| chartreuse | Also an update on the headphone jack detect, the 220k pullup at R144 seems to be working perfectly. Though since 9front handles it fine there must be a way of fixing it in software, and I suspect that involves using the anti-pop capacitor discharge resistors | 01:13 |
|---|---|---|
| jfred | Now that I'm getting the hang of kicad a bit more I'm noticing some things about the Pocket top panel. I see there's a big square in the B.Mask layer, is that there for heat dissipation or something since it'd be exposed copper? | 01:39 |
| jfred | And I guess it's fine to overhang the edge of the PCB in one of the mask layers? That might actually make my life a lot easier when importing my design haha | 01:40 |
| chartreuse | Masks can be bigger than the board outline use, makes it easy when you have curved corners to just use a rectangular mask if you need to have some exposed copper in the corners | 01:45 |
| chartreuse | Silkscreen can also overlap the edge and just be cut off | 01:45 |
| jfred | Excellent, that's helpful, thanks! Means I don't have to worry so much about getting the edges just right | 02:01 |
| jfred | I do still need to make sure I don't have silkscreen and mask drawings overlapping though but that should be much easier | 02:02 |
| - mtm (QUIT: Ping timeout: 264 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:03 | |
| jfred | a little preview of the top panel I've been working on https://usercontent.irccloud-cdn.com/file/j5gEo104/1000009921.png | 02:03 |
| chartreuse | Took me a while to find where 9front is doing the audio config stuff for the reform (/sys/src/9/imx8/sai.c). And in an older commit I can find the wm8960 at /sys/src/cmd/aux/wm8960.c but not finding that in the latest one so not sure where it got moved too | 02:04 |
| chartreuse | Kicad and similar have a export setting to subtract soldermask from silkscreen so you won't have it on the exported file, just would need to only worry about the overlap where you don't intend it to be cutoff | 02:05 |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:05 | |
| jfred | Oh neat | 02:06 |
| jfred | Although in this case I think I'd actually want the reverse haha since my silkscreen design is meant to look like it's obscuring the soldermask design | 02:08 |
| - mtm (QUIT: Ping timeout: 268 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:37 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:39 | |
| - xktr (QUIT: Quit: leaving) (~xktr@user/xktr) | 02:47 | |
| - nsc (QUIT: Ping timeout: 264 seconds) (~nicolas@105-99-142-46.pool.kielnet.net) | 03:05 | |
| + nsc (~nicolas@19-97-142-46.pool.kielnet.net) | 03:07 | |
| + lifning (~lifning@user/lifning) | 03:11 | |
| - lifning (QUIT: Quit: Lost terminal) (~lifning@user/lifning) | 03:52 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 04:03 | |
| - nsc (QUIT: Ping timeout: 268 seconds) (~nicolas@19-97-142-46.pool.kielnet.net) | 04:04 | |
| + nsc (~nicolas@220-99-142-46.pool.kielnet.net) | 04:05 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 04:05 | |
| - vagrantc (QUIT: Ping timeout: 272 seconds) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 04:51 | |
| + lifning (~lifning@user/lifning) | 04:57 | |
| lifning | should the headphone jack on my pocket reform be showing up in pavucontrol when i have something plugged into it? there's nothing but Speakers in the Port dropdown. tried re-seating the little cable to the side-board on both ends just in case, but no difference | 05:00 |
| - Gooberpatrol66 (QUIT: Remote host closed the connection) (~Gooberpat@user/gooberpatrol66) | 05:00 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 05:00 | |
| + photomattmills (~photomatt@184-23-21-40.fiber.dynamic.sonic.net) | 07:09 | |
| photomattmills | do I need to do anything to get serial console output on the pocket reform? I've tried both ports, nothing on either S1 or S2 | 07:17 |
| josch | photomattmills: Lukas usually joins this channel only in about 2-3 hours, so maybe ping minute about your issue | 07:32 |
| photomattmills | yeah, I think we're at crossed timezones. might just post on the forum so we can debug async | 07:33 |
| ch | chartreuse: i think sys/src/cmd/reform/audio.c is what you are looking for | 07:58 |
| - photomattmills (QUIT: Ping timeout: 250 seconds) (~photomatt@184-23-21-40.fiber.dynamic.sonic.net) | 08:27 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 08:39 | |
| - vagrantc (QUIT: Client Quit) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 08:43 | |
