| - staticbunny (QUIT: Ping timeout: 246 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 00:05 | |
| - cobra (QUIT: Ping timeout: 244 seconds) (~cobra@user/Cobra) | 00:05 | |
| - kensanata (QUIT: Quit: Ping timeout (120 seconds)) (~alex@user/kensanata) | 00:05 | |
| + kensanata (~alex@user/kensanata) | 00:06 | |
| + cobra (~cobra@user/Cobra) | 00:10 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 00:16 | |
| - staticbunny (QUIT: Ping timeout: 245 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 00:27 | |
| minute | ch: i hope you'll remind me tomorrow about this question | 00:34 |
|---|---|---|
| ch | :) | 00:34 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 00:44 | |
| - staticbunny (QUIT: Ping timeout: 248 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 00:48 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 00:53 | |
| + _justin_kelly (~justinkel@user/justin-kelly/x-6011154) | 01:10 | |
| - staticbunny (QUIT: Ping timeout: 245 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 01:15 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 01:16 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 01:30 | |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 01:35 | |
| - NanoCodeBug (QUIT: Quit: Client closed) (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 01:35 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 01:37 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 01:41 | |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 01:45 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:00 | |
| - mtm (QUIT: Ping timeout: 248 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:04 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:04 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:05 | |
| - midfavila (QUIT: K-Lined) (midfavila@freeshell.org) | 02:14 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:19 | |
| - staticbunny (QUIT: Ping timeout: 245 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:32 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:34 | |
| - aloo_shu (QUIT: Ping timeout: 252 seconds) (~aloo_shu@90.166.99.189) | 02:38 | |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:39 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:40 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:44 | |
| ch | ah NanoCodeBug has left already | 02:55 |
| - xktr (QUIT: Ping timeout: 248 seconds) (~xktr@user/xktr) | 02:56 | |
| + xktr (~xktr@user/xktr) | 02:58 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 02:59 | |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 03:04 | |
| ch | yeah there's some bad interaction between the usb 5V source and the charger | 03:06 |
| ch | more tomorrow | 03:06 |
| - qbit (QUIT: Remote host closed the connection) (~qbit@mail.suah.dev) | 03:13 | |
| + qbit (~qbit@mail.suah.dev) | 03:13 | |
| - nsc (QUIT: Ping timeout: 260 seconds) (~nicolas@170-97-142-46.pool.kielnet.net) | 03:15 | |
| + nsc (~nicolas@i5C74DD8F.versanet.de) | 03:17 | |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 03:18 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 03:18 | |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 03:32 | |
| - sir-photch (QUIT: Ping timeout: 252 seconds) (~m-hy5poy@static.93.70.235.167.clients.your-server.de) | 03:35 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 03:48 | |
| + cobra (~cobra@user/Cobra) | 03:51 | |
| + xha (~xha@user/xha) | 03:52 | |
| - xha (QUIT: Client Quit) (~xha@user/xha) | 03:52 | |
| + xha (~xha@user/xha) | 03:53 | |
| - xha (QUIT: Client Quit) (~xha@user/xha) | 03:56 | |
| + xha (~xha@user/xha) | 03:58 | |
| - Gooberpatrol_66 (QUIT: Ping timeout: 248 seconds) (~Gooberpat@user/gooberpatrol66) | 04:00 | |
| - paperManu (QUIT: Ping timeout: 265 seconds) (~paperManu@198.16.214.40) | 04:10 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:14 | |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 04:21 | |
| + cobra (~cobra@user/Cobra) | 04:24 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:26 | |
| - staticbunny (QUIT: Ping timeout: 246 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:38 | |
| - chrcav (QUIT: Quit: leaving) (~chrcav@user/chrcav) | 04:40 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:42 | |
| - yuu (QUIT: Read error: Connection reset by peer) (sid267332@ilkley.irccloud.com) | 04:45 | |
| + yuu (sid267332@id-267332.ilkley.irccloud.com) | 04:45 | |
| - staticbunny (QUIT: Ping timeout: 255 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:47 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 04:47 | |
| + chrcav (~chrcav@user/chrcav) | 04:51 | |
| - staticbunny (QUIT: Ping timeout: 246 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 05:06 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 05:15 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 05:28 | |
| - staticbunny (QUIT: Ping timeout: 248 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 05:32 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 06:14 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 06:28 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 06:56 | |
| - staticbunny (QUIT: Ping timeout: 246 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 07:01 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 07:33 | |
| - staticbunny (QUIT: Ping timeout: 255 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 07:38 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 07:38 | |
| - staticbunny (QUIT: Ping timeout: 246 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 07:46 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 08:01 | |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 08:05 | |
| + sir-photch (~m-hy5poy@static.93.70.235.167.clients.your-server.de) | 08:13 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 08:20 | |
| josch | could somebody with a pocket reform maybe try the latest system image here: https://source.mnt.re/reform/reform-system-image/-/jobs/5844/artifacts/browse | 08:21 |
| josch | and check whether the wallpaper gets displayed correctly | 08:21 |
| josch | and that the fonts are as they should (Inter and JetBrains) | 08:22 |
| josch | thank you in advance! | 08:23 |
| josch | (in other news: reform-tools 1.53 is out) | 08:23 |
