openocd.git
10 years agoMerge commit 'origin/master'
Øyvind Harboe [Mon, 12 Oct 2009 07:28:56 +0000 (09:28 +0200)]
Merge commit 'origin/master'

10 years agoIf halt times out, stop GDB. Allows e.g. manual reset via monitor commands.
Øyvind Harboe [Mon, 12 Oct 2009 07:27:27 +0000 (09:27 +0200)]
If halt times out, stop GDB. Allows e.g. manual reset via monitor commands.

10 years agoSupply default reset_config statement to make target scripts useful standalone and...
Øyvind Harboe [Mon, 12 Oct 2009 07:25:08 +0000 (09:25 +0200)]
Supply default reset_config statement to make target scripts useful standalone and provide sensible default

10 years agoxscale_load_ic cleanup
David Brownell [Sun, 11 Oct 2009 17:35:52 +0000 (10:35 -0700)]
xscale_load_ic cleanup

Remove unused and deprecated (in the arch spec) mode for loading
code into the *main* icache (vs the "mini" icache).  Disable some
extremely noisy (and rarely useful) low-level debug messages

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoxscale.c cleanup
David Brownell [Sun, 11 Oct 2009 17:35:28 +0000 (10:35 -0700)]
xscale.c cleanup

Declare almost everything as static.
Move stuff to remove most forward references.
Remove most forward declarations.
Warn if the unimplemented register functions get called.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoxscale bugfix to handler loading
David Brownell [Sun, 11 Oct 2009 17:06:08 +0000 (10:06 -0700)]
xscale bugfix to handler loading

Just fill out the rest of the cache line with NOPs; don't change
the record of how much data we consumed.  Otherwise the count of
how much data is left can roll over from positive to negative
("VERY positive") and skip the loop termination of zero.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoxscale minor cleanup
David Brownell [Sun, 11 Oct 2009 09:52:00 +0000 (02:52 -0700)]
xscale minor cleanup

Add a header comment referencing useful XScale specs.
Make most data static, and the tables readonly.
Scrub extra blank lines.
Return fault codes from one routine.
Remove a needless NOP methood.

(BUGFIX) When we update R0, mark R0 as dirty/valid ... not R15/PC!

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoprintf format warning fixes
David Brownell [Sat, 10 Oct 2009 18:32:39 +0000 (11:32 -0700)]
printf format warning fixes

Observed on a Cygwin build.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoFix reset delays and tinker with ID's
Wookey [Sat, 10 Oct 2009 07:08:06 +0000 (09:08 +0200)]
Fix reset delays and tinker with ID's

10 years agotweak new "translating ..." text
David Brownell [Fri, 9 Oct 2009 22:51:16 +0000 (15:51 -0700)]
tweak new "translating ..." text

Fix formatting and layout bugs in the new "translating configuration
files" bit.  Make it a section within the chapter about config files.
Add a crossreference.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoadd documentation about reset customization
David Brownell [Fri, 9 Oct 2009 19:52:42 +0000 (12:52 -0700)]
add documentation about reset customization

We added two overridable procedures; document them, and the
two jtag arp_* operations they necessarily expose.

Update the comment about the jtag_init_reset() routine; it's
been obsolete for as long as it's had SRST support.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoStop debug session if halt fails
Øyvind Harboe [Fri, 9 Oct 2009 08:31:40 +0000 (10:31 +0200)]
Stop debug session if halt fails

10 years agoARM11 error checking
Øyvind Harboe [Fri, 9 Oct 2009 08:00:05 +0000 (10:00 +0200)]
ARM11 error checking

10 years agoMerge commit 'origin/master'
Øyvind Harboe [Fri, 9 Oct 2009 07:14:27 +0000 (09:14 +0200)]
Merge commit 'origin/master'

10 years agoAdded tip in documentation on how to translate quirky syntax
Øyvind Harboe [Fri, 9 Oct 2009 07:03:53 +0000 (09:03 +0200)]
Added tip in documentation on how to translate quirky syntax

10 years agomake PXA255 targets enumerate sort-of-OK
David Brownell [Fri, 9 Oct 2009 06:51:50 +0000 (23:51 -0700)]
make PXA255 targets enumerate sort-of-OK

Startup now mostly works, except that the initial target state
is "unknown" ... previously, it refused to even start.

