| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 00:18 | |
| + pomel0 (~pomel0@user/pomel0) | 00:19 | |
| elb | huh well I had a new experience this evening; the Pocket rebooted, then it got stuck in a boot loop where after I put the password into unlock the SSD, it would brown out and reboot again some time between then and the login screen | 00:52 |
|---|---|---|
| elb | putting it on external power worked immediately | 00:52 |
| minute | elb: ah yes, this is the low voltage issue | 00:58 |
| minute | elb: i think that should be fixed by charger board v2 | 00:59 |
| elb | good to hear | 01:01 |
| elb | I've had it reboot spuriously before, but not cycle and be unable to boot again | 01:02 |
| - schalken (QUIT: Ping timeout: 256 seconds) (~schalken@117-118-178-69.gci.net) | 01:07 | |
| + schalken (~schalken@117-118-178-69.gci.net) | 01:08 | |
| minute | elb: yeah this cycling is very familiar to me from my tests in the last weeks | 01:29 |
| elb | understood, I look forward to hopefully not experiencing it again ;-) | 01:30 |
| + potash (~potash@user/foghorn) | 02:00 | |
| - chomwitt (QUIT: Ping timeout: 246 seconds) (~chomwitt@2a02:85f:9a5f:900:1ac0:4dff:fedb:a3f1) | 02:13 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 02:40 | |
| - schalken (QUIT: Ping timeout: 255 seconds) (~schalken@117-118-178-69.gci.net) | 02:41 | |
| + schalken (~schalken@117-118-178-69.gci.net) | 02:42 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@h193.131.19.98.dynamic.ip.windstream.net) | 03:10 | |
| - paperManu (QUIT: Ping timeout: 256 seconds) (~paperManu@198.58.139.163) | 03:17 | |
| - schalken (QUIT: Ping timeout: 244 seconds) (~schalken@117-118-178-69.gci.net) | 03:21 | |
| + schalken (~schalken@117-118-178-69.gci.net) | 03:23 | |
| - schalken (QUIT: Ping timeout: 256 seconds) (~schalken@117-118-178-69.gci.net) | 03:30 | |
| + schalken (~schalken@117-118-178-69.gci.net) | 03:31 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 03:48 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.6.130) | 05:13 | |
| + Ar|stote|is (~linx@149.210.6.130) | 05:22 | |
| - lexik (QUIT: Quit: No Ping reply in 180 seconds.) (~lexik@93.185.97.218) | 06:19 | |
| + lexik (~lexik@93.185.97.218) | 06:20 | |
| - digitalrane (QUIT: Ping timeout: 250 seconds) (~digitalra@user/digitalrane) | 07:26 | |
| - lanodan (QUIT: Ping timeout: 260 seconds) (~lanodan@2a01:e0a:d6:9930::35) | 07:28 | |
| + lanodan (~lanodan@2a01:e0a:d6:9930::35) | 07:32 | |
| - robin (QUIT: Ping timeout: 246 seconds) (~robin@user/terpri) | 07:37 | |
| + robin (~robin@user/terpri) | 07:42 | |
| - LainIwakura (QUIT: Quit: Client closed) (~LainIwaku@user/LainIwakura) | 08:02 | |
| - jogu (QUIT: Remote host closed the connection) (~jogu@user/jogu) | 08:33 | |
| + SpaceCheeseWizar (~SpaceChee@50.46.38.223) | 08:42 | |
| - SpaceCheeseWizar (QUIT: Quit: Client closed) (~SpaceChee@50.46.38.223) | 09:18 | |
| + gustav25 (~gustav@c-78-82-53-228.bbcust.telenor.se) | 10:03 | |
| + LainIwakura (~LainIwaku@user/LainIwakura) | 10:23 | |
| - deuill (QUIT: Ping timeout: 246 seconds) (~deuill@mail.deuill.org) | 10:45 | |
| - pasta161 (QUIT: Quit: WeeChat 4.7.1) (~pasta161@user/xha) | 12:12 | |
| + pasta161 (~pasta161@user/xha) | 12:17 | |
| - TadeusTaD (QUIT: Remote host closed the connection) (tadeustad@user/TadeusTaD) | 12:22 | |
| + TadeusTaD (tadeustad@psifactor.pl) | 12:34 | |
