| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 00:05 | |
| + pocketminute (~reform@91.64.136.147) | 00:20 | |
| - S0rin (QUIT: Ping timeout: 252 seconds) (~S0rin@user/s0rin) | 00:28 | |
| - pocketminute (QUIT: Quit: leaving) (~reform@91.64.136.147) | 00:31 | |
| + S0rin (~S0rin@user/s0rin) | 00:38 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~wielaard@gnu.wildebeest.org) | 00:43 | |
| + mjw (~wielaard@gnu.wildebeest.org) | 00:45 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 01:03 | |
| + robin (~robin@user/terpri) | 01:11 | |
| - robin (QUIT: Read error: Connection reset by peer) (~robin@user/terpri) | 01:20 | |
| + robin_ (~robin@user/terpri) | 01:20 | |
| * robin_ -> robin | 01:32 | |
| - mjw (QUIT: Ping timeout: 272 seconds) (~wielaard@gnu.wildebeest.org) | 01:47 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 02:06 | |
| - ajr (QUIT: Quit: WeeChat 3.7) (~ajr@user/ajr) | 02:37 | |
| - nsc (QUIT: Ping timeout: 268 seconds) (~nicolas@169-97-142-46.pool.kielnet.net) | 03:05 | |
| + nsc (~nicolas@205-96-142-46.pool.kielnet.net) | 03:07 | |
| * nsc -> Guest3598 | 03:08 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 03:09 | |
| cinap_lenrek | https://source.mnt.re/reform/reform-boundary-uboot/-/merge_requests/12 | 04:17 |
|---|---|---|
| cinap_lenrek | so, i suppose more people can now test this and give suggestions how to improve the patch | 04:19 |
| - wielaard (QUIT: Ping timeout: 260 seconds) (~mjw_@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 05:08 | |
| + bkeys1 (~Thunderbi@static-198-54-135-69.cust.tzulo.com) | 06:47 | |
| - bkeys1 (QUIT: Client Quit) (~Thunderbi@static-198-54-135-69.cust.tzulo.com) | 06:52 | |
| + wielaard (~mjw_@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 07:17 | |
| - bgs (QUIT: Remote host closed the connection) (~bgs@212-85-160-171.dynamic.telemach.net) | 08:21 | |
| + MajorBiscuit (~MajorBisc@83.232.58.225) | 09:24 | |
| - MajorBiscuit (QUIT: Read error: Connection reset by peer) (~MajorBisc@83.232.58.225) | 09:32 | |
| * wielaard -> mjw | 09:56 | |
| + MajorBiscuit (~MajorBisc@145.94.179.130) | 10:17 | |
| minute | cinap_lenrek: highly appreciated | 10:23 |
| - MajorBiscuit (QUIT: Quit: WeeChat 3.6) (~MajorBisc@145.94.179.130) | 10:33 | |
| + MajorBiscuit (~MajorBisc@145.94.179.130) | 11:01 | |
| + holo_memory (~holo_memo@dynamic-077-013-163-022.77.13.pool.telefonica.de) | 12:50 | |
| - mtm (QUIT: Ping timeout: 272 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 13:04 | |
| - holo_memory (QUIT: Quit: Client closed) (~holo_memo@dynamic-077-013-163-022.77.13.pool.telefonica.de) | 14:16 | |
| josch | nvme error-log /dev/nvme0n1 | wc -l --> 962 | 14:26 |
| josch | whoops... that would explain why my reform doesn't boot anymore... | 14:26 |
| josch | anybody else has seen this kind of errors during operation? http://paste.debian.net/1266872/ | 14:26 |
| josch | nvme smart-log shows 97% "available_spare" -- what other diagnostic could i run on the drive? | 14:28 |
| josch | (this is the Transcend 220S M.2 2280 which also comes with the reform by default) | 14:34 |
| c-keen[m] | I wonder whether not cleanly shutting the drive down damaged the block | 14:56 |
| c-keen[m] | which reminds be to automate the backups for the reform | 14:56 |
