| - ajr (QUIT: Ping timeout: 268 seconds) (~ajr@user/ajr) | 00:17 | |
| + ajr (~ajr@user/ajr) | 00:28 | |
| minute | josch: i also wasn't able to get any output from it. but putting in a sleep 10 stopped all execution for 10 seconds. | 00:33 |
|---|---|---|
| - chomwitt (QUIT: Ping timeout: 268 seconds) (~chomwitt@2a02:587:dc0c:c200:9e04:2be9:6643:df78) | 00:37 | |
| - S0rin (QUIT: Ping timeout: 265 seconds) (~S0rin@user/s0rin) | 01:04 | |
| + S0rin (~S0rin@user/s0rin) | 01:07 | |
| - Gooberpatrol66 (QUIT: Remote host closed the connection) (~Gooberpat@user/gooberpatrol66) | 01:11 | |
| - mtm (QUIT: Ping timeout: 265 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 02:04 | |
| - ajr (QUIT: Quit: WeeChat 3.6) (~ajr@user/ajr) | 02:08 | |
| - vagrantc (QUIT: *.net *.split) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 03:16 | |
| - robin (QUIT: *.net *.split) (~robin@user/terpri) | 03:16 | |
| - xktr (QUIT: *.net *.split) (~xktr@2602:fe3d:c01:10ca:1050:1ace:0:b) | 03:16 | |
| - chartreuse (QUIT: *.net *.split) (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 03:16 | |
| + xktr_ (~xktr@2602:fe3d:c01:10ca:1050:1ace:0:b) | 03:16 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 03:16 | |
| + chartreuse (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 03:16 | |
| - GNUmoon (QUIT: Quit: Leaving) (~GNUmoon@gateway/tor-sasl/gnumoon) | 04:06 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 04:07 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 04:09 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 04:53 | |
| josch | minute: ah good idea! Yes, that worked. :) | 04:57 |
| josch | minute: the answer is to add the option "debug" to the kernel boot args | 05:08 |
| josch | Then, after the system has booted, everything you printed on stdout and stderr and a complete "sh -x" output can be found in /run/initramfs/initramfs.debug | 05:08 |
| - S0rin (QUIT: Ping timeout: 265 seconds) (~S0rin@user/s0rin) | 05:18 | |
| josch | here one can see the modules being loaded first thing during boot: https://mister-muffin.de/p/4QJ0.txt | 05:19 |
| + S0rin (~S0rin@user/s0rin) | 05:20 | |
| josch | What content of https://source.mnt.re/reform/reform-tools/-/blob/main/etc/X11/xorg.conf is not relevant anymore? The /usr/local part is definitely unused, no? We are also shipping https://source.mnt.re/reform/reform-tools/-/blob/main/xorg/10-reform-etnaviv.conf | 05:33 |
| - S0rin (QUIT: Ping timeout: 246 seconds) (~S0rin@user/s0rin) | 05:40 | |
| josch | https://source.mnt.re/reform/reform-tools/-/merge_requests/27 | 05:40 |
| + S0rin (~S0rin@user/s0rin) | 05:49 | |
| + klardotsh (~klardotsh@98.97.113.58) | 05:52 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@static-198-54-135-69.cust.tzulo.com) | 06:27 | |
| + bkeys (~Thunderbi@198.54.135.69) | 06:27 | |
| - vkoskiv (QUIT: Ping timeout: 260 seconds) (~vkoskiv@89-166-62-97.bb.dnainternet.fi) | 06:39 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 06:45 | |
| + chomwitt (~chomwitt@2a02:587:dc0c:c200:5b5:50e2:356d:c45b) | 07:26 | |
| + bgs (~bgs@212-85-160-171.dynamic.telemach.net) | 07:26 | |
| + robin (~robin@user/terpri) | 07:59 | |
| + vkoskiv (~vkoskiv@89-166-62-97.bb.dnainternet.fi) | 08:06 | |
| - klardotsh (QUIT: Ping timeout: 248 seconds) (~klardotsh@98.97.113.58) | 08:37 | |
| - bkeys (QUIT: *.net *.split) (~Thunderbi@198.54.135.69) | 08:42 | |
| - mjw (QUIT: *.net *.split) (~mark@gnu.wildebeest.org) | 08:42 | |
| - frank2 (QUIT: *.net *.split) (~frank2@juicy.frank2.net) | 08:42 | |
| - joeyh (QUIT: *.net *.split) (~joeyh@kitenet.net) | 08:42 | |
| - lastebil (QUIT: *.net *.split) (truck@shell.suomiscene.fi) | 08:42 | |
| - natalie (QUIT: *.net *.split) (~natalie@user/natalie) | 08:42 | |
| + lastebil (truck@shell.suomiscene.fi) | 08:42 | |
| + joeyh (joeyh@2600:3c03::f03c:91ff:fe73:b0d2) | 08:42 | |
| + mjw (~mark@gnu.wildebeest.org) | 08:43 | |