| - TadeusTaD (QUIT: Changing host) (tadeustad@psifactor.pl) | 12:34 | |
| + TadeusTaD (tadeustad@user/TadeusTaD) | 12:34 | |
| grimmware | morning | 13:10 |
| - iank (QUIT: Quit: ZNC 1.8.2+deb2+deb11u1 - https://znc.in) (~iank@fsf/staff/iank) | 13:25 | |
| + iank (~iank@fsf/staff/iank) | 13:28 | |
| + paperManu (~paperManu@198.58.139.163) | 13:49 | |
| + paperManu_ (~paperManu@198.58.209.52) | 13:58 | |
| - paperManu (QUIT: Ping timeout: 264 seconds) (~paperManu@198.58.139.163) | 14:00 | |
| + chomwitt (~chomwitt@2a02:85f:9a5f:900:1ac0:4dff:fedb:a3f1) | 14:14 | |
| + bkeys (~Thunderbi@98.19.131.193) | 14:39 | |
| + erle (~erle@user/erle) | 15:42 | |
| + jogu (~jogu@user/jogu) | 16:02 | |
| josch | _o/ | 16:09 |
| grimmware | josch: can you just give me a quick sanity check - if I want to set a kernel configuration option I should be doing it in reform-debian-packages/linux/config, right? | 16:11 |
| grimmware | I'm running the kernel I built from https://source.mnt.re/grimmware/reform-debian-packages/-/commit/91850541266c0667caf54a795cfac3d4a28219fe but the config option doesn't appear to have been set | 16:15 |
| grimmware | I checked the build output for the configuration option as well and no dice | 16:16 |
| ch | pff almost cooked the pocket with a coffee spill | 17:15 |
| + FirefoxDeHuk (~FirefoxDe@109.108.69.106) | 17:29 | |
| - chomwitt (QUIT: Ping timeout: 256 seconds) (~chomwitt@2a02:85f:9a5f:900:1ac0:4dff:fedb:a3f1) | 17:35 | |
| josch | grimmware: theoretically, yes | 17:38 |
| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 17:38 | |
| + pomel0 (~pomel0@user/pomel0) | 17:38 | |
| josch | but CONFIG_DEBUG_INFO_BTF should've been enabled in the Debian arm64 kernel for about 4 years now: https://salsa.debian.org/kernel-team/linux/-/commit/929891281c61ce4403ddd869664c949692644a2f | 17:38 |
| josch | grimmware: but i also do not see CONFIG_DEBUG_INFO_BTF=y being set in my own /boot/config* | 17:39 |
| josch | there may be some special casing around kernels with debug info maybe | 17:39 |
| - LainIwakura (QUIT: Quit: Client closed) (~LainIwaku@user/LainIwakura) | 17:40 | |
| - FirefoxDeHuk (QUIT: Quit: Client closed) (~FirefoxDe@109.108.69.106) | 17:43 | |
| minute | super strange https://tooting.ch/@vimja/115435628986765190 | 18:04 |
| minute | everyone with rk3588, do you have "aes" etc in cat /proc/cpuinfo ? | 18:05 |
| minute | i do, but vimja doesn't | 18:05 |
| minute | reportedly also on latest system image | 18:05 |
| ch | no aes for me | 18:05 |
| ch | :o | 18:05 |
| minute | wtf | 18:05 |
| ch | i'm -somewhat- sure when i got it i saw it in a casual check | 18:05 |
| minute | is this a regression? wouldn't that cause a lot more CPU load? | 18:05 |
| ch | yeah thats probably terrible | 18:05 |
| minute | could it explain the weird reboots some people have? | 18:06 |
| minute | (probably not, but...) | 18:06 |
| ch | kinda doubt it? but i guess we'll miss other cpu features too? | 18:06 |
| minute | yes | 18:06 |
| minute | pmull | 18:06 |
| ch | is this something tf-a/uboot needs to setup? | 18:06 |
| minute | good question, i thought this can be queried from the processor directly | 18:07 |
| ch | https://paste.debian.net/plain/1402676 | 18:07 |
| minute | ch: wow bizarre | 18:07 |
| minute | i have: | 18:07 |
| minute | fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm lrcpc dcpop asimddp | 18:07 |
| minute | so you also miss atomics | 18:07 |