| c-keen[m] | josch: you could try to fsck that fs to backup data | 15:03 |
| Boostisbetter | <--- just dreaming about the Pocket Reform | 15:09 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 15:09 | |
| josch | yup, a fsck cleanup up tons of errors and deleted quite a number of files... luckily i have backups but it'd be nice to know more about this situation, like how it happened or whether the drive is broken now or not | 15:58 |
| c-keen[m] | what does the smart log say about unclean restarts? | 16:01 |
| josch | minute: what do you think about adding the nvme-cli tool to the reform-rescue image? It would require 1595 kB of space | 16:05 |
| josch | c-keen[m]: http://paste.debian.net/1266882/ | 16:06 |
| c-keen[m] | power_cycles: 74 | 16:12 |
| c-keen[m] | unsafe_shutdowns: 37 | 16:12 |
| c-keen[m] | that's more than half | 16:12 |
| c-keen[m] | but for your situation this is critical: media_errors: 857 | 16:12 |
| sigrid | 14:26 < josch> nvme error-log /dev/nvme0n1 | wc -l --> 962 | 16:15 |
| sigrid | do you actually see *errors* there though? | 16:15 |
| sigrid | wc -l is definitely not the way to count them | 16:16 |
| sigrid | and I believe media errors are normal for nvmes | 16:16 |
| c-keen[m] | [ 4748.023349] critical medium error, dev nvme0n1, sector 284581920 op 0x0:(READ) flags 0x80700 phys_seg 1 prio class 2 | 16:16 |
| c-keen[m] | full of these ^ | 16:16 |
| sigrid | unless ofc it skyrockets | 16:16 |
| sevan | to sidestep the issue I reboot and the power off from the keyboard when the screen goes dark. unsafe shutdown counter stopped incrementing after that. | 16:44 |
| sigrid | that's because how shutdown is down right now is wrong - it forcefully turns off power though lpc without kernel shutting down nvme first, as far as I understand | 16:45 |
| sigrid | *is done | 16:45 |
| sevan | I should have a nose around with nvmecli, I've been using smartmontools | 16:45 |
| sevan | sigrid: ack | 16:45 |
| c-keen[m] | isn't that just an ordering issue in the systemd shutdown sequence? | 16:46 |
| sigrid | no | 16:46 |
| - MajorBiscuit (QUIT: Ping timeout: 272 seconds) (~MajorBisc@145.94.179.130) | 17:18 | |
| + MajorBiscuit (~MajorBisc@145.94.179.130) | 17:23 | |
| sigrid | cinap_lenrek: minute: I am testing the lcdif uboot patch and it works so far, but with linux trying to use mipi dsi with lcdif still on (I guess) I see no output with linux | 17:25 |
| sigrid | I flashed the same one to emmc but for some reason it does not seem to boot. not sure if it's a different build I should make for it to work | 17:26 |
| sigrid | it boots fine from mmc :/ | 17:26 |
| sigrid | *sd card | 17:26 |
| cinap_lenrek | sigrid: interesting | 17:41 |
| sigrid | another thing is none of modifiers seem to work | 17:42 |
| sigrid | I can't type ":" | 17:42 |
| cinap_lenrek | sigrid: hahaha | 17:42 |
| cinap_lenrek | uboot quality! | 17:42 |
| sigrid | I am not sure whether it's keyboard firmware or u-boot :/ | 17:43 |
| sigrid | since it worked literally anywhere else, I assume it's u-boot | 17:43 |
| cinap_lenrek | uboot uses boot protocol i think | 17:43 |
| cinap_lenrek | not the report protocol | 17:43 |
| sigrid | ah ok | 17:43 |
| cinap_lenrek | (with the keyboard) | 17:43 |
