#
…

2023-01-14

00:02cylm has joined #asahi
00:05TheDcoder has joined #asahi
00:05TheDcoderHi, I'm new to Mac and I want to try Asahi Linux, but I have a few questions before I begin my journey :)
00:06TheDcoderHow do I make a complete backup of my MacBook (Air) so that I can restore it to the current state at a future date?
00:07TheDcoderYes but that was many months ago, so let me take a look again...
00:07tpw_rulestime machine is a decent way imo. there is other paid software that can do it
00:07j`eyfor backups you can use timemachine
00:07chadmedtime machine is probably the most seamless way to do so
00:07tpw_rulesnote that restoring it from complete disaster might require the use of DFU mode which requires another computer running linux or macos
00:08TheDcoderI'm looking for a more low-level backup, something between block level and FS, maybe something like partclone?
00:09tpw_rulesmacos doesn't really let you do that
00:09TheDcoderDFU mode sounds interesting
00:09chadmedi dont think thats really advisable given how the security model works too
00:09TheDcodertpw_rules: Oh, why is that? No root access?
00:10TheDcoderchadmed: Hmmm... I see. I do recall being forced to connect to the internet in order to factory reset my Mac.
00:10TheDcoderWhere can I learn more about this security stuff?
00:10tpw_rulesi don't guarantee it's impossible but it really goes against apple's whole model and it would be very difficult to actually restore from
00:10chadmedthats not the blocker, the sep wouldnt even let you boot if you mashed up all the volume keys and uuids by trying to do a block level restore without updating the boot policy
00:10chadmedwhich you wouldnt be able to do because doing a block level restore would mash up all the keys and prevent you from booting :P
00:11tpw_ruleswhy would that scramble keys actually?
00:11TheDcoderchadmed: I would backup the partition table as well, and that should preserve the uuids I think.
00:12tpw_rulesi do know that macos cannot write the partition table
00:12TheDcoderAnyway... does the MacBook have any other storage outside of the usual SSD?
00:12TheDcoderSomething to preserve security info? That's what I'm afraid of
00:13chadmedthe sep
00:13TheDcoderIn my good ol' x86 laptop I can just swap drives
00:13chadmedwhat exactly exists in your machine state that you cant preserve with time machine at the moment
00:13tpw_rulesi can vouch that it will restore properly
00:13TheDcoderwhat's sep?
00:13chadmedsecure enclave
00:14chadmedapples security coprocessor
00:14TheDcoder...is that like TPM?
00:14chadmedkinda but also not at all :P
00:14chadmedits much much much much better
00:14chadmedyknow what let me reboot into macos and hammer out this apple-security-for-dummies thing ive had in the works for about three months now
00:14TheDcoderI don't really have a reason, I just want to do it if possible, I do it with all new PCs I get.
00:15chadmedwell to be blunt these arent pcs and you shouldnt treat them as such
00:15TheDcoderchadmed: That would be awesome!
00:15TheDcoderMy MacBook isn't a PC?!
00:15chadmedunder the hood they look almost nothing like what one would consider a Personal Computer
00:15tpw_rulesIBM Personal Computer*
00:16chadmedtheyre close to an iphone in terms of platform architecture than they are to an ibm compatible
00:16tpw_rulesDFU will restore it to factory state if you want. there's a recovery partition but DFU will reinstall it
00:16tpw_rulesit's not possible to brick and there's no information on it from the factory really
00:17tpw_rulesnor is there interesting crapware on it :P
00:17chadmedin any case, theres really no point trying to preserve the state of the partition table or whatever on these machines because not only does it violate the security model but time machine backups will preserve the state of the APFS container youre backing up (which in a standard install is the only disk)
00:17TheDcoder...understood.
00:17tpw_rulesand contains all the state apple lets you change from macos
00:17chadmedthe system logical volume is sealed with a key and readonly so its not like you can clobber your filesystem so badly that the system becomes unbootable unless youre really _really_ trying
00:17TheDcoderWhat's the DFU?
00:18j`eyDevice Firmware Update
00:18TheDcoderchadmed: That's reassuring.
00:18TheDcoderj`ey: Oh, so is it the update that Mac is pestering me about? (I haven't updated yet)
00:19tpw_rulesno. in DFU mode the machine boots from ROM and downloads another OS over USB from a computer running idevicerestore or apple's equivalent. it's what will let you reinstall from a completely corrupted internal disk
00:20tpw_rulessadly internet recovery is no more, but as long as the disk isn't physically damaged that's how you factory restore it
00:21TheDcoderThere's a ROM?!
00:21chadmedtheres actually a few
00:21TheDcoderOh wait I guess that's just the BIOS/Firmware
00:22chadmedthe soc has a rom built in with the first stage bootloader on it, which loads the second stage bootloader from a flash rom on the board
00:23TheDcoderso can I restore my Mac from a completely corrupt internal storage with the DFU and an OS image without the internet?
00:23TheDcoderchadmed: Ah I see...
00:23chadmedi think you can feed an IPSW into idevicerestore
00:23TheDcodertpw_rules: re internet recovery, is that the same thing as the "connect to internet to factory reset" that I went through?
00:24tpw_rulesyeah, you can do the restore without internet
00:24tpw_rulesbut once the machine is restored it has to check in with apple over the internet once to activate during the first boot setup
00:24TheDcoderbut why? :(
00:24tpw_rules(presumably so you cannot DFU away e.g. the theft protection features)
00:24TheDcoderI don't want my mac calling home
00:25tpw_rulesit only has to do it once and only if you zorch it so bad you have to DFU restore
00:25TheDcoder...can I install Linux directly via the DFU?
00:25chadmed_ has quit [Remote host closed the connection]
00:25TheDcoderif I don't want to use Mac OS that i
00:25chadmedapple signs and encrypts everything
00:26chadmedalso we do not support removing macos at this time
00:26chadmedyou should absolutely 100% keep a small macos install on your machine
00:26TheDcoderYes I'm aware of that
00:26TheDcoderbut theoretically it's possible to remove Mac, right?
00:26tpw_rulesanyway internet recovery was a rather swanky feature on x86 macs where it could boot the macos installer over the internet even with a completely blanked disk
00:26chadmedtheoretically yeah
00:26chadmedbut not via dfu restore
00:26TheDcoderunderstood...
00:26chadmedthe machine checks that the image you feed it is signed with apples keys etc
00:27TheDcoderchadmed: can you share your "apple security for dummies"? that would be a great read for me :)
00:27chadmedyeah ill try and finish it today
00:27TheDcoderawesome... do ping my nick when you share it!
00:27chadmedand then get it peer reviewed of course :P
00:27TheDcoderI'll volunteer
00:28TheDcoderproof-read more like I guess
00:28chadmedi meant more for fact checking lmfao
00:28TheDcoderyes that why I'll just stick to proof-reading for spelling mistakes lol
00:28chadmedit should be pretty accurate but i am incredibly dense and its good to get a second opinion
00:28TheDcoderI don't mind honestly, the more the merrier
00:29TheDcodertpw_rules: I see... I wonder why don't still support it. It could come in handy still.
00:29chadmedthey do support it
00:29chadmedactually no, not in the same way
00:30TheDcoderbut tpw said it is not longer an option for M-series?
00:30chadmedyeah you need at least the system recovery volume intact
00:30chadmedits because iboot is incredibly dumb and simple to reduce the attack surface
00:30TheDcoderalso what file-system does Asahi use by default? Has anyone used FlashFS?
