DSiBrew - User contributions [en]

Talk:Main Page

2009-09-03T14:39:01Z

Scanlime: /* The DSiBrew Forums */

== Main Page color? ==

I'm planning on creating a WiiBrew-like Main Page, but I don't know what color it should be. As far as i can tell, the DSi only comes in black and white, both of which would not be good colors. Any thoughts?
:If you can do some type of shade of brown with yellow or white text might be good. I agree that black is not a very good option though. [[User:Pinball Wizard|Pinball Wizard]] 03:02, 27 January 2009 (UTC)

:Why not use the same Main Page as WiiBrew but with another logo, it looks and works great also makes the projects connect better. [[User:Tigertailz|Tigertailz]]

::''Pinball Wizard:'' What's the reasoning behind the brown? I don't really see any connection. If it were to be brown, it should definitely be a light shade. Black text is a must on a website where the entire background is white.
::''Tigertailz:'' I don't really think blue should be used again. This wiki's main page should be colored so that it is different from WiiBrew. I plan to keep the same page formatting, though.
::Anyways, once I get my admin powers transferred over here from WiiBrew, I'll transfer some templates and start working.
::<tt>--[[User:Blooper4912|'''Blooper''']] [[User talk:Blooper4912|(Talk)]]</tt> 00:58, 28 January 2009 (UTC)
:::The reasoning behind light brown is for the entire wiki to set it apart from other layouts. i do see your point though, We (you) should stick with something lighter but absolutely different, like you said to distinguish the wikis. I Suggest an uncommon color. [[User:Pinball Wizard|Pinball Wizard]] 01:29, 28 January 2009 (UTC)
::::I too vote for a similar looking page as Wiibrew--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 14:57, 25 March 2009 (UTC)
:I vote for a light blue, as the DSi was released in America with the colors being black and light blue instead of white.--winmaster

==logo contest==
since there is only an ugly logo without stating its the dsibrew logo i think there should be a better one. for wiibrew it was a nice startup. [[User:Bitflusher|Bitflusher]] 11:52, 27 January 2009 (UTC)

==IRC==
will there be a dsibrew irc?

We can use Xat.com, it's flash chat. [[User:Duhow|Duhow]]

See #dsihelp, #dsidev, and #dsidevot on the EFNET server.--[[User:Arikado|Arikado]] 00:00, 7 March 2009 (UTC)

==Forums==
I know this project just started but are we looking to set up a forum? [[User:Pinball Wizard|Pinball Wizard]] 22:58, 27 January 2009 (UTC) 
Yeah, I was thinking just that. Funnily enough, forum.dsibrew.org goes to the WiiBrew Forums... anyways, if this site did have a forum, I think it would be more idle speculation than it would be actual exciting DSi native code developments... --[[User:Zack|Zack]] 22:41, 11 June 2009 (UTC)

==Wiki==
I am having 2 issues that is low priority at the moment but will sooner or later need to be addressed: 1)The wiki takes 5-10 minutes to update from any changes. Not a biggy but Wiibrew used to have the issue for a few weeks then magically fixed itself. 2)I don't know how it works but on the wiibrew wiki you can use the command <youtube>XXXXXXXXXX</youtube> where the XXXX's represent the video id. Is it possible to get that added onto this wiki? This is very very very low on the list but could become useful. [[User:Pinball Wizard|Pinball Wizard]] 23:32, 28 January 2009 (UTC)

#Go into your preferences, click the Misc tab, and check the "Disable page caching" box. That should make the wiki update immediately, as long as you are logged in. It may make loading pages in general slower, though.
#None of the MediaWiki extensions have been installed on DSiBrew yet (see [http://dsibrew.org/wiki/Special:Version DSiBrew's extenstion list] as compared to [http://wiibrew.org/wiki/Special:Version WiiBrew's]. Hopefully they will be installed soon.
:Hope that helps. <tt>--[[User:Blooper4912|'''Blooper''']] [[User talk:Blooper4912|(Talk)]]</tt> 00:48, 30 January 2009 (UTC)
::Yeah that helps a lot. Now I can feel a little more useful on wiki work with changing things as needed. I figured it was a mediawiki extension. I don't know if you are coming up with a list of needed extensions though and wanted to make sure that the youtube one was at least considered. [[User:Pinball Wizard|Pinball Wizard]] 23:03, 30 January 2009 (UTC)

== Logo ==

Hi i was wondering if you needed a better logo for this site so i wiped one up for you. you don't have to use it but would be awesome if you did. --[[User:Thetooth|Thetooth]]

[[Image:DSiBrew.png]]

I have also made a logo =P
I could probably make it nicer with more time... --[[User:Bg4545|Bg4545]] 08:14, 23 February 2009 (UTC)
[[Image:Bg4545dsibrewlogo.png|thumb|left]]

: I like this << one --[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 14:59, 25 March 2009 (UTC)

::<strike>Thanks --[[User:Bg4545|Bg4545]] 01:24, 27 March 2009 (UTC)</strike>

My try --[[User:Mvit|Mvit]] 04:04, 26 March 2009 (UTC)[[Image:dsibrew.png]]
:Sorry, Bg4545, I like this one more>> maybe an official contest should be held --[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 22:13, 27 March 2009 (UTC)
The one above is my favourite! It rocks! [[User:TwoBladedKnight|TwoBladedKnight]]

