| - robin (QUIT: Remote host closed the connection) (~robin@user/terpri) | 00:05 | |
| + robin (~robin@user/terpri) | 00:06 | |
| + murphnj (~murph@user/murphnj) | 00:11 | |
| - NanoCodeBug (QUIT: Ping timeout: 260 seconds) (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 00:17 | |
| + NanoCodeBug (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 00:17 | |
| minute | josch: i think it's fine as it is... i just need(ed) to get back into things. i'm thinking about tooling that could help with moving patch stacks between reform-debian-packages and a matching monolithic linux kernel build for development though. | 00:45 |
|---|---|---|
| josch | in debian there is gbp-pq which translates patch stacks to git branches but i don't know if it can be shoehorned into our setup somehow... | 00:46 |
| josch | we can also completely drop the concept of directories with patches and switch to git branches | 00:47 |
| minute | josch: directories are fine | 00:54 |
| josch | okay, just reach out to me if you want this machinery to be refactored into something else | 00:55 |
| minute | josch: the problems i have are more like, the debian modular kernel is too unwieldly for the actual kernel development. so i do the development on another kernel with quick build/test cycle and then bring that work back to reform-debian-packages. hmm, maybe if there was a quick way to see if all patches apply, without all that build setup, that would help | 00:56 |
| josch | but that should only be a problem if you use a different git state of the kernel for your monolithic builds compared to the version in unstable, no? | 00:58 |
| josch | because the source is the same, only your config differs | 00:58 |
| minute | basically: 1. unpack debian kernel source 2. apply patch files in the same unforgiving way 3. maybe copy in my monolithic config, run an Image build to catch compile errors | 00:59 |
| minute | (step 2 because i would then quickly see the errors i only see now after 15 minutes or so in CI) | 01:00 |
| josch | you apply the patches using "quilt push -a"? | 01:00 |
| minute | josch: perhaps, i'm not familiar with these tools because i rarely directly had to use them | 01:02 |
| josch | yeah, i wouldn't call quilt the preferred tool of choice to do patch management in 2024 (i think git is much superior) but it's the tool which is able to handle directories of patches and is used in debian, so that's what the current method is based on -- this can definitely be changed of course | 01:04 |
| minute | i see! so i could either learn that or use find/a recursive glob + patch | 01:06 |
| - mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 01:06 | |
| minute | and it might be useful to version control my mono linux config for reform stuff that i'm carrying along in various places on disk | 01:07 |
| minute | (why do i work with a mono linux during development? 1. quick and simple to deploy to DUT in 1 second or so, it's just 1 file 2. greatly reduced build time) | 01:08 |
| + mjw (~mjw@gnu.wildebeest.org) | 01:08 | |
| minute | (no need to build an initramfs also) | 01:08 |
| josch | yes, i don't think you should have to wait for half an hour and then do debian-things only to try out one small change... | 01:09 |
| josch | minute: about quilt: its operation is quite simple. It uses the file debian/patches/series and applies all patches in it one after another | 01:10 |
| josch | if one patch does not apply, it stops and lets you investigate the situation | 01:11 |
| josch | it keeps track of its state (where it left off) by writing some data into the ./.pc directory | 01:11 |
| josch | but if all patches apply, it should be something similar to do doing: | 01:11 |
| josch | cat debian/patches/series | grep -v '^#' | while read patch; do patch -p1 < $patch; done | 01:12 |
| josch | (untested but i hope you get the idea) | 01:12 |
| + _justin_kelly8 (~justinkel@user/justin-kelly/x-6011154) | 01:14 | |
| - _justin_kelly (QUIT: Ping timeout: 252 seconds) (~justinkel@user/justin-kelly/x-6011154) | 01:16 | |
| * _justin_kelly8 -> _justin_kelly | 01:16 | |
| + reformer (~reformer@softboy.mntmn.com) | 01:22 | |
| + reformer (~reformer@softboy.mntmn.com) | 01:33 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.0.75) | 01:53 | |
| + Ar|stote|is (~linx@149.210.0.75) | 01:54 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 02:00 | |
| - mtm (QUIT: Ping timeout: 265 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:04 | |