00:30chadmedwhereas uefi is enormous and bulky and you can include fancy features like that, but its full of security holes
00:30nicolas17I think it just uses ext4
00:31chadmedext4 is the default, f2fs doesnt really work on 16k page size kernels afaik
00:31chadmedi use btrfs because im that guy
00:31TheDcoderI meant F2FS
00:31nicolas17the SSD controller on the Mac is good enough, shouldn't need another layer of wear-leveling tbh
00:31TheDcoderchadmed: I am that btrfs guy too
00:32chadmedim also running gentoo so my opinion is worth 40% less
00:32TheDcodernicolas17: I hope so, I recall something about MacOS doing a huge amount of writes per day, not bad enough to trash the SSD but enough to have an impact
00:32tpw_rulesis there a chart of distro vs. opinion
00:32tpw_rulesthat was a bug that got fixed
00:33TheDcoderchadmed: I've been an Arch user for several years but Gentoo still scares me
00:33nicolas17well the Asahi distro is Arch-based :)
00:33TheDcoderI know, that's another reason to try it! :)
00:34TheDcoderand that reminds me, is Asahi Linux just Arch Linux ARM with custom repos?
00:34chadmedyeah pretty much
00:34j`eynot many custom packages
00:34chadmedbasically just the kernel and some patches to a few things that were broken
00:34TheDcoderso are the packages same as well? or are they built separately?
00:35chadmedno the custom repo is basically an overlay atop the standard alarm repos
00:35chadmedso most things are just the vanilla alarm packages from their repos
00:36TheDcoderthat's nice!
00:37TheDcoderI already use ALARM on my RPi so it will be a smooth transition
00:37nicolas17I found a tool that seems to do the USB proxying I wanted
00:37nicolas17but it needs the raw-gadget module which asahi doesn't enable in its kernel, so time to build my own kernel, hmph
00:39TheDcoderKind of off-topic, but is KDE still buggy? I've had some bad experiences using it a couple of years ago. I want to know if there are any known bugs currently for Asahi Linux
00:39chadmednope and wayland support is fantastic now
00:39TheDcodergood to hear, but I've never used wayland before... can I expect any hiccups compared to a traditional X desktop?
00:40chadmedX11 kinda just doesnt really work and most people have issues with it (the way this hardware works is not really conducive to X11)
00:40chadmednah, Xwayland does all the heavy lifting for the few remaining apps these days without good wayland support
00:40TheDcoderoh X11 on Mac has issues?
00:40Tramtristbingo chadmed .. been using wayland for 2 years now without problems
00:41TheDcoderDoes Wayland support On-Screen Keyboard? I use that sometimes.
00:41chadmedyeah the display controller just doesnt work in a way that X11 is built to handle properly
00:41chadmedpretty sure OSDs work in wayland yeah
00:41chadmedand again, anything thats not wayland-y just gets run through xwayland
00:41chadmedwhich is a tiny x11 server running under wayland
00:41nicolas17chadmed: "injecting keystrokes into other apps" seems like the kind of thing wayland would restrict by default, so it's a fair question
00:41TheDcoderthe problem with Xwayland I imagine is that programs not running under X are simply "invisible" to other X tools
00:42* TheDcoderuses several tools related to X automation
00:43TheDcoderspecifically xdotool's type command, and another program that I wrote to hide GUIs via hotkeys, I guess none of those would work with native Wayland GUIs
00:44TheDcoderAnyway... what about GNOME, any issues there?
00:47j`eyI think some people are using GNOME
00:47chadmedalyssa dogfoods her mesa work on gnome afaik
00:47hertz has joined #asahi
00:48TheDcoderOkay, so should I just stick to KDE then? I'm kind of a sucker for Gnome/GTK stuff. I use Xfce on my main PC
00:49chadmednah by "dogfoods" i mean she daily drives gnome while testing her development work
00:49chadmedit works ~as well as kde
00:49j`eyisnt fedora gnome by default?
00:50TheDcoderbut they offer "spins" with different DEs
00:50eroc1990 has quit [Remote host closed the connection]
00:50bps2 has quit [Ping timeout: 480 seconds]
00:50delsol has quit [Remote host closed the connection]
00:50TheDcoderthey also offer editions with packages pre-installed for a specific target audience
00:50eroc1990 has joined #asahi
00:50TheDcoderI forgot what they are called
00:51j`eywonder if Linus is using gnome then
00:52TheDcoderI'm pretty sure he does, an interview of his caught a glimpse of his monitor screen
00:52TheDcoderit was Gnome if I recall correctly
00:53TheDcoderCan I *not* use GRUB with Asahi Linux? I prefer rEFInd just because how simple it is to configure, and it looks nicer by default.
00:54j`eyyes, you dont *need* GRUB
00:54hertz has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
00:54chadmedi tried refind a while ago and it didnt work too well
00:54chadmedi now use systemd-boot
00:54j`eyI couldn't get refind working on aarch64 (in a vm)
00:54TheDcoderchadmed: I see, I don't mind systemd-boot either.
00:54chadmedwhich is just as simple to configure but looks nowhere near as nice as refind
00:55nicolas17TheDcoder: you should do a default installation first :P
00:56TheDcoderdoes systemd-boot provide a GUI? I thought it was just a stub to load the kernel asap.
00:56TheDcodernicolas17: Nah, I like adventures
00:56nicolas17well, you want to replace grub
00:56TheDcoderalso it's Arch Linux tradition to DIY the install
00:57chadmedi wouldnt call it a gui but there is a boot picker yes
00:57nicolas17what is the previous stage that loads and runs grub?
00:57TheDcoderthat's nice I guess
00:59tpw_rulessystemd-boot provides a list of entries
00:59nicolas17TheDcoder: after all, ARM Macs don't have UEFI built in ;)
00:59tpw_rulesand lets you edit the kernel command line
00:59tpw_rulesbut to describe it as a GUI or even TUI would be kind of tenuous
00:59chadmedim with nicolas17. maybe get a minimal asahi install on your system first before you go about trying to clobber it :P
00:59TheDcodertpw_rules: cool, just like GRUB then! 👍
00:59chadmedit is significantly uglier than grub lmfao
01:00TheDcoderchadmed: we will see, it will depend on how scared I get
01:00chadmedyou cant really DIY an install in any case, theres no ability to boot from a USB
01:00TheDcoderI prefer ugly and configurable over average and PhD-required-for-config
01:00tpw_rulesboot the reference distro without first using the asahi installer to set up a uefi environment*
01:01chadmedagain i will reiterate that most issues people end up having here arise from them expecting this platform to behave like a PC
01:01chadmedit doesnt
01:01TheDcoderchadmed: but I can modify the install script >:)
01:01TheDcoderokay, I will do some more research
01:01nicolas17pretty sure the install script copies a disk image into the new partition :P
01:02chadmedyeah the install script just dds an ext4 disk image into free space that it sets up after creating the APFS stub
01:02TheDcoderI'll have to create my own image I guess
01:02chadmed[11:01] <chadmed> again i will reiterate that most issues people end up having here arise from them expecting this platform to behave like a PC
01:02chadmed[11:01] <chadmed> it doesnt
01:03nicolas17that dd is done from inside macOS right?
01:03TheDcoderI might do that if I have enough patience
01:03j`eyTheDcoder: any reason you cant change it.. after you installed it?
