| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.56.60) | 00:10 | |
| + Ar|stote|is (~linx@149.210.56.60) | 00:12 | |
| + cow321 (~deflated8@user/meow/deflated8837) | 00:29 | |
| - Gooberpatrol_66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 00:43 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 00:44 | |
| - gsora (QUIT: Ping timeout: 260 seconds) (~gsora@user/gsora) | 01:19 | |
| + gsora (~gsora@user/gsora) | 01:33 | |
| - cow321 (QUIT: Ping timeout: 252 seconds) (~deflated8@user/meow/deflated8837) | 02:26 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 02:33 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 02:33 | |
| * bkeys1 -> bkeys | 02:35 | |
| - erle (QUIT: Ping timeout: 246 seconds) (~erle@user/erle) | 02:44 | |
| + cow321 (~deflated8@user/meow/deflated8837) | 02:45 | |
| - mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 03:01 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 03:07 | |
| - nsc (QUIT: Ping timeout: 252 seconds) (~nicolas@53-99-142-46.pool.kielnet.net) | 03:26 | |
| + nsc (~nicolas@i5C74DC12.versanet.de) | 03:27 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 03:29 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 03:29 | |
| * bkeys1 -> bkeys | 03:31 | |
| - paperManu (QUIT: Ping timeout: 248 seconds) (~paperManu@198.16.214.40) | 03:41 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 04:05 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 04:06 | |
| * bkeys1 -> bkeys | 04:08 | |
| - spew (QUIT: Quit: spew) (~spew@201.141.99.170) | 04:28 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@173.16.175.75) | 04:33 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 04:38 | |
| - Gooberpatrol_66 (QUIT: Ping timeout: 276 seconds) (~Gooberpat@user/gooberpatrol66) | 04:39 | |
| - sts-q (QUIT: Ping timeout: 276 seconds) (~sts-q@2a11:fb80:3f0:8401:48b5:7b37:c759:8a59) | 04:48 | |
| + sts-q (~sts-q@89.58.73.41) | 04:54 | |
| - CandidCadaver (QUIT: Ping timeout: 248 seconds) (~aloo_shu@90.166.98.221) | 05:21 | |
| - Ar|stote|is (QUIT: Ping timeout: 276 seconds) (~linx@149.210.56.60) | 05:26 | |
| + Ar|stote|is (~linx@149.210.16.205) | 05:30 | |
| + CandidCadaver (~aloo_shu@90.166.98.221) | 05:33 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 06:03 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 06:43 | |
| - cobra (QUIT: Ping timeout: 246 seconds) (~cobra@user/Cobra) | 06:44 | |
| + blast007` (~blast@user/blast007) | 07:05 | |
| - blast007 (QUIT: Read error: Connection reset by peer) (~blast@user/blast007) | 07:05 | |
| + cobra (~cobra@user/Cobra) | 07:06 | |
| - grimmware (QUIT: Ping timeout: 276 seconds) (sid638840@id-638840.hampstead.irccloud.com) | 07:22 | |
| + grimmware (sid638840@hampstead.irccloud.com) | 07:23 | |
| + gustav28 (~gustav@c-5d34524e.019-141-67626730.bbcust.telenor.se) | 08:11 | |
| josch | minute: another thing: is the rk3588 loading u-boot from sd-card or from emmc by default? If the latter I suppose you are flashing the emmc of the modules before you ship them? Are you using reform-flash-uboot to do that flashing? If not, could you maybe try doing that once to make sure that the version of reform-flash-uboot on release does not mess anything up? | 08:23 |
|---|---|---|
| + jaume (~user@user/jaume) | 08:32 | |
| + erle (~erle@2a02:8109:da01:6400::430b) | 09:02 | |
| - erle (QUIT: Changing host) (~erle@2a02:8109:da01:6400::430b) | 09:02 | |
| + erle (~erle@user/erle) | 09:02 | |
| minute | josch: so far it appears that sd card has priority, if that has a bootloader. with no sd card inserted it tries to boot from emmc | 09:10 |
| josch | that's nice! | 09:17 |
| * blast007` -> blast007 | 09:29 | |
