marcan changed the topic of #asahi to: Asahi Linux: porting Linux to Apple Silicon macs | "Does XXX work yet?": https://alx.sh/fs | GitHub: https://alx.sh/g | Wiki: https://alx.sh/w | Topics: #asahi-dev #asahi-re #asahi-gpu #asahi-alt #asahi-stream #asahi-offtopic | Keep things on topic | Logs: https://alx.sh/l/asahi
jeisom_ has joined #asahi
john-cabaj has quit [Ping timeout: 480 seconds]
jeisom_ has quit [Ping timeout: 480 seconds]
kidplayer666 has quit [Quit: Connection closed for inactivity]
holiday has quit [Quit: WeeChat 4.1.1]
linuxgemini has quit [Quit: getting (hopefully fresh) air o/]
linuxgemini has joined #asahi
lewurm has quit [Ping timeout: 480 seconds]
linuxgemini has quit [Quit: getting (hopefully fresh) air o/]
linuxgemini has joined #asahi
Brainium has quit [Quit: Konversation terminated!]
lewurm has joined #asahi
linuxgemini has quit [Quit: getting (hopefully fresh) air o/]
linuxgemini has joined #asahi
milky has joined #asahi
denysonique has quit [Quit: .]
denysonique has joined #asahi
milky has quit [Remote host closed the connection]
denysonique has quit [Quit: .]
denysonique has joined #asahi
KxCORP has quit [Quit: Bye!]
KxCORP has joined #asahi
kdb424 has quit [Quit: The Lounge - https://thelounge.chat]
crabbedhaloablut has joined #asahi
hightower4 has joined #asahi
hightower3 has quit [Ping timeout: 480 seconds]
ten157237743247 has joined #asahi
ten15723774324 has quit [Ping timeout: 480 seconds]
ten157237743247 is now known as ten15723774324
vx^ has quit [Remote host closed the connection]
vx has joined #asahi
vx is now known as Guest7773
jmplans has joined #asahi
jmplans__ has quit [Ping timeout: 480 seconds]
jmplans_ has joined #asahi
jmplans has quit [Ping timeout: 480 seconds]
mxz has quit [Quit: cya]
mxz has joined #asahi
jnn has joined #asahi
marvin24 has joined #asahi
jn has quit [Ping timeout: 480 seconds]
marvin24_ has quit [Ping timeout: 480 seconds]
kdb424 has joined #asahi
eroc1990 is now known as Guest7779
eroc1990 has joined #asahi
Guest7779 has quit [Ping timeout: 480 seconds]
kidplayer666 has joined #asahi
frisvee has joined #asahi
kraem has quit [Remote host closed the connection]
kraem has joined #asahi
pthariensflame has joined #asahi
pthariensflame has quit []
maxkofler has joined #asahi
frisvee has quit [Remote host closed the connection]
ten15723774324 has quit [Remote host closed the connection]
ten15723774324 has joined #asahi
joske has joined #asahi
frisvee has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
joske has quit [Quit: joske]
frisvee has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
frisvee has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
joske has joined #asahi
joske has quit [Remote host closed the connection]
frisvee has joined #asahi
holiday has joined #asahi
jeisom_ has joined #asahi
jeisom_ has quit [Ping timeout: 480 seconds]
maxkofler_ has joined #asahi
maxkofler has quit [Ping timeout: 480 seconds]
kidplayer666 has quit [Quit: Connection closed for inactivity]
kidplayer666 has joined #asahi
asimpson has quit [Ping timeout: 480 seconds]
asimpson has joined #asahi
zzywysm has quit [Read error: Connection reset by peer]
zzywysm_ has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
maxkofler_ has quit [Quit: https://quassel-irc.org - Chat comfortably. Anywhere.]
