet session.
4221 @c note EXTREMELY ANNOYING word wrap at column 75
4222 @c even when lines are e.g. 100+ columns ...
4223 @c coded in startup.tcl
4224 @deffn {Command} help [string]
4225 With no parameters, prints help text for all commands.
4226 Otherwise, prints each helptext containing @var{string}.
4227 Not every command provides helptext.
4230 @deffn Command sleep msec [@option{busy}]
4231 Wait for at least @var{msec} milliseconds before resuming.
4232 If @option{busy} is passed, busy-wait instead of sleeping.
4233 (This option is strongly discouraged.)
4234 Useful in connection with script files
4235 (@command{script} command and @command{target_name} configuration).
4238 @deffn Command shutdown
4239 Close the OpenOCD daemon, disconnecting all clients (GDB, telnet, other).
4242 @anchor{debug_level}
4243 @deffn Command debug_level [n]
4244 @cindex message level
4245 Display debug level.
4246 If @var{n} (from 0..3) is provided, then set it to that level.
4247 This affects the kind of messages sent to the server log.
4248 Level 0 is error messages only;
4249 level 1 adds warnings;
4250 level 2 adds informational messages;
4251 and level 3 adds debugging messages.
4252 The default is level 2, but that can be overridden on
4253 the command line along with the location of that log
4254 file (which is normally the server's standard output).
4258 @deffn Command fast (@option{enable}|@option{disable})
4260 Set default behaviour of OpenOCD to be "fast and dangerous".
4262 At this writing, this only affects the defaults for two ARM7/ARM9 parameters:
4263 fast memory access, and DCC downloads. Those parameters may still be
4264 individually overridden.
4266 The target specific "dangerous" optimisation tweaking options may come and go
4267 as more robust and user friendly ways are found to ensure maximum throughput
4268 and robustness with a minimum of configuration.
4270 Typically the "fast enable" is specified first on the command line:
4273 openocd -c "fast enable" -c "interface dummy" -f target/str710.cfg
4277 @deffn Command echo message
4278 Logs a message at "user" priority.
4279 Output @var{message} to stdout.
4281 echo "Downloading kernel -- please wait"
4285 @deffn Command log_output [filename]
4286 Redirect logging to @var{filename};
4287 the initial log output channel is stderr.
4290 @anchor{Target State handling}
4291 @section Target State handling
4294 @cindex target initialization
4296 In this section ``target'' refers to a CPU configured as
4297 shown earlier (@pxref{CPU Configuration}).
4298 These commands, like many, implicitly refer to
4299 a current target which is used to perform the
4300 various operations. The current target may be changed
4301 by using @command{targets} command with the name of the
4302 target which should become current.
4304 @deffn Command reg [(number|name) [value]]
4305 Access a single register by @var{number} or by its @var{name}.
4307 @emph{With no arguments}:
4308 list all available registers for the current target,
4309 showing number, name, size, value, and cache status.
4311 @emph{With number/name}: display that register's value.
4313 @emph{With both number/name and value}: set register's value.
4315 Cores may have surprisingly many registers in their
4316 Debug and trace infrastructure:
4320 (0) r0 (/32): 0x0000D3C2 (dirty: 1, valid: 1)
4321 (1) r1 (/32): 0xFD61F31C (dirty: 0, valid: 1)
4322 (2) r2 (/32): 0x00022551 (dirty: 0, valid: 1)
4324 (164) ETM_CONTEXTID_COMPARATOR_MASK (/32): \
4325 0x00000000 (dirty: 0, valid: 0)
4330 @deffn Command halt [ms]
4331 @deffnx Command wait_halt [ms]
4332 The @command{halt} command first sends a halt request to the target,
4333 which @command{wait_halt} doesn't.
4334 Otherwise these behave the same: wait up to @var{ms} milliseconds,
4335 or 5 seconds if there is no parameter, for the target to halt
4336 (and enter debug mode).
4337 Using 0 as the @var{ms} parameter prevents OpenOCD from waiting.
4340 On ARM cores, software using the @emph{wait for interrupt} operation
4341 often blocks the JTAG access needed by a @command{halt} command.
4342 This is because that operation also puts the core into a low
4343 power mode by gating the core clock;
4344 but the core clock is needed to detect JTAG clock transitions.
4346 One partial workaround uses adaptive clocking: when the core is
4347 interrupted the operation completes, then JTAG clocks are accepted
4348 at least until the interrupt handler completes.
4349 However, this workaround is often unusable since the processor, board,
4350 and JTAG adapter must all support adaptive JTAG clocking.
4351 Also, it can't work until an interrupt is issued.
4353 A more complete workaround is to not use that operation while you
4354 work with a JTAG debugger.
4355 Tasking environments generaly have idle loops where the body is the
4356 @emph{wait for interrupt} operation.
4357 (On older cores, it is a coprocessor action;
4358 newer cores have a @option{wfi} instruction.)
4359 Such loops can just remove that operation, at the cost of higher
4360 power consumption (because the CPU is needlessly clocked).
4365 @deffn Command resume [address]
4366 Resume the target at its current code position,
4367 or the optional @var{address} if it is provided.
4368 OpenOCD will wait 5 seconds for the target to resume.
4371 @deffn Command step [address]
4372 Single-step the target at its current code position,
4373 or the optional @var{address} if it is provided.
4376 @anchor{Reset Command}
4377 @deffn Command reset
4378 @deffnx Command {reset run}
4379 @deffnx Command {reset halt}
4380 @deffnx Command {reset init}
4381 Perform as hard a reset as possible, using SRST if possible.
4382 @emph{All defined targets will be reset, and target
4383 events will fire during the reset sequence.}
4385 The optional parameter specifies what should
4386 happen after the reset.
4387 If there is no parameter, a @command{reset run} is executed.
4388 The other options will not work on all systems.
4389 @xref{Reset Configuration}.
4392 @item @b{run} Let the target run
4393 @item @b{halt} Immediately halt the target
4394 @item @b{init} Immediately halt the target, and execute the reset-init script
4398 @deffn Command soft_reset_halt
4399 Requesting target halt and executing a soft reset. This is often used
4400 when a target cannot be reset and halted. The target, after reset is
4401 released begins to execute code. OpenOCD attempts to stop the CPU and
4402 then sets the program counter back to the reset vector. Unfortunately
4403 the code that was executed may have left the hardware in an unknown
4407 @section I/O Utilities
4409 These commands are available when
4410 OpenOCD is built with @option{--enable-ioutil}.
4411 They are mainly useful on embedded targets,
4413 Hosts with operating systems have complementary tools.
4415 @emph{Note:} there are several more such commands.
4417 @deffn Command append_file filename [string]*
4418 Appends the @var{string} parameters to
4419 the text file @file{filename}.
4420 Each string except the last one is followed by one space.
4421 The last string is followed by a newline.
4424 @deffn Command cat filename
4425 Reads and displays the text file @file{filename}.
4428 @deffn Command cp src_filename dest_filename
4429 Copies contents from the file @file{src_filename}
4430 into @file{dest_filename}.
4434 @emph{No description provided.}
4438 @emph{No description provided.}
4442 @emph{No description provided.}
4445 @deffn Command meminfo
4446 Display available RAM memory on OpenOCD host.
4447 Used in OpenOCD regression testing scripts.
4451 @emph{No description provided.}
4455 @emph{No description provided.}
4458 @deffn Command rm filename
4459 @c "rm" has both normal and Jim-level versions??
4460 Unlinks the file @file{filename}.
4463 @deffn Command trunc filename
4464 Removes all data in the file @file{filename}.
4467 @anchor{Memory access}
4468 @section Memory access commands
4469 @cindex memory access
4471 These commands allow accesses of a specific size to the memory
4472 system. Often these are used to configure the current target in some
4473 special way. For example - one may need to write certain values to the
4474 SDRAM controller to enable SDRAM.
4477 @item Use the @command{targets} (plural) command
4478 to change the current target.
4479 @item In system level scripts these commands are deprecated.
4480 Please use their TARGET object siblings to avoid making assumptions
4481 about what TAP is the current target, or about MMU configuration.
4484 @deffn Command mdw addr [count]
4485 @deffnx Command mdh addr [count]
4486 @deffnx Command mdb addr [count]
4487 Display contents of address @var{addr}, as
4488 32-bit words (@command{mdw}), 16-bit halfwords (@command{mdh}),
4489 or 8-bit bytes (@command{mdb}).
4490 If @var{count} is specified, displays that many units.
4491 (If you want to manipulate the data instead of displaying it,
4492 see the @code{mem2array} primitives.)
4495 @deffn Command mww addr word
4496 @deffnx Command mwh addr halfword
4497 @deffnx Command mwb addr byte
4498 Writes the specified @var{word} (32 bits),
4499 @var{halfword} (16 bits), or @var{byte} (8-bit) pattern,
4500 at the specified address @var{addr}.
4504 @anchor{Image access}
4505 @section Image loading commands
4506 @cindex image loading
4507 @cindex image dumping
4510 @deffn Command {dump_image} filename address size
4511 Dump @var{size} bytes of target memory starting at @var{address} to the
4512 binary file named @var{filename}.
4515 @deffn Command {fast_load}
4516 Loads an image stored in memory by @command{fast_load_image} to the
4517 current target. Must be preceeded by fast_load_image.
4520 @deffn Command {fast_load_image} filename address [@option{bin}|@option{ihex}|@option{elf}]
4521 Normally you should be using @command{load_image} or GDB load. However, for
4522 testing purposes or when I/O overhead is significant(OpenOCD running on an embedded
4523 host), storing the image in memory and uploading the image to the target
4524 can be a way to upload e.g. multiple debug sessions when the binary does not change.
4525 Arguments are the same as @command{load_image}, but the image is stored in OpenOCD host
4526 memory, i.e. does not affect target. This approach is also useful when profiling
4527 target programming performance as I/O and target programming can easily be profiled
4532 @deffn Command {load_image} filename address [@option{bin}|@option{ihex}|@option{elf}]
4533 Load image from file @var{filename} to target memory at @var{address}.
4534 The file format may optionally be specified
4535 (@option{bin}, @option{ihex}, or @option{elf})
4538 @deffn Command {test_image} filename [address [@option{bin}|@option{ihex}|@option{elf}]]
4539 Displays image section sizes and addresses
4540 as if @var{filename} were loaded into target memory
4541 starting at @var{address} (defaults to zero).
4542 The file format may optionally be specified
4543 (@option{bin}, @option{ihex}, or @option{elf})
4546 @deffn Command {verify_image} filename address [@option{bin}|@option{ihex}|@option{elf}]
4547 Verify @var{filename} against target memory starting at @var{address}.
4548 The file format may optionally be specified
4549 (@option{bin}, @option{ihex}, or @option{elf})
4550 This will first attempt a comparison using a CRC checksum, if this fails it will try a binary compare.
4554 @section Breakpoint and Watchpoint commands
4558 CPUs often make debug modules accessible through JTAG, with
4559 hardware support for a handful of code breakpoints and data
4561 In addition, CPUs almost always support software breakpoints.
4563 @deffn Command {bp} [address len [@option{hw}]]
4564 With no parameters, lists all active breakpoints.
4565 Else sets a breakpoint on code execution starting
4566 at @var{address} for @var{length} bytes.