Mvit's one (the one which is slanted and without a DS) is hands-down the best. It doesn't rely on images of the DSi to work, and it looks pretty good and fits with the Wiibrew aesthetic. --[[User:Rhazes|Rhazes]] 20:14, 11 April 2009 (UTC)

[[Image:DSi_Brew_logo-Tabibito.png|200px]]

Does anyone have a template for the text? I don't have a matching font so I just used what I though looked best. I'm not really happy with the fonts though, I might redo it with new fonts if I ever get the chance. <tt>--[[User:Tabibito|'''Tabibito''']] [[User talk:Tabibito|>Talk]]</tt> 01:55, 4 April 2009 (UTC)

:On mine I just googled the DSi logo and pasted it. Then I used a type layer for "Brew". --[[User:Bg4545|bg4545]] 02:30, 4 April 2009 (UTC)

Hey everybody! Here is my logo I made for DSiBrew:

[[Image:Dsibrew-logo.png]] [[Image:Dsibrew-logo2.png]] [[Image:Dsibrew-logo2a.png]]

I'm not a great homebrew programmer but I know how to use Photoshop ;D
Hope you enjoy my logo, I would be really happy if you use it! --[[User:Makorus|Makorus]] 16:29, 8 April 2009 (UTC)
:i like the pic but why is the dsi covered in seaweed? o.O
--[[User:Mercluke|Mercluke]]
:Seaweed? This should be some kind of Matrix-effect.. But I'm happy that you like it
--[[User:Makorus|Makorus]] 10:44, 9 April 2009 (UTC)
:oh, wouldn't it make more sense to use a black DSi so it's like the old cmd promt pcs that use green text and black background?
--[[User:Mercluke|Mercluke]]

:I've added an inverted version of my logo... I think a white DSi with that Matrix-effect looks way better than a black one.

--[[User:Makorus|Makorus]] 10:34, 10 April 2009 (UTC)

:Updated the black matrix DSi. I just removed the headphone jack, because I think that the camera and the speakers are quite characteristic for the new DS.

--[[User:Makorus|Makorus]] 12:26, 12 April 2009 (UTC)

::How is this: [[Image:Wiiguy-dsibrewlogo.jpg]][[User:Wiiguy|Wiiguy]] 02:37, 10 April 2009 (UTC)
::Very good. Making it like Wiibrew is a good idea.--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 05:53, 10 April 2009 (UTC)

::I hope you guys replace the logo with something better soon. My vote goes to the black matrix DSi, but it needs to have its headphone jack colored black.--winmaster

<big>'''PLEASE NOW POST YOUR LOGOS IN THE PAGE [[DSiBrew:Contests]]'''</big> --[[User:CashMan|CashMan]] 12:52, 12 April 2009 (UTC)

==America Here==
So After April 5th DSi will be everywhere that its been announced. Looks like the fun's about to begin.BTW:Will it still be team twiizers or is everything going to change? --[[User:ChuckBartowski|ChuckBartowski]] 05:49, 29 March 2009 (UTC)
: SOMEONE needs to find a hack that lets you boot homebrew. Then a few more small steps and we have a DSi Homebrew Channel and easy ports of all the existing DS homebrew. Hopefully this creates new interest in developing for DS homebrew and we get a lot more apps with fixes for all the buggy old ones. Which brings me to a question, can people start working on making pages for existing DS homebrew projects?--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 15:27, 2 April 2009 (UTC)
::That's actually a good question. Is this wiki for only DSi homebrew or both DS and DSi homebrew? What if a homebrew works on the DS but not on the DSi? Should that be included on the wiki? I think this warrants some discussion. Personally I think it wouldn't be a bad idea to include information on all past DS homebrew because there currently is not one central place to find DS homebrew. <tt>--[[User:Blooper4912|'''Blooper''']] [[User talk:Blooper4912|(Talk)]]</tt> 02:45, 3 April 2009 (UTC)
:::I personally think we should include DS homebrew that works with the DSi, and have information on DS homebrew that people are porting to the DSi. However, I don't see any reason to cover homebrew for the DS that doesn't work on the DSi unless it's to solicit people to work on porting them. <tt>--[[User:Tabibito|'''Tabibito''']] [[User talk:Tabibito|>Talk]]</tt> 00:41, 4 April 2009 (UTC)
::I think you should include all DS hacks that work on the DSi for sure. You should also have a page dedicated to DS hacks that don't work on the DSi but don't require extra hardware. (not sure if there is any)--winmaster
: ?Team Twiizers is working on a hack if you read their blog, but I'm sure a lot of other parties will play a big role. --[[User:Zack|Zack]] 13:51, 16 June 2009 (UTC)

==DSi Hack page==
I created the [[DSi hack]] page so hackers can opt to update us on their project or get help on the [[talk:DSi hack|talk page]]. It's linked on the main page.--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 16:58, 4 April 2009 (UTC)

== New GUI ==

I've created a new GUI for the main page like WiiBrew --[[User:CashMan|CashMan]] 22:07, 4 April 2009 (UTC)
:looks good!--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 03:19, 5 April 2009 (UTC)

== flashcarts ==
Can we get proof that DSi flashcarts still work?--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 03:20, 6 April 2009 (UTC)
: I, for one, believe the Internet. If I get a DSi, however, I plan to test all flashcarts I have. --[[User:Zack|Zack]] 13:50, 16 June 2009 (UTC)

== InterWiki ==

Hey guys,

what do you think about InterWiki linking between WiiBrew and DSiBrew?

