bool is_rootless()
{
static bool gorn;
if (!gorn) {
// check
gorn = true;
}
}```#development
primal perch
primal perch
and is guaranteed to be 0
or .data one of the two
bool is_rootless()
{
static bool gorn;
static bool actual;
if (!gorn) {
// check
actual = access(...);
gorn = true;
}
return actual;
}```there
pre configured for the skid
indigo peak
bool is_rootless() {
static bool didRun;
static bool ret;
if (!didRun) {
ret = access("/var/jb/.procursus_strapped", F_OK) == 0;
didRun = true;
}
return ret;
}
set did run at the end and yes
indigo peak
die
primal perch
gotem
indigo peak
// before
if (access("/usr/lib/libhooker.dylib", F_OK) == 0)
// after
if (access(is_rootless() ? "/usr/lib/libhooker.dylib" : "/var/jb/usr/lib/libhooker.dylib", F_OK) == 0)
would it be like that
indigo peak
primal perch
ok
i would also write wrapper functions
get_lib_directory() + "libhooker.dylib" for example
or get_file_from_lib(...)
some bullshit like that
unkempt raft
@tepid olive nice
primal perch
std::string supports that perfectly 
indigo peak
access is C
primal perch
what
unkempt raft
```C
bool is_rootless() {
static bool didRun;
static bool ret;
if (!didRun...
this is the worst possible way to detect rootless
indigo peak
yeah
unkempt raft
test if you can delete /
primal perch
indigo peak
just update it to search through every file in teh system
and get the filepath from name
make a NSDictionary
primal perch
thats some captware
dont use a '
youre not talking about possession
learn ENGLISH
this is real
AEAssessmentIndividualConfiguration isEqual:
primal perch
indigo peak
@primal perch ```C
const char *new_path(const char *path) {
// check if rootless
if (access("/var/jb/.procursus_strapped", F_OK) == 0) {
// append /var/jb/ to path
NSString *new_path = [NSString stringWithFormat:@"/var/jb/%s", path];
return [new_path UTF8String];
}
return path;
}
something like that?
primal perch
<@726214891970625549> ```C
const char *new_path(const char *path) {
// check...
use is_rootless in this itself
indigo peak
use is_rootless in this itself
bool is_rootless() {
static bool didRun;
static bool ret;
if (!didRun) {
ret = access("/var/jb/.procursus_strapped", F_OK) == 0;
didRun = true;
}
return ret;
}
const char *new_path(const char *path) {
// check if rootless
if (is_rootless()) {
// append /var/jb/ to path
NSString *new_path = [NSString stringWithFormat:@"/var/jb/%s", path];
return [new_path UTF8String];
}
return path;
}
primal perch
```C
bool is_rootless() {
static bool didRun;
static bool ret;
if (!...
ya
im strapped daddy
based hayden
as always
primal perch
oh now you change your mind
cuz daddy hayden said so
capt is a bottom
hes a gamma male
gorn
unkempt raft
```C
bool is_rootless() {
static bool didRun;
static bool ret;
if (!...
you won't sleep well you know that right
indigo peak
yeah ill just wait until @lime pivot fixes it
kirb merge my theos PR challenge
failed
unkempt raft
indigo peak
thats valid
primal perch
MEOW MEOW MEOW @worn cradle
worn cradle
primal perch
real ios framework
worn cradle
becoming a jb dev now
primal perch
fr
unkempt raft
primal perch
unkempt raft
smh it does actually useful stuff
@indigo peak how does the valid cmd+c cmd+v going? 
primal perch
unkempt raft
the io kittens were a trap
primal perch
COCK
unkempt raft
Anybody knows how does iSH keep itself running an background and not get terminated? I remember reading somewhere when iSH just came out it was constantly requesting geo data from ios, but I'm not sure where to find that code and if it still works
lime pivot
Anybody knows how does iSH keep itself running an background and not get termina...
CLLocationManager is where you'd want to start looking
you can request coarse location, and never un-request it, so effectively you're asking for continuous background notifications forever
coarse won't drain battery because it's just triangulating based on cell towers and wifi SSIDs near the phone
well, it still has to go and send that data to Apple if it detects a significant change
unkempt raft
coarse won't drain battery because it's just triangulating based on cell towers ...
that's exactly what I was concerned about
thanks
lime pivot
but it's negligible when you're already asking for the CPU and radios to be kept alive I suppose
indigo peak
but it's negligible when you're already asking for the CPU and radios to be kept...
ok so do i pr cephei rootless
or will you do it
mr upside down
indigo peak
sick
@naive kraken would i be able to use you NSString* getRootPath(void) from CCSupport to update libCephei to work rootless
@lime pivot can i use a function from CCSupport of opa334 gives permission
lime pivot
should be fine
indigo peak
since its done by an actual dev the function will be actually good
and not shit
yk
hasty ruin
real
indigo peak
as always opa beat you to the punch
why is that so true
hasty ruin
bet
primal perch
indigo peak
should be fine
should i just make a header file called getRootPath
getRootPath.h & getRootPath.m
hasty ruin
No that’s bloat and will slow down compilation by multiple milliseconds
primal perch
only funny when shep says it
real
me downloading a 14GB file in 2 mins
thanks gigabit
hasty ruin
is it gorn by any chance
hasty ruin
🧢
lime pivot
should i just make a header file called `getRootPath`
let's do Rootless.h/m
indigo peak
bet
primal perch
cumless
indigo peak
@lime pivot
/home/fiore/theos/vendor/include/LightMessaging/LightMessaging.h:25:2: error: definition of configuration macro 'ROCKETBOOTSTRAP_LOAD_DYNAMIC' has no effect on the import of 'RocketBootstrap'; pass '-DROCKETBOOTSTRAP_LOAD_DYNAMIC=...' on the command line to configure the module [-Werror,-Wconfig-macros]
#include "../rocketbootstrap/rocketbootstrap.h"
trying to build cephei
lime pivot
<@239564780271173642>
```
/home/fiore/theos/vendor/include/LightMessaging/Light...
oh yeah I guess that's new
yeah so that define will need to be moved to the makefile
in ADDITIONAL_CFLAGS
then remove the #define line wherever it is
indigo peak
in ADDITIONAL_CFLAGS
ADDITIONAL_CFLAGS += ROCKETBOOTSTRAP_LOAD_DYNAMIC=1?
lime pivot
`ADDITIONAL_CFLAGS += ROCKETBOOTSTRAP_LOAD_DYNAMIC=1`?
add -D on the start of that
don't think it needs a value actually
just define it without =1
indigo peak
"___isOSVersionAtLeast", referenced from:```pins
unkempt raft
writing asm today
all I can say is good luck
indigo peak
i have no idea why this shit isnt working
// Rootless.m
#include "Rootless.h"
NSString* getRootPath(void) {
// blah
}
// Rootless.h
NSString* getRootPath(void);
// prefs/SomeController.m
#import "../Rootless.h"
getRootPath();
Undefined symbols for architecture armv7:
"_getRootPath", referenced from:
like
it should work
indigo peak
it works
lied again
it doesnt work
bro
im gonna end my life
this is so stupid
why wont it work
it should be tho
ocean raptor
Ew
@lime pivot what's your plan for building iphoneos-arm64 debs in theos?
