You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

396 lines
16KB

  1. // This file is part of the Doxygen Developer Manual
  2. /** @page tasks Pending and Open Tasks
  3. This page lists pending and open tasks being considered or worked upon
  4. by the OpenOCD community.
  5. @section thelist The List
  6. Most items are open for the taking, but please post to the mailing list
  7. before spending much time working on anything lists here. The community
  8. may have evolved an idea since it was added here.
  9. Feel free to send patches to add or clarify items on this list, too.
  10. @section thelisttcl TCL
  11. This section provides possible things to improve with OpenOCD's TCL support.
  12. - Fix problem with incorrect line numbers reported for a syntax
  13. error in a reset init event.
  14. - organize the TCL configurations:
  15. - provide more directory structure for boards/targets?
  16. - factor configurations into layers (encapsulation and re-use)
  17. - Fix handling of variables between multiple command line "-c" and "-f"
  18. parameters. Currently variables assigned through one such parameter
  19. command/script are unset before the next one is invoked.
  20. - Isolate all TCL command support:
  21. - Pure C CLI implementations using --disable-builtin-tcl.
  22. - Allow developers to build new dongles using OpenOCD's JTAG core.
  23. - At first, provide only low-level JTAG support; target layer and
  24. above rely heavily on scripting event mechanisms.
  25. - Allow full TCL support? add --with-tcl=/path/to/installed/tcl
  26. - Move TCL support out of foo.[ch] and into foo_tcl.[ch] (other ideas?)
  27. - See src/jtag/core.c and src/jtag/tcl.c for an example.
  28. - allow some of these TCL command modules to be dynamically loadable?
  29. @section thelistadapter Adapter
  30. This section list issues that need to be resolved in the Adapter layer.
  31. @subsection thelistadapterrework Code restructuring
  32. This section lists pending reworks to complete the restructure from the
  33. old JTAG centric implementation to a generic Adapter layer.
  34. This restructuring is very invasive and will prevent the merge of several
  35. changes pending in gerrit.
  36. - rename folder src/jtag/ to src/adapter/
  37. - rename var "jtag" to "adapter" in src/jtag/core.c
  38. - split content of src/adapter/ in the different protocols jtag.[ch],
  39. swd.[ch], ...
  40. - wrap the calls to adapter->transport_ops->api() with transport_api()
  41. and reduce the visibility of global var "adapter"
  42. - complete the migration of JTAG-only drivers to adapter->reset()
  43. - try to remove JTAG_SLEEP also from JTAG mode?
  44. - tap_set_state(TAP_RESET) is already done in src/jtag/core.c. No need
  45. to replicate it in the drivers, apart in case the driver sets TRST
  46. independently
  47. - separate SWIM from HLA and make it independent
  48. - add .hla_ops and .swim_ops to "adapter"
  49. - HLA is a API level (.hla_ops). Transport should simply be {jtag,swd},
  50. not {hla_jtag,hla_swd}.
  51. @subsection thelistadapterjtagcore JTAG Core
  52. The following tasks have been suggested for cleaning up the JTAG layer:
  53. - use tap_set_state everywhere to allow logging TAP state transitions
  54. - Encapsulate cmd_queue_cur_state and related variable handling.
  55. - add slick 32 bit versions of jtag_add_xxx_scan() that avoids
  56. buf_set_u32() calls and other evidence of poor impedance match between
  57. API and calling code. New API should cut down # of lines in calling
  58. code by 100's and make things clearer. Also potentially be supported
  59. directly in minidriver API for better embedded host performance.
  60. The following tasks have been suggested for adding new core JTAG support:
  61. - Improve autodetection of TAPs by supporting tcl escape procedures that
  62. can configure discovered TAPs based on IDCODE value ... they could:
  63. - Remove guessing for irlen
  64. - Allow non-default irmask/ircapture values
  65. - SPI/UART emulation:
  66. - (ab)use bit-banging JTAG interfaces to emulate SPI/UART
  67. - allow SPI to program flash, MCUs, etc.
  68. @subsection thelistadapterinterfaces Interface drivers
  69. There are some known bugs to fix in Interface drivers:
  70. - For JTAG_STATEMOVE to TAP_RESET, all drivers must ignore the current
  71. recorded state. The tap_get_state() call won't necessarily return
  72. the correct value, especially at server startup. Fix is easy: in
  73. that case, always issue five clocks with TMS high.
