Commit Graph
33 Commits
Author SHA1 Message Date
modeco80 8369de53ba qemu: log QEMU stderr as logger messages
Mostly for nicity, but also to make debugging start errors an actual possibility.
2024-06-23 02:56:17 -04:00
modeco80 2cc2c6ddf2 cvmts: "fix" panics in cvm-rs 2024-06-23 02:55:05 -04:00
modeco80 0df56cb5a4 qemu: cleanup/fix reset 2024-06-23 02:40:13 -04:00
modeco80 286f3eec62 cvm-rs: Actually throw Error on guac decode errors
we were throwing String or something before..
2024-06-23 02:11:23 -04:00
modeco80 183b17194e chore(README): cargo is usually a default part of a complete Rust toolchain so just specify "A Rust toolchain" 2024-06-22 21:46:37 -04:00
modeco80 dda72cad91 cvmts: quth => auth 2024-06-22 21:27:28 -04:00
modeco80 09d41617ed prettier reformat for merge (and remove jpeg-turbo Again) 2024-06-22 21:26:49 -04:00
modeco80 b8ed177885 cvm-rs: merge guac and jpeg libs together into one
doesn't really need to be two seperate libraries. also preperation for other funnies

the build script has been replaced with a much saner justfile which uses much saner "yarn workspace" invocations instead of blindly cding all over the place
2024-06-22 21:14:05 -04:00
modeco80 87a377a10f cvmts: replace jpeg-turbo native module with new rust module
This module also does threadpooling internally, so we don't need Piscina anymore (which I'm pretty sure was actually bottlenecking.)
2024-06-20 03:20:56 -04:00
modeco80 39521a4b1d misc stuff from production
(also refactors qemu a bit)
2024-06-19 23:30:29 -04:00
modeco80 97de887518 cvmts: Actually unref ipdata on disconnect 2024-06-19 18:26:27 -04:00
modeco80 fa23aa7432 cvmts: nope I have to fix it properly 2024-06-19 18:20:41 -04:00
modeco80 e798ff5c86 bandaid fix time 2024-06-19 18:16:16 -04:00
modeco80 0d34bb1c8e cvmts/qemu: support snapshots properly 2024-06-19 18:03:10 -04:00
modeco80 b485e7f689 cvmts: reimplement connection limit using ipdata 2024-06-19 17:56:55 -04:00
modeco80 ba8743f461 guac-rs: remove commented dead code
gits a SCM.
2024-06-19 02:34:38 -04:00
modeco80 b342d4874f remove license cargo key 2024-06-19 02:14:57 -04:00
modeco80 eefde464b4 why is yarn so picky 2024-06-19 01:49:12 -04:00
modeco80 d9ee611bc5 guac-rs: reformat im stupid 2024-06-19 01:37:17 -04:00
modeco80 4e50106585 cvmts: replace guacamole decoder with a node native module written in rust 2024-06-19 01:36:07 -04:00
modeco80 173ee8149f auth: Make more resilant to backend failures 2024-05-26 16:33:35 -04:00
modeco80 e184bfb085 qemu: fix qmp disconnection semi properly
this is actually something i need to push to crusttest as well, because
this will affect it as well, though not as badly because it will only break certain buttons
2024-05-22 17:56:04 -04:00
modeco80andyellows111 2e05504e4a hastily hand merge yellowcode vote patch thing
Co-Authored-By: yellows111 <ice_ice@email.com>
2024-05-01 08:08:43 -04:00
modeco80 e03bf57ede .. ok, i guess one node buffer can't be moved but the other magically can. 2024-04-24 04:38:47 -04:00
modeco80 a904f26961 abstract jpeg encoding away from "WSServer"
Additionally make thumbnail encoding threadpooled as well, just so it (probably) doesn't block as much.
2024-04-24 04:18:05 -04:00
modeco80 ddae307874 chore: reformat all code with prettier 2024-04-24 03:50:17 -04:00
modeco80 bcbf7db8d9 misc patches done to get everything to play ball
also adds editorconfig
2024-04-24 03:41:32 -04:00
modeco80 db97a62046 move jpeg encoding to a worker thread pool
this also switches cvmts back to building with tsc, mostly because
I couldn't get parcel's worker interop to work at all.
2024-04-23 19:43:42 -04:00
modeco80 59d5331b68 replace dirname with a hack that should behave like before
the '@ts-expect-error' is to silence tsc just in case, since this function does in fact exist during bundling. techinically the real solution would be to declare it but /shrug
2024-04-23 10:42:36 -04:00
modeco80 cb297e15c4 Giant refactoring (or at least the start)
In short:
- cvmts is now bundled/built via parcel and inside of a npm/yarn workspace with multiple nodejs projects
- cvmts now uses the crusttest QEMU management and RFB library (or a fork, if you so prefer).
- cvmts does NOT use node-canvas anymore, instead we opt for the same route crusttest took and just encode jpegs ourselves from the RFB provoded framebuffer via jpeg-turbo. this means funnily enough sharp is back for more for thumbnails, but actually seems to WORK this time
- IPData is now managed in a very similar way to the original cvm 1.2 implementation where a central manager and reference count exist. tbh it wouldn't be that hard to implement multinode either, but for now, I'm not going to take much time on doing that.

this refactor is still incomplete. please do not treat it as generally available while it's not on the default branch. if you want to use it (and report bugs or send fixes) feel free to, but while it may "just work" in certain situations it may be very broken in others.

(yes, I know windows support is partially totaled by this; it's something that can and will be fixed)
2024-04-23 09:57:02 -04:00
modeco80 820ef05832 chore: Remove reference to "fs" dummy package
Not needed.

This will end up landing again anyways in a major refactoring job,
but for now I'm making the change here because it's a one line change.
2024-04-23 07:14:13 -04:00
modeco80 28dddfc363 fix registered users' ability to bypass automute
ditto... oops.
2024-04-22 08:40:25 -04:00
modeco80 1637614c74 fix QMP deadlock 2023-02-07 22:38:41 -05:00