Do you know how you're going to do it, if at all?
lime pivot
Do you know how you're going to do it, if at all?
honestly, not really…
ocean raptor
Is the Architecture field set by theos or by the control file at the root?
lime pivot
Is the Architecture field set by theos or by the control file at the root?
control
I'm super undecided on how support for side-by-side arm and arm64 package building should be architected
(if it should even be a Theos feature at all)
because all the templates will have architecture iphoneos-arm in their control, unless you're building for 15+, in which case the control will be a lie and it's actually iphoneos-arm64
naive kraken
i have no idea why this shit isnt working
make it extern in the header maybe? idk, try putting wrong code into the getRootPath func and check if there is a compiler error, if not then the file isn't being compiled
tepid olive
what
vivid dew
i eat pussy like a starving animal
naive kraken
that won't work on linux
only on macOS and latest iOS
or you need to add like compiler-rt or something like that
indigo peak
.
indigo peak
that won't work on linux
i put the pinned code in a header file but now it says duplicated symbol or something
lime pivot
should stick it in a .m
otherwise it'll indeed be duplicating it for each time it's imported
hasty ruin
.
indigo peak
otherwise it'll indeed be duplicating it for each time it's imported
makes sense
i just dont know which one to put it in
lime pivot
.
restive ether
real
lime pivot
at this rate never
indigo peak
@lime pivot ```SH
==> Linking tool hbprefs (armv7)…
ld: library not found for -lcrt1.3.1.o
clang-10: error: linker command failed with exit code 1 (use -v to see invocation)
make[4]: *** [/home/fiore/theos/makefiles/instance/tool.mk:20: /home/fiore/libcephei/.theos/obj/debug/armv7/hbprefs] Error 1
make[3]: *** [/home/fiore/theos/makefiles/instance/tool.mk:20: /home/fiore/libcephei/.theos/obj/debug/armv7/hbprefs] Error 2
make[3]: *** Waiting for unfinished jobs....
==> Linking tool hbprefs (arm64)…
Undefined symbols for architecture arm64:
"start", referenced from:
implicit entry/start for main executable
ld: symbol(s) not found for architecture arm64
can't even get Cydia 2 in the loader
lime pivot
<@239564780271173642> ```SH
==> Linking tool hbprefs (armv7)…
ld: library not fo...
ugh not again
you'll have to just edit the makefiles to bump the TARGETs to something way higher
it's my fault for making it run on old ass iOSes
hasty ruin
Wen eta saily in loader.app 😭
After windows build
ocean raptor
it's my fault for making it run on old ass iOSes
Bro really said iOS 5
lime pivot
Bro really said iOS 5
bro really did, sadly
indigo peak
make it extern in the header maybe? idk, try putting wrong code into the getRoot...
i tried both ,and nothing changed
like it errored
lime pivot
I mean hey, rootless is my chance
indigo peak
and extern changed nothing
like its being compiled & erroring
whatever
made something up
somehow worked
lmao
restive ether
I mean hey, rootless is my chance
ALL THE BREAKING CHANGES TO MY HEARTS CONTENT
ocean raptor
Rootless is really our chance to fix everything we hate about the current setup. But no! Xina wants to use saily
faint stag
Rootless is really our chance to fix everything we hate about the current setup....
indeed, we should've had rootless years ago
indigo peak
@lime pivot id also need to change the shell scripts to account for rootless
ex: ln -s /usr/lib/Cephei
naive kraken
Rootless is really our chance to fix everything we hate about the current setup....
xina prefers Saily because the dev speaks chinese, xina has no problem with supporting other package managers though
ocean raptor
xina prefers Saily because the dev speaks chinese, xina has no problem with supp...
I know
I've been talking to him
naive kraken
ah
restive ether
xina prefers Saily because the dev speaks chinese, xina has no problem with supp...
support other package managers that will break your phone becuase they're made by people that never read docs 💪🏼💪🏼
hasty ruin
https://youtube.com/shorts/I_gFzWgC1Ss found @zenith hatch’s channel
zenith hatch
???
hollow scroll
Hey everyone, i have been a bit out of the loop for the past few months.
I am starting to get emails about some of my tweaks from people that use the new iOS 15 jailbreak.
they report that the preferences are not saving correctly. (when they change a value and go back to settings it wont save)
any ideas on what have changed on this new jailbreak? perhaps the location of saved preferences plist files?
naive kraken
Hey everyone, i have been a bit out of the loop for the past few months.
I am st...
where is your hook that's reading the preferences? maybe in a system app?
hollow scroll
where is your hook that's reading the preferences? maybe in a system app?
I am reading the preferences on the Springboard, but it sounds like the preferences file is not saved correctly so that's not the issue
One of the tweaks with this issue is open sourced: https://github.com/gilshahar7/ExactTime
naive kraken
I am reading the preferences on the Springboard, but it sounds like the preferen...
well I'm unsure about the specifics but directly reading a plist has always been discouraged
Ah
I know the issue
hollow scroll
I suspect that the issue is with the readPreferenceValue and setPreferenceValue on this file
https://github.com/gilshahar7/ExactTime/blob/master/exacttimeprefs/ExactTimeprefs.mm
naive kraken
In your pref bundle you write to /User/Library
/User has never been guaranteed to exist
And will never exist on iOS 15 due to SSV
So change that to /var/mobile/Library
hollow scroll
Awesome, that sounds reasonable. Thank you 🙂
tepid olive
https://youtube.com/shorts/I_gFzWgC1Ss found <@381538809180848128>’s channel
why isn't it using recursion
@grave sparrow tell me whats wrong now
i did more stuff
tardy narwhal
if that patch actually works it's fine for the first time writing raw assembly, adhering to conventions & proper assembly code-style should be secondary and the next stage in development
(imo)
ah, then rework is needed but I wouldn't necessarily push for the hard reset. correcting the mistakes as the code is now would probably be more fruitful, i'm not too deep into arm64 asm
the x64 bits are ... well they could use some love too
tepid olive
No
I am gonna clean this up after, I know I can’t use those registers
I don’t see why making a string with bytes is bad though?
Read the code
There are hardcoded addresses
Because I am not finished
I know
Stop thinking I’m this stupid
Lmao
The dyld slide doesn’t change so I hardcode it to accelerate development
Not in my experience
No
Yes
The code is right there
I still don’t see what’s wrong dude
Except the calling conv
Which I will fix
It does set the env correctly
Unused
It was used to test string alloc
No
It executes the first instructions before jumping back
Which isn’t a problem
The lr thing is temporary as well
This is a instruction of posix spawn
I can’t omit it
It has to be ran before jumping back
Otherwise it’s skipping 5 instructions
I am 100% sure that part is correct
I do
Dude
Please
What’s wrong with strings made with ascii bytes
Why?
