| - mtm (QUIT: Ping timeout: 256 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 01:02 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 02:49 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 03:08 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 03:08 | |
| - cobra (QUIT: Ping timeout: 256 seconds) (~cobra@user/Cobra) | 03:18 | |
| + cobra (~cobra@user/Cobra) | 03:49 | |
| - nsc (QUIT: Ping timeout: 260 seconds) (~nicolas@140-96-142-46.pool.kielnet.net) | 03:49 | |
| + nsc (~nicolas@186-97-142-46.pool.kielnet.net) | 03:51 | |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 04:13 | |
| + cobra (~cobra@user/Cobra) | 04:18 | |
| - GNUmoon (QUIT: Ping timeout: 260 seconds) (~GNUmoon@gateway/tor-sasl/gnumoon) | 08:05 | |
| + digitalrayne_ (~digitalra@vps-446f4f39.vps.ovh.ca) | 08:06 | |
| - digitalrayne (QUIT: Read error: Connection reset by peer) (~digitalra@vps-446f4f39.vps.ovh.ca) | 08:06 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 08:18 | |
| * digitalrayne_ -> digitalrayne | 08:22 | |
| + yankcrime (~nick@gw.tetromino.io) | 09:51 | |
| + mjw (~mjw@gnu.wildebeest.org) | 11:43 | |
| minute | we are looking for some feedback/spelling/grammar/factual error reports on the reform handbook 2nd edition: https://source.mnt.re/reform/reform-handbook/-/jobs/3539/artifacts/file/src/build/reform-handbook.pdf | 12:30 |
|---|---|---|
| josch | and i'll be looking for feedback/bugreports/feature-requests on the new system images that will probably be ready in 1.5 hours as I now have hopefully fixed the last bug in the reform-system-imx8mp branch of reform-system-images :) | 12:46 |
| josch | minute: how do you prefer to receive feedback? I don't see an open MR on the gitlab where I could comment directly on the diff. | 12:47 |
| minute | josch: good point, i will open an MR | 12:49 |
| minute | btw we will archive the old version and keep it online, it's just not 100% clear yet how (probably it will just get another URL) | 12:50 |
| minute | (possibly we will archive only the PDF version) | 12:50 |
| josch | minute: if you create a tag in the reform-handbook, then gitlab will keep its artifacts | 12:50 |
| josch | or create a release and attach the PDF to the release in gitlab? | 12:50 |
| minute | josch: https://source.mnt.re/reform/reform-handbook/-/merge_requests/15 | 12:52 |
| minute | josch: good idea @ tag+release | 12:53 |
| - mtm (QUIT: Ping timeout: 264 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 13:02 | |
| josch | minute: now it's your turn to test the 3 new system images: https://source.mnt.re/reform/reform-system-image/-/jobs/3737/artifacts/browse | 14:35 |
| josch | I'm very satisfied with the changes to reform-system-image from this branch. The tests I can now put into place together with the machine readable machine configuration files already have let me spot and fix multiple bugs in our flash-kernel patches, reform-tools, reform-system-image the Debian linux packaging (see my commits over the weekend). Now let me know if anything still boots or not. :D | 14:36 |
| - mjw (QUIT: Ping timeout: 240 seconds) (~mjw@gnu.wildebeest.org) | 15:04 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 15:08 | |
| + mjw (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 15:24 | |
| jjbliss | minute: I just took a quick look at the handbook. I can sign up to gitlab and leave some comments on the merge request, but I don't have an account yet. There are some formatting issues on pages 53 and 54, text continues off the page to the right. Also, the bluetooth section talks about the tray icon, but is in Chapter 6, Linux console basics. Everything else in the section avoids talking about graphical desktops, should th | 15:29 |
| minute | josch: awesome | 16:06 |
| minute | jjbliss: thank you. unfortunately your last sentence was cut off after "should th" | 16:07 |
| minute | haha when working on this i feel like working on a mainframe in the 70s via a teletype https://source.mnt.re/reform/reform-system-image/-/jobs/3738 | 16:13 |
| minute | "maybe this night my code might compile? we'll see tomorrow" | 16:13 |
| josch | minute: for testing purposes you can reduce the package set to be installed to those packages that are necessary for only a CLI environment. This will speed everything up tremendously. | 16:17 |
| josch | minute: another speed-up would be to make the CI runner arm64. Even my a311d is faster than the CI machine. :) | 16:17 |