| - Ar|stote|is (QUIT: Ping timeout: 252 seconds) (~linx@149.210.16.205) | 09:46 | |
| + Ar|stote|is (~linx@149.210.16.41) | 09:51 | |
| - Ar|stote|is (QUIT: Ping timeout: 260 seconds) (~linx@149.210.16.41) | 10:38 | |
| + jaume` (~user@119.15.112.1) | 10:40 | |
| - jaume (QUIT: Ping timeout: 245 seconds) (~user@user/jaume) | 10:41 | |
| + Ar|stote|is (~linx@149.210.16.24) | 10:42 | |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 11:04 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 11:04 | |
| + mjw (~mjw@gnu.wildebeest.org) | 11:40 | |
| + jaume`` (~user@67.218.253.121) | 12:37 | |
| - jaume` (QUIT: Ping timeout: 260 seconds) (~user@119.15.112.1) | 12:39 | |
| - robin (QUIT: Remote host closed the connection) (~robin@user/terpri) | 12:39 | |
| + robin (~robin@user/terpri) | 12:40 | |
| - mjw (QUIT: Ping timeout: 260 seconds) (~mjw@gnu.wildebeest.org) | 12:46 | |
| + paperManu (~paperManu@198.16.214.40) | 12:46 | |
| * tadeus_brick -> TadeusTaD | 13:00 | |
| - TadeusTaD (QUIT: Changing host) (tadeustad@psifactor.pl) | 13:00 | |
| + TadeusTaD (tadeustad@user/TadeusTaD) | 13:00 | |
| * Guest8415 -> mjw | 13:03 | |
| + tom_ (~tom@compara.offerte.fibra.casa) | 13:13 | |
| tom_ | still can't access your site | 13:13 |
| tom_ | checked ip on https://mioip.info and is ok | 13:13 |
| tom_ | did you rifrewall filter ip from italy? | 13:13 |
| gsora | wrong buffer tom_? | 13:14 |
| tom_ | dont know | 13:14 |
| tom_ | :\ | 13:14 |
| josch | tom_: i think you are in the wrong channel | 13:15 |
| tom_ | damn | 13:15 |
| tom_ | ! | 13:15 |
| tom_ | sorry | 13:15 |
| - tom_ (QUIT: Quit: I was getting bored) (~tom@compara.offerte.fibra.casa) | 13:15 | |
| + tom_ (~tom@madonna.che.casino) | 13:35 | |
| - tom_ (QUIT: Client Quit) (~tom@madonna.che.casino) | 13:38 | |
| minute | josch: ok so secondary HDMI on RK3588 strangely only works with clk_ignore_unused kernel parameter. that means some clock declaration is likely missing for the second hdmi. but i will add that to the kernel commandline then in uboot. and also the missing rotation for pocket rk3588. then the first system images will be ready | 14:08 |
| + ZylonMaster (~hjcs@syn-098-015-248-249.res.spectrum.com) | 14:31 | |
| josch | minute: okay, so for a311d u-boot we have: | 14:53 |
| josch | CONFIG_BOOTARGS="ro no_console_suspend console=ttyAML0,115200 pci=pcie_bus_perf libata.force=noncq nvme_core.default_ps_max_latency_us=0 console=tty1 fbcon=rotate:3" | 14:53 |
| josch | if you add something similar to rk3588 u-boot and tag it, then the hashes can be put into reform-tools and a new release can be made | 14:53 |
| ZylonMaster | will I get a notification if my mnt reform is coming to me? | 15:17 |
| ZylonMaster | just wondering | 15:17 |
| - jaume`` (QUIT: Remote host closed the connection) (~user@67.218.253.121) | 15:34 | |
| minute | ZylonMaster: yes | 15:45 |
| minute | josch: ah, i just need to tag, because i've done the uboot kernel cmdline fixes now and tested them with reform+pocket reform | 15:46 |
| minute | josch: tag 2024-10-10 pushed to reform-rk3588-uboot | 15:47 |
| ZylonMaster | from what I hear RK3588 can compete with the intel N100. Sounds like it must be blazingly fast for an ARM chip | 15:50 |
| ZylonMaster | its good that you picked that | 15:50 |
| minute | yes, i can't wait to put it in people's hands (working on that right now) | 15:50 |
| minute | we already got a lot of them through QA | 15:51 |
| ZylonMaster | dont blame you | 15:51 |
| ZylonMaster | thats some speed | 15:51 |
| ZylonMaster | how much you charging on crowdsupply is a good questionb | 15:51 |
| ZylonMaster | im sure some have asked that | 15:51 |
| minute | what do you mean? | 15:51 |
| ZylonMaster | well its US dollars | 15:51 |