How do you think ascii works
😭
No because that would require doing cursed disassembly to make the compiler work as intended
?
Ok
modern shuttle
Wow your knowledge is awesome
tepid olive
this is unnecessary pain
about this
i strongly doubt launchd has a env string set already
like it surely doesn’t
tepid olive
like it surely doesn’t
if it did then that would be after a userspace reboot, and it would be set to something the right path already
capt everything you’re talking about assumes that there is undefined behaviour but this is only one process that runs the same everywhere
tepid olive
How do i convert epoch time to yyyy:mm:dd hh:mm:ss
there is probably a github gist somewhere
with all the code
wheat grotto
I tried using the Time.h Library that the arduino ide should have by default, but "DateTime" was not declared in this scope
Copium
or it might not be updated
if so
im f dead
Poggers
neon jungle
Do I need to fix this "ld: warning: object file /home/me/work/sidestatus/.theos/obj/debug/arm64e/Tweak.xm.a6b474a9.o was built with an incompatible arm64e ABI compiler" if I want my tweak to work on arm64e?
wheat grotto
ig it doesnt support the esp8266
Tho im eventually moving to a uno with an eth shield, so maybe thats gonna fix it...?
tepid olive
You’re telling me that I’m supposed to consider that a user will debug launchd?
ocean raptor
@grave sparrow I need to intercept xpc messages on iOS, how do I do that?
tepid olive
PAC sucks
tepid olive
<@434097922762735618> I need to intercept xpc messages on iOS, how do I do that?
xpc spy
ocean raptor
frida-ps -U rebooted my device 💀
restive ether
`frida-ps -U` rebooted my device 💀
what jailbreak are you using
cause I know Frida just doesn't work on any coolstar jailbreak
ocean raptor
checkra1n
ocean raptor
cause I know Frida just doesn't work on any coolstar jailbreak
I got xpcspy working
restive ether
primal perch
the discord admin educates the advanced developer
hasty ruin
cause I know Frida just doesn't work on any coolstar jailbreak
it doesnt work on u0 either for me anymore
frida ios really fell off
restive ether
I know why it doesn't work on star's
but it was literally built for u0 more than anything so
ocean raptor
http://newosxbook.com/tools/XPoCe2.html
x86_64 only
tepid olive
@vivid dew hiii pac nerd can you help
how do I sign a pointer to a string on arm64e with raw instructions
restive ether
schizo
restive ether
Justin Trudeau
vivid dew
restive ether
vivid dew
vivid dew
IDK
primal perch
KDI
tepid olive
I am so tired of PAC why does it exist
primal perch
security
vivid dew
trolling purposes
vivid dew
just assume the pointer was never signed in the first place
primal perch
i mean is it though
vivid dew
it usually isn't
primal perch
its making whatever ur tryna do extra difficult
vivid dew
vivid dew
primal perch
capt inc writing a whole essay for one post
vivid dew
uhhh, you can start by hooking xpc_pipe_routine() and print the 2nd argument, which is the message to be sent
i’m not really sure about intercepting messages that are received, but you can start by hooking xpc_pipe_receive() and logging that? potentially?
extern int xpc_pipe_simpleroutine(xpc_object_t pipe, xpc_object_t message);
#if __has_feature(objc_arc)
extern int xpc_pipe_routine(xpc_object_t pipe, xpc_object_t message, __strong xpc_object_t *reply);
#else
extern int xpc_pipe_routine(xpc_object_t pipe, xpc_object_t message, xpc_object_t *reply);
#endif
#if __has_feature(objc_arc)
extern int xpc_pipe_receive(mach_port_t port, __strong xpc_object_t *message);
#else
extern int xpc_pipe_receive(mach_port_t port, xpc_object_t *message);
#endif
xpc_pipe_routine sends a message and waits for a reply synchronously. xpc_pipe_simpleroutine is similar, but does not expect a reply & returns immediately
note that xpc_pipe_routine() calls an underlying function: _xpc_pipe_routine() (prefixed underscore).
you may find better luck hooking the underlying function instead, but who knows. note that the underlying function has the message as the 3rd arg instead of the 2nd arg
primal perch
uhhh, you can start by hooking xpc_pipe_routine() and print the 2nd argument, which is the message to be sent
i’m not really sure about intercepting messages that are received, but you can start by hooking xpc_pipe_receive() and logging that? potentially?
extern int xpc_pipe_simpleroutine(xpc_object_t pipe, xpc_object_t message);
#if has_feature(objc_arc)
extern int xpc_pipe_routine(xpc_object_t pipe, xpc_object_t message, strong xpc_object_t reply);
#else
extern int xpc_pipe_routine(xpc_object_t pipe, xpc_object_t message, xpc_object_treply);
#endif
#if has_feature(objc_arc)
extern int xpc_pipe_receive(mach_port_t port, strong xpc_object_t message);
#else
extern int xpc_pipe_receive(mach_port_t port, xpc_object_tmessage);
#endif
xpc_pipe_routine sends a message and waits for a reply synchronously. xpc_pipe_simpleroutine is similar, but does not expect a reply & returns immediately
note that xpc_pipe_routine() calls an underlying function: _xpc_pipe_routine() (prefixed underscore).
you may find better luck hooking the underlying function instead, but who knows. note that the underlying function has the message as the 3rd arg instead of the 2nd arg
vivid dew
ocean raptor
#include <stdio.h>
#include <stdlib.h>
#include <xpc/xpc.h>
typedef xpc_object_t xpc_pipe_t;
#include "fishhook.h"
kern_return_t
xpc_pipe_routine(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply);
kern_return_t (*old_xpc_pipe_routine)(xpc_pipe_t, xpc_object_t, xpc_object_t*);
kern_return_t
hook_xpc_pipe_routine(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply)
{
kern_return_t ret = old_xpc_pipe_routine(pipe, request, reply);
char *requeststr = xpc_copy_description(request);
fprintf(stderr, "REQUEST: %s\n", requeststr);
free(requeststr);
char *replystr = xpc_copy_description(*reply);
fprintf(stderr, "REPLY: %s\n", replystr);
free(replystr);
return ret;
}
kern_return_t
xpc_pipe_routine_with_flags(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply, uint32_t flags);
kern_return_t (*old_xpc_pipe_routine_with_flags)(xpc_pipe_t, xpc_object_t, xpc_object_t*, uint32_t);
kern_return_t
hook_xpc_pipe_routine_with_flags(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply, uint32_t flags)
{
kern_return_t ret = old_xpc_pipe_routine_with_flags(pipe, request, reply, flags);
char *requeststr = xpc_copy_description(request);
fprintf(stderr, "REQUEST: %s\n", requeststr);
free(requeststr);
char *replystr = xpc_copy_description(*reply);
fprintf(stderr, "REPLY: %s\n", replystr);
free(replystr);
printf("FLAGS: %i\n", flags);
return ret;
}
__attribute__((constructor))
static void ctor(void) {
struct rebinding rebindings[] = {
{"xpc_pipe_routine_with_flags", hook_xpc_pipe_routine_with_flags, &old_xpc_pipe_routine_with_flags},
{"xpc_pipe_routine", hook_xpc_pipe_routine, &old_xpc_pipe_routine}
};
rebind_symbols(rebindings, 2);
fprintf(stderr, "Hooked!\n");
}
I figured it out without you
I couldn't get xpcspy to work
primal perch
did apple start symbolicating ios 16 kcaches or something
cuz the one i just dropped in has almost all the symbols as the ios 14 one 
vivid dew
rare based move from apple
tepid olive
just assume the pointer was never signed in the first place
it crashes with authentication failure
tepid olive
```c
#include <stdio.h>
#include <stdlib.h>
#include <xpc/xpc.h>
typedef xpc_obj...