| + bkeys (~Thunderbi@static-198-54-135-69.cust.tzulo.com) | 08:45 | |
| + frank2 (~frank2@2a05:f480:2000:1d0d:22ba:38ed:7fdc:2b5a) | 08:45 | |
| + natalie (~natalie@user/natalie) | 08:46 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by wielaard!~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440))) (~mark@gnu.wildebeest.org) | 10:02 | |
| * wielaard -> mjw | 10:02 | |
| + mark_ (~mark@gnu.wildebeest.org) | 10:03 | |
| + MajorBiscuit (~MajorBisc@145.94.164.10) | 10:30 | |
| q66 | minute: i made a new image with fixed u-boot and new kernel https://repo.chimera-linux.org/live/testing/chimera-linux-reform-imx8mq-20221012.img.gz | 11:39 |
| q66 | i wonder if that works now | 11:39 |
| josch | q66: curl: (56) OpenSSL SSL_read: error:0A000126:SSL routines::unexpected eof while reading, errno 0 | 12:36 |
| josch | q66: attempting to boot what I was able to download leads to this: http://paste.debian.net/1256769/ | 12:38 |
| - ex-parrot (QUIT: Ping timeout: 255 seconds) (~fincham@user/ex-parrot) | 12:38 | |
| + ex--parrot (~fincham@user/ex-parrot) | 12:38 | |
| q66 | josch: what, when do you get that message | 12:50 |
| josch | q66: when booting -- i included the whole log | 12:50 |
| q66 | i mean the curl message | 12:50 |
| josch | q66: I get it when I curl that url you posted | 12:51 |
| josch | vkoskiv: did you try out 5.18 yet? I get very similar (the same?) messages in dmesg with 5.18: | 12:52 |
| josch | [ 99.455059] nvme nvme0: I/O 15 QID 0 timeout, disable controller | 12:52 |
| josch | [ 99.503568] nvme nvme0: Identify Controller failed (-4) | 12:52 |
| josch | [ 99.513382] nvme nvme0: Removing after probe failure status: -5 | 12:52 |
| josch | [ 102.795780] nvme0n1: detected capacity change from 1953525168 to 0 | 12:52 |
| q66 | josch: that's weird | 12:52 |
| vkoskiv | Didn't yet, but that's what I was getting with 5.19 | 12:52 |
| q66 | the img.gz has a sha256sum of 8ede745a69351ea0f27fe0c48780d4efa37ff086135e129358ce09694957e8f9 | 12:52 |
| q66 | er | 12:52 |
| q66 | dd049dcfcd7631962a7ed9c261726e8b14d58fadb18b3e7f2c4dea4161dad1bb | 12:52 |
| josch | q66: I do not know the checksum because I piped directly from curl into dd | 12:52 |
| q66 | ah | 12:53 |
| q66 | no weird curl messages here | 12:54 |
| q66 | i wonder what the kernel message is about | 12:55 |
| josch | q66: did you apply these patches: https://source.mnt.re/reform/reform-debian-packages/-/tree/main/linux/patches | 12:56 |
| q66 | no, but they do nto seem relevant to bootup | 12:56 |
| josch | agreed but last time i built a Debian kernel without these patches I got a kernel panic as well | 12:57 |
| josch | i do not have the log anymore so i cannot say whether it was the same panic | 12:57 |
| josch | but it was with 5.19 | 12:57 |
| q66 | hm | 12:59 |
| q66 | i guess we could add special 5.19 reform kernel with the patchset | 12:59 |
| josch | you could try which of all of these patches is the one preventing the kernel panic | 13:00 |
| q66 | it seems 6.0 broke something at very least because with my vanilla 5.19 kernel minute was able to boot yesterday | 13:00 |
| q66 | just without display | 13:00 |
| josch | oh display did come up | 13:00 |
| q66 | yeah because i enabled the respective modules for it | 13:01 |
| q66 | in the 6.0 config | 13:01 |
| josch | good, then that worked | 13:01 |
| q66 | unless the problem is that enabling them results in the panic :P | 13:01 |
| q66 | and that could then be fixed by the patches | 13:01 |
| josch | oh wait, i'm getting slightly different output on the display than via serial | 13:02 |
| q66 | ah | 13:02 |
| josch | on the display it also says: | 13:03 |
| josch | run-init: /usr/bin/init: Exec format error | 13:03 |
| q66 | wth | 13:03 |
| q66 | hold on | 13:05 |
| q66 | does it get through the initramfs then | 13:05 |
| q66 | or does it die in initramfs | 13:06 |
| josch | q66: let me re-download... | 13:06 |