| sigrid | probably why :/ | 17:43 |
| cinap_lenrek | maybe that could be a difference? | 17:43 |
| sigrid | yes | 17:44 |
| cinap_lenrek | i can debug it | 17:44 |
| sigrid | I fucked up my emmc so I can't boot into 9front | 17:44 |
| cinap_lenrek | no problem | 17:44 |
| cinap_lenrek | boot from sdcard :) | 17:44 |
| sigrid | I will try | 17:45 |
| sigrid | thank god u-boot defaults to partition 0 | 17:46 |
| sigrid | since I can type "0" but not "0:0" :D | 17:46 |
| cinap_lenrek | hahaha | 17:47 |
| cinap_lenrek | okay | 17:47 |
| cinap_lenrek | lemme try to fix that | 17:47 |
| sigrid | another thing: since 9front reinits lcdif, it can end up shifted permanently | 17:47 |
| sigrid | I mean the screen image | 17:47 |
| cinap_lenrek | sigrid: thats 9fronts problem, no? | 17:48 |
| cinap_lenrek | i mean, maybe we need to reset it differently? | 17:48 |
| sigrid | probably | 17:49 |
| cinap_lenrek | i havnt experienced the shifting tho | 17:49 |
| cinap_lenrek | poco a poco | 17:49 |
| cinap_lenrek | lemme build system image to see if i can boot linux | 17:51 |
| cinap_lenrek | and also add some debugging for the keyboard | 17:51 |
| cinap_lenrek | i suppose keyboard is higher prio as if you cant type : you'r screwed | 17:52 |
| sigrid | usb errored after a bit in 9front | 17:52 |
| sigrid | maybe should turn it off in u-boot before booting. maybe it's 9front's problem | 17:52 |
| cinap_lenrek | yeah, possible | 17:52 |
| cinap_lenrek | especially xhci | 17:52 |
| cinap_lenrek | you really want that to be off as it is dmaing | 17:53 |
| cinap_lenrek | i suppose uboot should have infrastructure for this? | 17:53 |
| cinap_lenrek | to turn it off? | 17:53 |
| sigrid | pretty sure it does have that | 17:53 |
| cinap_lenrek | tho i didnt have issues with usb | 17:53 |
| sigrid | another thing: mmc command hangs every so often | 17:58 |
| sigrid | unless you type things in a very specific order it might just crash | 17:59 |
| josch | sigrid: hrm... i wanted to paste the output of "nvme error-log /dev/nvme0n1" but now it's suddenly filled with a lot of success messages. All the entries (i scrolled through it) were errors when i counted 962 lines. | 17:59 |
| sigrid | cinap_lenrek: maybe our boot.scr should have "usb stop f && bootm ..." | 18:02 |
| sigrid | so far it works fine that way | 18:02 |
| sigrid | josch: weird | 18:02 |
| josch | ugh... i wish i had saved the log... i guess somehow have to try and create the same errors again for the log to show them because they vanish after a power cycle? | 18:02 |
| cinap_lenrek | ok | 18:04 |
| cinap_lenrek | keyboard never sets modifier | 18:04 |
| cinap_lenrek | but it generates scancodes for the shift/ctl/alt keys | 18:05 |
| eery[m] | Will the SoM try booting from MMC if loading from SD fails, or does the DIP switch have to be flipped? | 18:07 |
| minute | eery[m]: dip switch has to be flipped | 18:07 |
| josch | eery[m]: the SoM will read u-boot from eMMC or SD-Card depending on the dip switch position and not fall back to the other | 18:07 |
| - MajorBiscuit (QUIT: Ping timeout: 260 seconds) (~MajorBisc@145.94.179.130) | 18:08 | |
| sigrid | "mmc rescan; mmc dev 1" -> hang | 18:09 |
