PS3 Hacks

#1 Spot for PS3 Hacks

Home | PS3 News | PS3 Hacks | PS3 Downloads | PS3 Saves

Folding@home | PS3-Hacks Live Chat | PS3 Reviews | Contact Us


You are not logged in.

#1  2009-04-29 05:28:20

OneSoul
PS3 Newbie In Training
Registered: 2009-04-25
Posts: 5

JTAG: if exists, WHERE is it?

Hi all!

Because I don't believe that the on-board flash programs by itself, probabily the factory use some kind of connection, like JTAG, to do it...

Where is a possible connector/header/pinout on the PS3 board?
Is it a JTAG or another standard? (see PSP...)

Do you have an idea?

Offline

 

#2  2009-04-29 07:16:08

Powerslave
Ruler of All
From: Alpha Quadrant
Registered: 2007-01-15
Posts: 12433
Website

Re: JTAG: if exists, WHERE is it?

OneSoul wrote:

Hi all!

Because I don't believe that the on-board flash programs by itself, probabily the factory use some kind of connection, like JTAG, to do it...
WRONG!

Where is a possible connector/header/pinout on the PS3 board?
Is it a JTAG or another standard? (see PSP...)
Don't need Jtag or LPC bus.  And not lets not see PSP, they are not the same, at all.

Do you have an idea?
For you? RESEARCH!

You can't be anymore incorrect (no research on your part), because the Infectus MODCHIP (useless right now, really), can write the flash without using a JTAG.  I think they have basically given up on this for PS3 at this point.

Also, and even more simpler, and in plain sight, are the SYSTEM UPDATES!  There is no hardware connection there either, you download the update, and the NAND, and other FLASH roms are updated, like the BR-D flashrom.  Same with XBOX360, we just do not have the ability to do it as they are. How can you miss something like that?

If you replace the FLASH on the PS3, and write the first 16MB with the data from the old one with a programmer before installing it (which is all you can read without hitting a block) the rest of FLASH is updated with a system update, PS3 will work

You have done absolutely no research to base your statements off of, because is is 100% false.

Infectus wrote:

The infectus will allow you to dump and reflash your NAND (firmware chip). It will NOT allow you to play backups. If a custom firmware comes out one day, you can use the infectus to flash it to the system. Due to Sony suing companies that make and sell modchips for any Sony system, development of PS3 modchips is slow to nonexistant. Don't ask if/when there will be a PS3 modchip for backups.

Offline

 

#3  2009-04-29 11:48:06

OneSoul
PS3 Newbie In Training
Registered: 2009-04-25
Posts: 5

Re: JTAG: if exists, WHERE is it?

oopsss...

sorry for the "stupid" question... I have read about INFECTUS but some concepts were not clear for me...

so... because I see you VERY EXPERT, can you explain me, and to all the comunity, the exact architecture of the on board flash including boot sequence, encryption ecc?

the scope of my cuiosity is just technical...

please... can you forgive me? wink

PD: a question is not a crime...

Offline

 

#4  2009-04-29 13:46:18

0m1kr0n
PS3 Hacks Kung Fu Is On
From: NC
Registered: 2008-03-05
Posts: 231

Re: JTAG: if exists, WHERE is it?

The only non-open-market chips on all the PS3 board revisions are The PS2 emulator chip/ES-GS, RSX, Sony Southbridge(which I seen available from an asian wholesaler but I don't know what the deal was,) and the CELL/CPU. The rest are are NIC, HDMI,NAND,RAM, and a lot of small inverter/timer/signal controller ICs and analog conductors(<-assumption about all small ICs.) The unknown ICs can have the PS3 unsigned initialization code in non-volatile memory, and any of them can be unsigned and have a unsecured stack allocation. You also have some code storage on the BR-D and HDD ASICs that are unknown in security.