| josch | speaking of the last point: consider runnig this locally. Your rk3588 will be even faster than my a311d. :) | 16:18 |
| minute | hmm > I: running --customize-hook in shell: sh -c '{ printf "[Service]\nExecStart=\nExecStart=-/sbin/agetty --autologin root --noclear %%I \$TERM\n" } > "$1"/etc/systemd/system/getty@tty1.service.d/override.conf' | 16:18 |
| minute | josch: any idea what's the issue here? | 16:19 |
| minute | it says > exec: 1: Syntax error: end of file unexpected (expecting "}") | 16:19 |
| josch | minute: you are missing a semicolon before the closing brace | 16:19 |
| minute | ohhh | 16:19 |
| josch | since you only have one command you don't even need the braces | 16:20 |
| minute | ah | 16:20 |
| minute | i thought there was some arcane reason for them | 16:20 |
| josch | maybe you copied from the other line in mkimage.sh where i used multiple echo commands | 16:20 |
| minute | yes | 16:21 |
| minute | ok, pushed again. now to your system images | 16:21 |
| josch | \o/ | 16:23 |
| josch | minute: if nothing boots, note that i just copied the u-boot offsets from the existing platforms and that might be wrong for imx8mp | 16:24 |
| minute | jjbliss: i approved your account and gave you reporter access to handbook project | 16:24 |
| jjbliss | Thanks! | 16:24 |
| minute | josch: oh @ imx8mp. i think that's a different offset | 16:24 |
| minute | anyway, gonna try pocket a311d first and then probably pocket imx8mp | 16:26 |
| minute | josch: big success, pocket a311d image boots _and_ the console is rotated correctly! incredible lol | 17:12 |
| josch | \o/ | 17:21 |
| josch | minute: you only added the console rotation bits into u-boot, right? | 17:21 |
| minute | josch: yep | 17:55 |
| josch | nice, this means that the recent changes that make boot.scr respect ${bootargs} work as intended | 17:56 |
| josch | it means that these lines in flash-kernel/ubootenv.d/00reform2_ubootenv are indeed only needed for old platforms and nothing needs to be added for new boards: https://source.mnt.re/reform/reform-tools/-/blob/staging/flash-kernel/ubootenv.d/00reform2_ubootenv?ref_type=heads#L6 | 17:58 |
| - hairu (QUIT: Remote host closed the connection) (m-uotkmd@user/hairu) | 18:05 | |
| + hairu (m-uotkmd@user/hairu) | 18:24 | |
| minute | finally repaired our trusty old sigma r19 3d printer... feels good that it's finally working again | 18:24 |
| + mark_ (~mjw@gnu.wildebeest.org) | 18:37 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 18:58 | |
| * mjw -> Guest9687 | 19:01 | |
| - Guest9687 (QUIT: Killed (silver.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:01 | |
| * mark_ -> mjw | 19:01 | |
| + Guest9687 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:01 | |
| minute | josch: debian has # CONFIG_FB_SM750 is not set | 19:15 |
| minute | josch: i wanted to try an SM750 based mPCIe graphics card with LS1028A | 19:15 |
| josch | minute: can you check with your amd64 Debian box? Mine also doesn't have CONFIG_FB_SM750 set so this doesn't seem to be arm64 specific. | 19:22 |
| josch | it seems that Debian linux sets CONFIG_FB_SM750=m for loongson (mips) | 19:22 |
| minute | josch: yep, also unset | 19:23 |
| josch | if your SM750 card works with LS1028A i guess we should enable it :) | 19:24 |
| josch | from the Debian linux commit message: "Some new Loongson servers are using Aspeed BMC, which has an GPU. Some other Loongson servers are using SM750 GPU instead of AMD's." | 19:25 |
| minute | josch: yeah, well now i guess i have to build a custom kernel/module first | 19:27 |
| minute | doing apt-install linux-source on ls1028a box | 19:28 |
| minute | also fun fact: ls1028a happily does PCIe hotplug | 19:28 |
| minute | :D | 19:28 |
| minute | (i didn't get any output from the kernel with the sm750 plugged in because i think PCIe/PCI has some VGA magic going on) | 19:29 |
| minute | (i.e. i _think_ vga character generator takes over) | 19:29 |
| minute | but i could hot plug the mpcie card (carefully) and then rescan the bus | 19:29 |
| minute | now that we have a 25m ethernet cable through the mnt studio i _could_ hook up ls1028a as another build server for certain things | 19:33 |
| minute | for example for reform-system-image...... | 19:33 |
| minute | > root@layerscape:~/kernel/linux-source-6.7# make M=drivers/staging/sm750fb | 19:35 |
| minute | > [ 142.583240] sm750fb 0001:01:00.0: enabling device (0000 -> 0002) | 19:40 |