| - staticbunny (QUIT: Ping timeout: 245 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 08:39 | |
| - Gooberpatrol_66 (QUIT: Read error: Connection reset by peer) (~Gooberpat@user/gooberpatrol66) | 08:44 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 08:44 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 08:55 | |
| - sterni (QUIT: ) (~quassel@user/sterni) | 09:01 | |
| + sterni (~quassel@user/sterni) | 09:03 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 09:55 | |
| ch | bmaptool is a python thing, right? anyone have a source link? | 10:23 |
| ch | josch: will try | 10:23 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 10:23 | |
| josch | ch: https://github.com/yoctoproject/bmaptool | 10:24 |
| josch | ch: that's also linked from its d/control, see https://tracker.debian.org/pkg/bmap-tools | 10:25 |
| josch | why is the package called bmap-tools and the upstream project and the binary is called bmaptool? Look no further: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1076501 | 10:25 |
| ch | oh my | 10:26 |
| - staticbunny (QUIT: Ping timeout: 255 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 10:28 | |
| ch | minute: when you're around i'd like to discuss about the 5V source chip vs charger OTG source | 10:30 |
| ch | "pocket-reform-system-imx8mp.img.gz.bmap.html", sure gitlab, sure... | 10:32 |
| minute | ch: thanks for the ping, i've written it down now to get back to you once in the office | 10:34 |
| minute | i just found this https://bugs.launchpad.net/ubuntu/+source/linux-firmware/+bug/1993609 | 10:34 |
| minute | they are talking about using ath10k-sdio with qca9377 on imx6ull, this is quite similar to our setup | 10:35 |
| ch | ah very interesting | 10:36 |
| minute | here's the firmware blob https://launchpad.net/ubuntu/+source/linux-firmware/1.187.35 | 10:36 |
| minute | can we deduce which kernel version they were using there in 2022? | 10:36 |
| minute | nov 2022 | 10:36 |
| ch | probably, i'll dig a bit | 10:37 |
| ch | ath10k/QCA9377/hw1.0/firmware-sdio-5.bin is the referenced fw file | 10:37 |
| ch | https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/ath10k/QCA9377/hw1.0/firmware-5.bin | 10:38 |
| ch | er | 10:38 |
| ch | https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/commit/ath10k/QCA9377/hw1.0?id=d7904d5b07a9e2c4cdd9f8b2c5a5faa9c6e665cf | 10:38 |
| minute | > WLAN.TF.1.1.1-00061-QCATFSWPZ-1 | 10:39 |
| minute | well, it's always firmware-5.bin | 10:39 |
| ch | firmware-sdio-5.bin | 10:39 |
| ch | (*sdio*) | 10:40 |
| minute | yes, sorry | 10:40 |
| minute | it is the same file that debian ships | 10:41 |
| minute | these people use kernel 5.10 https://www.mail-archive.com/ath10k@lists.infradead.org/msg15597.html | 10:43 |
| minute | so i could try forward-porting the ath10k driver from 5.10.... | 10:44 |
| ch | complicated situation | 10:44 |
| minute | ch: could you try just unloading qcacld2 and loading ath10k_sdio? | 10:45 |
| minute | maybe it just works nowadays? | 10:45 |
| ch | focal mainline had 5.4, but from the firmware commit we know it cant work before 5.7 | 10:45 |
| ch | focal hwe is/was 5.11 | 10:45 |
| minute | ch: thanks | 10:45 |
| ch | will try in a moment | 10:46 |
| minute | (it is not loading automatically because of an entry in /etc/modprobe.d/reform.conf) | 10:46 |
| ch | should have installed the firmware files first m* | 10:48 |
| ch | m( | 10:48 |
| + gustav28 (~gustav@c-4135524e.019-141-67626730.bbcust.telenor.se) | 10:48 | |
| minute | ch: woops | 10:49 |
| josch | minute: /etc/modprobe.d/reform.conf moved to /usr/lib/modprobe.d/reform.conf since reform-tools 1.47 | 10:49 |
| minute | josch: ah sorry. i'm running a very old and many times upgraded system :D | 10:49 |
| minute | (on reform) | 10:49 |
| josch | i know, just mentioning because if you still have /etc/modprobe.d/reform.conf then editing that might not be enough to remove the "blacklist" entry | 10:50 |
| ch | hm nah, ath10k_sdio complains about writing to some address with an error | 10:50 |
| ch | ah but it connected now!? | 10:50 |
| ch | but no ip traffic passes through it | 10:51 |
| ch | i'll reboot just to see if qcacld left a bad state | 10:52 |
| minute | ch: ah this was kind of the state that i remember... maybe. not sure if i was able to connect | 10:52 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 10:55 | |
| ch | reboot doesnt help | 10:59 |
| minute | ch: thanks for testing. then this is still the state it was in when i first tried to get it to work and resorted to qcacld2 | 11:00 |
| minute | but maybe there is a long standing regression since 5.10 or 5.11 that noone found? | 11:00 |
| ch | wouldnt find that very surprising tbh | 11:01 |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 11:03 | |
| ch | https://gist.githubusercontent.com/zeha/61a863ff8b6b8819d4d39f309bfdbc63/raw/760783606e971bd5f785cf7652ca375607b3d27f/gistfile1.txt | 11:07 |
| ch | https://wireless.wiki.kernel.org/en/users/drivers/ath10k/firmware talks about firmware api 6, but i think the fw is api5? | 11:08 |
| minute | ch: it doesn't look like there's a firmware api 6 for sdio cards, only for other variants | 11:11 |
| minute | i.e. there's no firmware-sdio-6.bin for qca9377 sdio | 11:11 |
| minute | a collection of many ath10k versions :D https://github.com/greearb/ath10k-ct?tab=readme-ov-file | 11:12 |
| minute | could be worth a shot to try older ones | 11:12 |
| ch | so strange | 11:12 |
| ch | wondering if we need a "board file" | 11:15 |
| minute | another thing i could try, boot boundary's image and see how they set up the wifi https://www.ezurio.com/resources/software-announcements/debian-12-bookworm-unified-image-for-nitrogen8-boards | 11:15 |
| ch | possibly from a boundary image | 11:15 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 11:15 | |
| minute | ah they're using qcacld2 of course | 11:16 |
| ch | :D | 11:16 |
| ch | might still be interesting to see if it works better | 11:16 |
| minute | as seen here https://www.ezurio.com/resources/software-announcements/customizing-ubuntu-debian-kernels-on-i-mx-8-boards | 11:16 |
| minute | yes, might indeed be interesting | 11:16 |
| ch | BD-SDMAC WiFi/BT modules support right | 11:16 |
| minute | yes | 11:19 |