Actually documenting all this would be the most productive effort in reversing the PS3 thus far. The rest of the people know now that the higher abstraction layers on the ps3 are obscured by a signature chain and good cryptography. The infectus just dumps NAND, and it doesn't even completely do that from what I've read on the PS3 NAND blocks segments.


Actual Answer(?): I think it emulates multiple interfaces in firmware on the Actel ProAsic3 http://www.actel.com/products/pa3/

Last edited by 0m1kr0n (2009-04-29 14:06:03)


http://sigster.100webspace.net/siggy.png

Offline

 

#5  2009-04-29 15:13:15

Powerslave
Ruler of All
From: Alpha Quadrant
Registered: 2007-01-15
Posts: 12433
Website

Re: JTAG: if exists, WHERE is it?

Right, the Infectus can only read the 1st 16MB of the flash, BUT, as it turns out, that is all you need if you should have to replace the flash, and you have a backup of that 16MB.  This has been done; with a replaced FLASH, with the 16MB flashed into the bank.  Then, it must do a SYSTEM UPDATE, and the rest of the NAND is filled (flashed) (updated) with what was missing.

That is with the NEW firmware updates, you were able to DOWNGRADE firmware before they locked up the NAND with newer firmware.

Offline

 

#6  2009-04-29 18:58:42

0m1kr0n
PS3 Hacks Kung Fu Is On
From: NC
Registered: 2008-03-05
Posts: 231

Re: JTAG: if exists, WHERE is it?

Powerslave wrote:

Right, the Infectus can only read the 1st 16MB of the flash, BUT, as it turns out, that is all you need if you should have to replace the flash, and you have a backup of that 16MB.  This has been done; with a replaced FLASH, with the 16MB flashed into the bank.  Then, it must do a SYSTEM UPDATE, and the rest of the NAND is filled (flashed) (updated) with what was missing.

That is with the NEW firmware updates, you were able to DOWNGRADE firmware before they locked up the NAND with newer firmware.

I wonder what firmware did to lock it. I read about the old problem, but I forget the details; think it had something to do with signatures and corruption. Chip makers I think put bit flag segment lockers in chips for security too me thinks.

I wouldn't mind seeing a list of all the ICs on the board.


http://sigster.100webspace.net/siggy.png

Offline

 

#7  2009-04-29 21:55:55

Powerslave
Ruler of All
From: Alpha Quadrant
Registered: 2007-01-15
Posts: 12433
Website

Re: JTAG: if exists, WHERE is it?

0m1kr0n wrote:

Powerslave wrote:

Right, the Infectus can only read the 1st 16MB of the flash, BUT, as it turns out, that is all you need if you should have to replace the flash, and you have a backup of that 16MB.  This has been done; with a replaced FLASH, with the 16MB flashed into the bank.  Then, it must do a SYSTEM UPDATE, and the rest of the NAND is filled (flashed) (updated) with what was missing.

That is with the NEW firmware updates, you were able to DOWNGRADE firmware before they locked up the NAND with newer firmware.

I wonder what firmware did to lock it. I read about the old problem, but I forget the details; think it had something to do with signatures and corruption. Chip makers I think put bit flag segment lockers in chips for security too me thinks.

I wouldn't mind seeing a list of all the ICs on the board.

I have it here in the forums: http://www.ps3-hacks.com/forums/viewtop … 015#p53015

Was all I could find...

Offline

 

#8  2009-04-29 21:59:21

Mick
Harsesus
From: Canada
Registered: 2007-01-31
Posts: 3334
Website

Re: JTAG: if exists, WHERE is it?

0m1kr0n wrote:

I wouldn't mind seeing a list of all the ICs on the board.

POST 13 wrote:

PS3 MAINBOARD HARDWARE - Originally by: 0m1kr0n - Updated by Powerslave

Updated 7/6/08

Items for the initial 20/60GB USA and JAP models are indicated as (USA/Jap Hardware Emu). Those parts are not in the EU/PAL versions, or the 80GB & 40GB models.  They are either removed, and/or replaced with another chip, and relocated.