| ZylonMaster | on crowdsupply | 15:52 |
| ZylonMaster | i wondered its differing vs your website | 15:52 |
| ZylonMaster | 500 euros is your website | 15:52 |
| minute | ah you mean if we offered rk3588 upgrade on crowd supply | 15:52 |
| minute | we haven't negotiated that yet | 15:52 |
| ZylonMaster | yeah | 15:52 |
| ZylonMaster | ah okay | 15:52 |
| ZylonMaster | probably 600 if its like the website tho | 15:52 |
| minute | we have to offer it and they have to agree to carry it etc | 15:52 |
| ZylonMaster | i would guess | 15:52 |
| ZylonMaster | but uh yeah | 15:55 |
| ZylonMaster | i wait on that | 15:55 |
| ZylonMaster | dont see why they wouldn't offer it | 15:55 |
| + bkeys (~Thunderbi@173.16.175.75) | 15:59 | |
| - ZylonMaster (QUIT: Quit: Leaving) (~hjcs@syn-098-015-248-249.res.spectrum.com) | 16:00 | |
| - bkeys (QUIT: Client Quit) (~Thunderbi@173.16.175.75) | 16:00 | |
| + bkeys (~Thunderbi@173.16.175.75) | 16:01 | |
| - bkeys (QUIT: Client Quit) (~Thunderbi@173.16.175.75) | 16:04 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 16:04 | |
| * bkeys1 -> bkeys | 16:06 | |
| - murphnj (QUIT: Quit: Leaving) (~murph@user/murphnj) | 16:10 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 16:41 | |
| ehmry | the stock debian install on my pocket broker afte I booted from an sdcard: https://paste.sr.ht/~ehmry/16620b56b0690176da538b0e5a3388d810fc562e | 16:43 |
| ehmry | … and the off switch for the power board just broke off | 16:45 |
| - mark_ (QUIT: Quit: Leaving) (~mjw@gnu.wildebeest.org) | 16:49 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 16:53 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 16:54 | |
| josch | minute: thank you! i'll combine that with fixing the expected kernel cmdline in the machines.conf files -- just to be sure, could you copypaste me your current /proc/cmdline with the updated u-boot? | 16:54 |
| josch | ehmry: is there anything interesting above the first "Kernel panic"? | 16:55 |
| josch | ah that's probably due to your loglevel setting | 16:55 |
| * bkeys1 -> bkeys | 16:56 | |
| - bkeys (QUIT: Client Quit) (~Thunderbi@173.16.175.75) | 16:57 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 16:57 | |
| josch | ehmry: your log shows 6.8.9-mnt-reform-arm64 -- that's quite old. What was your setup and what did you change? | 16:59 |
| * bkeys1 -> bkeys | 17:00 | |
| - Ar|stote|is (QUIT: Ping timeout: 276 seconds) (~linx@149.210.16.24) | 17:00 | |
| ehmry | josch: I didn't change anything | 17:03 |
| + Ar|stote|is (~linx@149.210.16.212) | 17:04 | |
| josch | ehmry: you said you booted from an sd-card. Are you saying that you didn't change your original install, then booted a rescue image, shut it off, took out the sd-card, booted again and then it was broken? | 17:04 |
| ehmry | yes | 17:05 |
| josch | that is... odd... | 17:05 |
| ehmry | the image on the external didn't sucessfully boot, it did a fallback to debian | 17:05 |
| ehmry | well hopefully I can get it to boot from sdcard | 17:07 |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 17:07 | |
| + bkeys (~Thunderbi@173.16.175.75) | 17:07 | |
| - bkeys (QUIT: Client Quit) (~Thunderbi@173.16.175.75) | 17:10 | |
| + bkeys (~Thunderbi@173.16.175.75) | 17:10 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 17:13 | |
| + bkeys (~Thunderbi@173.16.175.75) | 17:13 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 17:43 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 17:44 | |
| * bkeys1 -> bkeys | 17:46 | |
| + spew (~spew@201.141.99.170) | 17:52 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 17:55 | |
| + bkeys (~Thunderbi@173.16.175.75) | 17:55 | |