4567 This is a software breakpoint, unless @option{hw} is specified
4568 in which case it will be a hardware breakpoint.
4570 (@xref{arm9tdmi vector_catch}, or @pxref{xscale vector_catch},
4571 for similar mechanisms that do not consume hardware breakpoints.)
4574 @deffn Command {rbp} address
4575 Remove the breakpoint at @var{address}.
4578 @deffn Command {rwp} address
4579 Remove data watchpoint on @var{address}
4582 @deffn Command {wp} [address len [(@option{r}|@option{w}|@option{a}) [value [mask]]]]
4583 With no parameters, lists all active watchpoints.
4584 Else sets a data watchpoint on data from @var{address} for @var{length} bytes.
4585 The watch point is an "access" watchpoint unless
4586 the @option{r} or @option{w} parameter is provided,
4587 defining it as respectively a read or write watchpoint.
4588 If a @var{value} is provided, that value is used when determining if
4589 the watchpoint should trigger. The value may be first be masked
4590 using @var{mask} to mark ``don't care'' fields.
4593 @section Misc Commands
4596 @deffn Command {profile} seconds filename
4597 Profiling samples the CPU's program counter as quickly as possible,
4598 which is useful for non-intrusive stochastic profiling.
4599 Saves up to 10000 sampines in @file{filename} using ``gmon.out'' format.
4602 @deffn Command {version}
4603 Displays a string identifying the version of this OpenOCD server.
4606 @deffn Command {virt2phys} virtual_address
4607 Requests the current target to map the specified @var{virtual_address}
4608 to its corresponding physical address, and displays the result.
4611 @node Architecture and Core Commands
4612 @chapter Architecture and Core Commands
4613 @cindex Architecture Specific Commands
4614 @cindex Core Specific Commands
4616 Most CPUs have specialized JTAG operations to support debugging.
4617 OpenOCD packages most such operations in its standard command framework.
4618 Some of those operations don't fit well in that framework, so they are
4619 exposed here as architecture or implementation (core) specific commands.
4621 @anchor{ARM Hardware Tracing}
4622 @section ARM Hardware Tracing
4627 CPUs based on ARM cores may include standard tracing interfaces,
4628 based on an ``Embedded Trace Module'' (ETM) which sends voluminous
4629 address and data bus trace records to a ``Trace Port''.
4633 Development-oriented boards will sometimes provide a high speed
4634 trace connector for collecting that data, when the particular CPU
4635 supports such an interface.
4636 (The standard connector is a 38-pin Mictor, with both JTAG
4637 and trace port support.)
4638 Those trace connectors are supported by higher end JTAG adapters
4639 and some logic analyzer modules; frequently those modules can
4640 buffer several megabytes of trace data.
4641 Configuring an ETM coupled to such an external trace port belongs
4642 in the board-specific configuration file.
4644 If the CPU doesn't provide an external interface, it probably
4645 has an ``Embedded Trace Buffer'' (ETB) on the chip, which is a
4646 dedicated SRAM. 4KBytes is one common ETB size.
4647 Configuring an ETM coupled only to an ETB belongs in the CPU-specific
4648 (target) configuration file, since it works the same on all boards.
4651 ETM support in OpenOCD doesn't seem to be widely used yet.
4654 ETM support may be buggy, and at least some @command{etm config}
4655 parameters should be detected by asking the ETM for them.
4656 It seems like a GDB hookup should be possible,
4657 as well as triggering trace on specific events
4658 (perhaps @emph{handling IRQ 23} or @emph{calls foo()}).
4659 There should be GUI tools to manipulate saved trace data and help
4660 analyse it in conjunction with the source code.
4661 It's unclear how much of a common interface is shared
4662 with the current XScale trace support, or should be
4663 shared with eventual Nexus-style trace module support.
4666 @subsection ETM Configuration
4667 ETM setup is coupled with the trace port driver configuration.
4669 @deffn {Config Command} {etm config} target width mode clocking driver
4670 Declares the ETM associated with @var{target}, and associates it
4671 with a given trace port @var{driver}. @xref{Trace Port Drivers}.
4673 Several of the parameters must reflect the trace port configuration.
4674 The @var{width} must be either 4, 8, or 16.
4675 The @var{mode} must be @option{normal}, @option{multiplexted},
4676 or @option{demultiplexted}.
4677 The @var{clocking} must be @option{half} or @option{full}.
4680 You can see the ETM registers using the @command{reg} command, although
4681 not all of those possible registers are present in every ETM.
4685 @deffn Command {etm info}
4686 Displays information about the current target's ETM.
4689 @deffn Command {etm status}
4690 Displays status of the current target's ETM:
4691 is the ETM idle, or is it collecting data?
4692 Did trace data overflow?
4696 @deffn Command {etm tracemode} [type context_id_bits cycle_accurate branch_output]
4697 Displays what data that ETM will collect.
4698 If arguments are provided, first configures that data.
4699 When the configuration changes, tracing is stopped
4700 and any buffered trace data is invalidated.
4703 @item @var{type} ... one of
4704 @option{none} (save nothing),
4705 @option{data} (save data),
4706 @option{address} (save addresses),
4707 @option{all} (save data and addresses)
4708 @item @var{context_id_bits} ... 0, 8, 16, or 32
4709 @item @var{cycle_accurate} ... @option{enable} or @option{disable}
4710 @item @var{branch_output} ... @option{enable} or @option{disable}
4714 @deffn Command {etm trigger_percent} percent
4715 @emph{Buggy and effectively a NOP ... @var{percent} from 2..100}
4718 @subsection ETM Trace Operation
4720 After setting up the ETM, you can use it to collect data.
4721 That data can be exported to files for later analysis.
4722 It can also be parsed with OpenOCD, for basic sanity checking.
4724 @deffn Command {etm analyze}
4725 Reads trace data into memory, if it wasn't already present.
4726 Decodes and prints the data that was collected.
4729 @deffn Command {etm dump} filename
4730 Stores the captured trace data in @file{filename}.
4733 @deffn Command {etm image} filename [base_address] [type]
4734 Opens an image file.
4737 @deffn Command {etm load} filename
4738 Loads captured trace data from @file{filename}.
4741 @deffn Command {etm start}
4742 Starts trace data collection.
4745 @deffn Command {etm stop}
4746 Stops trace data collection.
4749 @anchor{Trace Port Drivers}
4750 @subsection Trace Port Drivers
4752 To use an ETM trace port it must be associated with a driver.
4754 @deffn {Trace Port Driver} dummy
4755 Use the @option{dummy} driver if you are configuring an ETM that's
4756 not connected to anything (on-chip ETB or off-chip trace connector).
4757 @emph{This driver lets OpenOCD talk to the ETM, but it does not expose
4758 any trace data collection.}
4759 @deffn {Config Command} {etm_dummy config} target
4760 Associates the ETM for @var{target} with a dummy driver.
4764 @deffn {Trace Port Driver} etb
4765 Use the @option{etb} driver if you are configuring an ETM
4766 to use on-chip ETB memory.
4767 @deffn {Config Command} {etb config} target etb_tap
4768 Associates the ETM for @var{target} with the ETB at @var{etb_tap}.
4769 You can see the ETB registers using the @command{reg} command.
4773 @deffn {Trace Port Driver} oocd_trace
4774 This driver isn't available unless OpenOCD was explicitly configured
4775 with the @option{--enable-oocd_trace} option. You probably don't want
4776 to configure it unless you've built the appropriate prototype hardware;
4777 it's @emph{proof-of-concept} software.
4779 Use the @option{oocd_trace} driver if you are configuring an ETM that's
4780 connected to an off-chip trace connector.
4782 @deffn {Config Command} {oocd_trace config} target tty
4783 Associates the ETM for @var{target} with a trace driver which
4784 collects data through the serial port @var{tty}.
4787 @deffn Command {oocd_trace resync}
4788 Re-synchronizes with the capture clock.
4791 @deffn Command {oocd_trace status}
4792 Reports whether the capture clock is locked or not.
4797 @section ARMv4 and ARMv5 Architecture
4801 These commands are specific to ARM architecture v4 and v5,
4802 including all ARM7 or ARM9 systems and Intel XScale.
4803 They are available in addition to other core-specific
4804 commands that may be available.
4806 @deffn Command {armv4_5 core_state} [@option{arm}|@option{thumb}]
4807 Displays the core_state, optionally changing it to process
4808 either @option{arm} or @option{thumb} instructions.
4809 The target may later be resumed in the currently set core_state.
4810 (Processors may also support the Jazelle state, but
4811 that is not currently supported in OpenOCD.)
4814 @deffn Command {armv4_5 disassemble} address [count [@option{thumb}]]
4816 Disassembles @var{count} instructions starting at @var{address}.
4817 If @var{count} is not specified, a single instruction is disassembled.
4818 If @option{thumb} is specified, or the low bit of the address is set,
4819 Thumb (16-bit) instructions are used;
4820 else ARM (32-bit) instructions are used.
4821 (Processors may also support the Jazelle state, but
4822 those instructions are not currently understood by OpenOCD.)
4825 @deffn Command {armv4_5 reg}
4826 Display a table of all banked core registers, fetching the current value from every
4827 core mode if necessary. OpenOCD versions before rev. 60 didn't fetch the current
4831 @subsection ARM7 and ARM9 specific commands
4835 These commands are specific to ARM7 and ARM9 cores, like ARM7TDMI, ARM720T,
4836 ARM9TDMI, ARM920T or ARM926EJ-S.
4837 They are available in addition to the ARMv4/5 commands,
4838 and any other core-specific commands that may be available.
4840 @deffn Command {arm7_9 dbgrq} (@option{enable}|@option{disable})
4841 Control use of the EmbeddedIce DBGRQ signal to force entry into debug mode,
4842 instead of breakpoints. This should be
4843 safe for all but ARM7TDMI--S cores (like Philips LPC).
4844 This feature is enabled by default on most ARM9 cores,
4845 including ARM9TDMI, ARM920T, and ARM926EJ-S.
4848 @deffn Command {arm7_9 dcc_downloads} (@option{enable}|@option{disable})
4850 Control the use of the debug communications channel (DCC) to write larger (>128 byte)
4851 amounts of memory. DCC downloads offer a huge speed increase, but might be
4852 unsafe, especially with targets running at very low speeds. This command was introduced
4853 with OpenOCD rev. 60, and requires a few bytes of working area.
4856 @anchor{arm7_9 fast_memory_access}
4857 @deffn Command {arm7_9 fast_memory_access} (@option{enable}|@option{disable})
4858 Enable or disable memory writes and reads that don't check completion of
4859 the operation. This provides a huge speed increase, especially with USB JTAG
4860 cables (FT2232), but might be unsafe if used with targets running at very low
4861 speeds, like the 32kHz startup clock of an AT91RM9200.
