<zmatt>
21:48:13 -!- Guest79 is now known as Gigi95
<buzzmarshall>
ya i thnk thats who i seen you talking to, it just stuck in my mind as i think outside of myself i dont think i have ever seen anyone here ask about jtag
<zmatt>
I explained to them in very unambiguous terms that JTAG is not useful for typical debugging on a beaglebone
<zmatt>
I really hate it when someone asks something, you explain it, they thank you for the information, and then days late come back and ask THE EXACT SAME FUCKING QUESTION
<buzzmarshall>
ya it get it and thats why i never bother trying to mod or get to publicly involved in most things these days
<buzzmarshall>
and i have a deep appreciation for the ones that do
mabb0ne has quit [Ping timeout: 268 seconds]
<zmatt>
I mean, modding and helping people are unrelated things... by modding I assume you mean being a channel operator (irc doesn't have "moderators")
<buzzmarshall>
back on novernet many years ago i was a op for awhile as well as a few private irc servers
<buzzmarshall>
but i also moded a bunch of forums and was part of a public group that provided a peice of software people were using all over the world
<zmatt>
I see being op just as looking after the channel, e.g. deal with things like spammers/bots if needed, and people if they really go too far (though that fortunately doesn't really happen here)
<buzzmarshall>
ya things are pretty peaceful here as i never see much really go on
<buzzmarshall>
irc isn't what it used to be so in most cases life is much more simple here
<buzzmarshall>
back in the day tho
<buzzmarshall>
irc and uucp was all there was
<buzzmarshall>
irc was depending on the channel like the old out west
<buzzmarshall>
lol
<buzzmarshall>
actully its funny cause up till i got the bbb aways back i had forgotten all about irc
<buzzmarshall>
then i setup a client when i seen you guys had a room here
<zmatt>
I've been on irc uninterrupted for... jeez I dunno, more than 3 decades
<buzzmarshall>
for some reason beagle stuff just isnt as plentiful as alot of other things
<jkridner>
what needs to be more plentiful?
<buzzmarshall>
after i set up to get your room on freenet i pretty much start the client up everyday and log in just to see whats going on
<zmatt>
useful uptodate documentation, uptodate examples that work
<buzzmarshall>
after years of being on the amlogic and rockchip boards where theres tons and tons of places to talk and find info i just found that when it came to the beagles theres a lot less
<buzzmarshall>
just an observation, not saying good or bad
<buzzmarshall>
most of my buds are still all big on the avr's and rpis
<buzzmarshall>
alot are 3d printer guys messin with the avr's on the printers and now how deep i was into atmel stuff so i started out looking at all the atmel code
<zmatt>
there's obviously also some positive feedback loop there... more people using something leads to more community resources making it easier and more attractive for new users
<buzzmarshall>
and after looking at marlin and the other stuff how the atmels no longer up to that task
<buzzmarshall>
i started looking at arm based boards like the rockchip pro and some amlogic stuff but theres to much proprieatary stuff and not a good ide for them
<buzzmarshall>
ive always liked and used TI stuff and thats how i came across the bbb
<buzzmarshall>
its small and got decent support
<buzzmarshall>
and the pru attracted me
<zmatt>
pru is the single best thing about the bbb
<buzzmarshall>
i would agree and its readily available
<buzzmarshall>
anyone looking for motion control needs a good pulse gen and without a fpga solution the pru is the best i have come across
<buzzmarshall>
ive looked at other boards trying to achieve something similar but its way to much work and in some cases i am not sure even doable without putting in some serious time trying to reverse somethings that are not published
<buzzmarshall>
ti and pru is all there for the looking
<buzzmarshall>
for the last month or so ive been looking at a allwinner board for someone as they thought they could its arisc processor to work like a pru but its not a good idea from wha i can see
<buzzmarshall>
oh well i learned alot more about the allwinner soc's but now i just started back into the bbb as my machine is just about built and i need to start thinking about making it work
<zmatt>
ah it sounds like the arisc is used for power management... like the cortex-m3 on the AM335x
<zmatt>
yeah, it's nothing like PRU
<zmatt>
it even has an instruction cache, which pretty much guarantees unpredictable instruction timing
<zmatt>
and of course allwinners's blob for it is closed-source (while TI's blob for the am335x cortex-m3 is open source)
<zmatt>
buzzmarshall: using the arisc for real-time signals is particularly funny, since linux-sunxi.org/AR100 mentions that safely accessing GPIO from the ARISC requires using hardware spinlocks to coordinate with linux, which is of course a great way to ruin your timing
<zmatt>
(I'm guessing the allwinner GPIO controller doesn't have atomic set/clear registers?)
