#
…

2022-11-23

00:01amarioguy2 has joined #asahi
00:04millenialhacker has quit [Ping timeout: 480 seconds]
00:07jannauwhere do you see percent values for the keyboard backlight? in my kde "battery and brightness" dialog is just a slider without scale
00:08jannaubut it behaves as expected and can turn the kbd backlight completely off
00:08jannaudoes `echo -n 0 > /sys/class/leds/kbd_backlight/brightness` work as root?
00:15nicolas17I enabled a keyboard shortcut for "toggle keyboard backlight" and pressed it, and that did turn it off
00:15nicolas17this might be a plasma settings thing
00:23nicolas17jannau: that echo command does turn it off, so yeah I blame plasma
00:27zshrc has joined #asahi
00:29zshrc has quit []
00:29jmalexan has joined #asahi
00:31jmalexan is now known as NotDisliked
00:32NotDislikedhello I've come to ask the questions that 50 people have probably already asked
00:33NotDislikedI saw that keyboard backlight support got added in linux-asahi-edge (which I'm now on), but I can't find a place in kde to control the keyboard brightness. am I blind or do I need to grab some extra module
00:35jannauthe keyboard backlight is also in the linux-asahi kernel. the "Battery and Brightness" tray item has controls for it. it might be hidden if charger is connected and the battery is full
00:35NotDislikedknew I was blind
00:36jannauno problem
00:36weitcis_ has quit [Ping timeout: 480 seconds]
00:37weitcis_ has joined #asahi
00:37NotDislikedmy other question is more niche, I was trying to modify the existing discord_arch_electron package in the AUR to compile for arm, I think I've got everything working fine but when I launch the app I get an error about "dri3 extension not supported". I just wanted to validate my suspicion that this is due to the lack of a GPU driver on asahi at present, or maybe someone happens to know that I've just messed something up in my
00:37NotDislikedfiddling with the package.
00:39NotDislikedthe full error is "[3498:1122/163056.590907:ERROR:gpu_memory_buffer_support_x11.cc(44)] dri3 extension not supported.", thus my suspicion about the GPU driver being the culprit
00:48user982492 has joined #asahi
00:55amarioguy2 has quit [Ping timeout: 480 seconds]
00:55amarioguy2 has joined #asahi
01:03vx has quit [Quit: G-line: User has been permanently banned from this network.]
01:03vx has joined #asahi
01:05amarioguy2 has quit [Ping timeout: 480 seconds]
01:07weitcis_ has quit [Quit: Konversation terminated!]
01:09axboejannau: I used to run with the out-of-tree patches for the kbd backlight and they worked fine, doesn't work for me in asahi-wip
01:09axboecaveat that this may be me missing something, but it seems to get killed at load time:
01:09axboe[ 0.188224] leds_pwm led-controller: failed to read period for kbd_backlight, default to off
01:14nicolas17hm that name sounds familiar :)
01:14nicolas17axboe: you work on block device stuff right?
01:15millenialhacker has joined #asahi
01:19axboenicolas17: yep, block stack, io_uring, bunch of different things
01:20axboebut I'm also an avid m1 linux user, and get to be the newbie here :)
01:21nicolas17yesterday I was looking for a way to write to a file and say "delay flushing this to disk (unless you're low on memory or something) because I'll delete this file soon anyway"
01:22nicolas17Windows has this flag on the CreateFile API: "Specifying the FILE_ATTRIBUTE_TEMPORARY attribute causes file systems to avoid writing data back to mass storage if sufficient cache memory is available, because an application deletes a temporary file after a handle is closed. In that case, the system can entirely avoid writing the data."
01:22nicolas17FreeBSD has this in mmap: "MAP_NOSYNC: Causes data dirtied via this VM map to be flushed to physical media only when necessary (usually by the pager) rather than gratuitously. Without this option any VM pages you dirty may be flushed to disk every so often (every 30-60 seconds usually)"
01:22nicolas17there doesn't seem to be anything like it on Linux, is there?