| ch | great, https://boundarydevices.com/product/bd_sdmac_wifi/ 404s, and https://www.ezurio.com/wireless-modules/wifi-modules-bluetooth doesnt show any ath10k thing :> | 11:19 |
| minute | yeah it's by silex | 11:19 |
| minute | it was before acquisition by laird/ezurio | 11:19 |
| minute | https://github.com/boundarydevices/qcacld-2.0/tree/boundary-CNSS.LEA.NRT_3.1-next | 11:19 |
| minute | there's a "next" branch | 11:19 |
| - staticbunny (QUIT: Ping timeout: 248 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 11:20 | |
| minute | https://github.com/boundarydevices/qcacld-2.0/commits/boundary-CNSS.LEA.NRT_3.1-next/ | 11:20 |
| + mjw (~mjw@gnu.wildebeest.org) | 11:20 | |
| minute | lol i think that's the outdated one | 11:22 |
| minute | because https://github.com/boundarydevices/qcacld-2.0/commits/boundary-CNSS.LEA.NRT_3.1 | 11:22 |
| ch | https://github.com/qca/qca-swiss-army-knife/blob/master/tools/scripts/ath10k/ath10k-bdencoder ugh | 11:25 |
| minute | ch: ah yeah i faintly remember dealing with that stuff | 11:25 |
| minute | i must have spent a lot of time with this but my memories are conveniently surpressed :D | 11:26 |
| ch | https://git.kernel.org/pub/scm/linux/kernel/git/ath/ath.git/commit/drivers/net/wireless/ath/ath10k?id=09b8cd69edcf2be04a781e1781e98e52a775c9ad | 11:26 |
| minute | anyway, it looks like i forked the boundary tree in late 2023 and it does not include all the latest commits from boundary... not sure if they would fix anything though. https://source.mnt.re/reform/qcacld2/-/commits/binary-deb?ref_type=heads | 11:26 |
| ch | looks like somebody had it working in 2021-11 :) | 11:26 |
| minute | ch: ha :D | 11:27 |
| minute | the problem is, ath10k has many files. which ones to copy from older kernels... | 11:28 |
| ch | josch: what exactly should i look for on the new system image? | 11:33 |
| ch | wtf https://forum.armbian.com/topic/20236-beelink-gt1-ultimate-s912-wifi-chip-qca9377-wifi-partially-works/ claims | 11:37 |
| ch | # replace special firmware (which seems more stable) | 11:37 |
| ch | ln -sf /lib/firmware/ath10k/QCA9377/hw1.0/firmware-sdio-5.bin /lib/firmware/ath10k/QCA9377/hw1.0/firmware-5.bin | 11:37 |
| minute | :D | 11:41 |
| josch | ch: does the wallpaper look as you'd expect it to look like or is it only filling the upper-left corner? | 11:42 |
| josch | ch: are the fonts correct? | 11:42 |
| ch | josch: wallpaper is fine | 11:43 |
| minute | ch: that ln doesn't make sense i think? it surely loads sdio5.bin? | 11:44 |
| ch | maybe superstition | 11:46 |
| ch | it looks for firmware-sdio-6.bin first, so i assume it loads firmware-sdio-5.bin then | 11:46 |
| ch | josch: https://per.namespace.at/~ch/x/IMG_5593.jpeg https://per.namespace.at/~ch/x/IMG_5594.jpeg https://per.namespace.at/~ch/x/IMG_5595.jpeg | 11:47 |
| ch | josch: doesnt look wrong to me font-wise | 11:48 |
| ch | josch: but i'm not the expert there :) | 11:48 |
| josch | ch: you can for example open gedit and in preferences -> fonts and colors | 11:48 |
| josch | it should say "use system fixed width font (JetBrains Mono 11)" | 11:49 |
| ch | ah like that | 11:49 |
| josch | wallpaper look good | 11:49 |
| ch | minute: https://gist.github.com/zerog2k/75e6e27aeef0494504ae9d56bf690097 this hints that armbian have 6.4 or 6.6 with sdio working | 11:50 |
| ch | josch: gedit shows what you say | 11:51 |
| josch | _hramrach: seems your recent suggestion regarding avoiding gsettings works now -- thank you for bringing that up! | 11:51 |
| josch | ch: thank you! | 11:51 |
| josch | minute: and it seems reform-wallpaper now also does what it should :) | 11:51 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 11:51 | |
| ch | josch: do you have an idea how to set the console font size correctly as soon as possible? | 11:52 |
| ch | josch: the cryptsetup prompt is still in tiny font | 11:52 |
| _hramrach | josch: thanks for making it haooen. Di not get to testing it yet | 11:52 |
| ch | at least on my install | 11:52 |
| josch | ch: i do not know out of the top of my head | 11:53 |
| ch | k | 11:53 |
| ch | easy solution is to not reboot :> | 11:53 |
| _hramrach | I also understand why upstream does not want to mess with the schema overrides: when they are distribution loose config files those are self-contained with the gsettings in them. When making a distribution package the situation is different. | 11:55 |
| _hramrach | ch: probably a service dependency problem, the cryptsetup whatever should depend on the console setup. but also the fonts may not be in the ramdisk at all, and would need to be addded | 11:56 |
| _hramrach | some distributions use plymouth which may affect this as well | 11:57 |
| ch | as if we had "services" | 11:57 |
| minute | josch: amazing @ wallpaper | 11:57 |
| minute | yeah no services in initramfs | 11:58 |
| _hramrach | there is compile-time default for the font but I think only one is supported. Changing to a very big fon would make the text too big on other systems | 11:58 |
| minute | you can probably include the font in the initramfs and activate it in a early hook script | 11:59 |
| - staticbunny (QUIT: Ping timeout: 245 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 11:59 | |
| josch | minute: yes, it would go into initramfs-tools/scripts/reform which already does the dmesg -n 7 for "MNT Reform 2 with LS1028A Module" | 12:04 |
| josch | those land in /usr/share/initramfs-tools/scripts/init-top and are thus run *before* the luks passphrase prompt | 12:05 |
| ch | gosh, the board-2.bin file contains the "eeprom" data for individual boards!? | 12:05 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 12:14 | |
| - mjw (QUIT: Ping timeout: 260 seconds) (~mjw@gnu.wildebeest.org) | 12:17 | |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 12:18 | |
| + _justin_kelly9 (~justinkel@user/justin-kelly/x-6011154) | 12:43 | |
| - _justin_kelly (QUIT: Read error: Connection reset by peer) (~justinkel@user/justin-kelly/x-6011154) | 12:45 | |
| * _justin_kelly9 -> _justin_kelly | 12:45 | |
| ch | does anyone know what fsl,sdio-async-interrupt-enabled does in the usdhc node in the dts? | 12:48 |
| + paperManu (~paperManu@198.16.214.40) | 12:48 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 12:49 | |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 12:56 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 13:11 | |