| - Gooberpatrol_66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 02:05 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:06 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 02:06 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-72-235.tukw.qwest.net) | 02:20 | |
| + reformer (~reformer@softboy.mntmn.com) | 02:23 | |
| + colinsane (~colinunin@97-113-72-235.tukw.qwest.net) | 02:24 | |
| + reformer (~reformer@softboy.mntmn.com) | 02:29 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 02:38 | |
| + jacobk (~quassel@64.189.202.173) | 02:58 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@64.189.202.173) | 03:06 | |
| - amospalla (QUIT: Ping timeout: 255 seconds) (~jordi@user/amospalla) | 03:16 | |
| + jacobk (~quassel@utdpat241106.utdallas.edu) | 03:22 | |
| - paperManu (QUIT: Ping timeout: 252 seconds) (~paperManu@198.16.214.40) | 03:23 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@utdpat241106.utdallas.edu) | 03:27 | |
| kop316 | is the USB PD charger from crowd supply known to work with the pocket reform? | 03:28 |
| - nsc (QUIT: Ping timeout: 260 seconds) (~nicolas@i5C74DEDC.versanet.de) | 03:29 | |
| + nsc (~nicolas@98-97-142-46.pool.kielnet.net) | 03:31 | |
| + amospalla (~jordi@user/amospalla) | 03:31 | |
| - Ar|stote|is (QUIT: Ping timeout: 276 seconds) (~linx@149.210.0.75) | 03:37 | |
| + Ar|stote|is (~linx@149.210.1.251) | 03:47 | |
| - NanoCodeBug (QUIT: Read error: Connection reset by peer) (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 04:03 | |
| + jacobk (~quassel@syn-076-184-114-053.res.spectrum.com) | 04:15 | |
| - CandidCadaver (QUIT: Ping timeout: 265 seconds) (~aloo_shu@85.51.17.35) | 04:24 | |
| + CandidCadaver (~aloo_shu@85.51.17.35) | 04:34 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@syn-076-184-114-053.res.spectrum.com) | 04:54 | |
| - sts-q (QUIT: Ping timeout: 272 seconds) (~sts-q@89.58.73.0) | 04:57 | |
| + sts-q (~sts-q@89.58.73.209) | 04:58 | |
| + NanoCodeBug (~NanoCodeB@c-73-35-191-67.hsd1.wa.comcast.net) | 04:58 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 06:00 | |
| - Gooberpatrol_66 (QUIT: Ping timeout: 248 seconds) (~Gooberpat@user/gooberpatrol66) | 06:00 | |
| + jaume (~user@user/jaume) | 07:29 | |
| + chomwitt (~chomwitt@2a02:587:7a07:9500:1ac0:4dff:fedb:a3f1) | 07:31 | |
| ch | kop316: with the current sysctl firmware, people have posted positive results; before, it didnt wor | 07:32 |
| ch | k | 07:32 |
| - jaume (QUIT: Remote host closed the connection) (~user@user/jaume) | 07:51 | |
| ryukazou | So after de-andoizing and re-andoizing, the screw hole will only got bigger like 0.02mm to 0.03mm. | 08:49 |
| + jacobk (~quassel@utdpat242016.utdallas.edu) | 09:18 | |
| hramrach | I do have some experience with managing kernel patch stacks in quilt format | 09:35 |
| hramrach | The thing is that i very rarely use quilt | 09:35 |
| hramrach | there is rapidquilt that not only can apply 20k patches in a few secnds but also most of the time gives detailded information why it does not apply https://github.com/openSUSE/rapidquilt which make it possible to change a line or two of the patch in place without going through a lot of extra steps | 09:38 |
| hramrach | with the quilt series in git it's possible to compare the patch stack between the different revisions of the kernel, and see the differences | 09:40 |
| hramrach | also it's possible to copy the original patch stack over what you have in git, and see the different from the original with git diff | 09:40 |
| hramrach | and aply the patches in git with something like gbp-pq to do manual rebase and also compare both the result and the patchset in git with git diff/range-diff | 09:42 |
| hramrach | basically what tracking quilt patchset in git gives you a meta-history that makes it possible to edit the history created by gbp-pq when the patches are applied, and track changes to it | 09:43 |
| + minute (~mntirc@softboy.mntmn.com) | 09:54 | |
| * ChanServ changed mode (+o, minute) | 09:54 | |
| minute | weird, sometimes i come back and weechat has disconnected and lost all channels | 09:56 |
| - chomwitt (QUIT: Ping timeout: 246 seconds) (~chomwitt@2a02:587:7a07:9500:1ac0:4dff:fedb:a3f1) | 10:32 | |
| + chomwitt (~chomwitt@2a02:587:7a02:eb00:1ac0:4dff:fedb:a3f1) | 10:34 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@utdpat242016.utdallas.edu) | 10:34 | |
| - Gooberpatrol66 (QUIT: Ping timeout: 276 seconds) (~Gooberpat@user/gooberpatrol66) | 10:43 | |
| josch | hramrach: if it were up to you, what would you change with how the patch management is done right now? | 11:18 |