01:23millenialhacker has quit [Ping timeout: 480 seconds]
01:25stsmwg has joined #asahi
01:26stsmwg has quit []
01:28stsmwg has joined #asahi
01:32NotDisliked has quit [Quit: Konversation terminated!]
01:35rvalue has quit [Quit: ZNC - https://znc.in]
01:40vx has quit [Quit: G-line: User has been permanently banned from this network.]
01:40vx has joined #asahi
01:52tpw_rulesjannau: ah, thanks for the info
01:56stsmwg has quit [Quit: Lost terminal]
02:36axboenicolas17: don't think so, nope
02:52millenialhacker has joined #asahi
02:54mac456 has joined #asahi
03:00millenialhacker has quit [Ping timeout: 480 seconds]
03:04chadmedaxboe: sorry to be a bother but if youre still interested in speakers on your j316, do you mind giving the latest asahi-audio a try with the latest kernel? we've fixed the problem with weird channel allocations ruining pipewire's routing so im curious as to what the sound quality is like on the 16"
03:05chadmedi havent had any feedback from 16" owners yet so idk if i need to acquire one to do new IRs for it or if we actually can get away with the 14" ones with reasonable results
03:05chadmedi feel like its going to be the former
03:15mac456 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
03:16nicolas17chadmed: is there M2 speaker support yet?
03:17chadmedno, the 13" machines use a totally different setup
03:17chadmedi have an A1708 that i could use to get a rough idea of what I need to do for them but the kernel driver for the later intel macbooks is dodgy and unmaintained
03:18nicolas17so you've been working only on 14" so far?
03:19chadmedyeah, its the only machine i have and i dont really have the means to just accumulate macs
03:19nicolas17oh for sure
03:20chadmedluckily the desktops dont really _need_ anything just for basic functionality so i can ignore those for now, and if the same set of IRs works fine on the 16" then thats the main devices sorted (hopefully) until the next redesign
03:21nicolas17all desktops have speakers, right?
03:21chadmedim banking on them finally retiring the 13" chassis after this generation. my hunch is that they just had a shitload of inventory or something or havent finished amortising the tooling for making them. but given the macbook air redesign and how fish-out-of-water the thing looks now i cant see them continuing the line
03:21chadmedyeah all the desktops have speakers
03:22chadmedthe only one im really interested in tuning is the imac because i think that has some crazy setup similar to the laptops, but the studio and mini are just a mono woofer cone. the most id need to do there is wrench down the mids to stop it from honking
03:22chadmedand ofc safety stuff
03:23nicolas17"pandemic-affected work products finally getting to the end of the pipeline" is another theory I have seen mentioned for weird product decisions, especially with respect to the latest iPads
03:24chadmedcertainly a possibility
03:25chadmedmy thinking is that the last line of intel 13" machines sold poorly because of their awful reputation so they are/were overstocked on chassis and tooling
03:25chadmedso why not chuck decent hardware in them and ship them out the door rather than go to the expense of melting them down/otherwise disposing of them
03:25user982492 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
03:25nicolas17oh did M1 MBP use the same chassis as Intel? I knew M1 and M2 were almost identical but didn't know last Intel too
03:26chadmedyeah the 13" chassis hasnt changed since like 2016 afaik
03:28beeblebrox has quit [Quit: Konversation terminated!]
03:29nicolas17 has quit [Quit: sleep]
03:35axboechadmed: not a bother at all! I can give it a whirl
03:35chadmedthanks!
03:35chadmedah you will need lsp-plugins installed too
03:36axboeinstalled
03:36axboeI ran mac-audio.sh
03:36chadmedyoull need to log out and log back in to restart pipewire
03:37ciggi has quit [Read error: Connection reset by peer]
03:37ciggi has joined #asahi
03:38axboeI just rebooted to be sure
03:38axboenot sure if I can tell a difference, but it worked pretty well before
03:38axboebut I also have no idea what osx sounds like...
03:39chadmedoh so the issues you had initially were already gone?
