#
…

2021-01-07

00:11ndom91 has quit [Remote host closed the connection]
00:36ransom has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
00:46msalter_ has joined #asahi
00:46msalter_ has quit [Client Quit]
00:47ransom has joined #asahi
01:14qyousef has joined #asahi
01:29rbenua has joined #asahi
01:59KarboniteKream has joined #asahi
02:44ganpa has joined #asahi
02:44bent__ has joined #asahi
02:47bent_ has quit [Ping timeout: 246 seconds]
02:53thecake21[m] has joined #asahi
02:54aratuk has joined #asahi
03:04blazra[m] is now known as blazra
03:09blazra has quit [Quit: authenticating]
03:10blazra has joined #asahi
03:30awordnot0 has joined #asahi
03:31awordnot has quit [Read error: Connection reset by peer]
03:31awordnot0 is now known as awordnot
03:39delogips[m] has joined #asahi
04:33KarboniteKream has quit [Ping timeout: 256 seconds]
04:55browzing has quit [Ping timeout: 256 seconds]
05:00maximus64marcan: you mention that TrustZone is impossible to implement securely. I'm curious why?
05:05marcanbecause unless you put half of your OS kernel in TrustZone, EL1 has immense power over EL3
05:05marcanand even if you do, there's the entire massive field of speculation/etc microarchitectural vulnerabilities
05:06marcane.g. if the EL1 kernel can do CPU voltage/frequency scaling, EL3 is just one clock glitch away from getting pwned
05:06marcansame with anything related to the memory subsystem
05:06marcanand modern SoCs are so complex it's *impossible* to get this right and plug all the holes
05:06maximus64ahh I see yea I remember same attack was done on Intel SGX
05:06marcanyes, SGX is the same nonsense
05:07marcanIntel: "hey, ARM had a terrible idea, why don't we copy it?" :-)
05:07maximus64lol yea I totally agree with you
05:13Shizif you put half of your OS kernel into trustzone, there's a limit->100% chance of one of the interface having vulns
05:13Shizlittle use of putting your stuff into an Even More Privileged Layer if that's where the vuln lives now :p
05:19maximus64I think with ARM64 you could run the whole linux kernel in trustzone ;)
05:21browzing has joined #asahi
05:26browzing has quit [Ping timeout: 246 seconds]
05:30JTLmaximus64: but why? :P
05:32maximus64I just say that you could doesn't mean that you should heh
05:33ransom has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
05:48TheJollyRogerGood morning!
05:49rootspring[m] has joined #asahi
05:56browzing has joined #asahi
06:01KarboniteKream has joined #asahi
06:02browzing has quit [Ping timeout: 272 seconds]
06:13bdju has joined #asahi
06:15browzing has joined #asahi
06:20browzing has quit [Ping timeout: 256 seconds]
06:40awordnot has quit [Quit: The Lounge - https://thelounge.chat]
06:41Alex[m]17 has joined #asahi
06:43rockinrobstar[m] has joined #asahi
06:53Alex[m]17i am excited for linux to unleash the potential of M1 chips
06:58aratuk_ has joined #asahi
07:00aratuk_ has quit [Remote host closed the connection]
07:06awordnot has joined #asahi
07:25KarboniteKream has quit [Ping timeout: 260 seconds]
08:04KarboniteKream has joined #asahi
08:26Rays42 has joined #asahi
08:27aratuk has quit [Remote host closed the connection]
08:46bear24rw has quit [Remote host closed the connection]
08:47bear24rw has joined #asahi
08:52bear24rw has quit [Ping timeout: 246 seconds]
08:56ephe_meral has joined #asahi
09:05aratuk has joined #asahi
09:05aratuk has quit [Client Quit]
09:15bear24rw has joined #asahi
09:18merbanan has joined #asahi
09:21aratuk has joined #asahi
09:27haddock has left #asahi [#asahi]
09:39_whitelogger has joined #asahi
10:02zopieux has joined #asahi
10:17browzing has joined #asahi
