#development
1 messages · Page 128 of 1
for once chatgpt said something helpful
but sould i use setClass or hook layoutsubviews ?
because I need to hook something, before procursus is initialized
what the fuck do you need object_setClass for
what are you trying to do?
not hooking layoutsubviews
apparently he only needs to change stuff only in one instance of a class
but like theres many other ways of doing this
honestly object_setClass is more cursed than hooking layoutSubviews
I guess I can use another hooking library tho
yea, basically i want to override the positioning of two views
https://github.com/Rprop/And64InlineHook I could probably just use this
but, the other place that would be help full is removing like little gliches in the border radius for the artowrkView
I've used it before
have fun
me?
yes
or you could just link it normally..
I don't know iOS's link paths!
@grave sparrow like fixing this
because I'm doing this before the dyld patching
what? it would work, and unless i hook layout subview to apply the radius, it does this
its a UIImageView
wait meta made fishhook??
yes
amazing
I can't use it anyways because it only works for hooking functions with symbols
I don't think I can use And64InlineHook cuz CFI/PAC
the problem is I need to hook a function before MobileSubstrate or any dynamically linked hooking library is available
no during process execution
(opainject)
nooo
I want to hook launchd
I inject my tweak using opainject
but there aren't any hooking libraries available at this point in the jailbreak flow
because I'm making a jailbreak, WITHOUT jbd because opa said launchd is the new way
why is that so damn hard to understand
cuz it’s capt
lmao
y'all aren't very helpful, I'll just figure it out myself
I literally just told you oh my god
what does this have to do with userspace reboot
the launchd hook is being used instead of jailbreakd, as opa TOLD ME that's the better way to do it
because it's what I'm doing?!
no the problem is
is this is before fakelib
I don't know iOS's link paths
where dyld searches for dynamic libraries??







oh my god
do you not understand how linking on *os works at all
i think you should learn how linking on *os works
to summarize, where to find dylibs is specified in the binary
ohhh
I see
sudo install_name_tool -add_rpath /var/jb/baseboin/ Jupiter.dylib
so basically just this
and it'll let it search in that path, correct
yes and no
oh does it not allow symlinks
Alot of things don't like symlinks
especially in *os, at least in my experience
here is my tweak
/var/jb/Library/MobileSubstrate/DynamicLibraries/Dahlia.dylib (compatibility version 0.0.0, current version 0.0.0)
/usr/lib/libMobileGestalt.dylib (compatibility version 1.0.0, current version 1.0.0)
@rpath/CydiaSubstrate.framework/CydiaSubstrate (compatibility version 0.0.0, current version 0.0.0)
/usr/lib/libSystem.B.dylib (compatibility version 1.0.0, current version 1292.100.5)
/System/Library/Frameworks/UIKit.framework/UIKit (compatibility version 1.0.0, current version 4218.1.100)
/System/Library/Frameworks/Foundation.framework/Foundation (compatibility version 300.0.0, current version 1775.118.101)
/usr/lib/libobjc.A.dylib (compatibility version 1.0.0, current version 228.0.0)
/usr/lib/libc++.1.dylib (compatibility version 1.0.0, current version 905.6.0)
ok i cut out a bunch of useless stuff
anyways
it will only search that rpath you add for stuff that specifically references rpath
aka @rpath/CydiaSubstrate.framework/CydiaSubstrate
otherwise it will just use the file path explicitly specified
ah
@rpath ftw
yea but that's just /S/L/F etc so pretty irrelevant
the fallback stuff is basically never used in today's world
so can I just directly tell it where the library is, because it should always be in the same place
ye
you can
generally the consensus is rpath
idk why actually
that wouldn't work because the env is already set, because the tweak is being loaded with opainject
once a program is fed the environment, you can't change it, right
so, like this sudo install_name_tool -add_rpath @rpath/var/jb/baseboin/ Jupiter.dylib
so i can add these rpaths
- /usr/local/lib
- /var/jb/usr/local/lib
etc
and if i specify the dylib name to link to as as @rpath/libabc.dylib
then it will search for /usr/local/lib/libabc.dylib and /var/jb/usr/local/lib/libabc.dylib
arguably if this is going to be for the jailbreak process just hardcode it
idk up to you
yeah lol
because if I do that then I also don't have to rename ellekit's dylib
sudo install_name_tool -change MobileSubstrate.dylib /var/jb/baseboin/ellekit.dylib Jupiter.dylib so something along these lines
/Library/Frameworks/CydiaSubstrate.framework/CydiaSubstrate (compatibility version 0.0.0, current version 0.0.0) this is the line I want to change, right
you can specify a couple rpaths
and it'll check each one
at least from what i understand
I know it's used in repack-rootless.sh but there are prob a lot of other cool uses for it
never hurts to have cool features ig
which dylib? this one, right libellekit.dylib
there's multiple and idk which one I need 😓
yeah that seems to be the right one
anyone know what the panic CS_KILLED initproc exited means?