Back of Mainboard:
- Samsung K9F1G08U0A is 1GB SLC NAND Flash
- NEC / TOKIN 0E128 is a Proadlizer - High Speed Decoupler

Top Side of Mainboard:
- Marvell 88E6108-LAR1 is an Ethernet Controller
- Silicon Image Sil9132CBU is HDMI Transmitter (output)
- Sony CXM4024R is only known as ASIC B outside Sony (USA/JAP Hardware Emu)

    Sony Computer and Entertainment Inc., CXM4024R, � & (M), 2005SCEI, 628D30V

- Sony CXD9280GP - ?  Next to ASIC B Chip
- Marvell 88SA8040-TBC1 is a SATA controller
- Sony CXD2973GB (Above CELL) "0611HAL" Hardware Abstraction Layer (USA/JAP Hardware Emu)
- Sony CXD2979GB (Above CELL) "0629HAL" Hardware Abstraction Layer (Software Emu Boards)
- Sony CXD2973GB (Left of RSX) Microcontroller - Part of Southbridge  (Software Emu Boards)
- Sony CXD4302GB is I/O Bridge Controller
- Samsung K4Y50164UC-JCB3 is 256MB of XDRTM DRAM

    Specs on ram say XDRTM DRAM 512Mbit XDR TM DRAM(C-die) - You need eight chips for 512MB total, and PS3 only has 512 total, 256/256 split.  The 256MB video ram is built into the RSX chip.

- Sony CXD2964GB Cell BBE/Core Processor.
- NEC/TOKIN 0E128 is a Proadlizer - High Speed Decoupler
- Sony CXD2971GB is the RSX GPU (w/256MB GDDR3 VRAM)
- Samsung K4J52324QC-SC14 256MB GDDR RAM (on GPU)

    Specs on ram say 512Mbit GDDR VRAM, same notation as above with XDR Ram.  It is false, there is only 256MB for CPU and 256MB for GPU.

- Sony CXR713120 - PS1 Backward Compatibility Related (USA/JAP Hardware Emu) - known as ASIC A. Here's a quote from an Asian parts site:

    Sony Computer and Entertainment Inc., CXR713120, � & (M) 2006SCEI,-201GB, 640A69W

- Sony CXD2953AGB is the Emotion Engine & Graphics Synthesizer Chip (USA/JAP Hardware Emu)
- Samsung K9F1G08U0A is 1GB SLC NAND Flash
- Toshiba (MFG Not printed on chip) CXD9208GP PS2 Legacy ASIC (NTSC/JAP Hardware Emu)
- ?MFG? 32 MB RDRAM "Rambus" (USA/JAP Hardware Emu)
- Genesys Logic GL852 - USB2.0 HUB CONTROLLER w/8BIT RISC Processor

Blu-Ray Board:
- Sony CXD5064R is an audio Decoder
- Spansion S99-50111-001 is 16MB Flash memory
- Sony CXA2720R is a Blu-Ray interface controller
- Sony CXD5063GG-1 ASIC / CPU - Video Decryption Chip
- Samsung K4S641632K-UC75 64MB SDRAM
- S!PWM is a timer chip - Pulse Width Modulator (controls laser and spindle) Has EEPROM and RAM

Wifi Board
- SCEI&SCR "D3261GG" (Bluetooth transceiver?)
- Spansion S99AL008D002 NOR FLASH 8MB 3.3V
- Marvell 88W8010-NNB1 Optimized RF-baseband transceiver, integrates at near 20 dBm
- ISSI IS42S32400B 128MB SRAM
- Marvell 88W8580-BAN1 802.11g 56Mbps WLAN

Controller Board Front
- SCEI&SCR "D3261GG" (Bluetooth transceiver?)
- 325A (?)