| josch | aha, i did a read-only run of badblocks on /dev/nvme0n1 and after a few seconds badblocks started being enumerated and now i have something in the error log: https://paste.debian.net/hidden/019fbee8/ | 18:09 |
| sigrid | so I can't access the sd card at all from u-boot | 18:09 |
| sigrid | maybe that's why linux did not show anything on the screen. pretty sure it did not even boot | 18:10 |
| sigrid | cinap_lenrek: could there be some conflict between sd and lcdif? | 18:10 |
| sigrid | I assume it does work for you? | 18:11 |
| cinap_lenrek | sigrid: no issues for me, no | 18:11 |
| cinap_lenrek | i get all the drives listed | 18:11 |
| cinap_lenrek | both emmc and sdcards | 18:11 |
| cinap_lenrek | lemme try yout command line | 18:12 |
| cinap_lenrek | => mmc rescan; mmc dev 1 | 18:12 |
| cinap_lenrek | switch to partitions #0, OK | 18:12 |
| cinap_lenrek | mmc1 is current device | 18:12 |
| cinap_lenrek | => | 18:12 |
| cinap_lenrek | oh | 18:13 |
| cinap_lenrek | but! | 18:13 |
| cinap_lenrek | the screen turns off | 18:13 |
| cinap_lenrek | hahahaha | 18:13 |
| sigrid | yeah... | 18:13 |
| cinap_lenrek | wtf | 18:13 |
| cinap_lenrek | and sometimes, mmc rescan hangs | 18:14 |
| cinap_lenrek | okay | 18:14 |
| - mjw (QUIT: Quit: Leaving) (~mjw_@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:26 | |
| cinap_lenrek | sigrid: https://source.mnt.re/reform/reform-boundary-uboot/-/merge_requests/12 | 18:46 |
| cinap_lenrek | sigrid: that fixes the keyboard | 18:46 |
| sigrid | nice, that's better than changing keyboard fw :) | 18:47 |
| cinap_lenrek | so, next lets see whats up with the mmc | 18:48 |
| sigrid | http://okturing.com/src/14808/body | 18:49 |
| sigrid | fixed two warnings | 18:49 |
| cinap_lenrek | lemme merge that | 18:50 |
| sigrid | I need to fix ext4srv speed, copying a linux kernel (20Mb) to emmc takes ages | 18:50 |
| sigrid | *31Mb :( | 18:51 |
| sigrid | +20Mb initramfs | 18:51 |
| cinap_lenrek | pushed. | 18:52 |
| cinap_lenrek | i'll squash the commits later | 18:52 |
| cinap_lenrek | this is wild hacker branch :D | 18:52 |
| sigrid | I am going to re-test linux manually | 18:52 |
| cinap_lenrek | i'm *STILL* compiling system image @_@ | 18:53 |
| sigrid | you could just download it, no? | 18:53 |
| cinap_lenrek | its fine | 18:53 |
| sigrid | yay, I can type :! | 18:57 |
| + ajr (~ajr@user/ajr) | 19:02 | |
| sigrid | retried linux manually and still nothing on the screen. it just gets cleared out | 19:03 |
| cinap_lenrek | gpio_request(GP_EMMC_RESET, "usdhc2_reset"); | 19:20 |
| cinap_lenrek | gpio_direction_output(USDHC2_PWR_GPIO, 0); | 19:20 |
| cinap_lenrek | haha | 19:20 |
| cinap_lenrek | nope. | 19:21 |
| cinap_lenrek | => mmc rescan | 19:37 |
| cinap_lenrek | unable to select a mode | 19:37 |
| cinap_lenrek | hm | 19:37 |
| cinap_lenrek | it doesnt hang sigrid | 19:37 |
| sigrid | with no changes? | 19:42 |
| sigrid | or is that a serial output | 19:42 |
| + mjw (~wielaard@gnu.wildebeest.org) | 19:46 | |
| cinap_lenrek | no | 19:48 |
| cinap_lenrek | it just times out | 19:48 |
| cinap_lenrek | also, mmc dev 1 is what kills the display | 19:48 |
| cinap_lenrek | so lemme debug what happens there | 19:48 |
| cinap_lenrek | mmc dev 0 works fine | 19:48 |
| sigrid | which is emmc for you, right? | 19:49 |