| minute | ch: yeah i faintly remember this. lots of hacks were involved with the sdio interface. for example, something about if tuning was used it can't work if there's also an sdio interrupt on one of the lines etc | 13:13 |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 13:16 | |
| minute | ok, i'm at a work computer and will look into your diode first ch | 13:18 |
| minute | strange, all kinds of apps are crashing here now on my pc, time for a reboot after many updates in the last days i guess | 13:18 |
| - Guest943 (QUIT: Ping timeout: 248 seconds) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 13:27 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 13:31 | |
| - staticbunny (QUIT: Ping timeout: 248 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 13:36 | |
| + Guest943 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 13:40 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 13:43 | |
| minute | ch: here's the explanation for sdio-async-interrupt-enabled https://github.com/TechNexion/linux-tn-imx/blob/80c1b69083bdfef4fc2b6bb68ec4ba2c16e7abfd/drivers/mmc/host/sdhci-esdhc-imx.c#L1642 | 13:46 |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 13:47 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 14:03 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:04 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:06 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 14:08 | |
| + mjw (~mjw@gnu.wildebeest.org) | 14:35 | |
| minute | ok, will now look into imx8mp again | 14:40 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 14:41 | |
| ch | ok back | 14:44 |
| ch | minute: i was wondering if the qca wifi problems are maybe caused by sth on the sdio side being wrong | 14:44 |
| ch | minute: re diode: i wrote on the forum because the diode causes the charger ic to not fast charge off my 30w charger | 14:45 |
| ch | minute: and imo it also prevents us from using the charger IC "otg mode" to provide 5v | 14:45 |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 14:45 | |
| ch | minute: i've removed the diode from my pocket and have a sysctl fw that uses the otg mode to provide the 5v for the usb host and that works very well! | 14:46 |
| minute | ch: yes, very possible @ sdio/mmc. that was also my hunch back then. but it is strange that the qcacld driver works with the same mmc/sdio config | 14:47 |
| ch | minute: and, using the otg mode has a few interesting advantages: 1) the charger ic doesn't see the 5v from U12 as an "AC" source, and also the measured discharge current looks more reasonable now (about 450mA vs 900mA) | 14:48 |
| ch | minute: i wonder if the qcacld papers over sdio problems. because even when it works its super slow for me | 14:49 |
| minute | ch: in my tests i got almost 200mibt/s from qcacld on pocket | 14:50 |
| minute | (that was maybe half a year ago though) | 14:50 |
| ch | i get maybe 200kB/s on qcacld. with the mt7612 and the same antenna i can get 40MB/s without a problem | 14:50 |
| minute | ok then there's some regression | 14:50 |
| ch | (but maybe thats another problem) | 14:50 |
| minute | ch: ok, back to the charger stuff for a moment, while my imx8mp image is flashing | 14:51 |
| ch | ok | 14:55 |
| ch | yeah so tl;dr imo D11 and U12 could (should?) be left out, the mps ic should probably get its OTG pin tied high | 14:59 |
| ch | bit hard to rework the last part tho :> | 14:59 |
| minute | ch: ok so i U12 is there so that we can toggle 5V output on that port and it also prevents backflow of USB_VBUS into +5V | 15:01 |
| minute | (because iirc it has built in back-to-back FETs) | 15:01 |
| minute | i didn't realize that the charger IC can supply 5V | 15:01 |
| minute | but... i need to double check if we're actually using that part number, i remember we had to change the part last minute because of availability | 15:02 |
| ch | i worked from the MP2762A datasheet | 15:02 |
| ch | ah in the charger BOM it says MP2650 | 15:03 |
| minute | yeah | 15:03 |
| minute | yeah so the actual chip is the MP2650 in this batch... it is very similar though | 15:04 |
| ch | i mean it works | 15:04 |
| minute | otg can be enabled via software, it seems, and you're doing that, yeah? | 15:04 |
| ch | yes | 15:05 |
| ch | so in the default setup, you need the OTG pin high and set OTG_EN | 15:05 |
| ch | i reconfigure the pin setup so the OTG pin acts as CMIN | 15:05 |
| minute | ok cool, well spotted. i'm thinking what to do about this | 15:05 |
| minute | i guess we need a list of things to go in a motherboard/charger board rev 2 | 15:05 |
| ch | which might be a problem if the pin is floating, but not sure what to do about that | 15:05 |
| minute | ok, i'll create an issue for hardware improvements in the pocket-reform repo that can go in the next rev, and that people can do on their own if they feel like hw hacking | 15:07 |
| ch | what do you think about generally leaving out d11? | 15:08 |
| ch | i'm not sure what its purpose is/was | 15:08 |
| minute | ch: i'm trying to remember why i wanted to prevent VCHG flowing into VBUS | 15:15 |
| minute | ch: i guess it was just some extra safety | 15:17 |
| minute | ch: did i get this right? https://source.mnt.re/reform/pocket-reform/-/issues/3 | 15:17 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 15:18 | |
| minute | meanwhile booting imx8mp pocket reform | 15:19 |
| minute | off the latest published system image | 15:19 |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 15:19 | |
| ch | MPS did a good job with the two ICs, to the host it makes no difference, except for the max charging current (which is so high the pocket doesnt want it anyway) | 15:20 |
| minute | yeah, we were mega lucky there | 15:21 |
| ch | minute: maybe note that the OTG pin should be tied high in this config | 15:21 |
| minute | ch: btw i gave you reporter access to that repo | 15:21 |
| minute | ch: so feel free to add comments or make new issues also | 15:21 |
| ch | \o/ | 15:22 |
| minute | ok, this pocket has 6.10.7. it sees our 5ghz network. i'll try to connect | 15:22 |
| ch | edited your text to add the otg pin hint | 15:23 |
| minute | thank you! | 15:23 |
| minute | going to speedtest.net | 15:23 |
| minute | i have to note i'm using the laird antenna because that was already installed in this dev pocket | 15:24 |