| minute | josch: i'm having some last minute issues with rk3588 uboot SPL. first of note: the rk doesn't seem to use the boot partitions but expects the bootloader on the main emmc partition (i stumbled over this before). then, for some reason the current uboot build's SPL only tries to boot from MMC1 (eMMC) and fails if that's not there, even the sd card version. i have an older sd card with uboot from | 18:21 |
| minute | Aug 28 and there it correctly tries to load uboot from MMC2. so i need to dig into that now | 18:21 |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 18:24 | |
| + bkeys (~Thunderbi@173.16.175.75) | 18:25 | |
| - NanoCodeBug (QUIT: Ping timeout: 252 seconds) (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 18:25 | |
| + NanoCodeBug (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 18:25 | |
| minute | reading https://github.com/u-boot/u-boot/blob/master/doc/usage/spl_boot.rst#image-load-methods | 18:28 |
| hramrach | there is some logic to determine from what device the SPL booted and try loading the next stage from that, maybe it broke | 18:31 |
| minute | the docs mention board_boot_order() but that's only defined in very old boards, so the docs are likely outdated | 18:33 |
| hramrach | depending on the SPL implementation there can be a fixed order or logic that tries to load the next stage from the same medium | 18:33 |
| minute | rockchip-common.h has #define BOOT_TARGETS "mmc1 mmc0 nvme scsi usb pxe dhcp spi" | 18:35 |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 18:36 | |
| + bkeys (~Thunderbi@173.16.175.75) | 18:37 | |
| minute | but BOOT_TARGETS are not referenced in spl code | 18:38 |
| minute | mhm, there is ./arch/arm/mach-rockchip/spl-boot-order.c | 18:40 |
| minute | ah, they have special fdt nodes | 18:43 |
| minute | u-boot,spl-boot-order | 18:43 |
| hramrach | also spl_decode_boot_device | 18:45 |
| minute | aha, there's a special rk3588-rock-5b-u-boot.dtsi | 18:46 |
| minute | and i changed to our own dts so that stuff is now gone | 18:47 |
| minute | but that dtsi has > u-boot,spl-boot-order = "same-as-spl", &sdmmc, &sdhci; | 18:47 |
| minute | is there a convention that auto includes -u-boot.dtsi for a given dts? because it's not explicitly included anywhere | 18:48 |
| minute | adding that to our uboot .dtses | 18:50 |
| hramrach | yes, I think there is some fixup that appends the foo-u-boot.dtsi to foo.dts or something like that | 18:52 |
| minute | mhm makes sense | 18:52 |
| hramrach | hm, there are a lot of device trees that include it explicitly | 18:53 |
| minute | hmmmmm | 18:53 |
| minute | ah maybe i grepped wrongly | 18:53 |
| minute | > rg rk3588-rock-5b-u-boot | 18:54 |
| minute | only yields board/radxa/rock5b-rk3588/MAINTAINERS | 18:54 |
| hramrach | but they often include a different one | 18:54 |
| minute | ok, but that was it, fixes SPL | 18:55 |
| hramrach | yes, it should work automagically, described in doc/develop/devicetree/control.rst | 18:55 |
| hramrach | the cases when the u-boot.dtsi is included are cases like yours when it is reused | 18:56 |
| hramrach | I suopose you could rename it to rk3588-u-boot.dtsi to get it included in all rk3588 boards | 18:58 |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 18:58 | |
| + bkeys (~Thunderbi@173.16.175.75) | 18:58 | |
| - bkeys (QUIT: Client Quit) (~Thunderbi@173.16.175.75) | 18:59 | |
| + bkeys (~Thunderbi@173.16.175.75) | 19:00 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:02 | |
| minute | i somehow broke one of the modules so that i now get "Found non-UXN mapping" during boot | 19:02 |
| minute | hramrach: thx | 19:02 |
| minute | this is in kernel_init -> ptdump_check_wx | 19:03 |
| minute | > watchdog: BUG: soft lockup - CPU#4 stuck for 104s! | 19:04 |
