- If communication fails during assert between assert/deassert and during
assert, warnings are printed. The warning suggests using srst_only if the
clock locks up as that would allow the reset vector to be set up before
asserting reset.
git-svn-id: svn://svn.berlios.de/openocd/trunk@544 b42882b7-edfa-0310-969c-e2dbd0fdcd60
as an optional parameter with default reset_init. This is to streamline
things w.r.t. the target library.
git-svn-id: svn://svn.berlios.de/openocd/trunk@540 b42882b7-edfa-0310-969c-e2dbd0fdcd60
for a bit more info before I can tell with confidence whether or not
this would have any effect.
- worked on error propagation and output for flash
git-svn-id: svn://svn.berlios.de/openocd/trunk@539 b42882b7-edfa-0310-969c-e2dbd0fdcd60
Wrote a default flash erase check fn which uses CFI's target algorithm
w/fallback to memory reads.
- "flash info" no longer prints erase status as it is stale.
- "flash erase_check" now prints erase status. erase check can take a
*long* time. Work in progress
- arm7/9 with seperate srst & trst now supports reset init/halt
after a power outage. arm7/9 no longer makes any assumptions
about state of target when reset is asserted.
- fixes for srst & trst capable arm7/9 with reset init/halt
- prepare_reset_halt retired. This code needs to be inside
assert_reset anyway
- haven't been able to get stm32 write algorithm to work. Fallback
flash write does work. Haven't found a version of openocd trunk
where this works.
- added target_free_all_working_areas_restore() which can
let be of restoring backups. This is needed when asserting
reset as the target must be assumed to be an unknown state.
Added some comments to working areas API
- str9 reset script fixes
- some guidelines
- fixed dangling callbacks upon reset timeout
git-svn-id: svn://svn.berlios.de/openocd/trunk@536 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- fix generic bitbang code to allow scans to end in Shift-[ID]R
- several CFI fixes (thanks to Michael Schwingen):
- buffer overflow when converting target code in cfi_intel_write_block -
cfi_fix_code_endian needs the number of words, not bytes, as size
argument.
- Spansion flash write was completely broken on big-endian targets - I
borrowed mechanisms from the intel driver, and moved some common code
into the cfi_command_val helper function. There is still more common code
that might be cleaned up.
- the buffer size check in cfi_write was broken for spansion flashes, where
cfi_write_words is not implemented. cfi_write_words is no only called if
the flash does have a buffer size >1.
- "flash info" printed CFI status information for non-CFI flashes, which is
confusing. It now only prints those when a real CFI flash is detected.
git-svn-id: svn://svn.berlios.de/openocd/trunk@517 b42882b7-edfa-0310-969c-e2dbd0fdcd60
is post reset = reset speed.
- removed infinite loop's and exit()'s upon poor arm7/9 communication
- cleaned up error messages a bit. Push ERROR() up into fn's that
fail and can say something meaningful about what failed.
git-svn-id: svn://svn.berlios.de/openocd/trunk@511 b42882b7-edfa-0310-969c-e2dbd0fdcd60
that all the jtag_add_xxx() are indeed intended to do the same thing.
git-svn-id: svn://svn.berlios.de/openocd/trunk@504 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- hopefully clarified the difference between TRST and TMS reset.
- added DEBUG() statements w.r.t. state changes
- TRST released and moving out of TAP_TLR are completely
different events. Only TRST released has a DEBUG() statement
git-svn-id: svn://svn.berlios.de/openocd/trunk@496 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- uses packed transfers for 8/16bit read/writes greater than 4bytes
- 8/16bit transfers now use address auto increment
git-svn-id: svn://svn.berlios.de/openocd/trunk@495 b42882b7-edfa-0310-969c-e2dbd0fdcd60
fail for two of it's return codes. A little bit weird, but compatible with
existing codebase.
