#asahi · 2022-02-16 02:10
kovI'm now wondering if it may have been caused by a tool I used to look at cpu frenquency, maybe it tried to be smart and set the policy#asahi · 2022-02-16 02:08
kovlet's see if I can reproduce later#asahi · 2022-02-16 02:07
kovshrugs#asahi · 2022-02-16 02:02
kovI wonder if it may have something to do with the battery being lower than a certain point, but I'd expect that to be lower than at least 30%, and it was around 70#asahi · 2022-02-16 02:02
kovlet me try another boot#asahi · 2022-02-16 02:01
kovsteev, yeah, well, it said it once#asahi · 2022-02-16 02:00
kovsteev, I don't think that's needed on the M1, no, file doesn't exist fwiw#asahi · 2022-02-16 02:00
kovweird, now the Air has the correct policy, building the kernel with -j8 is putting the e-cores on 2.06 and the p-cores at 3.2 like the Mini#asahi · 2022-02-16 01:56
kovhmm#asahi · 2022-02-16 01:42
kov current policy: frequency should be within 600 MHz and 1.96 GHz. wat#asahi · 2022-02-16 01:40
koveven weirder, on the mba it seems to only push the p-cores to 1.96Ghz *scratches head*#asahi · 2022-02-16 00:59
kovbetter link https://trac.webkit.org/changeset/289805/webkit/releases/WebKitGTK/webkit-2.34#asahi · 2022-02-16 00:58
kovwe should see it soon enough in distros#asahi · 2022-02-16 00:57
kovtpw_rules, 16k pages made it into the stable branch for webkitgtk / wpe https://svn.webkit.org/repository/webkit/releases/WebKitGTK/webkit-2.34/Source/WTF/wtf/PageBlock.h#asahi · 2022-02-16 00:45
kovgoing make -j8 e-cores at 2.6 and p-cores at 3.2#asahi · 2022-02-16 00:45
kovI guess it's expected that the cluster changes frequency in sync? but all of them showing 3.2 was unexpected#asahi · 2022-02-16 00:44
kovsometimes the job goes to the e-core it seems, then they all go to 2GHz#asahi · 2022-02-16 00:43
kovthe p-cores on the mini are all boosting to 3.2GHz while compiling linux with a single job#asahi · 2022-02-16 00:42
kovweird#asahi · 2022-02-16 00:37
kovmarcan, I'm only getting 2GHz on the mini it seems, I think j`ey noticed the same thing above#asahi · 2022-02-16 00:12
kovyeah, tbh my workloads are usually all-core, so I care more about that ;D#asahi · 2022-02-16 00:12
kovI thought it was lower, not too bad#asahi · 2022-02-16 00:12
kovgreat#asahi · 2022-02-16 00:12
kovmarcan, when things are properly enabled do you know how high the all-core frequency is?#asahi · 2022-02-16 00:11
kovj`ey, nice#asahi · 2022-02-15 23:03
kovj`ey, thank you!#asahi · 2022-02-15 23:02
kovah, I see, 3.2 is single core boost then#asahi · 2022-02-15 23:01
kovj`ey, out of curiosity why is it that the full 3.2 can't be reached yet?#asahi · 2022-02-15 18:30
kovtoo =D#asahi · 2022-02-14 22:19
kovhttps://trac.webkit.org/wiki/WebKitGTK/StableRelease#asahi · 2022-02-14 22:19
kovthat's tricky to answer, but I'll probably poke the maintainers of webkitgtk to backport that to the stable branch#asahi · 2022-02-14 22:18
kovhttps://bugs.webkit.org/show_bug.cgi?id=236564 landed, gnome is pretty much ready for 16k page size ;D#asahi · 2022-02-14 19:06
kovgnome-shell 41.3 on debian testing#asahi · 2022-02-14 19:05
kovfwiw gnome idles at 0% pretty much#asahi · 2022-02-13 23:28
kovGlanzmann, hmm I've seen a parameter to mkfs to overcome that limit#asahi · 2022-02-13 18:33
kovtpw_rules, yep, 1.3.2#asahi · 2022-02-13 18:33
kovdebian testing, let me check libunwind#asahi · 2022-02-13 18:32
kovtpw_rules, but my daily driver has been fedora for a while, I'll try and install that later on#asahi · 2022-02-13 18:32
kovtpw_rules, I have debian on the m1 atm#asahi · 2022-02-13 18:32
kovI guess potential memory bloat, though that would be minimal I guess?#asahi · 2022-02-13 18:31
kovis there any other concern btw other than binary size?#asahi · 2022-02-13 18:31
kovin my testing#asahi · 2022-02-13 18:30
kovtpw_rules, thanks! it worked with jsc and a couple apps that use webviews#asahi · 2022-02-13 18:21
kovtpw_rules, will you submit it upstream? =)#asahi · 2022-02-13 18:21
kovtpw_rules, this is yours I think? https://pastebin.com/NzJEQJDW#asahi · 2022-02-13 18:13
kovbut in the end, if there's no big impact on binary size (which seems to be their only motivation?) I think chromium will accept a change to 16k?#asahi · 2022-02-13 18:12
kovwhen linux and qemu with rosetta support#asahi · 2022-02-13 18:12
kovhaha I think even the xps 13 will be faster than that ;D#asahi · 2022-02-13 18:11
kovI was trying to get an electron build going here for vscode, but I think it'll be easier to cross compile on x86 =/ the whole thing assumes x86-64#asahi · 2022-02-13 18:09
kovI also wonder, I'll try to find out, the first mention I've seen of chromium failing was actually on a bug that mentioned RH kernels