PS3 Ps3Xploit - Now Install CFW from 4.82 OFW: NOR/NAND Software Flash Writer + NOR Dumper Released

UPDATE (Jan 2017): v2.0 update has been released
IMPORTANT UPDATE (MARCH-2019) v3.0.1 was released providing support for 4.84 HFW (a new hybrid firmware that restores PS3Xploit for 4.84 OFW user's (See Link to HFW))
.There has been alot of development going on around the PS3 since the discoveries of the PS3Xploit (webkit exploit) on 4.81 OFW, first we seen the IDPS dumper (4.81/2 OFW) released with some big news & hope to come along with it, like a Flash Writer (downgrader) for OFW, so if you have been reluctant to buy a Hardware Flasher such as the E3 Flasher and bust open your PS3, but have been wanting to get your PS3 FAT(PHAT) Console and your Downgradable PS3 SLIM Models (up to & including 25xx models with minimum installable version <= 3.56) on Custom Firmware, then here is your chance with a 100% SOFTWARE SOLUTION thanks to the work of PS3Xploit Team ( @bguerville, @esc0rtd3w & W) along with contributions from new team member @habib to help expedite this release. Essentially what this Software Solution does is write a patch to the CoreOS (on NOR/NAND Chip) and when the PS3 Console is then rebooted you can install a Custom Firmware directly, So downgrading back to 3.55 is not required in the process, rather "Direct OFW to CFW patching" is done to allow for Custom Firmware Installation. Since this exploit is executed from 4.82 OFW, you can only install to a 4.82 CFW, HOWEVER if you wish to use an earlier firmware such as REBUG 4.81 for example, once on 4.82 CFW you must TOGGLE QA using a toggle tool, which allows CFW user's to freely switch CFW version from past and present. Read more about this in the Frequently Asked Question (FAQ) and more info in the details provided:


UPDATE - View Public Warning (Beware of Imposter's)
BfEBwXb.jpg



Flash Writer Compatible with these PS3 Models:
Supports FAT Models CECHAxx/Bxx/Cxx/Exx/Gxx/Hxx/Jxx/Kxx/Lxx/Mxx/Pxx/Qxx
Supports SLIM Models 2xxx (minver 3.56 or lower ONLY, check with >>> minverchk.pup - SEE FAQ TAB for USAGE )
​
4.82 NAND / NOR WRITER NOR FLASH DUMPER F.A.Q & Additional Notes.

  • PS3 OFW 4.82 NAND/NOR FLASH WRITER v1.0
    ***** IMPORTANT DETAILS BELOW -- AVOIDING A BRICK *****
    ​
    WARNING: USE THE PROVIDED flsh.hex AS IS. DON'T PATCH IT OR MODIFY IT OR YOU WILL BRICK *****
    • Verify flsh.hex file on a flash drive and in the far right USB slot!
      • 4.82 flsh.hex MD5: 8E156C99101BF36EC3EDB832982AE46D
    • DO NOT USE ON CFW (Custom Firmware) (Only Supports OFW)
    • DO NOT USE ON PS3 Models 3xxx/4xxx (aka SuperSlims / Late Slim models) you will brick those console.
    • USE ONLY ON 4.82 OFW


    PLEASE READ FIRST:
    • It's essential not to flood the browser memory with junk before running the exploit. The reason for this is that due to javascript core memory usage limitations we are scanning several times a small range of browser memory (a few Mb) to find some essential data in RAM, if the memory is flooded then the range to scan becomes much larger & the probabilities that our data is found in the smaller range decrease dramatically..
    • So in short, never use the browser or set a homepage you cancel before running the exploit!
    • If you need to, set the homepage to 'blank', close the browser then reopen it to start the flash writer.