| minute | wow, firefox is not fast on this processor | 15:24 |
| ch | yeah :/ | 15:25 |
| minute | ok, 10 mbps here in the back | 15:25 |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 15:25 | |
| ch | ok maybe it just hates my specific wifi | 15:25 |
| minute | 10mbps is not good though, testing in the kitchen | 15:25 |
| ch | its still a lot more then it gives me here ;) | 15:26 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 15:38 | |
| ch | minute: do you have a suggestion where the new antennas should be put? | 15:39 |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 15:42 | |
| * mjw -> Guest7437 | 15:44 | |
| - Guest7437 (QUIT: Killed (erbium.libera.chat (Nickname regained by services))) (~mjw@gnu.wildebeest.org) | 15:44 | |
| * Guest943 -> mjw | 15:44 | |
| minute | ch: yes, one second | 15:50 |
| minute | ch: before it completely stalled i got 65mbit down in the kitchen, but i was also using firefox, which is like 2x slower than chromium on the imx8mp | 15:51 |
| ch | well that seems good enough | 15:52 |
| minute | but... it totally stalled after that test | 15:52 |
| minute | no more pings etc | 15:52 |
| minute | also, i forgot how damn hot imx8mp gets | 15:53 |
| minute | so it went to thermal shutdown (i don't have a backplate on) | 15:53 |
| minute | this doesn't even happen with rk3588... | 15:53 |
| minute | it'll happily work without any heatsink (unless you do like a big compile on it) | 15:53 |
| ch | yeah its not very comfy to use on a couch when charging and so on | 15:54 |
| ch | everything gets very warm very soon | 15:54 |
| minute | might be the overclocking though | 15:55 |
| minute | maybe we should roll that back | 15:55 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 15:56 | |
| minute | debian bug: cpupower-gui doesn't work... dependency missing | 15:56 |
| minute | (Namespace Handy not available) | 15:56 |
| minute | josch: debian's cpupower-gui has a missing dependency on gir1.2-handy-1 | 15:59 |
| - staticbunny (QUIT: Ping timeout: 245 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 16:01 | |
| josch | oh no... seems that cpupower-gui is badly maintained (it has no real maintainer) | 16:02 |
| minute | ;/ | 16:02 |
| josch | the good news is, that due to that i can upload a fix myself without waiting for others :) | 16:02 |
| minute | ha nice | 16:03 |
| ch | https://imgur.com/2u0Stpd 1200mhz, no heatsink | 16:03 |
| minute | ch: hottie hot | 16:03 |
| minute | the only bearable browser IMHO on imx8mp is chromium with graphics acceleration turned ~off~ | 16:04 |
| minute | > HARDWARE PROTECTION shutdown (Temperature too high) | 16:05 |
| minute | sigh | 16:05 |
| minute | maybe it's gpu freq, not cpu freq | 16:05 |
| ch | ah right. i noticed it draws a lot more power when wayland is running vs console | 16:07 |
| minute | yeah | 16:07 |
| minute | there's also ./devices/platform/soc@0/32700000.interconnect/devfreq/32700000.interconnect/governor | 16:12 |
| minute | sorry, /sys/ | 16:12 |
| minute | lol it has only 2 available freqs | 16:13 |
| minute | 200mhz and 1 ghz | 16:13 |
| minute | can be forced to 200mhz by echo 200000000 > max_freq | 16:13 |
| minute | that seems to help with temperature | 16:14 |
| minute | hmm, the default governor for that is userspace, but there's also simple_ondemand available | 16:15 |
| minute | now the freq hovers around 76 instead of going beyond 90 | 16:15 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 16:17 | |
| josch | minute: thank you for your cpupower-gui bug report -- i just uploaded a fixed version to unstable | 16:19 |
| josch | a dependency on gir1.2-glib-2.0 was missing as well :) | 16:20 |
| minute | josch: awesome, thank you! | 16:20 |
| - staticbunny (QUIT: Ping timeout: 248 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 16:22 | |
| minute | josch: we're not shipping wireless-regdb. that seems like a mistake | 16:32 |
| ch | simple_ondemand doesnt seem to switch to 200mhz for me | 16:34 |
| minute | hmm ok | 16:34 |
| minute | weird | 16:34 |
| ch | minute: wireless-regdb is in https://source.mnt.re/reform/reform-system-image/-/merge_requests/108 but josch asked to get the runner assigned to my fork, so the pipeline can run | 16:35 |
| minute | ch: oh ok | 16:35 |
| minute | ch: how about i just merge that | 16:35 |
| ch | fair :) | 16:36 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 16:36 | |
| minute | ch: thanks for that, btw | 16:37 |
| minute | also for the bluetooth fw, wowie | 16:37 |
| ch | sure, thanks for merging | 16:39 |
| ch | https://source.mnt.re/reform/pocket-reform/-/merge_requests/9 if you like another MR to maybe merge ;-) | 16:39 |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 16:41 | |
| minute | ch: you're right, it stays at 1ghz | 16:44 |
| minute | ch: ha, sorry, i coincidentally did part of that change, can you rebase !9? | 16:45 |
| ch | minute: rebased! | 16:46 |
| minute | thanks | 16:48 |
| minute | ok wifi just disconnected after 13 mins | 16:49 |
| minute | is not reconnecting by itself | 16:49 |
| minute | but can reconnect manuallu | 16:49 |
| minute | with normal heatsink on, getting 172 mpbs down | 16:52 |
| ch | josch: could have closed #1078019 :) | 16:52 |
| ch | or maybe you did, dunno | 16:52 |
| - q66 (PART: WeeChat 4.3.5) (~q66@q66.moe) | 16:53 | |
| minute | ok, wifi now stalling, but i connected to 2.4ghz as a test | 16:57 |
| minute | 5ghz is ok again | 16:58 |
| + minutepocket (~minute@p509936c9.dip0.t-ipconnect.de) | 16:59 | |
| minutepocket | hmm | 17:03 |
| minutepocket | indeed after around 30 minutes uptime, the wifi gets flaky | 17:04 |
| minute | so some pings go through normally, then few seconds pause, then again normally etc | 17:06 |
| minute | will now apply thermal pad to the wifi while the system is on | 17:06 |
| ch | ah yeah i think i saw that too | 17:06 |
| ch | the ram could also use one i guess :) | 17:06 |
| + ericsfraga (~user@2.28.19.2) | 17:07 | |
| minute | aha | 17:07 |
| ch | https://community.nxp.com/t5/i-MX-Processors/Reg-Wifi-Interface-Error/m-p/249389 this seems relevant for ath10k | 17:08 |
| minute | thermal pad applied to wifi, reconnected to 5ghz wifi and now pings are normal here | 17:08 |
| ch | at least the first error messages that i get with ath10k look very similar | 17:08 |