| cinap_lenrek | i dont know | 19:50 |
| - bkeys (QUIT: Ping timeout: 260 seconds) (~Thunderbi@static-198-54-135-69.cust.tzulo.com) | 19:50 | |
| sigrid | ls mmc 0 / | 19:50 |
| sigrid | guess if you see a shitload of linux files there, that's emmc | 19:50 |
| + bkeys (~Thunderbi@static-198-54-135-69.cust.tzulo.com) | 19:51 | |
| cinap_lenrek | i think i killed the fs there | 19:55 |
| cinap_lenrek | so yeah | 19:55 |
| cinap_lenrek | mmc 1 is sdcard | 19:55 |
| cinap_lenrek | mmc 0 is emmc | 19:55 |
| cinap_lenrek | i think all this is more related with the usb really | 19:56 |
| cinap_lenrek | let me retry with usb disabled | 19:56 |
| sigrid | fwiw I tried stopping it before switching to dev 1 and that did not help | 19:57 |
| sigrid | it's interesting that the same exact u-boot image booting from emmc never shows me anything, just loads into 9front | 20:14 |
| - mjw (QUIT: Ping timeout: 260 seconds) (~wielaard@gnu.wildebeest.org) | 20:32 | |
| eery[m] | Another dumb question, would a pcie wifi card not being used consume a measurable amount of power? | 20:52 |
| cinap_lenrek | ok | 20:54 |
| cinap_lenrek | the display turns off when: | 20:55 |
| cinap_lenrek | mmc_power_off(): calling regulator_set_enable() | 20:55 |
| cinap_lenrek | mmc_power_off(): regulator_set_enable() -> 0 | 20:55 |
| sigrid | minute: where does u-boot reside on the emmc? | 20:55 |
| sigrid | I located my serial, and I see that usb is not even started when I'm booting from emmc :/ | 20:56 |
| sigrid | looks like I wrote the updated u-boot to it but no idea what's actually booting | 20:56 |
| cinap_lenrek | fixed_regulator_set_enable: dev='regulator-wlan-vmmc', enable=0, delay=70000, has_gpio=1 | 21:03 |
| cinap_lenrek | fixed_regulator_set_enable: done | 21:03 |
| cinap_lenrek | hahaha | 21:03 |
| cinap_lenrek | wlan.... | 21:03 |
| cinap_lenrek | i'm pretty sure this is a bad devicetree file | 21:03 |
| sigrid | did not know u-boot has device trees too | 21:04 |
| cinap_lenrek | pinctrl_reg_wlan_vmmc: reg-wlan-vmmcgrp { | 21:04 |
| cinap_lenrek | fsl,pins = < | 21:04 |
| cinap_lenrek | #define GP_REG_WLAN_VMMC <&gpio3 20 GPIO_ACTIVE_HIGH> | 21:04 |
| cinap_lenrek | MX8MQ_IOMUXC_SAI5_RXC_GPIO3_IO200x16 | 21:04 |
| cinap_lenrek | >; | 21:04 |
| cinap_lenrek | }; | 21:04 |
| cinap_lenrek | yep | 21:04 |
| cinap_lenrek | and gpio3_20 is the bridge enable gpio | 21:04 |
| cinap_lenrek | no wonder the screen turns off | 21:04 |
| sigrid | whoops. | 21:04 |
| cinap_lenrek | i hate this shit so much | 21:05 |
| cinap_lenrek | so much indirection | 21:05 |
| cinap_lenrek | ok | 21:05 |
| cinap_lenrek | lets comment out this thing | 21:05 |
| sigrid | matryoshka-driven development | 21:05 |
| cinap_lenrek | so now is the question | 21:15 |
| cinap_lenrek | there seems to be a gpio pin for powering the sdcard slot | 21:15 |
| cinap_lenrek | gpio2_19 | 21:15 |
| cinap_lenrek | maybe it was intended to be that? | 21:15 |
| cinap_lenrek | sigrid: okay, pushed | 21:21 |
| cinap_lenrek | so | 21:23 |
| cinap_lenrek | remaining issue leenuks | 21:23 |
| cinap_lenrek | how to make it go? | 21:24 |
| minute | sigrid: it is on the magical boot partition | 21:24 |
| minute | mmcblk0boot0 | 21:24 |
