| sevan | for some reason dkms process ends up building the zfs module twice | 00:24 |
|---|---|---|
| sevan | first as part of the initial process to generate the new module then again on the installation attempt. | 00:25 |
| sevan | haven't investigated but perhaps the package maintainer has set it so it will always be built from source for whatever operation? (never make use of anything prepared) | 00:26 |
| sevan | anyway, new kernel installed, rebooting. | 00:27 |
| josch | fingers crossed! | 00:27 |
| josch | sevan: there was a reason but i forgot -- i think it was the same reason as to why the initramfs is built twice | 00:27 |
| sevan | josch: if the system has booted the display hasn't initialised | 00:30 |
| sevan | josch: re initramfs, ah ok, I see that too | 00:30 |
| josch | sevan: could you connect with a uart adapter to see what is printed via serial? | 00:30 |
| sevan | ye | 00:31 |
| sevan | p | 00:31 |
| sevan | need to find my adapters | 00:32 |
| + chomwitt (~chomwitt@ppp-2-84-146-212.home.otenet.gr) | 00:45 | |
| josch | sevan: do you have an estimate? i wonder whether i should go to bed or wait a bit longer :) | 00:59 |
| sevan | josch: sorry | 01:00 |
| sevan | josch: starting kernel & it just hangs there | 01:01 |
| sevan | josch: http://paste.debian.net/1321205/ | 01:01 |
| josch | okay, that's the same failure mode as on the pocket | 01:02 |
| sevan | ok | 01:02 |
| josch | in some way this relieves me a bit because it would've been weird if only the pocket was affected | 01:02 |
| sevan | ack | 01:03 |
| sevan | so keep eyes peeled here for a new kernel announcement? | 01:03 |
| josch | not really | 01:04 |
| josch | want to try out 6.9? | 01:04 |
| sevan | :D | 01:04 |
| josch | the failure happened on the *same* kernel version | 01:04 |
| sevan | sure | 01:04 |
| josch | the only difference was, when it was built | 01:04 |
| josch | so it must be something in the build environment that makes the difference | 01:04 |
| josch | i tried comparing the vmlinuz that worked with the one that did not | 01:05 |
| josch | maybe notably, the working vmlinuz is 31M large and the failing one only 26M | 01:05 |
| josch | so something in the build environment in the last 3 weeks broke boot and affected the kernel size | 01:05 |
| sevan | erm, that's a big difference | 01:06 |
| josch | yup | 01:06 |
| josch | but it could also be a red herring | 01:06 |
| josch | and maybe the size difference is weird but unrelated | 01:06 |
| sevan | maybe we can try tomorrow, there's going to be a 40+ minute turnaround because of the dkms thingy. | 01:06 |
| sevan | with the 6.9 kernel. ^ | 01:07 |
| josch | sevan: sure, you find the 6.9 kernel as a CI artifact | 01:07 |
| sevan | ok | 01:07 |
| sevan | btw I see the size difference | 01:08 |
| sevan | -rw-r--r-- 1 root root 30M Mar 9 23:22 vmlinuz-6.7.9-reform2-arm64 | 01:08 |
| sevan | -rw-r--r-- 1 root root 26M Jun 5 11:42 vmlinuz-6.8.12-mnt-reform-arm64 | 01:08 |
| josch | those are different versions but yes | 01:08 |
| sevan | we have a single kernel for reform2 & pocket? | 01:08 |
| josch | we have a single kernel for everything, yes | 01:08 |
| sevan | ok, must be the magic of dedupe :) | 01:09 |
| jfred | This sort of thing is why I'm happy that Guix tightly controls the build environment for packages so much... I still need to try out that Guix install on my Reform | 01:16 |
| josch | jfred: oh i can also look up which packages in the build environment different between the successful and failing build, thanks to reproducible builds | 01:17 |
| josch | using the buildinfo file, the debian linux kernel is bit-by-bit reproducible | 01:18 |
| josch | so i have no problem reproducing the failing and the successful kernel | 01:18 |
| josch | but that still does not tell me what is going on | 01:18 |
| ch | binutils-arm-linux-gnueabihf 2.42-4 -> 2.42.50.20240618-1 ? | 01:18 |
| josch | maybe that | 01:19 |
| josch | or maybe debhelper | 01:19 |
| josch | i have to try out some variations tomorrow | 01:19 |