03:39axboemaybe lacking some woofer, not sure
03:39axboethe half-speed stuff? yes that got magically fixed
03:40axboeand for some reason I ended up with "Pro Audio" and "Pro Audio 1"
03:40chadmedpro audio is the headphone jack, pro audio 1 is the raw speaker device
03:40chadmedso you can swap between pro audio 1 and macbook pro j31x to a/b test the difference
03:40axboemakes sense
03:40chadmedyoull want to keep pro audio 1 at 100% volume for normal use too and just use the filter chain to set the volume
03:41chadmedand also max out all the speaker sliders in alsamixer now, you dont need to keep them at 80 or whatever
03:41axboeit sounds like pro audio 1 is missing tweeters
03:41axboeand internal is just tweeters?
03:41axboeah perfect, I'll max them out - had them at 78 or something
03:43axboethis is way better than previously
03:43axboewas already pretty happy with it :)
03:43chadmedthats what i like to hear!
03:43chadmedare the mids still honky or has that been fixed?
03:44axboethanks so much for all your (and povik, et al) work on this
03:44axboeI'd need to try a bit of booting between osx and linux to check
03:44axboeI've never run osx on it
03:44axboebut definite difference between pro audio 1 and internal
03:44chadmedthats impressive
03:44axboelatter more legible imho
03:45axboepro audio 1 sounds somewhat muffle
03:45chadmedmacos still has better bass management but thats because they do a bunch of freaky stuff in coreaudio we cant really emulate (not that id want to, i think the bass in macos is a bit overbearing)
03:45chadmedyeah the raw speakers actually kinda suck, theres a boatload of EQ happening in the IRs
03:45axboefor me it's just a laptop and the audio will be so-so at best anyway
03:45chadmedthe woofers sound like an AM radio station
03:45axboeyeah that's basically what it sounds like
04:27mac456 has joined #asahi
04:33weitcis has joined #asahi
04:38weitcis has quit [Quit: Konversation terminated!]
04:48pthariensflame has joined #asahi
04:54pthariensflame has quit [Quit: Textual IRC Client: www.textualapp.com]
04:57ten157 has joined #asahi
04:57marvin24_ has joined #asahi
05:00marvin24 has quit [Ping timeout: 480 seconds]
05:02ten15 has quit [Ping timeout: 480 seconds]
05:02ten157 is now known as ten15
05:28Stroller has quit [Quit: Stroller]
05:45mac456 has quit [Ping timeout: 480 seconds]
05:49millenialhacker has joined #asahi
06:01millenialhacker has quit [Ping timeout: 480 seconds]
06:13tmlind has quit [Quit: leaving]
06:14tmlind has joined #asahi
06:14nickchan has joined #asahi
06:14nickchan has left #asahi [#asahi]
06:21Graypup_ has quit [Quit: meow]
06:21Graypup_ has joined #asahi
06:29lethalnyan has quit [Quit: kill -9 -1]
06:29lethalbit has joined #asahi
06:38nickchan has joined #asahi
06:38nickchan has left #asahi [#asahi]
06:44fossdd has joined #asahi
06:46Treibholz has joined #asahi
06:49SSJ_GZ has joined #asahi
07:04chadmed has quit [Quit: Konversation terminated!]