<zmatt>
(indeed it doesn't! lol)
yCrazyEdd has quit [Ping timeout: 250 seconds]
Shadyman has joined #beagle
charlie5 has quit [Quit: Leaving.]
starblue1 has joined #beagle
starblue has quit [Ping timeout: 255 seconds]
CCFL_Man has joined #beagle
charlie5 has joined #beagle
mabb0ne has joined #beagle
vagrantc has quit [Quit: leaving]
<buzzmarshall>
sorry was off doing something... its actually dedicated to running crust so i dont no where they got the idea that they could make use of it
mabb0ne has quit [Ping timeout: 255 seconds]
<buzzmarshall>
again i think its part of some stigma most that i know that seem to think the bbb's are overpriced for what it has to offer so they just don't look at it
<zmatt>
"The AR100 is not currently used for anything on mainline Linux, as power management there is implemented using native drivers"
<buzzmarshall>
allwinner loads crust into it
<zmatt>
but yeah, like I said it's clearly analogous to the power-management cortex-m3 on the AM335x, not PRU
<buzzmarshall>
ya i don't disagree i just no the ones bugging me won't look at the beagles as they think there over priced compared to other boards
<zmatt>
yeah so presumably mainline will have worse power management than allwinner's bsp kernel
<buzzmarshall>
i agree with you as the pru is another thing but i thought i'd play along and learned abunch about one of the soc's that up till now ive never looked at
<zmatt>
I mean, if price is all one cares about, they're probably right
<buzzmarshall>
they look at rockchip and rpi's and others with more resources and thats about as far as they see and think about
<buzzmarshall>
those boards are more media based stuff not cnc or motion control
<zmatt>
yup
<buzzmarshall>
but they just don't see that
<zmatt>
the BBB wouldn't be my choice if I wanted an SBC for use as media center ;)
<buzzmarshall>
true
<buzzmarshall>
since most of those other brands moved from android boxes into the sbc market theres to much bs and misinformation with how to use those sbcs
<renrelkha>
i am trying to use cloud9 debugger but it not stop on breakpoints, i must be doing something wrong? the led blink, but debugger not https://imgur.com/A8RVjAN.png
<buzzmarshall>
even though they may be more sbc like they really are intende for media not general purpose
<renrelkha>
any pointers?
<zmatt>
renrelkha: I wish I could help you, I don't use cloud9 myself
<renrelkha>
ok no problem
mabb0ne has joined #beagle
mabb0ne is now known as mattb0ne
sicelo has quit [Ping timeout: 265 seconds]
sicelo has joined #beagle
sicelo has joined #beagle
sicelo has quit [Changing host]
renrelkha has quit [Quit: bye]
renrelkha has joined #beagle
mattb0ne has quit [Ping timeout: 240 seconds]
buzzmarshall has quit [Quit: Konversation terminated!]
<kveremitz>
hrmph, haven't ordinarily needed to change this before, but where's the script that sets up usb networking?! one of the interfaces is coming up on the same class C network as my home lAN .. d'oh!
<kveremitz>
ip link set usb1 down has got me fixed temporarily, but I'd like to make a more persistent change :D
<zmatt>
kveremitz: pretty sure there's a config file in /etc/default/ somewhere
<zmatt>
or if you don't care about the usb gadget at all, sudo systemctl disable generic-board-startup
samael has joined #beagle
Guest4171 has joined #beagle
rob_w has joined #beagle
<Guest4171>
hi, i need some help with BB-AI input pins value dos not change at all
<Guest4171>
i did echo 76 to /sys/class/gpio/gpio76/export and "in" at /sys/class/gpio/gpio76/direction ith root
<Guest4171>
is there a common issue or something?
florian has joined #beagle
<zmatt>
Guest4171: is the pin muxed to gpio mode?