| q66 | i'm confused mainly because in initramfs there is no /usr/bin/init | 13:07 |
| josch | yes, this has to come from the rootfs | 13:07 |
| q66 | if it dies in the "real" system then you should have been getting messages from the initramfs | 13:07 |
| josch | it's probably corrupted because of the error i got | 13:07 |
| q66 | stuff like init-top running etc | 13:07 |
| q66 | yeah that's possible | 13:07 |
| josch | lets try again | 13:08 |
| josch | q66: I got a shell | 13:14 |
| josch | q66: want me to run anything in it? | 13:14 |
| q66 | josch: booted and logged in? | 13:17 |
| q66 | got display too? | 13:17 |
| josch | q66: yes, user root, password chimera | 13:17 |
| q66 | nice! | 13:17 |
| josch | yes, also logged in via usb keyboard and see the /bin/sh on the display | 13:17 |
| q66 | can you take me a photo | 13:19 |
| q66 | i will tweet it later | 13:19 |
| josch | sure, just of the shell or would you like me to run anything? | 13:20 |
| q66 | boot log + uname -a or /proc/cpuinfo or sth | 13:22 |
| josch | chimera needs a more fancy /etc/motd :) | 13:24 |
| q66 | it does have one for live isos, but i have not added one for platform images | 13:26 |
| q66 | since those are bootstrapped more or less straight from packages | 13:26 |
| q66 | i mean the live images are too, but they have a more involved live-only initramfs | 13:27 |
| q66 | i guess even though it boots i'll apply some of the patches | 13:27 |
| q66 | 6.0.1 is out anyway so | 13:27 |
| josch | q66: https://mister-muffin.de/p/aIuD.jpg | 13:28 |
| q66 | excellent | 13:28 |
| q66 | do you have a twitter handle so i can give you some credit | 13:28 |
| josch | q66: no, I am not on any of those platforms | 13:29 |
| q66 | alright | 13:29 |
| josch | just post away -- no need to credit me | 13:29 |
| q66 | https://source.mnt.re/reform/reform-debian-packages/-/blob/main/linux/patches/0002-pci-imx6-add-support-for-internal-refclk-imx8mq.patch this probably applies only to reform 1 | 13:30 |
| q66 | oh, no | 13:30 |
| q66 | imx8 too | 13:30 |
| q66 | hm, or maybe not | 13:31 |
| minute | ? | 13:32 |
| q66 | well it seemed imx6-specific | 13:33 |
| minute | q66: ah, well yeah because some of the IP in the chip is carried over from imx6 | 13:33 |
| minute | josch: quite some dust on your display ^^ | 13:34 |
| minute | cool that it works! | 13:35 |
| josch | minute: that just shows that I'm using it a lot ;) | 13:35 |
| minute | josch: indeed, indeed! | 13:35 |
| josch | my wife bought me a dedicated brush to clean mechanical keyboards because she couldn't see the dirt accumulating there XD | 13:36 |
| minute | very nice | 13:36 |
| q66 | minute: how important are the first four patches in https://source.mnt.re/reform/reform-debian-packages/-/tree/main/linux/patches ? | 13:37 |
| q66 | the others seem generic enough that i can just pick them as it is (some are also already in 6.0) | 13:37 |
| + bluerise (~bluerise@user/bluerise) | 13:43 | |
| q66 | actually since it boots, it's probably ok to keep it vanilla for now and clean up later if people pick it up | 13:50 |
| - mtm (QUIT: Ping timeout: 268 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:04 | |
| q66 | doing a 6.0.1 kernel, this afternoon i'll generate all the isos and platform images and since stuff is confirmed working enough i'll add the reform image to the set | 14:11 |
| minute | q66: you want the pcie patch, otherwise there is no wifi | 15:45 |
| q66 | minute: the pcie-interference one? | 15:49 |
| q66 | does not look related | 15:49 |
| q66 | i guess the internal-refclk one | 15:50 |
| sigrid | minute: considering keyboard has a connection to som via uart, has it been considered to use the oled as a dumb serial console? | 15:53 |
| sigrid | for all those cases where having an adapter is too much :) | 15:53 |
