ChanServ changed the topic of #asahi-dev to: Asahi Linux: porting Linux to Apple Silicon macs | General development | GitHub: | Wiki: | Logs:
chadmed has joined #asahi-dev
VinDuv has joined #asahi-dev
fossdd has quit [Remote host closed the connection]
fossdd has joined #asahi-dev
c10l has quit [Quit: Ping timeout (120 seconds)]
dingodoppelt has quit [Quit: ZNC 1.9.x-git-170-9be0cae1 -]
flying_sausages has joined #asahi-dev
c10l has joined #asahi-dev
ChaosPrincess has quit [Remote host closed the connection]
ChaosPrincess has joined #asahi-dev
Retr0id2 has joined #asahi-dev
Retr0id2 is now known as Retr0id
dingodoppelt has joined #asahi-dev
riker77_ has joined #asahi-dev
riker77 has quit [Ping timeout: 480 seconds]
riker77_ is now known as riker77
phiologe has joined #asahi-dev
PhilippvK has quit [Ping timeout: 480 seconds]
PyroPeter_ has joined #asahi-dev
PyroPeter has quit [Ping timeout: 480 seconds]
mini0n has joined #asahi-dev
kov has quit [Quit: Coyote finally caught me]
lawq has joined #asahi-dev
lawq has quit [Remote host closed the connection]
doggkruse has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
chadmed has quit [Ping timeout: 480 seconds]
chadmed has joined #asahi-dev
mini0n has quit [Remote host closed the connection]
MajorBiscuit has joined #asahi-dev
chadmed has quit [Ping timeout: 480 seconds]
chadmed has joined #asahi-dev
chadmed_ has joined #asahi-dev
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
the_lanetly_052__ has joined #asahi-dev
chadmed has quit [Read error: Connection reset by peer]
the_lanetly_052__ has quit [Ping timeout: 480 seconds]
<marcan> ha, that's a cute trick
<marcan> and that's cheaper than the ref boards
<marcan> hmm, tempting
<marcan> still need a drive enclosure to go with it, though
<chadmed_> i have a very cunning plan on that front, it's horrible and evil and revolting but it's cunning nonetheless
<marcan> there's always this kind of thing:
<marcan> not the prettiest though
<marcan> (and not a proper hotplug bay)
<marcan> or you could get a proper server backplane:
<ar> that looks like an HP-compatible cage
<marcan> yes, it is
<marcan> also, this one looks janky but more like what I'd want:
<ar> oh, it even says "DL380 G7"
<marcan> comes with the trays and no weird connections
<chadmed_> silverstone sell a micro atx case with a sas backplane on it. my idea was to erm..."secure" the mac mini where the motherboard goes, get an HBA and hook it up to one of those tb dev boards, take 5V from a USB port on the mac to a relay and have that trigger an ATX PSU to power the drives
<chadmed_> self contained, unsafe, impractical sleeper mac server :P
<chadmed_> perfect for all your production machine needs
<marcan> lol
<marcan> that works :')
<marcan> oh, they literally have just the trays, like that aliexpress link but less janky:
<marcan> I might want to go with that and just design/3D print something for the TB adapter bit
<rkjnsn> There are some SATA port multiplier enclosures that support hot plug and look pretty nice, if one is okay sharing one 6Gb/s link across multiple drives. The problem is that in my experience, Linux port multiplier support has been pretty terrible with every host controller + port multiplier combination I've ever tried it with.
<chadmed_> we have enough TB ports to just add more HBAs :P
<chadmed_> if youre adding enough drives to need more ports then the cost of a secondhand LSI card and a tb dev board (or your very nifty little solution) is vastly dwarfed by the cost of drives
<rkjnsn> Indeed.
<chadmed_> the case im looking at comes with 5 hotswap ports already with room to add 5 more using that very silverstone cage linked above so if youre spending $200 for the cage plus $x for the drives, another ~$200 on a tb board and HBA is worth it
<marcan> there's also this USB one which might not be bad?