  74. - amt_jtagaccel.c
  75. - arm-jtag-ew.c
  76. - bitbang.c
  77. - bitq.c
  78. - gw16012.c
  79. - jlink.c
  80. - usbprog.c
  81. - vsllink.c
  82. - rlink/rlink.c
  83. - bug: USBprog is broken with new tms sequence; it needs 7-clock cycles.
  84. Fix promised from Peter Denison openwrt at marshadder.org
  85. Workaround: use "tms_sequence long" @par
  86. https://lists.berlios.de/pipermail/openocd-development/2009-July/009426.html
  87. The following tasks have been suggested for improving OpenOCD's JTAG
  88. interface support:
  89. - rework USB communication to be more robust. Two possible options are:
  90. -# use libusb-1.0.1 with libusb-compat-0.1.1 (non-blocking I/O wrapper)
  91. -# rewrite implementation to use non-blocking I/O
  92. - J-Link driver:
  93. - fix to work with long scan chains, such as R.Doss's svf test.
  94. - Autodetect USB based adapters; this should be easy on Linux. If there's
  95. more than one, list the options; otherwise, just select that one.
  96. The following tasks have been suggested for adding new JTAG interfaces:
  97. - TCP driver: allow client/server for remote JTAG interface control.
  98. This requires a client and a server. The server is built into the
  99. normal OpenOCD and takes commands from the client and executes
  100. them on the interface returning the result of TCP/IP. The client
  101. is an OpenOCD which is built with a TCP/IP minidriver. The use
  102. of a minidriver is required to capture all the jtag_add_xxx()
  103. fn's at a high enough level and repackage these cmd's as
  104. TCP/IP packets handled by the server.
  105. @section thelistbs Boundary Scan Support
  106. - add STAPL support?
  107. - add BSDL support?
  108. A few possible options for the above:
  109. -# Fake a TCL equivalent?
  110. -# Integrate an existing library?
  111. -# Write a new C implementation a la Jim?
  112. Once the above are completed:
  113. - add support for programming flash using boundary scan techniques
  114. - add integration with a modified gerber view program:
  115. - provide means to view the PCB and select pins and traces
  116. - allow use-cases such as the following:
  117. - @b Stimulus
  118. -# Double-click on a pin (or trace) with the mouse.
  119. - @b Effects
  120. -# The trace starts blinking, and
  121. -# OpenOCD toggles the pin(s) 0/1.
  122. @section thelisttargets Target Support
  123. - Many common ARM cores could be autodetected using IDCODE
  124. - general layer cleanup: @par
  125. https://lists.berlios.de/pipermail/openocd-development/2009-May/006590.html
  126. - regression: "reset halt" between 729(works) and 788(fails): @par
  127. https://lists.berlios.de/pipermail/openocd-development/2009-July/009206.html
  128. - registers
  129. - add flush-value operation, call them all on resume/reset
  130. - mcr/mrc target->type support
  131. - missing from ARM920t, ARM966e, XScale.
  132. It's possible that the current syntax is unable to support read-modify-write
  133. operations(see arm966e).
  134. - mcr/mrc - retire cp15 commands when there the mrc/mrc commands have been
  135. tested from: arm926ejs, arm720t, cortex_a8
  136. - ARM7/9:
  137. - clean up "arm9tdmi vector_catch". Available for some arm7 cores? @par
  138. https://lists.berlios.de/pipermail/openocd-development/2009-October/011488.html
  139. https://lists.berlios.de/pipermail/openocd-development/2009-October/011506.html
  140. - add reset option to allow programming embedded ice while srst is asserted.
  141. Some CPUs will gate the JTAG clock when srst is asserted and in this case,
  142. it is necessary to program embedded ice and then assert srst afterwards.
  143. - ARM926EJS:
  144. - reset run/halt/step is not robust; needs testing to map out problems.
  145. - ARM11 improvements (MB?)
  146. - add support for asserting srst to reset the core.
  147. - Single stepping works, but should automatically
  148. use hardware stepping if available.
  149. - mdb can return garbage data if read byte operation fails for
  150. a memory region(16 & 32 byte access modes may be supported). Is this
  151. a bug in the .MX31 PDK init script? Try on i.MX31 PDK:
  152. mdw 0xb80005f0 0x8, mdh 0xb80005f0 0x10, mdb 0xb80005f0 0x20. mdb returns
  153. garabage.
  154. - implement missing functionality (grep FNC_INFO_NOTIMPLEMENTED ...)
