Nick Foster wrote:
> Thanks to both of you for quick and effective responses! Everything
> seems OK now on rev 986.
>
More importantly - thank you for confirming the problem is resolved.
Often I wonder if somebody figured it out - or if they gave up. You did
exactly what Eric Raymond's famous
Hi,
I tested with IAR EWARM 5.11KS thru GDB on LPC2138.
If I debug in Flash, everything is OK.
But if I debug in RAM, svn later than 919 will fail.
With svn919, it's OK but EWARM will use single-stepping due to lack of
breakpoints.
With svn 920, it's ONLY OK for the first time, openocd need to be
Thanks to both of you for quick and effective responses! Everything
seems OK now on rev 986.
Nick
On Fri, 2008-09-12 at 12:53 +0200, Øyvind Harboe wrote:
> On Fri, Sep 12, 2008 at 12:49 PM, Duane Ellis <[EMAIL PROTECTED]> wrote:
> > the tail end of 'process reset' needs to flush the target callba
On Fri, Sep 12, 2008 at 12:49 PM, Duane Ellis <[EMAIL PROTECTED]> wrote:
> the tail end of 'process reset' needs to flush the target callbacks.
> The new patch assumes this was done in 'keep_alive()"
>
> W/ the keep alive fix this needs to be added.
Committed.
Thanks!
--
Øyvind Harboe
http://w
the tail end of 'process reset' needs to flush the target callbacks.
The new patch assumes this was done in 'keep_alive()"
W/ the keep alive fix this needs to be added.
-Duane.
Index: src/target/target.c
===
--- src/target/targ
>Was caused by Oyvinds recent changes dealing with "keep_alive".
Actually I think the problem was made *worse* by my latest patch rather than
introduced by it.
>This has introduced a problem, keep_alive() needs a recursion block.
I've committed a patch that has reduced the job of keep_al
Committed.
- Fixed regression introduced in 890 when "fixing warnings" for
target_call_timer_callbacks_now().
target_call_timer_callbacks_now() did the same as target_call_timer_callbacks().
- Reduced keep_alive()'s job to only deal with GDB keep alive problems.
Index: C:/workspace/openocd/src