| q66 | the internal-refclk patch appears to require changes to the device tree as well so it will not work cleanly on top of mainline as far as i can tell | 16:02 |
| q66 | in the reform dts in that repo, it appears set for pcie0, which is not even defined in the upstream dts | 16:03 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:10 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 17:05 | |
| josch | sigrid: the usb keyboard does not work in uboot right now | 17:11 |
| sigrid | if usb is started by uboot, it's going to be at least powered, right? | 17:13 |
| sigrid | that would be enough for it to read uart, no? | 17:13 |
| josch | maybe -- all I observed so far was, that after enabling usb (uboot is able to load kernel and initramfs from a usb stick) the keyboard still remained unresponsive | 17:14 |
| sigrid | I don't think even usb start is needed | 17:14 |
| sigrid | otherwise how would the circle key work | 17:15 |
| sigrid | it would probably need a hackish way of switching between uart/usbhid mode though | 17:16 |
| sigrid | ie where the keys should go | 17:16 |
| josch | I bought a 1 TB Kingston NV2 NVMe PCIe SSD and this drive also errors out after resuming from suspend: http://paste.debian.net/1256804/ | 17:26 |
| josch | What nvme drives are you guys using that are known to work with suspend? | 17:26 |
| sigrid | wd blue sn550, also 570 | 17:27 |
| sigrid | I know people complained it does not work, but what actually does not work is linux | 17:28 |
| sigrid | specifically 5.19, it seems | 17:28 |
| minute | sigrid: yes it has been considered @ kbd console, would be quite interesting | 17:31 |
| minute | josch, sigrid: the keyboard has 2 power supplies. one is always on (3v3 standby). that's enough for serial+oled. the 5v is toggled on by USB. it also makes the USB keyboard function work. those two can work independently. | 17:32 |
| minute | (the backlight also needs the 5v, that's why it turns on when USB is probed) | 17:33 |
| minute | q66: yes, the internal refclk one is needed for wifi. | 17:33 |
| minute | q66: ah yeah, mainline doesn't have this in dts precisely because the internal refclk for pcie0 is not supported in mainline. | 17:34 |
| q66 | minute: if i take that patch and then add the pcie0 definitions in the upstream dts, that should be enough to get wifi going? | 18:03 |
| minute | q66: i think so | 18:04 |
| q66 | minute: https://gist.github.com/q66/e8ddf8070ff642973ea51a147487dfe9 | 18:10 |
| q66 | missing anything? | 18:10 |
| q66 | the non-dts part of the patch is just yours | 18:10 |
| q66 | i copied in the dts bits | 18:11 |
| * xktr_ -> xktr | 18:21 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark_!~mark@gnu.wildebeest.org))) (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 18:30 | |
| * mark_ -> mjw | 18:30 | |
| + wielaard (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 18:31 | |
| minute | q66: imx6_pcie->internal_refclk is never set | 18:34 |
| minute | q66: in the patch | 18:34 |
| q66 | minute: ? | 18:34 |
| minute | q66: so i think there's a snippet missing, where it sets that depending on dts? | 18:34 |
| - MajorBiscuit (QUIT: Ping timeout: 268 seconds) (~MajorBisc@145.94.164.10) | 18:34 | |
| q66 | line 138 of patch | 18:34 |
| minute | q66: ah sorry, you're right | 18:34 |
| minute | q66: so if there are no errors compiling the dts, it should be fine | 18:35 |
| minute | i think nowadays the whole anatop hacking can actually be removed and it can be done in dts somehow, but i don't have the time to figure it out atm | 18:36 |
| q66 | i think if there were errors they'd get printed to stdout | 18:37 |
| q66 | the log is silent though | 18:37 |
| q66 | so it should have compiled fine | 18:37 |
| minute | good! | 18:46 |
| josch | sigrid: but I see the same problem in 5.18 (see the error log I pasted above) | 19:29 |
| sigrid | I get the power state change failure but it's working fine | 19:29 |
| sigrid | no other errors | 19:29 |
