| + jacobk (~quassel@64.189.201.150) | 00:53 | |
| violet | yeah thats probably what i meant | 01:17 |
|---|---|---|
| violet | i think i mashed up "3399" and "3588" in my head | 01:17 |
| violet | i always have a hard time keeping the rk numbers straight | 01:17 |
| - mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 01:39 | |
| + nocko (~nock@user/nocko) | 01:40 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 02:03 | |
| unixpoet | that's fair, yeah. I have boards with both chips so the numbers stick in my head more | 02:05 |
| - XYZ (QUIT: Ping timeout: 240 seconds) (~XYZ@37-48-33-138.nat.epc.tmcz.cz) | 02:18 | |
| abortretryfail | i'm trying to make sense of this lpc nonsense with the ???. according to the manual, the communication from lpc shows up on /dev/ttymxc2, but i never see anything out of there except "wake\0" from pressing circle-space | 02:29 |
| + XYZ (~XYZ@37-48-33-138.nat.epc.tmcz.cz) | 02:39 | |
| - q66 (QUIT: Quit: WeeChat 4.0.2) (~q66@q66.moe) | 02:44 | |
| abortretryfail | does the kernel driver interfere with that somehow? | 02:52 |
| + q66 (~q66@q66.moe) | 03:39 | |
| - nsc (QUIT: Ping timeout: 264 seconds) (~nicolas@12-48-142-46.pool.kielnet.net) | 03:48 | |
| + nsc (~nicolas@103-99-142-46.pool.kielnet.net) | 03:50 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 04:09 | |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@64.189.201.150) | 04:50 | |
| + jacobk (~quassel@utdpat241106.utdallas.edu) | 05:00 | |
| - jacobk (QUIT: Quit: No Ping reply in 180 seconds.) (~quassel@utdpat241106.utdallas.edu) | 05:54 | |
| - svp (QUIT: Server closed connection) (sid537750@id-537750.uxbridge.irccloud.com) | 05:57 | |
| + svp (sid537750@id-537750.uxbridge.irccloud.com) | 05:57 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 06:05 | |
| + jacobk (~quassel@129.110.242.173) | 06:09 | |
| - jacobk (QUIT: Quit: No Ping reply in 180 seconds.) (~quassel@129.110.242.173) | 06:13 | |
| + jacobk (~quassel@129.110.242.173) | 06:15 | |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@129.110.242.173) | 07:35 | |
| Boostisbetter | When you do circle + space that should wake the Reform from standby. Usually what I see here is waking and a percentage, which then disappears, and the screen is then blank. | 08:14 |
| violet | i have this weird thing | 08:33 |
| violet | linux-headers-6.5.0-1-reform2-arm64 : Depends: gcc-13 but it is not installable | 08:33 |
| violet | why do the headers depend on gcc-13 lol | 08:33 |
| + jacobk (~quassel@108.218.212.124) | 08:34 | |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@108.218.212.124) | 08:52 | |
| - klardotsh (QUIT: Ping timeout: 252 seconds) (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 09:05 | |
| josch | violet: with which gcc would you like to compile it instead? | 09:10 |
| + jacobk (~quassel@64.189.201.150) | 09:15 | |
| violet | ah! 12 i think | 09:16 |
| violet | i think 12 is the latest in bookworm? | 09:16 |
| violet | that or my system is configured wrong | 09:17 |
| josch | violet: oh you want to compile unstable kernel on stable? | 09:17 |
| violet | right! well, sort of. i want to install the headers for it on stable | 09:18 |
| violet | so i can build the lpc dkms module | 09:18 |
| josch | okay, so are running stable with a 6.5 kernel, yes? | 09:18 |
| violet | yeah that's right | 09:18 |
| josch | violet: you know about reform.debian.net? | 09:19 |
| violet | i dont think so | 09:19 |
| josch | aha! :D | 09:19 |
| josch | violet: it's a service i maintain which might provide what you want :) | 09:19 |
| josch | that is, an apt repository for the reform targetting Debian stable instead of unstable | 09:19 |
| violet | oh ok | 09:19 |
| violet | that sounds great | 09:20 |
| josch | what you tried to do earlier is called backporting | 09:20 |
| josch | and it has been done for 6.4 (but not 6.5) and i'll incorporate backports into reform.d.net repos soon | 09:20 |
| violet | yeah i think i just have the unstable repo still in there from when i got this thing | 09:21 |
| josch | but it's currently not super urgent because for imx8mq you don't really need 6.5 | 09:21 |
| josch | 6.5 becomes necessary for the a311d, for example | 09:21 |
| violet | o.0 what's in 6.5 for the a311d | 09:22 |