fishhook 
vivid dew
ok now disassemble the function where it crashes and find what instruction is being used to auth the pointer
tepid olive
it crashes with authentication failure
it shouldn’t actually be crashing though
you can pass a pointer from arm64 processes and it works
ocean raptor
fishhook <:applecry:889699912453873664>
what
ocean raptor
I don't have any hooking lib installed
tepid olive
But it crashes somewhere else
vivid dew
trolling
tepid olive
KERN_INVALID_ADDRESS at 0xf1060001000b0000 -> 0x00000001000b0000 (possible pointer authentication failure)
0x00000001000b0000 is the posix_spawn pointer
What does it mean
vivid dew
you passed a signed pointer where you shouldnt have
vivid dew
fr
vivid dew
is the pointer signed or not
tepid olive
can mach_vm_address_t be signed
vivid dew
tepid olive
as in, the address where i've allocated memory
vivid dew
no
tepid olive
so basically i'm jumping to a executable page that i just allocated
should i be signing with the asia key
tepid olive
yes
vivid dew
if you pass it out to something following normal abi then you need to sign it
vivid dew
you need uhh
tepid olive
When I ret in the function where I jumped
vivid dew
wtf
vivid dew
where is it crashing now
tacit stump
faint stag
haven't seen capt like this since the beta alert
hasty ruin
xina crashes when anything is reading syslog
1.1.3.6 fixed that for me
indigo peak
you buying me a mac
hasty ruin
lime pivot
support other package managers that will break your phone becuase they're made b...
while poking over the cursed changes I reminded myself again that it gladly deletes the dpkg lockfiles
restive ether
and using force options in dpkg!
tepid olive
the ret makes it go back
restive ether
and they don't know how a dist repo works!
lime pivot
tepid olive
since it’s blr
lime pivot
Removing the dpkg lock files is never a correct solution
restive ether
https://wiki.debian.org/Teams/Dpkg/FAQ#Q:_What_can_be_done_when_the_dpkg_lock_is...
most of the time I've seen people get this is when they have two apt instances running
and newer versions of apt don't lock nearly as much regardless
lime pivot
which, turns out, isn’t supported, and shouldn’t be overridden to make it work!
it’s almost as if the feature exists for a reason, or something like that
indigo peak
@lime pivot im not going to try and get cephei on ios 15 working, that requires a lot of work idk how to do
lime pivot
if you wanna just PR that one thing that’s cool with me
or whatever you got working so far
lime pivot
lmao sounds good to me
restive ether
can we just agree to keep preferences in the /var/jb path instead of whatever is going on now
no /var/mobile whatever please 😭
primal perch
based
naive kraken
can we just agree to keep preferences in the /var/jb path instead of whatever is...
tell that to cfprefsd, lol?
restive ether
ill tell cfprefsd to get fucked
naive kraken
actually making a cfprefsd hook to save all prefs with a non apple prefix in /var/jb would be very easy
but then again hooking cfprefsd is broken on xina 🙂
also /var/jb/var points back to /var
in an ideal world we'd store prefs in /var/jb/var/mobile/Library/Preferences
restive ether
idk what Xina is doing, I'm not quite sure he does either
restive ether
in an ideal world we'd store prefs in /var/jb/var/mobile/Library/Preferences
yeah everything should just be whatever the old path was + jb prefix
naive kraken
I mean, he's looking into fixing cfprefsd now just so I can get Crane working so that's something
tepid olive
isn’t ret just mov pc lr
faint timber
indigo peak
@primal perch whats broken w birdpoop rn
tepid olive
okay
i just don’t get it anymore
there must be a stupid mistake somewhere
the lr triggers a pointer authentication failure
How would I sign my lr
i guess i use the discriminator
no
i mean right now i’m tryna do the direct jump
unsigned address
but it crashes
posix spawn + 20
skipping the first jump to the replacement
it never reached it though, the branch itself crashes the process
the one that jumps back to posix spawn
i am doing br x16 now
that’s the thing it just won’t go back
Okay wait
If I compile for arm64, posix spawn outputs bad address as an error
but it shows that something is wrong with envp
Ok it is fixed
Idk how the fuck
But it is
that's what i did
almost works
Pushed the current thing
Going to bed now
Too bad i guess
I’ll just do malloc memcpy
that’s work for later
primal perch
<@726214891970625549> whats broken w birdpoop rn
the offsets are off for like 1/2 of them
i updated them to what i thought was correct but tested and crashed
didnt test f urther
restive ether
neon jungle
Nvm I figured it out, my tweak couldn’t find the preference bundle and it causes respring loop on xina
The other error idk
ocean raptor
fishhook <:applecry:889699912453873664>
I removed fishhook
#include <stdio.h>
#include <stdlib.h>
#include <xpc/xpc.h>
typedef xpc_object_t xpc_pipe_t;
#define DYLD_INTERPOSE(_replacement,_replacee) \
__attribute__((used)) static struct{ const void* replacement; const void* replacee; } _interpose_##_replacee \
__attribute__ ((section ("__DATA,__interpose"))) = { (const void*)(unsigned long)&_replacement, (const void*)(unsigned long)&_replacee };
kern_return_t
xpc_pipe_routine(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply);
kern_return_t
hook_xpc_pipe_routine(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply)
{
kern_return_t ret = xpc_pipe_routine(pipe, request, reply);
char *requeststr = xpc_copy_description(request);
fprintf(stderr, "\033[32mREQUEST: %s\033[m\n", requeststr);
free(requeststr);
char *replystr = xpc_copy_description(*reply);
fprintf(stderr, "\033[31mREPLY: %s\033[m\n", replystr);
free(replystr);
return ret;
}
DYLD_INTERPOSE(hook_xpc_pipe_routine, xpc_pipe_routine);
#ifdef WITH_FLAGS
kern_return_t
xpc_pipe_routine_with_flags(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply, uint32_t flags);
kern_return_t
hook_xpc_pipe_routine_with_flags(xpc_pipe_t pipe, xpc_object_t request,
xpc_object_t *reply, uint32_t flags)
{
kern_return_t ret = xpc_pipe_routine_with_flags(pipe, request, reply, flags);
char *requeststr = xpc_copy_description(request);
fprintf(stderr, "\033[32mREQUEST: %s\033[m\n", requeststr);
free(requeststr);
char *replystr = xpc_copy_description(*reply);
fprintf(stderr, "\033[31mREPLY: %s\033[m\n", replystr);
free(replystr);
fprintf(stderr, "\033[32mFLAGS: %i\033[m\n", flags);
return ret;
}
DYLD_INTERPOSE(hook_xpc_pipe_routine_with_flags, xpc_pipe_routine_with_flags);
#endif
(the _with_flags function is only on iOS 13+ and I needed to run it on iOS 12)
gentle grove
Beat this
Same
Maybe github accidentally sent 4 emails to everyone
same content in all 4
untold drift
is mshookfunction working with xina jailbreak ?