| minute | lifning: no, this is still TODO, i don't know yet why the headphone output doesn't show up. but you can enable headphone output, volume etc in alsamixer (in a terminal) | 09:05 |
| lifning | ooh! thanks for the workaround :D | 09:52 |
| chartreuse | ch: Ah perfect, I didn't see that reform folder in cmd | 09:55 |
| josch | ex-parrot: did you test the debian-installer images on a imx8mq reform on an a311d reform? and if the former: did you try the bookworm or the bookworm-backports images? | 09:55 |
| chartreuse | A bit late for me tonight, but I'll dig into what registers that 9front is setting vs the linux drivers | 09:55 |
| josch | ex-parrot: for some reason the bookworm-backports kernel fails to load the xz-compressed modules | 09:56 |
| ex-parrot | josch: imx8mq, and "bookworm", the "bookworm-backports" image didn't boot | 09:56 |
| josch | aha! | 09:56 |
| ex-parrot | ok that would match | 09:56 |
| josch | yes, i know why now | 09:56 |
| ex-parrot | lovely | 09:56 |
| josch | bookworm-backports doesn't work in qemu either | 09:56 |
| ex-parrot | I haven't had a chance to re-test anything yet, I got sidetracked hacking on a different machine | 09:56 |
| josch | i'm currently trying to setup automated testing for debian-installer on the reform using openqa.debian.net | 09:56 |
| ex-parrot | nice! | 09:56 |
| josch | i have yet to find a way to tell qemu to create block devices that are named like eMMC devices because i want to be able to test that d-i can put /boot on the drive that is *not* sata or nvme | 09:57 |
| josch | hrm... maybe -drive if=sd suffices for now... | 09:59 |
| chartreuse | https://gitlab.com/qemu-project/qemu/-/issues/1747 | 10:04 |
| chartreuse | Doesn't sound like it has a emmc type, but it does have a sdhci type | 10:05 |
| josch | if=sd seems to be enough | 10:05 |
| josch | i now have a /dev/mmcblk0 | 10:05 |
| josch | magic options were -device sdhci-pci -device sd-card,drive=mydrive -drive id=mydrive,if=none,format=raw,file=emmc.img | 10:06 |
| + photomattmills (~photomatt@184-23-21-40.fiber.dynamic.sonic.net) | 10:09 | |
| - lifning (QUIT: Remote host closed the connection) (~lifning@user/lifning) | 10:26 | |
| josch | aha and apparently only one sd-card at once is possible but not multiple: https://gitlab.com/qemu-project/qemu/-/issues/54 | 10:27 |
| josch | but nice, i can also emulate nvme and now have a /dev/nvme0n1 -- this will be practical to test reform-setup-encrypted-nvme | 10:30 |
| - photomattmills (QUIT: Ping timeout: 250 seconds) (~photomatt@184-23-21-40.fiber.dynamic.sonic.net) | 10:38 | |
| josch | ex-parrot: could you, once you get the chance, take a photo of the partitioning menu on your imx8mq? | 11:10 |
| josch | i drew the partition editor using ansci escapes: https://mister-muffin.de/p/cvkC.html | 11:10 |
| josch | *ansi | 11:10 |
| josch | and now i need to edit the text so that it doesn't say QEMU anymore but whatever is shown on imx8mq instead | 11:11 |
| josch | i think if i put a screenshot on the reform.d.n page it might be more clear what should be done to get the partitioning right | 11:12 |
| josch | "screenshot" of d-i without using a raster graphic: https://reform.debian.net/d-i/ | 11:42 |
| josch | probably a png would be smaller than the HTML used to render this but the advantage is, that this makes the 'screenshot' editable later | 11:42 |
| - mjw (QUIT: Quit: Leaving) (~mjw@gnu.wildebeest.org) | 13:32 | |
| - Bertl (QUIT: Server closed connection) (herbert@IRC.13thfloor.at) | 13:35 | |
| + Bertl (herbert@IRC.13thfloor.at) | 13:35 | |
| * Guest3632 -> mjw | 13:58 | |
| - mtm (QUIT: Ping timeout: 264 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:03 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:05 | |
| - mtm (QUIT: Read error: Connection reset by peer) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 15:54 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 15:55 | |
| - inhji (QUIT: Quit: The Lounge - https://thelounge.chat) (~inhji@2a01:4f8:1c0c:42e4::1) | 17:27 | |
| josch | uff.. these nodejs dependencies remind me why we cannot yet ship the Iosevka font in debian... :( | 17:29 |
| + inhji (~inhji@2a01:4f8:1c0c:42e4::1) | 17:29 | |
| swivel | hrm I'm still having occasional problems with my imx8 reform becoming unresponsive after running for days-weeks in a mostly idle use case, doesn't seem to be prevented by turning off dpms/console blanking etc | 17:37 |