10:21browzing has quit [Ping timeout: 240 seconds]
10:21konstater has joined #asahi
10:27ts170[m] has joined #asahi
10:36j`eymarcan: do anything asahi+technical today? :-)
10:40marcanyup, was just readying to do a little livestream shortly actually :)
10:40marcangoing to work on scaffolding for speeding up the debug/test cycle
10:40marcan(I have learned the lesson many times that this pays off *big time* for this kind of project)
10:41j`eysweet, I will be ready
10:45flying_sausages has joined #asahi
10:45flying_sausagesthat's a big IRC room for 2 days of being up
10:51gruetzkopflast time i've seen that was the #$(console)dev channels
10:53bear24rw has quit [Remote host closed the connection]
11:00Mary_ has joined #asahi
11:03marcan"I'll just reply to emails and get on the stream" *2 hours later still replying to emails*
11:03bear24rw has joined #asahi
11:07jinen[m] has joined #asahi
11:08bear24rw has quit [Ping timeout: 264 seconds]
11:31CDFHLooking forward to a stream, will be interesting to have on in the background
11:36ephe_meral has quit [Ping timeout: 240 seconds]
11:41marcanalright, let me get set up
11:44TheJollyRoger has quit [Remote host closed the connection]
11:59Jamie[m]2 has joined #asahi
12:07ephe_meral has joined #asahi
12:11marcanI'll be on #asahi-dev for this stream.
12:16keystone[m] has left #asahi ["User left"]
12:35brinlysome familiar handles here
12:36j`eyanyone had issues connecting USB C monitor to their mac?
12:36j`eymy monitor shows as a billboard device :/
12:36plainbits has joined #asahi
12:53cdesaiohai brinly
13:00ksixty has joined #asahi
13:07sharpiro has quit [Ping timeout: 264 seconds]
13:11mah has joined #asahi
13:26TheJollyRoger has joined #asahi
13:45mah has quit [Ping timeout: 264 seconds]
14:00aratuk has joined #asahi
14:16aratuk has quit [Remote host closed the connection]
14:24riker77 has quit [Quit: Quitting IRC - gone for good...]
14:39mah has joined #asahi
14:40ceph3us has joined #asahi
14:42ceph3usmarcan: i did cat /etc/passwd on a real mac and there's no _sshd, just sshd, guess they probably built the recovery image with this done deliberately?
14:44davidrysk[m]ohhh this is the recovery image?
14:45j`ey"I'm in recovery" :P
14:50TheJollyRoger has quit [Ping timeout: 240 seconds]
14:52davidrysk[m]I joined late and will have to leave for a bit\
14:53TheJollyRoger has joined #asahi
14:54ceph3us has quit [Remote host closed the connection]
15:01ceph3us has joined #asahi
15:02mah is now known as mch
15:03mch is now known as mah
15:03mah has quit [Quit: Leaving]
15:10plainbits has quit [Quit: Go to sleep. Night!]
15:15ephe_meral has quit [Ping timeout: 260 seconds]
15:27Rays42 has quit [Read error: Connection reset by peer]
15:29Axenntio has joined #asahi
15:29Axenntio has quit [Client Quit]
15:29ephe_meral has joined #asahi
15:29Axenntio has joined #asahi
15:32Axenntio_ has joined #asahi
15:35Axenntio_ has quit [Remote host closed the connection]
15:36Axenntio has quit [Ping timeout: 264 seconds]
15:49davidrysk[m]marcan: so when sshd is unloaded by launchctl, I get connection refused (as expected), when it's loaded I get connection reset by peer (which is why I asked if that was what was expected)
15:49davidrysk[m]so let's see why it's failing
15:50davidrysk[m]oh yeah useful tip, if you need to use CLI tools that are unavailable from recovery you may be able to mount a macOS volume and run them from there
15:51davidrysk[m]and I can see that both su and sudo don't work
15:53brentr123[m] has joined #asahi
15:54riker77 has joined #asahi
16:01jn__nice :D
16:01spikebikeshows the priorities ;-)
16:07winocmI tried.
