| ch | do you have the oled on, showing battery status or so? does it work again after a few seconds? does it look like the keyboard resets? | 00:36 |
|---|---|---|
| ch | erle: ^ | 00:36 |
| erle | ch OLED is on. i can interact with that part of the device just fine. idk what a keyboard reset looks like. | 00:38 |
| josch | ch: i thought the connection between oled on and system hanges were only on new hardware like keyboard 4.0 -- erle has one of the very first reform classic imx8mq with keyboard 2 | 00:38 |
| erle | both turning it off and selecting reset work there | 00:39 |
| ch | ah i didnt know its a cöassic | 00:39 |
| ch | classic* | 00:40 |
| erle | i also once had the system start without the mouse being visible, but idk if that was from standby | 00:40 |
| ch | so nvm what i said | 00:40 |
| erle | well, still, what can i do? | 00:40 |
| erle | there isn't even a sysrq key | 00:40 |
| josch | erle: theory: what dtb are you on? | 00:41 |
| erle | how do i figure that out and what's your theory? | 00:41 |
| josch | cat /proc/device-tree/model | 00:42 |
| erle | one moment | 00:42 |
| josch | erle: with imx8mq i had problems with hard-locks and they were undeterministic -- switching to HDMI dtb made these go away | 00:42 |
| josch | but that was also a long while ago, things might've changed a lot | 00:42 |
| erle | josch it says “MNT Reform 2” | 00:43 |
| josch | okay, that's single display | 00:43 |
| erle | oh, the system also once locked up when i started flare-game | 00:43 |
| josch | erle: can you, just to try it out, run the hdmi-dtb for a while and see if it makes a difference? | 00:44 |
| erle | but then flare-game on the next startup was like “oopsie seems i crashed, i chose safe graphics settings” and no more crashies | 00:44 |
| erle | josch if you tell me just *how* to do that i could probably try it out. it would motivate me if you assure me it can easily be reverted and if it presents no risk to the hardware. | 00:45 |
| josch | there is no risk to the hardware | 00:45 |
| josch | is your /boot on emmc or on sd-card? | 00:46 |
| erle | i have no idea. one moment. | 00:46 |
| josch | findmnt --noheadings --evaluate --mountpoint /boot --output SOURCE | 00:48 |
| erle | i did “mount |grep /boot” | 00:48 |
| josch | sure | 00:48 |
| erle | i have no idea why your command is so much longer, does it do something more useful? | 00:48 |
| erle | in any case, /boot is on /dev/mmcblk1p1 | 00:49 |
| josch | your command shows all lines that contain the string "/boot" -- mine asks about the source of the mount point at a specific path | 00:49 |
| josch | okay, that's sd-card | 00:49 |
| josch | then run this: | 00:49 |
| josch | reform-display-config dual | 00:49 |
| josch | then reboot | 00:49 |
| josch | when you want to go back, run: | 00:49 |
| josch | reform-display-config single | 00:49 |
| erle | i do not have a second display though | 00:49 |
| erle | so what does this command change in material terms then if i have not? | 00:50 |
| josch | it will allow you to attach a hdmi monitor but you don't have to | 00:50 |
| erle | then why isn't it the default? | 00:51 |
| erle | is there some downside? | 00:51 |
| erle | josch so i should run that command and then report back if i get any hangs, correct? | 00:52 |
| josch | erle: The disadvantage of having HDMI enabled is that this results in the internal display being run by a less performant display engine (LCDIF instead of DCSS) and as a result, the panel is run at a lower frequency. | 00:56 |
| josch | erle: yes, try out running with the HDMI dtb for a while and report back if that changes something for you | 00:56 |
| erle | josch the result of that command is: the computer still boots and the GUI works, but when i suspend and resume, i see a cursor blinking and then … nothing. | 00:56 |