like <nowiki>[[wb:WUFE|Wii Update File Extractor]] or [[dsi:System_Menu|DSI's Menu]]</nowiki>?
: Don't see the point. The most they need are links on the sidebar to help users move around the homebrew wikis. --[[User:Zack|Zack]] 13:49, 16 June 2009 (UTC)

== DS(i) Download-Game-Funtion ==

The Nintendo DS(i) has a Function to Download games via WLAN/WIFI, is it possible to modify the "Nintendo Wi-Fi USB-Connector"-Software to start Homebrew, like a Installer for a "Homebrew-Channel"? It´s just a idea from me, but i think it can be easy. (I´m sorry for my bad english, but i´m from Germany and i´m not very good in english...) --[[User:Plinfa|Plinfa]] 23:49, 2 May 2009 (UTC)
: I was actually thinking of something like that, except for normal DS Download Play. You can modify a wireless network card to send out DS Download signals, however, as someone pointed out in the forums, very few cards support this. For the ones that do, it might be possible for the owners and their friends and family to get homebrew. Maybe they could take the network cards and a laptop to conventions and spread homebrew further. I don't know. --[[User:Zack|Zack]] 13:47, 16 June 2009 (UTC)

== Userboxes ==

I added a few crappy user boxes. You can see them in action on my page. I do not have a DSi, so I do not use the DSi userbox.--[[User:Zack|Zack]] 14:10, 16 June 2009 (UTC)

== The DSiBrew Forums ==

http://dsiforums.podtube.us/index.php

Currently unofficial. Try it out. --[[User:Zack|Zack]] 18:39, 17 June 2009 (UTC)

:You sure activation works? Still waiting for my e-mail. [[User:Muzer|Muzer]] 19:59, 17 June 2009 (UTC)

::Works now, sorry. --[[User:Zack|Zack]] 20:47, 17 June 2009 (UTC)

:::Theres really no need for a forum ATM <_< Also, if its unofficial, I'm not sure if it should be on the main page of DSiBrew. Your user page perhaps?--[[User:Arikado|Arikado]] 20:48, 18 June 2009 (UTC)

::::That's what I said above, but then I ended up making it anyways... I do see your point, but at least it's marked as unofficial. --[[User:Zack|Zack]] 03:33, 19 June 2009 (UTC)

:::::sorry, I think you should just make 'em official.--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 07:14, 20 June 2009 (UTC)

::::::I cant agree with that. WiiBrew, the WiiBrew Forums, and DSiBrew are all hosted on the same server AFAIK. Until we get a 'go-ahead' from someone like bushing, they should remain unofficial. So hop on the IRC already and get working on getting that 'go-ahead' :P--[[User:Arikado|Arikado]] 12:31, 20 June 2009 (UTC)

::::::Yeah, I was thinking about that, but I can't really imagine any sort of plea that wouldn't sound mildly retarded... "Hi, I made DSiBrew a forum. Any chance it could be official now? People are using it." "DSiBrew needed a forum, I made one. Here you go. Please make it official, people are on it." I'm a horrible conversationalist. --[[User:Zack|Zack]] 18:53, 20 June 2009 (UTC)

:::::::i will help. "I dont know if youve heard yet, but i have made an unnoficiall forum for DSi Brew. It is set up mostly like the wiibrew forums. There are a few users on it already, and is gaining in popularity. (then you put link to page here) thats the link. If you guys want to move it and/or make it "official" let me know (put email here)."--[[User:ChuckBartowski|ChuckBartowski]] 03:57, 24 June 2009 (UTC)

::::::::Sniff... it's... it's BEAUTIFUL! Thanks, Chuck! --[[User:Zack|Zack]] 04:13, 24 June 2009 (UTC)
:::::::::Your welcome. --[[User:ChuckBartowski|ChuckBartowski]] 22:51, 25 June 2009 (UTC)
::::::::::::Anyone mind telling me Why our link got removed? --[[User:Remadon|Remadon]] 03:15, 3 September 2009 (UTC)
:::::::::::::Sure. You ripped off the forum software and layout from the Wiibrew forum -- down to my witty forum subtitles -- and then filled it with silly topics like [http://dsiforums.podtube.us/read.php?4,1345 this] and [http://dsiforums.podtube.us/read.php?11,938 this]. -- [[User:Bushing|Bushing]] 05:57, 3 September 2009 (UTC)
::::::::::::::if you didn't like you forums being ripped off, why don't you make the DSi forum yourself?--[[User:Funkamatic|FUNKAMATIC]][[User talk:Funkamatic| ~talk]] 14:58, 3 September 2009 (UTC)
:::::::::::::::As Arikado said earlier, I really don't think we have much need for a forum at the moment.. right now DSi homebrew is still in its early stages, and we still mostly need people with a large amount of experience with reverse engineering, hardware, and embedded systems. So there isn't a really large audience right now for anything we're doing, and IRC has been serving pretty well for the reverse engineering efforts. -- [[User:Scanlime|Scanlime]] 16:39, 3 September 2009 (UTC)

== Want to Help-Box ==

The Text in the Box "Want to Help" should be replaced, because run Homebrew in DSi-Mode is now possible! --[[User:Plinfa|Plinfa]] 23:22, 16 July 2009 (UTC)
:You like it better now? --[[User:Zack|Zack]] 23:16, 18 July 2009 (UTC)

== About the news "DSi Download Center" ==

This is not legal I think...
It's a website where you pay for ROMS !
It's should be removed --[[User:Corenting|Corenting]] 06:01, 25 August 2009 (UTC)
:It's almost definitely a scam too. Should be removed --[[User:Shrodes|Shrodes]] 04:27, 29 August 2009 (UTC)
:: I fixed it up. News on flashcard firmware is irrelevant to DSiBrew too at this stage IMO. --[[User:Shrodes|Shrodes]] 04:32, 29 August 2009 (UTC)

List of DSi Homebrew

2009-09-02T14:37:45Z

Scanlime: Might as well add CookInject too. Can't let WinterMute have all the fun ;)

This '''list''' is designed to be a updated collection of current DSi mode homebrew.