Controller Board Back
- Toshiba T6UM2EFG-0103 (Gyroscope/Motion Sensor? Centered)

UNUSED PS3 MOTHERBOARD CHIPSET PORTS:
3x TSIF Digital Video In
Analog Video In
Analog Audio In
3x Video Out (HDMI,D4,A-DAC)
IEEE1394 Firewire
1x PATA (IDE)

KEY TO TERMS:
HAL = Hardware Abstraction Layer: Binary code that allows a portion of the operating system that lets programs deal with hardware directly. This allows programs needing more speed from the computer to bypass the standard OS calls to hardware.  The PS3 Hypervisor prevents this access needed for unlocking the full potential under Linux or with hacking the PS3.

ASIC = Application-Specific Integrated Circuit: An IC that is customized for a particular use, rather than intended for general-purpose use. For example, a chip designed solely to run on the PS3 is an ASIC.  In the case of the PS3, they are for the hardware PS2/1 emulation on the USA/JAP models with the EE+GS.

PROADLIZER = This device easily resolves two opposing needs: higher speeds and performance, and weight reduction and miniaturization.  This device replaces numerous capacitors used for decoupling circuits such as high-frequency resonant ceramic capacitors, large-capacity ceramic capacitors, and alminum (not aluminum) electrolytic capacitors. It was developed for CPU decoupling circuits, for PCs and servers, which have serious decoupling circuit problems.

NOR FLASH = http://en.wikipedia.org/wiki/Flash_memory#NOR_flash

NAND FLASH = http://en.wikipedia.org/wiki/Flash_memory#NAND_flash

SLC (FLASH) = Single-Level Cell: SLC NAND Flash�s control logic does a better job conserving energy than MLC (Multi-Layer Cell). This is primarily because the device only needs to manage the electrical charge for two states and one bit of stored data. As a result, SLC architectures offer a significantly greater cycle endurance as compared to MLC.

http://www.ps3-hacks.com/forums/viewtopic.php?id=1423

Offline

 

#9  2009-04-29 22:27:24

0m1kr0n
PS3 Hacks Kung Fu Is On
From: NC
Registered: 2008-03-05
Posts: 231

Re: JTAG: if exists, WHERE is it?

I don't think I got any of the small ICs, and didn't even look at the ones on the HDD and BD-R drive. I thought most of the small ones where just timers, gates, inverters, opamps etc..

Also in one of my other posts I remember mentioning that on the 40GB boards they actually obscured the switch traces on a middle PCB layer to keep people from mapping the initialization order. I don't know if all the other boards had that because I only had a 40GB board, and some high res angled imagery. That's usually the easiest way to reverse engineer embedded systems though.

The thing everyone is trying to do is trying to find the initialization code that in theory will be unsigned instructions, or some new hardware non-static cryptographic logic preferably on a die but detectable(unlikely,) or unsigned code that can get pass allocator security like the x360 DMA drive exploit. These are the only ways to do that in practicality because some form of memory corruption on the higher levels would most likely be blocked whether it's heap or stack variant, and even then you have a relativily secure chain of cryptographically signed data below that, and that's most likely got real time signature checking.

In theory a dynamic die-based logic system that could start a sig chain, and bus logic all signed and/or pseudo-random+locking allocator processed would make a embedded system relatively uncrackable without heavy cryptanalysis, which in this case is SHA1 and AES supposedly, so it'd be side channel attacks or nothing at all till some future technology comes along.

Last edited by 0m1kr0n (2009-04-29 22:49:11)


http://sigster.100webspace.net/siggy.png

Offline

 

#10  2009-04-29 22:44:45

Powerslave
Ruler of All
From: Alpha Quadrant
Registered: 2007-01-15
Posts: 12433
Website

Re: JTAG: if exists, WHERE is it?

0m1kr0n wrote:

I don't think I got any of the small ICs, and didn't even look at the ones on the HDD and BD-R drive. I thought most of the small ones where just timers, gates, inverters, opamps etc..

