| - ephase (QUIT: Ping timeout: 267 seconds) (~ephase@82-66-198-11.subs.proxad.net) | 00:23 | |
| + ephase (~ephase@82-66-198-11.subs.proxad.net) | 00:32 | |
| - mjw (QUIT: Ping timeout: 261 seconds) (~mjw@gnu.wildebeest.org) | 01:20 | |
| minute | interesting, a new shader compiler for mali (used in rk3588) is coming, called kraid | 01:39 |
|---|---|---|
| minute | > Building KRAID requires the "-Dpanfrost-rust" Meson build option as well as the "PAN_USE_KRAID" environment variable being set. | 01:40 |
| minute | experimental geom shaders support is also close to being merged :0 https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/44118 | 01:44 |
| - RandyK (QUIT: Remote host closed the connection) (~RandyK@user/randyk) | 01:47 | |
| + RandyK (~RandyK@user/randyk) | 01:47 | |
| - RandyK (QUIT: Remote host closed the connection) (~RandyK@user/randyk) | 01:49 | |
| + RandyK (~RandyK@user/randyk) | 01:49 | |
| - paperManu_ (QUIT: Ping timeout: 260 seconds) (~paperManu@72.10.153.237) | 01:53 | |
| - f_ (QUIT: Remote host closed the connection) (16abab341f@postmarketOS/funderscore) | 01:58 | |
| - jbmorley (QUIT: Remote host closed the connection) (~jbmorley@dhcp-72-253-8-30.hawaiiantel.net) | 02:19 | |
| + paperManu_ (~paperManu@72.10.153.237) | 02:19 | |
| loon | do you have a link for the kraid shader compiler? | 02:23 |
| loon | I found the collabora page, hidden among so many slop pages | 02:24 |
| - paperManu_ (QUIT: Ping timeout: 253 seconds) (~paperManu@72.10.153.237) | 02:39 | |
| - paperManu (QUIT: Ping timeout: 260 seconds) (~paperManu@72.10.153.237) | 02:46 | |
| - AnimaInvicta (QUIT: Ping timeout: 260 seconds) (~AnimaInvi@88-169-191-86.subs.proxad.net) | 02:46 | |
| + AnimaInvicta (~AnimaInvi@88-169-191-86.subs.proxad.net) | 02:47 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 02:47 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 02:58 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 02:59 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:06 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 03:06 | |
| - lidstah (QUIT: Remote host closed the connection) (~lidstah@gateway/tor-sasl/lidstah) | 06:53 | |
| + lidstah (~lidstah@gateway/tor-sasl/lidstah) | 06:53 | |
| + f_ (16abab341f@postmarketOS/funderscore) | 08:13 | |
| - colinsane (QUIT: Remote host closed the connection) (~colinunin@97-113-79-79.tukw.qwest.net) | 08:22 | |
| - ex-parrot (QUIT: Quit: _b) (~fincham@user/ex-parrot) | 08:24 | |
| + ex-parrot (~fincham@user/ex-parrot) | 08:25 | |
| + colinsane (~colinunin@97-113-79-79.tukw.qwest.net) | 08:26 | |
| rick_ | minutes: oh interesting.. since i'm not deep in the graphics stack.. could this help with open gl stuff like prusaslicer? | 08:29 |
| rick_ | minutes: oh interesting.. since i'm not deep in the graphics stack.. could this help with open gl stuff like prusaslicer? | 08:29 |
| rick_ | eh sorry for the spam.. hit the wrong key | 08:29 |
| rick_ | and for the typo of your name ^^v | 08:30 |
| + Chwoka (~Chwoka@76.27.251.172) | 08:44 | |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 09:56 | |
| - voltaire28_ (QUIT: Client Quit) (~jlafon@28.162.2.93.rev.sfr.net) | 09:57 | |
| + voltaire28_ (~jlafon@28.162.2.93.rev.sfr.net) | 09:57 | |
| - Chwoka (QUIT: Remote host closed the connection) (~Chwoka@76.27.251.172) | 11:00 | |
| + mjw (~mjw@gnu.wildebeest.org) | 11:23 | |
| + paperManu (~paperManu@72.10.153.237) | 11:31 | |
| - aperezdc (QUIT: Remote host closed the connection) (~aperezdc@46.23.89.43) | 12:06 | |
| + aperezdc (~aperezdc@2a03:6000:6e61:633::43) | 12:14 | |