ohhh
wait
It does that because it isn't launched as debugged
I'm guessing
alright
I know how to fix it lol
anyone have any clue where the dsc is on iOS 17
/private/preboot/Cryptexes/OS/System/Library/Caches/com.apple.dyld
thanks
i deleted this file and my device is buggin
please help
(i was running out of storage and this file was really big)
apfs exploit
guys, I'm trying to send an XPC message to launchd, but it just gives me an I/O error
anyone know why
well yes I knew that
I basically just copy pasted code from dopamine
and I'm not launchd
soooo
this is literally the code ```c
struct _os_alloc_once_s {
long once;
void *ptr;
};
struct xpc_global_data {
uint64_t a;
uint64_t xpc_flags;
mach_port_t task_bootstrap_port; /* 0x10 /
#ifndef _64
uint32_t padding;
#endif
xpc_object_t xpc_bootstrap_pipe; / 0x18 */
// and there's more, but you'll have to wait for MOXiI 2 for those...
// ...
};
extern struct _os_alloc_once_s _os_alloc_once_table[];
extern void *_os_alloc_once(struct _os_alloc_once_s *slot, size_t sz,
os_function_t init);
xpc_object_t launchd_xpc_send_message(xpc_object_t xdict) {
void *pipePtr = NULL;
if (_os_alloc_once_table[1].once == -1) {
pipePtr = _os_alloc_once_table[1].ptr;
} else {
pipePtr = _os_alloc_once(&_os_alloc_once_table[1], 472, NULL);
if (!pipePtr)
_os_alloc_once_table[1].once = -1;
}
xpc_object_t xreply = nil;
if (pipePtr) {
struct xpc_global_data *globalData = pipePtr;
xpc_object_t pipe = globalData->xpc_bootstrap_pipe;
if (pipe) {
int err = xpc_pipe_routine(pipe, xdict, &xreply);
if (err != 0) {
AppendLog(@"Error on sending message to launchd! %s",xpc_strerror(err));
usleep(500);
return nil;
}
}
}
return xreply;
}
and then I just get I/O error
do I need special entitlement?
@granite frigate
is 472 correct
¯_(ツ)_/¯
What are some good debuggers that can run directly on iOS to show the unmangled assembly?
lldb
thanks
@grave sparrow
I'VE ONLY GOTTEN MORE FUCKED SINCE OUR LAST MEETING COUNT
TWICE THE FUCKED DOUBLE THE FALL
I now can't feel my arms and hands

mangling? bro is debugging swift
I’ll steal this 
Fake
What
ellekit 1.1.1 thing
oh
Can anyone tell me how to extract sandbox profiles starting from an ipsw?
i remember Corellium had a sandblaster fork to extract them for iOS 17 betas, but idk
Okay thank you very much. One more question but task_for_pid doesn't work on iOS 16 anymore?