| ch | so basically i have none of the typical 'crypto' things | 18:07 |
| ch | ew | 18:08 |
| ch | 6.16.12-mnt-reform-arm64 ftr | 18:08 |
| minute | ah you have atomics | 18:08 |
| ch | i was wondering just a few moments ago why all disk stuff seemed so slow | 18:09 |
| minute | what's missing is: aes pmull sha1 sha2 | 18:09 |
| ch | no aes accel would explain that (for FDE) | 18:09 |
| minute | yes | 18:09 |
| minute | i also did a gnome-disks test on holo_memory's pocket last week and was wondering why the transfer rate looked pretty slow | 18:10 |
| minute | he also told me "the computer is booting more slowly now" iirc | 18:10 |
| minute | ch: what's your kernel commandline? | 18:10 |
| minute | ah mine seems old | 18:10 |
| minute | > Kernel command line: ro no_console_suspend cryptomgr.notests clk_ignore_unused cma=256M swiotlb=65535 fbcon=rotate:3 fbcon=font:TER16x32 console=tty1 splash | 18:10 |
| ch | ro no_console_suspend cryptomgr.notests clk_ignore_unused cma=256M swiotlb=65535 fbcon=rotate:3 fbcon=font:TER16x32 video=DSI-1:panel_orientation=left_side_up console=tty1 | 18:10 |
| minute | do we still have to rotate fbcon? :0 | 18:11 |
| minute | (but unrelated) | 18:11 |
| ch | i havent touched the cmdline in a while :) | 18:11 |
| ch | unrelated: https://github.com/hbiyik/rkddr | 18:13 |
| minute | ah yeah i also saw that. i also saw that the cpu clocks could probably go much lower than they can go now | 18:14 |
| minute | ok, what about that aes mess... | 18:14 |
| minute | looking at https://elixir.bootlin.com/linux/v6.16.12/source/arch/arm64/kernel/cpufeature.c#L3147 | 18:17 |
| + LainIwakura (~LainIwaku@user/LainIwakura) | 18:17 | |
| ch | which uboot do you have btw? | 18:18 |
| minute | hmm, how to find out on a booted system... | 18:19 |
| ch | currently looking at strings /dev/mmcblk0 | grep U-Boot | 18:19 |
| minute | ah yeah | 18:19 |
| minute | > U-Boot SPL 2024.10-g424c714eb247-dirty (May 06 2025 - 15:46:29 +0000) | 18:19 |
| ch | bit inconclusive but i want to say | 18:19 |
| ch | U-Boot SPL 2024.10-g424c714eb247-dirty (Nov 13 2024 - 21:36:24 +0000) | 18:19 |
| minute | ok yours seems older | 18:19 |
| minute | well, older build | 18:20 |
| minute | but same git checkout | 18:20 |
| - pomel0 (QUIT: Ping timeout: 240 seconds) (~pomel0@user/pomel0) | 18:21 | |
| - LainIwakura (QUIT: Quit: Client closed) (~LainIwaku@user/LainIwakura) | 18:21 | |
| + pomel0 (~pomel0@user/pomel0) | 18:21 | |
| ch | Features : fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm lrcpc dcpop asimddp | 18:22 |
| ch | after reform-flash-uboot emmc && poweroff | 18:22 |
| ch | U-Boot SPL 2024.10-g424c714eb247-dirty (May 06 2025 - 15:46:29 +0000) | 18:23 |
| + LainIwakura (~LainIwaku@user/LainIwakura) | 18:25 | |
| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 18:26 | |
| + pomel0 (~pomel0@user/pomel0) | 18:26 | |
| minute | ch: ok wat | 18:26 |
| minute | ch: uboot update brought you the instructions? | 18:27 |
| ch | yes | 18:27 |
| minute | wow | 18:27 |
| ch | newer tf-a in that build or so? | 18:27 |
| ch | i think the tf-a version isnt pinned, right? | 18:27 |
| minute | oh yes very possible | 18:27 |
| minute | normally cpu register ID_AA64ISAR0_EL1 has the bits to detect these features | 18:28 |
| ch | also i think building uboot in CI will fail now because rkbin.git bumped the versions and rk3588_ddr_lp4_2112MHz_lp5_2400MHz_v1.18.bin does not exist anymore | 18:31 |
