#asahi · 2022-02-18 17:05
axboeGlanzmann: yeah, that sucks...#asahi · 2022-02-18 17:05
axboeGlanzmann: yeah thath doesn't look great#asahi · 2022-02-18 17:03
axboeI generally find them pretty solid too, some models end up being dodgy but most are good#asahi · 2022-02-18 17:01
axboeif after a straight overwrite, defective hw for sure#asahi · 2022-02-18 17:01
axboethe way ssds do error correction these days, it should rewrite the block if the read confidence is marginal#asahi · 2022-02-18 17:00
axboeGlanzmann: I've been using samsung in other boxes and my other laptop, I've never seen that. I'm curious if it's some side effect of an issue with eg queued trim like you mentioned, because if not, the fw is pretty buggy#asahi · 2022-02-18 16:58
axboeGlanzmann: an ssd should never throw uncorrectable read errors, that's a huge red flag#asahi · 2022-02-18 16:10
axboejust for full disclosure#asahi · 2022-02-18 16:10
axboeGlanzmann: fwiw, I also run an m1n1 hack to have it work with the disabled cores#asahi · 2022-02-18 16:08
axboeGlanzmann: nothing#asahi · 2022-02-18 16:07
axboej`ey: with fire!#asahi · 2022-02-18 16:07
axboemostly good stuff in there, a few work-arounds that are specific to my setup#asahi · 2022-02-18 16:07
axboewhich does have the "kill dead cores on my m1 pro"#asahi · 2022-02-18 16:06
axboeGlanzmann: outside of that, it's just sitting on top of the stuff I currently run on it#asahi · 2022-02-18 16:06
axboeGlanzmann: that one was mostly just for sven to see the apple-nvme changes to support some of the effiency features#asahi · 2022-02-18 15:58
axboeI don't need to learn Yet Another One#asahi · 2022-02-18 15:58
axboemy main gripe with trying a new distro is the packaging system#asahi · 2022-02-15 20:28
axboeGlanzmann: sorry, meant nvme with fix that you tested#asahi · 2022-02-15 20:26
axboeGlanzmann: seems solid here now too#asahi · 2022-02-15 19:45
axboefind does indeed reproduce it pretty quickly here too#asahi · 2022-02-15 19:44
axboesounds like it#asahi · 2022-02-15 19:44
axboeGlanzmann: any timeouts in dmesg?#asahi · 2022-02-14 22:21
axboejannau: yep, 3 per cluster is what I see#asahi · 2022-02-14 22:18
axboeok I think the layout is right, it must be something with the freq driver then#asahi · 2022-02-14 21:38
axboeI might've just messed up the definitions#asahi · 2022-02-14 21:35
axboethat might explain it, if so#asahi · 2022-02-14 21:35
axboeonly [4 5] in the last one#asahi · 2022-02-14 21:34
axboeare the clusters [ 0 1 2 ] [ 3 4 5], or [ 0 1 2 3 ] [ 4 5 6 7 ]#asahi · 2022-02-14 21:34
axboesince I have 6 p cores#asahi · 2022-02-14 21:34
axboeso...#asahi · 2022-02-14 21:33
axboejannau: could be, yes#asahi · 2022-02-14 21:33
axboeor if m1n1 is setting up things wrong for the t600x#asahi · 2022-02-14 21:32
axboekettenis: not spending time on that, just trying to verify if there's something wrong with how I set it up#asahi · 2022-02-14 21:31
axboekettenis: yeah I know#asahi · 2022-02-14 21:29
axboenever worked on embedded like stuff, so just stabbing in the dark here and finding clues :)#asahi · 2022-02-14 21:28
axboethe ratio between them are pretty close to max-freq/boot-freq#asahi · 2022-02-14 21:27
axboep0..p2 seem to scale fine, p3..p5 do not#asahi · 2022-02-14 21:27
axboejannau: I dropped that m1n1 hack to set all P cores to 15 in cpufreq#asahi · 2022-02-14 21:08
axboethe fact that cpu_capacity is also missing for those 3 cpus makes me suspicious that something is wrong in the definition or bringup of them#asahi · 2022-02-14 21:07
axboeso they are most likely still running at full tilt, which would help explain why battery isn't as good as it should be#asahi · 2022-02-14 21:07
axboewas digging a bit, and seems like cpufreq isn't correctly set on cpus 5-7#asahi · 2022-02-14 19:13
axboehuh, cpu_capacity missing from 5..7#asahi · 2022-02-14 19:08
axboej`ey: says 54.7Wh#asahi · 2022-02-14 18:58
axboebut cpufreq made it definitely cooler on the lap#asahi · 2022-02-14 18:58
axboeGlanzmann: I've never run osx on this one except to do the setup, so not sure what it's like there#asahi · 2022-02-14 18:57
axboenot super bright though#asahi · 2022-02-14 18:57
axboemost likely#asahi · 2022-02-14 18:56
axboeso unless the idle loop is doing something stupid, it's not the cpus eating the power#asahi · 2022-02-14 18:56
axboenot sure what it is, looking at frequency it's ~6% at 600MHz, ~93% Idle#asahi · 2022-02-14 18:50
axboexfce4 definitely more idle, but still seem to get a pretty low battery life