4864 @deffn {Debug Command} {arm7_9 write_core_reg} num mode word
4865 @emph{This is intended for use while debugging OpenOCD; you probably
4868 Writes a 32-bit @var{word} to register @var{num} (from 0 to 16)
4869 as used in the specified @var{mode}
4870 (where e.g. mode 16 is "user" and mode 19 is "supervisor";
4871 the M4..M0 bits of the PSR).
4872 Registers 0..15 are the normal CPU registers such as r0(0), r1(1) ... pc(15).
4873 Register 16 is the mode-specific SPSR,
4874 unless the specified mode is 0xffffffff (32-bit all-ones)
4875 in which case register 16 is the CPSR.
4876 The write goes directly to the CPU, bypassing the register cache.
4879 @deffn {Debug Command} {arm7_9 write_xpsr} word (@option{0}|@option{1})
4880 @emph{This is intended for use while debugging OpenOCD; you probably
4883 If the second parameter is zero, writes @var{word} to the
4884 Current Program Status register (CPSR).
4885 Else writes @var{word} to the current mode's Saved PSR (SPSR).
4886 In both cases, this bypasses the register cache.
4889 @deffn {Debug Command} {arm7_9 write_xpsr_im8} byte rotate (@option{0}|@option{1})
4890 @emph{This is intended for use while debugging OpenOCD; you probably
4893 Writes eight bits to the CPSR or SPSR,
4894 first rotating them by @math{2*rotate} bits,
4895 and bypassing the register cache.
4896 This has lower JTAG overhead than writing the entire CPSR or SPSR
4897 with @command{arm7_9 write_xpsr}.
4900 @subsection ARM720T specific commands
4903 These commands are available to ARM720T based CPUs,
4904 which are implementations of the ARMv4T architecture
4905 based on the ARM7TDMI-S integer core.
4906 They are available in addition to the ARMv4/5 and ARM7/ARM9 commands.
4908 @deffn Command {arm720t cp15} regnum [value]
4909 Display cp15 register @var{regnum};
4910 else if a @var{value} is provided, that value is written to that register.
4913 @deffn Command {arm720t mdw_phys} addr [count]
4914 @deffnx Command {arm720t mdh_phys} addr [count]
4915 @deffnx Command {arm720t mdb_phys} addr [count]
4916 Display contents of physical address @var{addr}, as
4917 32-bit words (@command{mdw_phys}), 16-bit halfwords (@command{mdh_phys}),
4918 or 8-bit bytes (@command{mdb_phys}).
4919 If @var{count} is specified, displays that many units.
4922 @deffn Command {arm720t mww_phys} addr word
4923 @deffnx Command {arm720t mwh_phys} addr halfword
4924 @deffnx Command {arm720t mwb_phys} addr byte
4925 Writes the specified @var{word} (32 bits),
4926 @var{halfword} (16 bits), or @var{byte} (8-bit) pattern,
4927 at the specified physical address @var{addr}.
4930 @deffn Command {arm720t virt2phys} va
4931 Translate a virtual address @var{va} to a physical address
4932 and display the result.
4935 @subsection ARM9 specific commands
4938 ARM9-family cores are built around ARM9TDMI or ARM9E (including ARM9EJS)
4940 Such cores include the ARM920T, ARM926EJ-S, and ARM966.
4942 For historical reasons, one command shared by these cores starts
4943 with the @command{arm9tdmi} prefix.
4944 This is true even for ARM9E based processors, which implement the
4945 ARMv5TE architecture instead of ARMv4T.
4947 @c 9-june-2009: tried this on arm920t, it didn't work.
4948 @c no-params always lists nothing caught, and that's how it acts.
4950 @anchor{arm9tdmi vector_catch}
4951 @deffn Command {arm9tdmi vector_catch} [@option{all}|@option{none}|list]
4952 @cindex vector_catch
4953 Vector Catch hardware provides a sort of dedicated breakpoint
4954 for hardware events such as reset, interrupt, and abort.
4955 You can use this to conserve normal breakpoint resources,
4956 so long as you're not concerned with code that branches directly
4957 to those hardware vectors.
4959 This always finishes by listing the current configuration.
4960 If parameters are provided, it first reconfigures the
4961 vector catch hardware to intercept
4962 @option{all} of the hardware vectors,
4963 @option{none} of them,
4964 or a list with one or more of the following:
4965 @option{reset} @option{undef} @option{swi} @option{pabt} @option{dabt} @option{reserved}
4966 @option{irq} @option{fiq}.
4969 @subsection ARM920T specific commands
4972 These commands are available to ARM920T based CPUs,
4973 which are implementations of the ARMv4T architecture
4974 built using the ARM9TDMI integer core.
4975 They are available in addition to the ARMv4/5, ARM7/ARM9,
4976 and ARM9TDMI commands.
4978 @deffn Command {arm920t cache_info}
4979 Print information about the caches found. This allows to see whether your target
4980 is an ARM920T (2x16kByte cache) or ARM922T (2x8kByte cache).
4983 @deffn Command {arm920t cp15} regnum [value]
4984 Display cp15 register @var{regnum};
4985 else if a @var{value} is provided, that value is written to that register.
4988 @deffn Command {arm920t cp15i} opcode [value [address]]
4989 Interpreted access using cp15 @var{opcode}.
4990 If no @var{value} is provided, the result is displayed.
4991 Else if that value is written using the specified @var{address},
4992 or using zero if no other address is not provided.
4995 @deffn Command {arm920t mdw_phys} addr [count]
4996 @deffnx Command {arm920t mdh_phys} addr [count]
4997 @deffnx Command {arm920t mdb_phys} addr [count]
4998 Display contents of physical address @var{addr}, as
4999 32-bit words (@command{mdw_phys}), 16-bit halfwords (@command{mdh_phys}),
5000 or 8-bit bytes (@command{mdb_phys}).
5001 If @var{count} is specified, displays that many units.
5004 @deffn Command {arm920t mww_phys} addr word
5005 @deffnx Command {arm920t mwh_phys} addr halfword
5006 @deffnx Command {arm920t mwb_phys} addr byte
5007 Writes the specified @var{word} (32 bits),
5008 @var{halfword} (16 bits), or @var{byte} (8-bit) pattern,
5009 at the specified physical address @var{addr}.
5012 @deffn Command {arm920t read_cache} filename
5013 Dump the content of ICache and DCache to a file named @file{filename}.
5016 @deffn Command {arm920t read_mmu} filename
5017 Dump the content of the ITLB and DTLB to a file named @file{filename}.
5020 @deffn Command {arm920t virt2phys} va
5021 Translate a virtual address @var{va} to a physical address
5022 and display the result.
5025 @subsection ARM926ej-s specific commands
5028 These commands are available to ARM926ej-s based CPUs,
5029 which are implementations of the ARMv5TEJ architecture
5030 based on the ARM9EJ-S integer core.
5031 They are available in addition to the ARMv4/5, ARM7/ARM9,
5032 and ARM9TDMI commands.
5034 The Feroceon cores also support these commands, although
5035 they are not built from ARM926ej-s designs.
5037 @deffn Command {arm926ejs cache_info}
5038 Print information about the caches found.
5041 @deffn Command {arm926ejs cp15} opcode1 opcode2 CRn CRm regnum [value]
5042 Accesses cp15 register @var{regnum} using
5043 @var{opcode1}, @var{opcode2}, @var{CRn}, and @var{CRm}.
5044 If a @var{value} is provided, that value is written to that register.
5045 Else that register is read and displayed.
5048 @deffn Command {arm926ejs mdw_phys} addr [count]
5049 @deffnx Command {arm926ejs mdh_phys} addr [count]
5050 @deffnx Command {arm926ejs mdb_phys} addr [count]
5051 Display contents of physical address @var{addr}, as
5052 32-bit words (@command{mdw_phys}), 16-bit halfwords (@command{mdh_phys}),
5053 or 8-bit bytes (@command{mdb_phys}).
5054 If @var{count} is specified, displays that many units.
5057 @deffn Command {arm926ejs mww_phys} addr word
5058 @deffnx Command {arm926ejs mwh_phys} addr halfword
5059 @deffnx Command {arm926ejs mwb_phys} addr byte
5060 Writes the specified @var{word} (32 bits),
5061 @var{halfword} (16 bits), or @var{byte} (8-bit) pattern,
5062 at the specified physical address @var{addr}.
5065 @deffn Command {arm926ejs virt2phys} va
5066 Translate a virtual address @var{va} to a physical address
5067 and display the result.
5070 @subsection ARM966E specific commands
5073 These commands are available to ARM966 based CPUs,
5074 which are implementations of the ARMv5TE architecture.
5075 They are available in addition to the ARMv4/5, ARM7/ARM9,
5076 and ARM9TDMI commands.
5078 @deffn Command {arm966e cp15} regnum [value]
5079 Display cp15 register @var{regnum};
5080 else if a @var{value} is provided, that value is written to that register.
5083 @subsection XScale specific commands
5086 Some notes about the debug implementation on the XScale CPUs:
5088 The XScale CPU provides a special debug-only mini-instruction cache
5089 (mini-IC) in which exception vectors and target-resident debug handler
5090 code are placed by OpenOCD. In order to get access to the CPU, OpenOCD
5091 must point vector 0 (the reset vector) to the entry of the debug
5092 handler. However, this means that the complete first cacheline in the
5093 mini-IC is marked valid, which makes the CPU fetch all exception
5094 handlers from the mini-IC, ignoring the code in RAM.
5096 OpenOCD currently does not sync the mini-IC entries with the RAM
5097 contents (which would fail anyway while the target is running), so
5098 the user must provide appropriate values using the @code{xscale
5099 vector_table} command.
5101 It is recommended to place a pc-relative indirect branch in the vector
5102 table, and put the branch destination somewhere in memory. Doing so
5103 makes sure the code in the vector table stays constant regardless of
5104 code layout in memory:
5107 ldr pc,[pc,#0x100-8]
5108 ldr pc,[pc,#0x100-8]
5109 ldr pc,[pc,#0x100-8]
5110 ldr pc,[pc,#0x100-8]
5111 ldr pc,[pc,#0x100-8]
5112 ldr pc,[pc,#0x100-8]
5113 ldr pc,[pc,#0x100-8]
5114 ldr pc,[pc,#0x100-8]
5116 .long real_reset_vector
5117 .long real_ui_handler
5118 .long real_swi_handler
5120 .long real_data_abort
5121 .long 0 /* unused */
5122 .long real_irq_handler
5123 .long real_fiq_handler
5126 The debug handler must be placed somewhere in the address space using
5127 the @code{xscale debug_handler} command. The allowed locations for the
5128 debug handler are either (0x800 - 0x1fef800) or (0xfe000800 -
5129 0xfffff800). The default value is 0xfe000800.
5132 These commands are available to XScale based CPUs,
5133 which are implementations of the ARMv5TE architecture.
5135 @deffn Command {xscale analyze_trace}
5136 Displays the contents of the trace buffer.
5139 @deffn Command {xscale cache_clean_address} address
5140 Changes the address used when cleaning the data cache.
5143 @deffn Command {xscale cache_info}
5144 Displays information about the CPU caches.
5147 @deffn Command {xscale cp15} regnum [value]
5148 Display cp15 register @var{regnum};
5149 else if a @var{value} is provided, that value is written to that register.