| ch | https://tracker.debian.org/news/1537738/accepted-debhelper-1316-source-into-unstable/ seems unlikely | 01:22 |
| josch | indeed | 01:22 |
| josch | on the CI i triggered a run using testing as the base which has the same binutils but different debhelper -- trying out that gives us a few more data points and package versions to compare | 01:23 |
| josch | i'll do a local build over night with binutils-arm-linux-gnueabihf 2.42-4 | 01:24 |
| josch | i guess you are right that this is the most likely culprit | 01:24 |
| josch | especially given the file size difference as well... | 01:25 |
| ch | sizes in debians 6.8.12-1 vs 6.9.2-1~exp1 also look suspicious | 01:25 |
| josch | ch: in what way do you mean? | 01:25 |
| ch | 6.9.2-1~exp1 vmlinuz is 4M smaller | 01:25 |
| josch | right, but if that would be unbootable there would definitely be a bug open about that already | 01:26 |
| ch | its in experimental | 01:26 |
| josch | ch: i followed the development of 6.9 in salsa including successful test reports | 01:28 |
| josch | unless of course they built this on bookworm which also is unlikely :) | 01:28 |
| josch | anyways, must test more :) | 01:28 |
| ch | but did you follow the binutils development? | 01:28 |
| ch | 6.9.2-1~exp1 looks pretty much broken to me | 01:28 |
| josch | oh? | 01:29 |
| josch | does somebody know someone with an x86 box who would be willing to test 6.9.2-1~exp1? :) | 01:30 |
| ch | will do in a sec | 01:30 |
| josch | uh nice | 01:30 |
| josch | sleep delayed :D | 01:30 |
| jfred | josch: Ahh right! Makes sense re: reproducible builds | 01:30 |
| ch | arm64 was easier | 01:31 |
| josch | ch: what arm64 system do you have that uses the vanilla kernel? | 01:31 |
| - chomwitt (QUIT: Ping timeout: 256 seconds) (~chomwitt@ppp-2-84-146-212.home.otenet.gr) | 01:31 | |
| ch | utm on mac | 01:31 |
| ch | cheating, i know | 01:32 |
| josch | heh :) | 01:32 |
| josch | emulating x86 boot on my a311d reform is so slow that things time out left right and center during boot XD | 01:32 |
| josch | (using qemu) | 01:32 |
| ch | yeah dealing with the physical x86 hardware is also slower and annoying :) | 01:33 |
| ch | amd64 boots | 01:35 |
| ch | wow, the broken 6.9 managed to crash qemu/utm | 01:36 |
| josch | oh? | 01:36 |
| josch | i'm now trying this locally for arm64 using debvm as well | 01:37 |
| josch | i didn't even consider that it could be broken for everybody and not just for us... | 01:38 |
| josch | ch: same results for arm64 under qemu over here | 01:40 |
| + chomwitt (~chomwitt@2a02:587:7a1d:6300:1ac0:4dff:fedb:a3f1) | 01:44 | |
| ch | josch: are you going to file a bug against binutils + linux-image-6.9-arm64? | 01:45 |
| ch | not sure i can add any useful detail | 01:45 |
| josch | ch: we do not know yet whether it is binutils though | 01:46 |
| josch | i already wrote quite a bit of text in #debian-kernel over on OFTC | 01:46 |
| josch | if i do not get any responses there, i'll file a bug, yes | 01:46 |
| ch | hm, new binutils is already in testing | 01:47 |
| ch | no urgency then :D | 01:47 |
| josch | exactly | 01:49 |
| josch | ch: thanks a lot for your input -- without your insistence i would not've evaluated whether vanilla debian kernel exhibits the same issue or not | 01:49 |
| josch | but now i'm happy and can go to sleep with a good feeling :) | 01:49 |
| ch | sleep well! | 01:49 |
| josch | thank you! :) | 01:50 |
| + hjcfd (~hjcs@syn-098-015-248-249.res.spectrum.com) | 01:53 | |
| josch | ch: why must i only learn now that you are zeha!!! XD | 01:53 |
| ch | haha | 01:53 |
| josch | (yes, the nick name is funny) | 01:53 |
| ch | too obvious! | 01:53 |
| - ZylonMaster (QUIT: Ping timeout: 268 seconds) (~hjcs@syn-098-015-248-249.res.spectrum.com) | 01:56 | |
| - hjcfd (QUIT: Quit: Leaving) (~hjcs@syn-098-015-248-249.res.spectrum.com) | 02:03 | |
| - mtm (QUIT: Ping timeout: 264 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 02:03 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 02:06 | |