- tightend up error handling. Since the jtag_xxx() is a queue that is either
executed as things are added(hw queue) or a software queue, then
errors can only be caught during jtag_execute_queue(). No error
code is therefore returned from the queuing fn's.
git-svn-id: svn://svn.berlios.de/openocd/trunk@492 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- hmm..... did I screw up? Was XScale and not gdb_server busted here?
My thinking was that OpenOCD has a canonical internal representation
of registers that match GDB's expectations
git-svn-id: svn://svn.berlios.de/openocd/trunk@484 b42882b7-edfa-0310-969c-e2dbd0fdcd60
are up and running. The examination and validation is actually
"optional" from the point of view of GDB + telnet servers.
Multiple targets should work fine with this.
- jtag_speed is dropped(divisor is increased), if jtag examination and
validation fails.
- the chain is validated 10x to catch the worst jtag_speed offences
- added LOG_SILENT that can be used to shut up log. Feeble
ersatz for try+catch.
- GDB register packets are now always replied in order to make sure
that GDB connect works. If the target is not halted, then these
packets contain dummy values.
git-svn-id: svn://svn.berlios.de/openocd/trunk@483 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- Fixed memory leak in gdb_server.c
- pushed "Error:" statements up into fn's that know something about what went wrong
- load_image now fails if target_write_memory() fails
- only issue an asynchronous halt() upon connect of GDB. Synchronous halt/reset
doesn't really work as what's required to initialize the target might involve a
special monitor sequence for the target in question
- syntax error handling improved(fewer exit()'s)
git-svn-id: svn://svn.berlios.de/openocd/trunk@482 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- ERROR_FAIL - added to emphasize that information about what went wrong is
contained in Error: log statements and not in a 32 bit return value.
git-svn-id: svn://svn.berlios.de/openocd/trunk@481 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- added new commands arm11 mrc and arm11 mcr to generate these instructions during debug
Thanks Michael Bruck
git-svn-id: svn://svn.berlios.de/openocd/trunk@477 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- switch to synchronous halt during connect. This fixes the bug
where poll() was not invoked between halt() and servicing the
'g' register packet
- halt() no longer returns error code when target is already halted, just
logs a warning. Only the halt() implementation can say anything
meaningful about why a halt() failed, so error messages are pushed
up to halt()
- fixed soft_reset_halt infinite loop bug in arm7_9_common.c. The rest
of the implementations are still busted.
- by using USER() instead of command_print() the log gets the
source + line #. Nice.
- no longer invoke exit() if soft_reset_halt fails. A reset can often
fix the problem.
git-svn-id: svn://svn.berlios.de/openocd/trunk@475 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- force simulate_reset_on_next_halt when target state is initially detected
- print out method of debug entry
- fix VCR activation (didn't work before)
git-svn-id: svn://svn.berlios.de/openocd/trunk@452 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- added a check to jtag_add_pathmove() for legal path transitions
- tweaked jtag.h docs a little bit
- made some jtag bypass tests _DEBUG_JTAG_IO_
git-svn-id: svn://svn.berlios.de/openocd/trunk@448 b42882b7-edfa-0310-969c-e2dbd0fdcd60
Passes 94kBytes/s regression test.
Works with str912, which relies on bypass
git-svn-id: svn://svn.berlios.de/openocd/trunk@447 b42882b7-edfa-0310-969c-e2dbd0fdcd60
uninitialized buffer issue in handle_flash_info_command in flash.c, some formatting.
git-svn-id: svn://svn.berlios.de/openocd/trunk@446 b42882b7-edfa-0310-969c-e2dbd0fdcd60
* GDB remote server will now remain online even if the target
is in a funny state, e.g. if it requires a reset, it is
running while GDB is not in the continue or step packet,
e.g. via monitor resume/halt commands in GDB script.
* Added some _DEBUG_GDB_IO_ debug tools
* Fixed a couple of GDB server lockups, e.g. when O packets
detect a severed connection
* added ACK upon connection (send +).