| cinap_lenrek | minute: https://source.mnt.re/cinap_lenrek/reform-boundary-uboot/-/commit/438465fd5af2d183b1a66e7fa5ab42e6d05c3240 | 21:25 |
| cinap_lenrek | minute: it is probably not correct | 21:25 |
| cinap_lenrek | i just move the usdhc2 vdd to fixed3v3 | 21:25 |
| cinap_lenrek | it seems to still work | 21:25 |
| cinap_lenrek | but you are the wizzard here | 21:26 |
| cinap_lenrek | but GPIO3_IO20 is definitely the wrong gpio :) | 21:26 |
| sigrid | minute: I meant the offset on the device itself | 21:26 |
| sigrid | I keep writing over with bs=1024 and skip 33 blocks but it looks like it's still the same old uboot getting loaded | 21:27 |
| sigrid | I even read it all back and compared, and it looks different | 21:27 |
| sigrid | idk what's going on | 21:27 |
| sigrid | md5sum is the same too | 21:31 |
| - cinap_lenrek (QUIT: Read error: Connection reset by peer) (~cinap_len@ns3076381.ip-147-135-136.eu) | 21:34 | |
| + cinap_lenrek (~cinap_len@ns3076381.ip-147-135-136.eu) | 21:35 | |
| cinap_lenrek | sorry for the reconnect | 21:35 |
| cinap_lenrek | did i miss anything? | 21:35 |
| sigrid | not really | 21:36 |
| sigrid | cinap_lenrek: https://www.kernel.org/doc/Documentation/mmc/mmc-dev-parts.txt | 21:37 |
| sigrid | do you know whether those boot partitions supposed to be accessible as usual over /dev/sdO0/data? | 21:37 |
| cinap_lenrek | i think you need to issue special commands | 21:37 |
| cinap_lenrek | right now i'm not supporting this | 21:38 |
| sigrid | I am confused whether it's a feature of emmc itself or just some safety mechanism on linux | 21:38 |
| cinap_lenrek | i think it is a special thing in emmc | 21:38 |
| sigrid | oh well... | 21:38 |
| cinap_lenrek | like theres some magic boot protocol where bootloader just downloads a blob | 21:39 |
| cinap_lenrek | i'm not sure how this blob relates to the disk really | 21:39 |
| cinap_lenrek | if its just like some special reserved sectors or something | 21:40 |
| cinap_lenrek | i have no idea | 21:40 |
| cinap_lenrek | https://felloff.net/doc/comp/spec/JESD84-A43.pdf | 21:42 |
| sigrid | damn, yep | 21:43 |
| sigrid | looks like it's special regions | 21:43 |
| cinap_lenrek | yes | 21:43 |
| cinap_lenrek | its separate | 21:43 |
| sigrid | makes sense it kept the same bootloader | 21:43 |
| sigrid | it probably still writes over that address, but into user region | 21:44 |
| sigrid | so then it all checks out. | 21:44 |
| sigrid | lemme boot into linuks | 21:44 |
| cinap_lenrek | ok | 21:45 |
| cinap_lenrek | so we need some switcheroo | 21:45 |
| cinap_lenrek | set some ext_csd bits | 21:45 |
| cinap_lenrek | and then the boot partition becomes visible | 21:45 |
| sigrid | it has TWO boot partitions :) | 21:47 |
| sigrid | ok, updated uboot, it boots with backlight on but nothing shows on the lcd | 21:50 |
| cinap_lenrek | no idea | 21:52 |
| sigrid | but mmc 1 works now | 21:52 |
| cinap_lenrek | maybe theres some timing sensitivity? | 21:54 |
| cinap_lenrek | sigrid: remove the first line comment for #define DEBUG | 21:56 |
| cinap_lenrek | maybe some of the system clocks are different? | 21:58 |
| cinap_lenrek | i dont know | 21:58 |
| cinap_lenrek | http://okturing.com/src/14809/body | 21:59 |
| cinap_lenrek | these are the clocks when you boot from sdcard | 21:59 |