It was Cellebrite
Yes yes I found it later I will try. Now I was trying to figure out why task_for_pid doesn't work. In xnu it says it must be a privileged process, I start my process with root, like trollstorehelper does
Oops thanks
I was trying to find it and couldn't lmao
Damn i wonder how hard it'd be to intern there
Not too difficult if you try hard at projects and make yourself known
show code
I mean coming from TS dev... how hard was it to go from making regular apps to finding a CT bug from diffing lol
It wasn’t really that though
Making apps -> diffing CoreTrust took about a year
I also wrote Achilles and published a blog post on that
this spawn with root, and then kern_return_t kr = task_for_pid(mach_task_self(), pid, &task);
can it be related to the fact that I am starting my process from inside private/var/mobile ?
yes i have this entitlements:
<key>com.apple.system-task-ports.control</key>
<true/>
<key>com.apple.system-task-ports.token.control</key>
<true/>
<key>com.apple.system-task-ports</key>
<true/>
<key>task_for_pid-allow</key>
<true/>
<key>com.apple.security.get-movable-control-port</key>
<!-- Allows for task_get_special_port(..TASK_KERNEL_PORT..) -->
<true/>
get-task-allow
In my process yes, the process from which I want to get the task port no
you're missing the entitlement here though
I have this entitlements
on the child process?
yes in the process run as root, from this process I would like to take the Springboard task port for example
a so if my binary is not platform I can't take task ports of platform processes ?
i believe so
but in ios 15 it wasn't like that. and even in XNU I don't see this control.
IN XNU:
/*
- Routine: task_for_pid_posix_check
- Purpose:
-
Verify that the current process should be allowed to -
get the target process's task port. This is only -
permitted if: -
- The current process is root -
OR all of the following are true: -
- The target process's real, effective, and saved uids -
are the same as the current proc's euid, -
- The target process's group set is a subset of the -
calling process's group set, and -
- The target process hasn't switched credentials. - Returns: TRUE: permitted
-
FALSE: denied
*/
/*
- Routine: task_for_pid
- Purpose:
-
Get the task port for another "process", named by its -
process ID on the same host as "target_task". -
Only permitted to privileged processes, or processes -
with the same user ID. -
Note: if pid == 0, an error is return no matter who is calling. - XXX This should be a BSD system call, not a Mach trap!!!
you can check that with a non-platform binary
yes of course, which one do you say to test on ?
👁️
its beter to have less %hooks in tweaks right?
the same error
yes
@limpid pumice do you hve any snapchat tweaks?
my uid e my guid is 0
From what other path could I start it to check if this is the problem ?
from the /var/jb symlink since that'd be outside var
do check the console app logs though, there might be a clue there
Inside /var/containers/Bundle
hrtowii
you
never responded

I'll ask again; Why when sending a message to launchd the same way launchctl does it, why do I get an I/O error?
depends
but generally yes
like instde of hooking somthing just to hide it, do it from the super view wich is a "needed" hook
#import "server.h"
%hookf(void *,xpc_server_thing,void *a1,void *a2,xpc_object_t msg,void *a4) {
if(!server_hook(msg)) {
return %orig;
} else {
return 0x16;
}
}
void initme(void *addr) {
%init(xpc_server_thing = addr);
}
``` my hooks
lmao

i have no idea what that means
spinlock panics yay
what??
they do?
yeah
why
idk if the issue is fixed on iOS 16 though
I mean one or two hooks should be fine I guess
the more c func hooks you have, the greater the cahnce
yea that
Didn't opa say it was just an iOS 15 issue tho
¯_(ツ)_/¯
like i instead of %hooking the volume bar to hide it, i could just hide it from its superview hook
i mean it doesnt really matter that much
fair enough
hMMM
still confuzzled as to why I/O errors
but
¯_(ツ)_/¯
I don't think it depends on the path, on preboot I can't write, on /var/containers/Bundle, I get the same error
don't tell me what to do
wat
it's telling me getsegbyname is deprecated
nah it doesn't matter it still works lol
bad
also
either use camelCase or snake_case
writingvarnameslikethisisunreadable
I don't follow naming schemes, I'm too lazy
although there's probably a VSC extension to do it for me lmao
honestly probably yeah
@timid furnace May I asketh a question
iOS 15+
this should be the official forkfix icon
question
if theres entitlements that require a dev account to sign with, and i sign with a free acc will the installed binary on my phone have those dev acc entitlements?
