| + bibliocar (~EricShmar@195.82.99.14) | 00:08 | |
| chartreuse | Numbers after names -- When you need a source control for your source control :P | 00:18 |
|---|---|---|
| mntmn | :3 | 00:18 |
| bluerise | * The Arasan has a bugette whereby it may lose the content of successive | 00:42 |
| - bibliocar (QUIT: Read error: Connection reset by peer) (~EricShmar@195.82.99.14) | 00:42 | |
| bluerise | mhhh, baguette | 00:42 |
| sigrid | fixed that keys/mouse lockup | 00:46 |
| sigrid | it's not perfect, I think redrawing is supposed to run in a separate thread and the "done" callback of wayland should only signal that thread to redraw, but it fixed the freezes | 00:47 |
| sigrid | (the fix was in the drawterm logic) | 00:47 |
| mntmn | sigrid: oh! | 00:52 |
| sigrid | :) I would never think this could cause issues like that, but the logic basically dumps a new full hd xrgb32 image on each redraw, which I guess isn't exactly the best way on low-resource platforms | 00:54 |
| chartreuse | Guess you were spamming it at 60fps and it was drawing at slightly less than that? | 00:58 |
| chartreuse | Where the callbacks started building up | 00:59 |
| mntmn | oh yeah that probably consumes a bunch of memory bandwidth also | 00:59 |
| chartreuse | Ideally for low resources you'd only send the parts that have changed since the last frame, but not sure how that works on wayland | 01:00 |
| chartreuse | That's how you'd do it in X | 01:00 |
| chartreuse | Doesn't wayland maintain a framebuffer for each application to copy into? | 01:00 |
| sigrid | I am not sure | 01:03 |
| sigrid | first time I wrote any wayland code was yesterday | 01:04 |
| - mtm (QUIT: Ping timeout: 240 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 01:04 | |
| sigrid | anyway, apparently there was _another_ issue, with redrawing happening without asking wayland for "done" callback | 01:04 |
| sigrid | with that fixed it performs even better | 01:04 |
| chartreuse | Reducing the data sent would improve at least the memory performance, if you somehow maintain which parts of the screen are dirty and need redrawing | 01:06 |
| kfx | I guess this explains why I didn't run into trouble running drawterm under xwayland | 01:07 |
| - erlehmann (QUIT: Quit: Just say no, then the virus can not enter your body without your consent.) (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 01:11 | |
| sigrid | chartreuse: yeah, that's what I'm fixing now. it was always updating everything | 01:11 |
| sigrid | kfx: yep | 01:11 |
| - Christoph_ (QUIT: Remote host closed the connection) (~Christoph@p4fe73f2f.dip0.t-ipconnect.de) | 01:28 | |
| - chartreuse (QUIT: Ping timeout: 240 seconds) (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 02:37 | |
| + chartreuse (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 02:46 | |
| - chartreuse (QUIT: Client Quit) (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 02:47 | |
| + chartreuse (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 02:49 | |
| - Guest8449 (QUIT: Ping timeout: 240 seconds) (~nicolas@i5C7440A3.versanet.de) | 03:07 | |
| + nsc (~nicolas@i5C74409D.versanet.de) | 03:09 | |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 03:10 | |
| * nsc -> Guest5577 | 03:10 | |
| - chartreuse (QUIT: Remote host closed the connection) (~chartreus@S0106f0f249dfd9c3.cg.shawcable.net) | 04:47 | |
| - sts-q (QUIT: Ping timeout: 256 seconds) (~sts-q@212.53.219.148) | 04:50 | |
| + bibliocar (~g@195.82.99.14) | 04:50 | |
| bibliocar | wayland has a damage/damage_buffer for parts that need to be redrawn, but I have no idea how that all works. | 04:52 |
| + sts-q (~sts-q@91.200.108.229) | 05:02 | |