| ch | dunno why it would work with qcacld however | 17:08 |
| minute | hm, now i get 15mps here in the back room | 17:08 |
| jfred | if it does turn out to be a temp thing I wonder if you could make a back panel that's itself a big heatsink haha. suppose that would interfere with the antenna being on the other side | 17:09 |
| minute | ch: have you applied thermal pad to your wifi chip? | 17:09 |
| minute | jfred: the back panel _is_ a heatsink though | 17:09 |
| ch | no, forgot to order one | 17:09 |
| jfred | minute: sure, I mean with fins for greater surface area though | 17:09 |
| minute | jfred: one could make a mix of solid aluminum with some acrylic insets for the antennas maybe | 17:10 |
| minutepocket | still getting 170mbit now in the front room | 17:11 |
| minutepocket | uptime 37 min | 17:11 |
| - minutepocket (QUIT: Remote host closed the connection) (~minute@p509936c9.dip0.t-ipconnect.de) | 17:11 | |
| + minutepocket (~minute@p509936c9.dip0.t-ipconnect.de) | 17:11 | |
| minutepocket | link quality 90% | 17:11 |
| - minutepocket (QUIT: Remote host closed the connection) (~minute@p509936c9.dip0.t-ipconnect.de) | 17:13 | |
| minute | haha ok that was battery shutdown, unfortunately | 17:13 |
| jfred | haha oops, was wondering about that | 17:14 |
| ch | haha | 17:15 |
| minute | ok, but i will now move back to the stock molex antenna to compare | 17:15 |
| minute | in any case currently i don't believe there is a driver issue | 17:16 |
| ch | 25eur for the wuerth pad, wow | 17:18 |
| minute | huh | 17:23 |
| minute | probably a huge one? | 17:23 |
| minute | i mean we divide one into 25 pieces so it might make sense | 17:23 |
| ch | yeah your pn is 100x100 | 17:23 |
| minute | ok the molex antenna still gives decent mbits here in the back | 17:23 |
| minute | 15mbit down even | 17:23 |
| - mtm (QUIT: Read error: Connection reset by peer) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 17:23 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 17:24 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 17:28 | |
| + minutepocket (~minute@p509936c9.dip0.t-ipconnect.de) | 17:29 | |
| minutepocket | hello again | 17:29 |
| minutepocket | i am sitting outside the office, connected to 5ghz with the stock molex antenna | 17:30 |
| minutepocket | getting 54mbit down in speedteset | 17:30 |
| minutepocket | so i guess it's really the thermals | 17:30 |
| minutepocket | will let it charge and run for 20 more mins and test again | 17:30 |
| minute | someone on fedi says the flatpak of ungoogled-chromium also works well | 17:34 |
| ch | minute: do you get emails for things like https://source.mnt.re/reform/pocket-reform-handbook/-/merge_requests/3 or should i ping you? | 17:38 |
| minutepocket | ch: ping is always best. my email is most of the time a total swamp | 17:39 |
| ch | dont want to annoy you though | 17:39 |
| minutepocket | ch: no it's good | 17:39 |
| minutepocket | not annoying | 17:39 |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 17:41 | |
| - ericsfraga (QUIT: Quit: ERC 5.6.1-git (IRC client for GNU Emacs 31.0.50)) (~user@2.28.19.2) | 17:43 | |
| ch | cool | 17:45 |
| ch | wondering what else i should get from mouser or rs | 17:45 |
| ch | dont want to order sth again now and tomorrow discover i'll want yet another thing | 17:46 |
| minutepocket | always a conundrum | 17:47 |
| minutepocket | ch: beefy PD charger? :D | 17:47 |
| ch | heh. already got a few. but "interesting" PD chargers would be fun for testing | 17:48 |
| ch | maybe sth to get serial out | 17:49 |
| mhoye | Is there anything magic about making xorg stuff work on the firstgen Reform these days, or is that legacy abandonware now? | 17:53 |
| minutepocket | mhoye: xorg is def. legacy | 17:56 |
| minute | mhoye: esp on imx8mq i have no idea because probably noone tried that in a long time... you could use "cage" as a very thin wrapper though | 17:57 |
| ch | minute: about the antenna placement... do you have some hints? | 17:59 |
| minute | ch: for a single one, i put it in the middle of the headphone pcb where the original one goes | 18:01 |
| minute | ch: if you have dual antennas for the mediatek card, i put one there and the other facing it a bit offset on the inside of the port cover | 18:01 |
| ch | two antennas yeah. ok that feels a bit cramped :) | 18:06 |
| ch | (crammed?) | 18:07 |
| josch | minute: I was unable to merge 108 because it seems I lack permission to merge MRs with a failed pipeline. Thanks for merging. :) | 18:07 |
| josch | ch: #1078019 is closed manually now, thank you :) | 18:08 |
| josch | mhoye: there are some xorg configs in the reform-tools package and in the forum in case you want to give it a try | 18:12 |
| + andreas-e (~Andreas@2001:861:c4:f2f0::c64) | 18:20 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 18:32 | |
| minute | damn, i also have to test this... https://source.mnt.re/reform/pocket-reform/-/merge_requests/6/diffs | 18:35 |
| minute | but running out of time for today | 18:35 |
| ch | ah was wondering about sdk2.0 | 18:35 |
| ch | sysctl seems to just build | 18:35 |
| ch | used that by accident for a while :) | 18:36 |
| ch | build and work* | 18:36 |
| mhoye | minute: cage works great for wayland, I didn't know if that gave me x11 for free. I'll take a look. | 18:36 |
| minutepocket | mhoye: yeah it allows you to run xwayland fullscreen, that's very similar to running xorg | 18:37 |
| minutepocket | i mean, xwayland is xorg, just a different way of running it | 18:38 |
| josch | minute: once you got to look into pocket-reform!6 let me know because i also have to update the Debian packaging of pico-sdk to 2.0 and i'm currently holding off to do so to not break the source.mnt.re CI | 18:39 |
| - minutepocket (QUIT: Quit: Lost terminal) (~minute@p509936c9.dip0.t-ipconnect.de) | 18:42 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 18:44 | |
| ^alex | oh let me fix the merge conflict and `apt satisfy` syntax | 18:48 |
| _hramrach | If you can use the X11 input drivers with Xwayland then it's perfect :) | 18:49 |
| - staticbunny (QUIT: Ping timeout: 248 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 18:50 | |
| ^alex | iirc the CI uses a cloned copy of the picosdk, not the system packages? | 18:51 |