indigo peak
is mshookfunction working with xina jailbreak ?
No, you need to use LibHooker functions
untold drift
No, you need to use LibHooker functions
any documentation or example for it ?
indigo peak
I can give you an example
and link the docs
just give me a few
I just got out of bed
naive kraken
xina has both substitute and libhooker for whatever reason
indigo peak
Really?
timid furnace
naive kraken
so when you use MSHookFunction it calls to substitute and LHHookFunction goes to libhooker
indigo peak
I didnt get mshookmemory to work
naive kraken
Yeah substitute is half broken
timid furnace
no embed, sad
naive kraken
Xina should just drop it and include a substrate libhooker shim
that would be much better
untold drift
i tried to hook a mangled name class (from ida) but not working with %hookf
apps crashing when launch
elder scaffold
i need rootless injector for checkm8 devices
my arm64e is 15.4.1 so not ready yet
indigo peak
any documentation or example for it ?
#import <libhooker/libhooker.h>
struct LHMemoryPatch patch;
patch.destination = offset;
patch.data = new_memory;
patch.size = sizeof(new_memory);
patch.options = NULL;
LHPatchMemory(&patch, 1);
@naive kraken is this the "best" way to check if the device uses libhooker or should default to substitute
bool hasLibhooker(void) {
if (access("/.procursus_strapped", F_OK) == 0 || access("/var/jb/.procursus_strapped", F_OK) == 0) {
return true;
}
return false;
}
absolutely not
indigo peak
naive kraken
I just try to dlopen libhooker and dlsym LHHookFunctions
untold drift
thanks for your effort @indigo peak
naive kraken
and if that exists I use it, and if not I use MSHookFunction
(coolstar doesn't like this, but it seems to work for me so far)
indigo peak
cephei uses this
if (access("/usr/lib/libhooker.dylib", F_OK) == 0)
untold drift
whats alternative MSFindSymbol in libhooker ?
indigo peak
prob LHFindSymbols
tepid olive
Xina should just drop it and include a substrate libhooker shim
ellekit 
no symbol finder tho
naive kraken
symbol finders are largely broken anyways
tepid olive
it works otherwise
naive kraken
MSFindSymbol doesn't work correctly on libhooker for me
tepid olive
lhhookfunctions is more optimized than on libhooker
coolstar apparently doesn't even reuse pages
tepid olive
cephei uses this
`if (access("/usr/lib/libhooker.dylib", F_OK) == 0) `
why not RTLD_NOLOAD
untold drift
ellekit <:troll:794642746258096140>
wen eta ellekit xd
indigo peak
I just try to dlopen libhooker and dlsym LHHookFunctions
wouldnt just dlopen be enough
tepid olive
wen eta ellekit xd
it's out
indigo peak
wen eta ellekit xd
naive kraken
cephei uses this
`if (access("/usr/lib/libhooker.dylib", F_OK) == 0) `
problem is, you can't link both substrate and libhooker
naive kraken
So using LHHookFunctions directly will break substrate and substitute
unless you depend on like libhooker compatiblity shim
tepid olive
weak symbols exist
indigo peak
problem is, you can't link both substrate and libhooker
void HookMemory(Class class, SEL selector, uint64_t offset, uint32_t data) {
void *final_offset = (void *)[class instanceMethodForSelector:selector] + offset;
if (hasLibhooker()) {
struct LHMemoryPatch patch;
patch.destination = final_offset;
patch.data = &data;
patch.size = sizeof(data);
patch.options = NULL;
LHPatchMemory(&patch, 1);
} else {
MSHookMemory(final_offset, (const void *)&data, sizeof(data));
}
}
this seems to work
idk what youre talking about
tepid olive
```C
void HookMemory(Class class, SEL selector, uint64_t offset, uint32_t data) ...
you can also use mach apis
naive kraken
you can also use mach apis
(better to not do that)
tepid olive
why
naive kraken
never works correctly
tepid olive
that's what MS and LH do
naive kraken
and on some jailbreaks the memoryhooker functions do way more
signing pages and stuff
tepid olive
Nobody does this
naive kraken
substrate does
tepid olive
Libhooker doesn't and it's fine
indigo peak
If libhooker is not installed, the dylib won't load
i dont get it
naive kraken
only because of the jailbreak supporting it
tepid olive
Huh?
indigo peak
default to substitute
naive kraken
i dont get it
if you call LHHookFunctions directly, this means you link libhooker
naive kraken
that can test smth for me
do otool -L on the binary
no need to test anything
if libhooker is in there, it won't load if libhooker is not installed
if you want to make your dylib determine whether to use libhooker or substrate at runtime, you need to stop linking libhooker and use dlopen and dlsym to get the pointer to LHHookFunctions and then call that
but as I said, coolstar does not like this, not entirely sure why
indigo peak
@rpath/libhooker.dylib (compatibility version 0.0.0, current version 0.0.0)
naive kraken
yep
this means when your dylib loads, libhooker gets loaded too
and if libhooker doesn't exist, dyld will refuse to load your dylib
tepid olive
there u go both apis
indigo peak
oh i get it
indigo peak
and if libhooker doesn't exist, dyld will refuse to load your dylib
it makes sense now
naive kraken
why are you using lh api tho
because xina jb is unfinished and the substrate api is half broken on it
tepid olive
how
naive kraken
because instead of shipping a shim, it ships both libsubstitute and libhooker
tepid olive
what did they do
LMAO
hii
libellekit in /usr/lib
libsubstrate and libhooker symlink (in another package?) possibly?
libinjector in /usr/lib as well
it can't run completely independently sadly
almost done
It implements libsubstrate and libhooker
indigo peak
because instead of shipping a shim, it ships both libsubstitute and libhooker
is the symbol for LHPatchMemory just _LHPatchMemory
naive kraken
I just doubt anyone uses it?