| minute | > [ 142.608209] video memory phyAddr = 8040000000, size = 67108864 bytes | 19:40 |
| minute | system locks up though | 19:41 |
| minute | > sm750fb 0001:01:00.0: no specific g_option. | 19:46 |
| minute | driver still locks up the system though, sigh | 19:48 |
| minute | so it's probably not trivially compatible | 19:49 |
| minute | josch: omg https://lore.kernel.org/linux-devicetree/87shdfet90.fsf@linux.intel.com/ | 20:07 |
| minute | > So far the Layerscape SoC (such as LS1088A) has enabled this feature and I | 20:08 |
| minute | > have tested it. Once we add dma-coherent on DTS without this Patch, dwc3 | 20:08 |
| minute | > will fail on device enumeration as below: | 20:08 |
| minute | > [ 15.124031] xhci-hcd xhci-hcd.0.auto: Error while assigning device slot ID | 20:08 |
| minute | > [ 15.130912] xhci-hcd xhci-hcd.0.auto: Max number of devices this xHCI host supports is 127. | 20:08 |
| minute | > [ 15.139268] usb usb1-port1: couldn't allocate usb_device | 20:08 |
| minute | this is pretty much exactly what happens on ls1028a | 20:08 |
| minute | looks like this patch indeed fixes ls1028a usb issues | 20:30 |
| minute | josch: ok finally this built https://source.mnt.re/reform/reform-system-image/-/merge_requests/92 | 21:13 |
| minute | gonna test it now-ish | 21:13 |
| josch | hah, since i run my ls1028a headless and communicate over uart or via ssh i never saw usb issues i guess XD | 21:23 |
| minute | yeah sometimes the keyboard doesn't come on, which is annoying | 21:23 |
| josch | bummer that you still didn't get graphics to work | 21:23 |
| minute | yeah that card is just weird maybe | 21:23 |
| josch | i was in contact with Jasbir from http://jas-hacks.blogspot.com/ but didn't get much farther yet | 21:24 |
| minute | ohh | 21:24 |
| josch | maybe should send a frienly ping for my last message | 21:24 |
| josch | minute: i take it you also didn't yet try out the different rk3588 kernels i built for you? | 21:25 |
| josch | (just asking to make sure i didn't miss it) | 21:25 |
| minute | josch: ah no damn | 21:26 |
| minute | i have a bit much going on with system images | 21:26 |
| josch | i don't know what you could be talking about ;) | 21:26 |
| minute | hehe | 21:26 |
| josch | i'm currently learning rust packaging -- it's fun :) | 21:27 |
| minute | oh nice | 21:27 |
| josch | the debian rust team has an incredibly newbie friendly set of scripts to package their stuff -- i haven't seen the likes of it in any other language team in debian | 21:28 |
| josch | everything is very well documented and automated, really nice :) | 21:28 |
| minute | really cool! | 21:30 |
| minute | ok good news: my system image logged root in automatically | 21:31 |
| minute | bad news: it didn't start reform-setup | 21:31 |
| minute | i guess things in /etc/profile.d/ have to be marked executable? | 21:33 |
| josch | nothing in my local flatpak.sh is marked as executable | 21:33 |
| josch | how did that paste flatpak.sh? | 21:34 |
| josch | i meant to say /etc/profile.d/ | 21:34 |
| minute | :D | 21:34 |
| minute | yeah that's true, so that's not the reason then | 21:34 |
| minute | aha | 21:36 |
| minute | it has to be called something.sh | 21:36 |
| josch | funny, the bash man page doesn't document that fact it seems | 21:40 |
| minute | but yeah, the whole thing is ~almost~ working... it doesn't have the up-to-date cleanup script yet etc. will take care of that tomorrow | 21:50 |
| minute | also it looks like my updates to waybar aren't in debian-packages yet | 21:50 |
| minute | (mnt logo, 3 dots for the tray menu, ...) | 21:50 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 21:51 | |
| josch | minute: mnt logo and waybar settings would go into reform-tools not reform-debian-packages | 22:04 |
| * bluerise_ -> bluerise | 22:38 | |
| josch | minute: is there a good reason for the mnt-reform-setup-wizard to require sxd-document and sxd-xpath? Both projects have not seen any updates in the past five years and depend on another project (peresil) which saw its last release eight years ago (and last git commit four years ago). There seem to be a few much more popular and better maintained rust xml libraries out there. | 22:39 |
| josch | I'm hesitant about packaging something that seems to be unmaintained. | 22:39 |
| minute | josch: sorry yeah i meant reform-tools, but those are built by reform-debian-packages right | 22:41 |
| josch | yes | 22:41 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!