Hello all! :-)

>From all of the above posts Laurent it looks that you are trying to do
the TRANSPORT job. Please take a look at transport infrastructure. The
interface simply needs to be available for other things to work, so
there is no point in having handle ready with unusable interface. I
can understand this might be helpful, but also surely it will bring
confusion - if we work on the interface that has opened handle but
uninitialized interface how can you tell what happens next if you want
to work with other things (ie. transport or target) as init() can only
return OK or FAIL. This is why we consider such approach dangerous -
initialization can only succeed or fail, it cannot succeed with handle
and fail with interface. This is very simple and unabiguous - if
init() succeeds there is an interface read for use, otherwise it is
not and nothing else cannot work.

Regarding the malloc() problem it is already solved in simpler manner.

Regarding the deinitialization this is the problem to fix next - all
FT2232-based devices can have the same deinitialization procedure
because there is the same chip inside. This should be also trivial to
fix, as all interfaces have common quit() function, so no need to
create the separate one for each layout.

Maybe the good solution would be to create separate amontec.c driver
file, where you will test all your functionalities separately from all
other drivers without danger of breaking stuff for others. If some
solutions turn out to be better I am sure they will be also applied to
other drivers :-)


The SWD is on its way, almost done, but I want to do it elegant way
and bring additional benefit from it. This will be general
infrastructure, so you don't need to create again your own stuff from
scratch. Also the deinitialization procedure would be common for swd
and jtag ft2232 interfaces (all pins Hi-Z).  The ft2232_swd interface
is ready. I am working with KT-LINK interface that has been already
verified with UrJTAG. You can always watch the changes and even build
the code from http://repo.or.cz/w/openocd/libswd.git I will start
separate thread in a moment :-)

Best regards :-)
Tomek

-- 
CeDeROM, SQ7MHZ, http://www.tomek.cedro.info
_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to