| minute | oh strange, uboot says 16GB memory but this board has 32GB | 19:05 |
| minute | so perhaps wrong dram init | 19:05 |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 19:09 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 19:09 | |
| minute | ah no, that was a copypaste error, butttt the problem remains. so the problem is caused by the mainline SPL/ddr blob | 19:10 |
| minute | looks like the icore modules were always booting vendor SPL from emmc which was in turn booting the rest of our uboot from SD | 19:11 |
| minute | but now that i've overwritten that i get different output from ddr blob and then later memory errors | 19:11 |
| minute | vendor > DDR V1.10 75d050770f typ 23/02/28-20:47:10 | 19:11 |
| * bkeys1 -> bkeys | 19:11 | |
| minute | collabora > DDR Version V1.08 20220617 | 19:11 |
| minute | josch: btw: /scripts/init-top/reform: line 21: setfont: not found | 19:17 |
| minute | aha, it boots on the "broken" module with mem=1G which confirms it has to do with ddr setup | 19:18 |
| minute | so probably the ddr blob included in the collabora uboot build process can't deal with 32GB correctly, or with the DDR chips used on the firefly | 19:19 |
| josch | minute: that line in /scripts/init-top/reform was added by hramrach in https://source.mnt.re/reform/reform-tools/-/merge_requests/87 but that commit also copied /usr/bin/setfont into the initramfs... hrm... | 19:20 |
| minute | josch: that doesn't seem to work yeah | 19:21 |
| ch | huh, that should just work | 19:22 |
| josch | indeed | 19:22 |
| josch | it was also tested by hramrach | 19:22 |
| ch | is it possible the image doesnt have setfont in the first place? | 19:22 |
| josch | so i'm curious to find out how it broke for you | 19:22 |
| josch | ugh... and then the hook would *not* error out? | 19:23 |
| josch | sec... | 19:23 |
| ch | dunno, maybe | 19:23 |
| josch | everything is possible with mkinitramfs | 19:23 |
| ch | yeah | 19:23 |
| josch | would not be the first surprise... | 19:23 |
| josch | no, the vanilla system image has /usr/bin/setfont | 19:25 |
| josch | is this maybe a merged-/usr problem? | 19:26 |
| ch | please no | 19:26 |
| josch | the canonical path in the kbd package is /bin/setfont | 19:26 |
| ch | the initramfs is supposed to be merged, iirc | 19:26 |
| josch | this problem escaped me because i'm not testing the image on the pocket of course... :/ | 19:27 |
| josch | let me see if i can reproduce it in qemu... | 19:28 |
| josch | ah i see the bug! | 19:29 |
| josch | sorry! | 19:29 |
| ch | curious now | 19:30 |
| hramrach | hm, the vendor ddr blob claims it's newer. Can it be extracted? | 19:30 |
| josch | ch: https://source.mnt.re/reform/reform-tools/-/blob/main/initramfs-tools/hooks/reform?ref_type=heads#L35 | 19:30 |
| minute | hramrach: digging into that atm... | 19:31 |
| josch | this only copies in /usr/bin/setfont *if* the hook is run *on* the pocket | 19:31 |
| josch | ch: and the system image is not built on the pocket :) | 19:31 |
| minute | interesting, github code search for 75d050770f gives error 502 | 19:31 |
| ch | ugh | 19:31 |
| minute | aha https://github.com/rockchip-linux/rkbin/blob/master/doc/release/RK3588_EN.md | 19:31 |
| ch | josch: but that means the LS1028A workaround is also wrong? | 19:32 |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 19:32 | |
| ch | (running the hook -on- the pocket does work) | 19:32 |
| + bkeys (~Thunderbi@173.16.175.75) | 19:32 | |
| minute | so it's part of rkbin. and collabora cloned some old rkbin | 19:32 |
| ch | s/workaround/firmware copying/ | 19:32 |
| josch | ch: yes, but only very few people are running an ls1028a and those who do already have their machines set up and probably are not trying out new system images :) | 19:32 |