lmao
It wouldn’t install
yea
so if its signed with a free account then the dev entitlements would be removed?
alfie can I ask u question
The binary can have any entitlement as long as the provisioning profile has them too
i meant to say if u sideloaded w the free acc then u'd have the free acc entitlements and vice versa
No it wouldn't let you install it
It just wouldn’t let you install the app with those entitlements
Sure
I tried adding some dev account entitlement and it just refused
debatable
Do you know why when I send a message to launchd the same way as launchctl does, why it gives me an I/O error?
mf just called me out
Who said u can talk 😭😭
CGRect *rect = [[CGRect alloc] init];
(this launchctl https://github.com/ProcursusTeam/launchctl)
real
hmm tryna figure out how gameseagull for regular sideload
No idea sorry
and tryna patch the need for the team id
aw man :(

CGRect ***rect = *&[[CGRect alloc] init];
Hmm
Maybe it’s an iOS limitation
I told my beta testers it'd probably be ready this weekend but I didn't realize I'd get stuck here 
bro misunderstands pointers 💔
let me be
why do you make it a pointer and then dereference it
why not
CGRect *rect = [[CGRect alloc] initWithFrame:CGRectMake(20,20,20,20)];
@indigo peak send me an ipa i can test with enterprise cert
thats the point
talk ur shit
fixes:
word hunt crashing when hack enabled
mini golf not working when hack disabled
added:
darts aimbot
cup pong auto shoot button
archery target always at the closest distance
bro this mf
NW
Block
BRO
how can i tell if a view exsists other than like if(self.view)
web discord
Nightwind is such a goofy goober
@slender glade what is personal 3 reloaded
@granite frigate I know u did a bunch with launchd, do u know
its oop
advanced placement
remake of P3
Is persona 3 good
yes
Uh nope!
mAN!
its amazing, magnificent
@radiant idol , and other than if(view with tag 888)
apparently I'm just having some weird issue!
do u have a problem
yes
if (self.view)
Im tryna get into some of those anime story games like persona and danganronpa
play p5r or p3r first
I been replaying all the yakuza games for lad infinite wealth
if (self.view != nil) or if (self.view != NULL)
p3r is on gamepass so...
bro just do if (self.view) lol
Why not 1-2-3-4-5
if (self.view) { // Code to be executed if self.view exists // Use 'self.view' within this block } else { // Code to be executed if self.view is nil }
the most objc i know 
nah persona doesnt work like that
the first game is from 1994 and doesnt play out well now
if (!(!(!(!(self.view == nil))))) this is the other way that I know
mfs DRM'd the soundtrack 💀
big mistake
it's an exe
ok, this also might eb a fior/christian question but, rember how you like add your own view property to a view. Does that property only show up when you add it in the hook?
yea its a separate app
Then how tf ima know what happened first
for the collectors ed
each game plays out on its own they're not connected
Word
does persona use nexus drm
@indigo peak / @hexed knot this is a @indigo peak / @hexed knot question
97% of reviews are positive lol
yes
yes
when you do %property
Ok its a @indigo peak question
only way for it to actually exist
muchas gracias
@slender glade so 5 or 3 first
because if you just define it in the interface it and try to access it, youll crash
ofc nightwind says that
trust me
i trust fiore
God damn bruh
60 dollars
70 dollars
Guess ima get gamepass and play on my pc
Who the fuck pirates pc games
bad idea
fr
Aaron!!!!!!!!!!
trust capt instead
Whichever u can get access to the easiest
But I’d prefer 5 royal first
even worse idea
Yea I just use gamepass lol
bet
yup
what is a difference between task_name_for_pid vs task_for_pid?