| cinap_lenrek | maybe you can check if they differ when booting from emmc? | 21:59 |
| + XYZ (~XYZ@37-48-51-185.nat.epc.tmcz.cz) | 22:00 | |
| + mjw (~wielaard@gnu.wildebeest.org) | 22:00 | |
| cinap_lenrek | it is unlikely but who knows | 22:00 |
| sigrid | all the same | 22:01 |
| sigrid | oh. | 22:04 |
| sigrid | lcdif: video_hw_init(): i2c read error for edid | 22:04 |
| cinap_lenrek | interesting | 22:04 |
| cinap_lenrek | this is usually clock related | 22:05 |
| cinap_lenrek | maybe increase the delay? | 22:05 |
| cinap_lenrek | mdelay(200); | 22:05 |
| cinap_lenrek | maybe make this bigger? | 22:05 |
| cinap_lenrek | we can definitely talk to the bridge chip | 22:07 |
| cinap_lenrek | so the i2c is fine | 22:07 |
| cinap_lenrek | but it is not forwarding the aux channel stuff | 22:07 |
| sigrid | ok, I'll get the default system image on an sd card so I can flash u-boot in a better way | 22:23 |
| sigrid | the one I used doesn't like the fact it can't initialize the display | 22:23 |
| cinap_lenrek | working as fast as i can :) | 22:23 |
| sigrid | I am very slow because of my current setup | 22:24 |
| sigrid | reform's parts all around, wires. I have to keep my left arm raised because of the serial interface sticking out etc... | 22:25 |
| sigrid | my desk is not very useful for this | 22:25 |
| cinap_lenrek | sigrid: get a usbsdmux and a pdu :) | 22:34 |
| cinap_lenrek | my desk is covered in tabacoo :D | 22:34 |
| cinap_lenrek | ok | 22:34 |
| cinap_lenrek | so as a first step | 22:34 |
| cinap_lenrek | i'll try the system image | 22:34 |
| cinap_lenrek | then flash the new uboot with the lcdif not doing anything | 22:35 |
| cinap_lenrek | to see if its just that linux requires the dcss/hdmi shit from uboot or not | 22:35 |
| cinap_lenrek | or if we'r somehow breaking linux kms stuff | 22:35 |
| cinap_lenrek | if its the second issue, we can just undo the lcdif initialization before starting linux kernel | 22:36 |
| cinap_lenrek | yey | 22:37 |
| cinap_lenrek | flash done | 22:37 |
| cinap_lenrek | lets see what happens | 22:37 |
| sigrid | system image looks like works just fine | 22:41 |
| sigrid | so probably was some problem with the one I used before | 22:41 |
| cinap_lenrek | hahahaha | 22:42 |
| cinap_lenrek | WHAT | 22:42 |
| cinap_lenrek | linux boots fine | 22:42 |
| sigrid | yeah | 22:42 |
| cinap_lenrek | with screen | 22:42 |
| sigrid | sorry for the noise then | 22:42 |
| cinap_lenrek | its just that the panel is sifted | 22:42 |
| cinap_lenrek | espectacularrr! | 22:42 |
| sigrid | to the right? | 22:42 |
| cinap_lenrek | yes | 22:42 |
| sigrid | that's what I see sometimes | 22:42 |
| cinap_lenrek | with linux? | 22:43 |
| cinap_lenrek | or 9front? | 22:43 |
| sigrid | just 9front | 22:43 |
| cinap_lenrek | and linux you get black screen? | 22:43 |
| sigrid | when I worked on blanking I would get it _constantly_ | 22:43 |
| sigrid | nah, it's fine now | 22:43 |
| sigrid | with the official reform's image | 22:43 |
| sigrid | the shift got solved partially by that magic flag | 22:46 |
| sigrid | but I wonder if something goes wrong because things go out of sync | 22:46 |
| cinap_lenrek | wait | 22:47 |