<marcan> I might grab one and give it a spin, could easily move 4 drives into it from my existing NAS for now and see how they fare
<marcan> I wonder what chipset they use
<marcan> they at least clame sane speeds, unlike those chinese bottlenecked "raid" models
<Dementor[m]> marcan: some of these enclosures do their own raid which is really bad
<marcan> I know
<marcan> those all have terrible bottlenecks
<marcan> that one doesn't though
<marcan> I wonder what architecture it has, maybe just 4 controllers and a hub?
<Dementor[m]> I wonder if everything going out via one USB 3 cable is going to cause any bottleneck
<chadmed_> bandwidth wise on sequential reads probably not but id be interested in seeing random read/write performance
<Dementor[m]> Did anyone try to run Zfs on an M1?
<chadmed_> it probably wont even matter unless you put SSDs in there
<marcan> oh, a previous model had a USB3 DFP too, heh
<marcan> okay, that one is definitely a 5-port hub with 4 controllers then
<Dementor[m]> marcan: I assume that's a good thing as far as speeds right?
<marcan> chadmed_: I care about bandwidth, which is why I wouldn't use those terribad RAID enclosures (they are bottlenecked way, way lower than USB3 itself)
<marcan> but this one looks good
<marcan> if it can do >500MB/s aggregate that's good enough for me
<marcan> Dementor[m]: yes
<Dementor[m]> Also relaying on hardware raid can cause you to lose your data if the enclosure dies
<chadmed_> 99% of hw raid solutions just do not offer any real advantage over software anymore anyway apart from offloading the actual grunt work
<marcan> ah, the 5-bay model has the DFP
<marcan> so 6-port hub then?
<marcan> I don't use RAID anyway
<Dementor[m]> marcan: DFP?
<marcan> downstream facing port
<marcan> I'm going to grab the 4-bay model and give it a try
<marcan> 200 bucks for everything I need is reasonable (plus, being USB3, I can already use it with the Mac Minis on the A ports)
<marcan> (also good incentive to fix USB3 on the Mac Studio)
<Dementor[m]> marcan: are you going to try and run zfs on it?
<marcan> no, I use cephfs
<marcan> I'm going to get 3 of them and set up a proper 3-server cluster
<marcan> probably mac studio, m1 mac mini, and future m2 mac mini when it comes (maybe the 10G versions thereof)
<Dementor[m]> Isn't it a bit of an overkill haha
<marcan> not really, I already use cephfs, splitting it up among hosts is logical
<chadmed_> Dementor[m]: the question is never why, its why not
<Dementor[m]> I mean I have a home lab with a lot of over engineered stuff myself
<Dementor[m]> So I get it
<Dementor[m]> Would love to hear how well it works after you test it out
<marcan> just ordered one, if it works as advertised I'll get 2 more
<marcan> and probably just move my existing drives to them for now, and do the mac migration later
<Dementor[m]> We need a asahi-Nas channel 😂
<marcan> indeed :')
<Dementor[m]> marcan: the only thing you'll need is ipmi 😉
chadmed_ has quit [Quit: Konversation terminated!]
<marcan> well the 10G mini and mac studio kind of have that
<marcan> it's on the list to reverse engineer ;)
<maz> marcan: btw, do you have any write-up of what is needed to write the USB FW uploader on the Studio?
<marcan> not really, I need to look at it again
leitao has joined #asahi-dev
<marcan> it's just some ugly poking around the cfg space and BARs, throwing the blobs in in a weird interleaved form
<maz> OK. let me know if I can help, I have a vested interest in having it working! ;-)
<marcan> plus we need to write a firmware extractor for it
<marcan> I'll put it on my list of low-hanging fruit for the next couple of weeks :)
<dottedmag> marcan: Any idea why it is interleaved this way? I doubt Apple HW or SW engineer woke up one day and thought "let's interleave blobs because we like to make everything very complicated"
<maz> thanks!
