| 20 Dec 2025 |
 ghpzin | First one seems okay, pkgsMusl.xorg-server builds with bundle of local musl fixes.
Not sure how to do second one even for nixosTest.
I assume you need to feed pkgsMusl into everything at least xorg related.
"simple" approach of doing:
xorg-server = pkgsMusl.callPackage ../by-name/xo/xorg-server/package.nix {};
does not work properly:
xserver-wrapper[877]: (EE) Failed to load /nix/store/g89a4axc26znr6dwl5idw1wlc9c17j7s-xf86-input-libinput-1.5.0/lib/xorg/modules/input/libinput_drv.so: Error relocating /nix/store/38581mfz8czr1apmmnxc4iwzrrj6djfm-libinput-1.29.2/lib/libinput.so.10: __stpcpy_chk: symbol not found
| 12:33:10 |
 K900 | @emilazy:matrix.org do you want to test xorg on Darwin | 15:01:12 |
 emily | uh | 15:01:34 |
| emily | isn't that already broken? | 15:01:36 |
| emily | not sure I'll have time and it's so low-impact, but if you have a specific ref for me to build I can try | 15:01:55 |
| K900 | No | 15:01:57 |
| K900 | I have a PR | 15:01:57 |
| K900 | That makes it unironically almost normal | 15:02:01 |
| K900 | I don't need build I need run | 15:02:07 |
| K900 | I have build | 15:02:12 |
| K900 | https://github.com/NixOS/nixpkgs/pull/472679 | 15:03:00 |
| K900 | Look at it | 15:03:00 |
| K900 | It's beautiful | 15:03:00 |
| K900 | And cursed | 15:03:00 |
| emily | ok I will take a look. what kind of testing do you need? GLX? | 15:04:40 |
| emily | do we really need all those patches… | 15:04:56 |
| K900 | The patches are staying I don't trust them | 15:07:19 |
| K900 | I don't know if we actually need them | 15:07:26 |
| K900 | But yes I basically just want to see if it runs and has GLX | 15:07:36 |
| K900 | If it does that's probably enough | 15:08:01 |
| emily | so these patches are from https://github.com/XQuartz/xorg-server/commits/XQuartz-2.7.11. the latest patches are from https://github.com/XQuartz/xorg-server/commits/XQuartz-2.8.6_beta2. | 15:08:46 |
| emily | the first one we're fetching only applies for universal builds which we don't do, and only touches a Makefile that the Meson build presumably doesn't use. the second is a revert that isn't present on the more recent tags | 15:09:03 |
| emily | at least some of the patches in that latest tag are already upstream, e.g. "rootless: Ensure gResizeDeathPix is stored in locally-managed memory rather than re-using the implementation's backing store" went upstream in 2023 | 15:09:51 |
| K900 | If you can delete more of those please do | 15:09:53 |
| K900 | And sorry for sniping you into this | 15:10:21 |
| emily | well I was about to apply my time-box by asking if you can try cherry-picking the ones from that latest tag on top of our Xorg version to see which ones actually still do anything :P | 15:11:17 |
| K900 | But maybe if we can get it more normal we can not touch it again | 15:11:28 |
| emily | https://github.com/K900/nixpkgs/blob/d075ccf56d5c304ee6f7933d2eabb095bac5cd5c/pkgs/by-name/xo/xorg-server/darwin/find-cpp.patch is kinda gross, couldn't we just upstream using the Meson function to get the compiler | 15:14:57 |
| emily | I can try running it, but would be nice to get the patches sorted first | 15:15:28 |
| K900 | I couldn't get it working with cc.preprocess for whatever reason so I gave up | 15:16:19 |