| josch | violet: the list of patches required for the banana pi is very long: https://source.mnt.re/reform/reform-debian-packages/-/tree/main/linux/patches/meson-g12b-bananapi-cm4-mnt-reform2 | 09:23 |
| josch | violet: there would've been ~10 more but they have been integrated into 6.5 | 09:23 |
| violet | wow | 09:24 |
| - XYZ (QUIT: Ping timeout: 260 seconds) (~XYZ@37-48-33-138.nat.epc.tmcz.cz) | 09:29 | |
| + XYZ (~XYZ@37-48-2-164.nat.epc.tmcz.cz) | 09:42 | |
| + ephase (~ephase@2a01:cb1a:74:673:ad65:dd17:d92d:7803) | 09:49 | |
| ephase | Hi, there someone who manage wireless networks with iwd on the MNT Reform? | 09:50 |
| ephase | I have an issue, iwd complain about lack of kernel options activated : CONFIG_CRYPTO_USER_SKCIPHER, CONFIG_CRYPTO_ECB, CONFOG_CRYPTO_CBC ... | 09:53 |
| ephase | ant it refuse to start | 09:53 |
| Boostisbetter | Speaking of the 6.5 kernel, I am having REALLY good success on resume from standby with it. I'm easily over 20 resumes at this point. | 09:54 |
| Boostisbetter | and I should point out that this is without the dropping of the sway display on resume and reinitializing it. | 09:55 |
| Boostisbetter | This is all really good signs. | 09:55 |
| Boostisbetter | josch: I think you just need a new mainboard to get standby working for you. Although it seems like you have been making the most of your reform's processor time. | 09:56 |
| + mjw (~mjw@gnu.wildebeest.org) | 10:14 | |
| - ephase (QUIT: Quit: WeeChat 4.0.2) (~ephase@2a01:cb1a:74:673:ad65:dd17:d92d:7803) | 10:43 | |
| josch | ephase: i just checked on an amd64 machine and those kernel options are not set there either, so i don't think it's reform specific. Or are you successfully running iwd on another system? | 10:49 |
| josch | Boostisbetter: i'll buy the reform motherboard 2.5 as soon as it's out and let you know if anything changed. :) | 10:50 |
| josch | ah ephase left | 10:51 |
| abortretryfail | Boostisbetter: mine doesn't actually sleep, so i can't test it. I was expecting to see the LPC's responses on /dev/ttymxc2 like the manual says | 11:43 |
| abortretryfail | out of all of the functions the keyboard sends, wake is the only one i see a response for on the serial port. | 11:44 |
| abortretryfail | Oh I take that back, apparently 6.5 fixed suspend. :D | 11:45 |
| - mjw (QUIT: Ping timeout: 248 seconds) (~mjw@gnu.wildebeest.org) | 12:14 | |
| Boostisbetter | hmm, that is interesting. 5.12 had good performance, but so far 6.5 has been flawless. I remain optimistic. | 12:37 |
| abortretryfail | spoke too soon. sound is broken, lol | 12:47 |
| * mark_ -> mjw | 12:50 | |
| minute | ok, i need to understand where the bootcmd that our imx uboot uses comes from, because if i comment out the custom bootcmd of ls1028a, there's no bootcmd left at all | 12:57 |
| josch | minute: in 2dd6c0e7e2ee54784169d68933ab5b2e44881585 I removed bootcmd from reform-boundary-uboot | 13:00 |
| minute | josch: aha? how does it find boot.scr? | 13:00 |
| josch | minute: the commit message has some hints what will happen instead | 13:00 |
| josch | it uses the distro_boot mechanism | 13:00 |
| josch | what i would write now is part of the commit message of that commit :) | 13:00 |
| minute | mhm https://source.mnt.re/reform/reform-boundary-uboot/-/commit/2dd6c0e7e2ee54784169d68933ab5b2e44881585 | 13:01 |
| josch | minute: the BOOTENV define includes the important stuff | 13:02 |
| minute | ok, run distro_bootcmd doesn't work | 13:02 |
| minute | maybe because it wants to load from mmc 1 | 13:02 |
| minute | but sd card is mmc 0 on ls1028a | 13:02 |
| josch | you can change this with DISTRO_BOOT_DEV_MMC | 13:02 |
| josch | (also part of that commit) | 13:02 |
| minute | ok, something is borked still, will need to look into this again later | 13:12 |
| jfred | There's no widely available CHERI hardware yet but it would be super cool (albeit extremely niche at this point) to have a Morello or CHERI-RISC-V Reform module | 13:59 |
| jfred | Even the possibility of that is probably years away though | 13:59 |
| - mtm (QUIT: Ping timeout: 260 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:04 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 14:40 | |
| - mlarkin (QUIT: Ping timeout: 244 seconds) (~mlarkin@047-036-115-056.res.spectrum.com) | 14:52 | |