| bibliocar | I've been thinking about stuff like scrollbars and resizing windows. It's not my subject of expertise, though. But, I've been wondering about, like, if it's best to recalculate the dimensions of the bar, or have a picture of the scroll slider, and source over composite it over the image of the slider. | 05:04 |
| bibliocar | scroll bar, even, but also very offtopic | 05:08 |
| - bibliocar (QUIT: Quit: GrumpyChat v. 1.0.0. Such client. WOW. Much quit. https://github.com/grumpy-irc/grumpy) (~g@195.82.99.14) | 06:15 | |
| + chomwitt (~chomwitt@2a02:587:dc0b:200:12c3:7bff:fe6d:d374) | 07:16 | |
| - GNUmoon (QUIT: Ping timeout: 276 seconds) (~GNUmoon@gateway/tor-sasl/gnumoon) | 07:25 | |
| + erlehmann (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 07:36 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 08:38 | |
| - chomwitt (QUIT: Ping timeout: 268 seconds) (~chomwitt@2a02:587:dc0b:200:12c3:7bff:fe6d:d374) | 08:39 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:21:21:0:100b) | 09:05 | |
| + chomwitt (~chomwitt@athedsl-15695.home.otenet.gr) | 09:21 | |
| - erlehmann (QUIT: Read error: Connection reset by peer) (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 10:41 | |
| + erlehmann (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 10:41 | |
| + MajorBiscuit (~MajorBisc@wlan-145-94-218-113.wlan.tudelft.nl) | 10:48 | |
| + mjw (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 11:51 | |
| - erlehmann (QUIT: Quit: Just say no, then the virus can not enter your body without your consent.) (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 11:53 | |
| + erlehmann (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 11:53 | |
| - chomwitt (QUIT: Ping timeout: 256 seconds) (~chomwitt@athedsl-15695.home.otenet.gr) | 11:55 | |
| - mtm (QUIT: Ping timeout: 256 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 13:03 | |
| + chomwitt (~chomwitt@athedsl-15695.home.otenet.gr) | 13:19 | |
| + Christoph_ (~Christoph@p54bf63e2.dip0.t-ipconnect.de) | 13:25 | |
| - erlehmann (QUIT: Ping timeout: 240 seconds) (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 13:52 | |
| sigrid | mesa git doesn't need magic vars set nor patches applied to render properly. just tested that | 13:54 |
| + erlehmann (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 14:00 | |
| mntmn | sigrid: ah, they switched over to NIR then | 14:09 |
| mntmn | i say they like i hadn't commited to mesa myself ;3 | 14:10 |
| - erlehmann (QUIT: Ping timeout: 256 seconds) (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 14:11 | |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 15:09 | |
| + Nulo (~Nulo@user/nulo) | 16:04 | |
| - Ar|stote|is (QUIT: Ping timeout: 268 seconds) (~linx@149-210-24-194.mobile.nym.cosmote.net) | 16:06 | |
| + Ar|stote|is (~linx@149-210-32-252.mobile.nym.cosmote.net) | 16:10 | |
| + erlehmann (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 16:12 | |
| - Ar|stote|is (QUIT: Read error: Connection reset by peer) (~linx@149-210-32-252.mobile.nym.cosmote.net) | 16:20 | |
| + Ar|stote|is (~linx@149-210-32-252.mobile.nym.cosmote.net) | 16:29 | |
| - mjw (QUIT: Quit: Leaving) (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 18:36 | |
| Boostisbetter | mntmn: I saw the lovely stack of Reforms ready to go out on twitter. Most excellent! | 18:38 |
| Boostisbetter | I am sure a bunch of CS folks will be very happy | 18:38 |
| mntmn | yeaah | 18:39 |
| mntmn | finally it's moving | 18:39 |
| - MajorBiscuit (QUIT: Ping timeout: 256 seconds) (~MajorBisc@wlan-145-94-218-113.wlan.tudelft.nl) | 18:39 | |
| mntmn | josch: i'm ready to tackle this now... what's the current status from your POV? https://source.mnt.re/reform/reform-system-image/-/issues/15 | 18:54 |