5152 @deffn Command {xscale debug_handler} target address
5153 Changes the address used for the specified target's debug handler.
5156 @deffn Command {xscale dcache} (@option{enable}|@option{disable})
5157 Enables or disable the CPU's data cache.
5160 @deffn Command {xscale dump_trace} filename
5161 Dumps the raw contents of the trace buffer to @file{filename}.
5164 @deffn Command {xscale icache} (@option{enable}|@option{disable})
5165 Enables or disable the CPU's instruction cache.
5168 @deffn Command {xscale mmu} (@option{enable}|@option{disable})
5169 Enables or disable the CPU's memory management unit.
5172 @deffn Command {xscale trace_buffer} (@option{enable}|@option{disable}) [@option{fill} [n] | @option{wrap}]
5173 Enables or disables the trace buffer,
5174 and controls how it is emptied.
5177 @deffn Command {xscale trace_image} filename [offset [type]]
5178 Opens a trace image from @file{filename}, optionally rebasing
5179 its segment addresses by @var{offset}.
5180 The image @var{type} may be one of
5181 @option{bin} (binary), @option{ihex} (Intel hex),
5182 @option{elf} (ELF file), @option{s19} (Motorola s19),
5183 @option{mem}, or @option{builder}.
5186 @anchor{xscale vector_catch}
5187 @deffn Command {xscale vector_catch} [mask]
5188 @cindex vector_catch
5189 Display a bitmask showing the hardware vectors to catch.
5190 If the optional parameter is provided, first set the bitmask to that value.
5192 The mask bits correspond with bit 16..23 in the DCSR:
5195 0x02 Trap Undefined Instructions
5196 0x04 Trap Software Interrupt
5197 0x08 Trap Prefetch Abort
5198 0x10 Trap Data Abort
5205 @anchor{xscale vector_table}
5206 @deffn Command {xscale vector_table} [<low|high> <index> <value>]
5207 @cindex vector_table
5209 Set an entry in the mini-IC vector table. There are two tables: one for
5210 low vectors (at 0x00000000), and one for high vectors (0xFFFF0000), each
5211 holding the 8 exception vectors. @var{index} can be 1-7, because vector 0
5212 points to the debug handler entry and can not be overwritten.
5213 @var{value} holds the 32-bit opcode that is placed in the mini-IC.
5215 Without arguments, the current settings are displayed.
5219 @section ARMv6 Architecture
5222 @subsection ARM11 specific commands
5225 @deffn Command {arm11 mcr} pX opc1 CRn CRm opc2 value
5226 Write @var{value} to a coprocessor @var{pX} register
5227 passing parameters @var{CRn},
5228 @var{CRm}, opcodes @var{opc1} and @var{opc2},
5229 and the MCR instruction.
5230 (The difference beween this and the MCR2 instruction is
5231 one bit in the encoding, effecively a fifth parameter.)
5234 @deffn Command {arm11 memwrite burst} [value]
5235 Displays the value of the memwrite burst-enable flag,
5236 which is enabled by default.
5237 If @var{value} is defined, first assigns that.
5240 @deffn Command {arm11 memwrite error_fatal} [value]
5241 Displays the value of the memwrite error_fatal flag,
5242 which is enabled by default.
5243 If @var{value} is defined, first assigns that.
5246 @deffn Command {arm11 mrc} pX opc1 CRn CRm opc2
5247 Read a coprocessor @var{pX} register passing parameters @var{CRn},
5248 @var{CRm}, opcodes @var{opc1} and @var{opc2},
5249 and the MRC instruction.
5250 (The difference beween this and the MRC2 instruction is
5251 one bit in the encoding, effecively a fifth parameter.)
5252 Displays the result.
5255 @deffn Command {arm11 no_increment} [value]
5256 Displays the value of the flag controlling whether
5257 some read or write operations increment the pointer
5258 (the default behavior) or not (acting like a FIFO).
5259 If @var{value} is defined, first assigns that.
5262 @deffn Command {arm11 step_irq_enable} [value]
5263 Displays the value of the flag controlling whether
5264 IRQs are enabled during single stepping;
5265 they is disabled by default.
5266 If @var{value} is defined, first assigns that.
5269 @section ARMv7 Architecture
5272 @subsection ARMv7 Debug Access Port (DAP) specific commands
5273 @cindex Debug Access Port
5275 These commands are specific to ARM architecture v7 Debug Access Port (DAP),
5276 included on cortex-m3 and cortex-a8 systems.
5277 They are available in addition to other core-specific commands that may be available.
5279 @deffn Command {dap info} [num]
5280 Displays dap info for ap @var{num}, defaulting to the currently selected AP.
5283 @deffn Command {dap apsel} [num]
5284 Select AP @var{num}, defaulting to 0.
5287 @deffn Command {dap apid} [num]
5288 Displays id register from AP @var{num},
5289 defaulting to the currently selected AP.
5292 @deffn Command {dap baseaddr} [num]
5293 Displays debug base address from AP @var{num},
5294 defaulting to the currently selected AP.
5297 @deffn Command {dap memaccess} [value]
5298 Displays the number of extra tck for mem-ap memory bus access [0-255].
5299 If @var{value} is defined, first assigns that.
5302 @subsection ARMv7-A specific commands
5305 @deffn Command {armv7a disassemble} address [count [@option{thumb}]]
5307 Disassembles @var{count} instructions starting at @var{address}.
5308 If @var{count} is not specified, a single instruction is disassembled.
5309 If @option{thumb} is specified, or the low bit of the address is set,
5310 Thumb2 (mixed 16/32-bit) instructions are used;
5311 else ARM (32-bit) instructions are used.
5312 With a handful of exceptions, ThumbEE instructions are the same as Thumb2;
5313 ThumbEE disassembly currently has no explicit support.
5314 (Processors may also support the Jazelle state, but
5315 those instructions are not currently understood by OpenOCD.)
5319 @subsection Cortex-M3 specific commands
5322 @deffn Command {cortex_m3 disassemble} address [count]
5324 Disassembles @var{count} Thumb2 instructions starting at @var{address}.
5325 If @var{count} is not specified, a single instruction is disassembled.
5328 @deffn Command {cortex_m3 maskisr} (@option{on}|@option{off})
5329 Control masking (disabling) interrupts during target step/resume.
5332 @deffn Command {cortex_m3 vector_catch} [@option{all}|@option{none}|list]
5333 @cindex vector_catch
5334 Vector Catch hardware provides dedicated breakpoints
5335 for certain hardware events.
5337 Parameters request interception of
5338 @option{all} of these hardware event vectors,
5339 @option{none} of them,
5340 or one or more of the following:
5341 @option{hard_err} for a HardFault exception;
5342 @option{mm_err} for a MemManage exception;
5343 @option{bus_err} for a BusFault exception;
5346 @option{chk_err}, or
5347 @option{nocp_err} for various UsageFault exceptions; or
5349 If NVIC setup code does not enable them,
5350 MemManage, BusFault, and UsageFault exceptions
5351 are mapped to HardFault.
5352 UsageFault checks for
5353 divide-by-zero and unaligned access
5354 must also be explicitly enabled.
5356 This finishes by listing the current vector catch configuration.
5359 @anchor{Software Debug Messages and Tracing}
5360 @section Software Debug Messages and Tracing
5361 @cindex Linux-ARM DCC support
5365 OpenOCD can process certain requests from target software. Currently
5366 @command{target_request debugmsgs}
5367 is supported only for @option{arm7_9} and @option{cortex_m3} cores.
5368 These messages are received as part of target polling, so
5369 you need to have @command{poll on} active to receive them.
5370 They are intrusive in that they will affect program execution
5371 times. If that is a problem, @pxref{ARM Hardware Tracing}.
5373 See @file{libdcc} in the contrib dir for more details.
5374 In addition to sending strings, characters, and
5375 arrays of various size integers from the target,
5376 @file{libdcc} also exports a software trace point mechanism.
5377 The target being debugged may
5378 issue trace messages which include a 24-bit @dfn{trace point} number.
5379 Trace point support includes two distinct mechanisms,
5380 each supported by a command:
5383 @item @emph{History} ... A circular buffer of trace points
5384 can be set up, and then displayed at any time.
5385 This tracks where code has been, which can be invaluable in
5386 finding out how some fault was triggered.
5388 The buffer may overflow, since it collects records continuously.
5389 It may be useful to use some of the 24 bits to represent a
5390 particular event, and other bits to hold data.
5392 @item @emph{Counting} ... An array of counters can be set up,
5393 and then displayed at any time.
5394 This can help establish code coverage and identify hot spots.
5396 The array of counters is directly indexed by the trace point
5397 number, so trace points with higher numbers are not counted.
5400 Linux-ARM kernels have a ``Kernel low-level debugging
5401 via EmbeddedICE DCC channel'' option (CONFIG_DEBUG_ICEDCC,
5402 depends on CONFIG_DEBUG_LL) which uses this mechanism to
5403 deliver messages before a serial console can be activated.
5404 This is not the same format used by @file{libdcc}.
5405 Other software, such as the U-Boot boot loader, sometimes
5406 does the same thing.
5408 @deffn Command {target_request debugmsgs} [@option{enable}|@option{disable}|@option{charmsg}]
5409 Displays current handling of target DCC message requests.
5410 These messages may be sent to the debugger while the target is running.
5411 The optional @option{enable} and @option{charmsg} parameters
5412 both enable the messages, while @option{disable} disables them.
5414 With @option{charmsg} the DCC words each contain one character,
5415 as used by Linux with CONFIG_DEBUG_ICEDCC;
5416 otherwise the libdcc format is used.
5419 @deffn Command {trace history} (@option{clear}|count)
5420 With no parameter, displays all the trace points that have triggered
5421 in the order they triggered.
5422 With the parameter @option{clear}, erases all current trace history records.
5423 With a @var{count} parameter, allocates space for that many
5427 @deffn Command {trace point} (@option{clear}|identifier)
5428 With no parameter, displays all trace point identifiers and how many times
5429 they have been triggered.
5430 With the parameter @option{clear}, erases all current trace point counters.
5431 With a numeric @var{identifier} parameter, creates a new a trace point counter
5432 and associates it with that identifier.
5434 @emph{Important:} The identifier and the trace point number
5435 are not related except by this command.
5436 These trace point numbers always start at zero (from server startup,
5437 or after @command{trace point clear}) and count up from there.
5442 @chapter JTAG Commands
5443 @cindex JTAG Commands
5444 Most general purpose JTAG commands have been presented earlier.
5445 (@xref{JTAG Speed}, @ref{Reset Configuration}, and @ref{TAP Declaration}.)
5446 Lower level JTAG commands, as presented here,
5447 may be needed to work with targets which require special
5448 attention during operations such as reset or initialization.
5450 To use these commands you will need to understand some
5451 of the basics of JTAG, including:
5454 @item A JTAG scan chain consists of a sequence of individual TAP
5455 devices such as a CPUs.