07:04chadmed has joined #asahi
07:05millenialhacker has joined #asahi
07:13millenialhacker has quit [Ping timeout: 480 seconds]
07:18guillaume_g has joined #asahi
07:34jannauaxboe: "leds_pwm led-controller: failed to read period for kbd_backlight, default to off" seems to be not a problem. I get that as well but the kbd backlight works as expected
07:35Graypup_ has quit [Quit: meow]
07:35turo has quit [Remote host closed the connection]
07:35Graypup_ has joined #asahi
07:35millenialhacker has joined #asahi
07:35turo has joined #asahi
07:53millenialhacker has quit [Ping timeout: 480 seconds]
07:57Graypup_ has quit [Ping timeout: 480 seconds]
07:57turo has quit [Ping timeout: 480 seconds]
07:57tpw_rules has quit [Ping timeout: 480 seconds]
07:58reillyeon has quit [Ping timeout: 480 seconds]
08:00Graypup_ has joined #asahi
08:00reillyeon has joined #asahi
08:01turo has joined #asahi
08:02millenialhacker has joined #asahi
08:04tpw_rules has joined #asahi
08:05tobhe has joined #asahi
08:10tobhe_ has quit [Ping timeout: 480 seconds]
08:10millenialhacker has quit [Ping timeout: 480 seconds]
08:20SSJ_GZ has quit [Remote host closed the connection]
08:20SSJ_GZ has joined #asahi
08:21millenialhacker has joined #asahi
08:24tobhe_ has joined #asahi
08:25ggardet has joined #asahi
08:27Graypup_ has quit [Ping timeout: 480 seconds]
08:27turo has quit [Ping timeout: 480 seconds]
08:29reillyeon1 has joined #asahi
08:29reillyeon has quit [Ping timeout: 480 seconds]
08:29reillyeon1 is now known as reillyeon
08:29tpw_rules has quit [Ping timeout: 480 seconds]
08:29turo has joined #asahi
08:29tobhe has quit [Ping timeout: 480 seconds]
08:29millenialhacker has quit [Ping timeout: 480 seconds]
08:29Graypup_ has joined #asahi
08:30tpw_rules has joined #asahi
08:33guillaume_g has quit [Ping timeout: 480 seconds]
08:46capta1nt0ad has quit [Read error: Connection reset by peer]
09:02mlq has joined #asahi
09:09zimsneexh has joined #asahi
09:10millenialhacker has joined #asahi
09:14hendry has quit [Quit: WeeChat 3.7.1]
09:18millenialhacker has quit [Ping timeout: 480 seconds]
09:21guillaume_g has joined #asahi
09:28ggardet has quit [Ping timeout: 480 seconds]
09:38bluetail9 has joined #asahi
09:39MajorBiscuit has joined #asahi
09:40a_plastic_bag has joined #asahi
09:44leitao has joined #asahi
09:46bluetail has quit [Ping timeout: 480 seconds]
09:46bluetail9 is now known as bluetail
09:46a_plastic_bag has quit [Quit: WeeChat 3.7.1]
09:54leitao has quit [Ping timeout: 480 seconds]
10:12rvalue has joined #asahi
10:23millenialhacker has joined #asahi
10:25Misthios has quit [Quit: Misthios]
10:26Misthios has joined #asahi
10:30zimsneexh has quit [Quit: zimsneexh]
10:31millenialhacker has quit [Ping timeout: 480 seconds]
10:32Misthios has quit [Quit: Misthios]
10:33Misthios has joined #asahi
10:36Kasper has joined #asahi
10:46Kasper has quit [Remote host closed the connection]
10:49Misthios has quit [Quit: Misthios]
10:55Misthios has joined #asahi
10:56roy has joined #asahi
10:57CatDentures has joined #asahi
11:13roy has quit [Remote host closed the connection]
11:44rvalue has quit [Quit: ZNC - https://znc.in]
11:55rvalue has joined #asahi
12:11jvzr has joined #asahi
12:18jvzr has quit [Remote host closed the connection]
12:25millenialhacker has joined #asahi
12:31cylm has joined #asahi
12:33millenialhacker has quit [Ping timeout: 480 seconds]
12:39ten15 has quit [Ping timeout: 480 seconds]
12:39weitcis has joined #asahi
12:45pjakobsson has joined #asahi
12:46weitcis has quit [Remote host closed the connection]
13:30Stroller has joined #asahi
13:39jvzr has joined #asahi
13:42millenialhacker has joined #asahi
13:50millenialhacker has quit [Ping timeout: 480 seconds]
13:56Heyb4r has joined #asahi
13:58millenialhacker has joined #asahi
13:59axboejannau: ah ok, well then it's something else causing it to not work, because it doesn't :)
13:59axboewhereas the old patch worked fine
14:00j`eycan you check the DT, make sure there wasnt a mistake made there?