16:31bear24rw has joined #asahi
16:33brentr123[m]I dont know if anyone has asked this already, but will asahi have x86 emulation?
16:33winocm(I hate the taste of beer.)
16:33opticronbrentr123[m], that's kind of an odd question
16:34opticronyou'll be able to run qemu on it, so yes?
16:34j`eyi guess maybe they meant the TSO bit? which enables the x86 memory model
16:35bear24rw has quit [Ping timeout: 256 seconds]
16:35jn__brentr123[m]: through qemu-x86_64, if you install it: sure
16:36justMakumarcan mentioned earlier that he will try to get the TSO to work, apparently Linux kernel already kinda supports something similar, so shouldn’t be lots of work
16:36justMakuthat it’s after he figures out how to enable it on M1
16:37davidrysk[m]someone already figured it out and wrote a kext that does it
16:38davidrysk[m]that calls a sysctl though
16:39q3k|m... it implements a sysctl, doesn't call one
16:39davidrysk[m]wait, no, it implements a sysctl that sets the thread pointer contents
16:41bear24rw has joined #asahi
16:41davidrysk[m]will need to dig further
16:42davidrysk[m]q3k|m: yes but that's only setting the flag and then the kernel sets the register
16:44q3k|moh, isee
16:45q3k|mi thought this was some magical aarch64/m1 segment for per thread hw control
16:46bear24rw has quit [Remote host closed the connection]
16:47brentr123[m]i meant similar to how windows insider does it
16:47brentr123[m]the arm windows
16:50brentr123[m]or just "rosetta 2" but on linux
16:51j`eyqemu-user-x86 or whatever its called
16:51linkmauvebrentr123[m], you can use binfmt_misc to register your emulator of choice for your file format of choice.
16:52linkmauveThen chmod +x it, and ./the-foreign-binary will work.
16:52brentr123[m]thats not end-user friendly
16:52j`eyasahilinux is probably not starting out to be user friendly
16:53davidrysk[m]there are tools that make it easier, the problem is that to do everything you need a chroot with arm64 libs
16:53davidrysk[m]that said I was experimenting with mcsema to see if it could be used to AOT-recompile an aarch64 binary to x86_64
16:53davidrysk[m]but that's a low priority project :P
16:54bear24rw has joined #asahi
16:57Axenntio has joined #asahi
16:58carlosn has joined #asahi
17:00bear24rw has quit [Ping timeout: 264 seconds]
17:03ephe_meral has quit [Ping timeout: 256 seconds]
17:16bear24rw has joined #asahi
17:20davidrysk[m]yeah, it's being discussed in the -gpu channel
17:35marcanit's an MSR or something, I'll implement it as a prctl()
17:36marcanthe TSO thing
17:39davidrysk[m]marcan: it's probably an MSR
17:40marcanit should be like 10 lines of code if that, this will be trivial
17:40marcanonce basic userspace is usable I'll just do it at some point, feel free to remind me if I forget
17:40marcanmore fun will be using it in qemu :)
17:40marcan(cc agraf? :))
17:41marcandavidrysk[m]: oh good point on the macos tools... the OS should actually already be mounted in recovery mode
17:41marcanyou can only use signed apps, but of course those are signed
17:42marcanI can just copy script or whatever from osx, or from the net via curl
17:42marcanI completely forgot about that
17:42modwizcode has joined #asahi
17:42marcanI can't *make* tools but I can *copy* them
17:42marcanso yeah maybe I'll just give up on ssh, copy script over and use a netcat pipe
17:42davidrysk[m]marcan: programs that are ad-hoc signed on another computer just work
17:43davidrysk[m]even with SIP enabled
17:43marcanI thought recovery mode enforced proper apple codesigning
17:43marcana la iOS
17:43tiago_ has joined #asahi
17:43marcan(though I haven't tested it)
17:43marcan(just something I heard)
17:43marcanthe point of 1TR is that it is a trusted environment
17:44marcanif you can just throw code in you subvert that
17:44davidrysk[m]well, nothing in 1TRR is persistent
17:44davidrysk[m]it's a ramdisk that's created on boot
17:44Shizwell yeah, but it has more privileged access to stuff
17:44Shizif you can run arbitrary bins on it what's the point
17:45davidrysk[m]well, I compiled a tester program on another computer, copied it to an external usb, plugged the usb into the machine in recovery, and it ran
17:45davidrysk[m]the point is that physical access is required to enter 1TRR
17:45davidrysk[m]and that changes to recovery aren't persistent, so you can't install a persistent implant
17:45marcandavidrysk[m]: on M1?