don't think any tweak has ever used libsubstitute directly
not even bingners own tweaks lol
tepid olive
sounds good
There's only the Swift API for that
I'm not providing a stable API atm
indigo peak
@naive kraken
bool hasLibhooker() {
static void (*_LBHookMessage)(Class, SEL, void*, void*);
if (!_LBHookMessage) {
void *handle = dlopen("/usr/lib/libsubstrate.dylib", RTLD_LAZY);
_LBHookMessage = dlsym(handle, "LBHookMessage");
}
if (_LBHookMessage) {
return true;
}
return false;
}
i have 0 idea if thats right or not
naive kraken
hm
tepid olive
now i have to do little thingies and i can put it in launchd
naive kraken
yeah looks ok
indigo peak
now i need to make it check /usr/lib and or /var/jb/usr/lib
untold drift
<:fr:792934328236638267> there u go both apis
is that work with xina jailbreak ?
naive kraken
int (*__LHHookFunctions)(const struct LHFunctionHook *hooks, int count);
int HCHookFunctions(const struct LHFunctionHook *hooks, int count)
{
static dispatch_once_t onceToken;
dispatch_once (&onceToken, ^{
void* lhImage = dlopen((rootifyCPath("/usr/lib/libhooker.dylib")), RTLD_NOW);
if(lhImage)
{
NSLog(@"picking libhooker");
// this is illegal according to coolstar but it works
__LHHookFunctions = (void*)dlsym(lhImage, "LHHookFunctions");
NSLog(@"__LHHookFunctions = %p", __LHHookFunctions);
}
else
{
NSLog(@"not picking libhooker");
}
});
// if libhooker is available, use it
if(__LHHookFunctions)
{
return __LHHookFunctions(hooks, count);
}
// otherwise, fall back to substrate
else
{
for(int i = 0; i < count; i++)
{
struct LHFunctionHook hook = hooks[i];
MSHookFunction(hook.function, hook.replacement, hook.oldptr);
}
return 0;
}
}
this is what I do
tepid olive
huh
naive kraken
NSString* getRootPath(void)
{
static NSString* rootPath = nil;
static dispatch_once_t onceToken;
dispatch_once(&onceToken, ^
{
NSFileManager* fileManager = [NSFileManager defaultManager];
NSDictionary* attributes = [fileManager attributesOfItemAtPath:@"/var/jb" error:nil];
if(attributes)
{
NSString* fileType = attributes[NSFileType];
if([fileType isEqualToString:NSFileTypeSymbolicLink])
{
NSString* destination = [fileManager destinationOfSymbolicLinkAtPath:@"/var/jb" error:nil];
if(![destination isEqualToString:@"/jb"] && ![destination isEqualToString:@"/jb/"])
{
rootPath = destination;
}
}
}
if(!rootPath)
{
rootPath = @"/";
}
});
return rootPath;
}
NSString* rootifyPath(NSString* path)
{
return [getRootPath() stringByAppendingPathComponent:path];
}
const char* rootifyCPath(const char* cPath)
{
NSString* path = [NSString stringWithUTF8String:cPath];
return rootifyPath(path).fileSystemRepresentation;
}
whats the performance diff between dispatch_once & not using it at all
naive kraken
whats the performance diff between `dispatch_once` & not using it at all
uh idk
just good practice
just copy all of that tbh
works fine (at least from what I tested)
then just call HCHookFunctions instead of MSHookFunction / LHHookFunctions
ocean raptor
Or libiosexec
tepid olive
i can do this just send what i have to do
ocean raptor
tepid olive
ok i'll use a vmapple to test
untold drift
```
int (*__LHHookFunctions)(const struct LHFunctionHook *hooks, int count);
in...
assigning to 'int (*)(const struct LHFunctionHook *, int)' from incompatible type 'void *'
naive kraken
assigning to 'int (*)(const struct LHFunctionHook *, int)' from incompatible typ...
I assume this is because you are using it from objc++
just put a cast there
tepid olive
@naive kraken Any idea why I can't attach lldb to launchd?
I get attach failed ((os/kern) invalid address)
SIP off, arm64e enabled
ocean raptor
@tepid olive you may be able to dtrace it
indigo peak
@naive kraken converted it to hook memory & not hook functions
int HookMemory(const struct LHMemoryPatch *patches, int count)
{
static dispatch_once_t onceToken;
dispatch_once (&onceToken, ^{
void* lhImage = dlopen((rootifyCPath("/usr/lib/libhooker.dylib")), RTLD_NOW);
if(lhImage) {
__LHPatchMemory = (void*)dlsym(lhImage, "LHPatchMemory");
}
});
// if libhooker is available, use it
if(__LHPatchMemory) {
return __LHPatchMemory(patches, count);
} else {
for(int i = 0; i < count; i++)
{
struct LHMemoryPatch patch = patches[i];
MSHookMemory(patch.destination, patch.data, patch.size);
}
return 0;
}
}
right ty
tepid olive
i’m so close to having this work
the patch is installed, it just does nothing atm for some reason
untold drift
is there any way to call native c++ or c function in framework with tweak?
tepid olive
is there any way to call native c++ or c function in framework with tweak?
Yeah just put the function definition in a header
untold drift
function is not in header file. i found it in ida
timid furnace
is the symbol exported
untold drift
looking like symbol bcs start with _
timid furnace
well, if you are in ida, check the exports tab
the symbol starting with _ is not a good indicator of whether it is exported or not
untold drift
function looking in export
untold drift
ios 11.4
timid furnace
that's your SDK? and not your deployment target?
timid furnace
what is your SDK
ok
now open $THEOS/sdks/iPhoneOS15.0.sdk/System/Library/(Private)Frameworks/<FrameworkName>.framework/<FrameworkName>.tbd
untold drift
the framework is target application's framework
timid furnace
oh
ok
in that case, you would want to declare the function in a header somewhere
with ```extern "C" {
// function declaration
}
if not, just extern is fine
untold drift
@interface SomeFramework:NSObject extern _someHiddenFunction(); @end
is that enough ?
timid furnace
i thought you said this was a c/c++ function
you would just leave it bare
so
// if tweak is Obj-C, and target function is C
// or if tweak is C++ and target function is C++
extern void someFunc();
// if tweak is C++ and target function is C
extern "C" void someFunc2();
then in your constructor, replace %init; with %init(someFunc = MSFindSymbol(NULL, "_someFunc"));
replace MSFindSymbol with whatever hooking API you're using
another example here https://theos.dev/docs/logos-syntax
timid furnace
that probably means the framework isn't loaded yet and as such you have a null ptr
hm
maybe you can load the framework first
indigo peak
when i use MSFindSymbol, app is crash
arent you on xina
timid furnace
this is starting to get out of my wheelhouse though
indigo peak
try using libhooker instead of substitute
untold drift
thanks for your patience, and helpful supports
untold drift
try using libhooker instead of substitute
im not know how is working libhooker 😄 you tried to teach me but its complicated for me atm
indigo peak
@untold drift why cant you use hookf
char someFunc(int arg1, bool arg2);
%hookf(char, someFunc, int arg1, bool arg2) {
char orig = %orig;
// your code
return orig;
}
tried but crashing.. i think substitute not working well with xina
restive ether
don't worry nothing works with it really
indigo peak
@naive kraken do you think a wrapper for libhooker and subtrate would be useful
like it has a check for substrate libhooker
for all functions
naive kraken
<@208334999722786816> do you think a wrapper for libhooker and subtrate would be...