| josch | indeed i misremember! | 18:53 |
| josch | Contrary to my memory, this got never merged: https://source.mnt.re/reform/reform/-/merge_requests/69 | 18:53 |
| ^alex | also we noticed that when the keyboard reboots, it provokes a syntax error when it tries to talk to the keyboard | 18:55 |
| ^alex | so i think we're going to make the keyboard send a UART break when it comes up, then make the syscon reset the keyboard state machine when it gets one | 18:56 |
| ^alex | otherwise the keyboard gets squidgy on the USB port until it synchronizes with the syscon again | 18:56 |
| - theesm (QUIT: Ping timeout: 252 seconds) (~theesm@static.134.243.21.65.clients.your-server.de) | 19:05 | |
| * mjw -> Guest888 | 19:06 | |
| - Guest888 (QUIT: Killed (zirconium.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:06 | |
| * mark_ -> mjw | 19:06 | |
| + Guest888 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:07 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 19:11 | |
| _hramrach | ugh, the qcacld2 CI build uses manual download of packages, not an apr repository. Installing the kernel headers that way would be a pain, they require like at least half dozen of packages as dependencies | 19:11 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 19:18 | |
| - staticbunny (QUIT: Ping timeout: 276 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 19:23 | |
| + NanoCodeBug (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 19:25 | |
| ch | ^alex: !6 doesnt change the CI config to use sdk 2.0. is that expected? (iow: is the MR just to make it possible to build with 2.0, or should it also do that by default?) | 19:32 |
| Twodisbetter | it would be great if the keyboard backlight on the pocket reform would have an inactivity timeout and would turn the backlight off. When the mouse is moved or a key pressed the previous state of the keys would be restored. | 19:32 |
| ch | Twodisbetter: it does. but you have to update it first :) | 19:32 |
| Twodisbetter | ch: that functionality already exists in the latest firmware for the keyboard? | 19:33 |
| ch | ^alex: might also be a good idea to align install-fw-dependencies.sh and .gitlab-ci.yml | 19:33 |
| ch | Twodisbetter: yes | 19:33 |
| ch | Twodisbetter: https://source.mnt.re/reform/pocket-reform/-/commit/6f569d18f177dd6312ebfaf1c89ce3057e06262c did that | 19:35 |
| ch | imo the timeout could be shorter | 19:35 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 19:37 | |
| - staticbunny (QUIT: Ping timeout: 260 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 19:41 | |
| Twodisbetter | ch: that was from 2 months ago, how would I not have that on my pocket? I just updated yesterday, assuming i tcan be pulled using apt | 19:43 |
| ch | keyboard firmware doesnt get updated with apt | 19:44 |
| NanoCodeBug | ch: removing that diode is interesting, I was wondering why the mp2650 reported a lower input voltage than my pd meter. from a glance at the data sheet, the only code change to the firmware PR I have open would be to disable USB_SRC entirely, and replace instances of disabling/enabling USB_SRC with enabling/disabling OTG mode. | 19:45 |
| josch | Twodisbetter: the ultimate plan is to distribute the firmware for keyboard and motherboard via fwupd but somebody has to work on that first | 19:45 |
| ch | NanoCodeBug: yes thats what my fw branch is doing too, and it works very well | 19:45 |
| josch | Twodisbetter: the fwupd author would love to add support for that but it needs a volunteer on the MNT side to get the firmware ready for this | 19:45 |
| NanoCodeBug | ch: you bypassed the resistor, but didn't tie a pullup to the otg pin? and its behaving fine? | 19:45 |
| ch | NanoCodeBug: and yeah, i also found this because i was wondering about the differing voltages | 19:45 |
| ch | NanoCodeBug: yeah | 19:46 |
| NanoCodeBug | ch: nice | 19:46 |
| Twodisbetter | josch: well lets hope a volunteer comes forward because being able to push fw upodate to those components via the terminal would be awesome. | 19:46 |
| Twodisbetter | josch: it took me forever to update the keybaord firmware on my big Reform. | 19:47 |
| Twodisbetter | I held off on doing it because I just didn't have the time. | 19:48 |
| NanoCodeBug | I believe its been stated in the past that this is a explicit design decision - some folks flash custom keyboard firmware to have specific layouts or features, or system firmware, and apt would overwrite those changes. | 19:48 |
| NanoCodeBug | a tool in reform-tools that pulls the latest firmware from a stable build, so folk can update on-demand without too much friction would be nice though. | 19:50 |
| josch | Twodisbetter: with fwupd you can do it via the terminal but the idea is that it happens automatically | 19:50 |
| Twodisbetter | can't you just hold the update so that it doesnt get updated? To me it seems like it would be better to have the most current firmware pushed out and easily pullable. Still it isn't that serious to me. I mainly have the pocket plugged in. I think it is cool that it has already been added though | 19:50 |
| josch | NanoCodeBug: i'd prefer to use an established mechanism if possible instead of cooking up yet another custom script | 19:51 |
| Twodisbetter | josch: ahh ok, folks are using it over on the Framework side of the house for bios updates, so I though it was just a manually invoked porocess. | 19:51 |
| NanoCodeBug | josch: for sure | 19:51 |
| josch | Twodisbetter: tell me more | 19:52 |
| ch | i think if we get something nice like fwupd, then there's probably some way to flag custom fw | 19:52 |
| Twodisbetter | josch: ? | 19:52 |
| josch | ch: yes, i'd just use the version which fwupd needs to read anyways | 19:52 |
| ch | yeah | 19:52 |
| josch | and right now there is no way for userspace to even read a version | 19:52 |
| ch | yeah we also need that | 19:53 |
| josch | you can look at it on the oled but that's it | 19:53 |
| ch | and a lot many other things, hehe | 19:53 |
| ch | if theres some info on what fwupd needs/supports that would also be useful | 19:53 |
| ch | so we dont invent new stuff on the fw side | 19:53 |
| josch | Twodisbetter: can you expand on the last thing you said? i thought fwupd was automatic and not manual | 19:54 |
| ch | like rn its super convenient to update the sysctl fw with picotool, because that just works | 19:54 |