| minute | ch: i fixed that already | 18:31 |
| minute | didn't i push? :0 | 18:31 |
| ch | not on master | 18:31 |
| minute | huh oh maybe only in networking branch | 18:31 |
| ch | maybe only in the MR | 18:31 |
| ch | https://gitlab.collabora.com/hardware-enablement/rockchip-3588/trusted-firmware-a/-/commit/b833bbe6f088e3ee78037515d6c7c5ebb6d9a0cc | 18:32 |
| ch | thats older than november, but not sure we know which tf-a was in the november uboot build | 18:33 |
| - erle (QUIT: Quit: K-lined) (~erle@user/erle) | 18:33 | |
| bremner | no aes here in Features: | 18:33 |
| - LainIwakura (QUIT: Ping timeout: 250 seconds) (~LainIwaku@user/LainIwakura) | 18:34 | |
| bremner | I have "fp asimd evtstrm crc32 atomics fphp asimdhp cpuid asimdrdm lrcpc dcpop asimddp" | 18:35 |
| ch | might want to give `reform-flash-uboot emmc` a try | 18:39 |
| + LainIwakura (~LainIwaku@user/LainIwakura) | 18:39 | |
| ch | btw, cat /proc/device-tree/chosen/u-boot,version | 18:41 |
| ch | but that has only a subset of the info | 18:42 |
| ch | maybe it'd be good to make the uboot version 'better' | 18:42 |
| minute | serial boot log should have tf-a version | 18:42 |
| minute | bremner: btw this could also explain why your suspend/wake works worse | 18:43 |
| bremner | I wondered. | 18:48 |
| bremner | ch: reform-flash-uboot hint for me? | 18:51 |
| minute | bremner: sudo reform-flash-uboot emmc | 18:51 |
| bremner | OK, rebooting... | 18:53 |
| bremner | minute: it seems to fix the bug unloading the wifi module. I now have different issues with suspend: it reports [ 259.931465] nvme 0004:41:00.0: Unable to change power state from D3hot to D0, device inaccessible\n[ 259.994019] usb usb4: root hub lost power or was reset | 19:02 |
| bremner | hrm. Well no time to follow up now, but if you want me to try something specific I can later. | 19:04 |
| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 19:50 | |
| + pomel0 (~pomel0@user/pomel0) | 19:51 | |
| ch | https://pcbview.zeha.at/?project=pocket-reform%2Fpocket-reform-motherboard%2Fpocket-reform-motherboard.kicad_pro&file=pocket-reform-motherboard.kicad_sch | 19:58 |
| ch | https://github.com/fwupd/fwupd/blob/main/plugins/vbe/README.md | 20:21 |
| ch | https://docs.google.com/document/d/e/2PACX-1vTnlIL17vVbl6TVoTHWYMED0bme7oHHNk-g5VGxblbPiKIdGDALE1HKId8Go5f0g1eziLsv4h9bocbk/pub | 20:22 |
| ch | i wonder if we can use this to update uboot | 20:22 |
| minute | ch: woaaah @ pcbview | 20:25 |
| minute | ch: did you have this already or you just set that up? | 20:25 |
| ch | just hacked it together | 20:25 |
| ch | kicanvas is obviously 99.9999% of it | 20:26 |
| ch | your idea on community sounded good, and now i can stop using the handbook PDF, hopefully | 20:27 |
| minute | woah woah | 20:27 |
| minute | ch: that's very awesome. i wonder if this could be integrated in CI | 20:27 |
| minute | lmao it got my development snapshot from yesterday https://pcbview.zeha.at/?project=reform-next%2Fnref-ports-one%2Fnref-ports-one.kicad_pro | 20:28 |
| ch | well its a git clone from today ;) | 20:29 |
| ch | kicanvas directly reads the kicad files in the browser | 20:29 |
| ch | cant serve them directly from gitlab though, i think | 20:29 |
| ch | not sure why it renders ${REFERENCE} for some things, maybe kicanvas bug | 20:30 |