| - aperezdc (QUIT: Remote host closed the connection) (~aperezdc@2a03:6000:6e61:633::43) | 12:39 | |
| + aperezdc (~aperezdc@46.23.89.43) | 12:40 | |
| - mjw (QUIT: Ping timeout: 258 seconds) (~mjw@gnu.wildebeest.org) | 13:01 | |
| - paperManu (QUIT: Ping timeout: 248 seconds) (~paperManu@72.10.153.237) | 13:13 | |
| * Guest9413 -> mjw | 13:42 | |
| + paperManu (~paperManu@modemcable065.144-200-24.mc.videotron.ca) | 14:01 | |
| - ex-parrot (QUIT: Ping timeout: 257 seconds) (~fincham@user/ex-parrot) | 14:05 | |
| + ex-parrot (~fincham@user/ex-parrot) | 14:08 | |
| minute | rick_: yes it could @ prusa | 14:42 |
| MartiniMoe | uhm, is there a way to search the irc logs on mntre.com/reform-irc-logs? I am looking for the kernel patches that supposedly activate suspend on the rk3588 pocket reform | 14:50 |
| kensanata | Oh, I did not know about column(1). | 14:55 |
| minute | MartiniMoe: no, but maybe gordon1 and josch know about the latest state of that | 14:59 |
| - digitalrane (QUIT: Ping timeout: 254 seconds) (~digitalra@user/digitalrane) | 15:14 | |
| + digitalrane (~digitalra@user/digitalrane) | 15:18 | |
| + spew_ (~spew@user/spew) | 15:40 | |
| gordon1 | i don't, i kinda stalled on my suspend work tbh | 15:41 |
| - spew (QUIT: Remote host closed the connection) (e9e8135d87@user/spew) | 15:42 | |
| * spew_ -> spew | 15:42 | |
| josch | MartiniMoe: two answers: a) there is no official way to search. I am online 24/7 so i just search my irc log. For others, they could wget recursively the html archive. b) the latest state of suspend is that stephano wanted to look into it | 15:56 |
| josch | well, s2idle to be precise, not suspend | 15:56 |
| josch | MartiniMoe: and sigrid has s2idle working on 7.2.5 with alpine: https://ftrv.se/_/alpine/packages/community/aarch64/l | 15:57 |
| josch | inux-rk3588-7.2.5-r0.apk | 15:57 |
| qbit | hot dang! | 16:18 |
| + _justin_kelly4 (~justinkel@user/justin-kelly/x-6011154) | 16:29 | |
| - lidstah (QUIT: Ping timeout: 246 seconds) (~lidstah@gateway/tor-sasl/lidstah) | 16:38 | |
| - RandyK (QUIT: Ping timeout: 243 seconds) (~RandyK@user/randyk) | 16:38 | |
| + lidstah (~lidstah@gateway/tor-sasl/lidstah) | 16:39 | |
| + RandyK (~RandyK@user/randyk) | 16:39 | |
| - lidstah (QUIT: Ping timeout: 268 seconds) (~lidstah@gateway/tor-sasl/lidstah) | 16:48 | |
| + lidstah (~lidstah@gateway/tor-sasl/lidstah) | 16:49 | |
| + riabenko (~riabenko@2a01:d0:333::831) | 16:57 | |
| minute | hm, got the first interesting benchmark from cq8750s | 17:18 |
| minute | with cpuidle turned off. and still no heatsink, i get a total score of 33912 in "7z b" | 17:18 |
| - voltaire28_ (QUIT: Quit: WeeChat 4.6.3) (~jlafon@28.162.2.93.rev.sfr.net) | 17:18 | |
| minute | my intel i9-9900 pc here gets 63000 | 17:18 |
| qbit | that's compressing? KiB/s? | 17:47 |
| qbit | oh, total | 17:47 |
| qbit | neat, i get 43523 on my amd ryzen 7 | 17:50 |
| qbit | any idea what the next gets? | 17:53 |
| - riabenko (QUIT: Quit: riabenko) (~riabenko@2a01:d0:333::831) | 17:59 | |
| minute | qbit: next in terms of rk3588? | 18:06 |
| + Ar|stote|is (~linx@149.210.10.88) | 18:09 | |
| - Ar|stote|is (QUIT: Client Quit) (~linx@149.210.10.88) | 18:09 | |
| qbit | ya | 18:24 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 18:30 | |
| + Ar|stote|is (~linx@149.210.10.88) | 18:31 | |
| vagrantc | is anyon maintaining a 6.12.x branch for mnt/reform? or is it prtty much all on newer kernels? | 18:37 |
| + talos8 (~talos@2600:6c5d:f0:92c0:7d80:53cf:8d79:725f) | 18:38 | |
| qbit | heh, og reform (imx8) gets 5657 | 18:39 |
| - talos (QUIT: Ping timeout: 252 seconds) (~talos@2600:6c5d:f0:92c0:2333:e035:a1d:add3) | 18:40 | |
