| + furan (~ian@38.75.239.86) | 00:17 | |
| furan | random q - is everything needed to build a kernel for the reform in upstream? | 00:17 |
|---|---|---|
| furan | or do I need to build a linux-imx tree? | 00:18 |
| furan | aha I found it nm | 00:19 |
| mntmn | furan: everything is here, but currently in sysimage-v2 branch https://source.mnt.re/reform/reform-system-image/-/tree/sysimage-v2 | 00:20 |
| mntmn | furan: i should really merge it! it's what's shipped on your device | 00:20 |
| furan | thanks | 00:20 |
| mntmn | MR for that is open | 00:21 |
| furan | I want to write a simple app to configure DSI to scan out of an arbitrary block of physical memory and I'm not sure how complicated that is. I figured maybe there was an fbdev driver that does it, but maybe fbdev is using the DRM driver? | 00:21 |
| furan | (well what I mean is, maybe there's a configured fbdev driver that uses DRM) | 00:21 |
| furan | I think I need to read up more on how the display controller works | 00:22 |
| furan | :) | 00:22 |
| mntmn | furan: so it's one of two display engine chains. the default one is RAM -> DCSS -> NWL DSI -> Mixel DPHY | 00:23 |
| furan | ic | 00:23 |
| mntmn | furan: the alternative is LCDIF instead of DCSS (it's used when you switch to dual display, then DCSS goes to HDMI)_ | 00:24 |
| furan | is that so you can do stuff like clone? | 00:24 |
| mntmn | furan: this should also be interesting https://community.mnt.re/t/u-boot-and-internal-screen/295/2 | 00:24 |
| mntmn | furan: also i highly recommend to peek into "IMX8MDQLQRMref2.pdf" from NXP (requires NXP account) | 00:25 |
| mntmn | furan: or http://dump.mntmn.com/IMX8MDQLQRMref2.pdf ;) | 00:25 |
| furan | thx! | 00:25 |
| mntmn | furan: it is because each of DCSS and LCDIF can only output to one display | 00:26 |
| mntmn | (on this SoC) | 00:26 |
| furan | ok | 00:26 |
| mntmn | furan: DCSS is the more capable engine, it has compression and buffers and some fancy things, LCDIF is very old in spartan (was used in many old imx generations) | 00:27 |
| mntmn | very old _and_ spartan | 00:27 |
| furan | ic | 00:29 |
| mntmn | furan: and yeah do go back to your original question, either lcdif (also called mxsfb) or dcss (i think also called imx-dcss) are the kernel drivers that setup up /dev/fb* | 00:30 |
| mntmn | furan: dcss can feed either into dsi or hdmi (there is a register that toggles it), but lcdif can feed _only_ into dsi | 00:30 |
| furan | ok, and dcss can't drive two outputs at once, and that's why you have to use lcdif when hdmi is enabled (and using dcss)? | 00:34 |
| furan | so, when HDMI is enabled and you are driving the panel and HDMI, LCDIF is used for the panel, and DCSS for HDMI | 00:35 |
| mntmn | furan: correct | 00:35 |
| furan | cool | 00:35 |
| mntmn | furan: it's possible that as root and using a tool that can write to arbitrary memory, you might be able to poke arbitrary memory location into dcss's scanout address | 00:36 |
| mntmn | furan: all the IP registers are in 32-bit space | 00:36 |
| furan | devmem2 and all that | 00:36 |
| mntmn | yes | 00:36 |
| mntmn | furan: dcss is a bit more complicated than lcdif because it has some "kick" IRQ that has to load something every frame or so | 00:37 |
| furan | ok. I am going to try and make it scan out of PCIe BAR just to see it fall over. I think if I want to scan out of that, I will have to copy to RAM first. | 00:37 |
| mntmn | furan: i like your thinking :3 | 00:37 |
| furan | oh that synchronization is (maybe?) preferred | 00:37 |