| josch | i had a look into looking gbp-pq but that requires that the linux sources are checked into git which is not the case for the linux source package in debian | 11:18 |
| josch | so it would require an extra step to produce a synthetic git repository after having run "make -f debian/rules orig" and potentially "make -f debian/rules debian/control-real" | 11:19 |
| hramrach | josch: which linux sources need to be in git? If you quilt patches that are supposed to apply on top of Linux 6.10 you can check out v6.10 tag in a given git repository and quiltimport the patches. Is that not what gbp-pq does? | 11:22 |
| josch | maybe there are some magic options i don't know about but usually i'd use gbp-pq in a debian source package git. In case of src:linux, the debian linux maintainers only ship the ./debian directory in git. Maybe gbp-pq can be told to also operate on a upstream git tree without a ./debian dir but i never did that. | 11:23 |
| hramrach | I used to use this https://github.com/openSUSE/kernel-source/blob/master/scripts/series2git before the patch number grew to the point it's impractical and it's done by cron in advance | 11:27 |
| hramrach | the directory structure is different for the debian kernels so it won't work out of the box but should not be that difficult to adapt | 11:28 |
| josch | hrm... okay. I'd like to avoid adding another NIH solution. | 11:28 |
| hramrach | it's the solution for the case the source package only has the patches, not the upstream source | 11:31 |
| hramrach | Looks like gpb-pq is indeed incapable of using a separate repository for the sources, you would have to merge the upstream sources with the debian packaging | 11:33 |
| josch | yes, thereby creating a synthetic git repo just for the sake of being able to use gbp-pq | 11:34 |
| hramrach | that or craeting a new script/updating gbp-pq to be able to work with the existing debian kernel sources | 11:35 |
| hramrach | I wonder how the kernel packagers do this. I suspect they have such synthetic repo or rool for working with the serces somewhere | 11:35 |
| josch | i just asked over in #debian-kernel how the debian kernel packaging team usually handles their patch stack | 11:35 |
| + mjw (~mjw@gnu.wildebeest.org) | 11:36 | |
| - Kooda (QUIT: Quit: WeeChat 4.2.2) (~kooda@2a00:5884:8301:0:a5f7:ccb:ffd9:4d2a) | 11:43 | |
| + Kooda (~kooda@natsu.upyum.com) | 11:48 | |
| - chomwitt (QUIT: Ping timeout: 272 seconds) (~chomwitt@2a02:587:7a02:eb00:1ac0:4dff:fedb:a3f1) | 11:51 | |
| - mjw (QUIT: Ping timeout: 276 seconds) (~mjw@gnu.wildebeest.org) | 12:10 | |
| hramrach | I think generally having an easy way from quilt to git with applied patches is key for managing these patchstacks. The exact way it's done may vary | 12:28 |
| + paperManu (~paperManu@198.16.214.40) | 12:44 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 12:57 | |
| * Guest1898 -> mjw | 13:12 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:03 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:05 | |
| minute | ok, back to working on patches | 14:28 |
| ryukazou | Will the screw hole got bigger like 0.02mm to 0.03mm greatly affect the tidiness of m2 screw? | 14:31 |
| minute | ryukazou: no idea tbh! | 14:32 |
| - sts-q (QUIT: Ping timeout: 245 seconds) (~sts-q@89.58.73.209) | 15:12 | |
| + sts-q (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 15:19 | |
| - [tj] (PART: !!unknown attribute: msg!!) (sid609767@id-609767.lymington.irccloud.com) | 15:40 | |
| - Ar|stote|is (QUIT: Ping timeout: 265 seconds) (~linx@149.210.1.251) | 15:43 | |
| + Ar|stote|is (~linx@149.210.0.207) | 15:48 | |
| minute | josch: fuzz is forbidden, yes? https://source.mnt.re/reform/reform-debian-packages/-/jobs/6052#L1854 | 15:59 |
| minute | josch: ok so i notice that with my local build-linux.sh i actually get very quickly to the patch-checking stage. i just forgot this | 16:13 |
| + wielaard (~mjw@gnu.wildebeest.org) | 16:19 | |
| - sts-q (QUIT: Remote host closed the connection) (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 16:25 | |
| + sts-q (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 16:48 | |
| - wielaard (QUIT: Ping timeout: 276 seconds) (~mjw@gnu.wildebeest.org) | 17:04 | |
| + chomwitt (~chomwitt@2a02:587:7a02:eb00:1ac0:4dff:fedb:a3f1) | 17:29 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 17:31 | |
| josch | minute: yes, fuzz is forbidden | 18:22 |
| minute | alright, i'm close to finishing this patch operation... | 18:23 |
| josch | minute: maybe your build-linux.sh can somehow be added to the reform-debian-packages git? Last time I did this often was during imx8mq times but with an extra reform available soon, i might be in need of this as well :) | 18:23 |
