Homebrew ARM9Loader -- Technical Details and Discussion

[Supster131]

I'll wait on the screen init update until there's screen de-init implemented into the CFWs. I've gotten reports that the screen init brings back the 3D bug for n3DS users. So I'ma hold off on it for now.

That, and apparently Decrypt9 can't decrypt/encrypt.

 

[Mrrraou]

That, and apparently Decrypt9 can't decrypt/encrypt.

That is a known issue, and it's easy to know why, actually...

 

[Supster131]

That is a known issue, and it's easy to know why, actually...

Easy to know why, but is it easy to fix?

Once we have a complete working Decrypt9 and it doesn't break the 3D on the n3DS, then it'll be very ideal.

 

[Apache Thunder]

For now D9 not working from A9L is minor. I can just boot a CIA of it from my 9.2 Emunand. Works just fine from there.

 

[FenrirWolf]

Yeah. At first I was averse to the idea of switching everything over to sysnand, but I think I might just do it later tonight. So long as the FIRM write blocking code in AuReiNand works 100% of the time, it's not like there's any chance of a system update clobbering A9LH, right?

 

[ChaosRipple]

Yeah. At first I was averse to the idea of switching everything over to sysnand, but I think I might just do it later tonight. So long as the FIRM write blocking code in AuReiNand works 100% of the time, it's not like there's any chance of a system update clobbering A9LH, right?

I personally wouldn't risk it. I'm sticking to Emunand. It's like a safety net over the case where somehow you accidentally touch the FIRM or somehow Nintendo manages to fix the FIRM.

 

[democracy]

To summarize upcoming changes for arm9hax update for New 3DS;
-Sounds like the upcoming arm9hax with decrypt9 doesn't provide any more brick protection than the current release.
-We trade 3d on boot for decrypt9 on boot and 3D bug.
Correct?

Seems that New 3DS users are better served staying on old arm9hax until more advancements have been made.

 

[Mrrraou]

-Sounds like the upcoming arm9hax with decrypt9 doesn't provide any more brick protection than the current release.

wtf ? what brick protection ? which d9 ?

-We trade 3d on boot for decrypt9 on boot and 3D bug.
Correct?

Absolutely not... We trade the little 3D bug (that can be fixed by closing the lid for some seconds) on N3DS for a screen init from a9lhax, which means we can have for example the CakesFW menu from arm9loaderhax.

Seems that New 3DS users are better served staying on old arm9hax until more advancements have been made.

And this isn't released yet.

 

[Orkna]

3D? I keep my slider at the bottom at all time.

 

[TimX24968B]

I just read the info on the first pages and I've heard this allows for kernel in sysnand and such, but I can't understand half of what he says on the first page of this thread. I just read what the possibilities are.

 

[Urbanshadow]

Does a9lh modify sysnand's NCSD header?

 

[Mrrraou]

Does a9lh modify sysnand's NCSD header?

No. The NCSD is signed and checked by the bootROM.

 

[Urbanshadow]

No. The NCSD is signed and checked by the bootROM.

But we add our decrypted payload on the end of firm0, doesn't that carry NAND partition start by the payload size?

 

[Mrrraou]

But we add our decrypted payload on the end of firm0, doesn't that carry NAND partition start by the payload size?

The end of FIRM0 isn't signed. And FIRM0 is 8mb long, IIRC.

 

[Urbanshadow]

The end of FIRM0 isn't signed. And FIRM0 is 8mb long, IIRC.

Oh, but the firmware doesn't fill it or pad it. There's plenty of space left, but we don't need too much.

Then, technically a backup of the four things that compose a9lh (firm0,firm1,secret sector and 0x5C000 stage) should install/uninstall it pretty fast, instead of making a whole nand restore, is it?

 

[Mrrraou]

Oh, but the firmware doesn't fill it or pad it. There's plenty of space left, but we don't need too much.

Then, technically a backup of the four things that compose a9lh (firm0,firm1,secret sector and 0x5C000 stage) should install/uninstall it pretty fast, instead of making a whole nand restore, is it?

Yes, it would be pretty fast, and doing a backup of the CTRNAND would help too.

 

[Urbanshadow]

Yes, it would be pretty fast, and doing a backup of the CTRNAND would help too.

How big is it? By the offsets on 3dbrew is something around 180 KB but also says it weights around 800 MB (I guess it's wrong). Or is big endian?

 

[Mrrraou]

How big is it? By the offsets on 3dbrew is something around 180 KB but also says it weights around 800 MB (I guess it's wrong).

CTRNAND can be 1GB. CTRNAND is almost your NAND. It contains the most important stuff.

 

[Urbanshadow]

CTRNAND can be 1GB. CTRNAND is almost your NAND. It contains the most important stuff.

That's the slowy part, but I get the point. Just CTRNAND backups wouldn't touch a9lh state.

We can actually get the firm0/firm1 secret sector and 0x5C000 from a encrypted backup on pc?

 

[Mrrraou]

That's the slowy part, but I get the point. Just CTRNAND backups wouldn't touch a9lh state.

We can actually get the firm0/firm1 secret sector and 0x5C000 from a encrypted backup on pc?

Yeah, even if you can't decrypt firm0, firm1 and the secret sector. But you can.