<zmatt>
jkridner: is there any sort of ETA of when there will be new official images? particularly for the BBAI (with overlays/config-pin support) ?
<Guest4171>
zmatt how i can know?
xet7 has quit [Read error: Connection reset by peer]
<Guest4171>
zmatt i read a bit about them but i don't know how to check or apply
xet7 has joined #beagle
<zmatt>
yeah pin configuration is a big headache on the BBAI in released images, it initially required making a custom dts... someone fixed that and make it work basically the same as on the BBB but for some reason this isn't in the officially released images yet (maybe in the testing images? I'm not sure)
<Guest4171>
zmatt do you have some solution bookmarked?
<zmatt>
I don't, I haven't really followed it very closely. I don't really have any particular use for my BBAI so I haven't spent much time with it
<zmatt>
if a pin always seems to read as zero the pin config is probably wrong (pin not input-enabled)
<zmatt>
oh he left
rob_w has quit [Remote host closed the connection]
vagrantc has joined #beagle
_whitelogger has joined #beagle
rcn-ee has joined #beagle
<kveremitz>
zmatt: ty
<kveremitz>
rcn-ee: from earlier: <zmatt> jkridner: is there any sort of ETA of when there will be new official images? particularly for the BBAI (with overlays/config-pin support) ?
<kveremitz>
wtb new image too, although got the 2020-08 one working.. suspect PEBKAC
<kveremitz>
rcn-ee: it looked like you mighta been updating build scripts, etc :D
<kveremitz>
07-26 wow that's Fresh :D
<rcn-ee>
those have all the 4.19.x work from last years gsoc student for bbai..
<kveremitz>
zmatt: ^
<kveremitz>
rcn-ee: leave it with me, should have some results in <0.5h :D
<rcn-ee>
kveremitz, 'config-pin' will never fully work well on the BBAI, it's best to just configure the overlay thru u-boot..
<zmatt>
rcn-ee: in what way is an overlay better than config-pin?
<zmatt>
on the BBAI
<zmatt>
(the erratum applies to both)
<rcn-ee>
correct... i still feel we could do 'overlays' in SPL... (pre full u-boot) pre memory init....
<kveremitz>
zmatt: in your [vast] experience(!), how long does the BBB eMMC 'last' in 'average' usage?
<zmatt>
kveremitz: there's no 'average' usage
<kveremitz>
I'm starting to get errors on a 2GB model :/
<kveremitz>
it might benefit from some 'treatment' although I couldn't say what
<zmatt>
rcn-ee: theoretically sure, but the erratum has been ignored since the BBAI release and will be ignored until such a hypothetical system exists, and if ignoring it that long and widestream is acceptable then it's probably also fine to just continue to ignore it and have the convenience of config-pin
<zmatt>
:P
<kveremitz>
:D heh
<rcn-ee>
true true...
florian has quit [Quit: Ex-Chat]
<rcn-ee>
zmatt, i had a meeting with TI about a week ago, we were talking about pinmux/pru issues.. (TI: please i don't what to hear about that...) ;)
<zmatt>
(I also think it may be possible to temporarily reenter I/O isolation to perform pin configuration after memory init, but I haven't explored that idea either
<zmatt>
)
<zmatt>
rcn-ee: it would still be really nice to get more detailed info on the erratum, to be able to build a mental model of what _exactly_ the risks are in practice
<rcn-ee>
i agree, we were begging TI for full details at one point they were working on a application note, one engineer said possible "internal shorts"...
<zmatt>
and isolating the external i/o cells helps against that? weird
<zmatt>
and yes application note please much thanks
<rcn-ee>
atleast the whole thing was ripped out and redesigned in am6...
<zmatt>
that doesn't help am5 users
<kveremitz>
heh
<kveremitz>
rcn-ee: usb-gadget networking.. zmatt suggest config is in /etc/default?
<kveremitz>
I got a conflict yesterday with local LAN lol
<kveremitz>
if only we had a wiki .. lol :P *duck,cover,run*
<kveremitz>
(google failed me)
<kveremitz>
damnit need some caffeine.. brb
<rcn-ee>
So we tried that for BeagleBoard.org, by using github pages... turns out Google doesn't index Github very well... we should have never left elinux as the default wiki..
