Hacking Official Corbenik - Another CFW for advanced users (with bytecode patches!)

[gnmmarechal]

Wrong, it still working, I tested just now

It does? Nice then.

 

[ih8ih8sn0w]

It does? Nice then.

Modifications which don't rely on hardcoded offsets or modified code.bins are great

 

[The Catboy]

I suppose the Pokemon S/M outline patch needs to be changed for 1.1, correct?

I thought I posted that it worked, but it appears a sever error only saved the message and didn't post it. The cool thing about the patches for this is that they will always work, compared to the code.bin method

 

[Gray_Jack]

I made the SciresM's patches dynamic, seen to work, but I like some others testers to make sure I did nothing wrong

https://gist.github.com/GrayJack/fc5adeb74f99eab45924e3f80197195a

I decided to use gist cause it looks like a better way to show people the patch instead of creating a new branch, commit the patch and tell people to clone the entire repository to try it.


For the ones that wish to try the patch but don't have the environment set up to compile it or just don't want/can compile:

https://up1.secretalgorithm.com/#d9GCe0wAWzF1jxM461laPA

Just put the .vco file inside the SD:/corkenik/bin folder.

 

[chaoskagami]

Managed to get it to "work" with hs, but stuff wasn't actually getting beyond the nintendo 3ds screen with the exception of fbi 2.whatever the latest build is (this would go into a fatal error: Failed to set syscore CPU time limit: D900A82F). Without sigpatches, the Nintendo 3ds screen doesn't even appear, and it just black screens. Didn't get too much time to attempt to get further along and try titles outside of hs (I would need to install them while on 2.1), but I guess it semi-works?
Other random thing, launching it twice would corrupt the bottom screen display, but inputs from there still worked properly. Solved by launching another app.

Yeah, that's probably just application incompatibility. ctrulib assumes the system isn't running 2.x. I'm more surprised FBI even boots, honestly. Quick check; you're just using sigpatches, right? I'm doubtful loader injection is sane pre-rewrite on the 2.x FIRMs.

Here it is, then

Spoiler: error log

Cache: Region Free HOME (Loader)
Version: 10
cache: 0004003000008F02
cache: 0004003000008202
cache: 0004003000009802
cache: 000400300000A102
cache: 000400300000A902
cache: 000400300000B102
Cache: Title Downgrade Fix (11.0+ NFIRM)
Version: 10
cache: 0004013800000002
cache: 0004013820000002
Cache: Remove Outlines - Pokemon S/M (Loader)
Version: 1
cache: 0004000000164800
cache: 0004000000175E00
arena: 35317068
ordblks: 5
uordblks: 35314708
fordblks: 2360
arena: 1762636
ordblks: 2
uordblks: 1758572
fordblks: 4064
Firmlaunch failed, returning to menu
Cache: Remove Outlines - Pokemon S/M (Loader)
Version: 1
cache: 0004000000164800
cache: 0004000000175E00
arena: 1762636
ordblks: 3
uordblks: 1759132
fordblks: 3504
Firmlaunch failed, returning to menu
arena: 1762636
ordblks: 2
uordblks: 1758572
fordblks: 4064
Firmlaunch failed, returning to menu
arena: 1762636
ordblks: 2
uordblks: 1758612
fordblks: 4024
Firmlaunch failed, returning to menu
Cache: Region Free HOME (Loader)
Version: 10
cache: 0004003000008F02
cache: 0004003000008202
cache: 0004003000009802
cache: 000400300000A102
cache: 000400300000A902
cache: 000400300000B102
Cache: Settings Version String (Loader)
Version: 10
cache: 0004001000021000
cache: 0004001000020000
cache: 0004001000022000
cache: 0004001000026000
cache: 0004001000027000
cache: 0004001000028000
Cache: Block eShop Updates (Loader)
Version: 10
cache: 0004013000002C02
Cache: Block Cart Update / Cart RF (Loader)
Version: 10
cache: 0004013000008002
Cache: Remove Outlines - Pokemon S/M (Loader)
Version: 1
cache: 0004000000164800
cache: 0004000000175E00
arena: 1762636
ordblks: 5
uordblks: 1761452
fordblks: 1184
Firmlaunch failed, returning to menu

Did you run either o3ds_firm.sh or n3ds_firm.sh, and do you have slot0x11key96.bin on your SD? It's not even running the bytecode, so it's failing to load the firm for whatever reason. I'm inclined to believe it's improperly set up.

I made the SciresM's patches dynamic, seen to work, but I like some others testers to make sure I did nothing wrong

https://gist.github.com/GrayJack/fc5adeb74f99eab45924e3f80197195a

I decided to use gist cause it looks like a better way to show people the patch instead of creating a new branch, commit the patch and tell people to clone the entire repository to try it.

For the ones that wish to try the patch but don't have the environment set up to compile it or just don't want/can compile:

https://up1.secretalgorithm.com/#d9GCe0wAWzF1jxM461laPA

Just put the .vco file inside the SD:/corkenik/bin folder.

I'm more than slightly curious what's up with the large amount of 3F. Anyways, since I rather want to go on a island scan binge, I'll test.

On the topic of the 2.x sigchecks; I shuffled around the source tree some, so it's probably going to merge conflict if you attempt to PR it. Sorry, my bad. :<

I'll go ahead and merge it in from your fork manually. From the reports, it seems to work as intended.