| + mlarkin (~mlarkin@047-036-115-056.res.spectrum.com) | 14:54 | |
| + reform14743 (~cliffya@180-150-53-95.b49635.mel.static.aussiebb.net) | 14:55 | |
| - reform14743 (QUIT: Quit: reform14743) (~cliffya@180-150-53-95.b49635.mel.static.aussiebb.net) | 15:04 | |
| + cliffya (~CliffyA@180-150-53-95.b49635.mel.static.aussiebb.net) | 15:05 | |
| - cliffya (QUIT: Client Quit) (~CliffyA@180-150-53-95.b49635.mel.static.aussiebb.net) | 15:05 | |
| + cliffya (~CliffyA@180-150-53-95.b49635.mel.static.aussiebb.net) | 15:05 | |
| - mark_ (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 15:28 | |
| - Boostisbetter (QUIT: Ping timeout: 240 seconds) (4a410829d7@irc.cheogram.com) | 15:37 | |
| + Boostisbetter (4a410829d7@irc.cheogram.com) | 15:55 | |
| abortretryfail | could you make one with that FPGA board? | 16:00 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:10 | |
| - mtm (QUIT: Ping timeout: 258 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:19 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:19 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 16:56 | |
| jfred | Possibly, with of course a huge performance penalty haha | 17:02 |
| - cliffya (QUIT: Quit: WeeChat 4.0.2) (~CliffyA@180-150-53-95.b49635.mel.static.aussiebb.net) | 17:18 | |
| abortretryfail | maybe it's my blackpill prescription talking, but at this point I'm skeptical of the value of new hardware security features. | 17:45 |
| abortretryfail | It seems the threat model they're most effectively used to protect against is the owner/end-user of the machine. | 17:45 |
| eery | No more than any other hardware security feature (e.g. w^x pages, pointer tagging), in the sense it prevents exploitation past OS-enforced limitations | 17:59 |
| eery | now, to really take advantage of it, you'd need an OS and software stack built around it, and I'm not holding my breath for that one... | 17:59 |
| - jacobk (QUIT: Ping timeout: 240 seconds) (~quassel@64.189.201.150) | 18:26 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 18:34 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mjw@gnu.wildebeest.org))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:36 | |
| * mark_ -> mjw | 18:36 | |
| + mark_ (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:36 | |
| - erle (QUIT: Quit: Just say no, then Putin can not legally invade your nation without your consent.) (~erle@2a02:8109:da40:c4:3b3a:3a:1200:b9a7) | 18:50 | |
| - sevan (QUIT: Quit: rebooting, wish me luck!) (~sevan@user/venture37) | 19:01 | |
| + klardotsh (~klardotsh@c-67-170-115-80.hsd1.wa.comcast.net) | 19:05 | |
| + sevan (~sevan@user/venture37) | 19:45 | |
| + jacobk (~quassel@129.110.242.224) | 20:00 | |
| Zaba | minute: out of curiosity, how exactly is the lcd panel held in place in the lid of the reform? the panel doesn't seem to have any mounting holes or lugs, is it basically just sandwiched between the back and the front of the lid? | 20:10 |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@129.110.242.224) | 20:13 | |
| + ajr (uid609314@user/ajr) | 20:47 | |
| - jryans (QUIT: Server closed connection) (sid12941@id-12941.ilkley.irccloud.com) | 20:48 | |
| + jryans (sid12941@id-12941.ilkley.irccloud.com) | 20:48 | |
| minute | Zaba: sandwiched, and some small foam tape strips | 20:59 |
| minute | ok, back at ls1028a boot debug | 20:59 |
| minute | something is overwriting boot_scripts, with boot_scripts=ls1028ardb_boot.scr | 20:59 |
| minute | ah, ls1028ardb.h overwrites a lot of stuff | 21:01 |
| + jacobk (~quassel@129.110.242.224) | 21:06 | |
| - jacobk (QUIT: Ping timeout: 255 seconds) (~quassel@129.110.242.224) | 21:18 | |
| minute | it loads boot.scr now, and boot.scr boots the kernel, but it gets stuck, not sure where | 21:27 |
| minute | it gives root=/dev/ram0 to the kernel | 21:28 |
| minute | aha! | 21:32 |
| minute | CONFIG_BOOTARGS also contained additional junk polluting the cmdline... turned that off and now we have boot and display comes up | 21:32 |
| + jacobk (~quassel@129.110.242.224) | 21:46 | |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@129.110.242.224) | 21:55 | |
| minute | josch: this works https://source.mnt.re/reform/reform-ls1028a-uboot/-/commit/bb489fad7ed8a14e546a7e19edc8549f030b8361 | 22:02 |
| josch | uuuh, that's cool! | 22:06 |