| ch | right | 19:33 |
| hramrach | josch: I would hope that's handled by the /etc/flash-kernel/machine | 19:33 |
| minute | aha wait > export ROCKCHIP_TPL=../rkbin/bin/rk35/rk3588_ddr_lp4_2112MHz_lp5_2736MHz_v1.08.bin | 19:33 |
| minute | there's the version right there. so DDR init is in TPL | 19:33 |
| josch | hramrach: you might be right... maybe my conclusion was too quick | 19:34 |
| josch | i'll dissect an initramfs... | 19:34 |
| minute | but the rkbin fork is 2 years old anywy | 19:34 |
| ch | stupid initramfs brightness hack in case anyone wants it: https://gist.githubusercontent.com/zeha/121b12c1ef540ee17df807fae0288f57/raw/ee83a61dda413a6fc658abb83abc1a0942bcdad8/brightness | 19:36 |
| hramrach | thre is https://github.com/rockchip-linux/rkbin/blob/master/bin/rk35/rk3588_ddr_lp4_2112MHz_lp5_2736MHz_eyescan_v1.11.bin | 19:37 |
| minute | yeah, but firstly i will try https://github.com/rockchip-linux/rkbin/blob/master/bin/rk35/rk3588_ddr_lp4_2112MHz_lp5_2400MHz_v1.16.bin | 19:37 |
| ch | josch: i think your analysis is spot on | 19:37 |
| ch | ttyl | 19:37 |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 19:37 | |
| + bkeys (~Thunderbi@173.16.175.75) | 19:38 | |
| josch | ch: it might not because mkimage sets /etc/flash-kernel/machine | 19:38 |
| josch | as hramrach pointed out | 19:38 |
| josch | investigating... | 19:38 |
| minute | indeed the new ddr blob makes it work ;3 | 19:40 |
| josch | yup, setfont is missing from pocket-reform-system-imx8mp.img.gz | 19:40 |
| josch | minute: phew, glad you found the fix quickly! | 19:40 |
| minute | now by accident i caught the spurious crash in rk_hdptx_phy_runtime_resume... it's a null pointer deref. probably some race condition | 19:41 |
| minute | josch: yessss also very happy about that | 19:41 |
| minute | josch: first i thought i burnt the module... | 19:41 |
| josch | ch, hramrach: can either of you confirm that the currently published reform system images do not boot? https://source.mnt.re/reform/reform-system-image/-/jobs/6202/artifacts/browse | 19:41 |
| josch | minute: that would've been a big uff... :( | 19:42 |
| josch | ehmry: the currently published system images for the pocket reform might indeed be broken | 19:43 |
| minute | josch: the setfont stuff is not fatal though, it only means that the font stays small in initramfs | 19:44 |
| minute | note to self > [ 3.424539] pc : rk_hdptx_phy_runtime_resume+0x24/0xd8 [phy_rockchip_samsung_hdptx] | 19:44 |
| josch | minute: oh, so it just continues and just prints an error to the terminal? | 19:45 |
| minute | josch: yes, the setfont error is printed multiple times | 19:45 |
| minute | but it continues | 19:46 |
| josch | hah indeed the script is missing a set -e :) | 19:46 |
| josch | hramrach: okay, so the mkimage.sh script indeed cycles through all possible values for /etc/flash-kernel/machine contents but it then only runs flash-kernel and it does *not* run update-initramfs -u again | 19:48 |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 19:50 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 19:50 | |
| hramrach | I never tried to run another image. Does wwriting it to a SD card suffice to test it? | 19:50 |
| josch | yes, inserting an sd-card with the system image should show the issue | 19:50 |
| josch | hramrach: but minute already confirmed that the problem is not fatal, so i guess it's not that critical to reproduce it | 19:51 |
| josch | hramrach: but if you have some time, i'd appreciate a test of the fixed image later :) | 19:51 |
| hramrach | the setfont script is only cosmetic, there is no problem if it does not work unless you really want to see what's written to the display | 19:52 |