| - chomwitt (QUIT: Ping timeout: 268 seconds) (~chomwitt@2a02:587:7a1d:6300:1ac0:4dff:fedb:a3f1) | 03:18 | |
| - nsc (QUIT: Ping timeout: 264 seconds) (~nicolas@i5C74DF86.versanet.de) | 03:23 | |
| + nsc (~nicolas@i5C74DC29.versanet.de) | 03:24 | |
| - eschaton (QUIT: Remote host closed the connection) (eschaton@2600:3c01::f03c:91ff:fefd:5d92) | 03:51 | |
| + eschaton (eschaton@2600:3c01::f03c:91ff:fefd:5d92) | 03:51 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 03:59 | |
| - mtm (QUIT: Ping timeout: 255 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 05:56 | |
| - cobra (QUIT: Ping timeout: 256 seconds) (~cobra@user/Cobra) | 06:19 | |
| + cobra (~cobra@user/Cobra) | 06:28 | |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 07:04 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 07:20 | |
| - cobra (QUIT: Ping timeout: 268 seconds) (~cobra@user/Cobra) | 07:21 | |
| + chomwitt (~chomwitt@2a02:587:7a1d:6300:1ac0:4dff:fedb:a3f1) | 08:02 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 08:26 | |
| - chomwitt (QUIT: Ping timeout: 260 seconds) (~chomwitt@2a02:587:7a1d:6300:1ac0:4dff:fedb:a3f1) | 08:35 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 08:53 | |
| josch | ch: i found your theory that it could be binutils very convincing and prepared a MR which chooses old-enough binutils (and gcc-13 as new gcc does not work with old binutils) from snapshot.d.o: https://source.mnt.re/reform/reform-debian-packages/-/merge_requests/43 | 09:18 |
| josch | if somebody who wants to help out and for whom their reform is not their main machine wants to test this: feel free to try and install the linux-image that falls out of this pipeline | 09:19 |
| - megasquid (QUIT: Ping timeout: 268 seconds) (~megasquid@user/megasquid) | 09:27 | |
| ch | josch: i'm guessing the reform kernel should also boot in qemu? if so then this could also be a good first test? | 10:19 |
| josch | yes, it's on my todo list but it's family time today :) | 10:33 |
| josch | ch: and yes, we have a test script using tcl-exect which tests the kernel in qemu :) | 10:34 |
| - nybble (QUIT: Ping timeout: 246 seconds) (~sprang@user/csprng) | 10:36 | |
| ch | :) | 10:37 |
| josch | uuuuh would you look at that! | 10:38 |
| josch | with old binutils, the kernel size is 31 MB | 10:39 |
| josch | step aside daughter, daddy needs some time with qemu | 10:39 |
| ch | not gonna tell you what to do, but i know time with children is precious | 10:40 |
| josch | i blame Lukas for having made this laptop ;) | 10:41 |
| josch | yes! it boots in qemu! | 11:03 |
| josch | now, just to be sure, let me create a system image with the non-working kernel packages -- just to verify my test setup | 11:04 |
| josch | confirmed!! | 11:43 |
| josch | ch: thank you!! | 11:43 |
| ch | nice | 11:43 |
| + mjw (~mjw@gnu.wildebeest.org) | 11:45 | |
| josch | it's #1074111 now | 11:53 |
| josch | minute: the boot problem has a workaround now in this merge request: https://source.mnt.re/reform/reform-debian-packages/-/merge_requests/43 | 11:55 |
| josch | this makes it work in qemu and the kernel size is good again, so i'm quite confident that this also fixes it on the real hardware | 11:55 |
| josch | still, it would be nice if somebody could try this out for real using this package: https://source.mnt.re/reform/reform-debian-packages/-/jobs/4686/artifacts/file/repo/pool/main/l/linux/linux-image-6.8.12-mnt-reform-arm64_6.8.12-1+reform20240623T071545Z_arm64.deb | 11:56 |
| josch | thanks in advance! | 11:56 |
| josch | now finally out and afk for the rest of the day | 11:56 |
| josch | _o/ | 11:56 |
| + cobra (~cobra@user/Cobra) | 12:18 | |
| + reform25278 (~user@business-90-187-186-49.pool2.vodafone-ip.de) | 13:18 | |
| - sevan (QUIT: Ping timeout: 272 seconds) (~sevan@user/venture37) | 13:18 | |
| reform25278 | https://pastebin.com/wGwiDCpn | 13:18 |
| reform25278 | rebooting now | 13:18 |
| - reform25278 (QUIT: Remote host closed the connection) (~user@business-90-187-186-49.pool2.vodafone-ip.de) | 13:18 | |