| furan | what does it look like when display underrun happens? everything becomes a blurry gray blob? :) | 00:38 |
| mntmn | PCIe BAR is limited to like 128MB btw, a bit of a bummer... but still better than i.MX6's 14 MB | 00:38 |
| furan | oh bar size? | 00:38 |
| furan | that's ok | 00:38 |
| mntmn | yes | 00:38 |
| mntmn | furan: i have only seen it on LCDIF. it becomes VHS-style glitchy and if underrun persists for a second or so, it turns off | 00:39 |
| furan | ok | 00:39 |
| furan | what's the difference between IMX8MM and IMX8MQ? | 00:39 |
| furan | (which one is reform) | 00:39 |
| furan | https://community.nxp.com/t5/i-MX-Processors/PCIe-Bandwidth/m-p/1063269?lightbox-message-images-1063270=110867iCE486EC2A8705F45#M156271 | 00:39 |
| furan | looks like they benched DMA read and write | 00:40 |
| furan | for PCIe | 00:40 |
| mntmn | furan: reform is IMX8MQ | 00:40 |
| furan | excellent | 00:40 |
| mntmn | furan: handy table https://www.nxp.com/docs/en/supporting-information/IMX8LAYERCMPR.pdf | 00:40 |
| mntmn | (reform will also soon have LS1028A as an option which is in that table too) | 00:40 |
| mntmn | furan: IMX8MM is "mini" | 00:41 |
| furan | thx | 00:41 |
| - ephase (QUIT: Quit: WeeChat 3.2) (~ephase@2a01:e0a:168:1211:4ec7:a5dc:3f69:b03e) | 01:17 | |
| + furan_ (~ian@38.75.239.86) | 01:26 | |
| - furan (QUIT: Read error: Connection reset by peer) (~ian@38.75.239.86) | 01:26 | |
| * furan_ -> furan | 01:35 | |
| furan | any idea what's causing me this git redirect error? https://i.imgur.com/AvaHaGQ.png | 01:51 |
| furan | ah I think I fixed it changed URL to mnt.re one | 02:03 |
| mntmn | ah yeah, sorry, i thought i had already updated the url in that branch | 02:18 |
| furan | np. I built a new kernel which resulted in a new Image. I copied it to mmcblk1p1/Image, and now I boot to black screen, so I must have done something wrong | 02:31 |
| furan | only thing I changed was to enable realtek USB wifi builtin | 02:31 |
| furan | (this is a kernel from mkkernel.sh) | 02:32 |
| furan | I think it sort of works because the trackpad LEDs flash after a second or two | 02:33 |
| doppler | chatting from Reform… this thing is a blast | 02:34 |
| furan | \o/ | 02:34 |
| doppler | quite a bit of graphics flickering/glitching with PCI-E activity (seems to be, at least), so I'll have to get this thing up-to-date soon | 02:34 |
| doppler | (IIRC that's a known and fixed issue) | 02:35 |
| khm | doppler: do you have yours booting off sd card still | 02:35 |
| doppler | I do | 02:36 |
| - Neelfyn (QUIT: Quit: Connection closed for inactivity) (uid180106@id-180106.stonehaven.irccloud.com) | 02:45 | |
| furan | ah I think I forgot the .dtb files | 03:10 |
| mntmn | doppler: it is, but the fix is supposed to be in the v2 image... need to check on that tomorrow | 03:14 |
| mntmn | need to sleep, 3am here! i'll be available for questions again tomorrow | 03:15 |
| doppler | late night! hope you get some rest | 03:17 |
| furan | all set again, just needed to copy the right device tree binary | 03:23 |
| furan | night! | 03:23 |
| - jryans (QUIT: Quit: Reconnecting) (~jryansmat@2001:470:69fc:105::1d) | 04:07 | |
| + jryans (~jryans@2001:470:69fc:105::1d) | 04:07 | |
| jackhill | sbp: I've had my reform for 5 days, but have been busy this week. I just finished doing the "foreign distribution" installation as a package manager on top of the Debian system using the Guix-provided shell script. As expected, that worked well. | 06:31 |