<kveremitz>
how bad would it be to host eg. dokuwiki?
<kveremitz>
iirc thats same as elinux..
<rcn-ee>
the new forum.beagleboard.org index's really well and has wiki support.. it's just too new..
<kveremitz>
aha yes
<kveremitz>
that was a welcome shift f'sure
argonautx has quit [Quit: Leaving]
vagrantc has quit [Quit: leaving]
mattb0ne has quit [Ping timeout: 255 seconds]
LetoThe2nd has quit [Quit: Connection closed for inactivity]
<kveremitz>
is there a way of bridging IRC <-> element ?
<kveremitz>
probably just creating an IRC 'room' would work, and deploying the matrix bridge somehow
<kveremitz>
idk anything about matrix though.. drove me bats when I tried last
<zmatt>
there's a matrix bridge
<kveremitz>
zmatt: yeah I know about that .. just thinking about the builds.bb.org 'site'
<kveremitz>
I went there assuming some useful build[er]/status info .. and ended up with .. that, instead.
<zmatt>
it might be a server configuration mistake
<jkridner>
for sure it is a server config mistake....
<kveremitz>
LOL awesome :D
<jkridner>
happened when I (foolishly) allows certbot to rewrite rules.
<kveremitz>
ooof no, don't do that.
<kveremitz>
so what should each be?! :D
<jkridner>
the builds.beagleboard.org site hasn't been up for some time. wasn't really being utilized.
<kveremitz>
:sadpanda:
<zmatt>
debian.beagleboard.org links to it
<kveremitz>
^
<jkridner>
it was something I really tried to institute with the GSoC students last year for some kind of continuous integration, but they weren't really capable enough to leverage it.
<kveremitz>
hmm, they -should- be, but perhaps not targetted enough :)
<kveremitz>
what would be the underlying platform, something like jenkins/etc ?
<jkridner>
ah, I made the mistake of reusing the same port number. :-(
<kveremitz>
dude LOL
<jkridner>
I used to have buildbot running on that port, now I have matrix.
* kveremitz
sends jkridner a coffee refill
<kveremitz>
ah I'm conversant in buildbot :) its OK
LetoThe2nd has joined #beagle
<kveremitz>
(and nginx, LE, etc) :P
<jkridner>
yeah.... looks like I never even re-installed buildbot on this build of the server.
<jkridner>
I don't think anything links to builds.beagleboard.org does it? maybe debian.beagleboard.org
<jkridner>
anyway, I moved ports and now it is greeted by the extra friendly 502 (bad gateway)
<zmatt>
20:24 <@zmatt> debian.beagleboard.org links to it
<kveremitz>
LOL niiiice
<kveremitz>
been there also, rev-proxy ftw
<jkridner>
pointing to rcn-ee 's build server would probably be more productive.
<jkridner>
not sure he wants people linking to it.
<kveremitz>
well, that's what I was kinda thinking internally here ..
<kveremitz>
but like you say..
mattb0ne has joined #beagle
<kveremitz>
ok time for a site refresh .. LOL
<rcn-ee>
i don't mind, it's on linode, bandwidth is cheap...
<rcn-ee>
kveremitz, tony just sent a message on linux-omap, finally v5.15.x omap is starting the depreciation of that old serial driver..
<kveremitz>
and cos I'm a gentoo nerd, we need to nuke 'quiet' from kernel cmdline XD lol
<kveremitz>
rcn-ee: woop!
<zmatt>
kveremitz: that'll slow down boot though, might want to change your serial console speed to 460800 :P
thinkfat has joined #beagle
<kveremitz>
zmatt: rockchip use 1500000
<kveremitz>
115200 is ok for me .. but a notch up wouldn't be bad :P
<zmatt>
460800 is the highest that seems to be supported by everything hassle-free
<rcn-ee>
115200 is the defualt for u-boot/kernel for omap, best to change them all.. i think rockchip's boot rom defualts to that high.
<kveremitz>
good to know
<kveremitz>
rcn-ee: indeed it does
<rcn-ee>
maybe just disable printk and disable low level messages outright..