seb4nih14 has quit [Read error: Connection reset by peer]
alethkit has quit [Ping timeout: 480 seconds]
alethkit has joined #asahi
frisvee has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
seb4nih8 has joined #asahi
ciggi has joined #asahi
frisvee has joined #asahi
Moprius has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
holiday has quit [Ping timeout: 480 seconds]
Moprius has quit [Quit: bye]
Szadek636 has quit []
frisvee has joined #asahi
holiday has joined #asahi
Brainium has joined #asahi
rvalue has quit [Ping timeout: 480 seconds]
frisvee has quit [Ping timeout: 480 seconds]
frisvee has joined #asahi
holiday has quit [Ping timeout: 480 seconds]
holiday has joined #asahi
Szadek636 has joined #asahi
jeisom_ has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
rvalue has joined #asahi
frisvee has joined #asahi
frisvee has quit [Ping timeout: 480 seconds]
frisvee has joined #asahi
millefy has quit [Ping timeout: 480 seconds]
frisvee has quit []
holiday has quit [Quit: WeeChat 4.1.1]
SalimTerryLi has quit [Read error: Connection reset by peer]
sefidel has quit [Remote host closed the connection]
sefidel has joined #asahi
sefidel has quit [Remote host closed the connection]
sefidel has joined #asahi
SalimTerryLi has joined #asahi
Brainium has quit [Quit: Konversation terminated!]
john-cabaj has joined #asahi
jnn is now known as jn
john-cabaj has quit [Ping timeout: 480 seconds]
yuka has quit [Remote host closed the connection]
yuka has joined #asahi
nicolas17 has quit [Read error: Connection reset by peer]
nicolas17 has joined #asahi
delsol has joined #asahi
joshheyse has quit [Quit: ZNC 1.8.2+deb2build5 - https://znc.in]
holiday has joined #asahi
zealouscupuacu has joined #asahi
<zealouscupuacu> hey folks, is there a place where I can find information around compiling ARM programs using 64K page size vs other page sizes.
<tpw_rules> what information? there's not a lot to it
joshheyse has joined #asahi
<zealouscupuacu> I am trying to understand the impact of such change and arguments in favor of it. mainly to explain to package maintainers why such a change might be necessary and the pros and cons of it.
<tpw_rules> imho there's not a lot to that either. the pro is it runs on your system and the con is it would be hard to do if the program is not page-size-agnostic
<tpw_rules> "compiling ARM programs using 64K page size" is not really well defined anyway. if a program doesn't care about page sizes then that doesn't matter
<zealouscupuacu> would it cause the program to consume more memory or have any difference in performance?
<tpw_rules> page size is (in the general case) an attribute of the system the program is running on, not the program itself
<tpw_rules> but for some specific programs, those things could change depending how you compile them. do you have a specific example?
<zealouscupuacu> to be more specific, this might be an issue with all packages being built with GraalVM on NixOS. GraalVM by default uses the value of host for -H:PageSize and NixOS apparently runs on a 4K aarch64 kernel. Although they install fine, they fail to run altogether. Examples are babashka and clojure-lsp. By compiling those packages with -H:PageSize=64K they all work.
<zealouscupuacu> I proposed a fix for babashka specifically but the maintainer has a really good point that it should be fixed for all GraalVM packages. But there are also concerns of the impact of such changes.
<tpw_rules> you'd have to talk to the GraalVM flks
<tpw_rules> folks
<tpw_rules> https://github.com/AsahiLinux/docs/wiki/Broken-Software this is sort of the collective asahi wisdom on that subject
joshheyse has quit [Quit: ZNC 1.8.2+deb2build5 - https://znc.in]
<tpw_rules> i went through it with the nixpkgs folks to get jemalloc updated and they seemed okay with that
<zealouscupuacu> thanks for all the pointers and also thanks for the great work on nixos with asahi, I am having a blast with it :)
<tpw_rules> nice to hear from a satisfied user
<tpw_rules> (hm my memory might have been wrong about jemalloc)
<tpw_rules> (i think it was libunwind actually...)
<zealouscupuacu> I understand their concerns but I am also unable to come up with firm proof it's fine to do the change.
<zealouscupuacu> I will link to relates issues in other repositories and maybe that's enough for moving this forward.
<zealouscupuacu> could you maybe refer to the discussion in nixos about libunwind, that might be helpful to mention as well
tobhe_ has joined #asahi
tobhe has quit [Ping timeout: 480 seconds]
joshheyse has joined #asahi
sgoskz has joined #asahi
sgoskz has left #asahi [WeeChat 4.1.1]
ten15723774324 has quit [Remote host closed the connection]
ten15723774324 has joined #asahi
jeisom_ has quit [Quit: Leaving]
<thasti> firefox 120 (just upgraded from the alarm repos) is segfaulting on startup for me .. anyone else?