| minute | josch: oh sure! btw what is WORKDIR for exactly? | 18:24 |
| minute | i had to set it but i couldn't find a directory with that name afterwards | 18:25 |
| josch | minute: WORKDIR is populated in common.sh with a $(mktemp --directory --tmpdir="$(pwd)") | 18:25 |
| minute | oh uh | 18:25 |
| josch | and at the end of linux/build.sh it just rm -r it | 18:25 |
| minute | http://dump.mntmn.com/build-linux.sh.txt | 18:26 |
| minute | i have a patchstack that applies now, but we'll have to see if it builds... | 18:26 |
| josch | minute: instead of build-linux.sh.txt, you should be able to use the command at the bottom of the README | 18:27 |
| josch | i feel bad thinking how much time you are sinking into just fighting with quilt -- if if you tell me how you'd prefer doing it, we can surely set something up | 18:28 |
| - CandidCadaver (QUIT: Ping timeout: 265 seconds) (~aloo_shu@85.51.17.35) | 18:44 | |
| - chomwitt (QUIT: Ping timeout: 252 seconds) (~chomwitt@2a02:587:7a02:eb00:1ac0:4dff:fedb:a3f1) | 18:51 | |
| + CandidCadaver (~aloo_shu@85.51.17.35) | 18:52 | |
| hramrach | when it's something done only occasionally writing down how to do it is half of success | 19:02 |
| josch | well... i do it on every kernel version bump :) | 19:02 |
| hramrach | and that happens a few times a year, maybe every 2 months or so? | 19:03 |
| hramrach | I use large number of scripts, at least a dozen, to massage patches. But I use them at least once a week so I can remember what they are, and what they are for. | 19:05 |
| + wielaard (~mjw@gnu.wildebeest.org) | 19:39 | |
| minute | ah, now we're missing drm/display/drm_hdmi_state_helper.h | 19:43 |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 19:47 | |
| minute | i need the linux patch that introduced that in 6.11... | 19:47 |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 19:47 | |
| + gustav28 (~gustav@c-5834524e.019-141-67626730.bbcust.telenor.se) | 20:02 | |
| minute | ok my time is over | 20:07 |
| minute | and i will take 4 days off now | 20:07 |
| minute | and afterwards can try to backport that required feature from 6.11 to 6.10 (drm_hdmi_state_helper) or we can maybe go directly to 6.11 | 20:07 |
| hramrach | enjoy your vacation :) | 20:45 |
| grimmware | Yeah, hope you have a chill time :) | 20:52 |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 20:53 | |
| - Gooberpatrol66 (QUIT: Ping timeout: 252 seconds) (~Gooberpat@user/gooberpatrol66) | 20:53 | |
| - mjw (QUIT: Killed (erbium.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 20:59 | |
| * wielaard -> mjw | 20:59 | |
| + Guest8314 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 21:00 | |
| + chomwitt (~chomwitt@2a02:587:7a02:eb00:1ac0:4dff:fedb:a3f1) | 21:10 | |
| - sts-q (PART: !!unknown attribute: msg!!) (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 21:37 | |
| + sts-q (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 21:51 | |
| - sts-q (QUIT: Remote host closed the connection) (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 21:57 | |
| + sts-q (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 21:59 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-5834524e.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 22:16 | |
| - sts-q (QUIT: Remote host closed the connection) (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 22:38 | |
| Twodisbetter | minute: have a good 4 day. I'll be trying to do the same | 22:43 |
| - Gooberpatrol_66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 22:46 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 22:47 | |
| + sts-q (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 22:57 | |
| - sts-q (QUIT: Remote host closed the connection) (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 23:00 | |
| + sts-q (~sts-q@2a11:fb80:3f1:cb01:3453:af1b:2a:ed74) | 23:03 | |
| - cow321 (QUIT: Ping timeout: 255 seconds) (~deflated8@user/meow/deflated8837) | 23:07 | |
| + cow321 (~deflated8@user/meow/deflated8837) | 23:25 | |
| - chartreuse (QUIT: Ping timeout: 252 seconds) (~chartreus@S0106908d78501d1d.cg.shawcable.net) | 23:40 | |
| + chartreuse (~chartreus@S0106908d78501d1d.cg.shawcable.net) | 23:41 | |
| - eery (QUIT: Ping timeout: 252 seconds) (~eery@192.150.253.237) | 23:43 | |
| + iank_ (~iank@fsf/staff/iank) | 23:43 | |
| - iank (QUIT: Ping timeout: 252 seconds) (~iank@fsf/staff/iank) | 23:44 | |
| + eery (~eery@192.150.253.237) | 23:49 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!