  155. - Thumb2 single stepping: ARM1156T2 needs simulator support
  156. - Cortex-A8 support (ML)
  157. - add target implementation (ML)
  158. - Cortex-M3 support
  159. - when stepping, only write dirtied registers (be faster)
  160. - when connecting to halted core, fetch registers (startup is quirky)
  161. - Generic ARM run_algorithm() interface
  162. - tagged struct wrapping ARM instructions and metadata
  163. - not revision-specific (current: ARMv4+ARMv5 -or- ARMv6 -or- ARMv7)
  164. - usable with at least arm_nandwrite() and generic CFI drivers
  165. - ETM
  166. - don't show FIFOFULL registers if they're not supported
  167. - use comparators to get more breakpoints and watchpoints
  168. - add "etm drivers" command
  169. - trace driver init() via examine() paths only, not setup()/reset
  170. - MC1322x support (JW/DE?)
  171. - integrate and test support from JW (and DE?)
  172. - get working with a known good interface (i.e. not today's jlink)
  173. - AT91SAM92xx:
  174. - improvements for unknown-board-atmel-at91sam9260.cfg (RD)
  175. - STR9x: (ZW)
  176. - improvements to str912.cfg to be more general purpose
  177. - AVR: (SQ)
  178. - independently verify implementation
  179. - incrementally improve working prototype in trunk. (SQ)
  180. - work out how to debug this target
  181. - AVR debugging protocol.
  182. - FPGA:
  183. - Altera Nios Soft-CPU support
  184. - Coldfire (suggested by NC)
  185. - can we draw from the BDM project? @par
  186. http://bdm.sourceforge.net/
  187. or the OSBDM package @par
  188. http://forums.freescale.com/freescale/board/message?board.id=OSBDM08&thread.id=422
  189. @section thelistsvf SVF/XSVF
  190. - develop SVF unit tests
  191. - develop XSVF unit tests
  192. @section thelistflash Flash Support
  193. - finish documentation for the following flash drivers:
  194. - avr
  195. - pic32mx
  196. - ocl
  197. - str9xpec
  198. - Don't expect writing all-ones to be a safe way to write without
  199. changing bit values. Minimally it loses on flash modules with
  200. internal ECC, where it may change the ECC.
  201. - NOR flash_write_unlock() does that between sectors
  202. - there may be other cases too
  203. - Make sure all commands accept either a bank name or a bank number,
  204. and be sure both identifiers show up in "flash banks" and "nand list".
  205. Right now the user-friendly names are pretty much hidden...
  206. @subsection thelistflashcfi CFI
  207. - finish implementing bus width/chip width handling (suggested by NC)
  208. - factor vendor-specific code into separate source files
  209. - add new callback interface for vendor-specific code
  210. - investigate/implement "thin wrapper" to use eCos CFI drivers (ØH)
  211. @section thelistdebug Debugger Support
  212. - add support for masks in watchpoints? @par
  213. https://lists.berlios.de/pipermail/openocd-development/2009-October/011507.html
  214. - breakpoints can get lost in some circumstances: @par
  215. https://lists.berlios.de/pipermail/openocd-development/2009-June/008853.html
  216. - add support for masks in watchpoints. The trick is that GDB does not
  217. support a breakpoint mask in the remote protocol. One way to work around
  218. this is to add a separate command "watchpoint_mask add/rem <addr> <mask>", that
  219. is run to register a list of masks that the gdb_server knows to use with
  220. a particular watchpoint address.
  221. - integrate Keil AGDI interface to OpenOCD? (submitted by Dario Vecchio)
  222. @section thelisttesting Testing Suite
  223. This section includes several related groups of ideas:
  224. - @ref thelistunittests
  225. - @ref thelistsmoketests
  226. - @ref thelisttestreports
  227. - @ref thelisttestgenerichw
  228. @subsection thelistunittests Unit Tests
  229. - add testing skeleton to provide frameworks for adding tests
  230. - implement server unit tests
  231. - implement JTAG core unit tests
  232. - implement JTAG interface unit tests
  233. - implement flash unit tests
  234. - implement target unit tests
  235. @subsection thelistsmoketests Smoke Test Tools
  236. -# extend 'make check' with a smoketest app
  237. - checks for OOCD_TEST_CONFIG, etc. in environment (or config file)
  238. - if properly set, runs the smoke test with specified parameters
  239. - openocd -f ${OOCD_TEST_CONFIG}
  240. - implies a modular test suite (see below)
  241. - should be able to run some minimal tests with dummy interface:
  242. - compare results of baseline sanity checks with expected results
  243. -# builds a more complete test suite:
  244. - existing testing/examples/ look like a great start
  245. - all targets should be tested fully and for all capabilities
  246. - we do NOT want a "lowest common denominator" test suite
  247. - ... but can we start with one to get going?