14:06millenialhacker has quit [Ping timeout: 480 seconds]
14:13axboesure will do
14:14axboeI did look when I switched to the in-tree one and didn't spot anything, but it's easy to miss so I'll do another check
14:22Heyb4r has quit [Remote host closed the connection]
14:22Heyb4r has joined #asahi
14:24MajorBiscuit has quit [Ping timeout: 480 seconds]
14:30Heyb4r has quit [Ping timeout: 482 seconds]
14:32jannauaxboe: does /sys/class/leds/kbd_backlight/brightness exists? does anything happens when writing 255 into it?
14:40aykeIs it possible to have display brightness control while using the old framebuffer instead of the new DCP display support? (I'm guessing "no" but asking to be sure)
14:40Heyb4r has joined #asahi
14:40aykeDCP is unfortunately too slow for my workflow
14:42Heyb4r_ has joined #asahi
14:42Heyb4r has quit [Read error: Connection reset by peer]
14:44jannauit's not possible without larger hacks. the brightnes control is integrated in the dcp protocol. you have to wait for the gpu driver
14:44MajorBiscuit has joined #asahi
14:46as400dcp is fair bit faster with x11
14:47mpsas400: which DE you use
14:47as400mps: plasma
14:48mpsfor me it is fast enough, except when switching virtual screens in awesome wm
14:48as400mps: you're using xfce, right ?
14:49as400oh crap - awesome
14:49mpsright, awesome window manager (so strange generic name)
14:50Heyb4r_ has quit [Ping timeout: 480 seconds]
14:54Heyb4r has joined #asahi
14:55as400mps: awesome works only on x11, right ?
14:56j`eyas400: yeah, you could try sway for a tiling wayland wm
14:58as400j`ey: I tried sway some time before. I really liked it but with the time I started missing some stupid things that are quite easy to do in plasma. Like screenshots, keyboard backlight controls oob.
14:58as400You know - I'm lazy I guess
14:58j`eyIm sure you can do those things with sway
14:59aykejannau: ok, thnx
14:59as400j`ey: most certainly you're right.
15:00mpsaxboe: right
15:00Heyb4r has quit [Remote host closed the connection]
15:01aykeas400: I'm using KDE on X11, my problem is yakuake takes about a second to appear. And I'm too lazy to try and fix that when this will likely be fixed anyway in the not too distant future (I hope!)
15:01Heyb4r has joined #asahi
15:02mpsI use awesome wm for more than a decade and like it because I can config it to behave as I want, and it is easy to add applets with few lines of lua code, plus keybaord shortcuts
15:04as400mps: I used sway for a short time (like a month). But I must admit it was hard to go back to plasma.
15:04as400I got used to sway very quickly
15:09Heyb4r has quit [Ping timeout: 480 seconds]
15:18Heyb4r has joined #asahi
15:20MajorBiscuit has quit [Ping timeout: 480 seconds]
15:21delsol has joined #asahi
15:24Heyb4r_ has joined #asahi
15:24Heyb4r_ has quit [Read error: Connection reset by peer]
15:25sentinal8473 has joined #asahi
15:26Heyb4r has quit [Ping timeout: 480 seconds]
15:29BlueIn2Red has joined #asahi
15:29rvalue has quit [Remote host closed the connection]
15:30rvalue has joined #asahi
15:31panzone has joined #asahi
15:31MajorBiscuit has joined #asahi
15:33sentinal8473 has quit [Quit: Textual IRC Client: www.textualapp.com]
15:34beeblebrox has joined #asahi
15:40panzone has quit [Ping timeout: 480 seconds]
15:43millenialhacker has joined #asahi
15:47kujegersway and DCP works fine for me, just had to disable all transluceny effects in e.g. alacritty, or it would render at less than 1fps
15:57BlueIn2Red has quit [Quit: Konversation terminated!]