| Twodisbetter | I think they are using it with fwupdmgr | 19:54 |
| Twodisbetter | and now that I think about it, i think it might have been to just update the fingerprint reader firmware. | 19:54 |
| josch | ch: on fedi the fwupd maintainer said that there were super excited to have the reform supported, so maybe just open an issue and ask :) | 19:54 |
| ch | bios is probably easy, its not the first platform with a bios | 19:54 |
| ch | josch: link? | 19:55 |
| ch | (sooo many things to do for sysctl fw) | 19:56 |
| Twodisbetter | josch: totally agree. Let's get that ball rolling. | 19:56 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 19:56 | |
| ch | NanoCodeBug: btw, in my branch i started describing the i2c regs as structs, because thats quite nice to use from the fw. have only done it for the mps charger ic so far. | 19:56 |
| Twodisbetter | josch: also turns out I was wrong. It is automatic, if they have fwupdmgr installed. | 19:57 |
| Twodisbetter | pretty slick. | 19:57 |
| ch | NanoCodeBug: somehow we should get to a point where we pull out the refactors and get some of them merged first | 19:57 |
| ch | NanoCodeBug: maybe take a look at my branch if you like the registers as structs thingy | 19:57 |
| ch | .oO( i keep typing 'brunch' today ) | 19:57 |
| NanoCodeBug | ch: ah good, that makes it a lot easier to keep track of state. if its ok with you i'll try to integrate your interrupt and struct changes into the current PR. | 19:58 |
| ch | NanoCodeBug: ok. i think there's merit in pulling the refactors out first into clean commits and getting them merged, without doing functional changes | 19:58 |
| ch | NanoCodeBug: my branch is mostly in tryout mode, because i dont want to do refactors that wont get merged, and for the actual approaches (state machines for pd etc) i'm still experimenting to see what the fw really needs to do | 20:00 |
| ch | i feel like the fw needs to do a lot more than it does today, and also more than what i have in my branch (although that works!) | 20:00 |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 20:00 | |
| ch | OTG mode probably doesnt work when the diode is still in place, not sure what to do about that (maybe keep if #ifdefed out for now) | 20:01 |
| josch | ch: link: https://mastodon.social/@hughsie/113028622558551936 | 20:02 |
| minute | josch Twodisbetter: i've sent request for usb ids a few days ago to openmoko, waiting for reply. then we can integrate with fwupd | 20:03 |
| minute | if they won't do it we can ask openpids | 20:03 |
| minute | otherwise we'll squat some ids | 20:03 |
| NanoCodeBug | ch: I normally agree with that approach, but I'm not sure I'll have time to revert the PD logic so that the refactor is separate from the automatic pd toggle logic (and I haven't had time to test your branches interrupt logic on my chargers). I would like for folks to have a (sub-optimal without interrupts) firmware that fixes the issue sooner than | 20:03 |
| NanoCodeBug | latter. | 20:03 |
| minute | NanoCodeBug: i can test this weekend or monday btw | 20:03 |
| ch | ah bcdREV. good idea | 20:04 |
| minute | wanted to look into the 8mp wifi woes first (did that today) | 20:04 |
| ch | NanoCodeBug: right. optimizations from my branch can also just come later | 20:05 |
| josch | minute: i had read your message re openmoke a few days ago and i was surprised that something of them is still alive! :) | 20:05 |
| NanoCodeBug | ch: you are of course welcome to build off of the branch i have and supplant the current PR entirely | 20:06 |
| ch | NanoCodeBug: so maybe best if you get your stuff merged if it works now :) | 20:06 |
| ch | i want to get a lot of the fw "really right" but it will take a while | 20:07 |
| ch | minute: getting a new vid would be cooler though :) | 20:08 |
| NanoCodeBug | ch: having a detailed pd debugger is really nice, which model is it? | 20:09 |
| ch | NanoCodeBug: cypress (now infineon) cy4500 ez-pd | 20:09 |
| NanoCodeBug | ooo, digikey has them below retail price, tempting. | 20:11 |
| ch | do note that you might need an x86 windows to uodate it | 20:12 |
| ch | bbl | 20:12 |
| NanoCodeBug | i sadly have to use windows normally for weekday work -_- | 20:13 |
| NanoCodeBug | or do you mean the update drivers only work with a 32bit copy of windows? | 20:14 |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 20:31 | |
| _hramrach | What exactly does the standby power switch power off? | 20:43 |
| _hramrach | Would it make sense to use it tpo power off the machine when eg. adding/removing the disk? | 20:44 |
| ch | NanoCodeBug: 64bit windows works. just not arm ;) | 20:52 |
| - staticbunny (QUIT: Ping timeout: 246 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 21:11 | |
| ^alex | ch, good points, the both of 'em, we'll address that in a moment :) | 21:13 |
| - NanoCodeBug (QUIT: Ping timeout: 256 seconds) (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 21:26 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 21:42 | |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 21:49 | |
| + staticbunny (~staticbun@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 21:52 | |
| + NanoCodeBug (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 22:03 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-4135524e.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 22:36 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 22:37 | |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 22:43 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 22:43 | |
| minute | ch: yeah a VID would be cool but costs $6000 | 23:31 |
| minute | unfortunately we're always quite cash strapped | 23:31 |
| minute | _hramrach: yeah turning it off makes exchanging components safer. but also it conserves power that is otherwise consumed by the 2x rp2040s | 23:32 |
| + aloo_shu (~aloo_shu@90.166.193.180) | 23:41 | |
| ch | minute: yeah ok, dont need to be 6k cool :) | 23:45 |
| ch | the charge indicator in waybar is massively confused | 23:52 |
| ch | how does it know if the pocket is supposedly plugged in? | 23:52 |
| minute | ch: reform2-lpc driver | 23:57 |
| minute | ch: originally written by NanoCodeBug | 23:59 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
