| 11 Apr 2024 |
|  Matthias Görgens joined the room. | 01:30:58 |
|  sm2n joined the room. | 11:25:55 |
| 14 Apr 2024 |
 damentz | heftig: pushed the upcoming NT sync feature set to 6.8/main, looks like upstream is interested and already took part of the series for kernel 6.10 | 16:35:28 |
| damentz | getting in will be important when proton experimental and other wine derivatives implement the necessary plumbing to use it | 16:35:45 |
 heftig | ok, thanks | 16:35:48 |
| damentz | and btw, with it being a driver exposed at /dev/ntsync, I think this one is worth enabling with =y rather than as a module | 16:36:44 |
| damentz | not sure how it would work otherwise, maybe a udev rule or users need to add a line in /etc/modprobe.d? | 16:37:05 |
| heftig | damentz: modules.devname contains the list of static nodes | 22:32:56 |
| heftig | that gets translated to /run/tmpfiles.d/static-nodes.conf | 22:33:19 |
| damentz | interesting, I see, systemd-tmpfiles-setup-dev-early uses that file to make the static-nodes.conf file | 22:40:29 |
| heftig | that creates the devices; if a device gets accessed that the kernel doesn't have the module loaded for, it will execute kmod | 22:42:14 |
| heftig | e.g. if you cat /dev/vhci the hci_vhci module gets loaded | 22:42:29 |
| damentz | wow, I see | 22:42:50 |
| damentz | well, at least for ntsync, I don't see a devnames line, that's how the modules.devname file gets populated | 22:43:08 |
| damentz | for example for btrfs-control, there's a line that contains: MODULE_ALIAS("devname:btrfs-control"); | 22:43:48 |
| damentz | then searching for all devname references, there's a 1-for-1 in the file | 22:43:58 |
| damentz | alright then, I"ll switch it back to a module for Liquorix then | 22:44:57 |
| damentz | * alright then, I"ll switch it back to a module for Liquorix | 22:45:01 |
| damentz | although, the C file is just 26kB in size, so whether it's static or not probably doesn't matter | 22:45:40 |
| heftig | guess it's missing. report a bug? | 22:46:31 |
| damentz | probably, there's other issues | 22:46:59 |
| damentz | the file gets mounted in /dev as 0600 for root only | 22:47:11 |
| damentz | that means regular user space software can't use it I'm assuming | 22:47:30 |
| heftig | oh, the static-nodes stuff only works for non-dynamic minors, of course | 22:49:06 |
| damentz | I think I'll wait to give feedback for when regular generally available software gets the ntsync plumbing added, the main dev is focused on getting the rest of their patches into mainline | 22:49:33 |
| damentz | the first set is slated for 6.10 and none of the Wine-GE or Proton have it yet, getting it in early though means people can start testing it ASAP | 22:49:56 |
| damentz | also for people who have been testing it, there's not been any complaints activating it, so maybe it's all fine | 22:50:13 |
| damentz | heftig: ok ya, I confirmed, the ntsync device isn't loaded in /dev by default, and querying it just returns "No such file or directory" | 23:52:14 |
| damentz | but like I said, this is still super early, so I'll wait for a bug report on whether it should be static or not, or if the code needs to be changed | 23:52:33 |
| 25 Apr 2024 |
|  sjs joined the room. | 05:27:44 |