Jus gotta finish yakuza lad then i can play it
task names and tasks are different
a task name is probably literally the name of the task lmao
iCraze can’t even put that into perception
@indigo peak u listen to ji?
only a couple songs
💀
having the task name is not possible to get the task ?
mainly this one
Type shit
Sega getting a little too excited bro 70$ for new games
the task is an object represented by a number that lets you read and write it's memory and stuff like that
that's the new norm
what task are you trying to get
well that’s too much imo
I agree but even nintendo is doing it lol
literally
and now it's impossible to crack games cuz denuvo
Nah im cool with the 70 its worth it in this case, just fuck them for locking new game+ behind a 15 dollar paywall
first the erode the real value of games and now they increase the nominal cost too
wen eta actually good game 
getting double dipped
crazy transphobic russian women:
Oh AND some achievements are locked behind the dlc which is stupid
I would like to get SpringBoard, but task_for_pid does not work on any process even having both entitlements and uid 0
yeah literally only empress
And
whats your ultimate goal with springboard's task?
apparently
and empress only cracks the games they're interested in
or games that make her popular
yeah
apparently she hates rdr2 but cracked it twice cuz it made her famous
just going to read objects in memory, I don't have a specific goal I'm just studying how ios 16 works, but I can't understand why task_for_pid doesn't work
yeah lmao
my dad just bought rdr2 yesterday and he says it's amazing
Best rockstar game
I'm more of a square enix guy
Do they even make new games
I'm willing to pay for square's games
square enix needs to hire people to come with titles instead of ai
'triangle strategy'
'harvest villa'
One of the only AAA publishers I'll pay for
Last square enix game i can recall that isnt ff or some other shitty game is life is strange true colors
Amazing game
Last square enix game I can recall that isn't FF is KH3
Still waiting on updates on KH4
they announced it like 3 years ago and haven't said anything since

probably cuz they've been focusing on FF7 Rebirth
which I preordered a while ago
Damn i should replay life is strange true colors lowkey
I bought that one physical too
i've never played a ff game but considering there's like 20 it's definitely not their final fantasy
those mfs are living
Neither have I
Frl
lmao
but I had to preorder FF7 Rebirth because it looked really amazing
Start with FF7 afaik, it's the first of the modern ones iirc
although we're on
FF XIV rn
(16)
When you submit an app to TestFlight/App Store do you send the source?
can i %group method hooks ?
good question
I mean I'd assume so
No, they just want the binary
really?
Okay thought so
Yh
custom "Lets play 8ball! message"
It’s been a while since I’ve submitted something
so real

I mean would facebook really submit their src to app store
ur him
Wrong reply but yh


is there a list of all the initproc error codes somewhere?
@sonic totem is it better to do type manipulation in ida and then start hand editing, or should i do it straight from hand editing
Type manipulation?
better
fixing up types
Also imagine affording IDA 
ie. IDA thinks pointer is one type, it is actually another type
better
Oh hm
I just guess

no i mean
there's only two structs here
I take an educated guess based on symbols given
it's just is it faster to change the type in ida or to edit it in the source
i think i'll just do it in the source
Source?
decomp source
i already started editing it and shoving all of these macros into an enum for IDA is going to take a while
like if I know what types a function call takes, and it passes one of the arguments into another function, and I have that symbol
protip edit structs in the local types view
then I can easily infer the type
imagine having IDA Pro
yea but that doesn't help me when all these constants are defined as macros

Ghidra sucks
IDA ON TOP.