| josch | erle: switch back to single if you encounter problems | 00:56 |
| erle | so i can not effectively resume at all | 00:57 |
| josch | ooooh | 00:57 |
| josch | so it breaks suspend | 00:57 |
| josch | that's unfortunate | 00:57 |
| erle | is that news to you? | 00:57 |
| josch | yes | 00:57 |
| erle | well here we have the reason why it is not enabled by default i guess | 00:57 |
| josch | i have battled suspend on imx8mq for many, many hours but only with single display | 00:57 |
| josch | no, suspend never had something to do with it | 00:57 |
| josch | the reason is the display engine (see above) | 00:57 |
| erle | okay, here we have *another* reason ;) | 00:58 |
| josch | indeed :) | 00:58 |
| erle | so to go back to the previous state, i do what? | 00:58 |
| josch | reform-display-config single | 00:58 |
| erle | oki-doki | 00:59 |
| erle | we are in luck | 01:02 |
| erle | i suspended, it came up again | 01:02 |
| erle | and now i have the hang | 01:02 |
| erle | so what do you want to know about it? | 01:02 |
| erle | and how can i exfiltrate useful debug data? | 01:02 |
| josch | if we are talking about the same hang, then you will not even get any hint about what is going on in dmesg when connected via serial | 01:02 |
| josch | you could try if it's the same for you | 01:03 |
| erle | bold of you to assume i had the foresight to attach a serial connection anywhere | 01:03 |
| erle | i have a usbtty adapter in a box, but i doubt that will help here | 01:03 |
| josch | no foresight needed -- can alwyas be performed as an afterthought :) | 01:03 |
| josch | you need usb uart | 01:03 |
| erle | i think i will do it another day | 01:04 |
| erle | it's like 1 AM in the morning | 01:04 |
| erle | see, i even doubled that because i am of sleep | 01:04 |
| erle | thank you for your help so far | 01:06 |
| josch | yw | 01:07 |
| - mjw (QUIT: Ping timeout: 260 seconds) (~mjw@gnu.wildebeest.org) | 01:18 | |
| - erle (QUIT: Ping timeout: 250 seconds) (~erle@user/erle) | 01:54 | |
| + thejevans (~m-7r3qil@75-166-160-198.hlrn.qwest.net) | 01:55 | |
| - synnfynn (QUIT: Quit: Until next time..) (~synnfynn@user/synnfynn) | 02:50 | |
| - paperManu (QUIT: Ping timeout: 260 seconds) (~paperManu@198.16.179.107) | 03:24 | |
| - jahkosha (QUIT: Ping timeout: 250 seconds) (~quassel@user/jahkosha) | 03:51 | |
| - Ar|stote|is (QUIT: Ping timeout: 260 seconds) (~linx@149.210.0.217) | 03:55 | |
| - op_4 (QUIT: Remote host closed the connection) (~tslil@user/op-4/x-9116473) | 04:05 | |
| + op_4 (~tslil@user/op-4/x-9116473) | 04:05 | |
| + emdevt (~emdevt@2001:7e8:faf6:f101:70f9:10a4:805e:9fe1) | 04:08 | |
| - skipwich (QUIT: Remote host closed the connection) (~skipwich@user/skipwich) | 05:01 | |
| + skipwich (~skipwich@user/skipwich) | 05:02 | |
| - S0rin (QUIT: Ping timeout: 258 seconds) (~S0rin@user/s0rin) | 05:33 | |
| + S0rin (~S0rin@user/s0rin) | 05:34 | |
| + yqshao5 (~yqshao@206.168.214.200) | 05:34 | |
| - yqshao (QUIT: Read error: Connection reset by peer) (~yqshao@206.168.214.200) | 05:36 | |
| * yqshao5 -> yqshao | 05:36 | |
| + potash (~potash@user/foghorn) | 05:58 | |
| - klardotsh (QUIT: Ping timeout: 256 seconds) (~klardotsh@2601:600:c900:9443::72f) | 07:24 | |
| - potash (QUIT: Read error: Connection reset by peer) (~potash@user/foghorn) | 07:36 | |
| + potash (~potash@user/foghorn) | 07:36 | |
| + chomwitt (~chomwitt@2a02:85f:9a74:ce00:1ac0:4dff:fedb:a3f1) | 08:03 | |
| - Asmadeus (QUIT: Server closed connection) (~asmadeus@user/meow/Asmadeus) | 08:27 | |
| + Asmadeus (~asmadeus@user/meow/Asmadeus) | 08:27 | |
| - chomwitt (QUIT: Ping timeout: 256 seconds) (~chomwitt@2a02:85f:9a74:ce00:1ac0:4dff:fedb:a3f1) | 08:53 | |