5456 @item Control operations involve moving each TAP through the same
5457 standard state machine (in parallel)
5458 using their shared TMS and clock signals.
5459 @item Data transfer involves shifting data through the chain of
5460 instruction or data registers of each TAP, writing new register values
5461 while the reading previous ones.
5462 @item Data register sizes are a function of the instruction active in
5463 a given TAP, while instruction register sizes are fixed for each TAP.
5464 All TAPs support a BYPASS instruction with a single bit data register.
5465 @item The way OpenOCD differentiates between TAP devices is by
5466 shifting different instructions into (and out of) their instruction
5470 @section Low Level JTAG Commands
5472 These commands are used by developers who need to access
5473 JTAG instruction or data registers, possibly controlling
5474 the order of TAP state transitions.
5475 If you're not debugging OpenOCD internals, or bringing up a
5476 new JTAG adapter or a new type of TAP device (like a CPU or
5477 JTAG router), you probably won't need to use these commands.
5479 @deffn Command {drscan} tap [numbits value]+ [@option{-endstate} tap_state]
5480 Loads the data register of @var{tap} with a series of bit fields
5481 that specify the entire register.
5482 Each field is @var{numbits} bits long with
5483 a numeric @var{value} (hexadecimal encouraged).
5484 The return value holds the original value of each
5487 For example, a 38 bit number might be specified as one
5488 field of 32 bits then one of 6 bits.
5489 @emph{For portability, never pass fields which are more
5490 than 32 bits long. Many OpenOCD implementations do not
5491 support 64-bit (or larger) integer values.}
5493 All TAPs other than @var{tap} must be in BYPASS mode.
5494 The single bit in their data registers does not matter.
5496 When @var{tap_state} is specified, the JTAG state machine is left
5498 For example @sc{drpause} might be specified, so that more
5499 instructions can be issued before re-entering the @sc{run/idle} state.
5500 If the end state is not specified, the @sc{run/idle} state is entered.
5503 OpenOCD does not record information about data register lengths,
5504 so @emph{it is important that you get the bit field lengths right}.
5505 Remember that different JTAG instructions refer to different
5506 data registers, which may have different lengths.
5507 Moreover, those lengths may not be fixed;
5508 the SCAN_N instruction can change the length of
5509 the register accessed by the INTEST instruction
5510 (by connecting a different scan chain).
5514 @deffn Command {flush_count}
5515 Returns the number of times the JTAG queue has been flushed.
5516 This may be used for performance tuning.
5518 For example, flushing a queue over USB involves a
5519 minimum latency, often several milliseconds, which does
5520 not change with the amount of data which is written.
5521 You may be able to identify performance problems by finding
5522 tasks which waste bandwidth by flushing small transfers too often,
5523 instead of batching them into larger operations.
5526 @deffn Command {irscan} [tap instruction]+ [@option{-endstate} tap_state]
5527 For each @var{tap} listed, loads the instruction register
5528 with its associated numeric @var{instruction}.
5529 (The number of bits in that instruction may be displayed
5530 using the @command{scan_chain} command.)
5531 For other TAPs, a BYPASS instruction is loaded.
5533 When @var{tap_state} is specified, the JTAG state machine is left
5535 For example @sc{irpause} might be specified, so the data register
5536 can be loaded before re-entering the @sc{run/idle} state.
5537 If the end state is not specified, the @sc{run/idle} state is entered.
5540 OpenOCD currently supports only a single field for instruction
5541 register values, unlike data register values.
5542 For TAPs where the instruction register length is more than 32 bits,
5543 portable scripts currently must issue only BYPASS instructions.
5547 @deffn Command {jtag_reset} trst srst
5548 Set values of reset signals.
5549 The @var{trst} and @var{srst} parameter values may be
5550 @option{0}, indicating that reset is inactive (pulled or driven high),
5551 or @option{1}, indicating it is active (pulled or driven low).
5552 The @command{reset_config} command should already have been used
5553 to configure how the board and JTAG adapter treat these two
5554 signals, and to say if either signal is even present.
5555 @xref{Reset Configuration}.
5558 @deffn Command {runtest} @var{num_cycles}
5559 Move to the @sc{run/idle} state, and execute at least
5560 @var{num_cycles} of the JTAG clock (TCK).
5561 Instructions often need some time
5562 to execute before they take effect.
5565 @c tms_sequence (short|long)
5566 @c ... temporary, debug-only, probably gone before 0.2 ships
5568 @deffn Command {verify_ircapture} (@option{enable}|@option{disable})
5569 Verify values captured during @sc{ircapture} and returned
5570 during IR scans. Default is enabled, but this can be
5571 overridden by @command{verify_jtag}.
5574 @deffn Command {verify_jtag} (@option{enable}|@option{disable})
5575 Enables verification of DR and IR scans, to help detect
5576 programming errors. For IR scans, @command{verify_ircapture}
5577 must also be enabled.
5581 @section TAP state names
5582 @cindex TAP state names
5584 The @var{tap_state} names used by OpenOCD in the @command{drscan},
5585 and @command{irscan} commands are:
5588 @item @b{RESET} ... should act as if TRST were active
5589 @item @b{RUN/IDLE} ... don't assume this always means IDLE
5592 @item @b{DRSHIFT} ... TDI/TDO shifting through the data register
5594 @item @b{DRPAUSE} ... data register ready for update or more shifting
5599 @item @b{IRSHIFT} ... TDI/TDO shifting through the instruction register
5601 @item @b{IRPAUSE} ... instruction register ready for update or more shifting
5606 Note that only six of those states are fully ``stable'' in the
5607 face of TMS fixed (low except for @sc{reset})
5608 and a free-running JTAG clock. For all the
5609 others, the next TCK transition changes to a new state.
5612 @item From @sc{drshift} and @sc{irshift}, clock transitions will
5613 produce side effects by changing register contents. The values
5614 to be latched in upcoming @sc{drupdate} or @sc{irupdate} states
5615 may not be as expected.
5616 @item @sc{run/idle}, @sc{drpause}, and @sc{irpause} are reasonable
5617 choices after @command{drscan} or @command{irscan} commands,
5618 since they are free of JTAG side effects.
5619 However, @sc{run/idle} may have side effects that appear at other
5620 levels, such as advancing the ARM9E-S instruction pipeline.
5621 Consult the documentation for the TAP(s) you are working with.
5624 @node Boundary Scan Commands
5625 @chapter Boundary Scan Commands
5627 One of the original purposes of JTAG was to support
5628 boundary scan based hardware testing.
5629 Although its primary focus is to support On-Chip Debugging,
5630 OpenOCD also includes some boundary scan commands.
5632 @section SVF: Serial Vector Format
5633 @cindex Serial Vector Format
5636 The Serial Vector Format, better known as @dfn{SVF}, is a
5637 way to represent JTAG test patterns in text files.
5638 OpenOCD supports running such test files.
5640 @deffn Command {svf} filename [@option{quiet}]
5641 This issues a JTAG reset (Test-Logic-Reset) and then
5642 runs the SVF script from @file{filename}.
5643 Unless the @option{quiet} option is specified,
5644 each command is logged before it is executed.
5647 @section XSVF: Xilinx Serial Vector Format
5648 @cindex Xilinx Serial Vector Format
5651 The Xilinx Serial Vector Format, better known as @dfn{XSVF}, is a
5652 binary representation of SVF which is optimized for use with
5654 OpenOCD supports running such test files.
5656 @quotation Important
5657 Not all XSVF commands are supported.
5660 @deffn Command {xsvf} (tapname|@option{plain}) filename [@option{virt2}] [@option{quiet}]
5661 This issues a JTAG reset (Test-Logic-Reset) and then
5662 runs the XSVF script from @file{filename}.
5663 When a @var{tapname} is specified, the commands are directed at
5665 When @option{virt2} is specified, the @sc{xruntest} command counts
5666 are interpreted as TCK cycles instead of microseconds.
5667 Unless the @option{quiet} option is specified,
5668 messages are logged for comments and some retries.
5674 If OpenOCD runs on an embedded host(as ZY1000 does), then TFTP can
5675 be used to access files on PCs (either the developer's PC or some other PC).
5677 The way this works on the ZY1000 is to prefix a filename by
5678 "/tftp/ip/" and append the TFTP path on the TFTP
5679 server (tftpd). For example,
5682 load_image /tftp/10.0.0.96/c:\temp\abc.elf
5685 will load c:\temp\abc.elf from the developer pc (10.0.0.96) into memory as
5686 if the file was hosted on the embedded host.
5688 In order to achieve decent performance, you must choose a TFTP server
5689 that supports a packet size bigger than the default packet size (512 bytes). There
5690 are numerous TFTP servers out there (free and commercial) and you will have to do
5691 a bit of googling to find something that fits your requirements.
5693 @node GDB and OpenOCD
5694 @chapter GDB and OpenOCD
5696 OpenOCD complies with the remote gdbserver protocol, and as such can be used
5697 to debug remote targets.
5699 @anchor{Connecting to GDB}
5700 @section Connecting to GDB
5701 @cindex Connecting to GDB
5702 Use GDB 6.7 or newer with OpenOCD if you run into trouble. For
5703 instance GDB 6.3 has a known bug that produces bogus memory access
5704 errors, which has since been fixed: look up 1836 in
5705 @url{http://sourceware.org/cgi-bin/gnatsweb.pl?database=gdb}
5707 OpenOCD can communicate with GDB in two ways:
5711 A socket (TCP/IP) connection is typically started as follows:
5713 target remote localhost:3333
5715 This would cause GDB to connect to the gdbserver on the local pc using port 3333.
5717 A pipe connection is typically started as follows:
5719 target remote | openocd --pipe
5721 This would cause GDB to run OpenOCD and communicate using pipes (stdin/stdout).
5722 Using this method has the advantage of GDB starting/stopping OpenOCD for the debug
5726 To list the available OpenOCD commands type @command{monitor help} on the
5729 OpenOCD supports the gdb @option{qSupported} packet, this enables information
5730 to be sent by the GDB remote server (i.e. OpenOCD) to GDB. Typical information includes
5731 packet size and the device's memory map.
5733 Previous versions of OpenOCD required the following GDB options to increase
5734 the packet size and speed up GDB communication:
5736 set remote memory-write-packet-size 1024
5737 set remote memory-write-packet-size fixed
5738 set remote memory-read-packet-size 1024
5739 set remote memory-read-packet-size fixed
5741 This is now handled in the @option{qSupported} PacketSize and should not be required.
5743 @section Programming using GDB
5744 @cindex Programming using GDB
5746 By default the target memory map is sent to GDB. This can be disabled by
5747 the following OpenOCD configuration option:
5749 gdb_memory_map disable
5751 For this to function correctly a valid flash configuration must also be set
5752 in OpenOCD. For faster performance you should also configure a valid
5755 Informing GDB of the memory map of the target will enable GDB to protect any
5756 flash areas of the target and use hardware breakpoints by default. This means
5757 that the OpenOCD option @command{gdb_breakpoint_override} is not required when
5758 using a memory map. @xref{gdb_breakpoint_override}.