| swivel | i use it at work as a terminal server, mostly remotely accessed via ethernet+ssh | 17:37 |
| swivel | with all its usb ports connected to serial consoles via usb2serial adapters | 17:38 |
| sigrid | I've considered it might have been related to having a wifi module, and for quite a while I stopped having hangs on mine, but once the summer came (read: higher temperatures) the hangs appeared | 17:39 |
| swivel | it just randomly stops being pingable and when I try use it at the keyboard the screen is black and nothing works except circle key to cause a reset via the oled/lpc | 17:39 |
| sigrid | theories welcome | 17:39 |
| swivel | sigrid: hmm so you have hangs too, and suspect it's heat related? | 17:39 |
| sigrid | "for quite a while I stopped having hangs on mine" -> because I took the wifi module out | 17:39 |
| sigrid | yes | 17:39 |
| josch | swivel: you have it running sway or in a tty? | 17:40 |
| sigrid | in retrospect I think it'd make sense taking wifi module out would help in winter - it produces heat too | 17:40 |
| swivel | josch: nah it's just sitting at a console | 17:40 |
| swivel | josch: not even past the login prompt | 17:40 |
| sigrid | but it only ever seems to hang while in idle | 17:40 |
| sigrid | like, *hours* in idle | 17:41 |
| josch | the only input i can give in the matter is, that i had my imx8mq turned on 24/7 (in sway) and stopping to use dpms on/off completely made my random hangs go away (even in summer -- the thing is sitting on a couch cushion) | 17:41 |
| sigrid | maybe it's both, temperature and dpms? I can't imagine why though | 17:42 |
| josch | swivel: how do you turn the screen brightness off when you are in console? or do you never turn it off? | 17:42 |
| swivel | josch: it's just sitting there with the login prompt visible 24x7 | 17:43 |
| josch | okay, then indeed it cannot be dpms | 17:44 |
| swivel | i'd rather it didn't but was told that may be the cause of the hangs, didn't seem to change the hangs | 17:44 |
| josch | swivel: can you maybe try and see if you also get the hangs when you idle in sway instead of idling at the tty? | 17:45 |
| chartreuse | I'd be surprised if it was heat though, the reform doesn't get particularilly hot. I did add a thermal pad to mine between the heatsink and rear case so it's not 1:1, but mine basically sits pretty solid around 45-50C cpu temps | 17:45 |
| swivel | josch: i can | 17:47 |
| chartreuse | I've had lockups like that before too though, quite rarely though | 17:47 |
| josch | swivel: it's a super long shot but maybe at least something to try if there are no other things to try... | 17:48 |
| swivel | willing to try pretty much anything, since I use it at work to access the hardware on my desk when out of office... it's quite annoying when it's unavailable | 17:48 |
| chartreuse | I'm not using the wifi card that was offered with the reform (bought it as a kit), but it's also an Ath9k one (a 3x3 version from a mac) | 17:48 |
| josch | swivel: ah indeed i also am not using the wifi card supplied with the reform | 17:49 |
| swivel | i have the supplied wifi card installed but don't actually use it | 17:49 |
| chartreuse | If it's frequent enough to debug, maybe worth hooking up the serial port to another computer and logging it till it happens and see if anything is printed? | 17:50 |
| josch | i did that with mine when i had the dpms lockups and there was nothing in dmesg or over serial before the lockup -- would just stop | 17:53 |
| josch | not saying that you will not have anything either, but maybe do not get your hopes up too much ;) | 17:53 |
| swivel | i'd love to but don't really have time to mess with it | 17:55 |
| swivel | my day job is already an infinite backlog of debugging broken embedded linux junk | 17:55 |
| sigrid | maybe you can test booting with enable_wait_mode=off kernel param | 17:56 |
| sigrid | there is an errata item on imx8mq that looks suspiciously related | 17:56 |
| swivel | oh? cursory search makes it look more related to imx6? | 17:57 |
| swivel | e.g. https://community.nxp.com/t5/i-MX-Processors/Implications-of-setting-enable-wait-mode-off-in-3-0-35/m-p/234080 | 17:57 |
| sigrid | "e11174: CA53: Cannot enter WAIT mode" | 17:58 |
| sigrid | https://www.nxp.com/docs/en/errata/IMX8MDQLQ_0N14W.pdf | 17:58 |
| swivel | oof | 17:59 |
| sigrid | I can try doing the same here on 9front and see if it helps | 17:59 |
| sigrid | all I need to test this is to keep the laptop on at night | 17:59 |