Getting that far required fixing the ircapture value (which
can never have been correct!) and the default JTAG clock rate,
then providing custom reset script.

The "reset" command is still iffy.  DCSR updates, and loading
the debug handler, report numerous DR/IR capture failures.
But once that's done, "poll" reports that the CPU is halted
(which it shouldn't be, this was "reset run"!), due to the
rather curious reason "target-not-halted".

Summary:  you still can't debug these parts, but it's closer.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoadd overridable Tcl "init_reset"
David Brownell [Fri, 9 Oct 2009 06:51:50 +0000 (23:51 -0700)]
add overridable Tcl "init_reset"

This abstracts the "jtag arp_init-reset" call into a method
called from OpenOCD startup and reset processing.

Platforms which have different requirements for how such hard
resets must be performed can now override "init_reset" instead
of needing to rebuild custom hacked versions of the server.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agominor JTAG init messaging tweaks
David Brownell [Fri, 9 Oct 2009 06:51:50 +0000 (23:51 -0700)]
minor JTAG init messaging tweaks

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoFunction to flash SheevaPlug u-boot sectors
Rabeeh Khoury [Fri, 9 Oct 2009 00:10:52 +0000 (17:10 -0700)]
Function to flash SheevaPlug u-boot sectors

This function is used by the SheevaPlug installer to flash the
erase and re-flash the U-Boot environment in the NAND Flash.

10 years agoinitial builds of OSK5912 boards need srst_pulls_trst
David Brownell [Thu, 8 Oct 2009 20:15:18 +0000 (13:15 -0700)]
initial builds of OSK5912 boards need srst_pulls_trst

This is clearly noted in the hardware spec (section 5.2.3); it
works around a chip erratum:  "If the MPU_RESET signal is used,
it may cause the EMIFS bus to lock."

I seem to have a board with such an initial build.  The chip
is labeled XOMAP.  Presumably, parts without that "X" prefix
(eXperimental) resolve this.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoprevent abort via polling during jtag_reset
David Brownell [Thu, 8 Oct 2009 18:14:00 +0000 (11:14 -0700)]
prevent abort via polling during jtag_reset

Observed:

  openocd: core.c:318: jtag_checks: Assertion `jtag_trst == 0' failed.

The issue was that nothing disabled background polling during calls
from the TCL shell to "jtag_reset 1 1".  Fix by moving the existing
poll-disable mechanism to the JTAG layer where it belongs, and then
augmenting it to always pay attention to TRST and SRST.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoStop ignoring most scan chain validation errors
David Brownell [Thu, 8 Oct 2009 16:16:01 +0000 (09:16 -0700)]
Stop ignoring most scan chain validation errors

Among other things this causes startup errors to kick in the
fallback "reset harder" logic during server startup.  Comments
are also updated a bit, explaining what the various error paths
signify (in at least my observation).

There's one class of validation error that we can still plausibly
ignore:  when wrong IDCODE values are observed.

This change seems to have helped make an OMAP5912 behave much
more reliably.  There's still some post-reset flakiness, but
it's unrelated to scan verification.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoMerge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd...
David Brownell [Thu, 8 Oct 2009 14:32:47 +0000 (07:32 -0700)]
Merge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd/openocd

10 years agoTesting committing changes to .gitignore to ignore more build output
Øyvind Harboe [Thu, 8 Oct 2009 13:43:51 +0000 (15:43 +0200)]
Testing committing changes to .gitignore to ignore more build output

10 years agoAdd .project to .gitignore
Øyvind Harboe [Thu, 8 Oct 2009 13:27:09 +0000 (15:27 +0200)]
Add .project to .gitignore