upcoming xina version will probably fix this issue
besides, yeah I think this could work, but if done it should be done like the following: 2 separate packages, one that has like hookcompat.dylib that links libhooker and the other implements the same functions but links libsubstrate
and then when you have substrate or substitute it installs hookcompat (substrate) and when you have libhooker it installs hookcompat (libhooker)
this is also the only way to support all of this without using dlopen / dlsym
because you could just depend on hookcompat and link that
ocean raptor
@grave sparrow u up?
indigo peak
@naive kraken im just doing shit like this
int (*__LBHookMessage)(Class class, SEL selector, void *replacement, void *old_ptr);
int SHHookMessage(Class class, SEL selector, void *replacement, void *old_ptr) {
static dispatch_once_t onceToken;
dispatch_once(&onceToken, ^{
void *lhImage = dlopen(rootifyCString("/usr/lib/libhooker.dylib"), RTLD_LAZY);
if (lhImage) {
__LBHookMessage = dlsym(lhImage, "LBHookMessage");
}
});
if (__LBHookMessage) { // if we have libhooker, use it
return __LBHookMessage(class, selector, replacement, old_ptr);
} else { // otherwise use substitute
MSHookMessageEx(class, selector, (IMP)replacement, (IMP *)old_ptr);
return 0;
}
}
int (*__LHHookFunctions)(const struct LHFunctionHook *hooks, int count);
int SHHookFunctions(const struct LHFunctionHook *hooks, int count) {
static dispatch_once_t onceToken;
dispatch_once(&onceToken, ^{
void *lhImage = dlopen(rootifyCString("/usr/lib/libhooker.dylib"), RTLD_LAZY);
if (lhImage) {
__LHHookFunctions = dlsym(lhImage, "LHHookFunctions");
}
});
if (__LHHookFunctions) {
return __LHHookFunctions(hooks, count);
} else {
for (int i = 0; i < count; i++) {
struct LHFunctionHook hook = hooks[i];
MSHookFunction(hook.function, hook.replacement, hook.oldptr);
}
return 0;
}
}
and then just publisdh the .h/m
naive kraken
<@208334999722786816> im just doing shit like this
```C
int (*__LBHookMessage)(C...
before publishing this, maybe make sure it works on real libhooker
because I haven't done that yet
indigo peak
smart
ocean raptor
dms
indigo peak
@naive kraken it seems to work on reg libhooker
at least LHPatchMemory/MSHookMemory
clarity just tested it on taurine
naive kraken
<@208334999722786816> it seems to work on reg libhooker
nice
indigo peak
so i got HookMessage, HookFunctions and HookMemory with the rootless fix
idk what else to do
MSFindSymbol
that one
i think
indigo peak
xina's substitute is like half broken
so im making wrapper functions that detect libhooker and use that if its present
or substitute if its not present
basically making tweaks that use either one compatible with both libs
wheat grotto
Where do i edit the neofetch config on macos ? installed it via brew
neon jungle
How do I set this inside my objective C tweak?
wheat grotto
Where do i edit the neofetch config on macos ? installed it via brew
resolved
faint stag
Where do i edit the neofetch config on macos ? installed it via brew
same as linux
indigo peak
How do I set this inside my objective C tweak?
MSHookIvar<NSInteger>(object, "_orientation") = number;
// or
[self setValue:@(number), forKey:@"_orientation"];
// or
// look for a setter / getter for that and hook that getter/setter
// getter:
-(NSInteger)orientation {
return number;
}
// setter:
-(void)setOrientation:(NSInteger)arg1 {
%orig(number);
}
different ways i guess
neon jungle
```C
MSHookIvar<NSInteger>(object, "_orientation") = number;
// or
[self setVa...
Thanks Idk wtf I'm doing
indigo peak
any one of those would probably work
i said it earlier + less reading
i like helping people without actually describing what i did
gets the mind going
indigo peak
its alright, my girlfriend doesnt mind getting pinged
(its funny because shes property)
neon jungle
How do you set a property instead of a method?
indigo peak
How do you set a property instead of a method?
self.property = new_property;
neon jungle
Nope
indigo peak
gotta join that
shut stag
got it
indigo peak
why
drive?
perfect
that answers your question
get that drunk guy back in the server'
wheat grotto
im broke stfu
primal perch
raptor lake >>>
primal perch
so true
crude swallow
noble zodiac
Hello guys, I don't know why is my code returning this error...
noble zodiac
how can I fix this ? I don't care if user get nothing if error
how to throw an error object
@lime pivot
lime pivot
the first bit of this should help explain how error handling works in Swift: https://docs.swift.org/swift-book/LanguageGuide/ErrorHandling.html
well, the whole thing should be worth a read
gentle grove
lime pivot
well observed, bbaovanc#9999
noble zodiac
I just want to remove the throw thats it
not to fix it
but I don't know what to remove
lime pivot
I don't understand. comment it out then?
but the function still needs to return something, of course
either return a String or throw an Error
noble zodiac
either return a String or throw an Error
I'm returning a string
noble zodiac
return image["StatusBarCarrierName"] as! String
lime pivot
right, but the code won't reach that path 100% of the time
noble zodiac
wdym
lime pivot
hence why the throw is there, because there is a case where it can fail to find what you need
noble zodiac
yes and how can I fix that ?
lime pivot
if you comment out the throw line and try to build, you'll get an error that the function doesn't return a value on all paths
noble zodiac
if you comment out the throw line and try to build, you'll get an error that the...
yep
lime pivot
so, you need to make sure that final line of the func either returns a String, or throws an Error
hope that makes sense?
at the moment what you've got is trying to throw a String, which isn't possible
lime pivot
it's hard for me to really give you good advice since I don't know the codebase you're working with. but indeed there is no function like that by default, if you copied that from somewhere, that codebase is defining an extension on UIApplication for that
noble zodiac
Any other way to spawn an alert ?
lime pivot
since you're using SwiftUI, there's a proper way to do it with SwiftUI modifiers
noble zodiac
like ?