| minute | ok, so i could probably put a daily cronjob on our webserver... but how did you configure it? | 20:32 |
| - wickedshell (QUIT: Ping timeout: 256 seconds) (~wickedshe@c-174-56-77-83.hsd1.nm.comcast.net) | 20:32 | |
| ch | https://kicanvas.org/embedding/#multiple-files is the basic idea | 20:34 |
| ch | i glued some really bad php/js around it https://raw.githubusercontent.com/zeha/pcbviewer/refs/heads/main/htdocs/index.php | 20:34 |
| minute | explanation for missing crypto: https://review.trustedfirmware.org/c/TF-A/trusted-firmware-a/+/29363 | 20:34 |
| ch | not sure i can recommend the glue. probably better to use CI to write some html files | 20:35 |
| ch | but you get the idea | 20:35 |
| minute | (via #linux-rockchip) | 20:35 |
| ch | is there some public explainer on how the mmcblk0 partitions work on rk3588? | 20:40 |
| ch | /dev/mmcblk0boot0 and /dev/mmcblk0boot1 seem useless? | 20:41 |
| minute | ch: the boots seem useless yes. | 20:42 |
| minute | ch: no idea if they can have some function | 20:42 |
| minute | btw i think we'll stop the edk2 effort for now and focus on getting uboot to a more modern shape (incl uboot's EFI support) | 20:42 |
| minute | cc josch | 20:43 |
| minute | counterpillow in #linux-rockchip spoke out against edk2 as well | 20:43 |
| ch | any particular reason against edk2? | 20:44 |
| minute | to paraphrase: 1. we have to use devicetree anyway, ACPI will unlikely work out (and who wants it?) 2. uboot supports EFI anyway 3. uboot actually has a TUI mode similar to UEFI boot manager 4. edk2 codebase is extremely windows/dos like | 20:45 |
| ch | ok if 1 happens then edk2 is probably meh | 20:45 |
| minute | and the question if we use edk2 was prompted by the aes missing stuff | 20:46 |
| minute | suggesting that many things are broken with edk2 | 20:46 |
| minute | and we (mostly ailurux) tried pretty hard getting the display panel to work in edk2 and failed so far | 20:49 |
| ch | sigh | 21:02 |
| ch | the quectel i got is too long | 21:03 |
| ch | didnt occur to me to check this | 21:03 |
| - pomel0 (QUIT: Ping timeout: 246 seconds) (~pomel0@user/pomel0) | 21:22 | |
| + pomel0 (~pomel0@user/pomel0) | 21:23 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@98.19.131.193) | 21:29 | |
| ch | rkbin/tools/boot_merger: ELF 64-bit LSB executable, x86-64 | 21:42 |
| ch | got to be kidding me, rockchip | 21:42 |
| - LainIwakura (QUIT: Quit: Client closed) (~LainIwaku@user/LainIwakura) | 21:47 | |
| - antti (QUIT: Read error: Connection reset by peer) (~antti@user/antti) | 21:51 | |
| + antti (~antti@user/antti) | 21:52 | |
| + robin_ (~robin@user/terpri) | 21:55 | |
| - robin (QUIT: Read error: Connection reset by peer) (~robin@user/terpri) | 21:55 | |
| minute | ch: lol! box64 to the rescue. but i think uboot has a tool for that? | 21:59 |
| - aloo_shu (QUIT: Ping timeout: 256 seconds) (~aloo_shu@85.51.16.22) | 22:01 | |
| + aloo_shu (~aloo_shu@85.51.16.22) | 22:02 | |
| ch | yeah supposedly so. i think i can ignore it when not booting from emmc, lets see | 22:04 |
| josch | good evening | 22:09 |
| josch | re "18:19 <@minute> hmm, how to find out on a booted system..." -- sudo reform-check :) | 22:09 |
| minute | https://lists.denx.de/pipermail/u-boot/2025-January/577641.html | 22:10 |
| minute | josch: right! | 22:10 |
| ch | how do i boot from sd card again? | 22:11 |