| minute | josch: (besides getting backlight PWM to work) i need to do only one more thing for a working system image, which is to drop ls1028a-mhdpfw.bin in /lib/firmware. how should i do that? build a package for it in reform-debian-packages? | 22:14 |
| minute | in the long term we can figure out how to pack it into + load it from uboot | 22:15 |
| josch | the reform-tools package could build a firmware package as well | 22:15 |
| josch | hm... but you are already patching upstream u-boot -- would it be so much to add another patch to upstream uboot that loads the firmware? | 22:16 |
| josch | i don't know how much could that would be | 22:16 |
| josch | s/could/code/ | 22:16 |
| minute | i am extremely low on time atm (that's why i still sit here at 22h), and i have a working setup, so i would rather just finish that first (only have to drop a file in the system image...) | 22:17 |
| minute | i see some references to "hdp load" in mainline uboot... | 22:18 |
| minute | but i don't see where the command is defined | 22:18 |
| josch | if you put the file into a .deb, how would that .deb get onto the system? | 22:18 |
| josch | would it be an extra package installed by reform-system-image only for the ls1028a target? | 22:19 |
| minute | josch: with reform-system-image's build process | 22:19 |
| minute | josch: i'm also ok with just sticking it in reform-tools | 22:19 |
| josch | and if yes, then why have it in a package in the first place? | 22:19 |
| minute | but that would pollute its license | 22:19 |
| minute | so probably not good | 22:19 |
| josch | if it's in reform-tools, then all platforms would receive the firmware file | 22:19 |
| minute | josch: yeah, i agree. | 22:20 |
| minute | josch: can i use a line similar to the existing "--customize-hook="upload flash-$PLATFORM.bin /boot/flash.bin" \ in mkimage.sh? | 22:20 |
| josch | yes, that would work | 22:20 |
| josch | it would be the easiest solution i think | 22:20 |
| minute | ok! | 22:20 |
| violet | hmm switching to the reform.debian.net kernel broke my boot for some reason. probably the disk encryption stuff didnt make their way into the initramfs, not sure why yet | 22:21 |
| josch | and it would work if you accept that via that method, the firmware cannot get upgraded | 22:21 |
| minute | josch: yeah, it doesn't need upgrades | 22:21 |
| minute | i think :D | 22:21 |
| violet | need to do the ol swap over back to SD boot from eMMC whenever i feel like unscrewing this thing | 22:21 |
| josch | violet: do you have boot messages that tell you what is wrong? | 22:21 |
| josch | violet: is your u-boot new enough to have display output? | 22:21 |
| violet | oh it gets into the kernel and everything | 22:22 |
| josch | violet: did you consider install u-boot-menu so that you can pick which kernel to load on each boot? | 22:22 |
| violet | and i get the initramfs rescue image eventually | 22:22 |
| violet | my u-boot has display output but its not very useful because the display is extremely flickery/glitchy | 22:22 |
| violet | until it gets out of uboot and into the kernel | 22:22 |
| violet | not backlight flicker mind you, LCD data flickery | 22:23 |
| violet | and it obscures the text | 22:23 |
| josch | uh i've never seen such an effect | 22:23 |
| violet | mine almost always does that | 22:23 |
| violet | i can take a video later | 22:23 |
| violet | anyways its doing the same thing that happened before i installed the initramfs crypto package the first time around which is that its waiting around for a boot disk to show up forever but the disk doesnt. | 22:24 |
| josch | violet: which steps did you take to replace your kernel with the one from reform.d.n? | 22:25 |
| josch | i'd like to find out where it broke for you to prevent this happening to others | 22:26 |
| violet | i followed https://reform.debian.net/repo/ using extrepo method. and then what i did was apt install the 6.4 kernel, and then apt remove the 6.5 kernel, and then reboot | 22:27 |
| violet | and i am in fact booted into the 6.4 kernel i installed i believe because `uname -a` in the initramfs says Debian 6.4.1-~exp1+reform20230710Twhatever | 22:28 |
| josch | violet: you mean the 6.1 kernel, right? | 22:28 |
| violet | no, i mean 6.4 | 22:28 |