| pandora | Worked | 13:20 |
| pandora | Buuut | 13:36 |
| pandora | I don’t have network anymore | 13:36 |
| ch | like, no eth interface or it doesn't work? which nic? | 13:39 |
| ch | i wonder what 'Starting background process to update reform-qcacld2' actually does | 13:40 |
| pandora | No eth | 13:41 |
| pandora | Only the loopback and end1 . Guess the latter is the SIM card | 13:41 |
| ch | and your eth should be the qcacld2 thing? | 13:41 |
| pandora | Idk tbh | 13:42 |
| pandora | I got the reform | 13:42 |
| pandora | A few days back | 13:42 |
| pandora | I will upload a dmesg in a minute | 13:44 |
| ch | oO( my sane resistance to buy a pocket reform is waning every minute ) | 13:49 |
| pandora | can't find a usb-c stick rn :D | 13:50 |
| ch | maybe other q: your eth is supposed to be wired or wifi? | 13:51 |
| ch | (qcacld2 is the wifi driver) | 13:51 |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 13:51 | |
| pandora | wifi | 13:53 |
| pandora | but can't see either of the interfaces | 13:53 |
| pandora | here is the dmesg log | 13:55 |
| pandora | https://pastebin.com/tXe75R0F | 13:55 |
| ch | minute: is there a way of getting a reform purple + sleeve with an rk3588 cpumodule without getting the imx8 module and then swapping it? | 14:01 |
| - mtm (QUIT: Ping timeout: 240 seconds) (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 14:03 | |
| + akira (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 14:27 | |
| Zaba | end1 is what eth0 gets renamed to | 15:11 |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 15:21 | |
| + akira (~akira@ipb218fb58.dynamic.kabel-deutschland.de) | 15:22 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@ipb218fb58.dynamic.kabel-deutschland.de) | 15:24 | |
| + akira (~akira@ipb218fb58.dynamic.kabel-deutschland.de) | 15:24 | |
| - akira (QUIT: Ping timeout: 256 seconds) (~akira@ipb218fb58.dynamic.kabel-deutschland.de) | 15:31 | |
| pandora | yeah u r right ... but i can't test that because i don't have an industrial ethernet cable / adapter | 15:31 |
| + akira (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 15:32 | |
| + sevan (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 15:39 | |
| - sevan (QUIT: Changing host) (~sevan@2001:470:1f1d:1d6:5a55:caff:fe24:ed4) | 15:39 | |
| + sevan (~sevan@user/venture37) | 15:39 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 15:44 | |
| + akira (~akira@178.24.251.58) | 15:45 | |
| - akira (QUIT: Ping timeout: 264 seconds) (~akira@178.24.251.58) | 15:56 | |
| + akira (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 16:00 | |
| + mtm (~mtm@c-71-228-84-213.hsd1.fl.comcast.net) | 16:09 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 16:15 | |
| + akira (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 16:16 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a0e:73eb:1c49:80f0:acf9:2709) | 16:22 | |
| + akira (~akira@37.4.230.225) | 16:22 | |
| + chomwitt (~chomwitt@2a02:587:7a27:9000:1ac0:4dff:fedb:a3f1) | 16:31 | |
| minute | pandora: end1 is ethernet | 17:20 |
| minute | qcacld2 is the wifi driver | 17:20 |
| noam | ...I don't think I've ever used the ... key to the right of ctrl on the reform, lol | 17:36 |
| noam | i think i semiregularly realize that :P | 17:36 |
| truck | I remapped it to ctrl, because it was in the wrong place for my fingers, and there was no other use for it | 17:41 |
| truck | I think it's there for folks who are used to shift lock being there (I've seen .de layouts with both caps and shift lock) | 17:42 |
| truck | and, of course, for keeping a nice orthargonal layout (: | 17:42 |
| pandora | @minute: interesting. I don’t a see any mentions of qcacld2 in dmesg | 17:59 |
| noam | literally no idea what it's for lol | 18:03 |
| + nybble (~sprang@user/csprng) | 18:57 | |
| - cobra (QUIT: Ping timeout: 256 seconds) (~cobra@user/Cobra) | 19:20 | |