| josch | re "18:42 < ch> maybe it'd be good to make the uboot version 'better" -- yes, i asked minute about that here: https://community.mnt.re/t/uboot-mystery-value-up-to-date-but-not-using-it/3938/2 | 22:13 |
| josch | ch: by having boot.scr or extlinux on the first partition on your sd-card. u-boot (no matter where it comes from) will prefer sd-card over emmc if both have a boot.scr or extlunx.conf | 22:14 |
| ch | sorry, how do i boot u-boot from sd | 22:14 |
| - gustav25 (QUIT: Quit: Quit) (~gustav@c-78-82-53-228.bbcust.telenor.se) | 22:15 | |
| minute | ch: you need to erase the first few MB from emmc | 22:16 |
| ch | ugh :D | 22:16 |
| minute | ch: but don't kill your partition table i guess :D | 22:16 |
| minute | i guess you can flash /dev/zero with the size of uboot and the correct offset | 22:17 |
| josch | minute: what's the problem with edk2 re "i think we'll stop the edk2 effort for now"? | 22:18 |
| josch | minute: ah okay, should read the full backlog first before asking questions, sorry :) | 22:19 |
| josch | oh but didn't you have display output with edk2 and rk3588? or was that with hdmi output? | 22:19 |
| minute | josch: was hdmi only | 22:21 |
| minute | josch: which is a bit moot now that we primarily use dsi again | 22:21 |
| josch | indeed | 22:22 |
| josch | i only remembered your photos with edk2 and display (and i think windows booting?) but that was before dsi worked with rk3588 | 22:23 |
| josch | ch: the pcbview is interesting to me because the pocket reform motherboard schematic cannot be rendered by kicad on a311d | 22:28 |
| josch | but it does work in firefox on a311d with pcbview | 22:28 |
| ch | so maskrom might be using usb host 0 (not sure), but if it does, that goes to the tusb8041 | 22:29 |
| ch | which probably doesnt help me | 22:29 |
| minute | yooo barebox seems to have vop2 driver https://github.com/Danct12/barebox/blob/master/drivers/video/rockchip/rockchip_drm_vop2.c | 22:29 |
| minute | https://lore.pengutronix.de/barebox/20241028-vop2-rk3588-v1-0-a54bf7c28adc@pengutronix.de/ | 22:32 |
| josch | minute: do you agree to include the git tag in the EXTRAVERSION as i proposed in https://community.mnt.re/t/uboot-mystery-value-up-to-date-but-not-using-it/3938 to make the output of /proc/device-tree/chosen/u-boot,version more meaningful? | 22:32 |
| ch | it also needs to pin a tf-a and rkbin commit though, otherwise its still 'who knows' | 22:33 |
| minute | josch: sure | 22:33 |
| minute | and yeah ch is right | 22:34 |
| minute | hmmm does barebox support more linux driver apis or how did they do that... | 22:34 |
| josch | minute, ch: that's what this MR does: https://source.mnt.re/reform/reform-rk3588-uboot/-/merge_requests/10 | 22:34 |
| minute | josch: oh ok! sorry, that slipped past me somehow | 22:36 |
| josch | no worries | 22:36 |
| josch | it needs testing, so it needs somebody to have time :) | 22:36 |
| josch | as i learned in the backlog, updating u-boot can have really weird consequences | 22:36 |
| minute | josch: it can have good ones :D | 22:36 |
| josch | :) | 22:37 |
| minute | now i really want to try barebox... but should probably do that during worktime, meh :D | 22:37 |
| ch | dd if=/dev/zero of=/dev/mmcblk0 bs=512 seek=64 count=1 | 22:37 |
| minute | https://www.barebox.org/demo/?graphic=0 | 22:37 |
| - pomel0 (QUIT: Ping timeout: 256 seconds) (~pomel0@user/pomel0) | 22:37 | |
| ch | yay/nay? :) | 22:37 |