5760 To view the configured memory map in GDB, use the GDB command @option{info mem}
5761 All other unassigned addresses within GDB are treated as RAM.
5763 GDB 6.8 and higher set any memory area not in the memory map as inaccessible.
5764 This can be changed to the old behaviour by using the following GDB command
5766 set mem inaccessible-by-default off
5769 If @command{gdb_flash_program enable} is also used, GDB will be able to
5770 program any flash memory using the vFlash interface.
5772 GDB will look at the target memory map when a load command is given, if any
5773 areas to be programmed lie within the target flash area the vFlash packets
5776 If the target needs configuring before GDB programming, an event
5777 script can be executed:
5779 $_TARGETNAME configure -event EVENTNAME BODY
5782 To verify any flash programming the GDB command @option{compare-sections}
5785 @node Tcl Scripting API
5786 @chapter Tcl Scripting API
5787 @cindex Tcl Scripting API
5791 The commands are stateless. E.g. the telnet command line has a concept
5792 of currently active target, the Tcl API proc's take this sort of state
5793 information as an argument to each proc.
5795 There are three main types of return values: single value, name value
5796 pair list and lists.
5798 Name value pair. The proc 'foo' below returns a name/value pair
5804 > set foo(you) Oyvind
5805 > set foo(mouse) Micky
5806 > set foo(duck) Donald
5814 me Duane you Oyvind mouse Micky duck Donald
5816 Thus, to get the names of the associative array is easy:
5818 foreach { name value } [set foo] {
5819 puts "Name: $name, Value: $value"
5823 Lists returned must be relatively small. Otherwise a range
5824 should be passed in to the proc in question.
5826 @section Internal low-level Commands
5828 By low-level, the intent is a human would not directly use these commands.
5830 Low-level commands are (should be) prefixed with "ocd_", e.g.
5831 @command{ocd_flash_banks}
5832 is the low level API upon which @command{flash banks} is implemented.
5835 @item @b{ocd_mem2array} <@var{varname}> <@var{width}> <@var{addr}> <@var{nelems}>
5837 Read memory and return as a Tcl array for script processing
5838 @item @b{ocd_array2mem} <@var{varname}> <@var{width}> <@var{addr}> <@var{nelems}>
5840 Convert a Tcl array to memory locations and write the values
5841 @item @b{ocd_flash_banks} <@var{driver}> <@var{base}> <@var{size}> <@var{chip_width}> <@var{bus_width}> <@var{target}> [@option{driver options} ...]
5843 Return information about the flash banks
5846 OpenOCD commands can consist of two words, e.g. "flash banks". The
5847 @file{startup.tcl} "unknown" proc will translate this into a Tcl proc
5848 called "flash_banks".
5850 @section OpenOCD specific Global Variables
5854 Real Tcl has ::tcl_platform(), and platform::identify, and many other
5855 variables. JimTCL, as implemented in OpenOCD creates $HostOS which
5856 holds one of the following values:
5859 @item @b{winxx} Built using Microsoft Visual Studio
5860 @item @b{linux} Linux is the underlying operating sytem
5861 @item @b{darwin} Darwin (mac-os) is the underlying operating sytem.
5862 @item @b{cygwin} Running under Cygwin
5863 @item @b{mingw32} Running under MingW32
5864 @item @b{other} Unknown, none of the above.
5867 Note: 'winxx' was choosen because today (March-2009) no distinction is made between Win32 and Win64.
5870 We should add support for a variable like Tcl variable
5871 @code{tcl_platform(platform)}, it should be called
5872 @code{jim_platform} (because it
5873 is jim, not real tcl).
5877 @chapter Deprecated/Removed Commands
5878 @cindex Deprecated/Removed Commands
5879 Certain OpenOCD commands have been deprecated or
5880 removed during the various revisions.
5882 Upgrade your scripts as soon as possible.
5883 These descriptions for old commands may be removed
5884 a year after the command itself was removed.
5885 This means that in January 2010 this chapter may
5886 become much shorter.
5889 @item @b{arm7_9 fast_writes}
5890 @cindex arm7_9 fast_writes
5891 @*Use @command{arm7_9 fast_memory_access} instead.
5892 @xref{arm7_9 fast_memory_access}.
5895 @*An buggy old command that would not really work since background polling would wipe out the global endstate
5896 @item @b{arm7_9 force_hw_bkpts}
5897 @*Use @command{gdb_breakpoint_override} instead. Note that GDB will use hardware breakpoints
5898 for flash if the GDB memory map has been set up(default when flash is declared in
5899 target configuration). @xref{gdb_breakpoint_override}.
5900 @item @b{arm7_9 sw_bkpts}
5901 @*On by default. @xref{gdb_breakpoint_override}.
5902 @item @b{daemon_startup}
5903 @*this config option has been removed, simply adding @option{init} and @option{reset halt} to
5904 the end of your config script will give the same behaviour as using @option{daemon_startup reset}
5905 and @option{target cortex_m3 little reset_halt 0}.
5906 @item @b{dump_binary}
5907 @*use @option{dump_image} command with same args. @xref{dump_image}.
5908 @item @b{flash erase}
5909 @*use @option{flash erase_sector} command with same args. @xref{flash erase_sector}.
5910 @item @b{flash write}
5911 @*use @option{flash write_bank} command with same args. @xref{flash write_bank}.
5912 @item @b{flash write_binary}
5913 @*use @option{flash write_bank} command with same args. @xref{flash write_bank}.
5914 @item @b{flash auto_erase}
5915 @*use @option{flash write_image} command passing @option{erase} as the first parameter. @xref{flash write_image}.
5917 @item @b{jtag_device}
5918 @*use the @command{jtag newtap} command, converting from positional syntax
5919 to named prefixes, and naming the TAP.
5921 Note that if you try to use the old command, a message will tell you the
5922 right new command to use; and that the fourth parameter in the old syntax
5923 was never actually used.
5925 OLD: jtag_device 8 0x01 0xe3 0xfe
5926 NEW: jtag newtap CHIPNAME TAPNAME \
5927 -irlen 8 -ircapture 0x01 -irmask 0xe3
5930 @item @b{jtag_speed} value
5931 @*@xref{JTAG Speed}.
5932 Usually, a value of zero means maximum
5933 speed. The actual effect of this option depends on the JTAG interface used.
5935 @item wiggler: maximum speed / @var{number}
5936 @item ft2232: 6MHz / (@var{number}+1)
5937 @item amt jtagaccel: 8 / 2**@var{number}
5938 @item jlink: maximum speed in kHz (0-12000), 0 will use RTCK
5939 @item rlink: 24MHz / @var{number}, but only for certain values of @var{number}
5940 @comment end speed list.
5943 @item @b{load_binary}
5944 @*use @option{load_image} command with same args. @xref{load_image}.
5945 @item @b{run_and_halt_time}
5946 @*This command has been removed for simpler reset behaviour, it can be simulated with the
5953 @item @b{target} <@var{type}> <@var{endian}> <@var{jtag-position}>
5954 @*use the create subcommand of @option{target}.
5955 @item @b{target_script} <@var{target#}> <@var{eventname}> <@var{scriptname}>
5956 @*use <@var{target_name}> configure -event <@var{eventname}> "script <@var{scriptname}>"
5957 @item @b{working_area}
5958 @*use the @option{configure} subcommand of @option{target} to set the work-area-virt, work-area-phy, work-area-size, and work-area-backup properties of the target.
5966 @item @b{RTCK, also known as: Adaptive Clocking - What is it?}
5968 @cindex adaptive clocking
5971 In digital circuit design it is often refered to as ``clock
5972 synchronisation'' the JTAG interface uses one clock (TCK or TCLK)
5973 operating at some speed, your target is operating at another. The two
5974 clocks are not synchronised, they are ``asynchronous''
5976 In order for the two to work together they must be synchronised. Otherwise
5977 the two systems will get out of sync with each other and nothing will
5978 work. There are 2 basic options:
5981 Use a special circuit.
5983 One clock must be some multiple slower than the other.
5986 @b{Does this really matter?} For some chips and some situations, this
5987 is a non-issue (i.e.: A 500MHz ARM926) but for others - for example some
5988 Atmel SAM7 and SAM9 chips start operation from reset at 32kHz -
5989 program/enable the oscillators and eventually the main clock. It is in
5990 those critical times you must slow the JTAG clock to sometimes 1 to
5993 Imagine debugging a 500MHz ARM926 hand held battery powered device
5994 that ``deep sleeps'' at 32kHz between every keystroke. It can be
5997 @b{Solution #1 - A special circuit}
5999 In order to make use of this, your JTAG dongle must support the RTCK
6000 feature. Not all dongles support this - keep reading!
6002 The RTCK signal often found in some ARM chips is used to help with
6003 this problem. ARM has a good description of the problem described at
6004 this link: @url{http://www.arm.com/support/faqdev/4170.html} [checked
6005 28/nov/2008]. Link title: ``How does the JTAG synchronisation logic
6006 work? / how does adaptive clocking work?''.
6008 The nice thing about adaptive clocking is that ``battery powered hand
6009 held device example'' - the adaptiveness works perfectly all the
6010 time. One can set a break point or halt the system in the deep power
6011 down code, slow step out until the system speeds up.
6013 Note that adaptive clocking may also need to work at the board level,
6014 when a board-level scan chain has multiple chips.
6015 Parallel clock voting schemes are good way to implement this,
6016 both within and between chips, and can easily be implemented
6018 It's not difficult to have logic fan a module's input TCK signal out
6019 to each TAP in the scan chain, and then wait until each TAP's RTCK comes
6020 back with the right polarity before changing the output RTCK signal.
6021 Texas Instruments makes some clock voting logic available
6022 for free (with no support) in VHDL form; see
6023 @url{http://tiexpressdsp.com/index.php/Adaptive_Clocking}
6025 @b{Solution #2 - Always works - but may be slower}
6027 Often this is a perfectly acceptable solution.
6029 In most simple terms: Often the JTAG clock must be 1/10 to 1/12 of
6030 the target clock speed. But what that ``magic division'' is varies
6031 depending on the chips on your board.
6032 @b{ARM rule of thumb} Most ARM based systems require an 6:1 division;
6033 ARM11 cores use an 8:1 division.
6034 @b{Xilinx rule of thumb} is 1/12 the clock speed.
6036 Note: Many FTDI2232C based JTAG dongles are limited to 6MHz.
6038 You can still debug the 'low power' situations - you just need to
6039 manually adjust the clock speed at every step. While painful and
6040 tedious, it is not always practical.
6042 It is however easy to ``code your way around it'' - i.e.: Cheat a little,
6043 have a special debug mode in your application that does a ``high power
6044 sleep''. If you are careful - 98% of your problems can be debugged
6047 Note that on ARM you may need to avoid using the @emph{wait for interrupt}
6048 operation in your idle loops even if you don't otherwise change the CPU
6050 That operation gates the CPU clock, and thus the JTAG clock; which
6051 prevents JTAG access. One consequence is not being able to @command{halt}
6052 cores which are executing that @emph{wait for interrupt} operation.
6054 To set the JTAG frequency use the command:
6062 @item @b{Win32 Pathnames} Why don't backslashes work in Windows paths?
