[2022-03-21T00:24:15Z] hi sad_plan [2022-03-21T00:27:58Z] how you doing? C: [2022-03-21T00:51:32Z] pretty good, hbu? [2022-03-21T01:36:33Z] good good. Im good :D [2022-03-21T04:06:10Z] phoebos: did you mention using lilypond at some point or am i thinking of someone else? [2022-03-21T04:28:56Z] Hi [2022-03-21T04:29:24Z] @Guest6647:libera.chat: lol chromium wayland is broken on nvidia [2022-03-21T04:29:32Z] It flickers a lot unless u --disable-gpu [2022-03-21T04:31:19Z] What's wrong with firefox? Everything works properly there with hw accel and better than chromium [2022-03-21T04:38:58Z] ipv6 [2022-03-21T04:39:25Z] i mean for his use case [2022-03-21T04:40:02Z] ah [2022-03-21T05:20:16Z] Guest6647: https://termbin.com/xn1e [2022-03-21T05:21:19Z] testuser[m]: Got it, thanks [2022-03-21T05:21:43Z] Oh [2022-03-21T05:21:53Z] Works [2022-03-21T05:21:58Z] Hm [2022-03-21T05:22:05Z] show ur exact output [2022-03-21T05:22:09Z] what gpu [2022-03-21T05:22:53Z] https://termbin.com/u6ev [2022-03-21T05:25:22Z] I'll write later, I'm just in college right now C: [2022-03-21T05:25:29Z] ok [2022-03-21T07:25:32Z] illiliti i ran the sandbox unit tests [2022-03-21T07:25:55Z] musl https://termbin.com/ejjd [2022-03-21T07:25:56Z] glibc https://termbin.com/hzs7 [2022-03-21T07:26:15Z] 11 extra failures on musl + 1 timeout [2022-03-21T07:26:25Z] the trap message shows up in dmesg [2022-03-21T07:26:32Z] need to check which exact test causes that [2022-03-21T07:46:40Z] can you attach gdb into it? [2022-03-21T07:46:46Z] gdb shall catch that sigtrap [2022-03-21T07:54:23Z] yea it does but idk which exact test to look for, since some tests trap on glibc too [2022-03-21T07:59:06Z] wait [2022-03-21T07:59:08Z] ill write a script [2022-03-21T08:19:40Z] o there are some tests where its supposed to die with a sigtrap so idk how tf to differentiate between an actual trap and an intended one [2022-03-21T08:19:41Z] bruh [2022-03-21T08:51:44Z] i ran chromium with --disable-seccomp-filter-sandbox and it crashes on medium.com almost immediately [2022-03-21T08:51:50Z] so the bug is not in seccomp if i understand correctly [2022-03-21T08:59:59Z] bruh there are no extra traps in musl tests, same tests trap on glibc and same on musl (intended) [2022-03-21T09:00:10Z] the extra failing tests on musl are realted to filesystem permission stuff [2022-03-21T09:00:13Z] weird [2022-03-21T09:00:38Z] i used this garbage script https://termbin.com/jpuq [2022-03-21T09:01:36Z] https://termbin.com/op0xx these failures are unique to musl [2022-03-21T09:01:46Z] only the ForkInNewPidNamespace.ExitWithSignal seems sus cuz it gets stuck [2022-03-21T09:08:11Z] i see [2022-03-21T09:08:18Z] is there a way to enable verbose logs? [2022-03-21T09:08:31Z] like what assert fails [2022-03-21T09:10:25Z] yea they use gtest so it tells u what value mismatched [2022-03-21T09:10:30Z] but this test just gets stuck so it doesnt know that [2022-03-21T09:10:33Z] ill have to put some printfs [2022-03-21T09:10:57Z] can you disable that test? [2022-03-21T09:11:12Z] wdym [2022-03-21T09:11:15Z] im running all of them one by one [2022-03-21T09:12:39Z] ah [2022-03-21T09:12:43Z] i thought that test prevents other tests from printing logs [2022-03-21T09:18:49Z] btw that threadpoolforeg thing belongs to glib [2022-03-21T09:18:54Z] the one that keeps segfaulting [2022-03-21T09:20:02Z] hmm no [2022-03-21T09:42:15Z] bruh im trying to build with symbol_level set but the obj files dont seem to have any debug info [2022-03-21T09:42:28Z] ig ill have to pass -g manually [2022-03-21T09:44:15Z] what does is_debug do? [2022-03-21T09:45:37Z] also set this: dcheck_always_on = true [2022-03-21T09:47:20Z] non debug build just enables more optimization [2022-03-21T09:47:25Z] nothing with symbols [2022-03-21T09:49:32Z] btw you should use -ggdb rather than -g [2022-03-21T09:51:21Z] or even -ggdb3, since chromium is highly polluted with macros [2022-03-21T09:52:17Z] it'll probably take a whole day to build with that [2022-03-21T09:52:31Z] yeah [2022-03-21T09:52:37Z] whole week [2022-03-21T09:58:33Z] https://termbin.com/serb i checked the args and symbol_level does seem to be set properly, no idea how tf they're passing it to the compiler.. [2022-03-21T09:58:50Z] nothing in ninja file or in /proc/pid/cmdline of gcc [2022-03-21T10:01:07Z] have you set is_debug=true? [2022-03-21T10:07:48Z] doesn't seem to be, it's not present in alpine's debug build file either [2022-03-21T10:07:57Z] They just set symbol level [2022-03-21T10:08:01Z] Seem to be required* [2022-03-21T10:10:35Z] blink_symbol_level=2 [2022-03-21T10:11:14Z] Yeah [2022-03-21T10:11:28Z] v8_symbol_level=2 [2022-03-21T10:11:30Z] Both that and symbol_level without blink_ prefix [2022-03-21T10:20:56Z] is_component_build=true [2022-03-21T10:33:09Z] Wat [2022-03-21T10:55:57Z] noocsharp: not me [2022-03-21T13:22:08Z] http://0x0.st/oNO5.png [2022-03-21T13:22:20Z] illiliti: looks like the debug symbols were enabled after all [2022-03-21T13:27:03Z] Lol the final link took up everything including swap (24gb) and crashed [2022-03-21T13:51:10Z] mold? [2022-03-21T13:54:20Z] gnoo ld [2022-03-21T13:54:38Z] i dont think ram usage will be reduced unless mold uses mmap or someth [2022-03-21T13:54:49Z] 30gb seems to be enough, been linking for 20 mins now.. [2022-03-21T13:56:08Z] oh god i forgot i would have to relink if i tried to fix the bug [2022-03-21T13:56:21Z] u can mix linkers right? [2022-03-21T13:58:12Z] should be possible if lto/pgo isn't used [2022-03-21T13:58:23Z] oh i forgot its just a collection of obj files in static libs before final link, so it shouldnt matter [2022-03-21T13:58:33Z] ill wait 10 more mins else switch to mold [2022-03-21T14:05:45Z] lol mold finished it in under a minute and didnt even touch swap [2022-03-21T14:08:43Z] the exe is 2gb it'll take some time to upload it [2022-03-21T14:08:50Z] good luck opening it under gdb lol [2022-03-21T14:21:32Z] 2gb omg [2022-03-21T14:21:47Z] is there a way to disable crashing on this fucking DCHECK [2022-03-21T14:22:09Z] the check is in a public header so it'll need a ton of recompilation to comment it out [2022-03-21T14:22:27Z] why it crashes? [2022-03-21T14:23:26Z] https://github.com/chromium/chromium/blob/75c2ebf35d3ae28207dfb8f3f679cde9d2800ff5/mojo/public/cpp/bindings/remote.h#L261 [2022-03-21T14:23:27Z] god knows [2022-03-21T14:23:42Z] --disable-dcheck ? [2022-03-21T14:24:38Z] doesnt do anything [2022-03-21T14:31:39Z] well, you have to recompile then [2022-03-21T14:32:06Z] but, iirc gdb can import debug symbols from other files [2022-03-21T14:33:04Z] so you can run gdb chromium_without_dcheck and then import debug symbols from chromium_with_dcheck [2022-03-21T14:33:30Z] https://termbin.com/3z88 [2022-03-21T14:33:44Z] bruh [2022-03-21T14:33:53Z] i took a snapshot of this tree, lets see [2022-03-21T14:35:01Z] [584/9304] [2022-03-21T14:35:02Z] no [2022-03-21T14:35:10Z] 1/3rd of the full build [2022-03-21T14:35:11Z] no [2022-03-21T14:37:40Z] how many cores do you have? [2022-03-21T14:38:33Z] 6 [2022-03-21T14:40:41Z] Tbh our time would be better spent figuring out broken ipv6 in Firefox [2022-03-21T14:51:08Z] i agree [2022-03-21T14:51:23Z] but let's deal with chromium first [2022-03-21T14:55:19Z] I grepped for DCHECK in sandbox/linux there's barely 5-10 of them so if more dchecks get triggered after rebuild im just gonna rebuild with dcheck disabled [2022-03-21T15:53:35Z] Fuck i could've just ld preloaded the abort function [2022-03-21T16:13:32Z] nvm `#define TRAP_SEQUENCE1_() asm volatile("int3")` [2022-03-21T16:16:30Z] i can still do that lol i forgot the macro is called from a function [2022-03-21T16:16:44Z] im going crazy [2022-03-21T17:07:52Z] Sha256 80e211ba5678c4f3c4a296e5a5816a5f2322a8efa40a3a8628d11090f4767afd [2022-03-21T17:07:54Z] https://bin.git-bruh.duckdns.org/chromium.tar.xz [2022-03-21T17:08:08Z] illiliti: it won't trap for any DCHECK failure now [2022-03-21T17:09:08Z] Btw i saw a tab crash and some dcheck was failing for nodejs mutex stuff [2022-03-21T19:34:29Z] hey guys [2022-03-21T19:34:40Z] does mpv in repo support lua scripts? [2022-03-21T19:38:16Z] if you install lua before mpv [2022-03-21T19:38:55Z] because then the mpv build will detect lua is installed and build it with lua suppost [2022-03-21T19:40:12Z] Hi all again [2022-03-21T19:40:32Z] acheam: I understand that my mpv does not open videos via youtube-dl, could this be because\ I did not install lua before building mpv? [2022-03-21T19:45:09Z] to only use youtubedl just install luajit in community [2022-03-21T19:45:24Z] and rebuild mpv [2022-03-21T19:47:11Z] Thanks [2022-03-21T20:07:45Z] acheam: just luajit do the work [2022-03-21T20:07:57Z] i put the scripts in the wrong location [2022-03-21T20:08:57Z] Guest69: yes, as rohan said [2022-03-21T20:09:10Z] rohan: not sure what you mean by "just luajit do the work" [2022-03-21T20:09:25Z] you said to install lua [2022-03-21T20:09:33Z] community have lua and luajit [2022-03-21T20:10:01Z] just using luajit can make the scripts run [2022-03-21T20:10:11Z] ah [2022-03-21T20:10:22Z] mpv works with lua 5.1, 5.2, and luajit [2022-03-21T20:10:50Z] idk what version is in community, I used my own lua5.2 package [2022-03-21T20:29:49Z] acheam: ok, i'll try it