alright back to crying in decomp i go
what r u re'ing
stuff related to arm64 installd bug
the problem isn't that i don't know which type it is
the problem is ida doesn't know and picked the wrong type
ghidra:
fr
ie
it thinks v82 is fat_arch*
it is actually fat_arch_64*
sizeof(fat_arch_64) == 32
that looks like a struct
that is what i mean by fixing all the types
but doing this in ida, reexporting, and then fixing up all the constants is going to take too long
so i will just do it by hand
i know what you meant
but i dont think there's a quicker way to do this
I've scoured the internet but apparently noone uses IDA on mac
true then
(wait I can just use wine)
Jtool
the latest version of mac ida leaked is 7.0
use 8.3 in wine
oui
I use jtool lmao
(wait oh no does wine work on AMD hackintoshes)
I’m only on Mac because it’s the only correct choice
yeah and I'm also on mac cuz app support of Windows, but still *nix
jtool is cool
I don't like windows
how much money do you have
ida 100% discount:
piracy
Didn’t work
rip
But are you sure that with fastpathsign all signature entitlements are enabled?
You have to sign with the entitlements
There’s a couple that are restricted
See TrollStore README
hackintosh 
yup
Heard he’s a scammer
Idk that’s what iCraze told me
Had to change original message
Too mean
I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing I hate diffing
yeah
Yes. I still use Wine for my Hackintoshes.
Me when CoreTrust
OkOk taskforpid is not blocked. But it doesn't work
Alfie how do I see the actual side by side diff between two matched functions from two bins in Ida
I think I’m dumb
BinDiff 
Yes I have that
But how do I actually see them side by side
Do I just open another window with the other bin or what
no arm decompilers 
does anyone know what happened to my ida?
i used the exact same version of ida to decomp both these binaries, one was 1-2 years ago and the other was recent
it happens on any binary now, not just this particular one
is the old one ^
is the new one^
Export the .BinExport for the two bins
Create a new diff in BinDiff
And see
you could also use google's thing
7.7 arm decompiler + https://github.com/x0rloser/ida_dll_shim
Enables using HexRays Decompiler v7.6 and v7.7 with IDA 8.x - GitHub - x0rloser/ida_dll_shim: Enables using HexRays Decompiler v7.6 and v7.7 with IDA 8.x
🥱
BANGER REPO
That is what we are talking about
oh lmao
frfr
old binary or old screenshot?
thats the old binary in the same ida

not crack drug
well
actually
either will do
someoen sent me 8.3.7 crack but idk if i trust it lmao
it is real
wtf is 8.3.7
yeah it's 8.3 lol
but ida doesn't sign binaries and all pro versions are watermarked so hashes are diff
no way to easily verify that doesn't have some juicy malware
so now to figure out what the actual diff is
run it in vm or gamble
No
Best you can do is find the function in IDA and copy pseudo
frfr
if u can't google for ida pro u dont deserve ida pro
all the copies are for windows
it's simply the rules
i did
all for windows
i looked on the bay as well
none for mac
yeah because the latest leaked version for other operating systems is 7.0
use 8.3 in wine
k
what os ?
😭
`- (BOOL)isSelected {
if([vibrationDefault boolForKey: @"silent-vibrate"]){
return NO;
}
else{
return YES;
}
}
-
(void)setSelected:(BOOL)selected {
if (selected) {
// Your module turned selected/on, do something
[vibrationDefault setBool:NO forKey:@"silent-vibrate"];} else {
// Your module turned unselected/off, do something
[vibrationDefault setBool:YES forKey:@"silent-vibrate"];
}[super setSelected: selected];
}`
ccsupport tweak. would this on its own handle the selection ?
because it doesn't when i test. the suite is "com.apple.springboard"
or must i use NSNotificationCenter ?
@timid furnace how do i move my 7.7 decompilers to 8.3
the crack doesnt have arm decompilers
the decompilers are in plugins
hexarm
hexarm64
smth like that
good use of snapper for the vid
gg
i wouldnt have known othwerwise
he has the funny s name tweak
Can i pirate Nexus then??