| jfred | Something odd I've noticed... when I plug my ebook reader (Kobo, forget which model) into my Reform while it's charging, the Kobo's touchscreen goes unresponsive until I unplug the Reform from power. Weird grounding thing maybe? | 19:30 |
| pandora | Is it a 2 or 3 prong connector in the power plug? | 19:53 |
| pandora | I think there is a long forum post on this | 19:53 |
| pandora | Other laptops with 2 prong power supply behave similar. On some u can get a tingling feeling if the laptop is fully metal | 19:54 |
| pandora | https://community.mnt.re/t/psu-not-suitable-for-use-with-metal-housing-resolved/1713/3 | 19:56 |
| + cobra (~cobra@user/Cobra) | 20:07 | |
| noam | um | 20:26 |
| noam | should i be concerned if the keyboard light shut off out of nowhere? >_< | 20:27 |
| noam | everything else is totally fine lol | 20:27 |
| - akira (QUIT: Ping timeout: 252 seconds) (~akira@37.4.230.225) | 20:50 | |
| + akira (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 20:52 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 20:59 | |
| + akira (~akira@81.92.17.144) | 20:59 | |
| - akira (QUIT: Ping timeout: 268 seconds) (~akira@81.92.17.144) | 21:03 | |
| + akira (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 21:05 | |
| minute | pandora: take a look at /sbin/reform-hw-setup and search for qcacld2, then you'll see how that is loaded and probably not found in your case | 21:27 |
| pandora | it says insmod -f "/opt/reform/qcacld2/qcacld2-${KERNVER}.ko" | 21:37 |
| pandora | if i execute that manually the wifi starts to work | 21:39 |
| pandora | for some reason the "automatic" part broke after I installed | 21:39 |
| pandora | https://source.mnt.re/reform/reform-debian-packages/-/jobs/4686/artifacts/file/repo/pool/main/l/linux/linux-image-6.8.12-mnt-reform-arm64_6.8.12-1+reform20240623T071545Z_arm64.deb | 21:39 |
| pandora | i thik the issue is ${KERNVER} | 21:43 |
| pandora | it is different than what i find in /opt/reform/qcacld2/ | 21:44 |
| + reform12450 (~user@business-90-187-186-49.pool2.vodafone-ip.de) | 21:45 | |
| reform12450 | user@mnt:~$ dpkg-query --show --showformat '${Version}' "linux-image-$(uname -r)" | 21:45 |
| reform12450 | 6.8.12-1+reform20240623T071545Zuser@mnt:~$ ls /opt/reform-qcacld2/ | 21:45 |
| reform12450 | qcacld2-6.8.12-1+reform20240605T104210Z.ko | 21:45 |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 22:14 | |
| josch | pandora: /sbin/reform-hw-setup does not use set -e, so if parts of it fail, the script will not fail | 22:15 |
| josch | pandora: you might be interested in helping with improving https://source.mnt.re/reform/reform-tools/-/merge_requests/70 | 22:16 |
| josch | in your case it sounds like you are missing an upgrade to the latest qcacld2 package | 22:17 |
| josch | pandora: can you also show me the output of this: apt-cache policy reform-qcacld2-6.8.12-mnt-reform-arm64 | 22:21 |
| - reform12450 (QUIT: Remote host closed the connection) (~user@business-90-187-186-49.pool2.vodafone-ip.de) | 22:22 | |
| + reform12631 (~user@90.187.186.49) | 22:29 | |
| reform12631 | user@mnt:~$ apt-cache policy reform-qcacld2-6.8.12-mnt-reform-arm64 | 22:29 |
| reform12631 | reform-qcacld2-6.8.12-mnt-reform-arm64: | 22:29 |
| reform12631 | Installed: 20240605T114456Z | 22:29 |
| reform12631 | Candidate: 20240605T114456Z | 22:29 |
| reform12631 | Version table: | 22:29 |
| reform12631 | *** 20240605T114456Z 100 | 22:29 |
| josch | oh right, because your manually installed kernel does not come from any repo | 22:30 |
| pandora | yeah probably | 22:31 |
| josch | reform12631: the situation regarding reform-qcacld2 is quite unfortunate. I'd call the current situation a hack but i do not know how to improve on it | 22:31 |
| josch | the fundamental problem is that qcacld2 exists in the first place | 22:31 |
| josch | ideally, the driver would be in-tree and then we'd not be having this problem | 22:31 |
| pandora | wouldn't that be something that should be upstreamed in the kernel? | 22:36 |
| josch | yes | 22:36 |
| josch | if you want to work on that please do! | 22:36 |