| jackhill | sbp: that will be a stepping stone for installing Guix system, but as I mentioned in that thread, I want to use Guix-managed kernels, so that will be more work. Particularly since this is my first time interacting with uboot :) | 06:32 |
| jackhill | mtm: ^^ I'm happy to chat/swap notes about getting Guix working. | 06:33 |
| jackhill | sbp: oh, and I haven't done uart stuff either. I'm waiting on the µArt from Crowd Supply which I figure will be helpful when hacking on the uboot config | 06:35 |
| jackhill | mntmn: I beleive I've asked this before, but the answer is not in my brain … does Reform/IMX8 support sending audio along with video over HDMI? | 06:41 |
| jackhill | mntmn: oh, also, I took a nap today, that after making a more-awesome laptop than Apple, you were making a more-awesome car than Tesla. All open hardware, software, etc. Funny subconscious. | 06:43 |
| jackhill | sbp: of sourse I could have done an `apt install guix` now thanks to vagrantc's work, but I stuck with the tried-and-true shellscript method. | 06:44 |
| + odnes (~odnes@109-178-242-34.pat.ren.cosmote.net) | 06:46 | |
| ex-parrot | mntmn: https://img.hotplate.co.nz/ocNU3Ml9 | 07:08 |
| ex-parrot | The tracking just abruptly stopped working after it entered NZ, then this appeared a couple of hours ago :) | 07:09 |
| ex-parrot | https://img.hotplate.co.nz/Fueuf7fm | 07:13 |
| ex-parrot | this machine is incredible. such a beautiful design, the unboxing and manual and everything are perfect. amazing work mntmn and team | 07:24 |
| ex-parrot | absolutely worth the wait | 07:24 |
| ex-parrot | I haven’t read the manual yet but is the keyboard PCB meant to be kind of “floating” in the case? | 07:24 |
| - furan (QUIT: Ping timeout: 268 seconds) (~ian@38.75.239.86) | 07:26 | |
| indefini[m] | ex-parrot: I dont think so.. | 07:36 |
| indefini[m] | did you build it? | 07:37 |
| indefini[m] | or maybe yes if by floating you mean there is some space under :) | 07:38 |
| ex-parrot | No it came assembled | 07:46 |
| ex-parrot | I’ll make a video shortly. By “floating” I mean the keyboard doesn’t seem totally fixed in place, it slides around in the housing a little bit | 07:47 |
| scops | O.o the keys mounted with a little play...? to you mean this? the keyboard pcb should be fixed with 6 screws | 07:53 |
| scops | (12 if you count the frame too) | 07:53 |
| ex-parrot | Huh ok mine is definitely | 07:55 |
| ex-parrot | Not quite right then | 07:55 |
| ex-parrot | The whole PCB slides around in there | 07:55 |
| ex-parrot | ACTION fetches a screwdriver | 07:56 |
| scops | yeah... you need 6 screws the same size as the frame screws to fix the keyboard pcb in place | 07:57 |
| ex-parrot | Mm ok I guess maybe that got missed in assembly | 07:57 |
| scops | (the ones on the pcb normaly havent a flatten head... besides that they are identically) | 07:58 |
| ex-parrot | Hmmmm a screw in the bottom case was almost entirely out too. I wonder if they just vibrated loose in shipping | 08:00 |
| ex-parrot | ah I see the problem; the keyboard screws are just missing entirely :D mntmn you may want to check that on other units | 08:06 |
| ex-parrot | https://img.hotplate.co.nz/bivHUxJd I assume those holes were meant to have screws in them 😂 | 08:12 |
| scops | ô.o strange... | 08:36 |
| ex-parrot | It’s no big deal, I can just order a set of the correct screws here. In the mean time I found a few mismatched ones of the right thread / length to secure it with | 08:41 |
| khm | Mine wasn't screwed down either, but I got the DIY version. | 08:57 |