==Regarding the current state of DSi mode homebrew in general==
There is currently no practical way to run homebrew in DSi-mode. Save game hacks exist, but these have severe limitations:
* The SD slot and NAND flash are inaccessible.
* The amount of code space is limited to the available space in the game's save file. (A few kilobytes)
** This restriction can be lifted with external hardware like [[EEPUART]] or [[spime]]. This is how [http://svn.navi.cx/misc/trunk/nds/dsi/cookinject/ larger chunks of code] or [http://www.flickr.com/photos/micahdowty/3794878172/in/set-72157621023570420/ image files] have been loaded by developers so far.

==Exploits==
===The Drunken Coder's savegame exploit===
This exploit allows DSi homebrew coders to run unsigned ARM asm code in DSi-mode utilizing a buffer overflow in Cooking Coach. The hack can be found [http://drunkencoders.com/2009/08/dsi-hack-update/ here].

===CookInject===
This is [[User:Scanlime]]'s exploit for Cooking Coach. It works with external hardware to bootstrap larger amounts of C code into system memory. Its primary feature is that it's possible to modify code from the game's ARM9 or ARM7 binaries, then return from the exploit back to the game. This allows insertion of hooks that modify or trace the game's normal behaviour. The code can be found [http://svn.navi.cx/misc/trunk/nds/dsi/cookinject/ here].

==Homebrew==

==Computer applications==
===pyftdi===
svpe's python binding of libftdi. Grab it and read more information [http://svenpeter.blogspot.com/2009/08/pyftdi.html here].

User:Scanlime

2009-09-02T07:23:59Z

Scanlime: New page: I'm just a random hardware/software geek who's been interested in the DSi lately. I've done a little bit of reverse engineering, some savegame hacking, NAND tracing/reading/writing, and m...

I'm just a random hardware/software geek who's been interested in the DSi lately.

I've done a little bit of reverse engineering, some savegame hacking, NAND tracing/reading/writing, and more recently some slightly more complex hardware hacks.

Spime

2009-09-02T07:10:33Z

Scanlime: New page: Short for the SPI Memory Emulator. This is User:Scanlime's approach to bootstrapping code via savegame hacks and providing a debug channel back to the PC. Whereas EEPUART is a ser...

Short for the SPI Memory Emulator.

This is [[User:Scanlime]]'s approach to bootstrapping code via savegame hacks and providing a debug channel back to the PC. Whereas [[EEPUART]] is a serial port, spime is a shared memory buffer.

It is implemented on a Xilinx Spartan 3 FPGA board in Verilog. A 1MB SRAM buffer is concurrently accessible by the PC over a 3 MBit serial port, and by the DS over an SPI EEPROM emulator. This buffer can store normal save files, plus it can be used as a fast communication channel for memory dumps, commands to execute, etc.

This shared memory buffer can also be used to send simple RPC-like commands to the DSi, implementing a very simple debug monitor for poking at registers in real-time.

Source code is under a BSD-style license, available from a [http://svn.navi.cx/misc/trunk/nds/spi-mem-emulator/ public Subversion repository].

List of DSi Homebrew

2009-09-02T07:05:52Z

Scanlime: Well, it isn't that homebrew is impossible, just impractical...

This list is designed to be a updated collection of current DS and DSi [[homebrew]].

= DSi-mode homebrew =

There is currently no practical way to run homebrew in DSi-mode. Save game hacks exist, but these have severe limitations:
* The SD slot and NAND flash are inaccessible.
* The amount of code space is limited to the available space in the game's save file. (A few kilobytes)
** This restriction can be lifted with external hardware like [[EEPUART]] or [[spime]]. This is how [http://svn.navi.cx/misc/trunk/nds/dsi/cookinject/ larger chunks of code] or [http://www.flickr.com/photos/micahdowty/3794878172/in/set-72157621023570420/ image files] have been loaded by developers so far.

= DS-mode homebrew =

== Games ==

== Emulators ==

== Applications ==

NAND

2009-08-31T18:15:51Z

Scanlime: Add partition starts, note that diag area is blank straight from factory.

The DSi uses a 256MB Samsung moviNAND flash chip, which is a NAND flash chip with a built-in controller that implements a MMC (SDIO) interface. In many ways, it's like an SD card in BGA packaging, and some people have successfully read it with modified SD card readers.

Addressing is done in terms of 512-bytes sectors. All wear-levelling and bad-block-mapping is handled transparently inside the chip by the controller. Most sectors are encrypted with a per-console key.

