|
From: Andreas F. <and...@gm...> - 2016-05-31 22:32:22
|
On Tue, May 31, 2016 at 5:13 PM, Uwe Bonnes < bo...@el...> wrote: > Hello, > > what is the reason to have all this --expected-id stuff in the > configuration > files. If we have a problem in these number, like at the moment in > stm32f7x.cfg, it is hard to see the problem for devellopers with SWD-only > debuggers or debugger targets. > > >From src/jtag/core.c, I see that if no --expected-id is provided, > jtag_examine_chain_match_tap() also returns success. > > For a production environment, I see a purpose for --expected-id, but for > the > distributed configuration this option is hard and tedious to maintain. > > So I propose to remove all --expected-id from our distributed > configurations. > > A similar discussion started around http://openocd.zylin.com/#/c/3377/ but > to no definite outcome. > > > Also see http://openocd.zylin.com/3385 I don't mind removing the BSTAPID blocks from STM32 configs, since the BS TAP is rarely used with OpenOCD, so complaining about a mismatch is mostly an annoyance. I wouldn't go so far as to remove *all* -expected-ids from *all* target configs. /Andreas |