| mntmn | btw who here is running kernel 5.15 (or newer) and where can i see your patch stack? | 19:01 |
| sigrid | ruff was doing that. I did that too | 19:05 |
| sigrid | I took ruff's stuff and turned it into a void package | 19:05 |
| sigrid | https://source.mnt.re/rufform/reform-system-image | 19:06 |
| mntmn | sigrid: thanks! | 19:06 |
| mntmn | i'm starting to prepare sysimage-v3 | 19:06 |
| sigrid | note that the kernel used there is not mainline, it's librem's fork | 19:07 |
| mntmn | ah, that's a problem | 19:08 |
| mntmn | but interesting nonetheless | 19:08 |
| sigrid | I tried to fix those patches up on mainline but could not get display to work | 19:09 |
| mntmn | sigrid: ok, i'm gonna attempt it | 19:10 |
| + vagrantc (~vagrant@2600:3c01:e000:21:21:21:0:100b) | 19:37 | |
| bluerise | mntmn: glad to see those reforms getting shipped | 19:48 |
| mntmn | bluerise: me too :3 | 19:50 |
| bluerise | hope my friends jcs gets his soon | 19:50 |
| mntmn | heard that texas has some winter trouble atm? | 19:52 |
| bluerise | Bavaria has some winter trouble atm | 19:52 |
| bluerise | a friend just stayed an hour at my place 'cause Blitzeis | 19:53 |
| mntmn | interesting https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/14466 | 19:54 |
| mntmn | bluerise: oof | 19:54 |
| sigrid | niiice | 19:54 |
| sigrid | i am going to try that one now | 19:55 |
| bluerise | if I was evil I would have ordered food for delivery ;) | 19:55 |
| mntmn | also https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9780#note_1213436 | 19:55 |
| Boostisbetter | mntmn: Speaking of kernel 5.15, am i able to update? Or are we still advised to hold? | 19:57 |
| mntmn | Boostisbetter: i have never tested it, building for the first time now | 20:07 |
| + bibliocar (~EricShmar@195.82.99.14) | 20:10 | |
| bibliocar | Haha, looking at those assembly pictures. I have to thank you for not making assembling the top part of the kit. I had tremendous difficulty snaking that wire around on the olimex laptop. | 20:11 |
| mntmn | bibliocar: yup ^^ | 20:11 |
| vagrantc | ACTION finally learned why vagrantc couldn't build upstream imx8mq ATF ... | 20:20 |
| vagrantc | i always build with debug, and it's grown too large to do so | 20:21 |
| vagrantc | disabling debug allows it to build! though ... no idea if it works yet :) | 20:21 |
| mntmn | vagrantc: woah | 20:21 |
| bluerise | argh | 20:22 |
| bluerise | you sure upstream ATF is better? | 20:22 |
| bluerise | I use the NXP sources | 20:22 |
| vagrantc | there's a vague note in upstream's readme that implies it's basically unsupported upstream, and that nxp will maintain a fork ... helpfully leaving out *where* that fork is | 20:23 |
| vagrantc | bluerise: no idea! | 20:23 |
| bluerise | codeaurora I think, 1s | 20:23 |
| vagrantc | yeah, that seems to be the place | 20:24 |
| bluerise | https://source.codeaurora.org/external/imx/imx-atf/ | 20:24 |
| vagrantc | just upstream docs are strangely unhelpful | 20:24 |
| sigrid | mesa git plus the one of those mentioned prs gives me 581fps over 541fps in glmark2 | 20:24 |
| sigrid | not so much in quake :) | 20:24 |
| vagrantc | ACTION shuffles around a few things on various backburners | 20:24 |
| - erlehmann (QUIT: Ping timeout: 240 seconds) (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 20:35 | |
| + erlehmann (~erle@ip5f5bd566.dynamic.kabel-deutschland.de) | 20:36 | |
| mntmn | sigrid: well, that's a little bit more | 20:48 |
| Boostisbetter | sigrid, you got Quake running on the Reform? Do tell | 20:48 |
| sigrid | not much to say, I just installed quakespasm | 20:51 |