10 years agoStop GDB when polling fails, srst assert or powerdropout is detected
Øyvind Harboe [Thu, 8 Oct 2009 12:53:25 +0000 (14:53 +0200)]
Stop GDB when polling fails, srst assert or powerdropout is detected

10 years agoUpdate copyright statements. Make it easier to sync with Jim Tcl
Øyvind Harboe [Thu, 8 Oct 2009 11:56:57 +0000 (13:56 +0200)]
Update copyright statements. Make it easier to sync with Jim Tcl

10 years agomake OMAP5912 resets more reliable
David Brownell [Thu, 8 Oct 2009 06:29:18 +0000 (23:29 -0700)]
make OMAP5912 resets more reliable

Without some extra delay after releasing SRST, we seemed to
be trying to talk to the TAP before it was ready to respond.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoiMX25 target support
John Rigby [Thu, 8 Oct 2009 04:19:32 +0000 (22:19 -0600)]
iMX25 target support

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
10 years agoHouston, we have Mirror!
David Brownell [Thu, 8 Oct 2009 02:11:37 +0000 (19:11 -0700)]
Houston, we have Mirror!

10 years agoMerge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd...
David Brownell [Thu, 8 Oct 2009 00:14:58 +0000 (17:14 -0700)]
Merge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd/openocd

10 years agoChange most in-tree references from SVN to GIT.
dbrownell [Thu, 8 Oct 2009 00:13:50 +0000 (00:13 +0000)]
Change most in-tree references from SVN to GIT.
Also, talk about "mainline" not "trunk".

The release.txt and release.sh files need more updates.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2825 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoForce sane SRST and TRST initialization
dbrownell [Wed, 7 Oct 2009 19:02:21 +0000 (19:02 +0000)]
Force sane SRST and TRST initialization

At least some FT2232 based adapters don't necessarily come up
in the expected state, with SRST and TRST disabled.  Since
other adapters could suffer the same problem, let's avoid
needing to patch every driver and just force *all* adapters
to initialize those values properly at server startup.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2824 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoBetter fix for TAPs violating the JTAG spec for IR-Capture.
dbrownell [Wed, 7 Oct 2009 18:51:11 +0000 (18:51 +0000)]
Better fix for TAPs violating the JTAG spec for IR-Capture.

Instead of just assuming all IDCODE-deprived TAPs violate the
JTAG spec (they don't!), just require TAPs with such problems
to be declared with proper ircapture/irmask values.  Example,
with mask and value of zero.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2823 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoRemove much #ifdeffery around _DEBUG_JTAG_IO_ usage.
dbrownell [Wed, 7 Oct 2009 17:30:51 +0000 (17:30 +0000)]
Remove much #ifdeffery around _DEBUG_JTAG_IO_ usage.
Have DEBUG_JTAG_IO() always trigger necessary warnings.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2822 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agobuildfix
dbrownell [Wed, 7 Oct 2009 16:28:44 +0000 (16:28 +0000)]
buildfix

git-svn-id: svn://svn.berlios.de/openocd/trunk@2821 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMerge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd...
David Brownell [Wed, 7 Oct 2009 16:16:39 +0000 (09:16 -0700)]
Merge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd/openocd

10 years agoRemove pointless "target library" chapter.
dbrownell [Wed, 7 Oct 2009 16:15:21 +0000 (16:15 +0000)]
Remove pointless "target library" chapter.