15:58millenialhacker has quit [Ping timeout: 480 seconds]
15:58millenialhacker has joined #asahi
15:59zimsneexh has joined #asahi
15:59bluetailkujeger thats weird that you say that since translucency seemed to not affect my m1 mac mini at all with the default kde plasma installment
16:00bluetailbut I forgot I dont know about what the DCP support changes
16:07j`eybluetail: changes to use the hardware display processor, rather than a 'dumb' framebuffer
16:26millenialhacker has quit [Ping timeout: 480 seconds]
16:28kujegeryeah, the translucency works fine without the DCP changes. I'm guessing that once the gpu driver appears, it'll work fine with dcp as well.
16:29kettenis_ has quit [Remote host closed the connection]
16:34LinuxM1 has joined #asahi
16:36panzone has joined #asahi
16:36panzone has quit []
16:52guillaume_g has quit []
17:32MajorBiscuit has quit [Ping timeout: 480 seconds]
17:33panzone has joined #asahi
17:34panzone has quit []
17:42ah_ has quit [Quit: Lost terminal]
17:44nicolas17 has joined #asahi
17:50zimsneexh_ has joined #asahi
17:51LinuxM1 has quit [Quit: Leaving]
17:54panzone has joined #asahi
17:56aykeHmm, interesting. I disabled translucency in yakuake but I guess some part of it still does the transparency operation (but now with opacity=100%)
17:56zimsneexh has quit [Ping timeout: 480 seconds]
17:56zimsneexh_ is now known as zimsneexh
18:10Stroller has quit [Quit: Stroller]
18:15millenialhacker has joined #asahi
18:22jamespmorgan has joined #asahi
18:26zotanmew has quit [Quit: goodbye: ~zotan via destiny.zotan.network]
18:28jamespmo_ has quit [Ping timeout: 480 seconds]
18:29zotanmew has joined #asahi
18:55lewurm has quit [Quit: ZNC 1.8.2+deb2build5 - https://znc.in]
18:56millenialhacker has quit [Read error: Connection timed out]
18:57zimsneexh has quit [Quit: zimsneexh]
18:59millenialhacker has joined #asahi
19:01lewurm has joined #asahi
19:13millenialhacker has quit [Ping timeout: 480 seconds]
19:18___nick___ has joined #asahi
19:18___nick___ has quit []
19:18riker77 has quit [Quit: Quitting IRC - gone for good...]
19:20___nick___ has joined #asahi
19:27riker77 has joined #asahi
19:40panzone has quit [Remote host closed the connection]
19:40panzone has joined #asahi
19:40panzone_ has joined #asahi
19:43psykose has joined #asahi
19:45Stroller has joined #asahi
20:03millenialhacker has joined #asahi
20:08riker77 has quit [Quit: Quitting IRC - gone for good...]
20:11psykose has quit [Ping timeout: 480 seconds]
20:11millenialhacker has quit [Ping timeout: 480 seconds]
20:19riker77 has joined #asahi
20:36panzone_ has quit [Remote host closed the connection]
20:36panzone_ has joined #asahi
20:38millenialhacker has joined #asahi
20:42panzone has quit [Ping timeout: 480 seconds]
20:44panzone_ has quit [Ping timeout: 480 seconds]
20:46millenialhacker has quit [Ping timeout: 480 seconds]
20:52panzone has joined #asahi
20:53panzone1 has joined #asahi
20:56psykose has joined #asahi
20:56jluthra has quit [Remote host closed the connection]
20:57jluthra has joined #asahi
21:08___nick___ has quit [Ping timeout: 480 seconds]
21:08giskard has quit [Read error: Connection reset by peer]
21:09leitao has joined #asahi
21:10leitao has quit []
21:20mac456 has joined #asahi
21:21Stroller has quit [Ping timeout: 480 seconds]
21:24mac456What are all the /dev/nvme0n* devices on j413 (M2 Air)? I have nvme0n1, nvme0n2, and nvme0n3.
21:25mac456What are 2 and 3? I thought there was only one hard drive.
21:27tpw_rulesnvme namespaces. i believe they are used by the core boot and firmware stuff
21:29mac456What are nvme namespaces? It's a single nvme chipset, right? But the kernel module exposes it to Linux kernel as 3 separate devices?