can i pirate typecentry
lmao
iCraze DRM is bulletproof
its manually obfuscated to
fuck new ida
Register dump:
rip:00006ffffd026be0 rsp:000000000041f958 rbp:0000000000000000 eflags:00010246 ( R- -- I Z- -P- )
rax:0000000002f38b20 rbx:000000000041fd70 rcx:0000000000000000 rdx:00000000fbffffff
rsi:0000000000000000 rdi:0000000002f38b20 r8:00000000bfffffff r9:0000000000ac01e0 r10:000000000000000b
r11:0000000000000080 r12:0000000000000000 r13:0000000000000000 r14:0000000000000000 r15:0000000000000000
Stack dump:
0x0000000041f958: 00000001401218fe 000000000041fd70
0x0000000041f968: 0000000000000000 0000000000000080
0x0000000041f978: 000000000041faa0 0000000000000000
0x0000000041f988: 0000000000000000 0000000000000000
0x0000000041f998: 000000014011ff38 00000000ffffffff
0x0000000041f9a8: 00000000ffffffff 000000000041faa0
0x0000000041f9b8: 0000000000000000 000000000156a6b0
0x0000000041f9c8: 0000000000000001 0000000002f94b70
0x0000000041f9d8: 0000000000000007 0000000000000007
0x0000000041f9e8: 00006ffffd3a6928 0000000140297da8
0x0000000041f9f8: 0000000000000000 0000000140297d38
0x0000000041fa08: 0000000000000100 0000000000000000
Backtrace:
=>0 0x006ffffd026be0 in qt5core (+0x1c6be0) (0000000000000000)

wine why do u do dis to me
Nice
i made an app for writing to the framebuffer in iOS but its ppl-protected
i think
either that or i suck at coding
You can do it with CoreTrust bug
Serotonin does it
PPL can't exist on MacOS afaik
Verbose boot
It exists
wait really
Yes
yah
ohhh wait I'm thinking of KTRR
i did it in ios 15 lol
Apple Platform Security Guide is wrong
Also exists
oh
oh wait
oh so appl lied to me
They changed framebuffer a lot in 16
wait until you realise you can’t create an IOSurface for the framebuffer on A15+
SPTM
Display CoProcessor
I was manually writting the bytes to the vinfo struct
LOL
tf you mean LOL
ayyy I got it to work
i guess the alternative to controlling the screen is stripping down the default environment and make room for my own stuff
i hate iOS 17.0 i hate iOS 17.0 i hate iOS 17.0 i hate iOS 17.0 i hate iOS 17.0 i hate iOS 17.0 i hate iOS 17.0 i hate iOS 17.0
Yea just seen it
Not that important
Me when i had [REDACTED] but it wouldn’t work
I love [REDACTED]
shoo
Oh dear
me when apple changed like 2 cells in the settings app so it broke my tweak
go out of development
oh no
only TWO
???????
nexus
IDA is on crack
yes !
a14 also has dcp
and you're able to create iosurfaces on those
also for a15, it's what's used for stable krw isn't it
Yeah but an IOSurface for the framebuffer idk
hum, idk
if anything, something broke with newer versions
I couldn't create IOSurfaces on iOS 8 and 9 either, but could just fine on 10 and above (same device and all)
spaces, dock, mission control, all have to go bye bye (though i should be able to disable most of this via hooking the dock itself)
I go bye bye
so real
fr
Can someone help plz. Cant find anything about how to use ccsupport
L
do you hardcode every cell????
what
why else would apple adding more cells break your tweak other than if you hardcoded the cell ids lol
because
those cells have been there
since
like
ios 13
but
apple only changed those two cells t use a different kind of cell
idk
its stupid
Get all your photos
All ic drive shit
no its not that
?
its that i cant drop another £1200 on a windows machine and another £1000 on a samsung phone
maybe when i turn 16 and get a job lol
Could this method although old work on iOS 16?
https://newosxbook.com/articles/PST2.html
capt likes men
correct
@restive ether you buying?
Debugserver with trollstore cannot be used?
i’m interested
Is this still available?
You COMPLETELY miss the point of #development . #development is not “ @slim bramble @acoustic imp hey how do I do this .” #development is not HD’s intellectual discussion. #development is not, StackOverflow or Theos server. #development is a place for people to be monsters – the horrible, senseless, uncaring monsters that they really are.