Layout:

{| border="1" cellpadding="3" cellspacing="0"
! Offset !! Size !! Description
|-
| 0 || 0x200 || PC-style MBR, encrypted with a per-console key
|-
| 0x200 || 0x200 || Stage 2 info block 1 (see below)
|-
| 0x400 || 0x200 || Stage 2 info block 2
|-
| 0x600 || 0x200 || Stage 2 info block 3
|-
| 0x800 || 0x4DC00 || Stage 2 bootloader (encrypted with a universal key)
|-
| 0x4E400 || 0x400 || Stage 2 footer -- unknown format, but first 10 bytes are build number of Stage 2 bootloader
|-
| 0x4E800 || ? || Unused (all 00)
|-
| 0xffa00 || 0x400 || Diagnostic area. (often contains build date of device in plaintext) Blank in never-before-booted DSi. Might be written to during firmware updates.
|-
| 0x10ee00 || ? || First sector of main FAT32 partition
|-
| 0xcf09a00 || ? || First sector of photo FAT32 partition
|}

NVRAM

2009-08-27T06:34:28Z

Scanlime: New page: {{stub}} The NVRAM is a 128 kilobyte SPI Flash memory chip located on the DSi's wifi board. On the DS, this was the system's firmware, plus it included writeable areas for user preferenc...

{{stub}}

The NVRAM is a 128 kilobyte SPI Flash memory chip located on the DSi's wifi board.

On the DS, this was the system's firmware, plus it included writeable areas for user preferences and wifi connection settings. On the DSi, this chip is maintained for backward compatibility, but it is mostly empty.

There is some new unknown data near the beginning of NVRAM which is involved in the boot process. The NVRAM is read very early in boot, before the NAND flash is initialized. An unreadable NVRAM chip will hang the boot process in an infinite retry loop, and errors in this portion of NVRAM will cause a stage1 [[Bootloader]] error.

Stage2

2009-08-27T06:30:23Z

Scanlime: Mention NVRAM, mention that stage2 key is fixed, be more certain that stage2 is really a bootloader

== Stage 1 ==

[[Image:boot-stage1-error.jpeg|frame|When the Stage 1 bootloader (in ROM) fails, it displays a 32-bit hexadecimal number on the top screen.]]

The first stage of the DSi's bootloader lives in ROM, presumably on the CPU die. It loads further encrypted (and probably signed) stages from NAND flash, starting with a (partially unencrypted) offset table in the sector at 0x200.

Not much is known about this bootloader yet, but it presumably knows how to:
# Initialize the encryption hardware
# Read blocks (but not files) from the NAND flash
# Perform some variety of integrity check on all data it reads (signature, CRC, ?)
# Display basic hexadecimal error codes
# Possibly factory-programming the NAND flash?
# Might also do basic power-on self test of peripherals

Known error codes:

{| border="1" cellpadding="3" cellspacing="0"
! Error Code !! Description
|-
| 0000FE00 || Error communicating with NAND chip. (It's missing, CLK is shorted, etc.)
|-
| 0000FEFC || Integrity error in first block of Stage 2 (address at 0x220)
|-
| 0000FEFD || Integrity error in second block of Stage 2 (address at 0x230)
|-
| 0000FEFE || Boot sector integrity error (Sector 0x200 not valid), or error in [[NVRAM]] contents.
|}

== Stage 2 ==

[[Image:boot-stage2-error.jpeg|frame|This may have been a Stage 2 bootloader error.]]

Unlike the stage1 bootloader, which must be small enough to fit in ROM (probably several kilobytes), the stage2 bootloader has about a megabyte of NAND flash reserved for it. The stage2 bootloader understands partitions and filesystems, and it is capable of loading the DSi menu. It also must understand the encryption used on filesystem blocks in the NAND, and it must understand how to load and validate title metadata.

The Stage 2 loader was not modified by the [[System Menu 1.4]] update. This is still earlier in the boot process than the "Health and Safety" warning.

The first stage bootloader reads sector 0x200 in order to find a table of offsets to the Stage 2 bootloader:

00000220 00 08 00 00 10 64 02 00 00 80 7b 03 00 66 02 00 |.....d....{..f..|
00000230 00 6e 02 00 88 75 02 00 00 80 7b 03 00 76 02 00 |.n...u....{..v..|
00000240 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|

This appears to be describing two chunks of the stage2 loader, one 0x26410 bytes in length at address 0x800, and one 0x27588 bytes at address 0x26e00.

Note that this sector (and two similar ones at 0x400 and 0x600) appear to be the only unencrypted blocks on the NAND flash.

It is unclear why there are two pieces which are nearly but not quite the same size. Passive traces of the boot sequence confirm that the 0x26e00 chunk is slightly larger, and it's loaded first. The 0x800 chunk is read immediately after the 0x26e00 chunk.

Whereas the filesystem data in NAND is encrypted using a unique key for every DSi, the stage2 bootloader is identical on every DSi tested so far. This probably means that it is encrypted using a fixed key included in stage1.

After Stage 2 is loaded:
# The NAND flash is partially re-initialized
# Sector 0 is read from the NAND. This appears to be an (encrypted) DOS-style MBR.
# The MBR signature and the type of the first partition are verified.
# Filesystem metadata is read from sectors starting around 0x100000. The metadata appears to be in FAT32 format with long filenames.
# Multiple files are loaded from the filesystem. The exact read addresses will vary depending on your DSi's firmware version and the state of its filesystem when you performed the last firmware update. On a brand new DSi, it appears that the DSi Menu itself is loaded from 0xb20000 after two small metadata files are read from 0xb1c000 and 0x7a0000.