| * talos8 -> talos | 18:40 | |
| minute | qbit: we've come a long way | 18:41 |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.10.88) | 18:44 | |
| qbit | indeed! | 18:44 |
| josch | qbit: rk3588 says 20391 here | 18:44 |
| + Ar|stote|is (~linx@149.210.10.88) | 18:45 | |
| qbit | sweet! ty! | 18:52 |
| - _justin_kelly4 (QUIT: Ping timeout: 241 seconds) (~justinkel@user/justin-kelly/x-6011154) | 18:53 | |
| + _justin_kelly4 (~justinkel@user/justin-kelly/x-6011154) | 19:01 | |
| - streifen (QUIT: Quit: Ping timeout (120 seconds)) (~streifen@84.215.253.62) | 19:16 | |
| + streifen (~streifen@84.215.253.62) | 19:16 | |
| - paperManu (QUIT: Ping timeout: 260 seconds) (~paperManu@modemcable065.144-200-24.mc.videotron.ca) | 19:16 | |
| + paperManu (~paperManu@modemcable065.144-200-24.mc.videotron.ca) | 19:18 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.10.88) | 19:18 | |
| + Ar|stote|is (~linx@149.210.10.88) | 19:20 | |
| + wielaard (~mjw@gnu.wildebeest.org) | 20:02 | |
| marty | does NetBSD install on the original diy iMX? | 20:03 |
| marty | has anyone tried that successfully? | 20:03 |
| MartiniMoe | josch: thanks! Going to look into how Sigrid achieved it ^^ | 20:10 |
| MartiniMoe | josch: oh. The link does not seem to work :/ | 20:11 |
| MartiniMoe | Oh, it continues in the next line :D | 20:14 |
| josch | MartiniMoe: this is the conversation from the logs: https://mntre.com/reform-irc-logs/2026-09-13.log.html#t14:53:09 | 20:14 |
| josch | sorry for the bad paste :( | 20:14 |
| MartiniMoe | Awesome, thanks a lot! | 20:15 |
| * mjw -> Guest3219 | 20:23 | |
| - Guest3219 (QUIT: Killed (osmium.libera.chat (Nickname regained by services))) (~mjw@2001:1c00:f482:400:c901:9e4f:ea0c:d6f4) | 20:23 | |
| * wielaard -> mjw | 20:23 | |
| + Guest3219 (~mjw@2001:1c00:f482:400:c901:9e4f:ea0c:d6f4) | 20:23 | |
| + riabenko (~riabenko@185.143.146.30) | 20:35 | |
| - paperManu (QUIT: Ping timeout: 260 seconds) (~paperManu@modemcable065.144-200-24.mc.videotron.ca) | 20:44 | |
| + paperManu (~paperManu@modemcable065.144-200-24.mc.videotron.ca) | 20:46 | |
| minute | uff i finally found the key to getting rid of gpu crashes on cq8750s... > echo -1 > /sys/devices/platform/soc@0/3d00000.gpu/power/autosuspend_delay_ms | 20:58 |
| minute | something goes wrong with autosuspending the gpu (which happens like all the time) | 20:58 |
| minute | and i also found out that the cpu is super slow when cpuidle is active (default) | 20:59 |
| minute | so weird | 20:59 |
| sigrid | MartiniMoe: it's 7.2.13 and the patches are all in https://git.sr.ht/~ft/aports | 21:00 |
| gordon1 | okay, i think components on the top of the keyboard are not observable without backlight, so i think in my case it is acceptable, but for anyone willing to have backlight - they probably will be noticeable | 21:10 |
| josch | gordon1: your keyboard doesn't have the leds for backlight? | 21:11 |
| gordon1 | josch: nope, i prefer blank keycaps and sorta see no point in it | 21:13 |
| gordon1 | josch: but i have some rgb leds for indication in a trench if that's what you're asking! | 21:13 |
| josch | haha :D | 21:13 |
| josch | okay, i see :) | 21:14 |
| gordon1 | (tbh i have a strong suspicion that that is the sole reason why kbd sometimes drops off the usb and all my other "mitigations" with crystal and flash routing are futile and it just needs a massive cap for 80 LEDs) | 21:22 |
| gordon1 | i wonder where can i get a classic mnt reform keyboard for experiments to confirm my theory | 21:23 |
| gordon1 | ah yes, i will have one soon :) | 21:23 |