6064 OpenOCD uses Tcl and a backslash is an escape char. Use @{ and @}
6065 around Windows filenames.
6078 @item @b{Missing: cygwin1.dll} OpenOCD complains about a missing cygwin1.dll.
6080 Make sure you have Cygwin installed, or at least a version of OpenOCD that
6081 claims to come with all the necessary DLLs. When using Cygwin, try launching
6082 OpenOCD from the Cygwin shell.
6084 @item @b{Breakpoint Issue} I'm trying to set a breakpoint using GDB (or a frontend like Insight or
6085 Eclipse), but OpenOCD complains that "Info: arm7_9_common.c:213
6086 arm7_9_add_breakpoint(): sw breakpoint requested, but software breakpoints not enabled".
6088 GDB issues software breakpoints when a normal breakpoint is requested, or to implement
6089 source-line single-stepping. On ARMv4T systems, like ARM7TDMI, ARM720T or ARM920T,
6090 software breakpoints consume one of the two available hardware breakpoints.
6092 @item @b{LPC2000 Flash} When erasing or writing LPC2000 on-chip flash, the operation fails at random.
6094 Make sure the core frequency specified in the @option{flash lpc2000} line matches the
6095 clock at the time you're programming the flash. If you've specified the crystal's
6096 frequency, make sure the PLL is disabled. If you've specified the full core speed
6097 (e.g. 60MHz), make sure the PLL is enabled.
6099 @item @b{Amontec Chameleon} When debugging using an Amontec Chameleon in its JTAG Accelerator configuration,
6100 I keep getting "Error: amt_jtagaccel.c:184 amt_wait_scan_busy(): amt_jtagaccel timed
6101 out while waiting for end of scan, rtck was disabled".
6103 Make sure your PC's parallel port operates in EPP mode. You might have to try several
6104 settings in your PC BIOS (ECP, EPP, and different versions of those).
6106 @item @b{Data Aborts} When debugging with OpenOCD and GDB (plain GDB, Insight, or Eclipse),
6107 I get lots of "Error: arm7_9_common.c:1771 arm7_9_read_memory():
6108 memory read caused data abort".
6110 The errors are non-fatal, and are the result of GDB trying to trace stack frames
6111 beyond the last valid frame. It might be possible to prevent this by setting up
6112 a proper "initial" stack frame, if you happen to know what exactly has to
6113 be done, feel free to add this here.
6115 @b{Simple:} In your startup code - push 8 registers of zeros onto the
6116 stack before calling main(). What GDB is doing is ``climbing'' the run
6117 time stack by reading various values on the stack using the standard
6118 call frame for the target. GDB keeps going - until one of 2 things
6119 happen @b{#1} an invalid frame is found, or @b{#2} some huge number of
6120 stackframes have been processed. By pushing zeros on the stack, GDB
6123 @b{Debugging Interrupt Service Routines} - In your ISR before you call
6124 your C code, do the same - artifically push some zeros onto the stack,
6125 remember to pop them off when the ISR is done.
6127 @b{Also note:} If you have a multi-threaded operating system, they
6128 often do not @b{in the intrest of saving memory} waste these few
6132 @item @b{JTAG Reset Config} I get the following message in the OpenOCD console (or log file):
6133 "Warning: arm7_9_common.c:679 arm7_9_assert_reset(): srst resets test logic, too".
6135 This warning doesn't indicate any serious problem, as long as you don't want to
6136 debug your core right out of reset. Your .cfg file specified @option{jtag_reset
6137 trst_and_srst srst_pulls_trst} to tell OpenOCD that either your board,
6138 your debugger or your target uC (e.g. LPC2000) can't assert the two reset signals
6139 independently. With this setup, it's not possible to halt the core right out of
6140 reset, everything else should work fine.
6142 @item @b{USB Power} When using OpenOCD in conjunction with Amontec JTAGkey and the Yagarto
6143 toolchain (Eclipse, arm-elf-gcc, arm-elf-gdb), the debugging seems to be
6144 unstable. When single-stepping over large blocks of code, GDB and OpenOCD
6145 quit with an error message. Is there a stability issue with OpenOCD?
6147 No, this is not a stability issue concerning OpenOCD. Most users have solved
6148 this issue by simply using a self-powered USB hub, which they connect their
6149 Amontec JTAGkey to. Apparently, some computers do not provide a USB power
6150 supply stable enough for the Amontec JTAGkey to be operated.
6152 @b{Laptops running on battery have this problem too...}
6154 @item @b{USB Power} When using the Amontec JTAGkey, sometimes OpenOCD crashes with the
6155 following error messages: "Error: ft2232.c:201 ft2232_read(): FT_Read returned:
6156 4" and "Error: ft2232.c:365 ft2232_send_and_recv(): couldn't read from FT2232".
6157 What does that mean and what might be the reason for this?
6159 First of all, the reason might be the USB power supply. Try using a self-powered
6160 hub instead of a direct connection to your computer. Secondly, the error code 4
6161 corresponds to an FT_IO_ERROR, which means that the driver for the FTDI USB
6162 chip ran into some sort of error - this points us to a USB problem.
6164 @item @b{GDB Disconnects} When using the Amontec JTAGkey, sometimes OpenOCD crashes with the following
6165 error message: "Error: gdb_server.c:101 gdb_get_char(): read: 10054".
6166 What does that mean and what might be the reason for this?
6168 Error code 10054 corresponds to WSAECONNRESET, which means that the debugger (GDB)
6169 has closed the connection to OpenOCD. This might be a GDB issue.
6171 @item @b{LPC2000 Flash} In the configuration file in the section where flash device configurations
6172 are described, there is a parameter for specifying the clock frequency
6173 for LPC2000 internal flash devices (e.g. @option{flash bank lpc2000
6174 0x0 0x40000 0 0 0 lpc2000_v1 14746 calc_checksum}), which must be
6175 specified in kilohertz. However, I do have a quartz crystal of a
6176 frequency that contains fractions of kilohertz (e.g. 14,745,600 Hz,
6177 i.e. 14,745.600 kHz). Is it possible to specify real numbers for the
6180 No. The clock frequency specified here must be given as an integral number.
6181 However, this clock frequency is used by the In-Application-Programming (IAP)
6182 routines of the LPC2000 family only, which seems to be very tolerant concerning
6183 the given clock frequency, so a slight difference between the specified clock
6184 frequency and the actual clock frequency will not cause any trouble.
6186 @item @b{Command Order} Do I have to keep a specific order for the commands in the configuration file?
6188 Well, yes and no. Commands can be given in arbitrary order, yet the
6189 devices listed for the JTAG scan chain must be given in the right
6190 order (jtag newdevice), with the device closest to the TDO-Pin being
6191 listed first. In general, whenever objects of the same type exist
6192 which require an index number, then these objects must be given in the
6193 right order (jtag newtap, targets and flash banks - a target
6194 references a jtag newtap and a flash bank references a target).
6196 You can use the ``scan_chain'' command to verify and display the tap order.
6198 Also, some commands can't execute until after @command{init} has been
6199 processed. Such commands include @command{nand probe} and everything
6200 else that needs to write to controller registers, perhaps for setting
6201 up DRAM and loading it with code.
6203 @anchor{FAQ TAP Order}
6204 @item @b{JTAG TAP Order} Do I have to declare the TAPS in some
6207 Yes; whenever you have more than one, you must declare them in
6208 the same order used by the hardware.
6210 Many newer devices have multiple JTAG TAPs. For example: ST
6211 Microsystems STM32 chips have two TAPs, a ``boundary scan TAP'' and
6212 ``Cortex-M3'' TAP. Example: The STM32 reference manual, Document ID:
6213 RM0008, Section 26.5, Figure 259, page 651/681, the ``TDI'' pin is
6214 connected to the boundary scan TAP, which then connects to the
6215 Cortex-M3 TAP, which then connects to the TDO pin.
6217 Thus, the proper order for the STM32 chip is: (1) The Cortex-M3, then
6218 (2) The boundary scan TAP. If your board includes an additional JTAG
6219 chip in the scan chain (for example a Xilinx CPLD or FPGA) you could
6220 place it before or after the STM32 chip in the chain. For example:
6223 @item OpenOCD_TDI(output) -> STM32 TDI Pin (BS Input)
6224 @item STM32 BS TDO (output) -> STM32 Cortex-M3 TDI (input)
6225 @item STM32 Cortex-M3 TDO (output) -> SM32 TDO Pin
6226 @item STM32 TDO Pin (output) -> Xilinx TDI Pin (input)
6227 @item Xilinx TDO Pin -> OpenOCD TDO (input)
6230 The ``jtag device'' commands would thus be in the order shown below. Note:
6233 @item jtag newtap Xilinx tap -irlen ...
6234 @item jtag newtap stm32 cpu -irlen ...
6235 @item jtag newtap stm32 bs -irlen ...
6236 @item # Create the debug target and say where it is
6237 @item target create stm32.cpu -chain-position stm32.cpu ...
6241 @item @b{SYSCOMP} Sometimes my debugging session terminates with an error. When I look into the
6242 log file, I can see these error messages: Error: arm7_9_common.c:561
6243 arm7_9_execute_sys_speed(): timeout waiting for SYSCOMP
6249 @node Tcl Crash Course
6250 @chapter Tcl Crash Course
6253 Not everyone knows Tcl - this is not intended to be a replacement for
6254 learning Tcl, the intent of this chapter is to give you some idea of
6255 how the Tcl scripts work.
6257 This chapter is written with two audiences in mind. (1) OpenOCD users
6258 who need to understand a bit more of how JIM-Tcl works so they can do
6259 something useful, and (2) those that want to add a new command to
6262 @section Tcl Rule #1
6263 There is a famous joke, it goes like this:
6265 @item Rule #1: The wife is always correct
6266 @item Rule #2: If you think otherwise, See Rule #1
6269 The Tcl equal is this:
6272 @item Rule #1: Everything is a string
6273 @item Rule #2: If you think otherwise, See Rule #1
6276 As in the famous joke, the consequences of Rule #1 are profound. Once
6277 you understand Rule #1, you will understand Tcl.
6279 @section Tcl Rule #1b
6280 There is a second pair of rules.
6282 @item Rule #1: Control flow does not exist. Only commands
6283 @* For example: the classic FOR loop or IF statement is not a control
6284 flow item, they are commands, there is no such thing as control flow
6286 @item Rule #2: If you think otherwise, See Rule #1
6287 @* Actually what happens is this: There are commands that by
6288 convention, act like control flow key words in other languages. One of
6289 those commands is the word ``for'', another command is ``if''.
6292 @section Per Rule #1 - All Results are strings
6293 Every Tcl command results in a string. The word ``result'' is used
6294 deliberatly. No result is just an empty string. Remember: @i{Rule #1 -
6295 Everything is a string}
6297 @section Tcl Quoting Operators
6298 In life of a Tcl script, there are two important periods of time, the
6299 difference is subtle.