| josch | so... of those people who claim to have decent suspend with the reform -- what nvme drive are you using? | 19:56 |
| vkoskiv | I had a 600P. Suspend and resume still do work on the bad WD drive, but the drive doesn't come up | 20:06 |
| + ajr (~ajr@user/ajr) | 20:13 | |
| minute | josch: transcend | 20:26 |
| josch | minute: and your transcend drive has resume mostly working on 5.19? | 21:00 |
| minute | josch: yes | 21:40 |
| josch | ACTION clicks the buy button | 21:40 |
| minute | there are different versions of the drive though i think | 21:40 |
| minute | with and without DRAM buffer for example | 21:41 |
| josch | minute: which are all called MTE220S? | 21:41 |
| minute | i have 256GB TS256GMTE220S | 21:42 |
| minute | this has the dram cache | 21:42 |
| minute | yeah | 21:42 |
| josch | I'm going to try the 1 TB version TS1TMTE220S | 21:44 |
| josch | The product description also says DDR3 SLC-Cache for that one | 21:44 |
| minute | ok, good luck! | 21:45 |
| josch | Thanks! Will report back my findings once I have it. | 21:45 |
| minute | cool! | 21:46 |
| josch | minute: you already hinted that you might offer the 3d printed trackball buttons in your shop -- could the same be done for the compex wifi card? | 21:46 |
| minute | we decided today to keep the changes in the new motherboard rev minimal | 21:46 |
| minute | josch: sure, i'll write that down | 21:46 |
| josch | nice! | 21:47 |
| vkoskiv | Which chips did you need to swap out? | 21:47 |
| minute | vkoskiv: so far only TUSB8041 (to TUSB8040) and INA260 | 21:47 |
| minute | we decided to stay with LPC11U24 instead of migrating to RP2040 because it's not worth the reeingineering | 21:48 |
| minute | and having then 2 different firmwares to be maintained | 21:49 |
| minute | in general compatibility is a big factor | 21:49 |
| minute | so the focus is mainly on bugfixes | 21:49 |
| minute | also the flashing connector will be usb-c instead of microusb but that's a tiny detail | 21:50 |
| vkoskiv | Yeah firmware maintenance burden would certainly a bummer | 21:50 |
| minute | exactly | 21:50 |
| vkoskiv | Are microusb hard to get these days | 21:50 |
| minute | no, but it's just a worse connector | 21:50 |
| vkoskiv | The most cursed connector in existence is micro B usb 3.x | 21:51 |
| minute | haha yes it is closely related | 21:51 |
| minute | also, several fixes for the power system will be integrated | 21:51 |
| vkoskiv | I recently found an external drive with that connector, and it has an USB C to that horrendous connector cable | 21:51 |
| minute | and the headphone bass fix | 21:51 |
| vkoskiv | Cool stuff. Good chance to fix things with a new rev | 21:52 |
| vkoskiv | Is the one hand-tweaked SMD cap/resistor on my board one of these fixes? :D | 21:52 |
| vkoskiv | R53 | 21:52 |
| josch | vkoskiv: I assume you are familiar with xkcd's "cursed connector" series? ;) | 21:53 |
| josch | (micro B usb 3.x looks so bad, that first time i saw it i assumed it was some proprietary connector) | 21:53 |
| minute | the funny thing about micro B usb 3 is that it is backwards compatible to micro usb | 21:55 |
| vkoskiv | Yeah it just has that normal micro b bit with the extra tumor on the side | 21:58 |
| - Nulo (QUIT: Ping timeout: 246 seconds) (~Nulo@user/nulo) | 22:11 | |
| - jjbliss (QUIT: Read error: Connection reset by peer) (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 22:13 | |
| + jjbliss (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 22:29 | |
| vagrantc | ok, going to the mnt/reform meetup in seattle on the 6th! | 22:48 |
| minute | vagrantc: awesome | 23:03 |
| - bgs (QUIT: Remote host closed the connection) (~bgs@212-85-160-171.dynamic.telemach.net) | 23:20 | |
| - chomwitt (QUIT: Ping timeout: 268 seconds) (~chomwitt@2a02:587:dc0c:c200:5b5:50e2:356d:c45b) | 23:51 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