* added keep-alive messages to reset so GDB protocol remains happy.
* fixed crash when timing out connection to GDB
git-svn-id: svn://svn.berlios.de/openocd/trunk@445 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- Added simulate_reset_on_next_halt that can be extended to do all sorts
of cleanups for systems without proper reset. Right now it just writes 0
to the control register to disable caches.
- Step skips over Wait for Interrupt instruction
- fix for count
- fix for printf format errors
git-svn-id: svn://svn.berlios.de/openocd/trunk@439 b42882b7-edfa-0310-969c-e2dbd0fdcd60
Small cosmetic fixes in the license header to make them all look the
same, fix some typos, update README.
git-svn-id: svn://svn.berlios.de/openocd/trunk@396 b42882b7-edfa-0310-969c-e2dbd0fdcd60
More precise, less code.
- removed unused code in fileio.c
- Windows should now find debug_handler.bin
git-svn-id: svn://svn.berlios.de/openocd/trunk@390 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- open_file_from_path() now uses logging feature
- open_file_from_path() no longer uses paths for anything but "r" mode. Fixes
bug waiting to happen.
git-svn-id: svn://svn.berlios.de/openocd/trunk@389 b42882b7-edfa-0310-969c-e2dbd0fdcd60
arm11
--- Added burst memory transfer mode
This does not explicitly query command execution but rather uses a small
delay produced by the FT2232 on certain TAP commands.
A potential failure of this process is detected afterwards and the
program terminates with an error.
'arm11 memwrite burst disable'
can be used to switch this feature off.
'arm11 memwrite error_fatal disable'
can be used to prevent the program to exit on an memory write error
--- Added support for interrupt breaking via VCR register
Use 'arm11 vcr' command to set.
--- Cleaned up the handling of
halt/resume/step/poll, target->state, target->debug_reason,
target_call_event_callbacks() at least as far as I could guess the
intended behaviour from other targets.
Did some overall positive tests with GDB.
--- Added support for breakpoints
Hardware breakpoints only. All breakpoints will be treated as hardware
breakpoints.
All ARM11's seem to have at least 6 hardware breakpoints.
--- Stepping over BKPT added
Modification to PC without touching the target.
--- Stepping over a B or BL to self will do nothing
git-svn-id: svn://svn.berlios.de/openocd/trunk@385 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- fix indentation of multi-level commands
- make help command work with multi-level commands
git-svn-id: svn://svn.berlios.de/openocd/trunk@384 b42882b7-edfa-0310-969c-e2dbd0fdcd60
this patch fixes an issue in jtag_init - shall not reset jtag to NULL.
This causes jtag interface (USB device) not to be closed properly if jtag chain validation fails.
Once the underlaying jtag interface is initialized, jtag pointer shall be left untouched,
even in case of an error, so that proper cleanup can be done, see exit_handler in openocd.c
git-svn-id: svn://svn.berlios.de/openocd/trunk@372 b42882b7-edfa-0310-969c-e2dbd0fdcd60
provides a better impeadance match for debug output char fn's, e.g. eCos.
- Line endings are now added at the caller site of command_print*(). command_print()
still adds a line ending
- echo of commands in scripts are now available via debug_level instead of
forced echo
- Added a USER_SAMELINE() for printing without a lineend.
git-svn-id: svn://svn.berlios.de/openocd/trunk@364 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- bugfix in server.c
- removed unused parameter from jtag_add_ir_scan et al. This
wasn't necessary in hindsight but anyway.
- arm11 source committed but not not in Makefile.am/target.c for now.
git-svn-id: svn://svn.berlios.de/openocd/trunk@341 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- some more flash cleanup of checking halted state
- moved output handler into options.c
- very slightly tweaked server.c to make it a bit more compatible with eCos
- retired arch_state. Not quite sure how I managed to leave that out last time.
git-svn-id: svn://svn.berlios.de/openocd/trunk@338 b42882b7-edfa-0310-969c-e2dbd0fdcd60
command and reduce confusion with "flash write_image".