01:03chadmedjust install asahi minimal which gives you a minimal ALARM install
01:03TheDcoderj`ey: I like adventures like I said
01:03chadmedand then customise from there
01:04j`eyI dont really see why it's less of an adventure to do the minimal install
01:04chadmedthat way you can at least get a feel for how the system is designed
01:04nicolas17TheDcoder: me too, which is why I recently built the kernel from source to try an untested patch from github
01:04chadmedand then have a go at trying to load up your own thing
01:04nicolas17but I did that *after* I had a working asahi system
01:04tpw_rulesif you want the usb boot experience, may i recommend nixos
01:04TheDcoderokay I'll just do the minimal install and remove GRUB after
01:05j`eyyou have to do most of what the minimal install does anyway!
01:05chadmedi didnt dare try to get gentoo working until i understood the default setup enough to know what was going on
01:05TheDcodernicolas17: nice 👍
01:05nicolas17TheDcoder: what component loads and runs GRUB? it seems like a bad idea to try to replace GRUB with something else without knowing that ;)
01:05chadmedand even then it relies on the minimal install to do all the platform specific stuff :P
01:05j`eynicolas17: u-nooy
01:05j`eyu-boot
01:06chadmedto paraphrase marcan, theres no point trying to reinvent this particular wheel
01:06tpw_rulesi think that was a quiz question
01:06TheDcodernicolas17: I don't know, some Mac thing I guess
01:06nicolas17j`ey: I was waiting for their wrong guess first! :D
01:06j`eyohh sorry
01:06TheDcoderchadmed: I still don't dare compiling my own kernel in gentoo on my PC
01:07tpw_ruleschadmed: do you have an installation guide or at least notes somewhere?
01:07nicolas17TheDcoder: the initial "Mac thing" doesn't support EFI
01:07chadmedyeah so again bluntly if you cant even do that then dare i say youre not ready at all to try and do a custom install on this platform :/
01:07nicolas17there's no EFI on these machines by default
01:07TheDcodernicolas17: so how does GRUB load?
01:08j`eyspoilers in chat :P
01:08TheDcoderchadmed: understandable, but one can hope...
01:08chadmedread this
01:08chadmedit has all the answers you seek
01:08nicolas17see that's why you should do a default install first ;) but if you want the dirty details there's the wiki
01:08TheDcoderchadmed: thanks, I was actually read that!
01:08seeeath has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
01:09chadmed oh yeah thats the one :P
01:09seeeath has joined #asahi
01:10TheDcoderthanks, reading all those articles
01:11TheDcoderTIL you can compile your own MacOS kernel
01:11chadmedthe open source components of XNU are extremely gimped since you dont get any of the apple secret source
01:11nicolas17recently some Apple hackers had to practically *beg* for the Kernel Development Kit to be released
01:12chadmediirc from my hackintosh days its basically _just_ enough XNU to dump you to a console
01:12nicolas17poking a security employee who would reply "oh wtf the new version is still not out? I'll go ping a senior manager I guess"
01:12TheDcoderwtf is XNU?
01:12chadmedNot UNIX
01:12chadmedits the name of apples kernel
01:12TheDcoderIsn't it based on Darwin or something?
01:12chadmedX is Not UNIX
01:13chadmedDarwin is the OS
01:13chadmedDarwin is to XNU as GNU is to Linux
01:14chadmedit's got an interesting history but its mostly useless as open source software
01:14TheDcoderThat's kind of confusing because GNU is like XNU
01:14TheDcoder*sounds like
01:14TheDcoderbut I get it
01:15nicolas17chadmed: did you see this desperate attempt at getting updated KDKs? https://twitter.com/_saagarjha/status/1592051045781606401
01:16TheDcodernicolas17: why is he paying $50?
01:17nicolas17he actually filed a TSI https://developer.apple.com/support/technical/ but I *think* he had one left over from his developer subscription so he didn't actually pay for it individually
01:17TheDcoderSecure Enclave Processor is optional and disabled by default? 🤔
01:18chadmedno its required and always on
01:18chadmedwe just cant access it from macos
01:18chadmeds/macos/linux
01:18chadmedamarioguy is on the case
01:18TheDcoderThe wiki says otherwise... "With the exception of the SEP (Secure Enclave Processor, a TPM equivalent), which is optional and disabled by default"
01:19chadmedwhich page
01:19chadmedoh i found it
01:19TheDcoderIntro to AS
01:21chadmedyeah so the SEP does stuff at boot time (enforces boot policies etc) then shuts off until its woken up again by the OS
01:21TheDcoderAS has not backdoor for real?! I thought we'd never see the day of modern processors without backdoors...
01:21TheDcoderchadmed: I see... got it.
01:21chadmedwell without getting out an SEM and account for every single transistor whos to say for certain
01:21chadmedbut these things are for all intents and purposes unbreakable
01:21marcantpw_rules: you cannot restore without the internet, it needs to get the tickets for your machine since it always restores in full security mode (that's not just the activation)
01:21TheDcoderwhat's an SEM?
01:22chadmedscanning electron microscope
01:22TheDcoderalso that's good enough I guess
01:22TheDcoderright... I recall Hector mentioning getting that
01:22marcanthe tickets are fetched by dedicated code in idevicerestore, while the activation stuff is just a TCP proxy thing
01:22chadmedthe bootloader is just too dumb to do anything other than go "yep this thing is signed by apple i will load it now"
01:22amarioguyTheDcoder: working on understanding the SEP init sequence btw if you're wondering how that's going
01:22marcanback when I was first fixing idevicerestore for AS, the TCP proxy thing was broken so I ended up with a successful but not activated restore and then got the phone home prompt on first boot (which went into recovery)
01:22marcanbut it definitely got the tickets to make it that far
01:24TheDcoderamarioguy: keep up the good work! 👍
01:24marcanTheDcoder: to answer your actual question about block-level restores: the SEP has a secure EEPROM that is used for anti-rollback, and a full block level restore is a massive rollback that will trip that and break xARTs which almost certainly means you can't access any of your data, and quite likely won't boot at all if the remote policy nonce has rolled over since the backup was taken
01:24TheDcodermarcan: it's the legend himself!... also does this mean you can spoof the home phone call with idevicerestore?
01:25marcani.e. this is how the FBI was cracking iPhones, by doing block level restores
01:25marcanso Apple fixed it
01:25marcanso you can't do it any more
01:25Zopolis4 has joined #asahi
01:25marcan(I wrote about this a while ago before they had this mitigation in)
01:26marcanyou can't spoof the call-home for activation, that's HTTPS. I think you used to be able to cache the tickets but I don't think that's the case any more, pretty sure there's nonces now
01:26amarioguymarcan: there's NVRAM vars you can set for AP side anyways :P
01:26chadmedyou used to be able to inject your own SHSH blobs and such to force a rollback but that went away like a decade ago at this point
01:26TheDcodermarcan: I see, thanks for the juicy details! I appreciate that Apple is preventing 3 letter agencies from accessing our data... at-least on a hardware level
01:27amarioguything is OTAs for AP side need to pin the nonce because you can't exactly request a ticket in iboot for those
01:27amarioguySEP side... it's complicated
01:27TheDcodertoo many tickets...
01:28amarioguyTheDcoder: in practice apple combines all the authorizations into one ticket
01:28TheDcoderalso does the phone home step require logging into an apple account or is that just a false memory I have?
01:28amarioguythe authorized install stuff is just an HTTP sequence
01:28marcanno, as long as the device isn't iCloud locked
01:29marcanthe tickets are just HTTP, the activation could fail and require iCloud login if the machine is locked
01:29marcan(that's the anti theft thing)
01:29marcanso many people have come here asking if Asahi could be installed on the cheap stolen machines they got off of ebay...