lime pivot
really that code using a custom alert() function is a bit of a cheat, it works but it's not the "best" way to do it
https://sarunw.com/posts/how-to-present-alert-in-swiftui-ios15/ this has more detail on displaying variables in the alert, like the error message in your case
noble zodiac
see here: https://developer.apple.com/documentation/swiftui/alert
but yeah it worked
lime pivot
probably have a syntax error in ContentView.swift
lime pivot
that's not quite how Alert() works - you need to use it as a modifier on the view, so SwiftUI knows it needs to be displayed
otherwise it's kind of just a statement that does nothing, you create an Alert() but don't do anything with it
lime pivot
https://sarunw.com/posts/how-to-present-alert-in-swiftui-ios15/ this has more de...
definitely would recommend giving this a read, it explains it pretty well
the .alert(…) bit is what you're missing, that attaches the alert to your view, so SwiftUI is aware of it
when your presentAlert state variable becomes true, SwiftUI knows it has to display the alert
noble zodiac
.alert("Hit !", isPresented: true, actions: {
// actions
}, message: {
Text("Please reboot your device. Made by C22")
})
} catch {
.alert("Title", isPresented: true, actions: {
// actions
}, message: {
Text("Message")
})
I get Value of tuple type '()' has no member 'alert'
@lime pivot
for the moment, i'm too lazy to figure out with alert, but i've put a print @lime pivot , however, i still get cannot find contentview in scope
turbid fjord
```
.alert("Hit !", isPresented: true, actions: {
...
You will want to make an @State var that’s a Boolean for the alerts isPresented value like it’s shown here
https://www.hackingwithswift.com/quick-start/swiftui/how-to-show-an-alert
Then set the variable true/false depending on what your trying to do
If the variable is true then the alert will be presented
noble zodiac
Hey devs, how can I make an app restart the device that host it with non sandbox and root entitlements ?
sour ruin
Hey guys doing some SpriteKit but cant figure out why this doesnt work (Trying to move player)
unkempt raft
Hey guys doing some SpriteKit but cant figure out why this doesnt work (Trying t...
you haven't ran addChild on the scene to add a sprite. The instance is created, but isn't added to the scene
I also suggest you moving the player and move_Right (also rename it to moveRight) to ViewController for them to be instance variables
sour ruin
oh sorry to clarrify that was the ViewController
unkempt raft
oh sorry to clarrify that was the ViewController
I see. I don't have much knowledge with SpriteKit, but I do remember at least on iOS there were touchesBegan, ended etc. Try looking for similar methods found on mac?
sour ruin
you haven't ran `addChild` on the scene to add a sprite. The instance is created...
might sound dumb but idk how to do this
unkempt raft
might sound dumb but idk how to do this
if ran from ViewController, you need to first obtain scene reference (will write how in a moment) and call addChild function
you can create a var scene: GameScene! at the top of ViewController and at the line where there's let sceneNode = scene.rootNode written you can write self.scene = sceneNode
this way the scene will now be set whenever it loads
unkempt raft
if ran from ViewController, you need to first obtain scene reference (will write...
then in keyDown you can reference it the way I described above
var scene: GameScene!
override func viewDidLoad() {
...
if let sceneNode = scene.rootNode as? GameScene {
self.scene = sceneNode
}
}
func keyDown(...) {
scene.movePlayer()
}
I made a mistake by the way, you rather have to addChild from the GameScene
you can think of viewcontroller as being completely unrelated to game logic
never write anything inside it to move, add sprites
let me know if something's not clear
I'll be glad to help
sour ruin
ok is codeUnit the correct way to identify what key is pressed, if so is there documentation on what key has wha code?
unkempt raft
ok is `codeUnit` the correct way to identify what key is pressed, if so is ther...
found this useful struct https://gist.github.com/swillits/df648e87016772c7f7e5dbed2b345066
sour ruin
found this useful struct https://gist.github.com/swillits/df648e87016772c7f7e5db...
Thanks Im kind of new to swift do you mind expanding on the addChild
unkempt raft
Thanks Im kind of new to swift do you mind expanding on the `addChild`
yes
Your scene (GameScene.swift usually contains GameScene in the template) has a function addChild
you can run the function addChild when the scene loads
in didLoad e.g.
addChild(spriteReference)
sour ruin
and then the node i would put in brackets is the just any sprite in the scene or...
tardy narwhal
I love you Craig but as an Apple SVP I expect you to know your own documentations!!
sour ruin
https://developer.apple.com/documentation/spritekit/drawing_spritekit_content_in...
i know that
im trying to move a sprite after a keypress
@unkempt raft this is where im at still no movement either
thorn ether
Is there a way to enable landscape mode on a portrait mode locked app via Filza?
untold drift
why i get error: excess elements in scalar initializer
trying to find symbols and hook a function with libhooker.
doesnt any example of libhooker in github. its so hard to understand for me 😦
whats dc trick for show highlight code
anway doenst work xd
gentle grove
```
Code
```
untold drift
thanks 😄
gentle grove
\`\`\`
Code
\`\`\`
This
gentle grove
oh, what key board is it
untold drift
TR qwerty
const struct LHFunctionHook *hooks= {getC2MChecksum,rep_getC2MChecksum,0,options};
gentle grove
Maybe this https://apple.stackexchange.com/a/290304
untold drift
Maybe this https://apple.stackexchange.com/a/290304
its windows os
gentle grove
Try it
untold drift
copy paste work always xd
gentle grove
Oh
untold drift
```void* getC2MChecksum = functions[1];
```
i think there is an error here
symbols looking like an array
why no one share an example of libhooker 😦
thats looking like an array in array
and return error about of that
i will try another way. i want to learn libhooker but its so hard for code. substrate is very easy :d
if not function return type void ? or function has a parameters?
you did very good example. i will play with them all of night 🙂
im very thankful
i have last question about that. we found the symbol's address. but how can i call it ?
void *addressOfNSLog = addresses[0];
addressOfNSLog(); is enough?
typedef const char* (*FunctionType)(const char*);
FunctionType function = (FunctionType)addressOfgetC2MChecksum;
const char* res = function("test");
i tried that but crashing app when launch :d
yes
how can i found ptrauth_key_function_pointer?
ah okay i will try
i tried that but addresses same after authentication
untold drift
i cant access crash log bcs using windows without type c connection for phone 😦 i looked cr4shed tweak but not working.
tepid olive
still so weird seeing this channel actually used for development
thorn ether
tytyty
thorn ether
worked flawlessly for me on my ipad for a trollstore app, didnt have to resideload it either
noble zodiac
Useless but who knows ?
https://github.com/c22dev/TrollstoreDetector
timid furnace
hasty ruin
hardy hearth
I am trying to use theos/bin/nic.pl but keep getting "Killed: 9" Are there any solutions? iOS 14
I tried both manually installing and theosinstaller and the issue still persists
restive ether
delete everything and start over
and make sure you actually read step by step
and definitely don't use a theos installer becuase they're all wrong
hardy hearth
delete everything and start over
I did rm -rf /var/theos is there anything else I need to delete?
Ye you are right I ran it as root and I reckon that was the issue, let me try without