08:15 <Hammdist> JTAG scan chain interrogation failed: all ones <-- when attempting to connect to blue pill. this is bad right?

08:18 <Hammdist> Warn : 'srst_nogate' reset_config option is required

08:21 <Hammdist> srst_only separate srst_nogate srst_open_drain connect_assert_srst <-- fixed the latter but still "all ones"

08:24 <Hammdist> Warn : There are no enabled taps. AUTO PROBING MIGHT NOT WORK!! <-- is this expected?

08:29 <Hammdist> https://pastebin.com/raw/1hYBNBkR <-- the config I'm trying to use

08:34 <Hammdist> got connected, whew. bad wiring

08:42 <PaulFertser> Hammdist: when you're doing auto-detection, without target config, SRST won't be touched either way because JTAG doesn't include SRST at all.

08:43 <PaulFertser> Hammdist: and if you added -f target/stm32f1x.cfg then srst_nogate would be enabled automatically as the target config includes that.

11:00 <mrec> https://dpaste.com/7CQWTR2FF

11:00 <mrec> how do I know that the firmware is successful uploaded?

11:00 <mrec> it seems like every version of openocd introduces some significant changes here which breaks my firmware tool

11:00 <mrec> the firmware is uploaded in that case, in earlier versions it seems like there was an output "wrote ..."

11:00 <mrec> that seems to be gone now?

11:04 <Haohmaru> you have a tool which uses openocd "under the hood" to flash firmware?

11:05 <Haohmaru> and your tool parses the openocd output to figure out what happened?

11:05 <mrec> yes

11:06 <Haohmaru> tricky i guess... i've seen the output messages depend a lot on the openocd options and/or target

11:07 <mrec> first I thought the HW is dead but it's just the openocd output which differs

11:07 <Haohmaru> wait till the experts wake up

11:08 <PaulFertser> mrec: -c "program" should finish with an appropriate exit code, you're not supposed to parse OpenOCD output normally.

11:09 <PaulFertser> And "wrote" wasn't the right word to look for anyway, you're more interested in verify results I guess.

11:09 <Haohmaru> "exit code" <- the return value from main() ?

11:09 <Haohmaru> yeah "verify" is something.. "wrote" doesn't matter if your pen was out of ink

11:12 <PaulFertser> Hammdist: yes, the "exit status of the child" per wait(2).

11:13 <Haohmaru> moi ;P~

11:15 <Haohmaru> PaulFertser, ever "-c ..." command is executed one after the other unless one of them fails, then the program stops (returning something bad), and if all such commands finish properly it returns "good", right?

11:15 <Haohmaru> "good" probably == 0 (EXIT_SUCCESS, or what was it)

11:16 <karlp> so yes, use -c program... it has a verify and exit options...

12:47 <mrec> I do a full device check after uploading the firmware, so if something's wrong I will see it in the endcheck

12:58 <PaulFertser> It shouldn't harm to verify flash contents after writing, and occassionally can help.

12:59 <Haohmaru> verifying should be the default, skipping verification should be done with an additional option (IMO)

13:01 <PaulFertser> Send a patch

13:10 <Hammdist> PaulFertser: did you check my config? I added a reset_config line. it should be issuing reset when probing (at least I'd like it to). won't it?

13:14 <PaulFertser> Hammdist: not without target defined, as JTAG has no notion of system reset.

17:51 <borneoa___> Hammdist: without a target, you can use 'adapter assert srst' and 'adapter deassert srst', but only after 'init'