- retired deprecated "flash erase" & "flash write".
- added flash_driver_protect/write/erase() that are wafer thin frontend
functions to low level driver functions. They implement checks
that were inconsistently handled by the drivers, e.g. check for
target halted was done in a spotty fashion.
- use return ERROR_COMMAND_SYNTAX_ERROR to print out
syntax of command instead of having lots of inlined replicas of
the command line syntax(some of which were wrong).
- use logging instead of dubious translation of error values to
human understandable explanations of why things failed.
The lower levels log the precise reason and the higher
levels can ammend context as the error propagates up
the call stack.
- simplified flash API slightly with logging instead of
allocating and returning information that the caller then
has to translate into print statements.
git-svn-id: svn://svn.berlios.de/openocd/trunk@337 b42882b7-edfa-0310-969c-e2dbd0fdcd60
access by disabling "unecessary" checks.
- arm926ejs. Added missing type->mmu and type->virt2phys fn's.
for now these are used by working_area when specifying
mmu enabled and mmu not enabled address
git-svn-id: svn://svn.berlios.de/openocd/trunk@335 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- added arch_state to show status of currently selected target
- simplified target->type->arch_state() api.
- clean up telnet output a bit
- fixed GDB output for arch_state
- removed a couple of unecessary exit()'s
- cleaned up error propagation a bit in a few places
git-svn-id: svn://svn.berlios.de/openocd/trunk@332 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- added virtual address to working_area.
- Improved error messages in a number of places
- Added ERROR_COMMAND_SYNTAX_ERROR that commands can return to have syntax printed
- Added help for some config commands
- Added verification of sw breakpoints with ERROR() message
- Removed a couple of exit()'s and replaced with error message
- cosmetic fix to armv4_5.c, easier to read
- added polymorphic(with default) virt2phys and mmu enable query function to target.h
- added virt2phys command that uses target->type->virt2phys() fn
Thanks to Øyvind Harboe
git-svn-id: svn://svn.berlios.de/openocd/trunk@310 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- add configuration option "gdb_report_data_abort <enable|disable> to enable reporting of data aborts during gdb memory read packets
- thanks to Øyvind Harboe for identifying the problem with GDB and data aborts
- fix some warnings in gdb_server.c
git-svn-id: svn://svn.berlios.de/openocd/trunk@307 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- Allow all commands to be executed during config stage
- Help now works for config commands
- make var args handling follow the rules more closely
Thanks Øyvind Harboe
git-svn-id: svn://svn.berlios.de/openocd/trunk@305 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- now uses jtag_state_pathmove() instead of making assumptions about implementation of jtag_statemove().
- fixed a couple of bugs in timeout handling
- removed megabytes of log output when communication is failing.
- sleep is now 300ms as documented instead of 3000ms
- fixed error path of bulk write
- debug_handler can now be issued during normal operation + has help text.
Thanks Øyvind Harboe
git-svn-id: svn://svn.berlios.de/openocd/trunk@304 b42882b7-edfa-0310-969c-e2dbd0fdcd60
- Faster DEBUG/INFO() when they are disabled
- target_read/write_buffer() now uses 16 and 32 bit access for single word aligned requests. Other requests are serviced as quickly as possible.
- *much* faster read/write GDB packets, removing timeout problems.
- GDB read/write packets w/single word aligned 32/16 bit access now use 32/16 bit word access.
- working area can now be changed on the fly. Provides a way to move working area about as MMU is enabled/disabled.
- cleaned up error messages for verify_image.
Thanks Øyvind Harboe
git-svn-id: svn://svn.berlios.de/openocd/trunk@302 b42882b7-edfa-0310-969c-e2dbd0fdcd60
oharboe
ntfreak
drath
vpalatin
mifi