01:30TheDcoderso if I login with an iCloud account is it automatically locked?
01:30amarioguyTheDcoder: if you enable Find My
01:30amarioguythen it'll lock to your account
01:30marcanalso, as far as I know Apple are basically guaranteeing this phone home stuff works forever
01:30TheDcoderI did not if I recall correctly
01:30marcanIIRC it still works for the original iPhone, for the latest update available for it
01:30amarioguymarcan: just wait until TSS goes down :P
01:31marcanamarioguy: one would hope they'd publish generic tickets if they ever actually take it down...
01:31marcanthat would be one massive ewaste problem if they don't, not good for their image
01:31TheDcoderyes I agree
01:31marcan(pretty sure they *can* publish generic tickets that would eliminate the phone home requirement, if they chose to deprecate it)
01:31TheDcoderanyway, so are stolen macs just e-waste if not returned to apple or original owner?
01:31marcanwell, you can use them for parts
01:31chadmedyeah, the crooks dont care though because its basically a scam at this point
01:32marcanthey sell them to unsuspecting buyers
01:32chadmedtheyll keep stealing them to sell to poor undergrads who just want a laptop
01:32TheDcoderthat's good because I'm not a crook.
01:32marcanI actually got an iCloud locked motherboard recently, going to rip the SoC off and image it optically :p (and maybe with a SEM later)
01:32marcanthey're good for that at least!
01:32amarioguycould always unfuse one if you got the time and or cash :)
01:33marcanI have a SEM, not a FIB :p
01:33TheDcodercan you actually figure out the circuitry with an SEM?
01:33marcannot with the one I have access to at 5nm, no
01:33marcantoo small
01:33marcanbut we could get some pretty pictures of higher level structures
01:33TheDcoderwhat size are the transistors in AS?
01:33chadmedi think the feature size on tsmc n5 is like 18nm or something
01:34amarioguyusually "N5" and such doesn't track with actual feature size
01:34amarioguymore of a node change indication
01:34chadmedits all marketing wank
01:34marcangiven the images I got of a 40nm chip, pretty sure 5nm won't be reasonably resolvable, but I guess we'll see
01:34marcan40nm defintely was
01:34chadmedintel are spruiking process nodes in angstroms now lmfao
01:35TheDcoderisn't angstrom a really large unit?
01:35chadmedno its 0.1nm
01:36chadmed"yeah look how advanced our 18 angstrom node is! oh the feature size? well its umm like uhhjjj well its mhhhhhh 20nm gate to gate"
01:36chadmedits the ship that made the kessell run in less than 12 parsecs etc etc
01:36amarioguychadmed: they can't even launch a server processor without high numbers of steppings...
01:36marcan40nm: gate pitch 168nm. 5nm: gate pitch 50nm
01:36marcanokay *maybe* I'll be able to see something?
01:37marcanwe'll see
01:37marcanpretty sure that SEM could use some repair too
01:37TheDcoderI hope so!
01:37amarioguyfrom what i've heard SEMs take a lot of maintenance
01:37chadmedamarioguy: i mean id rather them tape out something that works than release A0 silicon with broken units
01:37marcanthis one hasn't been touched in years, but I did get it a maint kit near last time it was used and we have extra filaments. I just don't know what state it's in. we'll see.
01:38Brainium has joined #asahi
01:38amarioguychadmed: ofc i'm more knocking on the whole "SPR delayed" stuff more than high numbers of tapeouts
01:38amarioguythey have their own fabs it's less of a hit for them
01:39marcanthe turbopump oil reserves annoy me, those are unobtainium to get official ones for. might need to figure out how to refurb them.
01:39marcanfilaments, I think there are services that will refurb those and we have a nice pile of dead ones
01:39marcanI think everything else is mostly stuff you can clean / refurb yourself if you must
01:40marcanthe scariest thing is the turbopump tbh, I think it's hard to really break anything else but that thing will definitely tear itself to bits if things go wrong
01:40marcanbut I think the noise/resolution problem we have is electrical
01:41marcanthere's some kind of sine wave looking noise in the X direction
01:41marcanat high mag
01:41tpw_rulesmarcan: ah sorry
01:41marcanbut anyway, for the M1 my plan is to do an optical scan first after my funny razor delayering process, and try to get a nice high res stitch done
01:41TheDcoderis that the pump responsible for pumping out electron?
01:41marcansince apparently nobody has published die shots yet
01:42marcanTheDcoder: air
01:42marcanit's basically a jet engine in reverse, driven by an electric motor
01:42TheDcoderoh, is the oil thing DRMed?
01:42marcanbut it's funny shaped cartridges
01:42TheDcodernot 3D printable?
01:42tpw_rulesturbopumps are pretty wild
01:43marcanI don't think 3D printers are suitable for making stuff around turbopumps...
01:43marcanyou really don't want any particles or other junk getting in there
01:43tpw_rules(and the good ones will tear YOU to bits too!)
01:43TheDcoderokay, makes sense
01:43marcanwe can reuse the old cases, but they have some cotton or something inside holding the oil
01:43marcanwe'd probably have to figure out a compatible replacement for that, as well as how to open the case and close it cleanly
01:43marcanplus there's an o-ring involved
01:44marcanplus figuring out what oil type is appropriate
01:44marcantpw_rules: thankfully this turbopump is pretty smol and nominally behind a couple layers of steel so I don't think it's going to blow anyone up
01:45TheDcodercan't we bribe some scientists with access to an SEM and do the imaging there? sounds like it would be cheaper than doing all the DIY repairs and taking risks
01:45chadmedyeah but not as fun
01:45chadmedalso unis arent fun anymore
01:45amarioguyi can attest to that... i tried to get access to a SEM at my uni
01:46chadmedwhen i was still studying you had to have a valid and current assignment tasksheet and present that to lab staff to even use an oscope
01:46chadmedyou couldnt even go into the lab itself without scanning the barcode on the assignment sheet
01:46TheDcoderThis whole new AS architecture sounds too good to be true...
01:46marcanmind you, not that there's anything I'd be looking for in the M1 with a SEM
01:47marcanthis is just for curiosity
01:47marcanand it's not like you could reverse engineer circuits in practice
01:47marcantoo complicated
01:47TheDcoderyes I understand, just to take some pretty pictures
01:47marcanbut it's fun looking at specific structures
01:47marcanand if nothing else you can count bitlines and figure out SRAM block sizes and things like that
01:47TheDcodermaybe we can make an AI to RE the circuits, make it do all the labour
01:47TheDcoderif we can get enough high-res info
01:47amarioguyTheDcoder: some people did that for ROMs
01:47amarioguymaybe not AI
01:47marcanthe problem is the imaging, even automating it, you need to image all layers perfectly
01:48marcanit's a massive project for modern chips
01:48amarioguythere's also just.... a lot of transistors in general
01:48marcanI have this idea that I want to do full circuit extraction of the Gameboy Advance SoC some day
01:48marcanthat's an ARM7TDMI
01:48marcanand even that ought to be quite the project
01:48TheDcoderamarioguy: yeah, we're going to need a super-computer
01:48marcan(but I think it's within the realm of doable these days)
01:49amarioguyTheDcoder: it's not the compute power you have
01:49amarioguythe pure time commitment
01:49TheDcodermarcan: very interesting project, GBA is one of my favourite handheld consoles!