| + dd88 (~dd88@2601:646:8683:7820:244c:1966:ae3f:1647) | 21:26 | |
| minute | gordon1: hmm but the power rails for the leds and the RP are separate right? 3v3 and 5v | 21:27 |
| gordon1 | minute: yes, but i have a suspicion that it is a common mode interference that comes over the usb lines that just stalls the bus, i.e. if you look at dmesg - device does not disappear meaning rp2040 didn't reset, it just some sort of a bus stall | 21:29 |
| gordon1 | maybe it's a weird ground loop thing when half of the return current goes over the UART connector's GND and USB is very unhappy about it | 21:29 |
| gordon1 | just a theory tho, have to confirm it | 21:30 |
| vagrantc | hrm. i have had some issues with my keyboard occasinally spitting out lots of extra characters occasionally (noticed suspiciously close to when it was "charging" plugged into well over 32v, but not sure if that was it) ... | 21:31 |
| vagrantc | I guess i could swap out my keyboard with the mnt/reform external keyboard and see if the problem follows the keyboard | 21:32 |
| vagrantc | might also just need a cleaning or something | 21:32 |
| vagrantc | :) | 21:32 |
| gordon1 | vagrantc: yes, and if you look at your dmesg you'll see something along the line "resetting USB device" or something, that is what i'm talking about | 21:32 |
| gordon1 | and i would bet you can recreate it if you do cirlce + F3/F4 and hold it to cycle between colors | 21:33 |
| josch | i have been using keyboard v4 for two months now -- how do i re-create the problem? | 21:38 |
| vagrantc | my problems tend to be on the c and v keys | 21:38 |
| vagrantc | but ... maybe not exclusively | 21:38 |
| vagrantc | min is a v3 keyboar | 21:39 |
| vagrantc | d | 21:39 |
| minute | gordon1: interesting theory @ ground loop | 21:42 |
| gordon1 | minute: there is another thing i found, but i would need your help with that since i'm not really familiar with the code, but there is this thing in the keyboard fw: https://source.mnt.re/reform/reform/-/blob/main/reform2-keyboard4-fw/src/remote.c?ref_type=heads#L287-292 | 21:44 |
| gordon1 | i believe i found the reason for it and there is a race in LPC | 21:44 |
| minute | gordon1: in any case i have zero problems with kbd4 standalone on usb for many months of sustained everyday use... and also not in the next so far. and no issues with pocket reform standalone keyboard 2.0 over usb. | 21:44 |
| gordon1 | but i have no way of verifying it | 21:44 |
| gordon1 | but i'll send MR shortly | 21:45 |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.10.88) | 21:51 | |
| gordon1 | minute: https://source.mnt.re/reform/reform/-/merge_requests/105 pls | 21:51 |
| - streifen (QUIT: Ping timeout: 243 seconds) (~streifen@84.215.253.62) | 21:59 | |
| + streifen (~streifen@62.141.134.177) | 22:00 | |
| - paperManu (QUIT: Ping timeout: 247 seconds) (~paperManu@modemcable065.144-200-24.mc.videotron.ca) | 22:05 | |
| minute | gordon1: what would be able to race there? i.e. which other path can access pcb->rxfifo.len ? | 22:12 |
| minute | gordon1: is something else calling uartRxBufferRead() from an irq? | 22:13 |
| + paperManu (~paperManu@72.10.153.237) | 22:13 | |
| minute | gordon1: and in that case line 78 would have the same issue no? | 22:14 |
| minute | ah you write "uartRxBufferWrite function is called from IRQ handler" | 22:14 |
| minute | gordon1: i think the double-power-on command is more about the state of the command parser in the lpc. | 22:16 |
| minute | gordon1: it doesn't time out the characters it receives, for example. so if there's a random character in it, and it's not a newline, the command parser state machine is not in the expected place. the double power on could probably be replaced with sending "\n1p\n" | 22:17 |