| + gustav2 (~gustav@c-92-32-82-4.bbcust.telenor.se) | 10:02 | |
| josch | on rk3588 (but this does not happen on a311d), when i start qemu with -device virtio-gpu-gl-pci -display gtk,gl=on then I get the console output during boot, but once cage starts, the display turns black and I get: | 10:36 |
| josch | (qemu:4927): Gdk-WARNING **: 10:29:06.280: gdk_gl_context_set_required_version - GL context versions less than 3.2 are not supported. | 10:36 |
| josch | vrend_compile_shader: context error reported 3 "cage" Illegal shader 0 | 10:36 |
| josch | Shader failed to compile | 10:36 |
| josch | 0:1(10): error: GLSL 1.50 is not supported. Supported versions are: 1.10, 1.20, 1.30, 1.40, 1.00 ES, and 3.00 ES | 10:36 |
| josch | full log with the qemu command i used here: https://paste.debian.net/1397499 | 11:00 |
| - V (QUIT: Remote host closed the connection) (~v@ircpuzzles/2022/april/winner/V) | 11:39 | |
| - yqshao (QUIT: Changing host) (~yqshao@206.168.214.200) | 11:46 | |
| + yqshao (~yqshao@user/yqshao) | 11:46 | |
| + manis (01a66df340@185.72.67.185) | 11:55 | |
| - BAndiT1983 (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@46.101.193.235) | 12:00 | |
| - anuejn (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@46.101.193.235) | 12:00 | |
| - se6astian (QUIT: Quit: se6astian) (~quassel@46.101.193.235) | 12:00 | |
| + chomwitt (~chomwitt@2a02:85f:9a74:ce00:1ac0:4dff:fedb:a3f1) | 12:01 | |
| + mjw (~mjw@gnu.wildebeest.org) | 12:11 | |
| - chomwitt (QUIT: Ping timeout: 248 seconds) (~chomwitt@2a02:85f:9a74:ce00:1ac0:4dff:fedb:a3f1) | 13:32 | |
| + paperManu (~paperManu@198.16.179.107) | 13:39 | |
| - shtrophic (QUIT: Ping timeout: 245 seconds) (~m-hrdsqi@user/shtrophic) | 13:45 | |
| + V (~v@ircpuzzles/2022/april/winner/V) | 14:43 | |
| + reformer (~reformer@softboy.mntmn.com) | 14:51 | |
| - emdevt (QUIT: Ping timeout: 252 seconds) (~emdevt@2001:7e8:faf6:f101:70f9:10a4:805e:9fe1) | 15:10 | |
| + erle (~erle@user/erle) | 15:23 | |
| + TheProtoHobo (~dfbarth@mobile-166-175-60-240.mycingular.net) | 15:30 | |
| + se6astian (~quassel@46.101.193.235) | 15:45 | |
| + BAndiT1983 (~quassel@46.101.193.235) | 15:45 | |
| + anuejn (~quassel@46.101.193.235) | 15:45 | |
| - thejevans (QUIT: Remote host closed the connection) (~m-7r3qil@75-166-160-198.hlrn.qwest.net) | 16:16 | |
| + thejevans (~m-7r3qil@75-166-160-198.hlrn.qwest.net) | 16:17 | |
| - brennen (QUIT: Server closed connection) (~brennen@user/brennen) | 16:23 | |
| + brennen (~brennen@user/brennen) | 16:24 | |
| elb | I don't know if anybody's tried it before, but hte pocket reform charges at 15 W off an iFixit FixHub | 16:29 |
| - jogu (QUIT: Quit: Konversation terminated!) (~jogu@user/jogu) | 16:31 | |
| + jogu (~jogu@user/jogu) | 16:31 | |
| - thejevans (QUIT: Ping timeout: 245 seconds) (~m-7r3qil@75-166-160-198.hlrn.qwest.net) | 16:57 | |
| + thejevans (~m-7r3qil@75-166-160-198.hlrn.qwest.net) | 16:58 | |
| + foobar1234979 (~quassel@2a02:1210:766d:4400:6f2d:e842:e16f:ba24) | 17:07 | |
| - foobar1234979 (QUIT: Client Quit) (~quassel@2a02:1210:766d:4400:6f2d:e842:e16f:ba24) | 17:07 | |
| + jahkosha (~quassel@user/jahkosha) | 17:08 | |
| - jahkosha (QUIT: Client Quit) (~quassel@user/jahkosha) | 17:09 | |
| - thejevans (QUIT: Ping timeout: 260 seconds) (~m-7r3qil@75-166-160-198.hlrn.qwest.net) | 17:09 | |
| + jahkosha (~jahkosha@user/jahkosha) | 17:09 | |
| - jahkosha (QUIT: Client Quit) (~jahkosha@user/jahkosha) | 17:09 | |
| + jahkosha (~jahkosha@user/jahkosha) | 17:11 | |
| + thejevans (~m-7r3qil@97-118-249-201.hlrn.qwest.net) | 17:11 | |
| + synnfynn (~synnfynn@user/synnfynn) | 17:46 | |
| + AnimaInvicta (~AnimaInvi@88-120-179-216.subs.proxad.net) | 18:16 | |
| + chomwitt (~chomwitt@2a02:85f:9a74:ce00:1ac0:4dff:fedb:a3f1) | 19:05 | |