6302 @item Evaluation Time
6305 The two key items here are how ``quoted things'' work in Tcl. Tcl has
6306 three primary quoting constructs, the [square-brackets] the
6307 @{curly-braces@} and ``double-quotes''
6309 By now you should know $VARIABLES always start with a $DOLLAR
6310 sign. BTW: To set a variable, you actually use the command ``set'', as
6311 in ``set VARNAME VALUE'' much like the ancient BASIC langauge ``let x
6312 = 1'' statement, but without the equal sign.
6315 @item @b{[square-brackets]}
6316 @* @b{[square-brackets]} are command substitutions. It operates much
6317 like Unix Shell `back-ticks`. The result of a [square-bracket]
6318 operation is exactly 1 string. @i{Remember Rule #1 - Everything is a
6319 string}. These two statements are roughly identical:
6323 echo "The Date is: $X"
6326 puts "The Date is: $X"
6328 @item @b{``double-quoted-things''}
6329 @* @b{``double-quoted-things''} are just simply quoted
6330 text. $VARIABLES and [square-brackets] are expanded in place - the
6331 result however is exactly 1 string. @i{Remember Rule #1 - Everything
6335 puts "It is now \"[date]\", $x is in 1 hour"
6337 @item @b{@{Curly-Braces@}}
6338 @*@b{@{Curly-Braces@}} are magic: $VARIABLES and [square-brackets] are
6339 parsed, but are NOT expanded or executed. @{Curly-Braces@} are like
6340 'single-quote' operators in BASH shell scripts, with the added
6341 feature: @{curly-braces@} can be nested, single quotes can not. @{@{@{this is
6342 nested 3 times@}@}@} NOTE: [date] is a bad example;
6343 at this writing, Jim/OpenOCD does not have a date command.
6346 @section Consequences of Rule 1/2/3/4
6348 The consequences of Rule 1 are profound.
6350 @subsection Tokenisation & Execution.
6352 Of course, whitespace, blank lines and #comment lines are handled in
6355 As a script is parsed, each (multi) line in the script file is
6356 tokenised and according to the quoting rules. After tokenisation, that
6357 line is immedatly executed.
6359 Multi line statements end with one or more ``still-open''
6360 @{curly-braces@} which - eventually - closes a few lines later.
6362 @subsection Command Execution
6364 Remember earlier: There are no ``control flow''
6365 statements in Tcl. Instead there are COMMANDS that simply act like
6366 control flow operators.
6368 Commands are executed like this:
6371 @item Parse the next line into (argc) and (argv[]).
6372 @item Look up (argv[0]) in a table and call its function.
6373 @item Repeat until End Of File.
6376 It sort of works like this:
6379 ReadAndParse( &argc, &argv );
6381 cmdPtr = LookupCommand( argv[0] );
6383 (*cmdPtr->Execute)( argc, argv );
6387 When the command ``proc'' is parsed (which creates a procedure
6388 function) it gets 3 parameters on the command line. @b{1} the name of
6389 the proc (function), @b{2} the list of parameters, and @b{3} the body
6390 of the function. Not the choice of words: LIST and BODY. The PROC
6391 command stores these items in a table somewhere so it can be found by
6394 @subsection The FOR command
6396 The most interesting command to look at is the FOR command. In Tcl,
6397 the FOR command is normally implemented in C. Remember, FOR is a
6398 command just like any other command.
6400 When the ascii text containing the FOR command is parsed, the parser
6401 produces 5 parameter strings, @i{(If in doubt: Refer to Rule #1)} they
6405 @item The ascii text 'for'
6406 @item The start text
6407 @item The test expression
6412 Sort of reminds you of ``main( int argc, char **argv )'' does it not?
6413 Remember @i{Rule #1 - Everything is a string.} The key point is this:
6414 Often many of those parameters are in @{curly-braces@} - thus the
6415 variables inside are not expanded or replaced until later.
6417 Remember that every Tcl command looks like the classic ``main( argc,
6418 argv )'' function in C. In JimTCL - they actually look like this:
6422 MyCommand( Jim_Interp *interp,
6424 Jim_Obj * const *argvs );
6427 Real Tcl is nearly identical. Although the newer versions have
6428 introduced a byte-code parser and intepreter, but at the core, it
6429 still operates in the same basic way.
6431 @subsection FOR command implementation
6433 To understand Tcl it is perhaps most helpful to see the FOR
6434 command. Remember, it is a COMMAND not a control flow structure.
6436 In Tcl there are two underlying C helper functions.
6438 Remember Rule #1 - You are a string.
6440 The @b{first} helper parses and executes commands found in an ascii
6441 string. Commands can be seperated by semicolons, or newlines. While
6442 parsing, variables are expanded via the quoting rules.
6444 The @b{second} helper evaluates an ascii string as a numerical
6445 expression and returns a value.
6447 Here is an example of how the @b{FOR} command could be
6448 implemented. The pseudo code below does not show error handling.
6450 void Execute_AsciiString( void *interp, const char *string );
6452 int Evaluate_AsciiExpression( void *interp, const char *string );
6455 MyForCommand( void *interp,
6460 SetResult( interp, "WRONG number of parameters");
6464 // argv[0] = the ascii string just like C
6466 // Execute the start statement.
6467 Execute_AsciiString( interp, argv[1] );
6471 i = Evaluate_AsciiExpression(interp, argv[2]);
6476 Execute_AsciiString( interp, argv[3] );
6478 // Execute the LOOP part
6479 Execute_AsciiString( interp, argv[4] );
6483 SetResult( interp, "" );
6488 Every other command IF, WHILE, FORMAT, PUTS, EXPR, everything works
6489 in the same basic way.
6491 @section OpenOCD Tcl Usage
6493 @subsection source and find commands
6494 @b{Where:} In many configuration files
6495 @* Example: @b{ source [find FILENAME] }
6496 @*Remember the parsing rules
6498 @item The FIND command is in square brackets.
6499 @* The FIND command is executed with the parameter FILENAME. It should
6500 find the full path to the named file. The RESULT is a string, which is
6501 substituted on the orginal command line.
6502 @item The command source is executed with the resulting filename.
6503 @* SOURCE reads a file and executes as a script.
6505 @subsection format command
6506 @b{Where:} Generally occurs in numerous places.
6507 @* Tcl has no command like @b{printf()}, instead it has @b{format}, which is really more like
6513 puts [format "The answer: %d" [expr $x * $y]]
6516 @item The SET command creates 2 variables, X and Y.
6517 @item The double [nested] EXPR command performs math
6518 @* The EXPR command produces numerical result as a string.
6520 @item The format command is executed, producing a single string
6521 @* Refer to Rule #1.
6522 @item The PUTS command outputs the text.
6524 @subsection Body or Inlined Text
6525 @b{Where:} Various TARGET scripts.
6528 proc someproc @{@} @{
6529 ... multiple lines of stuff ...
6531 $_TARGETNAME configure -event FOO someproc
6532 #2 Good - no variables
6533 $_TARGETNAME confgure -event foo "this ; that;"
6534 #3 Good Curly Braces
6535 $_TARGETNAME configure -event FOO @{
6538 #4 DANGER DANGER DANGER
6539 $_TARGETNAME configure -event foo "puts \"Time: [date]\""
6542 @item The $_TARGETNAME is an OpenOCD variable convention.
6543 @*@b{$_TARGETNAME} represents the last target created, the value changes
6544 each time a new target is created. Remember the parsing rules. When
6545 the ascii text is parsed, the @b{$_TARGETNAME} becomes a simple string,
6546 the name of the target which happens to be a TARGET (object)
6548 @item The 2nd parameter to the @option{-event} parameter is a TCBODY
6549 @*There are 4 examples:
6551 @item The TCLBODY is a simple string that happens to be a proc name
6552 @item The TCLBODY is several simple commands seperated by semicolons
6553 @item The TCLBODY is a multi-line @{curly-brace@} quoted string
6554 @item The TCLBODY is a string with variables that get expanded.
6557 In the end, when the target event FOO occurs the TCLBODY is
6558 evaluated. Method @b{#1} and @b{#2} are functionally identical. For
6559 Method @b{#3} and @b{#4} it is more interesting. What is the TCLBODY?
6561 Remember the parsing rules. In case #3, @{curly-braces@} mean the
6562 $VARS and [square-brackets] are expanded later, when the EVENT occurs,
6563 and the text is evaluated. In case #4, they are replaced before the
6564 ``Target Object Command'' is executed. This occurs at the same time
6565 $_TARGETNAME is replaced. In case #4 the date will never
6566 change. @{BTW: [date] is a bad example; at this writing,
6567 Jim/OpenOCD does not have a date command@}
6569 @subsection Global Variables
6570 @b{Where:} You might discover this when writing your own procs @* In
6571 simple terms: Inside a PROC, if you need to access a global variable
6572 you must say so. See also ``upvar''. Example:
6574 proc myproc @{ @} @{
6575 set y 0 #Local variable Y
6576 global x #Global variable X
6577 puts [format "X=%d, Y=%d" $x $y]
6580 @section Other Tcl Hacks
6581 @b{Dynamic variable creation}
6583 # Dynamically create a bunch of variables.
6584 for @{ set x 0 @} @{ $x < 32 @} @{ set x [expr $x + 1]@} @{
6586 set vn [format "BIT%d" $x]
6590 set $vn [expr (1 << $x)]
6593 @b{Dynamic proc/command creation}
6595 # One "X" function - 5 uart functions.
6596 foreach who @{A B C D E@}
6597 proc [format "show_uart%c" $who] @{ @} "show_UARTx $who"
6601 @node Target Library
6602 @chapter Target Library
6603 @cindex Target Library
6605 OpenOCD comes with a target configuration script library. These scripts can be
6606 used as-is or serve as a starting point.
6608 The target library is published together with the OpenOCD executable and
6609 the path to the target library is in the OpenOCD script search path.
6610 Similarly there are example scripts for configuring the JTAG interface.
6612 The command line below uses the example parport configuration script
6613 that ship with OpenOCD, then configures the str710.cfg target and
6614 finally issues the init and reset commands. The communication speed
6615 is set to 10kHz for reset and 8MHz for post reset.
6618 openocd -f interface/parport.cfg -f target/str710.cfg \
6619 -c "init" -c "reset"
6622 To list the target scripts available:
6625 $ ls /usr/local/lib/openocd/target
6627 arm7_fast.cfg lm3s6965.cfg pxa255.cfg stm32.cfg xba_revA3.cfg
6628 at91eb40a.cfg lpc2148.cfg pxa255_sst.cfg str710.cfg zy1000.cfg
6629 at91r40008.cfg lpc2294.cfg sam7s256.cfg str912.cfg
6630 at91sam9260.cfg nslu2.cfg sam7x256.cfg wi-9c.cfg
6635 @node OpenOCD Concept Index
6636 @comment DO NOT use the plain word ``Index'', reason: CYGWIN filename
6637 @comment case issue with ``Index.html'' and ``index.html''
6638 @comment Occurs when creating ``--html --no-split'' output
6639 @comment This fix is based on: http://sourceware.org/ml/binutils/2006-05/msg00215.html
6640 @unnumbered OpenOCD Concept Index
6644 @node Command and Driver Index
6645 @unnumbered Command and Driver Index