#
✕ nick:marcan
#asahi-re · 2022-02-12 11:35
marcan>>> ((41-0x44) & 0xff)
#asahi-re · 2022-02-12 11:35
marcan>>> ((41-0x44) & 0x7ff)
#asahi-re · 2022-02-12 11:35
marcan(why they did it this way, not the faintest clue)
#asahi-re · 2022-02-12 11:35
marcan* *11-bit*
#asahi-re · 2022-02-12 11:35
marcan68 is the additive inverse of 1980 in *11* math, not 8. maybe they have a char cast/variable somewhere where they meant a short, and 3 bits fell off.
#asahi-re · 2022-02-12 11:34
marcanbut for some bizarre reason, instead of adding 1980, they subtract 68 instead, then truncate to 8 bits and use that as the number
#asahi-re · 2022-02-12 11:33
marcanthey use a standard SBS command to get the date, and the format is standard; year is in years since 1980
#asahi-re · 2022-02-12 11:33
marcanit's a bug, lol
#asahi-re · 2022-02-12 11:33
marcanso I'm looking at SMC firmware and now the 290915 mfg date for my battery makes "sense"
#asahi-re · 2022-02-08 01:51
marcanthis is *especially* true for big unknown subsystems like graphics
#asahi-re · 2022-02-08 01:50
marcansee the policy
#asahi-re · 2022-02-08 01:50
marcanalso reminder that looking at kext code disqualifies you from working on the linux driver / etc unless I personally know I can trust your reversing hygiene
#asahi-re · 2022-02-08 01:49
marcanchadmed: kexts aren't obfuscated, they're just like that the way iokit works
#asahi-re · 2022-01-20 11:21
marcanlol the lua files and json files are encrypted?
#asahi-re · 2022-01-20 11:14
marcanheh
#asahi-re · 2022-01-20 11:14
marcan#endif
#asahi-re · 2022-01-20 11:14
marcan#define kPearlRepairSPC @"IPHONE COMP FACEID"
#asahi-re · 2022-01-20 11:14
marcan#if !RC_HIDE_D16 || !RC_HIDE_D17 || !RC_HIDE_D63 || !RC_HIDE_D64
#asahi-re · 2022-01-20 11:14
marcanwas not expecting to see *headers* in here
#asahi-re · 2022-01-20 11:13
marcan// CoreRepairKit
#asahi-re · 2022-01-20 11:13
marcan// CRFDR
#asahi-re · 2022-01-20 11:13
marcanso I managed to pull the field diagnostics image from Apple's CDN (normally loads only in field diags mode, but the loader is in the recovery mode image of course)
#asahi-re · 2022-01-20 08:07
marcanif you just want the Apple logo back while running m1n1, just do p.fb_restore_logo() or p.fb_shutdown(True)
#asahi-re · 2022-01-20 08:04
marcanyou can extract the im4p with img4tool and the image with something like https://github.com/realnp/ibootim
#asahi-re · 2022-01-20 08:04
marcanif you're looking for the actual Apple logo used by iBoot, it's in ipsw updates, e.g. Firmware/all_flash/applelogo@1x~mac-USBc.im4p, in a custom format
#asahi-re · 2022-01-20 08:02
marcanwhere are you trying to render it and how?
#asahi-re · 2022-01-20 08:02
marcannull-nop[m]: that file seems to be unused
#asahi-re · 2022-01-20 04:19
marcanI'm not sure exactly what you want to do
#asahi-re · 2022-01-20 04:19
marcannull-nop[m]: that code has nothing to do with M1s. BootX is for PowerPC Macs.
#asahi-re · 2022-01-16 03:56
marcanlove it when I'm trying to work out what some fields in a struct do, I look a the DCP code, and they are just... never initialized
#asahi-re · 2022-01-14 04:59
marcanI should probably write an article or give a talk about this some day... it's something that doesn't get talked about much in RE circles
#asahi-re · 2022-01-14 04:59
marcanthings get interesting when you get into the finer details of what is copyrightable and what is not; e.g. I've heard a legal opinion that register names are not copyrightable (but I'd still much rather write my own anyway, to be safe and because the official ones often suck)
#asahi-re · 2022-01-14 04:57
marcaninstead I'd much rather people always stick to safe things, like looking at input/output including MMIO and other activity traces
#asahi-re · 2022-01-14 04:56
marcan(and I've seen it go very wrong when someone didn't know what they're doing, and put an entire project in jeopardy, so I can't just let random people do it)
#asahi-re · 2022-01-14 04:55
marcanthere are a few people I trust to peek at binaries occassionally and still be able to write non-infringing code (including myself), but there is no good way to evaluate that other than through trust and watching people work over time
#asahi-re · 2022-01-14 04:54
marcanh_: that is correct, as a general rule
#asahi-re · 2021-12-17 10:08
marcanthey did indeed have something like this
#asahi-re · 2021-12-17 10:07
marcanJamie[m]1: sounds like this is the 32-bit era mailbox
#asahi-re · 2021-12-16 16:15
marcanJamie[m]1: I've looked at the firmware thing; I'm going to have some kind of fwcutter for the kernel in the installer
#asahi-re · 2021-12-16 07:28
marcanwhich is something we're going to want to use once we support this (probably via v4l2 m2m?)
#asahi-re · 2021-12-16 07:28
marcanffmpeg also now has support for the prores stuff, which means its prores decoding has scaffolding to hook into hwdec now
#asahi-re · 2021-12-16 07:26
marcan"leghorn" is what they call it
#asahi-re · 2021-12-16 07:26
marcan(on macos)
#asahi-re · 2021-12-16 07:26
marcanalso ffmpeg already has support for it, thanks to rcombs
#asahi-re · 2021-12-16 07:26
marcanthey do announce it in the plist though!
#asahi-re · 2021-12-16 07:24
marcanI guess apple decided they were allergic to naming the open devil codec since they have their hands so deep in the MPEG-LA pie :p
#asahi-re · 2021-12-16 07:24
marcanJamie[m]1: LGH is VP9
#asahi-re · 2021-12-14 14:57
marcanI think practically every reg write is strigified like this
#asahi-re · 2021-12-14 14:57
marcansven: whoever wrote that auto stringifying macro was a very nice person
#asahi-re · 2021-12-14 14:57
marcan "CS42L84_HPOUT_DACB_PATH_VOL_MSB",0x3007,"signBit",uVar1);

plus ›