| Wizzup | freemangordon: looks like our qemu/ci image builder doesn't want ot do aarch64 | 12:38 |
|---|---|---|
| Wizzup | maybe I should dist-upgrade the underlying thing | 12:38 |
| Wizzup | I'm building a PP chimaera image for rafael | 12:38 |
| Wizzup | it works on my laptop it seems | 12:38 |
| Wizzup | freemangordon: did you get this? | 13:44 |
| Wizzup | /bin/sh: 1: cd: can't cd to /home/user/qtwebengine/src/3rdparty/chromium/third_party/jstemplate/ | 13:44 |
| Wizzup | ERROR - Cannot read: exports.js | 13:44 |
| Wizzup | ERROR - Cannot read: jsevalcontext.js | 13:44 |
| Wizzup | ERROR - Cannot read: jstemplate.js | 13:44 |
| Wizzup | ok, chimaera boots on PP | 14:07 |
| Wizzup | (first try :)) | 14:09 |
| Wizzup | freemangordon: this qt web engine build will take days on our CI I think | 14:27 |
| Wizzup | freemangordon: maybe this is silly, but we could inherit qapplication and override the platformName call | 14:31 |
| Wizzup | freemangordon: I think that qtwebbrowser could just have platformName() method | 14:36 |
| Wizzup | unless this is protected somehow | 14:36 |
| Wizzup | I asked in #qt | 14:40 |
| norayr | btw, i have a problem: got update for surf in HAM, but it doesn't install because of webkit2gtk problem. | 14:56 |
| Wizzup | beowulf? | 14:56 |
| norayr | chimaera | 14:56 |
| Wizzup | probably relates to elogind? | 14:57 |
| norayr | exact error in HAM: Conflict with application packages: libwebkit2gtk-4.0-37 (2.38.2-1~deb11u1) | 14:59 |
| norayr | no idea, i missed what was going on with elogin here. | 14:59 |
| Wizzup | I think to solve this we need more exact errors from apt | 15:00 |
| Wizzup | HAM isn't very useful typically | 15:00 |
| Wizzup | for this at least | 15:00 |
| norayr | minute, i'll check with apt. make sure easylist update works, and check surf with apt. | 15:01 |
| norayr | btw i am not able to make it rotate. :/ and maybe my menubar removal was very ugly, i was doing elsif, elif with preprocessor for Q_WS_MAEMO_5 but it wasn't picking it up. | 15:03 |
| norayr | ok checking the surf. | 15:04 |
| norayr | apt upgrade surf results in | 15:04 |
| norayr | hildon-base: confligts: elogind but 246.10-2 is to be installed. | 15:04 |
| norayr | sicelo: you run enlightenment on what? (: | 15:10 |
| Wizzup | norayr: I think you just use menuBar->hide(); ? | 15:14 |
| Wizzup | norayr: yeah, like I said, we have to solve the elogind thing, it's a known problem | 15:14 |
| norayr | i did that and it works in the new update, you can check. | 15:14 |
| norayr | ah, it is not solved yet, okay! | 15:14 |
| norayr | just i am afraid i did it not the most beautiful way, because there were ifdefs for symbian and there was else, and i wanted to add ifdef for maemo, but my elif didn't work, etc. so i just commented out everything assuming this will run only on maemo, and used the hide(). | 15:15 |
| norayr | but if you check rotation one day that'd be good. | 15:16 |
| Wizzup | norayr: let me check, btw, maybe just set auto rotate, rather than requiring user to select 'rotate' from menu | 15:17 |
| Wizzup | actually I can't check right now, will try later today, if I forget poke me | 15:17 |
| norayr | not urgent at all... will poke you in several days. | 15:17 |
| norayr | i hid the menubar in jan 6, which i wanted to do since dec 21. | 15:18 |
| Wizzup | :) | 15:21 |
| norayr | sorry folks i cannot find what is icd2 on the internet. | 15:36 |
| norayr | it heats the droid and eats 40-80% of cpu. | 15:37 |
| norayr | ok it belongs to the package icd2 | 15:37 |
| norayr | hm this icd2 crashed d4, eventually. the screen got blurred and it became unresponsive. then the power button was not able to halt it, i had to do power plus volume down. | 15:45 |
| Wizzup | what happened exactly? this is probably because (1) chimaera is mostly untested (2) you probably dist-upgraded'd, which also can result in funky setups | 15:54 |
| Wizzup | but still icd2 should behave | 15:54 |
| norayr | internet connectivity daemon, i understand. | 16:11 |
| norayr | i think i used offline mode, and then normal mode. | 16:11 |
| norayr | that is maybe the reason. | 16:11 |
| sicelo | norayr: no, i don't run enlightenment | 16:16 |
| sicelo | or you wanted to say i3wm? yes i do use i3 frequently on n900 | 16:19 |
| Wizzup | freemangordon: ok, I'm giving up on qtwebbrowser for now | 16:22 |
| freemangordon | Wizzup: why? | 16:43 |
| freemangordon | also, we cannot subclass QApplication in every application around | 16:43 |
| Wizzup | freemangordon: I think with apt-get source it might work, let me check it | 16:47 |
| Wizzup | freemangordon: apt-get source seems to work, I will try to import that into CI tonight | 17:03 |
| freemangordon | thanks | 17:04 |
| freemangordon | will you fix it as well? | 17:04 |
| freemangordon | as it does not make sense to build twice | 17:04 |
| Wizzup | yes, I am testing now in VM if it builds with fix | 17:05 |
| freemangordon | thanks! | 17:05 |
| dsc_ | could someone issue & share `cat /proc/cpuinfo` from a droid4 or n900 (or both) | 21:29 |
| dsc_ | my droid4 device is in some sort of limbo state :P | 21:30 |
| sicelo | heh | 21:41 |
| sicelo | droid 4: https://paste.debian.net/1266365/ | 21:44 |
| dsc_ | ty | 21:46 |
| sicelo | n900: https://paste.debian.net/1266366/ | 21:49 |
| sicelo | how do these help you with the problem however? | 21:49 |
| dsc_ | just confirming the presence of NEON feature | 21:51 |
| dsc_ | also eyeballing' if https://github.com/torch/TH/blob/master/CMakeLists.txt#L64-L68 would work (probably not) | 21:51 |
| dsc_ | because https://github.com/torch/TH/blob/master/cmake/FindARM.cmake#L29 | 21:52 |
| dsc_ | im porting some libs to ARM that make heavy use of SIMD instructions | 21:53 |
| dsc_ | like AVX, SSE | 21:53 |
| dsc_ | Wizzup: | 22:50 |
| dsc_ | [ 83%] Linking CXX static library ../libmarian.a | 22:50 |
| dsc_ | [100%] Built target marian | 22:50 |
| dsc_ | :O | 22:50 |
| dsc_ | (on cortex a72) | 22:54 |
| buZz | omap4430 indeed has NEON, see the datasheet | 22:57 |
| dsc_ | ye | 22:57 |
| buZz | (or try PCSXrearmed, it -flies- because of NEON) | 22:57 |
| dsc_ | i am 'in the market' for some CMake that can detect Cortex A8 (n900) or Cortex A9 (droid4) | 22:58 |
| dsc_ | but this FindARM.cmake that I linked is targeted at omap stuff, I missed that | 22:58 |
| buZz | no clue, cmake is 10th ring of hell imho | 22:58 |
| dsc_ | yes | 22:58 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!