| sigrid | some other quake engine might work faster, I haven't tested yet | 20:51 |
| mntmn | yeah i also used that one | 20:51 |
| sigrid | it runs at 30fps lowest on dm6 in full screen | 20:51 |
| mntmn | and ioquake3, but there was also some raspi-optimized version that was faster IIRC | 20:51 |
| mntmn | i also played DOOM3 (dhewm3), but it's super slow in some complex scenes | 20:52 |
| Boostisbetter | no way | 21:13 |
| Boostisbetter | that's awesome | 21:13 |
| josch | mntmn: I think the next step is to update the repository. My patch stack is here: https://source.mnt.re/reform/reform-debian-packages/-/tree/main/patches | 21:26 |
| - bibliocar (QUIT: Quit: Leaving) (~EricShmar@195.82.99.14) | 21:29 | |
| mntmn | josch: oh yeah i see | 21:30 |
| erlehmann | sigrid how many fps in minetest | 21:31 |
| erlehmann | with, say, repixture game | 21:31 |
| mntmn | josch: btw i have rebased the patches for 5.15 now (but not yet tested, doing some builds atm) | 21:31 |
| mntmn | josch: i'm wondering what is my todo list for integrating reform-debian-packages into reform-system-image... do you have an idea already? | 21:34 |
| josch | mntmn: i think that reform-debian-packages should generate the contents of https://mntre.com/reform-debian/sid/ and then reform-system-image can just apt-get install from there | 21:36 |
| mntmn | josch: gotcha. and the kernel should also be a .deb living there, yeah? | 21:42 |
| mntmn | plomlompom: heads up that we should talk about this tomorrow | 21:42 |
| mntmn | ok finally have the hdmi driver building again | 21:46 |
| sigrid | erlehmann: 25-40 | 21:47 |
| sigrid | on lowest settings but in fullscreen | 21:47 |
| erlehmann | sigrid which renderer though? | 21:47 |
| erlehmann | i bet you get a speed boost if you use OGLES | 21:48 |
| sigrid | how do I do that? | 21:48 |
| mntmn | does the void package for minetest ship ogles? | 21:48 |
| sigrid | I do not know :) I don't play minetest | 21:49 |
| mntmn | haha this question was loosely directed towards erlehmann | 21:49 |
| mntmn | erlehmann: https://github.com/void-linux/void-packages/blob/master/srcpkgs/minetest/template | 21:50 |
| erlehmann | let me look | 21:51 |
| erlehmann | mntmn i think it depends on irrlicht having it enabled or not | 21:52 |
| erlehmann | you can select ogles in minetest rendering | 21:52 |
| erlehmann | but i am not sure if it has an effect in 5.4.1 | 21:52 |
| - GNUmoon (QUIT: Ping timeout: 276 seconds) (~GNUmoon@gateway/tor-sasl/gnumoon) | 21:56 | |
| josch | mntmn: Yes, the kernel is a stock Debian kernel with Debian patches plus the patches needed for the reform. | 21:57 |
| josch | Thanks to flash-kernel, it will then be possible to apt-get install/upgrade the kernel and have u-boot find it. | 21:57 |
| mntmn | josch: ok, interesting. i wonder what patches debian has for the kernel | 21:57 |
| mntmn | josch: neat | 21:57 |
| josch | mntmn: https://sources.debian.org/src/linux/5.15.15-1/debian/patches/ -- most of it doesn't concern us but in the past there have always been Debian specific things like kernel.unprivileged_userns_clone where it makes sense if the reform system behaves like any other "normal" Debian system | 22:14 |
| ex-parrot | oooo josch | 22:24 |
| ex-parrot | that's cool | 22:24 |
| ex-parrot | I had been meaning to look at doing something similar myself and get flash-kernel up etc | 22:24 |
| - vagrantc (QUIT: Ping timeout: 268 seconds) (~vagrant@2600:3c01:e000:21:21:21:0:100b) | 22:42 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:21:21:0:100b) | 22:42 | |
| + GNUmoon (~GNUmoon@gateway/tor-sasl/gnumoon) | 22:56 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!