| + pomel0 (~pomel0@user/pomel0) | 22:38 | |
| minute | ch: count=1 seems not much :D | 22:38 |
| ch | well i hope the rk rom looks for a signature and then doesnt find it? | 22:38 |
| ch | but maybe it checks for more bytes | 22:38 |
| + wickedshell (~wickedshe@c-174-56-77-83.hsd1.nm.comcast.net) | 22:38 | |
| ch | count=18428 is probably a full uboot | 22:39 |
| josch | ch: if you want to know for sure i can look it up in the shell history of my rk3588 reform -- want me to? | 22:40 |
| ch | if it feels vaguely right i'm ok | 22:40 |
| ch | :) | 22:40 |
| minute | ok barebox seems way cooler than i expected :0 we might've missed out | 22:41 |
| josch | from memory, what i did was to wipe the memory starting at byte 32768 and stopping such that i don't touch anything starting byte 16777216 | 22:41 |
| ch | 2025.10+zeha-gfd976ff3a233-dirty ok then | 22:42 |
| ch | no extra cpu features have appeared ;-) | 22:42 |
| minute | ch: do you mean beyond aes? | 22:43 |
| ch | Features : fp asimd evtstrm aes pmull sha1 sha2 crc32 atomics fphp asimdhp cpuid asimdrdm lrcpc dcpop asimddp | 22:44 |
| ch | i didnt expect that to change now from the 2024.10 May uboot | 22:44 |
| josch | minute: if aes (and others) can disappear because of the u-boot version, is this maybe something that reform-check should investigate (because it's very cheap to perform the test) | 22:44 |
| ch | mainline u-boot just boots is i think the result from my experiment | 22:44 |
| minute | josch: it's because of the tf-a version :3 | 22:46 |
| minute | josch: but yeah, totally! | 22:46 |
| minute | ch: ah ok! but which board did you select? | 22:46 |
| ch | so, mainline + pocket-reform defconfig + dts + mnt-reform-series-rk3588.c | 22:48 |
| ch | but it doesnt need the collabora uboot | 22:48 |
| ch | i diffed the defconfig against the jaguar_rk3558 but not sure that was necesary | 22:49 |
| minute | ch: ok very good to know! | 22:49 |
| ch | (and tf-a mainline and rkbin head) | 22:49 |
| minute | ch: https://github.com/barebox/barebox?tab=readme-ov-file#features | 22:50 |
| minute | make ARCH=arm rockchip_v8_defconfig | 22:53 |
| minute | this is mini outdated but similar https://www.barebox.org/doc/latest/boards/rockchip.html#id1 | 22:53 |
| minute | make menuconfig has Radxa ROCK5 | 22:53 |
| ch | Receiving objects: 57% (194325/340920), 64.80 MiB | 237.00 KiB/s | 22:54 |
| ch | lalala | 22:54 |
| minute | ch: echo 1 > /proc/sys/net/ipv6/conf/all/disable_ipv6 | 22:54 |
| minute | :D | 22:54 |
| ch | hah | 22:55 |
| ch | nah | 22:55 |
| minute | (i'm half joking, but this is what i usually have to do at home and at work because of ipv6 issues) | 22:55 |
| minute | that bizarrely slow down especially gitlab | 22:55 |
| - andreas-e (QUIT: Quit: Leaving) (~Andreas@2a02-8434-b6a3-e901-facc-8e87-8e54-890e.rev.sfr.net) | 23:01 | |
| + chomwitt (~chomwitt@2a02:85f:9a5f:900:1ac0:4dff:fedb:a3f1) | 23:04 | |
| minute | > ** barebox-rock5b.img skipped due to missing firmware ** | 23:04 |
| minute | cp rk3588_ddr_lp4_2112MHz_lp5_2400MHz_v1.19.bin arch/arm/boards/radxa-rock5/sdram-init.bin | 23:08 |
| minute | cp rk3588_bl32_v1.20.bin firmware/rk3588-bl32.bin | 23:08 |
| minute | cp rk3588_bl31_v1.51.elf firmware/rk3588-bl21.bin | 23:08 |
| minute | eh, -bl31.bin | 23:09 |
| minute | images built: | 23:09 |
| minute | barebox-rock5b.img | 23:09 |