| sigrid | with blanking I had to stop lcdif between frames iirc | 22:47 |
| cinap_lenrek | so linux works fine for you? | 22:47 |
| sigrid | ys | 22:47 |
| sigrid | was a bad image I used before | 22:47 |
| cinap_lenrek | maybe my system image is too old? | 22:47 |
| sigrid | I used https://source.mnt.re/reform/reform-system-image/-/jobs/1009/artifacts/file/reform2-imx8mq/reform-rescue-system.img.gz | 22:47 |
| cinap_lenrek | pdating 4b78b93..0d905f4 | 22:48 |
| cinap_lenrek | Fast-forward | 22:48 |
| cinap_lenrek | reform2-imx8mq/mkuserland.sh | 4 ++-- | 22:48 |
| cinap_lenrek | reform2-imx8mq/mkuserland3.sh | 4 ++-- | 22:48 |
| sigrid | heh. | 22:48 |
| cinap_lenrek | for me linux screen is always shifted | 22:49 |
| sigrid | oh it worked | 22:50 |
| sigrid | but also shifted now :) | 22:50 |
| sigrid | 9front - all good | 22:52 |
| cinap_lenrek | yeah | 22:52 |
| sigrid | so looks like msleep(200) -> msleep(2000) helped | 22:52 |
| sigrid | lemme disable debug again | 22:53 |
| cinap_lenrek | i think for linux, we need to disable the bridge | 22:53 |
| sigrid | yeah | 22:53 |
| sigrid | I wonder if perhaps linux keeps it on without purpose? | 22:54 |
| cinap_lenrek | where is uboot loading linux? | 22:54 |
| sigrid | I recall something awkward in the device tree | 22:54 |
| sigrid | from boot.scr | 22:54 |
| sigrid | there's also two slightly different configurations - with hdmi and without | 22:55 |
| cinap_lenrek | yeah | 22:55 |
| cinap_lenrek | and uboot has some code to load some firmware blobs | 22:55 |
| cinap_lenrek | hdmi can fuck itself | 22:55 |
| sigrid | disabled debug -> stuff works | 22:57 |
| sigrid | so http://okturing.com/src/14811/body helped | 22:58 |
| cinap_lenrek | here we go | 23:14 |
| + inhji777 (~inhji777@ip-095-223-184-073.um35.pools.vodafone-ip.de) | 23:14 | |
| cinap_lenrek | you need to disable mipi dsi core | 23:14 |
| sigrid | before booting into linux? | 23:15 |
| cinap_lenrek | yes | 23:19 |
| cinap_lenrek | theres a cleanup_before_linux() function where i put the call | 23:20 |
| sigrid | when hdmi is enabled linux is using lcdif, so that naturally results in black screen | 23:21 |
| sigrid | ie: http://okturing.com/src/14812/body | 23:21 |
| cinap_lenrek | pfff | 23:23 |
| cinap_lenrek | now linux fails to initialize sdcard | 23:23 |
| cinap_lenrek | like fuck all this bullshit | 23:23 |
| - inhji777 (QUIT: Remote host closed the connection) (~inhji777@ip-095-223-184-073.um35.pools.vodafone-ip.de) | 23:26 | |
| cinap_lenrek | ah | 23:34 |
| cinap_lenrek | found it | 23:34 |
| cinap_lenrek | gating off display clock killed it | 23:34 |
| cinap_lenrek | so really, all linux needs todo is assert the mipi resets | 23:36 |
| cinap_lenrek | i think | 23:37 |
| sigrid | Device 0: unknown device | 23:54 |
| sigrid | Card did not respond to voltage select! | 23:54 |
| sigrid | mmc_init: -95, time 37 | 23:54 |
| sigrid | I get this all the time | 23:54 |
| sigrid | it just hangs there | 23:55 |
| sigrid | if I do "mmc dev 0" it boots fine afterwards. that's without the sd card present | 23:55 |
| minute | you're doing the good work i see. | 23:57 |
| sigrid | https://ftrv.se/_/mntre_uboot.jpg | 23:59 |
| minute | sigrid: awesome | 23:59 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