| + xktr (~xktr@user/xktr) | 17:59 | |
| swivel | i'll add it on mine... is this supposed to be a power saving feature? | 18:00 |
| sigrid | yeah | 18:01 |
| swivel | mine's constantly plugged into AC so not particularly concerned with battery life, but eventually I wanted to get a second reform and use it at cafes etc. | 18:01 |
| sigrid | idk what's the impact on battery when disabling this. I imagine one could make a workaround that keeps WAIT mode enabled but makes the core not enter it when there *are* pending interrupts | 18:02 |
| swivel | considering the librem5 *is* power sensitive, I'm surprised this isn't prioritized for fixing properly | 18:03 |
| swivel | i guess they haven't really sold enough of those to build a thriving community | 18:03 |
| sigrid | hmmm enable_wait_mode isn't even in mainline linux | 18:10 |
| sigrid | maybe it's indeed only available for imx6* on old kernels | 18:12 |
| swivel | bummer | 18:28 |
| swivel | i spose another option is to just get a different SoM like the cm4 and see if the lockups go away | 18:28 |
| swivel | here I was thinking going imx8 was going to be the most mature and stable option | 18:28 |
| sigrid | I've disabled scu+l2 power down here and will see if it's related in any way | 18:41 |
| josch | swivel: in terms of stability, my a311d seems to be as stable as the imx8mq (if you disable dmps for both) | 19:01 |
| josch | (i still have my reform on 24/7 because there is no suspend) | 19:01 |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:14 | |
| * mjw -> Guest3664 | 19:16 | |
| - Guest3664 (QUIT: Killed (platinum.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:16 | |
| * mark_ -> mjw | 19:16 | |
| + Guest3664 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:16 | |
| - Gooberpatrol66 (QUIT: Ping timeout: 246 seconds) (~Gooberpat@user/gooberpatrol66) | 19:37 | |
| + gustav28 (~gustav@c-1134524e.019-141-67626730.bbcust.telenor.se) | 20:02 | |
| mhoye | Random q: is there a way to use the little 32-bit core on the imx8 ? | 20:30 |
| - midfavila (QUIT: Quit: Lost terminal) (midfavila@sdf.org) | 20:31 | |
| mhoye | Setting aside the whole 'why would you want that' question | 20:34 |
| chartreuse | Didn't know about that core and was just looking looking through the manual at it now. I'm guessing that it's going completely unused under linux. You'd probably need to write a kernel module to bring it up and point it at some code, and then set up the messaging unit for talking to it afterwords | 21:49 |
| - tretinha (PART: !!unknown attribute: msg!!) (3a571d9f43@2a03:6000:1812:100::1151) | 22:01 | |
| + tretinha (3a571d9f43@2a03:6000:1812:100::1151) | 22:01 | |
| op_4 | hey all just recieved my pocket reform and having a great time. Looking over the module page https://mntre.com/modularity.html i'd like to get a sense for what would be the least power hungry platform. As things stand with some effort i can obtain 6h ish from the pocket | 22:06 |
| op_4 | so is this information recorded somewhere? TDP/current/flops per W or something? | 22:06 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-1134524e.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| + hl_ (~hl@user/hl) | 23:18 | |
| - hl (QUIT: Ping timeout: 256 seconds) (~hl@user/hl) | 23:20 | |
| - hl_ (QUIT: Client Quit) (~hl@user/hl) | 23:22 | |
| minute | chartreuse mhoye: i think what you're looking for is remoteproc | 23:24 |
| + hl (~hl@user/hl) | 23:24 | |
| minute | op_4: not really measured yet but my current rec would be the rcm4+bpi (a311d), also available at crowd supply atm for a good price | 23:25 |
| svp | was wondering: from what i understand, since the modules' single-pcie-bus designs seem to prefer the m.2 slot for their pcie bus, that would mean the mpcie slot is going to be useless, right? | 23:25 |
| minute | chartreuse mhoye: https://lore.kernel.org/linux-remoteproc/20210805163943.GE3205691@p14s/T/ | 23:26 |
| minute | svp: on which module? in reform most modules have pcie on the mpcie slot | 23:27 |
| minute | svp: and we ship an m2 adapter for it | 23:27 |
| svp | rcm4, but i guess i was hasty in my guess | 23:28 |
| svp | still, i'm just a little puzzled at the decision not to connect mpcie 38 and 36 (usb d+/d-), it wouldve probably made the slot more useful in case the situation i was expecting was the case | 23:29 |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 23:38 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