| minute | ch: if you wanna try (i currently don't have my emmc erased) https://minute.social/secret-room/barebox-rock5b.img | 23:10 |
| minute | (haven't added our dts etc) | 23:10 |
| ch | should that have hdmi? i dont have a serial atm | 23:14 |
| minute | ch: i think so... | 23:14 |
| minute | ch: but it's a gamble of course | 23:14 |
| + reform12606 (~heiko@2a11:fb80:3cc:f401:e981:d025:1a25:befe) | 23:15 | |
| ch | gamble is fine | 23:23 |
| ch | doesnt obviously boot | 23:27 |
| minute | meh :D | 23:27 |
| ch | would need a serial... really got to find a solution for that | 23:28 |
| minute | > barebox@Radxa ROCK 5B:/ | 23:30 |
| minute | whoo | 23:30 |
| minute | (i found a sd card adapter) | 23:30 |
| ch | whoo | 23:30 |
| minute | and erased emmc on my classic reform | 23:30 |
| minute | (which has convenient built-in usb to serial) | 23:30 |
| ch | josch: i found a usb type c hub with PD passthroguh | 23:31 |
| ch | ... it exposes new interesting bugs, i guess | 23:32 |
| ch | it works if you plug in the power first and then the hub into the pocket | 23:32 |
| ch | other way round not so much | 23:32 |
| minute | heh, barebox has built-in vi | 23:33 |
| ch | heh | 23:33 |
| ch | thats quite useful | 23:33 |
| ch | not very 'bare'! | 23:33 |
| minute | not sure how to exit that vi though | 23:33 |
| ch | ordered an intel be200ngw to compare | 23:34 |
| minute | haha one has to enter ZQ | 23:34 |
| ch | uh | 23:34 |
| minute | not a genuine vi | 23:34 |
| minute | ch: nice @ be200 | 23:34 |
| minute | barebox@Radxa ROCK 5B:/ mount /dev/mmc1.1 /mnt | 23:35 |
| minute | ext4 ext40: EXT2 rev 1, inode_size 256, descriptor size 32 | 23:35 |
| minute | i like this | 23:35 |
| gordon2 | i have aes btw, is that expected? | 23:35 |
| gordon2 | U-Boot 2024.10-g424c714eb247-dirty (May 06 2025 - 15:46:29 +0000) | 23:35 |
| minute | it has tab completion and everything | 23:35 |
| minute | gordon2: it's good in any case :D | 23:35 |
| ch | gordon2: yeah thats the good uboot | 23:35 |
| gordon2 | minute: well, it is better if it is unexpected :D | 23:35 |
| gordon2 | science and stuff | 23:36 |
| minute | ok how to enable display on barebox... | 23:36 |
| minute | ah maybe i didn't compile fbconsole in | 23:36 |
| minute | i did tho... | 23:37 |
| minute | haha | 23:39 |
| minute | barebox has dts/src/arm64/rockchip/rk3588-mnt-reform2.dts | 23:39 |
| ch | lol | 23:39 |
| minute | but that uses hdmi0, we need hdmi1 | 23:40 |
| gordon2 | so i missed the big news it seems, s2idle works on 6.17 doesn't it? | 23:42 |
| minute | it also has arch/arm/boards/mnt-reform/board.c | 23:44 |
| minute | gordon2: on 6.16.12 but only on my monolithic test kernel so far | 23:44 |
| gordon2 | can i have config pls? | 23:44 |
| gordon2 | (no rush) | 23:45 |
| minute | gordon2: source.mnt.re/reform/reform-kernel-dev | 23:45 |
| gordon2 | nice, thanks | 23:45 |
| minute | grimmware: prepare.sh and then build.sh should build it | 23:46 |
| minute | sorry, wrong highlight | 23:46 |
| minute | gordon2: i meant you :D | 23:46 |
| gordon2 | ok, i'll give diff a careful study and will try to port it to my monolithic kernel | 23:47 |
| gordon2 | i think last time i tried s2idle it just refused to wake up even after i set wakup source | 23:47 |
| - pomel0 (QUIT: Ping timeout: 240 seconds) (~pomel0@user/pomel0) | 23:53 | |
| + pomel0 (~pomel0@user/pomel0) | 23:59 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!