<cyrinux> yes, me too
<thasti> ok, thanks
jeisom has joined #asahi
<janneg> same here on M1 Ultra. NULL ptr deref in SetPHCState(), possibly compile-time runtime page size mismatch or config issue
jeisom has quit [Ping timeout: 480 seconds]
<janneg> fedora built firefox 120 might be fine, trying to confirm whether 8GB ram devices are ok in general
ChaiTRex has quit [Quit: ChaiTRex]
zealouscupuacu has quit [Ping timeout: 480 seconds]
ChaiTRex has joined #asahi
<janneg> alarm built firefox works on 8GB devices, fedora built firefox is unaffected: spotted in th fedora's build script 'echo "ac_add_options --disable-jemalloc" >> .mozconfig'
<janneg> no such thing in alarm's PKGBUILD, sigh
<thasti> will try to build locally with this option
<cy8aer> Does someone experience a segfault with firefox 1.120? It crashes for me from the actual debian package (x86 works fine).
<psykose> read right above
<cy8aer> (facepalm - sorry)
Moprius has joined #asahi
<janneg> pretty certain that it is jemalloc build on a 4k host: this seems to be upstream change which caused this https://bugzilla.mozilla.org/show_bug.cgi?id=1854550
<janneg> does someone know where to manually turn off Firefox Nimbus feature flags? https://hg.mozilla.org/mozilla-central/rev/54717c8e85b2 should disable this and avoid the crash
Fischmie- has joined #asahi
Fischmiep has quit [Read error: Connection reset by peer]
<psykose> the key in setPref: is the one in about:config
<psykose> and so also in prefs.js i think
<psykose> in something like /usr/lib/firefox/defaults/pref/channel-prefs.js pref("memory.phc.enabled", false); might work, but i didn't try
<psykose> or /usr/lib/firefox/browser/defaults/preferences/vendor.js if you have it, i don't know what is actually used from all the prefs js files..
<psykose> seems there's also a prefs.js in .mozilla/profilename/
<psykose> * ~/.mozilla/firefox/profilename/
<cy8aer> I tried to set memory.phc.enabled in my profile's prefs.js. Did not work for me.
<janneg> /usr/lib/firefox/browser/defaults/preferences/vendor.js but adding the pref there has no effect
<cy8aer> ahem disabled (so set false of course)
<psykose> i can confirm that it's read, i.e. if i toggle it in the file then the state is reflected in about:config
<psykose> it probably doesn't fix the crash though
<kidplayer666> hey there yall
<kidplayer666> am i the only one who is unable to open ungoogled chromium regardless of repo used
<kidplayer666> but able to open normal chromium?
<janneg> yes, it can't fix the crash because the call trying to set the pref will crashes
<psykose> :D
<kidplayer666> ah crap
<kidplayer666> cause every now and then i need to use a chromium based browser
<kidplayer666> and really didn't want any of the google crap attached
<kidplayer666> I'm guessing I'll either have to apply the patches myself
<kidplayer666> Or use stuff like Brave
<cy8aer> There is a upstream bug report needed...
<thasti> build with --disable-jemalloc works fine, no segfault
<janneg> eric_engestrom: this is your crash
Brainium has joined #asahi
<thasti> i'll try to build on the mac with out that flag, just to confirm whether it's the build machine page size that matters
<janneg> there's already a bugzilla ticket but I can't easily link to it since firefox is broken
<kidplayer666> wait
<kidplayer666> I'm apparently on firefox 120 with no problems??
<kidplayer666> On an M1 on Fedora with 8GB of ram??
<thasti> only the arch linux arm builds appear to be affected (see above)
<kidplayer666> aaaah
<kidplayer666> ok
<cy8aer> ... and the debian one!