| * bkeys1 -> bkeys | 19:52 | |
| josch | hramrach: right, i feared that it was connected with the problem ehmry mentioned earlier and i missed that the initramfs script is indeed not run with set -e so nothing can break it :) | 19:53 |
| hramrach | it does return a failure in the end if it returns the last command return value but initramfs does not seem to mind | 19:56 |
| minute | interesting, the rk3588 ram blob is ARM64 | 19:56 |
| hramrach | it might be code that is executed on the main CPU | 19:57 |
| hramrach | some older rockchip SoCs have driver for this in u-boot and don't need a blob, iirc | 19:58 |
| minute | yeah, that's nice... on imx8m* and a311d it's a special ISA (ARCompact) that's hard to RE | 19:59 |
| - bkeys (QUIT: Ping timeout: 260 seconds) (~Thunderbi@173.16.175.75) | 20:01 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-72-235.tukw.qwest.net) | 20:02 | |
| + colinsane (~colinunin@97-113-72-235.tukw.qwest.net) | 20:06 | |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 20:15 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 20:15 | |
| + bkeys (~Thunderbi@173.16.175.75) | 20:21 | |
| - bkeys (QUIT: Client Quit) (~Thunderbi@173.16.175.75) | 20:22 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 20:22 | |
| * bkeys1 -> bkeys | 20:25 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 20:31 | |
| josch | untested: https://source.mnt.re/reform/reform-system-image/-/merge_requests/111 | 20:31 |
| josch | afk for a few hours now | 20:31 |
| + bkeys (~Thunderbi@173.16.175.75) | 20:32 | |
| - mjw (QUIT: Killed (tantalum.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 20:32 | |
| * mark_ -> mjw | 20:32 | |
| + Guest3540 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 20:33 | |
| minute | meh, why does CONFIG_BOOTDELAY=0 not work | 20:34 |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 20:34 | |
| + bkeys (~Thunderbi@173.16.175.75) | 20:34 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 20:45 | |
| + bkeys (~Thunderbi@173.16.175.75) | 20:46 | |
| minute | something puts bootdelay=2 into the env | 20:48 |
| minute | strange because "bootdelay="__stringify(CONFIG_BOOTDELAY)"\0" in env_default.h | 20:48 |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 20:54 | |
| + bkeys (~Thunderbi@173.16.175.75) | 20:54 | |
| - Ar|stote|is (QUIT: Ping timeout: 276 seconds) (~linx@149.210.16.212) | 20:58 | |
| minute | ah, nevermind, all good | 20:59 |
| minute | SPL hash checking seems slowish | 21:00 |
| + Ar|stote|is (~linx@149.210.16.41) | 21:02 | |
| ehmry | is there a way to boot the pocket from uboot on an sdcard rather than the onboard u-boot? | 21:03 |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 21:05 | |
| + bkeys (~Thunderbi@173.16.175.75) | 21:05 | |
| minute | ehmry: not the imx8mplus pocket | 21:07 |
| ehmry | ok | 21:07 |
| minute | ehmry: if there's a problem with the emmc flash bootloader you have to resort to USB reflashing which is documented in community | 21:08 |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 21:09 | |
| minute | ehmry: see also https://community.mnt.re/t/flashing-emmc-bootloader-unbricking-imx8mp/2287/7 | 21:09 |
| - Gooberpatrol66 (QUIT: Ping timeout: 276 seconds) (~Gooberpat@user/gooberpatrol66) | 21:10 | |
| ehmry | okay, no problem with my emmc bootloader yet so I won't worry about that | 21:12 |
| minute | ok | 21:14 |
| minute | ok emmc boot time on rk3588 pocket now down to 10 seconds. not awesome but also not bad | 21:15 |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 21:24 | |
| + jacobk_ (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 21:24 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 21:29 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 21:29 | |