It had a very little bit of content; move that to the more extensive
chapter on config file guidelines, and give more current "ls" output
to show the available library code.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2820 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoNote bug in handling of variables through command line parameters.
dbrownell [Wed, 7 Oct 2009 15:56:11 +0000 (15:56 +0000)]
Note bug in handling of variables through command line parameters.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2819 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoUpdates for "reset_config":
dbrownell [Wed, 7 Oct 2009 15:31:33 +0000 (15:31 +0000)]
Updates for "reset_config":

 - revert to previous default: don't talk JTAG during SRST
 - add "srst_nogates" flag, the converse of "srst_gates_jtag"
 - with no args, display the current configuration

And update the User's Guide text with bullet lists to be a bit more clear.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2818 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agofirst stab at imx35 reset init script
oharboe [Wed, 7 Oct 2009 12:21:58 +0000 (12:21 +0000)]
first stab at imx35 reset init script

git-svn-id: svn://svn.berlios.de/openocd/trunk@2817 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoremove recrusive reset invocation from reset init callback
oharboe [Wed, 7 Oct 2009 12:20:06 +0000 (12:20 +0000)]
remove recrusive reset invocation from reset init callback

git-svn-id: svn://svn.berlios.de/openocd/trunk@2816 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoadd timeouts and fix syntax error handling of mrc/mcr commands.
oharboe [Wed, 7 Oct 2009 12:19:17 +0000 (12:19 +0000)]
add timeouts and fix syntax error handling of mrc/mcr commands.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2815 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoadd timeouts and fix syntax error handling of mrc/mcr commands.
oharboe [Wed, 7 Oct 2009 12:05:37 +0000 (12:05 +0000)]
add timeouts and fix syntax error handling of mrc/mcr commands.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2814 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoincrease pause before reboot so web interface remains responsive when issuing a reboo...
oharboe [Wed, 7 Oct 2009 11:03:56 +0000 (11:03 +0000)]
increase pause before reboot so web interface remains responsive when issuing a reboot of zy1000

git-svn-id: svn://svn.berlios.de/openocd/trunk@2813 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoDo not check ir capture if there is no IDCODE
oharboe [Wed, 7 Oct 2009 11:03:01 +0000 (11:03 +0000)]
Do not check ir capture if there is no IDCODE

git-svn-id: svn://svn.berlios.de/openocd/trunk@2812 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years ago1.55 snapshot
oharboe [Wed, 7 Oct 2009 06:53:44 +0000 (06:53 +0000)]
1.55 snapshot

git-svn-id: svn://svn.berlios.de/openocd/trunk@2810 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoChange version labels to work better with GIT
dbrownell [Wed, 7 Oct 2009 05:43:05 +0000 (05:43 +0000)]
Change version labels to work better with GIT

 - The guess-rev.sh script is now a tweaked version of "setlocalversion" as
   seen in Linux, U-Boot, and various other projects.  When it finds source
   control support (git, hg, svn) it uses IDs from there.  Else (specific
   to this project) it reports itself as "-snapshot", e.g. from gitweb.

   I verified this new "guess-rev.sh" script runs under Cygwin.

 - Also update the generic version strings to be like "0.3.0-dev" (during
   development) instead of the very long "0.3.0-in-development".  These also
   show up in the PDF docs.  For better tracking, we might eventually change
   these strings to include the version IDs too.

 - Change the startup banner version strings so they include the guess-rev
   output.  Development and release versions with GIT will be like

    Open On-Chip Debugger 0.3.0-dev-00282-g7191a4f-dirty (2009-10-05-20:57)
    Open On-Chip Debugger 0.3.0 (2009-10-05-20:57)

   instead of the previous SVN-specific (even when using git-svn!)

    Open On-Chip Debugger 0.3.0-in-development (2009-10-05-01:39) svn:exported
    Open On-Chip Debugger 0.3.0 (2009-10-05-01:39) Release

git-svn-id: svn://svn.berlios.de/openocd/trunk@2809 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMerge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd...
David Brownell [Tue, 6 Oct 2009 22:58:06 +0000 (15:58 -0700)]
Merge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd/openocd

10 years agoAdd basic support for DM6446 EVM board.
dbrownell [Tue, 6 Oct 2009 22:56:52 +0000 (22:56 +0000)]
Add basic support for DM6446 EVM board.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2808 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoDragonite has the same EICE affliction as feroceon.
dbrownell [Tue, 6 Oct 2009 22:55:46 +0000 (22:55 +0000)]
Dragonite has the same EICE affliction as feroceon.