| + jnerula__ (~jnerula@li1009-93.members.linode.com) | 19:13 | |
| + timide_ (~timide@user/timide) | 19:14 | |
| + BAndiT1983_ (~quassel@46.101.193.235) | 19:14 | |
| + synnfynn_ (~synnfynn@user/synnfynn) | 19:15 | |
| + anuejn_ (~quassel@46.101.193.235) | 19:16 | |
| - chomwitt (QUIT: Ping timeout: 250 seconds) (~chomwitt@2a02:85f:9a74:ce00:1ac0:4dff:fedb:a3f1) | 19:16 | |
| - synnfynn (QUIT: *.net *.split) (~synnfynn@user/synnfynn) | 19:18 | |
| - BAndiT1983 (QUIT: *.net *.split) (~quassel@46.101.193.235) | 19:18 | |
| - anuejn (QUIT: *.net *.split) (~quassel@46.101.193.235) | 19:18 | |
| - yqshao (QUIT: *.net *.split) (~yqshao@user/yqshao) | 19:18 | |
| - S0rin (QUIT: *.net *.split) (~S0rin@user/s0rin) | 19:18 | |
| - timide (QUIT: *.net *.split) (~timide@user/timide) | 19:18 | |
| - jnerula_ (QUIT: *.net *.split) (~jnerula@li1009-93.members.linode.com) | 19:18 | |
| - colinsane (QUIT: *.net *.split) (~colinunin@97-113-128-166.tukw.qwest.net) | 19:18 | |
| + yqshao (~yqshao@user/yqshao) | 19:24 | |
| + S0rin (~S0rin@user/s0rin) | 19:24 | |
| + colinsane (~colinunin@97-113-128-166.tukw.qwest.net) | 19:27 | |
| + klardotsh (~klardotsh@2601:600:c900:9443::72f) | 19:36 | |
| * plomlompom -> plomlomp0m | 19:47 | |
| * plomlomp0m -> plomlompom | 19:47 | |
| * synnfynn_ -> synnfynn | 20:14 | |
| + thelounge0670 (~thelounge@gidzit.org) | 21:06 | |
| - thelounge0670 (QUIT: Quit: The Lounge - https://thelounge.chat) (~thelounge@gidzit.org) | 21:12 | |
| - TheProtoHobo (QUIT: Read error: Connection reset by peer) (~dfbarth@mobile-166-175-60-240.mycingular.net) | 21:19 | |
| + TheProtoHobo (~dfbarth@mobile-166-175-60-240.mycingular.net) | 21:24 | |
| erle | sometimes the reform hangs right after systemd says that udev is loading. any idea how to debug that? | 21:26 |
| + chomwitt (~chomwitt@2a02:85f:9a74:ce00:1ac0:4dff:fedb:a3f1) | 21:27 | |
| - TheProtoHobo (QUIT: Ping timeout: 244 seconds) (~dfbarth@mobile-166-175-60-240.mycingular.net) | 21:40 | |
| mhoye | In terms of stretching battery life as far as it will go, are we generally confident that "ondemand" is the best choice of governor or is there room for discussion there? | 21:51 |
| + aloo_shu_ (~aloo_shu@90.166.193.247) | 21:59 | |
| - mlarkin (QUIT: Ping timeout: 256 seconds) (~mlarkin@syn-076-081-194-027.biz.spectrum.com) | 22:00 | |
| - lexik (QUIT: Ping timeout: 256 seconds) (~lexik@93.185.97.218) | 22:00 | |
| - shdw (QUIT: Ping timeout: 256 seconds) (~shdw@static.218.156.216.95.clients.your-server.de) | 22:00 | |
| - f_ (QUIT: Ping timeout: 256 seconds) (16abab341f@postmarketOS/funderscore) | 22:00 | |
| + f_ (16abab341f@postmarketOS/funderscore) | 22:00 | |
| - aloo_shu (QUIT: Ping timeout: 256 seconds) (~aloo_shu@90.166.193.247) | 22:01 | |
| * aloo_shu_ -> aloo_shu | 22:01 | |
| + shdw (~shdw@static.218.156.216.95.clients.your-server.de) | 22:01 | |
| + lexik (~lexik@93.185.97.218) | 22:02 | |
| - gsora (QUIT: Read error: Connection reset by peer) (~gsora@user/gsora) | 22:04 | |
| + gsora_ (~gsora@user/gsora) | 22:05 | |
| * gsora_ -> gsora | 22:05 | |
| + mlarkin (~mlarkin@syn-076-081-194-027.biz.spectrum.com) | 22:06 | |
| minute | elb: good to know @ ifixit hub! | 22:06 |
| minute | mhoye: "powersave" should be the one. and i think ondemand has been succeeded by "schedutil", but haven't really benchmarked the two | 22:08 |
| minute | my pocket has been up for 10 hours 42 minutes and dmesg has been absolutely quiet, keyboard didn't reset in this time | 22:09 |
| + TheProtoHobo (~dfbarth@mobile-166-175-60-240.mycingular.net) | 22:09 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-128-166.tukw.qwest.net) | 22:12 | |
| - gustav2 (QUIT: Quit: Quit) (~gustav@c-92-32-82-4.bbcust.telenor.se) | 22:15 | |
| - chrcav (QUIT: Ping timeout: 256 seconds) (~chrcav@user/chrcav) | 22:21 | |