17:46davidrysk[m]well, I didn't test running that binary on M1. I had better :)
17:46davidrysk[m]physical access is required to access 1TRR on M1.
17:46marcanI mean my info might be wrong :p
17:46marcananyway, sleep now ;)
17:46marcanmore experiments tomorrow
17:46davidrysk[m]I'm using the M1 as my primary so I wanted to avoid rebooting unnecessarily
17:46davidrysk[m]but I sure can test it :)
17:46Shizoyasumi \o
17:58browzing has joined #asahi
18:04davidrysk[m]marcan: yeah indeed on Intel you can run anything but on M1 they enforce code signing in recovery
18:04davidrysk[m]Still, useful for testing
18:08davidrysk[m]programs signed by apple from the root volume do still work
18:08davidrysk[m]I didn't test a third party apple developer signed program
18:08Axenntio_ has joined #asahi
18:08Axenntio has quit [Remote host closed the connection]
18:08davidrysk[m](meaning, one with a legit non-adhoc sig)
18:09Axenntio_ has quit [Remote host closed the connection]
18:09Axenntio has joined #asahi
18:09Axenntio has quit [Remote host closed the connection]
18:17Calchan has joined #asahi
18:18bloom has joined #asahi
18:18bloom has left #asahi [#asahi]
18:20stormclad has joined #asahi
18:20bear24rw has quit [Remote host closed the connection]
18:21bear24rw has joined #asahi
18:23tiago_ is now known as tmartins
18:24bear24rw has quit [Remote host closed the connection]
18:24bear24rw has joined #asahi
18:25bear24rw has quit [Remote host closed the connection]
18:25Baughn has joined #asahi
18:25bear24rw has joined #asahi
18:25tmartins is now known as tiagom
18:26ghantaz has joined #asahi
18:26tiagom has quit [Quit: tiagom]
18:26tiagom has joined #asahi
18:27tbodt has joined #asahi
18:29ghantaz has quit [Remote host closed the connection]
18:33bear24rw has quit [Remote host closed the connection]
18:33bear24rw has joined #asahi
18:41gua has joined #asahi
18:42rafaelmartins has joined #asahi
18:57Axenntio has joined #asahi
18:58Axenntio_ has joined #asahi
18:58Axenntio_ has quit [Remote host closed the connection]
18:59King_InuYasha has joined #asahi
18:59guajust a heads up, i looked into discord and IRC bridges a month or two ago and the clear current leader is the go-discord-irc project: https://github.com/qaisjp/go-discord-irc
19:00gua(for if and when a discord bridge is set up, i would want it to not get a bad reputation from using inferior software)
19:01Axenntio has quit [Remote host closed the connection]
19:01King_InuYasha is now known as Conan_Kudo
19:01Conan_Kudo is now known as King_InuYasha
19:02Eighth_Doctor has joined #asahi
19:02gua has quit [Remote host closed the connection]
19:05Axenntio has joined #asahi
19:06Axenntio_ has joined #asahi
19:06Axenntio_ has quit [Remote host closed the connection]
19:08Axenntio_ has joined #asahi
19:09Axenntio_ has quit [Remote host closed the connection]
19:14Axenntio has quit []
19:15Axenntio_ has joined #asahi
19:15veyron has quit [Remote host closed the connection]
19:28Axenntio_ has quit [Remote host closed the connection]
19:31ConeOfAttack[m] has joined #asahi
19:41fl35[m] has joined #asahi
19:42plainbits has joined #asahi
20:07davidrysk[m]okay, next step is to pull the dyld_shared_cache off the recovery and see if it contains the needed functions
20:08davidrysk[m]is there a good dyld_shared_cache extractor available?