From: Nicolas Pitre <nico@fluxnic.net>

git-svn-id: svn://svn.berlios.de/openocd/trunk@2807 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoDragonite target script
dbrownell [Tue, 6 Oct 2009 22:55:04 +0000 (22:55 +0000)]
Dragonite target script

From: Nicolas Pitre <nico@fluxnic.net>

git-svn-id: svn://svn.berlios.de/openocd/trunk@2806 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMerge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd...
David Brownell [Tue, 6 Oct 2009 15:29:23 +0000 (08:29 -0700)]
Merge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd/openocd

10 years agoIntroduced jtag_init and "jtag arp_init" to allow target scripts more control over...
oharboe [Tue, 6 Oct 2009 08:10:57 +0000 (08:10 +0000)]
Introduced jtag_init and "jtag arp_init" to allow target scripts more control over how OpenOCD starts up and initializes the target.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2805 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agostop using targetnum
oharboe [Tue, 6 Oct 2009 08:10:05 +0000 (08:10 +0000)]
stop using targetnum

git-svn-id: svn://svn.berlios.de/openocd/trunk@2804 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoTake "patches" out of .gitignore ... Win32 is case-crazy
dbrownell [Tue, 6 Oct 2009 05:28:30 +0000 (05:28 +0000)]
Take "patches" out of .gitignore ... Win32 is case-crazy

git-svn-id: svn://svn.berlios.de/openocd/trunk@2803 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMerge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd...
David Brownell [Mon, 5 Oct 2009 23:49:51 +0000 (16:49 -0700)]
Merge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd/openocd

10 years agoAdd initial ".gitignore".
dbrownell [Mon, 5 Oct 2009 23:45:14 +0000 (23:45 +0000)]
Add initial ".gitignore".

git-svn-id: svn://svn.berlios.de/openocd/trunk@2802 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMerge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd...
David Brownell [Mon, 5 Oct 2009 08:27:43 +0000 (01:27 -0700)]
Merge branch 'master' of ssh://dbrownell@openocd.git.sourceforge.net/gitroot/openocd/openocd

10 years agoImprove jtag_validate_ircapture() diagnostics.
dbrownell [Mon, 5 Oct 2009 08:23:33 +0000 (08:23 +0000)]
Improve jtag_validate_ircapture() diagnostics.

Bugfix the error message so it shows the disliked value, and add
a debug message showing each TAP's IR capture value, all N bits.

This just changes diagnostics ... it still ignores the parameters
given to us at TAP declaration time.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2801 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoAdd a new JTAG "setup" event; use for better DaVinci ICEpick support.
dbrownell [Mon, 5 Oct 2009 08:20:28 +0000 (08:20 +0000)]
Add a new JTAG "setup" event; use for better DaVinci ICEpick support.

The model is that this fires after scanchain verification, when it's
safe to call "jtag tapenable $TAPNAME".  So it will fire as part of
non-error paths of "init" and "reset" command processing.  However it
will *NOT* trigger during "jtag_reset" processing, which skips all
scan chain verification, or after verification errors.

ALSO:
 - switch DaVinci chips to use this new mechanism
 - log TAP activation/deactivation, since their IDCODEs aren't verified
 - unify "enum jtag_event" scripted event notifications
 - remove duplicative JTAG_TAP_EVENT_POST_RESET

git-svn-id: svn://svn.berlios.de/openocd/trunk@2800 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMinor cleanup to ARM926 debug entry:
dbrownell [Mon, 5 Oct 2009 08:18:17 +0000 (08:18 +0000)]
Minor cleanup to ARM926 debug entry:

 - don't needlessly export this function
 - handle "case 0" debug method-of-entry better (silent by default)