| + chrcav (~chrcav@user/chrcav) | 22:22 | |
| josch | 10 hours? o0 | 22:25 |
| josch | ah on ac, right? | 22:25 |
| minute | josch: yeah ac. just wanted to say that my new kbd firmware didn't show any issues yet | 22:26 |
| josch | sweet! have something i can flash? :) | 22:28 |
| minute | absolutely | 22:30 |
| chorc | mhoye: powersave is the one, gives me about 45 extra minutes compared to schedutil; also about 5 degrees cooler, I'm on powersave even when plugged in most of the time | 22:31 |
| minute | josch: https://source.mnt.re/reform/pocket-reform/-/jobs/13221/artifacts/browse/pocket-reform-keyboard-fw/pocket-hid/build/ | 22:31 |
| josch | minute: didn't we learn yeasterday that it should rather be bool gfx_init(void)? :) | 22:33 |
| josch | this is quite a colorful bag of changes | 22:34 |
| minute | josch: no @ void | 22:36 |
| minute | josch: most of the changes are harmless, it's just cppcheck style stuff. the interesting thing is the strdup removal | 22:37 |
| josch | huh, then i misunderstood yesterday -- i thought if one does not put (void) into the function definition then you can supply any args (and i tested this with gcc successfully) | 22:38 |
| minute | josch: the person who brought this up here did not tell the whole story | 22:39 |
| minute | josch: https://www.open-std.org/jtc1/sc22/wg14/www/docs/n2841.htm | 22:40 |
| minute | josch: this old nonsense behaviour is being phased out | 22:40 |
| josch | oh wonderful | 22:40 |
| josch | i was prepared to pass on the wisdom that we need to put (void) in our headers everywhere next semester... | 22:41 |
| minute | please no :D | 22:41 |
| minute | this channel is not a reliable source of information sometimes :D | 22:42 |
| josch | i was relying on gcc behaviour | 22:42 |
| josch | and wasn't aware that it's about to get changed | 22:42 |
| minute | i think ch brought that up | 22:42 |
| minute | (that it got changed) | 22:42 |
| josch | but nice find with the strstr() missing a free() | 22:43 |
| josch | errr... strdup | 22:44 |
| josch | lets hope that this was it :) | 22:44 |
| - AnimaInvicta (PART: !!unknown attribute: msg!!) (~AnimaInvi@88-120-179-216.subs.proxad.net) | 22:50 | |
| Zaba | josch: personally I’ve been trained by years of working wirh sketchy proprietary compilers to be pessimistic about latest standard availability and to always write (void) in C code. if you only care about modern toolchains then it’s enough to ensure you’re actually compiling for C23 | 22:56 |
| - synnfynn (QUIT: Quit: until next time..) (~synnfynn@user/synnfynn) | 22:57 | |
| Zaba | and I would still consider this worth explaining to students if only because any real C code will have that in it and it’s good to know what it means | 23:00 |
| minute | https://source.mnt.re/reform/mnt-reform-setup-wizard/-/commit/853605523f1c59bf21a26ed5214418d500f39da6 | 23:19 |
| minute | josch: this fixes the "stuck at exit" issue | 23:20 |
| josch | wow, gains are being made tonight :) | 23:22 |
| josch | once, the pipeline finished i can build a new system image and see what happens | 23:23 |
| minute | yeah. i tested it now by putting a real rsync in the dummy script, but it's not ideal: one can still close the app window and rsync will continue running in the background. i've removed the window close button for now because it didn't make sense anyway in the OOBE | 23:33 |
| minute | (new push) | 23:33 |
| elb | minute: josch: so ... per C99, which is _the relevant C that everyone actually uses_, an empty formal parameter list on a function means "the formal aprameters for this function are not specified here", not "this function accepts no arguments" | 23:57 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!