01:49amarioguythe resources you'd need
01:49amarioguylike i heard from someone
01:49amarioguyyou'd likely go insane
01:49amarioguyat even one layer in
01:49TheDcoderamarioguy: I have seen some crazy people... maybe one of them could take up the time commitment
01:50amarioguyanti-delayering doesn't help matters
01:50TheDcoderThe guy who made Holy C comes to mind
01:50amarioguyTheDcoder: i don't think you realize how large timescales i'm talking here
01:50amarioguynot as large as some other ones
01:50TheDcoderHmmm... yes
01:50chadmedwill riker voice "no you CANT dont even TRY"
01:50TheDcoderha ha ha ha
01:51TheDcodermaybe it will be a team of crazy people funded by an equally crazy billionaire?
01:51amarioguyit's just... no
01:51amarioguythe people who could do that
01:51amarioguyare better off
01:51amarioguytheir own design fab
01:51amarioguyand there's a ton of IP issues to sort out
01:51amarioguyif one were stupid enough to try
01:52TheDcoderI'd imagine they'd do it just for the LOLs and to see if there are indeed any backdoors
01:52marcaneven if you could image it, how would you look for backdoors?
01:52amarioguyspoiler alert: what people are calling backdoors are really not with any amount of introspection
01:52chadmedthe barriers to entry for designing a modern performant IC in 2023 are high enough as is without trying to RE some existing one
01:52marcanaudit the entire chip?
01:52TheDcoderthey might even upload it to pirate bay, who knows...
01:52chadmedits just harebrained
01:53amarioguyi feel like we've lost track of the main topic lmao
01:53TheDcodermarcan: Create AI to do the RE and auditing
01:53amarioguymy gosh lol
01:53TheDcoderbut yes, at this point I'm just shit posting
01:53marcan"AI"s work by training with examples, there is no training set for "silicon backdoors"
01:53TheDcoderI don't trust Apple or anyone to make modern chips without backdoors
01:54amarioguyTheDcoder: see above spoiler alert
01:54marcanwell, then that just means you can't trust any modern hardware period :p
01:54TheDcodermarcan: well I guess we'll have to perfect self-improving AIs then :P
01:54TheDcoderyes that's correct, I don't trust any modern hardware
01:54amarioguyok i think we've reached well beyond the point of sanity with this thread
01:55TheDcoderbut I don't have to worry because I'm not a crook... or some mentally-ill paranoid person.
01:55TheDcoderI agree amarioguy
01:55amarioguyi mean i shitpost about my other project occasionally :P
01:56amarioguyimpdef PMUs...
01:56amarioguybut that's a topic for another blog post
01:56TheDcoderamarioguy: what's your fediverse address? I want to follow :)
01:57TheDcodercan I access the "Factory test logs" in the NOR Flash?
01:57amarioguyi mean you can dump the NOR at any time from linux
01:58amarioguyi would really avoid messing with NOR if you can avoid it
01:58amarioguythere's info in there that's can be difficult to recover if you're not careful
01:58TheDcoderis it okay if I just dump it read-only?
01:58amarioguyyes RO is fine
01:58chadmedthe tradeoff with this platform being "too good to be true" is you just have to accept that some things are best left not done by the user
01:59chadmedtouching the NOR or any of the BOH processes are those things
01:59marcanI still need to test a full NOR wipe on one of my minis
01:59TheDcoderyes I agree but that wasn't what I was talking about. I was concerned about hardware backdoors like Intel ME or AMD PSP
01:59marcanthere are none
01:59chadmedthose do not exist
01:59TheDcoder...we can't be sure
01:59chadmedwe can actually
01:59amarioguypeople have RE'd both
01:59amarioguyand nothing of the sort was found
02:00DragoonAethis has quit [Quit: hej-hej!]
02:00amarioguyalso most of these coprocs
02:00amarioguydon't even share peripherals with main cores
02:00TheDcoderbut... I thought we could never RE the chip
02:00chadmedyou dont need to RE the SoC
02:00marcanIntel ME and AMD PSP aren't hardware backdoors
02:00marcanthey are software backdoors running on separate CPUs
02:00marcanthe software is right there in flash
02:00DragoonAethis has joined #asahi
02:00amarioguymarcan: add HSP in there too :P
02:00amarioguy(ms thing)
02:00TheDcodermarcan: right...
02:00chadmedevery peripheral is behind an IOMMU which masks off addresses that the firmware on the copro is not explicitly allowed to touch, and most firmwares are stored in the clear
02:01TheDcoderbut there could be a more sophisticated backdoor in the AS SoC
02:01amarioguyon apple to clarify
02:01chadmedthe machine just straight up reboots if you try to make a peripheral touch memory its not supposed to
02:01marcanand we have the t8103 keys upto some 13.0 beta so we have decrypted versions of everything else except SEP
02:01marcanand we know SEP is optional and not even booted beyond the ROM by default
02:01TheDcoderI'm talking hardware level, no visible firmware etc.
02:01amarioguyTheDcoder: my dude
02:01amarioguya lot of time
02:01amarioguyto implement
02:01chadmed[11:55] <TheDcoder> but I don't have to worry because I'm not a crook... or some mentally-ill paranoid person.
02:01amarioguywe are definitely hitting paranoia
02:02TheDcoderamarioguy: I'm assuming FBI etc. would want some kind of access in any case and are willing to "invest" to make sure Apple doesn't go into a loss implementing a BD like that
02:02amarioguyi'll take this to dms
02:02TheDcoderbut I'm just saying
02:02chadmedapple specifically and routinely tell the alphabets to get fucked when they ask for this stuff
02:03TheDcodernot trying to start a conspiracy theory, just putting it out that we can't rule out anything.
02:03nicolas17Apple recently added an opt-in end-to-end-encryption thing for iCloud backups
02:03chadmedin fact the FBI sued them over it in 2014
02:03marcanlike the whole security design has been an arms race against the TLAs
02:03nicolas17the FBI is *pissed*
02:03chadmedand lost
02:04marcanthey tried to get Apple to implement PIN code bruteforcing for them, they told them to piss off
02:04marcanso they found exploits to do it themselves
02:04marcanthen Apple fixed those
02:04TheDcoderGood Apple 👍
02:05marcanand this is all just about seized devices and such
02:05nicolas17marcan: is that when Apple updated the SEP hardware mid-cycle?
02:05TheDcoderI used to be a Apple hater before this... still kind of am though.
02:05amarioguymarcan: they actually did make that harder fwiw
02:05amarioguypart of user data
02:05marcannicolas17: nah, this was when they started using an anti-rollback EEPROM and later a proper secure EEPROM
02:05amarioguythe key to unlock user data now incorporates the SEPOS hash iirc
02:06amarioguynicolas17 can elaborate further but "million rounds of AES" :P
02:06marcanyes, they also do all the OS-tying thing now
02:06amarioguy(wasn't me who discovered it btw)
02:07nicolas17marcan: yeah that's what I mean, the security guide talks about A12 before and after "Fall 2020" having different secure EEPROM
02:07nicolas17A12 and A13
02:07marcanI'm not sure, what I'm talking about was before there was even a secure EEPROM
02:08marcanhaven't followed all that closely
02:08nicolas17"A12, A13, S4, and S5 products first released in Fall 2020 have a 2nd-generation Secure Storage Component, whereas while earlier products based on these SoCs have a 1st-generation Secure Storage Component." spinning up a new variant of those SoCs to update the SEP's secure storage, must have been important...
02:08nicolas17or is it a separate chip?
