06:49 <PlasmaHH> PaulFertser: guess the idea of flipping the default value for the gdb all z ero option needs some more persuasion ... sadly I was unable to dig up the corresponding gdb bugreport..

06:50 <PlasmaHH> I am pretty sure it was fixed no far after it was reported since I never had any problems with any setup in the last I don't know how many years to get stacktraces...

07:26 <PaulFertser> PlasmaHH: no problem, the switch will happen.

07:26 <PaulFertser> Thank you for being persistent.

07:27 <PlasmaHH> PaulFertser: you would be surprised what I have to deal with at work ;)

08:11 <PlasmaHH> PaulFertser: https://sourceware.org/bugzilla/show_bug.cgi?id=9500 seems that is the bug now... nothing mentioned about a fix, but I also cannot reproduce it, though I have no idea what the exact circumstances would have been

08:11 <PaulFertser> PlasmaHH: you finally found it. I couldn't even though I tried.

08:12 <PlasmaHH> PaulFertser: a guy in the gdb channel found gnat bug 1 and then we added 2395 or whatever it was and found it that way...

08:13 <PaulFertser> PlasmaHH: I'd say for all practical purposes it makes sense to flip the switch by default.

08:13 <PaulFertser> Even if we do not know for sure it was fixed in GDB fully.

08:25 <PlasmaHH> PaulFertser: I tried the mentioned stack info depth mi command there and didn't have any problems, though it is not totally clear to me under which circumstances it made problems, my guess is that it happened when during stack unwinding a read error occurred, which I right now have no idea how to provoke

08:26 <PaulFertser> PlasmaHH: yeah, as GDB now with the correct debug info shouldn't try reading whatever it's not allowed to. Probably you can manually mangle "lr" to an invalid address before trying?

08:27 <PlasmaHH> PaulFertser: I see a few "Failed to read memory" messages in ocd but its not clear to me if they would have caused errors in the older verison

08:27 <PlasmaHH> my gdb version is usually rather new ;)

20:14 <PlasmaHH> PaulFertser: persistence++;

20:20 <PaulFertser> PlasmaHH: :))

20:21 <PaulFertser> PlasmaHH: your points are really nicely outlined and I think fully valid.

20:21 <PaulFertser> Thank you for taking time to write that.

20:23 <PlasmaHH> PaulFertser: you're welcome, human. Actually for me the outcome is quite irrelephant since I now know about the option and have it in all my openocd configs but I think its quite a trap for young players..

20:23 <PaulFertser> Sure thing!

20:31 <PlasmaHH> PaulFertser: alright, I am out, I forgot that contrary to 20 years ago these days people don't accept any kind of tickets without all the hours of effort in a patch

20:33 <PaulFertser> PlasmaHH: fwiw, Tommy isn't a maintainer

20:33 <PaulFertser> PlasmaHH: neither is me

20:33 <PaulFertser> PlasmaHH: I can send a patch flipping the switch and include your rationale in the commit message

20:35 <PlasmaHH> PaulFertser: I am probably to old for this, but I really don't like this "its open source go fix it yourself" tone whenever you report a bug anywhere... I am not demanding anyone to spend any time with it, so thats a stupid comeback... people don't seem to realize that others may not want to spend all their free time fixing bugs they discover all over the place...

20:36 <PlasmaHH> I discover several bugs in all kinds of software in a week, I would need 40 hours a day to fix them all...

20:36 <PlasmaHH> </rant>

20:38 <PaulFertser> PlasmaHH: I'm not sure Tommy meant it exactly like that. In this particular case the patch itself is trivial, and sometimes it's easier to get the maintainers' attention by sending one rather than by discussing an issue on the bugtracker.

20:44 <PlasmaHH> PaulFertser: the patch itself may or may not be trivial, I don't know whats involved in the project, necessity to write test cases etc. ( in some projects you get patches thrown back at you when your whitespace doesn't match their style ). register mailinglist, present it there, follow up ... yes, this instance is possibly one of the smaller ones, but this grumpy guy here got a bit sensitive over

20:44 <PlasmaHH> the decades to all these requirements. and quite frankly, if that change is so small then maybe applying the patch is more effort than replicating it by hand ;)

20:48 <PaulFertser> Understandable

20:49 <PaulFertser> For such a patch the most valuable would be the commit message.

20:49 <PlasmaHH> Absolutley, it would have to be a concise version of all the discussion and reasoning

20:49 <PlasmaHH> Something I am extraordinarily bad with, you can ask everyone I worked with ;)