21:30tpw_rulesyes, the nvme spec lets you divide the physical drive at some level lower than partitions
21:31mac456tpw_rules Thanks!
21:34mac456And the mac itself also has some "hidden" area of the SSD, right? Or is that just one of these alternate namespaces?
21:35tpw_ruleswhat do you mean? the nvme0n2 and nvme0n3 namespaces are small and used by very core apple boot software. i don't recall which off the top of my head. the higher level boot stuff including your macos install and the mandatory boot and recovery partitions are just gpt partitions on nvme0n1
21:36millenialhacker has joined #asahi
21:36tpw_rulesnone of the ssd is hidden to linux. and you can indeed stop the machine from booting by disturbing the stuff needed for boot.
21:37delsol^^^ THIS. The extra namespace stuff provided by NVME just lets apple/everyone put the other (boot/recovery/firmware/driver/tool options) in extra space.
21:37mac456Ok, I get it. I had just heard there was a hidden partition (hidden from macos), but I think that is referring to nvme0n2/3.
21:38tpw_rulesmacos is pretty cagey about letting you access arbitrary partitions
21:39mac456One thing that confuses me. The SSD on j413 is supposed to be 256 GB. But I only see 251 GB (nvme0n1) + 1 MB (nvme0n2) + 128 MB (nvme0n3). Where is the other 5 GB?
21:39dbadams has joined #asahi
21:39mac456(I mean GB as in base-10, not GiB as in base 2)
21:39tpw_rulesi don't know how the alternate namespaces show up in macos
21:39panzone1 has quit [Ping timeout: 480 seconds]
21:39delsolGB vs GiB vs reserved space?
21:39tpw_rulesvs marketing
21:39panzone has quit [Ping timeout: 480 seconds]
21:40rayanamukami3 has quit [Quit: https://www.youtube.com/watch?v=BA11qbcnPc0]
21:40delsolJust like the the old "internet recovery" mode on other machines, where you could remove the hard drive/ssd.... and still boot the machine with a basic kernel, connect to ethernet or wifi, download and install over the internet onto your replacement drive... even if it took allllll daaaaammmmmnnnnnnn daaaaaaaaaaaayyy.
21:40mac456I don't think it's marketing. The drive is called AP0256Z, that's not a marketing term.
21:41delsolmac456: perhaps its 256GB of NAND, but some of it is treated as SLC instead of MLC/TLC/QLC for write performance.
21:41delsolreducing capacity a bit
21:41mac456delsol: You're saying there's an additional 5 GB on the SSD devoted to Internet Recovery and Linux can't see it?
21:42delsolmac456: No.. I'm saying that the older machines had extra storage where that shit was, and that rather than having to build in a "fake bootable USB stick" or other workarounds for that kind of shit, now they can just use extra bits of the NVME namespace to do similar.
21:42delsoldells with their hardware function test stuff that was built into EFI/BIOS/etc
21:43tpw_rulesi think that's not a good usecase imo
21:43tpw_ruleson the old machines it was a physically separate chip so you could indeed remove the drive
21:43mac456delsol: And you can't see SLC in linux?
21:43tpw_rulesnow if the drive dies or you take it out the namespaces go too and your machine is hosed
21:43mac456What I don't get is why you can't see the full 256 GB in Linux? There seems to be 5 GB missing. Seems like a problem with the driver.
21:44millenialhacker has quit [Ping timeout: 480 seconds]
21:44tpw_rulesdelsol is saying there may be 256GB of physical chips, but it's not all given to the processor by the nvme controller
21:44delsolmac456: OK... so do you know how SLC/MLC/TLC/QLC works?
21:44mac456Oh, ok, I see what you mean
21:44delsolSLC stores 1 bit per cell.
21:44tpw_rulesand one way is that the controller can trade off performance vs capacity by in some sense double-storing data
21:44mac456delsol: no
21:44delsolfull empty.
