03:45 <limonene0> Does anyone know how to program a STM32g030 chip using an STLINK/V2 on openocd 0.10.0 ? I'm currently getting an error message: Error: BUG: current_target out of bounds

06:40 <PaulFertser> limonene0: hi

06:40 <PaulFertser> limonene0: what's your invocation command line? Please also use some pastebin to show full output.

06:41 <PaulFertser> limonene0: 0.12.0 is the current release and should be available in all modern distros.

15:57 <limonene0> Here is my config, invocation, and output: https://pastebin.com/raw/yGAXsUQw

15:57 <limonene0> I'm on Linux Mint 19.3 . What is the best way to get openocd version 0.12.0?

16:02 <limonene0> No wait, it's actually 20.3

16:12 <Boban> Hi there

16:13 <Boban> I'm looking to implement CTI support for peripheral halt on Cortex-M33. I'd like to know if OpenOCD supports this.

16:17 <Boban> The documentation doesn't explicitly mention CTI support for ARMv8-M but just ARMv8. Does this include ARMv8-M as well or only ARMv8-A parts?

16:17 <blathijs> Hey all. I've fixed a bug and submitted the patches to gerrit, but I'm not quite familiar with Gerrit (more used to pull requests on github or gitlab). I've separated my changes into three logically separated commits and pushed that per instructions, but now it seems that gerrit treats them as completely independent patches, meaning I cannot find a way to refer to the entire set (i.e. to add a

16:17 <blathijs> link to the bug report) and it reports a merge conflict because (I think) gerrit does not realize that the commits have a certain order. Any guidance? Here are the patches: https://review.openocd.org/q/is:open+owner:matthijskooijman+-is:wip

16:20 <blathijs> Hm, it does seem that the "Relation chain" displayed for each of these patches links them together, though.

16:25 <blathijs> Ah, it seems that Gerrit is configured for the cherry-pick submit type, which AFAICS means that gerrit does not track dependencies but expects each commit/patchset to be separately cherry-pickable, it seems.

16:29 <blathijs> I also cannot find a way to actually add a comment to the patchset, other than adding an inline comment that responds to a particular part of the patch, but maybe that's by design (I would want to comment that the merge conflict is because of the dependency, and something about why the build seemingly failed)

16:55 <limonene0> How do I write a target configuration file for STM32G030 ? There isn't one currently.

19:18 <PaulFertser> limonene0: the best way to get 0.12.0 is to ask yyour distro maiintainers.

19:19 <PaulFertser> blathijs: hey! Thank you so much for fixing and sending it properly to Gerrit!

19:21 <PaulFertser> blathijs: when a patch has +2 from one of the maintainer, it's possible to "submit" it, and our merging strategy is cherry-pick. So indeed that can result in a merge conflict if the previous patch wasn't cherry-picked before that. For OpenOCD project this strategy seems to be better for the maintainers, so Gerrit shows merge conflicts for all the patches that can't be applied right away to master

19:21 <PaulFertser> HEAD.

19:22 <PaulFertser> blathijs: Gerrit always tracks dependencies for the patchsets as it's working on top of real git. It's just that it doesn't take it into account when computing conflicts with this submit strategy.

19:23 <PaulFertser> blathijs: to comment press "Reply" on top, and add your "patchset comment" there.

19:24 <PaulFertser> limonene0: there's target configuration called tcl/target/stm32g0x.cfg added in v0.10.0-1134-gba131f30a.

19:25 <PaulFertser> limonene0: see, Debian ships 0.11.0 in current stable and 0.12.0 in testing which will become stable very soon: https://packages.debian.org/search?keywords=openocd

19:26 <PaulFertser> limonene0: so if your distro is stuck at 0.10.0 please talk to your maintainers. If they have issues packaging, the OpenOCD team is always happy to help.

19:28 <PaulFertser> blathijs: so don't worry about the merge conflict, no need to comment on it.

19:40 <PaulFertser> blathijs: regarding the failed build, no need to comment on it either, it's indeed pushing another revision just aborted the build and a new one will eventually happen and override the verified label.

20:53 <limonene0> PaulFertser: I managed to get upgraded to openocd 0.11 and the stm32g0x.cfg file appears to be working. Thank you. If you're ever in Minneapolis, I'll buy you a beer.

20:54 <PaulFertser> limonene0: welcome, and thank you :)

20:56 <blathijs> PaulFertser: Thanks for your elaborate reply, very helpful. I totally missed the "Reply" button, I see it now. I guessed as much that there was no real reason to comment, but wanted to do so just in case, but I'll refrain from it now. Looks indeed like the builds are now all ok, so I think I'm done for now - and can now redirect attention to implementating a workaround to work with the released

20:56 <blathijs> openocd versions :-)

21:02 <PaulFertser> blathijs: btw, to actually send per-line comments you'd need to press the reply anyway, they're always part of the bigger whole reply.

21:05 <PaulFertser> It's strange build time is an hour, guess I need to check if ccache is working properly soon.