| + benj234 (~benj234@p5099baa2.dip0.t-ipconnect.de) | 09:03 | |
| - benj234 (PART: Killed buffer) (~benj234@p5099baa2.dip0.t-ipconnect.de) | 09:03 | |
| + benj234 (~benj234@p5099baa2.dip0.t-ipconnect.de) | 09:04 | |
| benj234 | hi, if I order now from germany, do I still get it within ~4 weeks? | 09:06 |
| ex-parrot | I think the backlog of crowd funding orders need to be completed first | 09:28 |
| + reform13374 (~fincham@118.149.85.192) | 09:30 | |
| - _Bnu (QUIT: Read error: Connection reset by peer) (~beeanyew@89-160-120-72.cust.bredband2.com) | 09:30 | |
| reform13374 | hello | 09:30 |
| reform13374 | the machine lives :) | 09:31 |
| reform13374 | there's a bit of backlight bleed around the bottom edge, hopefully I can tweak the panel mounting a bit to improve that | 09:32 |
| reform13374 | otherwise it's running perfectly | 09:32 |
| reform13374 | ok time to get xwayland going so I can run bb :) | 09:38 |
| - reform13374 (QUIT: Quit: Leaving) (~fincham@118.149.85.192) | 09:38 | |
| ex-parrot | next question: how do I get audio working? Pulse sees no output devices | 09:47 |
| ex-parrot | Oh that’s in the FAQ, nm | 09:47 |
| - odnes (QUIT: Read error: No route to host) (~odnes@109-178-242-34.pat.ren.cosmote.net) | 09:51 | |
| ex-parrot | ah the aspect ratio of the case is super nice, just like a pre-touchpad thinkpad :P | 10:10 |
| scops | <reform13374 "there's a bit of backlight bleed"> i`ve got the same... if you can fix it tell us how you have done it :) | 10:10 |
| scops | hmp | 10:10 |
| scops | g | 10:10 |
| scops | x.x => hmpf. | 10:10 |
| ex-parrot | on the Novena I got rid of it by just loosening the mounts for the display and shimming it slightly, it’s very sensitive to any kind of torque at the mounts | 10:11 |
| scops | <ex-parrot "ah the aspect ratio of the case "> indeed... i wish the trackpad area would be bigger but it's ok. the size of the hole device is great imho | 10:11 |
| ex-parrot | For me the size is perfect, I can type without taking off a wristwatch :P | 10:11 |
| scops | <ex-parrot "on the Novena I got rid of it by"> hmmm i will try that... | 10:12 |
| ex-parrot | I can’t promise this display mounting arrangement would allow for similar manipulations thoighn | 10:12 |
| ex-parrot | How do I get the MNT logo back on the keyboard oled? :P | 10:12 |
| scops | xD | 10:13 |
| ex-parrot | I guess it would be only proper if use the machine to order some screws for itself | 10:16 |
| mntmn | ex-parrot: i apologize, assembly error that i didn't catch @ missing keyboard pcb screws | 10:17 |
| ex-parrot | no worries! I’m not bothered about it | 10:17 |
| ex-parrot | you’ve made it extremely easy to address :P | 10:18 |
| mntmn | ooof!! | 10:18 |
| mntmn | ok so about the logo on the OLED, this is shown only after poweron/reset | 10:18 |
| ex-parrot | Ah I figured as much | 10:19 |
| mntmn | backlight bleed surprises me tbh... did you try to untighten the screws on the screen frame in that place? | 10:20 |
| ex-parrot | No I haven’t yet, I will do | 10:20 |
| ex-parrot | I’ll try get a photo of it | 10:20 |
| mntmn | otherwise it could be a transport issue | 10:20 |
| mntmn | ok thanks | 10:20 |
| ex-parrot | how do I adjust the display backlight from sway? | 10:21 |
| mntmn | scops: same for you, did you try loosening the screw(s) of the display a bit? | 10:21 |
| mntmn | ex-parrot: super + f1/f2 | 10:21 |
| ex-parrot | here’s a picture of the backlight at the bottom: https://img.hotplate.co.nz/ojX8eYey | 10:23 |
| ex-parrot | It’s not quite as bad IRL as it looks in the photo | 10:23 |