Also in one of my other posts I remember mentioning that on the 40GB boards they actually obscured the switch traces on a middle PCB layer to keep people from mapping the initialization order. I don't know if all the other boards had that because I only had a 40GB board, and some high res angled imagery. That's usually the easiest way to reverse engineer embedded systems though.

The ones on the HDD should be irrelevant, since you can upgrade to any 2.5" Sata drive.

The thing with hiding traces in a middle layer, is; a PAD will be somewhere on the motherboard to connect it to the upper or lower layer.  No trace can be completely hidden in middle layers, as they HAVE TO tie in somewhere on the bottom or top layer.

Offline

 

#11  2009-04-29 23:00:18

0m1kr0n
PS3 Hacks Kung Fu Is On
From: NC
Registered: 2008-03-05
Posts: 231

Re: JTAG: if exists, WHERE is it?

Powerslave wrote:

0m1kr0n wrote:

I don't think I got any of the small ICs, and didn't even look at the ones on the HDD and BD-R drive. I thought most of the small ones where just timers, gates, inverters, opamps etc..

Also in one of my other posts I remember mentioning that on the 40GB boards they actually obscured the switch traces on a middle PCB layer to keep people from mapping the initialization order. I don't know if all the other boards had that because I only had a 40GB board, and some high res angled imagery. That's usually the easiest way to reverse engineer embedded systems though.

The ones on the HDD should be irrelevant, since you can upgrade to any 2.5" Sata drive.

The thing with hiding traces in a middle layer, is; a PAD will be somewhere on the motherboard to connect it to the upper or lower layer.  No trace can be completely hidden in middle layers, as they HAVE TO tie in somewhere on the bottom or top layer.

Yeah but that could be any of close to ~1,400 surface mount pads or through holes. It'd be a nightmare doing signal testing even with good equipment, and acid stripping the PCB layers might not work either.

The ones in the HDD might still be relevent. I guess it depends on what type of data a SATA controller can return to the host that controls branching. It goes through the Sony southbridge I think. There should be an ASIC on the drive that has a buffer role and firmware that's between all the motor controlling and laser decoding and the SATA interface in the Sony southbridge.

Last edited by 0m1kr0n (2009-04-29 23:03:01)


http://sigster.100webspace.net/siggy.png

Offline

 

#12  2009-04-29 23:22:19

Powerslave
Ruler of All
From: Alpha Quadrant
Registered: 2007-01-15
Posts: 12433
Website

Re: JTAG: if exists, WHERE is it?

You mean BR-D, not HDD, as you are referencing motor controlling, but more to the point; and laser decoding and the SATA interface in the Sony southbridge.

The HDD is a regular 2.5" HDD, nothing like what M$ did to theirs.

This is why the data on the Game partition is encrypted, so you can upgrade with any 2.5" HDD.  However, once you format a new one in the PS3, the previous one will no longer work in the PS3.

Offline

 

#13  2009-04-30 04:16:29

OneSoul
PS3 Newbie In Training
Registered: 2009-04-25
Posts: 5

Re: JTAG: if exists, WHERE is it?

so... the TWO flash on board work like a BIOS + mini-OS mixed together... similar to LinuxBIOS project...
probabily the FIRST 16MB is just del "BIOS", and the remaining the OS...
ALL the code is encrypted, or the first 16MB are in clear?

Take a look at this:
http://www.ps3-hacks.com/forums/viewtopic.php?id=7655

If the idea in the other topic could be good... If someone achieves to built a custom FW...
is it possible to put in the ps3? or the console rejec ANY type of different code?

Offline

 

#14  2009-04-30 06:04:51

The_Wii_Nes_Boy
Testing the Stargate
From: Scotland
Registered: 2008-07-02
Posts: 357

Re: JTAG: if exists, WHERE is it?