The "case 0" is a valid debug entry mode so it doesn't deserve the
warning int now gets.  But it probably means that OpenOCD confused
itself somehow; or that it confused the ARM9EJS target.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2799 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoUpdate the NEWS file to cover more of the user-visible changes.
dbrownell [Mon, 5 Oct 2009 08:14:57 +0000 (08:14 +0000)]
Update the NEWS file to cover more of the user-visible changes.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2798 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoGet rid of needless OMAP and Davinci target config options
dbrownell [Mon, 5 Oct 2009 08:13:00 +0000 (08:13 +0000)]
Get rid of needless OMAP and Davinci target config options
so they provide better examples and are easier to maintain.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2797 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoUpdated reset event handling in omap3530 cfg
mlu [Fri, 2 Oct 2009 16:52:02 +0000 (16:52 +0000)]
Updated reset event handling in omap3530 cfg

git-svn-id: svn://svn.berlios.de/openocd/trunk@2796 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoIt is not possible to invalidate I-Cache on memory writes while the target is running
mlu [Fri, 2 Oct 2009 16:50:42 +0000 (16:50 +0000)]
It is not possible to invalidate I-Cache on memory writes while the target is running

git-svn-id: svn://svn.berlios.de/openocd/trunk@2795 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMake sure that DSCR_DTR_RX is not full before writing
mlu [Fri, 2 Oct 2009 16:44:16 +0000 (16:44 +0000)]
Make sure that DSCR_DTR_RX is not full before writing

git-svn-id: svn://svn.berlios.de/openocd/trunk@2794 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMore error reporting in Cortex_a8 execute_opcode
mlu [Fri, 2 Oct 2009 16:39:50 +0000 (16:39 +0000)]
More error reporting in Cortex_a8 execute_opcode

git-svn-id: svn://svn.berlios.de/openocd/trunk@2793 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoAdded asser_reset and deassert_reset for cortex_a8
mlu [Fri, 2 Oct 2009 16:37:22 +0000 (16:37 +0000)]
Added asser_reset and deassert_reset for cortex_a8

git-svn-id: svn://svn.berlios.de/openocd/trunk@2792 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoAdded asser_reset and deassert_reset for cortex_a8
mlu [Fri, 2 Oct 2009 16:36:03 +0000 (16:36 +0000)]
Added asser_reset and deassert_reset for cortex_a8

git-svn-id: svn://svn.berlios.de/openocd/trunk@2791 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMinor ETB and ETM bugfixes and doc updates
dbrownell [Fri, 2 Oct 2009 09:19:03 +0000 (09:19 +0000)]
Minor ETB and ETM bugfixes and doc updates

 - ETB
    * report _actual_ hardware status, not just expected status
    * add a missing diagnostic on a potential ETB setup error
    * prefix any diagnostics with "ETB"
 - ETM
    * make "etm status" show ETM hardware status too, instead of
      just traceport status (which previously was fake, sigh)
 - Docs
    * flesh out "etm tracemode" docs a bit
    * clarify "etm status" ... previously it was traceport status
    * explain "etm trigger_percent" as a *traceport* option

ETM+ETB tracing still isn't behaving, but now I can see that part of
the reason is that the ETB turns itself off almost immediately after
being enabled, and before collecting any data.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2790 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoARMv7A: Report fault status registers when in Abort state
mlu [Thu, 1 Oct 2009 17:39:13 +0000 (17:39 +0000)]
ARMv7A: Report fault status registers when in Abort state

git-svn-id: svn://svn.berlios.de/openocd/trunk@2789 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoUpdate release process documentation.
zwelch [Wed, 30 Sep 2009 23:20:52 +0000 (23:20 +0000)]
Update release process documentation.
- Improve and clarify the wording of the introduction.
- Add section on version taggging.
- Some other minor corrections.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2788 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoImprove release script version tag management commands.
zwelch [Wed, 30 Sep 2009 23:20:43 +0000 (23:20 +0000)]
Improve release script version tag management commands.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2787 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoAdd numeric version tag support to release script.
zwelch [Wed, 30 Sep 2009 23:20:36 +0000 (23:20 +0000)]
Add numeric version tag support to release script.