| ex-parrot | Loosening the screws makes it wobble a little but doesn’t change it much | 10:24 |
| mntmn | ex-parrot: hmm, that's not ideal. i hope the panel wasn't damaged in transit. can you open the laptop 180 degrees (like an open book) and then remove the screen frame? (better turn it off first) | 10:25 |
| ex-parrot | It’s probably comparable to my old Lenovo X250, maybe a tiny bit worse | 10:25 |
| ex-parrot | Sure | 10:25 |
| mntmn | normally bleed happens when something is squeezing the display, as you said | 10:26 |
| ex-parrot | Ok with the bezel off it’s a lot better looking (I’m being careful not to tilt the display too much with it off) https://img.hotplate.co.nz/IGUDKKa1 | 10:29 |
| ex-parrot | just slightly “warm” in the bottom left | 10:30 |
| ex-parrot | As I lifted the bezel off there was a bit of a “ping” like it was under some stress or something. I’ll put it back on and see if it’s better now | 10:31 |
| ex-parrot | It’s definitely a little | 10:31 |
| ex-parrot | warped but nothing major. It’s warped “outwards” in the middle at the bottom slightly | 10:32 |
| mntmn | oh ok. thanks for checking! i see you have the "side" connector version of the panel | 10:33 |
| ex-parrot | this is unit 37 if that’s useful to know :P | 10:34 |
| mntmn | the backlight looks a bit uneven in the photo, is it visible like that in real life? | 10:34 |
| ex-parrot | yes but it’s more subtle than in the photo | 10:34 |
| mntmn | hmm. we should probably send you another panel to compare then | 10:35 |
| ex-parrot | if you want to you could, but this one is really not too bad | 10:35 |
| ex-parrot | I’m just putting the bezel back on now, I’ll see if the backlight bleed comes back | 10:35 |
| mntmn | ok | 10:36 |
| ex-parrot | I’m assuming a bit of bent ness in the aluminium panels is expected? | 10:36 |
| ex-parrot | Ok yep the backlight bleed comes back when I install the bottom LCD cover screws | 10:39 |
| mntmn | yes, aluminum frame warping is normal | 10:39 |
| ex-parrot | Good 👍 | 10:39 |
| ex-parrot | They’re really pleasantly tactile parts to hold somehow | 10:40 |
| mntmn | ex-parrot: ok, i will talk to my colleague about the screen. i think this version of the panel has some tape to hold it in place somewhere. | 10:40 |
| ex-parrot | Ah right | 10:40 |
| ex-parrot | I’m really not too worried about it, it’s a similar level of bleed to what I’ve seen on eg Lenovo machines | 10:41 |
| mntmn | maybe it's too thick or the panel is otherwise pressed, i will try to replicate here | 10:41 |
| mntmn | ok | 10:41 |
| ex-parrot | https://img.hotplate.co.nz/vrCu34hc a bit of finessing with the screws has improved it too, a quarter turn looser helped | 10:41 |
| ex-parrot | But yeah if I can just remove some excess padding or something and have it luxuriously perfect then I wouldn’t complain either :P | 10:41 |
| ex-parrot | I’m very happy with the machine, glad it made it unscathed on its long journey | 10:42 |
| ex-parrot | for some reason connman really doesn’t want to connect to my wifi but it will talk to my phone hotspot at least, I’ll need to dig in to that | 10:43 |
| ex-parrot | https://img.hotplate.co.nz/zsEIJvx0 ordering myself some screws for the keyboard \m/ | 10:45 |
| ex-parrot | other nice things: the ctrl / space bar keys land exactly where my hands expected them | 10:54 |
| ex-parrot | I am enjoying the nice industrial feel of the trackball as well, it’s so good to have a pointer with some physical inertia again | 10:54 |
