#asahi-re · 2026-03-04 18:31
nicolas17https://developer.apple.com/services-account/download?path=/macOS/Kernel_Debug_Kit_26.3.1_build_25D2128/Kernel_Debug_Kit_26.3.1_build_25D2128.dmg#asahi-re · 2026-03-03 19:08
nicolas17https://github.com/apple-oss-distributions/xnu and other repos were updated last night#asahi-re · 2026-02-19 20:20
nicolas17oh >.>#asahi-re · 2026-02-19 20:07
nicolas17amarioguy: why not, employer?#asahi-re · 2025-10-28 21:26
nicolas17not the kexts#asahi-re · 2025-10-28 21:26
nicolas17also the source dump is for the kernel alone#asahi-re · 2025-10-28 21:19
nicolas17I'm actually not sure if ghidra can use complete debugging symbols (with line numbers and local variable names) even if they were present...#asahi-re · 2025-10-28 21:18
nicolas17you can cross-reference it to the source code then#asahi-re · 2025-10-28 20:51
nicolas17if you're lucky, there's symbols and functions get names :D#asahi-re · 2025-10-28 20:51
nicolas17usually local variables and parameters never get names when decompiling native code#asahi-re · 2025-10-16 20:12
nicolas17I only have a J493#asahi-re · 2025-10-16 20:11
nicolas17and you need it filled in by the boot process rather than extracted from the ipsw?#asahi-re · 2025-10-16 17:43
nicolas17IIRC there's A19 and A19 Pro benchmarks showing different latency for certain things that would suggest they are not using "the same" cores#asahi-re · 2025-10-16 17:43
nicolas17chadmed: I think they stopped naming their CPU cores and it's now "the p-core of SoC X"#asahi-re · 2025-10-16 02:11
nicolas17there's now exclaves on mac#asahi-re · 2025-10-16 01:32
nicolas17wait Apple screwed up and didn't upload the fcs/wkms keys so you won't be able to mount the fs -.- can look at kernelcache and boot components though#asahi-re · 2025-10-16 00:39
nicolas17MacBook Pro 14" M5 (J704AP, Mac17,2): https://updates.cdn-apple.com/2025FallFCS/fullrestores/093-49300/3AC98479-C75E-46E8-8AF4-1DDB5EAE564D/UniversalMac_26.0.1_25A8364_Restore.ipsw#asahi-re · 2025-02-22 23:49
nicolas17mood#asahi-re · 2025-02-22 23:46
nicolas17so I wrote a script to erase the data volume from the disk image, going through the APFS object map to figure out what blocks belong to that volume#asahi-re · 2025-02-22 23:45
nicolas17and that's useless data since you can't decrypt it anyway#asahi-re · 2025-02-22 23:45
nicolas17but the disk image takes a lot of space because the data volume is encrypted and can't be compressed#asahi-re · 2025-02-22 23:44
nicolas17scenario: new Mac model comes out, it comes preinstalled with a special macOS build that is not available for download, researchers / data hoarders want a copy of it so they take a raw disk dump#asahi-re · 2025-02-22 23:43
nicolas17I wrote a hackish purpose-specific apfs parser in Python a while ago#asahi-re · 2025-02-22 23:43
nicolas17lol#asahi-re · 2025-02-22 23:41
nicolas17the second device doesn't have to be a Mac (it might need USB-C?)#asahi-re · 2025-02-22 23:39
nicolas17let me know if you want some small apfs disk images to poke at :p#asahi-re · 2025-02-22 23:39
nicolas17do you have no Mac hardware currently?#asahi-re · 2025-02-22 23:34
nicolas17chaos_princess: we should actually write down somewhere the errata for Apple-File-System-Reference.pdf :D#asahi-re · 2025-02-22 23:31
nicolas17I once tried to trace wifi and got quickly overwhelmed#asahi-re · 2025-02-22 23:30
nicolas17:D#asahi-re · 2025-02-22 23:29
nicolas17ignoring stuff made it able to keep up?#asahi-re · 2025-02-22 23:29
nicolas17how were the timeouts solved? just optimizing the tracer code?#asahi-re · 2025-02-22 23:27
nicolas17chaos_princess: I wonder if installing to an external volume would reduce the SKS spam... afaik you don't get SEP-tied data protection then#asahi-re · 2025-02-22 23:25
nicolas17ugh do you need gigalocker to do *anything* with the sep?#asahi-re · 2025-02-22 23:22
nicolas17so yeah you absolutely cannot avoid the SEP, it's impossible to eg. get a fingerprint image from the sensor, the secure enclave does all the matching#asahi-re · 2025-02-22 23:21
nicolas17that's how they communicate, code in the main CPU has to transport the data between the two while being unable to decrypt it#asahi-re · 2025-02-22 23:20
nicolas17the secure enclave replies with another encrypted blob which the driver passes to the sensor#asahi-re · 2025-02-22 23:20
nicolas17instead, the macOS kernel driver reads some encrypted data from the sensor and passes it untouched to the secure enclave#asahi-re · 2025-02-22 23:19
nicolas17Zavec: iirc the fingerprint sensor and the secure enclave are not physically connected#asahi-re · 2024-10-25 06:17
nicolas17https://updates.cdn-apple.com/private-cloud-compute/a1245d6d56b3de6009d1bbfe452262ff2e953034636e9b47d83e4c2a2d2540f7 I think this is the most recent pcc ipsw#asahi-re · 2024-10-25 06:03
nicolas17and VM images, including cleartext sepOS#asahi-re · 2024-10-25 06:03
nicolas17Apple has released Private Cloud Compute security docs https://security.apple.com/documentation/private-cloud-compute#asahi-re · 2024-08-14 16:30
nicolas17riiiight#asahi-re · 2024-08-14 16:30
nicolas17"cracking/reversing" "payment in cryptocurrency" ... "legitimate interests"#asahi-re · 2024-08-14 01:19
nicolas17Hoidine90: what software?#asahi-re · 2024-08-08 15:57
nicolas17they promised cleartext sepOS in the PCC images but there are no PCC images yet#asahi-re · 2024-07-23 19:56
nicolas17(note that SEP is still encrypted on macOS/iOS, and there are no PCC images yet)#asahi-re · 2024-07-23 19:56
nicolas17guess they realized keeping it encrypted would be even less useful than it already is, if PCC's was public#asahi-re · 2024-07-23 19:54
nicolas17but we thought they would only publish it for PCC#asahi-re · 2024-07-23 19:54
nicolas17yeah