<zmatt>
kveremitz: at least iirc linux had issues with higher than 460800, but it's been many many years since I fiddled with it so maybe higher is less of a problem nowadays
<kveremitz>
zmatt: any research/tips always useful :D
<kveremitz>
cool that looks good enough for me .. now I gotta migrate packages from *wheezy* to buster, and replace my old install
<kveremitz>
but first..
<kveremitz>
TEA! and some food :/
<zmatt>
back then I was doing using the serial console a bit more because iirc I was fiddling with some networking stuff, and I got sufficiently annoyed by the slowness of vim at 115200 that I ended up digging into it and making the console 460800 (including u-boot)
<zmatt>
and it's possible I submitted this upstream, or tried to, I vaguely remember trying to but I don't recall whether I was successful or whether it got lost in distraction or whatever
<zmatt>
the driver fix I mean
<kveremitz>
zmatt: the code seems to have been restructured since that :(
javaJake has quit [Ping timeout: 265 seconds]
<kveremitz>
if you have a few cycles to take a peek :)
thinkfat has quit [Ping timeout: 240 seconds]
<kveremitz>
jkridner: so where does the matrix/element gateway go now?! lol
<kveremitz>
just gotta fix up builds. and we're good again :D lol
<kveremitz>
oh element is riot..... ahhh that figures
<kveremitz>
looks like 2FA is .. missing so far
<zmatt>
kveremitz: I mean, I haven't looked at u-boot since then so I'd be at square one too... why don't you tell me where the setting has moved once you find it? :P
mru has quit [Ping timeout: 255 seconds]
mru has joined #beagle
<rcn-ee>
there are only 4 of us on chat.beagleboard.org, but it's up. ;)
thinkfat has joined #beagle
<kveremitz>
lol
<kveremitz>
zmatt: aritey - leave it with me :p a git checkout and grep probably works ... :p
<kveremitz>
rcn-ee: some sessions need timing out :)
akaWolf has quit [Ping timeout: 256 seconds]
thinkfat has quit [Ping timeout: 268 seconds]
akaWolf has joined #beagle
set_ has joined #beagle
<kveremitz>
zmatt: gettin there .. helps when your desktop doesn't trash copy/pasta support :( wtf KDE :/
LetoThe2nd has quit [Quit: Connection closed for inactivity]
<CCFL_Man>
anyone have issues getting a display on the mini hdmi port?
<kveremitz>
micro* ? or did they go up a size :D
<zmatt>
it's micro-HDMI yeah, not mini
<CCFL_Man>
micro, my apologies
<zmatt>
CCFL_Man: anyway, are you doing a poll whether anyone ever had any problems with display output under any circumstances? :P
<zmatt>
I very rarely use video output on a BBB, but whenever I've used it it worked fine as far as I remember
<kveremitz>
zmatt: ok made a local copy of my notes, whilst cache was valid :D I'll try a re-push of revised patch to gitlab, see what happens. Can I add you in a S-o-B line to the commit possibly?
<kveremitz>
probably wont' get actioned tonight .. but probably this week
<set_>
Me!
<zmatt>
I'm not sure what the appropriate policy is for derived/ported commits
<zmatt>
for S-o-B I mean
<set_>
Me?
<kveremitz>
zmatt: yeah its a classic grey area
<kveremitz>
I'm applying the "just stick it on the bottom a-la Gentoo" model :/
<zmatt>
kveremitz: I mean, maybe check the exact wording in the SubmittingPatches
<kveremitz>
which is by far not the most appropriate/legally correct
<zmatt>
well S-o-B means something fairly specific
<kveremitz>
yup will give that a checkover
<zmatt>
if significant change is required to the patch it might be more appropriate to drop it (since then I didn't actually sign off on the patch as submitted)
<zmatt>
but I'm not sure about that either, it depends on the wording
<kveremitz>
yeah, technically it becomes a Suggested-By I believe
<kveremitz>
or something.
<kveremitz>
going roughly by the linux kernel policy
<kveremitz>
iirc
<kveremitz>
my take on this generally being .. if the lawyers are getting involved, and this shit actually matters... its gone too far already..