You should try reading whats been wrote a bit better, there is only ONE flash of which ONLY the first 16Mb can be read.


http://i299.photobucket.com/albums/mm311/The_Wii_Nes_Boy/Sigs/TEKKEN-1.gif
Games I'm definitely gonna buy this month -
GOT: Uncharted 2: Among Thieves / STILL TO GET: Ratchet & Clank: A Crack In Time and Tekken 6

Offline

 

#15  2009-04-30 09:59:26

0m1kr0n
PS3 Hacks Kung Fu Is On
From: NC
Registered: 2008-03-05
Posts: 231

Re: JTAG: if exists, WHERE is it?

Powerslave wrote:

You mean BR-D, not HDD, as you are referencing motor controlling, but more to the point; and laser decoding and the SATA interface in the Sony southbridge.

The HDD is a regular 2.5" HDD, nothing like what M$ did to theirs.

This is why the data on the Game partition is encrypted, so you can upgrade with any 2.5" HDD.  However, once you format a new one in the PS3, the previous one will no longer work in the PS3.

Yeah I know the hard drive isn't optical I just didn't differentiate in that post. Still though the HDD and BR-D both have controller boards that have chips for formatting bit streams and buffering. http://www.hardwaresecrets.com/article/177/3

The only thing I was trying to point out is that the firmware on all the HDD on board chips returns some kind of data into a region of RAM on the ps3 that's most likely DMA and maybe not on a locked page. I don't know what the HDD controller returns to the south bridge on that layer though.

The Sony bridge could also have some type of security filtering too, but who knows. I think that same chip is used in other stuff besides the PS3. It shows up a lot on asian wholesale sites, but I don't know anything about who they're going to or whatever.

Also they might salt a crypto key with a hardware ID or something on the HDD, and that might go into a partition table somewhere. I don't have a clue about that though.

Last edited by 0m1kr0n (2009-04-30 10:04:36)


http://sigster.100webspace.net/siggy.png

Offline

 

#16  2009-04-30 14:36:44

Powerslave
Ruler of All
From: Alpha Quadrant
Registered: 2007-01-15
Posts: 12433
Website

Re: JTAG: if exists, WHERE is it?

Oh, you mean the chipsets on the PS3 motherboard, ok, I thought you meant ON the HDD itself, it wasn't all that clear, to me anyway...  There is nothing special about 2.5" hard drives, any one will work, so there isn't anything special on the HDD logic boards itself, else you would have to use Sony approved HDDs.  There is no security LOCK either, like XB1 did, they are open, just game O/S is encrypted.

Offline

 

#17  2009-04-30 21:36:47

0m1kr0n
PS3 Hacks Kung Fu Is On
From: NC
Registered: 2008-03-05
Posts: 231

Re: JTAG: if exists, WHERE is it?

Powerslave wrote:

Oh, you mean the chipsets on the PS3 motherboard, ok, I thought you meant ON the HDD itself, it wasn't all that clear, to me anyway...  There is nothing special about 2.5" hard drives, any one will work, so there isn't anything special on the HDD logic boards itself, else you would have to use Sony approved HDDs.  There is no security LOCK either, like XB1 did, they are open, just game O/S is encrypted.

No I meant the all of the above. The drives use IRQ on the PS3 too I'm pretty sure and the controller board buffer I think pushes some of itself on SDRAM in the south bridge which processes it somehow and then pushes it on main RAM for the NAND firmware to fetch from a segment defined in the map of the firmware hierarchy.

Anyway the IRQ triggers signals in the NAND firmware from H.A.L or whatever layer abstraction they use and whatever initialized a reading fetches the buffer from RAM. I know all the drives are standardized interfaces, but they all have their own vendor and hardware ID in the controller chip that could be used to salt keys for the HDD image security you mentioned.

Last edited by 0m1kr0n (2009-04-30 21:37:42)


http://sigster.100webspace.net/siggy.png

Offline

 
Home | PS3 News | PS3 Hacks | PS3 Downloads | PS3 Saves

Folding@home | PS3-Hacks Live Chat | PS3 Reviews | Contact Us


Board footer

Powered by PunBB
© Copyright 2002–2008 PunBB