| * bkeys1 -> bkeys | 21:31 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 21:52 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 21:52 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 21:52 | |
| * bkeys1 -> bkeys | 21:54 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-72-235.tukw.qwest.net) | 21:55 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 21:58 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 21:58 | |
| + colinsane (~colinunin@97-113-72-235.tukw.qwest.net) | 21:59 | |
| * bkeys1 -> bkeys | 22:00 | |
| ch | josch: do you still need sth tested? | 22:08 |
| - Ar|stote|is (QUIT: Ping timeout: 248 seconds) (~linx@149.210.16.41) | 22:14 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-5d34524e.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| + Ar|stote|is (~linx@149.210.16.147) | 22:18 | |
| - jacobk_ (QUIT: Ping timeout: 276 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 22:26 | |
| josch | ch: it should be working now: https://source.mnt.re/reform/reform-system-image/-/jobs/6220/artifacts/browse | 22:30 |
| josch | hramrach: if you want to test something ^ | 22:37 |
| ch | dd'ing it now | 22:40 |
| ch | maybe i should get an usb3 sd card reader/writer | 22:41 |
| josch | i heard there is this new sd-card standard | 22:44 |
| ch | :( | 22:45 |
| ch | i already dont know what any of the icons on the cards mean | 22:45 |
| josch | the one i probably mean is ultra extreme pro 2000 or so | 22:46 |
| ch | §B47 | 22:46 |
| ch | SD Express | 22:46 |
| ch | image boots with correct font, resizing atm | 22:48 |
| josch | cool! | 22:48 |
| josch | minute: the "setfont" problem should be fixed with newer system images -- do you have some /proc/cmdline output with more recent u-boot for me? :) | 22:49 |
| ch | sigh | 22:49 |
| ch | at some point the font is switched back | 22:50 |
| ch | but i vaguely remember this is not a new problem :> | 22:50 |
| - mjw (QUIT: Remote host closed the connection) (~mjw@gnu.wildebeest.org) | 22:51 | |
| + mjw (~mjw@gnu.wildebeest.org) | 22:51 | |
| ch | eh, if i close the setup wizard with the X in the top right, i'm left in a broken sway with no way to quit(?) | 22:51 |
| josch | i thought i had opened an issue about that already but it seems not | 22:53 |
| josch | ch: this is for you: https://source.mnt.re/reform/mnt-reform-setup-wizard/-/issues/ | 22:53 |
| ch | thx, made #19 | 22:55 |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 22:56 | |
| minute | ch: bmaptool also helps! (re: imaging speed) | 22:57 |
| + bkeys1 (~Thunderbi@173.16.175.75) | 22:57 | |
| * bkeys1 -> bkeys | 22:59 | |
| ch | dpkg-reconfigure console-setup -> Do not change the boot/kernel font -> much nicer | 23:15 |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 23:16 | |
| + bkeys (~Thunderbi@173.16.175.75) | 23:16 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 23:26 | |
| - Gooberpatrol_66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 23:29 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 23:30 | |
| josch | that should be console-setup/use_system_font | 23:34 |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.16.147) | 23:42 | |
| + Ar|stote|is (~linx@149.210.16.147) | 23:44 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@173.16.175.75) | 23:49 | |
| + bkeys (~Thunderbi@173.16.175.75) | 23:50 | |
| - bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@173.16.175.75) | 23:55 | |
| + bkeys1 (~Thunderbi@173.16.175.75) | 23:55 | |
| * bkeys1 -> bkeys | 23:57 | |
| - erle (QUIT: Quit: K-lined) (~erle@user/erle) | 23:57 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!