All errors show before the health and safety screen. It appears that stage2 errors from a cold power-on always cause the DSi to hang at a black screen, whereas stage2 errors after reset (pressing but not holding the power button) will give an error message screen. Known errors:

{| border="1" cellpadding="3" cellspacing="0"
! Text !! Description
|-
| "Error: 1-2435-8325" || Invalid signature or partition type in MBR, invalid starting LBA.
|-
| "Error: 3-2435-8325" || DSi Menu integrity checks failed
|-
|}

Stage2

2009-08-18T02:59:38Z

Scanlime:

= Stage 1 =

[[Image:boot-stage1-error.jpeg|frame|When the Stage 1 bootloader (in ROM) fails, it displays a 32-bit hexadecimal number on the top screen.]]

The first stage of the DSi's bootloader lives in ROM, presumably on the CPU die. It loads further encrypted (and probably signed) stages from NAND flash, starting with a (partially unencrypted) offset table in the sector at 0x200.

Not much is known about this bootloader yet, but it presumably knows how to:
# Initialize the encryption hardware
# Read blocks (but not files) from the NAND flash
# Perform some variety of integrity check on all data it reads (signature, CRC, ?)
# Display basic hexadecimal error codes
# Possibly factory-programming the NAND flash?
# Might also do basic power-on self test of peripherals

Known error codes:

{| border="1" cellpadding="3" cellspacing="0"
! Error Code !! Description
|-
| 0000FE00 || Error communicating with NAND chip. (It's missing, CLK is shorted, etc.)
|-
| 0000FEFC || Integrity error in first block of Stage 2 (address at 0x220)
|-
| 0000FEFD || Integrity error in second block of Stage 2 (address at 0x230)
|-
| 0000FEFE || Boot sector integrity error (Sector 0x200 not valid)
|}

= Stage 2 =

[[Image:boot-stage2-error.jpeg|frame|This may have been a Stage 2 bootloader error.]]

Not much concrete information is known about the second-stage bootloader. This may be an actual bootloader, or it's possible that this stage is the DSi Menu itself. The Stage 2 loader was not modified by the [[System Menu 1.4]] update. This is still earlier in the boot process than the "Health and Safety" warning.

The first stage bootloader reads sector 0x200 in order to find a table of offsets to the Stage 2 bootloader:

00000220 00 08 00 00 10 64 02 00 00 80 7b 03 00 66 02 00 |.....d....{..f..|
00000230 00 6e 02 00 88 75 02 00 00 80 7b 03 00 76 02 00 |.n...u....{..v..|
00000240 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|

This appears to be describing two chunks of the stage2 loader, one 0x26410 bytes in length at address 0x800, and one 0x27588 bytes at address 0x26e00.

Note that this sector (and two similar ones at 0x400 and 0x600) appear to be the only unencrypted blocks on the NAND flash.

It is unclear why there are two pieces which are nearly but not quite the same size. Passive traces of the boot sequence confirm that the 0x26e00 chunk is slightly larger, and it's loaded first. The 0x800 chunk is read immediately after the 0x26e00 chunk.

After Stage 2 is loaded:
# The NAND flash is partially re-initialized
# Sector 0 is read from the NAND. This appears to be an (encrypted) DOS-style MBR.
# The MBR signature and the type of the first partition are verified.
# Filesystem metadata is read from sectors starting around 0x100000. The metadata appears to be in FAT32 format with long filenames.
# A file is loaded from 0x790000 (just below 8MB). This may be the DSi Menu. If the "stage 2" loader ''is'' the DSi Menu, this may be data for the photo application that runs on the top screen.

All errors show before the health and safety screen. It appears that stage2 errors from a cold power-on always cause the DSi to hang at a black screen, whereas stage2 errors after reset (pressing but not holding the power button) will give an error message screen. Known errors:

{| border="1" cellpadding="3" cellspacing="0"
! Text !! Description
|-
| "Error: 1-2435-8325" || Invalid signature or partition type in MBR, invalid starting LBA.
|-
| "Error: 3-2435-8325" || Modified the file at 0x790000 (likely the DSi Menu)
|-
|}

Stage2

2009-08-18T02:37:43Z

Scanlime: More about the stage2 bootloader errors.

= Stage 1 =

[[Image:boot-stage1-error.jpeg|frame|When the Stage 1 bootloader (in ROM) fails, it displays a 32-bit hexadecimal number on the top screen.]]

The first stage of the DSi's bootloader lives in ROM, presumably on the CPU die. It loads further encrypted (and probably signed) stages from NAND flash, starting with a (partially unencrypted) offset table in the sector at 0x200.

Not much is known about this bootloader yet, but it presumably knows how to:
# Initialize the encryption hardware
# Read blocks (but not files) from the NAND flash
# Perform some variety of integrity check on all data it reads (signature, CRC, ?)
# Display basic hexadecimal error codes
# Possibly factory-programming the NAND flash?
# Might also do basic power-on self test of peripherals

Known error codes:

{| border="1" cellpadding="3" cellspacing="0"
! Error Code !! Description
|-
| 0000FE00 || Error communicating with NAND chip. (It's missing, CLK is shorted, etc.)
|-
| 0000FEFC || Integrity error in first block of Stage 2 (address at 0x220)
|-
| 0000FEFD || Integrity error in second block of Stage 2 (address at 0x230)
|-
| 0000FEFE || Boot sector integrity error (Sector 0x200 not valid)
|}

= Stage 2 =

[[Image:boot-stage2-error.jpeg|frame|This may have been a Stage 2 bootloader error.]]