The release process itself does not use this support yet, but it allows
packagers to automate the process of managing their own tags, if they
patch the source code before releasing binaries.  The release processes
should be revised to incorporate this feature to support -rc packages.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2786 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoImprove release script version handling commands.
zwelch [Wed, 30 Sep 2009 23:20:20 +0000 (23:20 +0000)]
Improve release script version handling commands.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2785 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoAdd workaround to release script to update source code URL keyword.
zwelch [Wed, 30 Sep 2009 23:20:13 +0000 (23:20 +0000)]
Add workaround to release script to update source code URL keyword.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2784 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoDocument and automate signature creation for the release archives.
zwelch [Wed, 30 Sep 2009 23:20:06 +0000 (23:20 +0000)]
Document and automate signature creation for the release archives.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2783 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoFix release script bugs after experience from 0.2.0:
zwelch [Wed, 30 Sep 2009 23:19:53 +0000 (23:19 +0000)]
Fix release script bugs after experience from 0.2.0:
- The NEWS step failed due to an empty commit comment.
- The final release step would have failed, because the steps to switch
  from the secure to insecure repository (and back again) require both
  switch and a URL relocation steps

git-svn-id: svn://svn.berlios.de/openocd/trunk@2782 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoRemove annoying EOL whitespace (again, sigh).
dbrownell [Wed, 30 Sep 2009 23:02:15 +0000 (23:02 +0000)]
Remove annoying EOL whitespace (again, sigh).

git-svn-id: svn://svn.berlios.de/openocd/trunk@2781 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoAdd DSCR_DTR_RX_FULL bit define
mlu [Wed, 30 Sep 2009 17:24:43 +0000 (17:24 +0000)]
Add DSCR_DTR_RX_FULL bit define

git-svn-id: svn://svn.berlios.de/openocd/trunk@2780 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agostrip gdb config options
oharboe [Wed, 30 Sep 2009 05:40:40 +0000 (05:40 +0000)]
strip gdb config options

git-svn-id: svn://svn.berlios.de/openocd/trunk@2779 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agomichal smulski <michal.smulski@ooma.com> reset now works
oharboe [Wed, 30 Sep 2009 05:37:32 +0000 (05:37 +0000)]
michal smulski <michal.smulski@ooma.com> reset now works

git-svn-id: svn://svn.berlios.de/openocd/trunk@2778 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoStreamline Capture-IR validation code
dbrownell [Tue, 29 Sep 2009 19:28:08 +0000 (19:28 +0000)]
Streamline Capture-IR validation code

 - Don't issue needless JTAG resets ... only do them after
   errors. Normal exit now leaves every TAP in BYPASS.

 - Fix an unlikely memory leak on one fault path.

 - Remove the oddball limitation that invalid capture LSBs
   trigger errors only for TAPs that support IDCODE.

Re the JTAG reset:  there are too many of them, and they can
(and do!) change system state.  So the needless ones should
get removed.  This one was especially pointless.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2777 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoARM11 command handling fixes
dbrownell [Tue, 29 Sep 2009 18:30:06 +0000 (18:30 +0000)]
ARM11 command handling fixes

 - Commands were supposed to have been "arm11 memwrite ..."
   not "memwrite ..."
 - Get rid of obfuscatory macros
 - Re-alphabetize
 - Add docs for "arm11 vcr"

git-svn-id: svn://svn.berlios.de/openocd/trunk@2776 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoMake "-expected-id 0" suppress warnings; not unlike it used to do.
dbrownell [Tue, 29 Sep 2009 18:26:18 +0000 (18:26 +0000)]
Make "-expected-id 0" suppress warnings; not unlike it used to do.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2775 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoDoc updates: add section on target software changes, minor fixes
dbrownell [Tue, 29 Sep 2009 18:20:30 +0000 (18:20 +0000)]
Doc updates:  add section on target software changes, minor fixes

git-svn-id: svn://svn.berlios.de/openocd/trunk@2774 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoETM: fix build issue on MinGW.
dbrownell [Tue, 29 Sep 2009 18:08:16 +0000 (18:08 +0000)]
ETM: fix build issue on MinGW.

