#asahi-re · 2026-11-02 03:42
nickchanon newer devices at least#asahi-re · 2026-11-02 03:41
nickchaniboot has used EL0/1 since iOS 11, maybe you are referring to iBoot actually using TTBR1_EL1 since iOS 16/macOS 13?#asahi-re · 2026-07-21 07:10
nickchanso there's precedent#asahi-re · 2026-07-21 07:10
nickchanthe controls to modify the cpu p-state definitions are disabled when the kernel boots on t8103 already#asahi-re · 2026-07-19 08:55
nickchanthat doesn't have an hardware indiactor#asahi-re · 2026-07-19 08:54
nickchanDon't think so except on like macbook neo#asahi-re · 2026-07-11 04:19
nickchanor formally, any method that includes running the bare metal kernel, but not running sptm#asahi-re · 2026-07-11 04:18
nickchansptm emulator#asahi-re · 2026-07-11 04:13
nickchanBut yaeh it's a reset on the work#asahi-re · 2026-07-11 04:12
nickchanthat itself isn't banned#asahi-re · 2026-07-11 04:12
nickchancould still use ersatz sptm method if one desires that#asahi-re · 2026-06-30 02:22
nickchanthat paper has to be slop#asahi-re · 2026-06-29 17:29
nickchanthere's some other mistakes like saying the host page size is 4KB as well#asahi-re · 2026-06-18 07:47
nickchantop of the line performance per watt is always going to suck but it's still possible to be reasonably efficient if just a bit lower#asahi-re · 2026-06-18 07:46
nickchanthe need to have such core type#asahi-re · 2026-06-18 07:46
nickchanif there's no need for top of the line performance than I can see another core type that gives reasonable efficiency at higher performance states#asahi-re · 2026-06-18 07:45
nickchanecores are only efficiency at low to medium so I can see that mattering#asahi-re · 2026-06-18 07:45
nickchanhigher than a certain performance level the ecores is less efficient than the pcores#asahi-re · 2026-06-18 07:43
nickchandie size maybe#asahi-re · 2026-06-18 07:42
nickchanand ecores has smaller caches too so the architecure must not be much slower#asahi-re · 2026-06-18 07:42
nickchanbut at isofreq pcore and ecores are close#asahi-re · 2026-06-18 07:41
nickchanthe ecores are clocked lower#asahi-re · 2026-06-18 06:39
nickchanmcore is the middle core#asahi-re · 2026-06-18 06:37
nickchanusing the enginnering terms, the MPIDR looks like: PCORE 0x80010x0y, ECORE 0x80000x0y, MCORE 0x80040x0y; x is cluster y is core in cluster#asahi-re · 2026-06-18 06:35
nickchanand the inconsistency would show in the MPIDR numbers in the device tree#asahi-re · 2026-06-18 06:34
nickchan...if we adopt the marketing terms then we either have cpu_s0 for the a19 pro neo which apple does not think it is or say that it's cpu_p0 performance core depsite it being the same type of core as m5#asahi-re · 2026-06-18 06:34
nickchanapple enginnering don't think so either if you look at the new source drop, the new core type is all "MCORE"#asahi-re · 2026-06-18 06:33
nickchanyes#asahi-re · 2026-06-18 06:32
nickchaneven though it's literally the same type of core as the ones in M5/M5 Pro/M5 Max#asahi-re · 2026-06-18 06:32
nickchanthen if an A19 Pro macbook neo ever existed, Apple would consider its biggest cores to be "Performance cores"#asahi-re · 2026-06-18 06:31
nickchanif we refer to the biggest core type in M5, M5 Pro, M5 Max as super cores or cpu_s0 or whatever in the dts file#asahi-re · 2026-06-18 06:31
nickchan1. biggest core type in A-series and M-series SoC is the same class of core from M1-M5 #asahi-re · 2026-06-18 06:30
nickchanthis is the real problem, if we used references to "super cores" in any not marketing material, it's going to become a problem if A19 Pro macbook neo ever exist #asahi-re · 2026-06-18 06:29
nickchanand according to strings on the A19 Pro iphone kernels, apple does not mark a19 pro having super cores#asahi-re · 2026-06-18 06:29
nickchanso m5; m5 pro; m5 max has super cores#asahi-re · 2026-06-18 06:29
nickchancluster types#asahi-re · 2026-06-18 06:29
nickchanhttps://github.com/apple-oss-distributions/xnu/blob/ac9718fb1af618d5ce8678d0dc6e8a58f252216f/osfmk/arm/machine_routines_common.c#L68#asahi-re · 2026-06-18 06:29
nickchannew xnu source drop, now I really do not recommend referencing "super cores" in any dts file or non-marketing material: #asahi-re · 2026-05-13 15:18
nickchanhabiby: checkra1n ios15 wasn't coming out but i want to jailbreak iphone x so i download ghidra from github and disassembled xnu and amfi patched some stuff without really knowing what im doing and repeat until it worked#asahi-re · 2026-05-13 15:05
nickchanto see if it holds a very specific value during bootup if so corrupt that register to a usable sp and restore the specifc value when im done#asahi-re · 2026-05-13 15:05
nickchanwell ""solution""#asahi-re · 2026-05-13 15:04
nickchanand my solution is to check a register for a very specific value through 64 tbz/tbnz chain#asahi-re · 2026-05-13 15:04
nickchanthere's also the issue that around when xnu is going to enable mmu xnu doesn't have a usable sp but I need to save state#asahi-re · 2026-05-13 14:57
nickchan (unfortunately patching pmap_enter is not sufficient to trace things properly)#asahi-re · 2026-05-13 14:57
nickchani would rather do shadow paging to trace things in same el then to disassemble xnu and pretend i know iokit#asahi-re · 2026-05-13 14:34
nickchansome IDA register names has always feel a bit sus#asahi-re · 2026-05-13 14:34
nickchanchaos_princess: erm where is that file supposed to go#asahi-re · 2026-05-13 14:29
nickchanso anything that use varargs is just broken#asahi-re · 2026-05-13 14:29
nickchandisassembling apple bins in ghidra kind of suck though it doesn't even support the call convention properly#asahi-re · 2026-05-04 14:16
nickchan+ running windows guests assuming hacr_el2 is also like that but it is not like running vms work at all in raw mode