| ex-parrot | the case is also really nice! Feels really premium. Reminds me a bit of a TRS-80 Model 100 case | 10:55 |
| ex-parrot | and now I’m going to bed. Thanks everyone | 10:55 |
| scops | <mntmn "scops: same for you, did you try"> yeah it worked. but still i have two dead (white) pixels... :/ *shrug maybe i replace the panel somewhere in the future if you offer a oled or something like this xD heh | 10:56 |
| mntmn | scops: dead pixels? oof. since the beginning? | 10:58 |
| scops | yes | 10:58 |
| mntmn | thank you and good night ex-parrot! | 10:58 |
| scops | <ex-parrot "and now I’m going to bed. Thanks"> gn8 | 10:58 |
| mntmn | scops: mind to show me a photo? i would like to see what i missed ^^ | 10:59 |
| scops | check :) i have a meeting i send it to you in a hour or so | 10:59 |
| - benj234 (QUIT: Read error: Connection reset by peer) (~benj234@p5099baa2.dip0.t-ipconnect.de) | 11:12 | |
| + _Bnu (~beeanyew@89-160-120-72.cust.bredband2.com) | 11:15 | |
| + mjw (~mjw_@2001:1c06:2487:f800:9e5c:8eff:fe8f:a440) | 11:28 | |
| - wiedi (QUIT: Read error: Connection reset by peer) (~wiedi@37.228.191.159) | 11:29 | |
| + wiedi_ (~wiedi@37.228.191.159) | 11:29 | |
| - erlehmann (QUIT: Ping timeout: 268 seconds) (~erle@dynamic-046-114-037-060.46.114.pool.telefonica.de) | 12:10 | |
| - wiedi_ (QUIT: Read error: Connection reset by peer) (~wiedi@37.228.191.159) | 12:48 | |
| + Neelfyn (uid180106@id-180106.stonehaven.irccloud.com) | 13:10 | |
| + odnes (~odnes@109-178-242-34.pat.ren.cosmote.net) | 13:11 | |
| scops | upgraded to wayfire 0.7.2 (needed a recompile of drm, wlroots,... new libseat...) i really hope wayfire will find it's way to official repos once it is more mature... | 13:44 |
| mntmn | scops: i think it will, i just need to set aside some time to thoroughly test it. maybe you can do a small video (screen capture) of your setup if you like | 13:45 |
| + erlehmann (~erle@dynamic-046-114-033-094.46.114.pool.telefonica.de) | 14:49 | |
| - odnes (QUIT: Read error: Connection reset by peer) (~odnes@109-178-242-34.pat.ren.cosmote.net) | 14:51 | |
| + odnes (~odnes@109-178-242-34.pat.ren.cosmote.net) | 14:51 | |
| + wiedi (~wiedi@37.228.191.159) | 15:08 | |
| - odnes (QUIT: Ping timeout: 265 seconds) (~odnes@109-178-242-34.pat.ren.cosmote.net) | 15:24 | |
| - wiedi (QUIT: Read error: Connection reset by peer) (~wiedi@37.228.191.159) | 18:33 | |
| + furan (~ian@38.75.239.86) | 18:34 | |
| + wiedi (~wiedi@37.228.191.159) | 19:07 | |
| - mjw (QUIT: Quit: Leaving) (~mjw_@2001:1c06:2487:f800:9e5c:8eff:fe8f:a440) | 19:09 | |
| - furan (QUIT: Ping timeout: 258 seconds) (~ian@38.75.239.86) | 19:42 | |
| - wiedi (QUIT: Read error: Connection reset by peer) (~wiedi@37.228.191.159) | 20:52 | |
| + ephase (~ephase@2a01:e0a:168:1211:ca5d:edf5:d218:6f8d) | 21:28 | |
| + odnes (~odnes@109-178-242-34.pat.ren.cosmote.net) | 22:10 | |
| + wiedi (~wiedi@37.228.191.159) | 22:23 | |
| + mjw (~mark@herd.wildebeest.org) | 22:36 | |
| - odnes (QUIT: Read error: Connection reset by peer) (~odnes@109-178-242-34.pat.ren.cosmote.net) | 22:37 | |
| + odnes (~odnes@109-178-242-34.pat.ren.cosmote.net) | 22:46 | |
| scops | mntmn: in your thread about recompiling ffmpeg you should add the switch "--enable-openssl" ... so that youtube (...) can be directly played with mpv :) | 23:08 |
| scops | *recompiling | 23:08 |
| mntmn | scops: oh ok | 23:23 |
| - odnes (QUIT: Read error: Connection reset by peer) (~odnes@109-178-242-34.pat.ren.cosmote.net) | 23:45 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