The Win32 global namespace is rather cluttered...

git-svn-id: svn://svn.berlios.de/openocd/trunk@2773 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoETB: cleanup needless symbol exports and forward decls.
dbrownell [Tue, 29 Sep 2009 18:06:26 +0000 (18:06 +0000)]
ETB: cleanup needless symbol exports and forward decls.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2772 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agohandle single threading
oharboe [Tue, 29 Sep 2009 14:24:02 +0000 (14:24 +0000)]
handle single threading

git-svn-id: svn://svn.berlios.de/openocd/trunk@2771 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoreentry assert
oharboe [Tue, 29 Sep 2009 14:23:42 +0000 (14:23 +0000)]
reentry assert

git-svn-id: svn://svn.berlios.de/openocd/trunk@2770 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agohttpd smoketest info
oharboe [Tue, 29 Sep 2009 13:46:26 +0000 (13:46 +0000)]
httpd smoketest info

git-svn-id: svn://svn.berlios.de/openocd/trunk@2769 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoadded t/nsrst_assert_width commands
oharboe [Tue, 29 Sep 2009 06:07:50 +0000 (06:07 +0000)]
added t/nsrst_assert_width commands

git-svn-id: svn://svn.berlios.de/openocd/trunk@2768 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoif srst pulls trst, then set state to TAP_RESET.
oharboe [Mon, 28 Sep 2009 12:22:20 +0000 (12:22 +0000)]
if srst pulls trst, then set state to TAP_RESET.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2767 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agospelling fix
oharboe [Mon, 28 Sep 2009 12:13:49 +0000 (12:13 +0000)]
spelling fix

git-svn-id: svn://svn.berlios.de/openocd/trunk@2766 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoShrink symbols exported from arm9tdmi.c and remove a forward ref.
dbrownell [Mon, 28 Sep 2009 04:55:23 +0000 (04:55 +0000)]
Shrink symbols exported from arm9tdmi.c and remove a forward ref.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2765 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoAdd list of JTAG adapter drivers with TAP_RESET statemove bug.
dbrownell [Sun, 27 Sep 2009 17:08:42 +0000 (17:08 +0000)]
Add list of JTAG adapter drivers with TAP_RESET statemove bug.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2764 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoUpdate FT2232 driver so that it reliably enters TAP_RESET.
dbrownell [Sun, 27 Sep 2009 16:55:52 +0000 (16:55 +0000)]
Update FT2232 driver so that it reliably enters TAP_RESET.

When the OpenOCD server starts up it records its state as TAP_RESET,
even though it could be anything.  Then when it starts to examine
the scan chain, it calls jtag_add_tlr() which sees it doesn't have
any work to do, and so it does nothing.  This can make the next
operations fail because they start from the wrong TAP state...

Instead of caring about the current recorded state, always enter
TAP_RESET by forcing five clocks with TMS high.

(NOTE:  it seems most other JTAG adapter drivers have this same bug.)

git-svn-id: svn://svn.berlios.de/openocd/trunk@2763 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoDon't provide invalid OMAP5912 IR capture value/mask attributes
dbrownell [Sun, 27 Sep 2009 07:59:10 +0000 (07:59 +0000)]
Don't provide invalid OMAP5912 IR capture value/mask attributes

git-svn-id: svn://svn.berlios.de/openocd/trunk@2762 b42882b7-edfa-0310-969c-e2dbd0fdcd60

10 years agoOn DM355 EVM board, associate NAND chips with $_TARGETNAME
dbrownell [Sat, 26 Sep 2009 22:08:50 +0000 (22:08 +0000)]
On DM355 EVM board, associate NAND chips with $_TARGETNAME
instead of a target number.

git-svn-id: svn://svn.berlios.de/openocd/trunk@2761 b42882b7-edfa-0310-969c-e2dbd0fdcd60