21:44mac456Yes, I know that
21:44mac456Yes, I know that
21:44mac456SLC is faster
21:44mac456I don't know MLC, TLC, QLC
21:45delsolMLC stores TWO bits per cell... glass is empty, full, 1/3 or 2/3rds full.... 00, 01, 10, 11
21:45delsolTLC stores 3 bits per cell.... QLC 4 bits per cell.
21:45tpw_rulesbut there is only a fixed number of cells which apple might quote as 256GB if used in QLC mode. the controller can write faster by trying to cram less bits into each cell
21:45delsolif you have 8GB of QLC.... and you use it as SLC... thats only 1 GB of space... in that 8GB chunk...
21:45delsolso you just "lost" 7GB.
21:46delsolbut you get to read/write/erase it 100000 cycles instead of 500.
21:46mac456Yeah, I see your point. So, the 5 GB could be more redundant for better longevity and store firmware needed to boot the machine even when the rest of the SSD has failed?
21:47delsolno. You use it as a write cache
21:47tpw_rulesno, that stuff is stored on a separate chip
21:47delsolscratch space to combine writes, shit to disk quick, and then re-shuffle in place elsewhere later.
21:47delsolsince you can't full speed "edit" whats stored on NAND.
21:48panzone has joined #asahi
21:48delsolyou read it, copy it elsewhere, erase it in larger chunks, then write the new contents back.
21:48delsolvery slow.
21:48delsolhaving small space to shit-writes down to very fast makes a world of difference.
21:48mac456Ok, so now I think you are saying the extra 5 GB not exposed to the AP could be a write cache for writes to the rest of the SSD?
21:49mac456Ok, I see what you mean. That makes sense.
21:49delsolAlso, usable space vs raw space
21:49delsolit may be keeping a small reserve for reallocations and cell failures and wear leveling
21:49cylm has quit [Ping timeout: 480 seconds]
21:50mac456Well, there are spares no matter what, independent of what the stated chip capacity is.
21:50delsol... not always
21:50mac456Ok, yeah, could be.
21:51delsolSee all the 256GB SSD's that show up as 228/229/230gb
21:51delsolwhen it should be "238" GB
21:51mac456Mmm, see what you mean, yeah.
21:52delsolbecause of the spare 8-10GB being held back for reserve capacity.... (and that was even back in the MLC days before using SLC caches)
21:53delsolI did the same back then too. Had plenty of drives that didn't have shit for spare capacity... just always left an extra 4-8-10GB unpartitioned at the end for the drive to wear level with as it pleased.
21:53delsolSeemed to make a difference in lifespan
21:54mac456Interesting. Thanks - this illuminated a lot. Gotta run but thanks for the replies. Talk to you later.
21:55mac456 has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
21:56panzone has quit [Ping timeout: 480 seconds]
21:57panzone has joined #asahi
21:57panzone_ has joined #asahi
21:59Stroller has joined #asahi
22:04millenialhacker has joined #asahi
22:05panzone has quit [Ping timeout: 480 seconds]
22:05panzone_ has quit [Ping timeout: 480 seconds]
22:06panzone has joined #asahi
22:12panzone_ has joined #asahi
22:16panzone has quit [Ping timeout: 480 seconds]
22:19millenialhacker has quit [Ping timeout: 480 seconds]
22:36psykose has quit [Remote host closed the connection]
22:37millenialhacker has joined #asahi
22:42jvzr has quit [Remote host closed the connection]
22:45millenialhacker has quit [Ping timeout: 480 seconds]
22:48millenialhacker has joined #asahi
22:53rayanamukami3 has joined #asahi
22:55panzone_ has quit [Remote host closed the connection]
22:56millenialhacker has quit [Ping timeout: 480 seconds]
23:12osaka1990 has quit [Remote host closed the connection]
23:16CyrinuxHi everyone, do you know if someone is working on webcam support? Or is it planned, complicated work?
23:16millenialhacker has joined #asahi
23:22SSJ_GZ has quit [Ping timeout: 480 seconds]
23:25millenialhacker has quit [Ping timeout: 480 seconds]
23:37osaka1990 has joined #asahi
23:37delsol has quit [Remote host closed the connection]
…