<marcan> something something SRAM blocks
<marcan> some silicon engineer decided to take a shortcut
<marcan> this isn't apple, it's asmedia
<dottedmag> I see
<marcan> entirely possible this was intended as a "debug" interface only for bring-up or whatever, and apple just decided to use it in production because it fit their purposes better
<maz> they probably saved a handful of gates by doing so.
bisko has joined #asahi-dev
<sven> alright, I guess it’s time to submit Bluetooth this weekend and then get back to thunderbolt and see if it hates me less this time
Stary has quit [Quit: ZNC -]
<maz> sven: I'm sure it has mellowed down over time... ;-)
Stary has joined #asahi-dev
<_jannau_> that USB 5 bay storage case looks identical to sabrent DS-SC5B (US) or raidsonic ib-3805-c31 (EU). I didn't find any information which hub or usb2sata chips they use
<kettenis> wonder if someone could lobby asmedia to provide firmware dor that usb chip that is redistributable
chengsun_ has quit [Quit: Quit]
chengsun has joined #asahi-dev
MajorBiscuit has quit [Ping timeout: 480 seconds]
chadmed has joined #asahi-dev
MajorBiscuit has joined #asahi-dev
leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nsklaus_ has quit [Quit: Textual IRC Client:]
bisko has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
nsklaus has joined #asahi-dev
leitao has joined #asahi-dev
jluthra has quit [Remote host closed the connection]
jluthra has joined #asahi-dev
<Foxboron> Not sure if this is relevant, but Linus got an arch64 laptop and apparently building the Asahi kernel :)
<Dcow[m]> oh my ;) that's cool
<j`ey> Foxboron: yeah, an m2 one
<Foxboron> j`ey: Ah, so he is running Asahi. Cool
<j`ey> asahi kernel, but fedora
<mps> hm, is M1 better than M1pro
<mps> uh, is M2 better than M1pro
<Dcow[m]> in some way yes, in some no
<Tramtrist> got my M2 yesterday... according to the forums its a huge debate
<Tramtrist> screen is better M1pro for sure.. as is write speed of SSD
<Dcow[m]> I'd saw it's better than M1
<Tramtrist> though fanless/weight of M2 is big selling point
<Dcow[m]> say*
<mps> I'm waiting answer from reseller to see will I buy M1pro or M2
<Tramtrist> this is probably OT
<j`ey> yes OT for -dev at least
<mps> OT but helped me :)
<Tramtrist> ~~~~~~
mini0n has joined #asahi-dev
MajorBiscuit has quit [Quit: WeeChat 3.5]
roxfan2 has quit [Remote host closed the connection]
roxfan has joined #asahi-dev
___nick___ has joined #asahi-dev
___nick___ has quit []
___nick___ has joined #asahi-dev
<Dementor[m]> <marcan> "well the 10G mini and mac studio..." <- they have ipmi ?
roxfan has quit [Remote host closed the connection]
roxfan has joined #asahi-dev
leitao has quit [Quit: My MacBook has gone to sleep. ZZZzzz…]
SuperED has joined #asahi-dev
leitao has joined #asahi-dev
SuperED has quit [Remote host closed the connection]
___nick___ has quit [Ping timeout: 480 seconds]
<ping_47> Maybe is this the key to "rosetta" like experience?
<j`ey> yeah that package already installs the binfmt stuff
<ar> ping_47: that's what i basically do on my x86_64 desktop to emulate arm64 stuff
<ar> (when i'm too lazy to properly crosscompile stuff, and decide to let my cpu just brute-force through the emulation)
<ping_47> I'm trying to fix the qemu-user-static that's failing with current version of qemu, checking if the rc-0 of 7.1 fix it
<j`ey> although that package installs a binfmt thing for arm64 binaries, someone had an issue with that here
<ping_47> I'm step before trying to make the qemu static libraries to work first
<ping_47> one step*
dingodoppelt has quit [Quit: ZNC 1.9.x-git-170-9be0cae1 -]
<mps> ping_47: not sure for Arch but all this works fine on alpine
dingodoppelt has joined #asahi-dev
leitao has quit [Ping timeout: 480 seconds]
amarioguy has joined #asahi-dev
Wisp404 has joined #asahi-dev
psykose has quit [Remote host closed the connection]
psykose has joined #asahi-dev
ptrc has quit [Remote host closed the connection]
ptrc has joined #asahi-dev
Wisp404 has quit [Remote host closed the connection]