Capt types a message and we laugh. Nightwind loses his sanity we laugh.
Developer leaves we laugh, and ask for more. Weird code, swift, capt – we laugh. 1984, assembly, OOP, *phobia, and baseless hate – we laugh. We are mindless “me-too”ism; we are irrational preference; we are pointless flamewars; we are the true face of the r/jailbreak.
?
yappin
You COMPLETELY miss the point of #development . #development is not “ @slim bramble @acoustic imp hey how do I do this .” #development is not HD’s intellectual discussion. #development is not, StackOverflow or Theos server. #development is a place for people to be monsters – the horrible, senseless, uncaring monsters that they really are.
Capt types a message and we laugh. Nightwind loses his sanity we laugh.
Developer leaves we laugh, and ask for more. Weird code, swift, capt – we laugh. 1984, assembly, OOP, *phobia, and baseless hate – we laugh. We are mindless “me-too”ism; we are irrational preference; we are pointless flamewars; we are the true face of the r/jailbreak.
You gotta see what I'm adding
"iboot patcher in python isn't real, it can't hurt you"
iboot patcher in python: https://github.com/m1stadev/eyepatch/blob/master/eyepatch/iboot/iboot64.py
eta lua eyepatcher
dn
unironically that'd be way more annoying to do
are there even be lua bindings for capstone
yes
ok nvm prob not hard then
i actually have no idea but i’m assuming someone’s wasted their time doing it
actually huge trolling: fuck capstone bindings, just require an internet connection and export disassembly to armconverter's API
that'd be so godawful slow
those lua bindings are old as shit they probably don't even work with 5.x
java :
aren't there java bindings in upstream
????
their website is outdated af
very
lets update it more
their docs are also mid
need to get the ball rolling on py apple
mf i've already been pulling a sisyphus
cuz at this point I really don't want to use limd anymore
well I don't really know py all too well and I'm taking a py class this simester so that may help
sure
i wanna talk to doronz sometime about maybe splitting up pymd3 into a few projects bc it's really starting to get bloated
tss can use mine 
just over 300 lines of code
super lightweight
ima move servers
@next wadi you care if i archive applereleases repo or do u want me to just transfer it to you
how many people are on that dev team anyway, they can't seem to keep up
also:
- ci is there but doesn't have all the targets
- packages maintained by other people and are out of date
literally just niklas iirc
wtf
stfu
v8
what
what is ARMv9
@radiant idol what do you think apple is gonna call ARMv9-A when they eventually switch to it 
arm64 OMEGA
arm64p
arm65e
yeah but it's still arm64
not arm65
also the next arch
would be arm128
because it's always multiplied by 2
Exactly that
I doubt it
arm64x
maybe
arm64f
Whos the orange name who has diabetes
you
@pine holly i might have diabetes
idrm
Whatever u wanna do big dog
whats this
python futurerestore
What's the difference between core, processor,cluster and CPU?
I see
so a cluster of apple cores
@faint timber you know what the "ok2pwrdwn" register does, right
exactly what the name suggests
yes but what does that actually, like do
not needs really though only boot cpu has the mask to prevent power down
completely powers off the core
ah
like I'm trying to find where this is actually used ARM64_REG_ACC_OVRD_disL2Flush4AccSlp
but it's only used in the deep sleep routine it seems
but would that mean I can only use it in that state
doesn't matter if its used or not or does it?
I guess not
but I'm not entirely sure if it'd do what I want it to do
maybe I should ask one of the Asahi devs
wym what you want it to do
well
what I want it to do is disable L2 cache flushing
I know it says it does that
mhm
I just don't want to put time into doing anything with it before I know if it'd work lmao
true
except
I'm not entirely sure how lmao
cuz like I can't find any documentation on this
yes
yeah but how do I read from the msr in userspace lmao
golb
golb
you can write to mmio right?
does golb work on arm64e
yes
then you don't need the other golb stuff
just the exec which can be done with coresight mmio
if it doesn't work so be it
but can't hurt to try
see the thing is though