| pandora | until then wouldn't it be better to just load the latest available version of qcacld2? | 22:37 |
| josch | pandora: you mean as a dkms module? | 22:37 |
| pandora | not but use something like this: | 22:39 |
| pandora | latest_driver=$(ls /opt/reform-qcacld2/qcacld2-*.ko | sort -V | tail -n 1) | 22:39 |
| pandora | insmod -f "$latest_driver" | 22:39 |
| pandora | to make sure the highest version is loaded | 22:40 |
| josch | pandora: and what do you do if the currently running kernel is not equal to the latest version you find in /opt/reform-qcacld2/? | 22:40 |
| pandora | u could try to load the exact match if there is any otherwise load the highest availale version | 22:40 |
| josch | yes, a fallback mechanism if there was no match is something we could do | 22:40 |
| josch | pandora: would you like to try writing a patch that does that? | 22:41 |
| josch | pandora: alternatively, would you like to receive and test a diff i send you? | 22:41 |
| josch | pandora: you could even make it such that it iterates through all /opt/reform-qcacld2/qcacld2-*.ko files in decending order until it finds one that succeeds to load, no? | 22:42 |
| pandora | yes i could do that. not sure about a "patch". normally at work i just create a branch. fix something and create a mr | 22:43 |
| pandora | but i do machine learning normally :D | 22:43 |
| josch | pandora: you can also submit a MR -- do you have an account at source.mnt.re? | 22:44 |
| pandora | lmc | 22:45 |
| pandora | its waiting for approval | 22:46 |
| josch | Lukas is currently on holidays, so we may need to go the patch route :) | 22:46 |
| josch | just run "git format-patch -1" once your commit is ready and put the resulting file into a pastebin or send it by mail | 22:46 |
| pandora | let me clone the repo and try :D | 22:48 |
| josch | pandora: i would be very happy (but this is not a requirement) if you also submit a fix for the pocket reform part of https://source.mnt.re/reform/reform-tools/-/merge_requests/70 as part of your patch | 22:49 |
| josch | by having your patch do "set -e" at the top or in the pocket branches | 22:49 |
| + akira (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 22:50 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 22:51 | |
| pandora | let me try to write a patch for it first :D | 22:51 |
| + akira (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 22:54 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 23:00 | |
| + akira (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 23:07 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 23:08 | |
| + akira (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 23:10 | |
| - akira (QUIT: Read error: Connection reset by peer) (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 23:10 | |
| - reform12631 (QUIT: Ping timeout: 246 seconds) (~user@90.187.186.49) | 23:11 | |
| - mjw (QUIT: Ping timeout: 264 seconds) (~mjw@gnu.wildebeest.org) | 23:18 | |
| + akira (~akira@2a01:599:a2b:1eff:c85d:fd50:5b36:418) | 23:24 | |
| + jacobk (~quassel@2600:1700:9e1e:7800:2c94:5a14:bbc3:b280) | 23:31 | |
| pandora | ok wrote a little patch and it seems to work on my reform | 23:38 |
| minute | pandora: what's your username on source? sorry for delays, but we only let in people after they ask here or via support email | 23:53 |
| minute | (to not have spam repos etc) | 23:53 |
| minute | (yes i'm on vacation but can click that button later) | 23:54 |
| pandora | pandora | 23:55 |
| pandora | @josch: send you a patch via mai | 23:55 |
| pandora | 23:55 | |
| pandora | "It works on my machine" | 23:56 |
| pandora | it contains the set -e too | 23:56 |
| pandora | tho i am not sure if that is really a good idea. if something fails it will just exit i guess without trying to do all the other stuff in the script | 23:58 |
| josch | pandora: with set -e you have to do correct error handling | 23:58 |
| minute | pandora: acc approved | 23:58 |
| pandora | merci | 23:58 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!