02:08marcanit's a separate chip
02:08marcanit has to be
02:09marcanthis is why it's so hard
02:09marcanyou can't do EEPROM/Flash in modern SoC processes, it's incompatible
02:09marcanso it has to be external
02:09marcanand then you need to secure the interface
02:09nicolas17okay, releasing a v2 is less crazy then
02:10amarioguycool people call this Lynx :)
02:10marcanis 1st gen the insecure one? (the one that was a dumb EEPROM)
02:10marcanor was that earlier?
02:10nicolas17it's dumb EEPROM, then Secure Storage gen 1, then secure storage gen 2
02:12nicolas17there's some stupid bug on the apple support website where it looks at my browser settings and loads in spanish, I change the language to english and it works, then I click any link and it redirects to spanish again /o\
02:14possiblemeatball has joined #asahi
02:17Brainium has quit [Quit: Konversation terminated!]
02:19TheDcoderwhat are xARTS?
02:19derzahl has quit [Remote host closed the connection]
02:19julio7359 has joined #asahi
02:20Emantor has quit [Quit: ZNC - http://znc.in]
02:20amarioguyTheDcoder: keybags
02:20amarioguykey/value storage
02:20Emantor has joined #asahi
02:20amarioguyfor SEP to hold keys to decrypt user data
02:21TheDcoderunderstood!
02:24TheDcoderis there just one huge APFS partition for all the user data?
02:24marcanby default yes
02:24marcancontaining multiple volumes
02:24chadmedoh right yeah, one container with multiple logical volumes
02:24amarioguyxARTs/gigalockers get their own volume
02:24amarioguyof course undumpable from macos
02:24TheDcodergot it. and yes I was asking about the default macOS state
02:24nicolas17an APFS partition ("container") can have multiple volumes, like btrfs
02:25TheDcoderamarioguy: what are gigalockers?
02:25marcanthere is one container for system stuff, one container for the OS/data, and one container for system recoveryOS
02:25marcanby default
02:25chadmedsystem data is immutable and sealed and once the SEP verifies its integrity it's unioned with the user data volume so that it's all presented as one filesystem
02:25marcanTheDcoder: xARTs
02:25TheDcoderunderstood regarding APFS
02:25marcanxARTs is why we're going to have to implement APFS at some point to make SEP work :p
02:26amarioguychadmed: immutable unless you allow root hash mismatches :P
02:26marcanthankfully it's a very limited subset of APFS
02:26TheDcoderchadmed: uhhh... but how? isn't it part of the same partition?
02:26marcanTheDcoder: separate volume
02:26amarioguyTheDcoder: merkle tree
02:26chadmedTheDecoder: consider an APFS container like an LVM set of volumes
02:26amarioguythere's also LwVM for you old folks that used to do ios back before APFS took over :)
02:26amarioguybut that's a meme moving on
02:26TheDcoderyes but how can the SEP make that immutable?
02:26amarioguyTheDcoder: not SEP
02:27amarioguyiBoot verifies this
02:27marcanthe kernel verifies it actually
02:27TheDcodersince all the system data is on the same partition... albeit in a container
02:27marcaniBoot just verifies the root hash
02:27marcanit's not "immutable"
02:27marcanit's just that if you mutate it it stops working
02:27TheDcoderyep, makes sense
02:27amarioguybelieve there's a SIP property to disable this
02:27amarioguybut either way
02:27amarioguyyou can't mount root as rw
02:27amarioguyon release
02:28amarioguyyou need to modify the snapshot
02:28marcanyes, with stuff in permissive security mode you can in principle make your own snapshots
02:28nicolas17(except dm-verity is at the block device level)
02:28TheDcoderIs all the APFS stuff done under the Asahi Linux umbrella as well?
02:28marcanyou can mount the FS as rw, you just can't mount it as root
02:28nicolas17afaik Asahi doesn't touch APFS currently
02:28marcan(that's how updates work, mount it live as RW, make changes, take snapshot, re-seal)
02:28marcannicolas17: correct
02:28marcanother than what the installer does
02:29TheDcoderoh, so there's no kernel driver for APFS?
02:29marcannot upstream
02:29TheDcodereven for RO support?
02:29marcanthere is a downstream one
02:29TheDcoderobviously not upstream :)
02:29marcannot sure what state it's in
02:29nicolas17and there's apfs-fuse
02:29marcanwe don't ship it yet
02:29TheDcoderokay, but it's not related to us
02:29marcanonce SEP comes into the picture we'll have to decide what to do
02:29nicolas17which last I checked still gives I/O errors on transparently-compressed files
02:29TheDcodernicolas17: already?
02:30marcanmight end up shipping it but with some big gate so people don't use it on normal volumes if it's not stable enough
02:30amarioguymarcan: can't wait to use my fork to test xART loads tomorrow :P
02:30TheDcodermarcan: I imagine it would be a whole new can of worms
02:30marcanthankfully xARTs is a super restricted use case
02:30nicolas17apfs-fuse might predate Apple Silicon macs actually
02:30marcanso much more likely that we can guarantee stability *just* for that
02:30TheDcodernicolas17: oh right, forgot that APFS is pre-AS
02:30nicolas17I just use it to grab stuff from iOS firmware images but the compression thing is a pain
02:31TheDcodermarcan: but I doubt we can get something like that upstreamed in Linux
02:31amarioguydo we even need to use apfs? i haven't checked whether SEP fetches the GL from the volume itself or whether AP has to preload it somehow
02:31marcanI think apple even uses bypasses for xARTs and doesn't actually hit the FS layer, but does direct writes from the kernel or something
02:31amarioguybut the fact there's no shmem access there
02:31amarioguymight explain it
02:31marcanamarioguy: AP does it
02:31marcanAP has to broker xARTs access
02:31marcanSEP doesn't know about APFS and it doesn't have any back channels to ANS
02:31marcanother than that key provisioning thing
02:31nicolas17amarioguy: I think the problem is we need to share the volume with macOS
02:31marcannicolas17: yes exactly
02:32marcanwe could just copy the file to Linux for experiments
02:32TheDcoderbut we aren't sharing the volume with macOS?
02:32marcanbut then macos would break
02:32amarioguyah right because tampering with that is bad news
02:32marcanso we have to share the APFS volume to make it work properly
02:32TheDcoderbut that's just the stub loader right?
02:32nicolas17TheDcoder: this would be the xARTs volume alone (which is currently not used at all on Linux)
02:32TheDcoderto boot the actual Linux
02:32marcanTheDcoder: no, I mean real macos
02:32marcanbut also iBoot potentially loads xARTs too
02:32tpw_ruleshas anybody figured out why apple does not just distribute the same update image to everyone
02:32marcanI believe Boot Policy stuff is stored there too
02:32TheDcoderuhhh... pretty sure I missed something important
02:33nicolas17tpw_rules: they don't?
02:33amarioguymarcan: erm how would that work with "recovery means no user data access"?
02:33TheDcodermarcan: wiki says yes
02:33amarioguythey flip a bit to disable the hw keys to decrypt user data
02:33amarioguyentirely possible
02:33marcannow I'm lost
02:33amarioguythat there's no user xART access in recovery
02:33tpw_rulesnicolas17: i mean like dd type updates. why go through the whole snapshot and start modifying files business
02:33TheDcodermarcan: :handshake:
02:33marcanoh you mean locking it out of recoveryOS
02:34nicolas17I think ipsws work that way
02:34amarioguythe whole boot progress register stuff
02:34marcannever noticed that but sure that wouldn't be hard to do
02:34marcanSEP knows what the boot state is
02:34TheDcoderrecoveryOS is an actual OS?