20:10choozy has joined #asahi
20:10Axenntio has joined #asahi
20:11Axenntio has quit [Remote host closed the connection]
20:12awordnotdavidrysk[m]: https://github.com/antons/dyld-shared-cache-big-sur along with the PR that enables arm64 support
20:13Axenntio has joined #asahi
20:18davidrysk[m]awordnot: I see that that project has not been updated for the dyld that was just released by Apple
20:19awordnotdavidrysk[m]: yeah I'm not sure how far it'll get you with the latest dyld. I was researching this recently and this was the most up-to-date tool I could find
20:19ransom has joined #asahi
20:23plainbits has quit [Ping timeout: 272 seconds]
20:27TheJollyRogerAhoy awordnot! Good to see you!
20:28ransom has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
20:28awordnotTheJollyRoger: oh hi JR :)
20:29Axenntio has quit [Ping timeout: 240 seconds]
20:29TheJollyRogerGreat to see more familiar faces here, hehehe.
20:30sharpiro has joined #asahi
20:31Axenntio has joined #asahi
20:32browzing has quit [Ping timeout: 246 seconds]
20:32sharpiro_ has joined #asahi
20:33davidrysk[m]found the problem, you need to edit /etc/master.passwd, not /etc/passwd
20:33davidrysk[m]the manpage for passwd(5) tells us that
20:33davidrysk[m]just add an alias in it, duplicate the line for _sshd with sshd and use the same uid/gid :P
20:34davidrysk[m]there's another issue though, you need to place hostkeys on the system
20:34Axenntio has quit [Remote host closed the connection]
20:35davidrysk[m]still getting connection reset though!!
20:35sharpiro has quit [Ping timeout: 256 seconds]
20:41davidrysk[m]it can't chroot, due to EACCES
20:41bloom has joined #asahi
20:42bloom has left #asahi [#asahi]
20:43davidrysk[m]hm not EACCES but EPERM
20:48dottedmagMary_: thog.eu cert looks expired
20:48Mary_yeaaah I know
20:48Mary_not using this domain anymore tbh
20:48tiagom has quit [Quit: tiagom]
20:49tbodtis it not possible to compile a custom sshd?
20:49dottedmagMary_: except e-mail for copyrights?
20:49dottedmagtbodt: signature check
20:50tbodtdoes it do that for everything in recovery?
20:50Mary_I mean the mail server is still up dottedmag, just too lazy to move stuffs to my new domain
20:52Mary_dottedmag, fixed the cert
20:53Mary_Anyhow I also tried some stuffs under recovery for sshd but without much success
20:54davidrysk[m]Mary_: I got as far as having it try and fail to chroot
20:54davidrysk[m]if you have an intel mac you can run self-signed binaries in recovery which is useful for testing
20:55plainbits has joined #asahi
20:55davidrysk[m]on M1, yes it enforces sigchecking in recovery. Apple-signed system binaries work, though, but I haven't tested running a dev-signed binary. Also the dyld_shared_cache is much-reduced.
20:56brentr123[m]Is there anyway to spoof the sig?
20:59jozzle has joined #asahi
20:59browzing has joined #asahi
21:06ky0ko has joined #asahi
21:10davidrysk[m]ConeOfAttack: when you send a too-long message from matrix it gets turned into a link on IRC
21:10jn__ConeOfAttack[m]: AFAIK (from https://github.com/AsahiLinux/docs/wiki/Codenames) the M1 has a "performance" core and a "efficiency" core, which is probably something to consider
21:10davidrysk[m]Personally, it would be nice if someone did a ton more uarch and memory testing. We don't even know how large the shared (L3?) cache is.
21:11davidrysk[m]it's not a traditional L3 cache because it's shared between the CPU, GPU, ML, and other cores
21:12ConeOfAttack[m]david.rysk: bummer! Any idea how many characters I have to work with? And yes, hopefully more individuals will obtain hardware and start testing.