The cleanup was more-than-slightly required, since some of the stuff in external isn't external, and I'm in the process of sorting out all the source dependencies.

I thought I posted that it worked, but it appears a sever error only saved the message and didn't post it. The cool thing about the patches for this is that they will always work, compared to the code.bin method

Short of severe rewriting by Nintendo / Game Freak / Whatever, anyways. And TWL is the one thing using fixed offsets. I should *really* fix that.

 

[Gray_Jack]

I'm more than slightly curious what's up with the large amount of 3F

Me too, I had a big help of a friend who knows assembler, but he/she doesn't want any mention of him/her.

Anyways, since I rather want to go on a island scan binge, I'll test.

Please, report if it worked. :3

On the topic of the 2.x sigchecks; I shuffled around the source tree some, so it's probably going to merge conflict if you attempt to PR it. Sorry, my bad. :<

I'll go ahead and merge it in from your fork manually. From the reports, it seems to work as intended.

The cleanup was more-than-slightly required, since some of the stuff in external isn't external, and I'm in the process of sorting out all the source dependencies.

Don't worry, I would never be mad for you making you software even better.
And that is another reason to use gist to yet experimental patches instead of a branch. :3

 

[chaoskagami]

Me too, I had a big help of a friend who knows assembler, but he/she doesn't want any mention of him/her.

Please, report if it worked. :3

I suspect it won't from looking at it and SciresM's code. Honestly, I'm going to apply this manually and figure out what it actually does first. The 3F nonsense doesn't look right to me. (Ugh, need to complete the VM rewrite so I can actually patch branches in VM...)

Don't worry, I would never be mad for you making you software even better.
And that is another reason to use gist to yet experimental patches instead of a branch. :3

Well, the way I merged wasn't just slapping the patch in my tree; I preserved the individual commit and used a merge commit to fix up the path. I prefer to not squash and lose attribution.

 

[Gray_Jack]

I suspect it won't from looking at it and SciresM's code. Honestly, I'm going to apply this manually and figure out what it actually does first. The 3F nonsense doesn't look right to me. (Ugh, need to complete the VM rewrite so I can actually patch branches in VM...)

Actually, It is very similar with the kitling patch , the set parts, except that my friend and I made it in more parts, now that I checked his gist.
And there is a lot of 3F too.

 

[Gray_Jack]

@chaoskagami
I'm having compilation issues with the new commits, not sure if is due the fact that I updated the devkitARM to r46 or if is a new compilation bug

 

[BARNWEY]

I made the SciresM's patches dynamic, seen to work, but I like some others testers to make sure I did nothing wrong

https://gist.github.com/GrayJack/fc5adeb74f99eab45924e3f80197195a

I decided to use gist cause it looks like a better way to show people the patch instead of creating a new branch, commit the patch and tell people to clone the entire repository to try it.


For the ones that wish to try the patch but don't have the environment set up to compile it or just don't want/can compile:

https://up1.secretalgorithm.com/#d9GCe0wAWzF1jxM461laPA

Just put the .vco file inside the SD:/corkenik/bin folder.

Hmmm, I'm kind of interested.... Quick question: Did you test this on corbenik or skeith?

 

[Gray_Jack]

Hmmm, I'm kind of interested.... Quick question: Did you test this on corbenik or skeith?

Skeith, but I don't remember the VM being updated after the last stable release, so I guess it should work with the Stable corbenik too

 

[BARNWEY]

Skeith, but I don't remember the VM being updated after the last stable release, so I guess it should work with the Stable corbenik too

Alright, I'll test it out...

 

[ih8ih8sn0w]

Yeah, that's probably just application incompatibility. ctrulib assumes the system isn't running 2.x. I'm more surprised FBI even boots, honestly. Quick check; you're just using sigpatches, right? I'm doubtful loader injection is sane pre-rewrite on the 2.x FIRMs.

I was only using sigpatches and firm protection. Going to try to get the ancient sysdowngraders to work so I can at least have access to signed test menu, devmenu, and some other titles... Fwiw, I tried using a cn dump, but that failed... Might be due to me installing it while on 11.x?

 

[BARNWEY]

Alright, I'll test it out...

My 3DS crashed when I tried to load Moon XD

 

[Sonic Angel Knight]

What is the easiest way to update this without needing a pc?

 

[BARNWEY]

What is the easiest way to update this without needing a pc?

Using Corbenik Updater RE:

 

[Gray_Jack]

My 3DS crashed when I tried to load Moon XD

What update? 1.0 or 1.1?

 

[Sonic Angel Knight]

Using Corbenik Updater RE:

Will this replace the default a9lh bin file on the sd card? I usually have luma as default and chain loaded the sketh or corbenik... which ever is supposed to be called, by holding down on d-pad during boot.

 

[BARNWEY]

Will this replace the default a9lh bin file on the sd card? I usually have luma as default and chain loaded the sketh or corbenik... which ever is supposed to be called, by holding down on d-pad during boot.

Yes, it will replace it...

--------------------- MERGED ---------------------------

What update? 1.0 or 1.1?

1.1

 

[Sonic Angel Knight]

Is there a option to maybe redirect the path to update with a text file or something, so i can just direct it to the other directory where is located? That be nice.