Not much concrete information is known about the second-stage bootloader. This may be an actual bootloader, or it's possible that this stage is the DSi Menu itself. The Stage 2 loader was not modified by the [[System Menu 1.4]] update. This is still earlier in the boot process than the "Health and Safety" warning.

The first stage bootloader reads sector 0x200 in order to find a table of offsets to the Stage 2 bootloader:

00000220 00 08 00 00 10 64 02 00 00 80 7b 03 00 66 02 00 |.....d....{..f..|
00000230 00 6e 02 00 88 75 02 00 00 80 7b 03 00 76 02 00 |.n...u....{..v..|
00000240 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|

This appears to be describing two chunks of the stage2 loader, one 0x26410 bytes in length at address 0x800, and one 0x27588 bytes at address 0x26e00.

Note that this sector (and two similar ones at 0x400 and 0x600) appear to be the only unencrypted blocks on the NAND flash.

It is unclear why there are two pieces which are nearly but not quite the same size. Passive traces of the boot sequence confirm that the 0x26e00 chunk is slightly larger, and it's loaded first. The 0x800 chunk is read immediately after the 0x26e00 chunk.

After Stage 2 is loaded:
# The NAND flash is partially re-initialized
# Sector 0 is read from the NAND. This may be an (encrypted) partition table.
# Filesystem metadata is read from sectors starting around 0x100000. The metadata appears to be in FAT32 format with long filenames.
# A file is loaded from 0x790000 (just below 8MB). This may be the DSi Menu. If the "stage 2" loader ''is'' the DSi Menu, this may be data for the photo application that runs on the top screen.

All errors show before the health and safety screen. It appears that stage2 errors from a cold power-on always cause the DSi to hang at a black screen, whereas stage2 errors after reset (pressing but not holding the power button) will give an error message screen. Known errors:

{| border="1" cellpadding="3" cellspacing="0"
! Text !! Description
|-
| "Error: 1-2435-8325" || Modified byte 0x1c2 of sector 0.
|-
| "Error: 3-2435-8325" || Modified the file at 0x790000 (likely the DSi Menu)
|-
|}

Stage2

2009-08-18T02:29:38Z

Scanlime:

= Stage 1 =

[[Image:boot-stage1-error.jpeg|frame|When the Stage 1 bootloader (in ROM) fails, it displays a 32-bit hexadecimal number on the top screen.]]

The first stage of the DSi's bootloader lives in ROM, presumably on the CPU die. It loads further encrypted (and probably signed) stages from NAND flash, starting with a (partially unencrypted) offset table in the sector at 0x200.

Not much is known about this bootloader yet, but it presumably knows how to:
# Initialize the encryption hardware
# Read blocks (but not files) from the NAND flash
# Perform some variety of integrity check on all data it reads (signature, CRC, ?)
# Display basic hexadecimal error codes
# Possibly factory-programming the NAND flash?
# Might also do basic power-on self test of peripherals

Known error codes:

{| border="1" cellpadding="3" cellspacing="0"
! Error Code !! Description
|-
| 0000FE00 || Error communicating with NAND chip. (It's missing, CLK is shorted, etc.)
|-
| 0000FEFC || Integrity error in first block of Stage 2 (address at 0x220)
|-
| 0000FEFD || Integrity error in second block of Stage 2 (address at 0x230)
|-
| 0000FEFE || Boot sector integrity error (Sector 0x200 not valid)
|}

= Stage 2 =

[[Image:boot-stage2-error.jpeg|frame|This may have been a Stage 2 bootloader error.]]

Not much concrete information is known about the second-stage bootloader. This may be an actual bootloader, or it's possible that this stage is the DSi Menu itself. The Stage 2 loader was not modified by the [[System Menu 1.4]] update. This is still earlier in the boot process than the "Health and Safety" warning.

The first stage bootloader reads sector 0x200 in order to find a table of offsets to the Stage 2 bootloader:

00000220 00 08 00 00 10 64 02 00 00 80 7b 03 00 66 02 00 |.....d....{..f..|
00000230 00 6e 02 00 88 75 02 00 00 80 7b 03 00 76 02 00 |.n...u....{..v..|
00000240 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|

This appears to be describing two chunks of the stage2 loader, one 0x26410 bytes in length at address 0x800, and one 0x27588 bytes at address 0x26e00.

Note that this sector (and two similar ones at 0x400 and 0x600) appear to be the only unencrypted blocks on the NAND flash.

It is unclear why there are two pieces which are nearly but not quite the same size. Passive traces of the boot sequence confirm that the 0x26e00 chunk is slightly larger, and it's loaded first. The 0x800 chunk is read immediately after the 0x26e00 chunk.

After Stage 2 is loaded:
# The NAND flash is partially re-initialized
# Sector 0 is read from the NAND. This may be an (encrypted) partition table.
# Filesystem metadata is read from sectors starting around 0x100000. The metadata appears to be in FAT32 format with long filenames.
# A file is loaded from 0x790000 (just below 8MB). This may be the DSi Menu. If the "stage 2" loader ''is'' the DSi Menu, this may be data for the photo application that runs on the top screen.