| gordon1 | minute: uartRxBufferWrite is called from an irq and uartRxBufferRead is not, and i guess if IRQ with uartRxBufferWrite happens when uartRxBufferRead is executed you can get off-by-one thing, and it is definitely something that i sorta experience - my reform refuses to turn on with ZMK from first try (it does not repeat the command), only from second, but it is not random, it is very much | 22:17 |
| gordon1 | deterministic, i.e. in my case i need to tap circle+1 twice, then i tap circle+0 and then it i need to tap circle+1 twice again | 22:17 |
| minute | gordon1: you'll see that reform lpc probably responds with syntax error to the first command | 22:18 |
| gordon1 | hmm | 22:18 |
| gordon1 | but if i send 1p\n and it receives p\n (because off by one) - that's a syntax error, is it not? | 22:18 |
| minute | sorry i meant \r instead of \n (no idea how this happened to become \r) | 22:18 |
| gordon1 | or is p a command? | 22:19 |
| minute | gordon1: i don't remember, need to check the parser code in the reform 2 lpc code. | 22:19 |
| minute | like, the number could be optional | 22:19 |
| minute | it would then equal 0p | 22:20 |
| gordon1 | anyway, i still sorta think that it is worth wrapping the thing in disable_irq/enable_irq just for the potential to corrupt the thing | 22:20 |
| gordon1 | it is not? | 22:20 |
| gordon1 | hmm, i wonder how can debug that without rendering my reform unusable | 22:21 |
| minute | gordon1: some kind of mutex for that access would be ok, but i'm not convinced it's the correct place | 22:22 |
| gordon1 | yeah probably better to put it where the uartRxBufferRead is called? | 22:22 |
| minute | i wouldn't put it in there at all before confirming that this is really a problem | 22:23 |
| minute | gordon1: do you have some other thing that can talk to a UART? (a usb-uart)? | 22:23 |
| minute | then you can just connect that to the lpc pins and talk to it manually and see what it says | 22:24 |
| minute | or console mode in the newest keyboard 4 fw? | 22:24 |
| gordon1 | i do but having another machine i can plug it into is much more difficult | 22:24 |
| gordon1 | oh, is it some sort of console over to sysctl on oled? | 22:25 |
| gordon1 | if so that might work | 22:25 |
| minute | gordon1: yes | 22:25 |
| minute | gordon1: you can also enable echo mode with 1e | 22:25 |
| gordon1 | okay, let me find out how to get it running | 22:26 |
| minute | https://source.mnt.re/reform/reform/-/blob/main/reform2-lpc-fw/src/boards/reform2/board_reform2.c?ref_type=heads#L945 | 22:26 |
| minute | gordon1: Console is a new oled menu entry in the kbd4 fw. could be that i only enabled it in NEXT mode but you can change that if necessary | 22:26 |
| gordon1 | my kbd fw is couple of months old, probably from beginning of this year | 22:27 |
| minute | rebase rebase | 22:29 |
| Esi | been using my pocket reform a lot at the Reykjavík fablab. the staff there has been asking me a lot of questions about it. if you get some more orders from Iceland it might be my fault :P | 22:35 |
| + paperManu_ (~paperManu@142.169.16.42) | 22:46 | |
| + erle (~erle@user/erle) | 23:24 | |
| - erle (QUIT: Client Quit) (~erle@user/erle) | 23:26 | |
| - paperManu_ (QUIT: Ping timeout: 252 seconds) (~paperManu@142.169.16.42) | 23:46 | |
| - murphnj (QUIT: Ping timeout: 257 seconds) (~murphnj@user/murphnj) | 23:57 | |
| + murphnj (~murphnj@user/murphnj) | 23:58 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!