| josch | that's not from the reform.d.n repo | 22:28 |
| violet | where is it from then lol | 22:29 |
| josch | violet: maybe you forgot to disable the old repo? | 22:29 |
| violet | i disabled it, i know this because before rebooting i tried to use my usb headphones and the driver was missing (because i removed the running kernel and this its modules). so i tried to reinstall the 6.5 kernel. but i couldnt because i had removed the repo | 22:30 |
| violet | so i know that the old repo was in fact disabled | 22:30 |
| josch | 6.4.1-~exp1+reform20230710Twhatever is from experimental, so it definitely cannot come from reform.d.n which only contains stable kernels | 22:32 |
| josch | that's what "exp1" means in the version string | 22:32 |
| josch | violet: can you boot from sd card, mount everything and chroot into the broken system to investigate? | 22:32 |
| violet | yeah | 22:32 |
| violet | that is the plan, i just need to get at the switch | 22:32 |
| josch | violet: why do you need to get to the switch? | 22:32 |
| violet | dont i have to switch it back to SD booting? | 22:33 |
| josch | violet: the switch chooses where u-boot comes from | 22:33 |
| violet | oh. well how do i tell u-boot to boot the SD | 22:33 |
| josch | but independent on whether the imx loads u-boot from, u-boot will probe the sd-card before emmc | 22:33 |
| violet | AH | 22:33 |
| violet | ok, cool | 22:33 |
| josch | the boot order in u-boot is fixed | 22:33 |
| josch | first usb, then sd, then emmc | 22:33 |
| violet | thats very good to know. i am used to my own bad uboot installations where i dont set it up like that haha | 22:34 |
| violet | conveniently my SD card is configured to boot this system anyway so i dont even need to chroot, i can just boot | 22:35 |
| josch | oh cool | 22:37 |
| violet | i wonder is there a way i can check to tell you specifically what repo a package i have installed came from | 22:38 |
| violet | hmm yeah you are right about the repo only providing 6.1 | 22:44 |
| violet | i have no idea how i pulled off what i did | 22:47 |
| violet | but i cant reproduce it | 22:47 |
| - vagrantc (QUIT: Ping timeout: 240 seconds) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 22:49 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 22:51 | |
| violet | aaaaand boot works fine now | 22:51 |
| violet | *shrug* dunno what that was all about, but good to know i can just pop this SD whenever i break something | 22:52 |
| josch | violet: how did you fix it? | 22:52 |
| violet | uninstalled 6.4, installed 6.1 via linux-image-arm64 | 22:54 |
| violet | then rebooted | 22:54 |
| josch | oh wow | 22:55 |
| josch | remember that you can have multiple kernels installed at the same time | 22:55 |
| violet | ayyy and now my battery indicator in my status bar finally works! | 22:55 |
| josch | you do not necessarily remove the old one immediately | 22:55 |
| josch | personally, i use u-boot-menu which will boot the latest kernel | 22:55 |
| josch | but will also let me choose an older one in case the newest kernel broke something | 22:56 |
| violet | yeah | 22:56 |
| josch | vagrantc: i have a flash-kernel question. Kernel 6.5 messed up caam_jr and another option to ${bootargs} is needed to avoid it. Would it be correct to temporarily add something to /usr/share/flash-kernel/ubootenv.d/00reform2_ubootenv which append that option to ${bootargs}? | 23:03 |
| vagrantc | josch: what's caam_jr ? | 23:07 |
| josch | vagrantc: crypto stuff: https://community.mnt.re/t/changes-introduced-by-kernel-6-5-rc7-1-exp1-reform20230831t205634z1/1651 | 23:08 |
| vagrantc | ah. | 23:08 |
| vagrantc | josch: no strong opinion ... those hooks are there to fiddle with things :) | 23:09 |
| josch | vagrantc: i don't know how much shell i can do in u-boot shell -- can i do version comparison easily? | 23:10 |
| vagrantc | you can do equality ... i think that is about it | 23:22 |
| vagrantc | been a while since i tried anything complicated | 23:23 |
| - Gooberpatrol66 (QUIT: Ping timeout: 245 seconds) (~Gooberpat@user/gooberpatrol66) | 23:47 | |
| josch | minute: this should limit passing cryptomgr.notests only for the affected kernel version: https://source.mnt.re/reform/reform-tools/-/merge_requests/52 | 23:48 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!