All errors show before the health and safety screen. Known errors:

{| border="1" cellpadding="3" cellspacing="0"
! Appearance !! Description
|-
| Black screen || Error in sector 0, error in filesystem metadata, error in file at 0x790000.
|-
| "Error: 3-2435-8325" on both screens || XXX: I got this error once when modifying the stage2 loader, but haven't been able to reproduce it since.
|-
| "Error: 1-2435-8325" on both screens || Modified byte 0x1c2 of sector 0.
|-
|}

Main Page/Navigation

2009-08-17T02:08:21Z

Scanlime: Add link to Bootloader

{{Main page box|Navigation|Main Page/Navigation}}
<div style="margin: -.3em -1em -1em -1em;">
{| width="100%" bgcolor="#fff" border="0" cellpadding="2px" cellspacing="2px" style="margin:auto;"
|- align="center" bgcolor="#e7eef6"
! width="33%" | '''General'''
! width="34%" | '''DSi hardware'''
! width="33%" | '''DSi software'''
|- valign="top" style="background: #F5FAFF;"
|
*[[DSiBrew:Contests|Contests]]
*[[DSi exploits]]
*[[Glossary]]
*[[FAQ]]
*[http://dsiforums.podtube.us/index.php Unofficial Forum]
|
*[[Hardware|DSi Hardware]]
*[[Cameras]]
|
*[[Nintendo Software]]
*[[Title list]]
*[[Title metadata]]
*[[SD Filesystem]]
*[[Flash Filesystem]]
*[[Bootloader]]
|}
</div>
{{box-footer-empty}}

Stage2

2009-08-17T02:07:48Z

Scanlime: New page: = Stage 1 = When the Stage 1 bootloader (in ROM) fails, it displays a 32-bit hexadecimal number on the top screen. The first stage of the DSi's boo...

= Stage 1 =

[[Image:boot-stage1-error.jpeg|frame|When the Stage 1 bootloader (in ROM) fails, it displays a 32-bit hexadecimal number on the top screen.]]

The first stage of the DSi's bootloader lives in ROM, presumably on the CPU die. It loads further encrypted (and probably signed) stages from NAND flash, starting with a (partially unencrypted) offset table in the sector at 0x200.

Not much is known about this bootloader yet, but it presumably knows how to:
# Initialize the encryption hardware
# Read blocks (but not files) from the NAND flash
# Perform some variety of integrity check on all data it reads (signature, CRC, ?)
# Display basic hexadecimal error codes
# Possibly factory-programming the NAND flash?
# Might also do basic power-on self test of peripherals

Known error codes:

{| border="1" cellpadding="3" cellspacing="0"
! Error Code !! Description
|-
| 0000FE00 || Error communicating with NAND chip. (It's missing, CLK is shorted, etc.)
|-
| 0000FEFC || Integrity error in first block of Stage 2 (address at 0x220)
|-
| 0000FEFD || Integrity error in second block of Stage 2 (address at 0x230)
|-
| 0000FEFE || Boot sector integrity error (Sector 0x200 not valid)
|}

= Stage 2 =

[[Image:boot-stage2-error.jpeg|frame|This may have been a Stage 2 bootloader error.]]

Not much concrete information is known about the second-stage bootloader. This may be an actual bootloader, or it's possible that this stage is the DSi Menu itself. The Stage 2 loader was not modified by the [[System Menu 1.4]] update. This is still earlier in the boot process than the "Health and Safety" warning.

The first stage bootloader reads sector 0x200 in order to find a table of offsets to the Stage 2 bootloader:

00000220 00 08 00 00 10 64 02 00 00 80 7b 03 00 66 02 00 |.....d....{..f..|
00000230 00 6e 02 00 88 75 02 00 00 80 7b 03 00 76 02 00 |.n...u....{..v..|
00000240 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................|

This appears to be describing two chunks of the stage2 loader, one 0x26410 bytes in length at address 0x800, and one 0x27588 bytes at address 0x26e00.

Note that this sector (and two similar ones at 0x400 and 0x600) appear to be the only unencrypted blocks on the NAND flash.

It is unclear why there are two pieces which are nearly but not quite the same size. Passive traces of the boot sequence confirm that the 0x26e00 chunk is slightly larger, and it's loaded first. The 0x800 chunk is read immediately after the 0x26e00 chunk.

After Stage 2 is loaded:
# The NAND flash is partially re-initialized
# Sector 0 is read from the NAND. This may be an (encrypted) partition table.
# Filesystem metadata is read from sectors starting around 0x100000. The metadata appears to be in FAT32 format with long filenames.
# A file is loaded from 0x790000 (just below 8MB). This may be the DSi Menu. If the "stage 2" loader ''is'' the DSi Menu, this may be data for the photo application that runs on the top screen.

All errors show before the health and safety screen. Known errors:

{| border="1" cellpadding="3" cellspacing="0"
! Appearance !! Description
|-
| Black screen || Error in sector 0, error in filesystem metadata, error in file at 0x7a0000.
|-
| "Error: 3-2435-8325" on both screens || XXX: I got this error once when modifying the stage2 loader, but haven't been able to reproduce it since.
|-
|}

File:Boot-stage2-error.jpeg

2009-08-17T01:46:09Z

Scanlime: "Stage 2" bootloader error.

"Stage 2" bootloader error.

File:Boot-stage1-error.jpeg

2009-08-17T01:09:25Z

Scanlime: Photo of a DSi displaying an error from the "stage 1" ROM bootloader.

Photo of a DSi displaying an error from the "stage 1" ROM bootloader.