  248. - probably requires one test configuration file per board/target
  249. - modularization can occur here, just like with targets/boards/chips
  250. - coverage can increase over time, building up bundles of tests
  251. -# add new 'smoketest' Makefile target:
  252. - calls 'make check' (and the smoketest app)
  253. - gather inputs and output into a report file
  254. @subsection thelisttestreports Test Feedback Tools
  255. These ideas were first introduced here: @par
  256. https://lists.berlios.de/pipermail/openocd-development/2009-May/006358.html
  257. - provide report submission scripts for e-mail and web forms
  258. - add new Makefile targets to post the report:
  259. - 'checkreportsend' -- send to list via e-mail (via sendmail)
  260. - 'checkreportpost' -- send web form (via curl or other script)
  261. @subsection thelisttestgenerichw Generic Hardware Tester
  262. - implement VHDL to use for FPGA-based JTAG TAP testing device
  263. - develop test suite that utilizes this testing device
  264. @section thelistautotools Autotools Build System
  265. - make entire configure process require less user consideration:
  266. - automatically detect the features that are available, unless
  267. options were specifically provided to configure
  268. - provide a report of the drivers that will be build at the end of
  269. running configure, so the users can verify which drivers will be
  270. built during 'make' (and their options) .
  271. - eliminate sources of confusion in @c bootstrap script:
  272. -# Make @c bootstrap call 'configure --enable-maintainer-mode \<opts\>'?
  273. -# Add @c buildstrap script to assist with bootstrap and configure steps.
  274. - automatically build tool-chains required for cross-compiling
  275. - produce mingw32, arm-elf, others using in-tree scripts
  276. - build all required target code from sources
  277. - make JTAG and USB debug output a run-time configuration option
  278. @section thelistarchitecture Architectural Tasks
  279. The following architectural tasks need to be accomplished and should be
  280. fairly easy to complete:
  281. - use dynamic allocations for working memory. Scan & fix code
  282. for excessive stack allocations. take linux/scripts/checkstack.pl and
  283. see what the worst offenders are. Dynamic stack allocations are found
  284. at the bottom of the list below. Example, on amd64:
  285. $ objdump -d | checkstack.pl | head -10
  286. 0x004311e3 image_open [openocd]: 13464
  287. 0x00431301 image_open [openocd]: 13464
  288. 0x004237a4 target_array2mem [openocd]: 4376
  289. 0x0042382b target_array2mem [openocd]: 4376
  290. 0x00423e74 target_mem2array [openocd]: 4360
  291. 0x00423ef9 target_mem2array [openocd]: 4360
  292. 0x00404aed handle_svf_command [openocd]: 2248
  293. 0x00404b7e handle_svf_command [openocd]: 2248
  294. 0x00413581 handle_flash_fill_command [openocd]: 2200
  295. 0x004135fa handle_flash_fill_command [openocd]: 2200
  296. - clean-up code to match style guides
  297. - factor code to eliminate duplicated functionality
  298. - rewrite code that uses casts to access 16-bit and larger types
  299. from unaligned memory addresses
  300. - libopenocd support: @par
  301. https://lists.berlios.de/pipermail/openocd-development/2009-May/006405.html
  302. - review and clean up interface/target/flash APIs
  303. The following strategic tasks will require ambition, knowledge, and time
  304. to complete:
  305. - overhaul use of types to improve 32/64-bit portability
  306. - types for both host and target word sizes?
  307. - can we use GDB's CORE_TYPE support?
  308. - Allow N:M:P mapping of servers, targets, and interfaces
  309. - loadable module support for interface/target/flash drivers and commands
  310. - support both static and dynamic modules.
  311. - should probably use libltdl for dynamic library handing.
  312. @section thelistadmin Documentation Tasks
  313. - Develop milestone and release guidelines, processes, and scripts.
  314. - Develop "style" guidelines (and scripts) for maintainers:
  315. - reviewing patches
  316. - committing to git
  317. - Review Users' Guide for documentation errors or omissions
  318. - "capture" and "ocd_find" commands
  319. - Update Developer's Manual (doxygen output)
  320. - Add documentation describing the architecture of each module
  321. - Provide more Technical Primers to bootstrap contributor knowledge
  322. */
  323. /** @file
  324. This file contains the @ref thelist page.
  325. */