09:31 <borneoa___> PaulFertser: sorry for bothering again. Gerrit is not working

09:51 <PaulFertser> borneoa___: sorry for not having figured out the root cause yet.

09:52 <PaulFertser> borneoa___: started the daemon, and will work on making it stable soon.

09:52 <PaulFertser> sorry :(

09:57 <borneoa___> PaulFertser: it's me that should feel sorry for bothering you during the weekend. Thanks for the prompt action

10:27 <PaulFertser> borneoa___: the weekends are great to help free software community, and you're doing so much work every week, I'm amazed again and again. Absolutely no need to be sorry on your side :)

10:36 <borneoa___> PaulFertser: my available time is getting less and less. The backlog is growing faster than I can handle. Luckily some of the developer is also participating in the review

10:48 <PaulFertser> borneoa___: at one point there might be a need to get sloppy with the review but that might turn out to be even worse in the long term. I have no good idea what to do, alas.

11:55 <blietaer> So if the OpenOCD GitHub is not the place to discuss and suggests features, where should I post ?

11:57 <PaulFertser> blietaer: the mailing list

11:57 <PaulFertser> blietaer: or here

11:57 <PaulFertser> blietaer: but suggesting features usually do not work unless you provide patches too.

11:57 <blietaer> Ah OK !

11:58 <blietaer> So mature enough for a PR

11:58 <blietaer> (if there is such thing on Gerrit)

11:58 <PaulFertser> blietaer: it's called a Changeset on Gerrit.

11:58 <PaulFertser> blietaer: discussing on ML is always good no patch needed.

11:59 <PaulFertser> But it's not like OpenOCD has many devs and someone is going to implement a new feature just after reading an idea.

12:00 <PaulFertser> blietaer: so what is it there on your mind?

12:00 <blietaer> Ah yes, many thanks

12:01 <blietaer> right, agreed, this is not my intend.

12:01 <blietaer> So, I am working on the SAM RH707 MCU, from the big Microchip SAM family, elaborating the flash algorithm for both OpenOCD and probe-rs (for Rust Embedded)

12:02 <blietaer> I have the (very) expensive dev. kit here just next to me so it is easy to code & test

12:02 <PaulFertser> blietaer: in OpenOCD it's supposed to be great when a flash algorithm is constantly reading from a circular RAM buffer in lock-free fashion while OpenOCD fills it up via JTAG/SWD without stopping the core.

12:03 <blietaer> Ah good, I'll keep that in mind

12:03 <blietaer> All I have as a starting point (not so bad...) is the dedicated RH707 flashing python script from Microchip MPLAB IDE dev. env.: this is a bit of a nightmare to run on itself, but at least the code is there to be reverse engineered in order to deduce the flashing sequence.

12:03 <blietaer> So, nothing mature enough for a PR, but I would be already super happy to have your bless (and hints?) about the official process to get my SAM RH707 support officially covered within your beautiful (and powerful) probe-rs tool.

12:04 <blietaer> ....and openocd tool

12:04 <blietaer> It is fun & efficient to work in the OpenOCD code though: I already have a stable atsamrh707.c code in the flash/ dir

12:05 <blietaer> inspired by previous atsam.c code, but need a lot of adaptation to support this one (even if this is 'just' a stupid cortex m7 in the end..)

12:05 <PaulFertser> blietaer: usually the essential part of implementing support for flash algorithm is to /not/ include some vendor code so that openocd would all be under a free software licence.

12:51 <karlp> oops, triggered ;)