#
✕ nick:davidrysk[m]
#asahi · 2021-01-08 04:19
davidrysk[m]ifixit took off the shields
#asahi · 2021-01-08 04:18
davidrysk[m]https://www.ifixit.com/News/46884/m1-macbook-teardowns-something-old-something-new
#asahi · 2021-01-08 04:14
davidrysk[m]it's likely that the FL1100 chips are covered with shields
#asahi · 2021-01-08 04:12
davidrysk[m]that's the more interesting of the teardowns
#asahi · 2021-01-08 04:12
davidrysk[m]https://egpu.io/forums/desktop-computing/teardown-late-2020-mac-mini-apple-silicon-m1-thunderbolt-4-usb4-pcie-4/
#asahi · 2021-01-08 04:11
davidrysk[m]yeah, script is there too
#asahi · 2021-01-08 04:10
davidrysk[m]okay :)
#asahi · 2021-01-08 04:10
davidrysk[m]macos does come with screen
#asahi · 2021-01-08 04:10
davidrysk[m]it forks and then sandboxes and then chroots
#asahi · 2021-01-08 04:10
davidrysk[m]there's something with the sshd sandboxing mechanism that doesn't play nice with recovery
#asahi · 2021-01-08 04:09
davidrysk[m]I'm still poking at sshd. how much effort should I put into that?
#asahi · 2021-01-08 04:09
davidrysk[m]well yeah
#asahi · 2021-01-08 04:08
davidrysk[m]what matters more, USB on the mini only, or USB on all the M1 machines? :)
#asahi · 2021-01-08 04:07
davidrysk[m]I mean, I wouldn't cry if they weren't supported initially
#asahi · 2021-01-08 04:06
davidrysk[m]which matches the teardown iirc
#asahi · 2021-01-08 04:06
davidrysk[m]yep
#asahi · 2021-01-08 04:06
davidrysk[m]I'm testing on a MBP
#asahi · 2021-01-08 04:06
davidrysk[m]the 2x TBT3/USBC ports are not
#asahi · 2021-01-08 04:06
davidrysk[m]the 2x USB 3.0 ports are attached via PCIe
#asahi · 2021-01-08 04:06
davidrysk[m]well that's why
#asahi · 2021-01-08 04:06
davidrysk[m]marcan: is that perchance the mac mini?
#asahi · 2021-01-08 03:43
davidrysk[m]this probably is a n00b question, but how does the devicetree show that the USB is connected via PCIe?
#asahi · 2021-01-08 03:22
davidrysk[m]hm I should attach a thunderbolt device and see where it appears in the ioregistry
#asahi · 2021-01-08 02:49
davidrysk[m]marcan: I extracted the device tree (needed to use tihmstar's img4 extractor first) and it's not exactly easy to read :)
#asahi · 2021-01-08 01:43
davidrysk[m]should I keep poking at ssh into recovery?
#asahi · 2021-01-08 01:40
davidrysk[m]I'm wondering if "usb-drd0" is a mapping to something else
#asahi · 2021-01-08 01:39
davidrysk[m]IOService:/AppleARMPE/arm-io@10F00000/AppleT810xIO/usb-drd0@82280000/AppleT8103USBXHCI@00000000/usb-drd0-port-hs@00100000/USB 3.0 FD@00100000/IOUSBHostInterface@0/IOUSBMassStorageInterfaceNub/IOUSBMassStorageDriverNub/IOUSBMassStorageDriver/IOSCSILogicalUnitNub@0/IOSCSIPeripheralDeviceType00/IOBlockStorageServices/IOBlockStorageDriver/PNY USB 3.0 FD Media
#asahi · 2021-01-08 01:39
davidrysk[m]I have a USB drive plugged in and it appears at
#asahi · 2021-01-08 01:35
davidrysk[m]I see that this gives away which firmware binaries the BRCM card requires
#asahi · 2021-01-08 01:35
davidrysk[m]I think the device tree but it tends to be super-confusing
#asahi · 2021-01-08 01:31
davidrysk[m]but the USB itself is connected to arm-io
#asahi · 2021-01-08 01:30
davidrysk[m]so there's an I2C connection to the USB, probably for control purposes
#asahi · 2021-01-08 01:29
davidrysk[m]They are super-handy for things like this
#asahi · 2021-01-08 01:29
davidrysk[m]there's a bunch of not so well known tools part of the "Additional Tools for Xcode" that are on the Apple More Downloads site.
#asahi · 2021-01-08 01:28
davidrysk[m]marcan: you gave me an idea, lemme grab the IORegistryExplorer :)
#asahi · 2021-01-07 21:15
davidrysk[m]ConeOfAttack: I'd just summarize it; people can click on the link if they need to
#asahi · 2021-01-07 21:15
davidrysk[m]the "fabric" has an L3-like cache that is shared by the various components
#asahi · 2021-01-07 21:14
davidrysk[m]the DRAM is on-package, not on-die
#asahi · 2021-01-07 21:13
davidrysk[m]I'd like to see more exhaustive tests like these
#asahi · 2021-01-07 21:13
davidrysk[m]https://lemire.me/blog/2021/01/06/memory-access-on-the-apple-m1-processor/
#asahi · 2021-01-07 21:12
davidrysk[m]I believe something around 500 but I'm not sure
#asahi · 2021-01-07 21:11
davidrysk[m]it's not a traditional L3 cache because it's shared between the CPU, GPU, ML, and other cores
#asahi · 2021-01-07 21:10
davidrysk[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.
#asahi · 2021-01-07 21:10
davidrysk[m]ConeOfAttack: when you send a too-long message from matrix it gets turned into a link on IRC
#asahi · 2021-01-07 20:56
davidrysk[m]no
#asahi · 2021-01-07 20:55
davidrysk[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.
#asahi · 2021-01-07 20:54
davidrysk[m]if you have an intel mac you can run self-signed binaries in recovery which is useful for testing
#asahi · 2021-01-07 20:54
davidrysk[m]Mary_: I got as far as having it try and fail to chroot
#asahi · 2021-01-07 20:43
davidrysk[m]hm not EACCES but EPERM
#asahi · 2021-01-07 20:41
davidrysk[m]it can't chroot, due to EACCES

plus ›