00:04 <pedro_a> the fact is that i already did this in the past but did not took notes , today i had to flash again the binary because the hardware was giving unknown device on windows after changing multiples usb cables to test . After the flash windows detected correctly the hardware again .

00:06 <pedro_a> for some unknown reason to me this hardware every 2 years needs a flash :D , not sure what are the conditions to affect this change in the flash memory in hardware , probably some power peaks when i start it up under some unknown conditions , not sure , anyway , it is fixed .

00:10 <pedro_a> this is not the only hardware i have this issue , recently i had to reflash a focusritte soundcard because flash in eprom was damaged in the first bits , so cpu was always in boot load . My advice to you all i to extract all your devices firmwares while they are working , just in case you need to flash them again in future instead throwing a good

00:10 <pedro_a> piece of hardware to the garbage because the brand do not send you the flash .

10:22 <RolfNoot> Hi all. Can someone help with this failing build: https://review.openocd.org/c/openocd/+/7983/7..8

10:23 <RolfNoot> It fails because of the flash loaders which are not found in contrib/loaders/psoc6

10:24 <PaulFertser> Why are they not there?

10:24 <RolfNoot> however I submitted patch https://review.openocd.org/c/openocd/+/7984 to included the loaders

10:24 <RolfNoot> They are there in patch 7984

10:25 <karlp> did you push them as one branch?they should have linked as dependents

10:25 <karlp> did you add them in the second patch, not the first patch that actually needs them?

10:25 <RolfNoot> yes I pushed them as one branch

10:25 <RolfNoot> yes but I swapped them

10:26 <RolfNoot> with git rebase -i origin/master so that the patch 7984 is first

10:27 <PaulFertser> First meaning on the first line or on the last?

10:27 <PaulFertser> Check git log locally prior to pushing.

10:27 <RolfNoot> 7984 is the last now

10:28 <PaulFertser> But you need origin/master -> 7984 -> 7983 , right?

10:28 <PaulFertser> Do you see that in local git log?

10:29 <RolfNoot> i mean the last line in git rebase -i origin/master

10:29 <RolfNoot> and the first in git log now

10:29 <PaulFertser> But the first (topmost) in git log should be 7983, right?

10:30 <RolfNoot> https://pastebin.com/rKJtwgtd

10:30 <PaulFertser> RolfNoot: but you need loaders _before_ the code that needs them.

10:30 <PaulFertser> Or even in the same commit actually.

10:32 <RolfNoot> Yes, so the order in the log is how it migrates?

10:32 <RolfNoot> I understand that the loaders should be there before

10:33 <RolfNoot> So what shall I do now? Merge them in one commit?

10:33 <PaulFertser> git log shows the latest commits topmost.

10:34 <PaulFertser> IIRC maintainers are OK with having loaders in the same commit as the code in src/flash that needs them.

10:34 <RolfNoot> yes, and that's the same order as the builds are tested?

10:34 <PaulFertser> Not exactly. For testing a build a commit is "checked out". And each git commit is by definition a snapshot of all the sources.

10:35 <PaulFertser> So it's hard to answer your question as it seems to be assuming wrong things about git.

10:35 <RolfNoot> Okay, I then better will do them in one commit

10:36 <RolfNoot> So I can "fixup" the loader to merge with the flash driver with git rebase -i origin/master?

10:36 <PaulFertser> RolfNoot: if you reason about git as something that stores _all the sources_ for _every commit_ (and not the differences or other incremental things) then you'll have the right model about what's happening.

10:37 <RolfNoot> Okay that makes thanks

10:38 <RolfNoot> that makes sense *

10:39 <PaulFertser> And each commit has metadata too, and in this metadata there's a list of "parent commits IDs" (a list because when doing merges you have more than one parent). And that's basically it.

10:45 <RolfNoot> I am getting a grip, trying to visualize. And it can get complex.

10:51 <RolfNoot> I saw that 7983 patchset 7 noted a rebase which I found strange as I didn't change anything for that commit. I probably have changed the order with by accident.

10:51 <RolfNoot> * by accident

15:12 <RolfNoot> How can I set my code 'ready for review'?

15:38 <karlp> RolfNoot: what do you mean by that?

15:38 <karlp> submitting it to gerrit _is_ "ready for review" unless you specifically mark them as draft or WIP or something...

15:45 <RolfNoot> I mean I fixed all issues and I was thinking I needed to do something to get it reviewed (not knowing that it is by default)

16:11 <karlp> well, you have a lot of "unresolved" comments on your diffs that you'll want to look at next I suppose...

16:36 <RolfNoot> where do I find these? The ones I found I marked with resolved.

16:38 <RolfNoot> Oh I need to press 'Reply" I guess

16:38 <RolfNoot> They are drafts by now

16:40 <PaulFertser> RolfNoot: right, all the line-by-line comments are submitted together when you do top level reply.

16:41 <RolfNoot> Yeah got it. Feel a bit silly, but will remind for next time.

17:02 <karlp> yeah, the need for the top level reply is cumbersome at itmes when you only have one comment too :|