21:12davidrysk[m]I believe something around 500 but I'm not sure
21:13davidrysk[m]I'd like to see more exhaustive tests like these
21:13tpw_rulesi would assume newlines also trigger the too-long link
21:13ConeOfAttack[m]I expect that studying the cache will be tricky too, since the DRAM is on the die, but the controller also has some sort of last layer cache. Is that an L3? Victim cache? L4? Weird stuff.
21:13brentr123[m]I might get a m1 soon, how could a noob like me with only minimal python knowledge test things?
21:13spikebikeI wrote a microbenchmark to explore latency and parallelism and while I don't have a M1, I've been having people run it on the M1
21:14davidrysk[m]the DRAM is on-package, not on-die
21:15davidrysk[m]the "fabric" has an L3-like cache that is shared by the various components
21:15spikebikeyes, but the slow cores have a seperate L3
21:15ConeOfAttack[m]david.rysk: thanks for correcting me. Also, is it worth breaking my first post into smaller chunks?
21:15davidrysk[m]ConeOfAttack: I'd just summarize it; people can click on the link if they need to
21:17tiagom has joined #asahi
21:21Yuzuslow cores have separate L3? afaik they use the same System Level Cache as everything else, they just have their own shared L2
21:24spikebikeah, seperate L2 would explain the graph as well
21:24Yuzuyeah, it's 12MB shared L2 for perf cluster, 4MB shared L2 for efficiency cluster
21:25spikebikeah, sure, the total cache size differing by 4MB with all cores in use fits well.
21:28jjanzic has quit [Remote host closed the connection]
21:28ConeOfAttack[m]Anandtech's uArch tests suggest the M1 ROB is 600+ deep. Rename stations have ~350 entries for INT & FP each. This should be quite power hungry, esp. the CAM, so I speculate that Apple's engineers have a novel implementation.
21:28jjanzic has joined #asahi
21:28ConeOfAttack[m]Mitch Alsup and lkcl have a modified CDC 6600 scoreboard, fully OoO with exceptions. The design is stated to avoid a CAM entirely. It also seems suitable for an esp. wide CPU. https://libre-soc.org/3d_gpu/architecture/6600scoreboard
21:30skg has joined #asahi
21:30_plainbits_ has joined #asahi
21:30bear24rw has quit [Remote host closed the connection]
21:31bear24rw has joined #asahi
21:32bear24rw has quit [Remote host closed the connection]
21:32xerpi[m] has joined #asahi
21:32bear24rw has joined #asahi
21:32ransom has joined #asahi
21:34plainbits has quit [Ping timeout: 272 seconds]
21:43ransom_ has joined #asahi
21:45ransom has quit [Ping timeout: 272 seconds]
21:47aratuk has joined #asahi
21:50aratuk_ has joined #asahi
21:52FFY00 has quit [Remote host closed the connection]
21:53FFY00 has joined #asahi
21:53aratuk has quit [Ping timeout: 256 seconds]
21:59jevinskie[m] has joined #asahi
22:01agrafmarcan: let's leave the TSO QEMU fun for after you have PCIe up and running :)
22:01ransom_ has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
22:03ransom has joined #asahi
22:03hspak2 has joined #asahi
22:04_plainbits_ has quit [Ping timeout: 272 seconds]
22:05plainbits has joined #asahi
22:07hspak has quit [Ping timeout: 272 seconds]
22:07hspak2 is now known as hspak
22:17choozy has quit [Remote host closed the connection]
22:20krbtgt has joined #asahi
22:27plainbits has quit [Quit: Go to sleep. Night!]
22:46stormclad has quit [Remote host closed the connection]
22:47stormclad has joined #asahi
23:03ransom has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
23:04ransom has joined #asahi
23:19assusdan[m] has joined #asahi
23:39modwizcode has quit [Quit: Later]
23:54jamadazi has joined #asahi
23:55Lightsword has joined #asahi
…