<kidplayer666> yep
<kidplayer666> i love debian
<kidplayer666> one of the first distros i daily drived
<kidplayer666> with xfce
<kidplayer666> on an old 2007 laptop
<janneg> devices with 8GB RAM are also not affected: https://searchfox.org/mozilla-central/source/xpcom/base/PHCManager.cpp#37
<kidplayer666> one of the few distros i did not manage to bork
<kidplayer666> gets a special place in my heart
<eric_engestrom> Thanks janneg ! I ended up taking this "opportunity" to finally wipe my disk and reinstall to fedora instead of alarm because I'm just fed up with all these papercuts, and right now I'm having some issue with btrfs so I don't have a laptop right now, just my phone 😅
Moprius has quit [Remote host closed the connection]
<kidplayer666> eric_engestrom: moving to fedora? One of us, one of us
billak has joined #asahi
john-cabaj has joined #asahi
<eric_engestrom> Hehe
<kidplayer666> first time i used fedora tbh
<kidplayer666> for a "novice" user, there isn't that much of a difference tbh
<kidplayer666> unless it's compared to arch, where i heard the aur is kinda nuts, and the wiki is always useful
<eric_engestrom> I'm also new to fedora, but from what I understand COPR is a good equivalent to the AUR
<janneg> eric_engestrom: I'd say it's luck that fedora is not affected or someone in fedora systematically disables jemalloc after https://github.com/jemalloc/jemalloc/issues/467
<kidplayer666> I don't know to be honest, I mostly stuck with Debian based distros throughout my Linux roamings
<kidplayer666> With the exception of manjaro
john-cabaj has quit [Ping timeout: 480 seconds]
<cy8aer> Ok I filed a debian bug... let's see what happens. As I know the debian package maintainer is a firefox developer.
julio7359 has joined #asahi
<thasti> local build without --disable-jemalloc segfaults the same way the one from the alarm repo does
holiday has quit [Ping timeout: 480 seconds]
holiday has joined #asahi
holiday has quit [Quit: WeeChat 4.1.1]
<janneg> it's not jemalloc bug but a genuine firefox bug
<janneg> I guess PHC is not built when jemalloc is disabled
ten15723774324 has quit [Remote host closed the connection]
ten15723774324 has joined #asahi
<janneg> thasti: kPageSize is defined to 4096 (except for macos arm64) and not build dependent
<janneg> so building firefox on a 16k is not going to help
<cy8aer> janneg: I'll do as soon as the bug is out
holiday has joined #asahi
rootbeerdan has quit [Quit: rootbeerdan]
maximbaz has quit [Quit: bye]
maximbaz has joined #asahi
billak has quit [Quit: Konversation terminated!]
maximbaz has quit [Ping timeout: 480 seconds]
i509vcb has quit [Quit: Connection closed for inactivity]
jjanzic has quit [Ping timeout: 480 seconds]
ydalton has joined #asahi
<ydalton> hey, how can i downgrade the mesa packages? i'm trying to see if there was a regression
jjanzic has joined #asahi
<j`ey> ydalton: i believe its just: dnf downgrade mesa
<j`ey> presumably it'll tell you the versions before doing anything
<ydalton> i'm guessing `dnf downgrade mesa*`
<ydalton> that worked, thanks!
ydalton has left #asahi [ERC 5.5.0.29.1 (IRC client for GNU Emacs 29.0.92)]
chadmed has joined #asahi
julio7359 has quit [Ping timeout: 480 seconds]
julio7359 has joined #asahi
lynndotpy has quit [Quit: bye bye]
lynndotpy has joined #asahi
alexaltea_ has quit [Ping timeout: 480 seconds]
<kidplayer666> ydalton: figured out yet what version you're in?
<kidplayer666> (of openGL)
<kidplayer666> also, given the now existant opengl 3.3 support, what is the expected roadmap for future API support? will the driver continue to slowly climb the openGL version or will it "jump" to vulkan earlier?
julio7359 has quit []
<j`ey> kidplayer666: probably will just continue as it is
<j`ey> people are working on vulkan
<j`ey> maybe l_ina will eventually join thst effort too
i509vcb has joined #asahi
kuruczgy has joined #asahi
alexaltea has joined #asahi
<kidplayer666> Alright
lynndotpy has quit [Quit: bye bye]
lynndotpy has joined #asahi
pthariensflame has joined #asahi
pthariensflame has quit []
delsol has quit [Remote host closed the connection]
maximbaz has joined #asahi
holiday has quit [Quit: WeeChat 4.1.1]
<ayke> Well I'm glad I switched to Fedora yesterday. No sudden issue with Firefox not working (even though this appears to be more luck than anything).
holiday has joined #asahi
crabbedhaloablut has quit []
jjanzic has quit [Quit: Bridge terminating on SIGTERM]