Øyvind Harboe wrote:
> I'm seing pitiful performance when programming STM32 flash(9kBytes/s).
>
> It appears OpenOCD uses most of it's time in target_wait_state(), haven't
> really done much investigations yet.
>
> Has anyone done any work / measurements on Cortex target_write_memory()/
> flash p
On Tue, Feb 24, 2009 at 7:06 AM, Rob Brown wrote:
> Øyvind Harboe wrote:
>> I'm seing pitiful performance when programming STM32 flash(9kBytes/s).
>>
>> It appears OpenOCD uses most of it's time in target_wait_state(), haven't
>> really done much investigations yet.
>>
>> Has anyone done any work
I'm seing pitiful performance when programming STM32 flash(9kBytes/s).
It appears OpenOCD uses most of it's time in target_wait_state(), haven't
really done much investigations yet.
Has anyone done any work / measurements on Cortex target_write_memory()/
flash programming performance they would l
On Mon, Feb 23, 2009 at 10:25 PM, Joern Kaipf wrote:
> Rick Altherr wrote:
>
>> You should be able to handle this like all the other chips that have
>> RTCK. They use 'jtag_khz 0'.
>
> I think I need to buy new glasses ;) Thanks for this hint, I saw what
> you meant and did it that way.
Also try
Holger Schurig wrote:
> For example, my Olimex ARM-USB-TINY has ftdic.type == 2 and
> ftdi_read_chipid() == 0xa5abd789.
Okay, they use different number for the chip in the libftdi. I looked
into the git repo and it seems that they do not have support for the new
FTDI devices at the moment.
Joe
Committed.
tinkered a bit with performance for Cortex flash programming. Mainly
make it easier to profile as a start.
### Eclipse Workspace Patch 1.0
#P openocd
Index: src/target/cortex_swjdp.c
===
--- src/target/cortex_swjdp.c (r
Rick Altherr wrote:
> You should be able to handle this like all the other chips that have
> RTCK. They use 'jtag_khz 0'.
I think I need to buy new glasses ;) Thanks for this hint, I saw what
you meant and did it that way.
> form of intelligence in the driver about what device it is talking t
When I connected to my target and issue the "shutdown" command
via telnet, my target immediately freezes. However, I'd like to
use my target device, no matter if openOCD is debugging it or
not.
Ctrl-C doesn't send the target to nirvana, thought.
I tracked this down to a ftdi_disable_bitbang()
On Sunday 22 February 2009 22:50:35 Joern Kaipf wrote:
> - A TCK of 30 Mhz is now possible and that's my main problem.
> How to handle this. I thought about three options:
>
...
> * try to implement a sort of autodetect. I currently do not
> know if it is possible, just an idea.
I assume you use l