02:34marcanit's macOS
02:34chadmedeven the boot picker is macos
02:34marcanthe boot picker is also macOS
02:35amarioguymacosbasesystem
02:35amarioguyit's *very* integrity checked
02:35chadmedspecifically a fullscreen macos app running in recoveryOS
02:35amarioguymore so than the actual os at points
02:35marcanit's all one big ramdisk
02:35tpw_rulesnicolas17: then why not just splat down the ipsw to do an update? it seems like it would be orders of magnitude faster
02:35marcanunless you're booting System RecoveryOS
02:35marcanthat one mounts the image directly
02:35TheDcoderI was expecting something like UEFI
02:35TheDcoderbut more Apple-ish
02:35marcanpresumably because they want to let you nuke OS partitions from recoveryOS
02:35amarioguyTheDcoder: you mean protocol hell :/
02:35marcanTheDcoder: they ditched UEFI because it's impossible to secure
02:36marcanthat was a very explicit decision
02:36nicolas17I believe the only graphical thing iBoot does is the Apple logo
02:36amarioguyEFI is a mess of protocols pretending to be remotely sane for booting
02:36chadmedthis has been mentioned multiple times already
02:36amarioguytrust me on this one
02:36TheDcoderyes but they could've still implemented their own thing instead of making macOS the bootloader as well :-/
02:36marcannicolas17: and the "entering startup options..." message, and the "oh no please restore me" screen, and the battery charging indicator
02:36marcanthat's about it
02:36amarioguyiboot is the bootloader
02:36chadmedamarioguy: enjoying windows bringup i see :D
02:36marcanTheDcoder: macOS is not the bootloader, their bootloader just does not support picking an OS
02:37marcanso they boot macOS to do that and tell the bootloader what to boot, then reboot
02:37amarioguychadmed: making a ramdisk to test boot is... well 'fun' might be a way to describe it
02:37amarioguyimpdef PMUs too
02:37TheDcoderI see... so in theory we can make our own iBoot-compliant bootloader
02:37marcanno, because iBoot is signed
02:37nicolas17did you read how Asahi boot works yet? P
02:37marcanclearly not :p
02:38TheDcodernope, still reading the Intro to AS wiki
02:38chadmedim almost done with the security quick rundown...
02:38TheDcoderchadmed: 👍
02:39amarioguyfwiw i'm in bootmgfw/cdboot now so not much longer before i get to write up a toot showcasing remote progress :)
02:39amarioguyjust gotta figure out how to trap this one MSR access...
02:39TheDcoderamarioguy: are you trying to run Windows ARM?
02:39amarioguyTheDcoder: it's my other project :P
02:39nicolas17tpw_rules: ipsw is 12GB, delta OTA updates are as small as 493MB
02:40TheDcoderamarioguy: yes but why? hardly anyone uses Windows ARM :P
02:40amarioguybecause fun
02:40amarioguyand challenge
02:40amarioguyand masochism
02:40TheDcoderI approve 👍
02:40nicolas17"The real reason, of course, is that it's obviously evil and wrong and therefore fun for hack value."
02:40TheDcoderha ha ha ha
02:48amarioguyone thing i don't think i completely get is apple's custom error msrs
02:48amarioguylike why is "undefined opcode" not just... something in ISS and is shunted off to "L2C_ERR_STS"
03:27TheDcoderI have to go to sleep, so will resume my journey tomorrow. Thanks to everyone for being so kind and answering my questions in detail :)
03:45jeffmr has joined #asahi
03:54mini0n has joined #asahi
03:58julio7359 has quit [Ping timeout: 480 seconds]
04:00derzahl has joined #asahi
04:01jeffmr has quit [Ping timeout: 480 seconds]
04:05hertz has joined #asahi
04:12amarioguy_ has joined #asahi
04:22amarioguy_ has quit [Remote host closed the connection]
04:27amarioguy_ has joined #asahi
04:34amarioguy_ has quit [Remote host closed the connection]
04:35julio7359 has joined #asahi
05:17mini0n has quit [Quit: Leaving]
05:25Zopolis4 has quit [Quit: Connection closed for inactivity]
05:59balor has quit [Quit: balor]
06:00balor has joined #asahi
06:09nicolas17 has quit [Quit: Konversation terminated!]
06:20Zopolis4 has joined #asahi
06:23rvalue has quit [Read error: Connection reset by peer]
06:23rvalue has joined #asahi
06:41hertz has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
07:01possiblemeatball has quit [Quit: Leaving]
07:15julio7359 has quit [Remote host closed the connection]
07:19hertz has joined #asahi
07:25balor has quit [Quit: balor]
07:25balor has joined #asahi
07:43SSJ_GZ has joined #asahi
07:53leitao has joined #asahi
07:55leitao has quit []
08:23hertz has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
08:42Techcable has quit [Ping timeout: 480 seconds]
08:50Techcable has joined #asahi
08:55capta1nt0ad has joined #asahi
09:05tim has joined #asahi
09:06tim is now known as Guest1321
09:07ihaveamac has quit [Quit: fail]
09:07capta1nt0ad has quit [Ping timeout: 480 seconds]
09:08ihaveamac has joined #asahi
09:12kujeger has quit [Ping timeout: 480 seconds]
09:26capta1nt0ad has joined #asahi
10:04bps2 has joined #asahi
10:19kujeger has joined #asahi
10:43jluthra has quit [Remote host closed the connection]
10:44jluthra has joined #asahi
11:10SSJ_GZ has quit [Read error: No route to host]
11:13landscape15 has joined #asahi
11:16jacksonchen666 has quit [Remote host closed the connection]
11:17jacksonchen666 has joined #asahi
11:22landscape15_ has joined #asahi
11:25landscape15 has quit [Ping timeout: 480 seconds]
11:26landscape15 has joined #asahi
11:30landscape15_ has quit [Ping timeout: 480 seconds]
11:36landscape15 has quit [Ping timeout: 480 seconds]
11:56axt has joined #asahi
12:09Guest1321 has quit [Quit: Guest1321]
12:33landscape15 has joined #asahi
12:35landscape15 has left #asahi [#asahi]
12:51chadmed_ has joined #asahi
12:52sah4ez has joined #asahi
12:54SSJ_GZ has joined #asahi
12:58sah4ez has quit [Read error: Connection reset by peer]
13:23axt has quit [Quit: Leaving.]
13:27mini_ has quit [Quit: ZNC closing...]
13:27mini_ has joined #asahi
13:37landscape15 has joined #asahi
13:42landscape15 has left #asahi [WeeChat 3.8]
13:43landscape15 has joined #asahi
13:54landscape15Can someone tell me how to know which recovery i booted into?
13:56capta1nt0ad has quit [Remote host closed the connection]
14:00landscape15 has left #asahi [WeeChat 3.8]
14:03SSJ_GZ has quit [Remote host closed the connection]
14:04SSJ_GZ has joined #asahi
14:08amarioguylandscape15: bputil -d
14:09amarioguychoose any UUID if prompted
14:09amarioguythe output will tell you if your recoveryOS is "paired"
14:09amarioguyif monterey and later
14:10amarioguybut *generally* speaking you're usually booted into the recoveryOS that's paired with the blessed volume/startup disk
15:04landscape15 has joined #asahi
15:05Zopolis4 has quit [Quit: Connection closed for inactivity]
15:07tim has joined #asahi
15:07tim is now known as Guest1344
15:09landscape15amarioguy: Thank you. What does it say if i booted into fallback recovery?
15:10sah4ez has joined #asahi
15:11landscape15 has quit [Quit: WeeChat 3.8]
15:15sah4ez has quit [Remote host closed the connection]
15:25TheDcoderAre any cool features powered via the SEP currently supported in Asahi Linux?
15:27TheDcoderAlso how screwed am I if I forget my macOS password without having an iCloud account linked?
15:29j`eyno SEP is only used for random number seed currently
15:29j`ey*no, SEP..
15:30TheDcoderbetter than nothing I guess... I'm looking forward to being able to sign stuff by just using my fingerprint :)
15:32faruk has joined #asahi
15:41possiblemeatball has joined #asahi
15:42derzahl has quit [Remote host closed the connection]
15:59amarioguylandscape15: not sure actually
16:00amarioguyTheDcoder: we are a *long* way from using touch id
16:08TheDcoderAlso why is the update taking 30 minutes after the download?
16:08amarioguyTheDcoder: Apple's ota update process is notoriously long
16:09TheDcoderWhy though...
16:10jamespmorgan has joined #asahi
16:10TheDcoderWhy haven't Windows and Apple figured out easy updates like Linux yet? :P
16:12roxfan has quit [Ping timeout: 480 seconds]
16:13jacksonchen666finger slipped
16:14flying_sausages has quit [Ping timeout: 480 seconds]
16:30TheDcoderCrap, I accidentally downloaded the Monterey update instead of the Ventura update...
17:00bps2 has quit [Ping timeout: 480 seconds]
17:03c10l has quit [Quit: Bye o/]
17:06c10l has joined #asahi
17:14ciggi has joined #asahi
17:14ciggi_ has quit [Read error: Connection reset by peer]
17:14bps2 has joined #asahi
17:26roxfan has joined #asahi
17:29bps2 has quit [Ping timeout: 480 seconds]
17:34mini0n has joined #asahi
17:35jamespmorgan has quit [Remote host closed the connection]
17:52bcrumb has joined #asahi
17:58faruk has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
18:04bcrumb has quit [Ping timeout: 480 seconds]
18:38faruk has joined #asahi
18:41systwi has quit []
19:05lewurm`` has quit [Read error: Connection reset by peer]
19:05lewurm has joined #asahi
19:31WindowPa- has joined #asahi
19:35WindowPain has quit [Ping timeout: 480 seconds]
20:13possiblemeatball has quit [Quit: Leaving]
20:21c10l has quit [Read error: Connection reset by peer]
20:21c10l has joined #asahi
20:21seeeath has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
20:23seeeath has joined #asahi
20:23faruk has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
20:24seeeath has quit []
20:25faruk has joined #asahi
20:29possiblemeatball has joined #asahi
20:32jamespmorgan has joined #asahi
20:36mkurz has joined #asahi
20:37mkurzWhen using Asahi edge on a MacBook Pro 14" with the Pro chip, is it possible to somehow connect an external Display / Beamer for doing a presentation?
20:38mkurzHDMI isn't working according to the wiki, but is there another way, like using USB3 or something?
20:38ChaosPrincessi think displaylink will work
20:38ChaosPrincessbut displaylink is kinda crap
20:39seeeath has joined #asahi
20:44ChaosPrincesscan't see that tweet
20:44j`eymkurz: that's not usuable yet
20:44mkurzj`ey: yeah, just realized that
20:44ChaosPrincessin case it was talking about displayport - thats not supported yet
20:45ChaosPrincessbut displaylink != displayport
20:45svenI guess that was when I got DisplayPort via usb-c working with a bunch of hacks
20:45mkurzChaosPrincess: I guess you must be logged in to Twitter and follow Sven Peter to see the tweet
20:45ChaosPrincessyea, im not doing the first part of those requirements :P
20:46sventhe second part also wont work because I made my account private and havent logged in in ages ;)
20:47svenanyway, that entire DP altmode us still WIP and I’ll probably only get back to it once thunderbolt is up
20:47mkurzsven: Are you working on Thunderbolt or someone else?
20:48mkurzsven: Nice! Also thanks for doing that!
20:58arsven: just wondering, the DP part has some dependencies on thunderbolt, or is TB just more interesting?
21:00svenall of that goes through atcphy one way or another
21:00svenand the dcpext side is essentially the same
21:00sven(For dp over tbt tunneling)
21:04bluetaildo you by chance know where I can find the mac mini m1 endurance for the integrated SSD?
21:04bluetailwondering if I should write my IRC logs to my external storage instead
21:05svenif the ssd was reaching its lifetime limit with a few mb per day at most that would be a pretty horrible ssd ;)
21:05jamespmorgan has quit [Remote host closed the connection]
21:07bluetailsven I am in a tons of channels
21:07bluetailI think 30
21:07bluetailand I have precache enabled for thelounge... which means I download everything below 2MiB somebody posts
21:07seeeath has quit [Quit: Textual IRC Client: www.textualapp.com]
21:08bluetailimages, videos... perhaps I should disable that
21:09faruk has quit [Quit: Textual IRC Client: www.textualapp.com]
21:10bluetailheh, its "only" 758.6MiB worth of logs
21:10bluetailso yes, probably fine
21:26irth has quit [Remote host closed the connection]
21:26irth has joined #asahi
21:26irth has quit [Remote host closed the connection]
21:29ChaosPrincessy'all should chill w.r.t ssd durability
21:30chadmed_yeah its not 2008
21:33nclssd endurance doesnt even feel like much of a.. thing
21:33ncltbw ratings are for warranties not for "it will die after this many writes"
21:40nclthere's maybe something about probability of corruption going up but the entire MO of ssds is error correction on fast unreliable media
21:44amarioguy has quit [Remote host closed the connection]
21:56Guest1344 has quit [Quit: Guest1344]
22:12joske has joined #asahi
22:20joske has quit [Quit: Leaving]
22:25elvishjerricco has joined #asahi
22:29hertz has joined #asahi
22:34possiblemeatball has quit [Ping timeout: 480 seconds]
22:50possiblemeatball has joined #asahi
23:00arndbluetail: "smartctl -a /dev/nvme0n1" reports values for "Data Units Written" and "Percentage Used" that you can use to estimate the expected life. I'm mostly using an external drive for write intensive work, so the it's only at 3.24TB written on the internal 1TB drive, which is rounded to 0%.
23:08flying_sausages has joined #asahi
23:13Droop has quit []
23:13faruk has joined #asahi
23:14Droop has joined #asahi
23:17possiblemeatball has quit [Quit: Leaving]
23:25balor has quit [Quit: balor]
23:25balor has joined #asahi
23:30WindowPa- has quit [Quit: ZNC 1.8.2 - https://znc.in]
23:30WindowPain has joined #asahi
23:39SSJ_GZ has quit [Ping timeout: 480 seconds]
23:48derzahl has joined #asahi
23:55Brainium has joined #asahi
23:56TheDcoderHi, the installers tells me that the selected partition has significant overhead (around 12G), I believe this is the redundant macOS Monterey updated that I download by accident. I did not install it but the update did finish downloading.
23:56cy8aer has joined #asahi
23:57TheDcoderI downloaded the latest Ventura update after that and completed the upgrade successfully
23:57TheDcoderbut now I don't know how I can get rid of the overhead... what should I do?
23:57TheDcoderI don't have Time Machine enabled
23:57derzahlhey jannau, did you ever have a chance to look at my asahi-diagnose output to see if you noticed any clues as to how i managed to break my display backlight? for reference: https://pastebin.com/avTF7qww
23:58cy8aer has quit [Remote host closed the connection]
23:58cy8aer has joined #asahi
…