    v1.0.0 - Initial Release
    • Supports Direct OFW to CFW patching for All Phat and 2xxx Slim (minver 3.56 Dec 2010 and lower)
    • the NOR/NAND writer will just copy 3Mb of CoreOS data to both ros0 & ros1 in the flash memory.
    • There is only one version released for 4.82. The same hex patch file can be used on nor & nand.
    • It's as safe as possible, with a check for usb device & patch file making the exploit hang instead of corrupting flash if file is not found.
    • In case of corruption (extremely rare but could always happen), it's only a partial brick because no per console info ever gets erased so a hardware flasher could still be used if ever a recovery reboot was impossible.
    ​
    Usage Tips:
    1) Try using a LAN connection or a solid WiFi connection during exploitation. A weak signal can cause problems.
    2) If the exploit takes more than 5 minutes to work, reload page, browser, or restart console and try again.
    3) If you are using a LAN connection and experience network issues, make sure all cables to router are in working order.​
    ​

    Steps:
    1. Setup a small Web server on pc or smartphone. A custom miniweb application (from: https://sourceforge.net/projects/miniweb/files/) with small changes to the JaveScript, and supplied to host files if you would like to use it. Don't come to us for explanations about how to run a http server though. Google it.

    2. Extract the files from release to your http server root folder.
    2a- To use the miniweb.exe server, it is necessary to create a folder: htdocs
    2b- The files *.html and *.js included in the zip files should be copied/moved to htdocs    ​
    3. Copy the "flsh.hex" file from release folder to root of flash drive.

    4. Put a FAT32 USB key in port closest to BD Drive (/dev_usb000).

    5. DOUBLE-CHECK your flash drive on XMB to make sure it shows up under Music, Photos, Videos, etc.

    6. Open the PS3 browser File Address window, write the IP address of your server (and the port if not 80) & press the Start button.

    7. Select the appropriate button for your console and wait for PS3 to power down. DO NOT STOP THE PROCESS ONCE STARTED!!

    8. Once PS3 has powered down, reboot console and install CFW matching OFW version. If installing through XMB does not work, boot to recovery and install.​

  • PS3 4.81/4.82 NAND/NOR Flash Dumper v1.0
    THE CORRECT FIRMWARE VERSION BETWEEN 4.81 and 4.82 IS AUTOMATICALLY SELECTED!

    UPDATE - ALSO Ported to earlier Official Firmware     >>> Link    ​

    PLEASE READ FIRST:
    • It's essential not to flood the browser memory with junk before running the exploit. The reason for this is that due to javascript core memory usage limitations we are scanning several times a small range of browser memory (a few Mb) to find some essential data in RAM, if the memory is flooded then the range to scan becomes much larger & the probabilities that our data is found in the smaller range decrease dramatically....
    • So in short, never use the browser or set a homepage you cancel before running the exploit! If you need to, set the homepage to 'blank', close the browser then reopen it to start the flash writer.

    v1.0.0 - Initial Release.
    • Supports Dumping NOR on both 4.81 & 4.82.
    • bguerville tried to produce a release that was easy to port & he succeeded. Anyone able to search for offsets in IDA can add support to any firmware version in the dumper in a matter of minutes.
    • For technical reasons, the Full NAND dumper release is postponed. We will now be focusing on self execution & if we succeed there will be no need for the extra ROP work to do the NAND dumper. If we fail, I will finish it in ROP.
    • A lot of time has been invested into making the javascript + UI more efficient, as well as the trigger phase faster & more stable. I hope you enjoy the result.
    Usage Tips:
    1) Try using a LAN connection or a solid WiFi connection during exploitation. A weak signal can cause problems.
    2) If the exploit takes more than 5 minutes to work, reload page, browser, or restart console and try again.
    3) If you are using a LAN connection and experience network issues, make sure all cables to router are in working order.​

    Steps:
    1. Setup a small Web server on pc or smartphone. A custom miniweb application (from: https://sourceforge.net/projects/miniweb/files/) with small changes to the JaveScript, and supplied to host files if you would like to use it. Don't come to us for explanations about how to run a http server though. Google it.

    2. Extract the files from release to your http server root folder.
    2a- To use the miniweb.exe server, it is necessary to create a folder: htdocs
    2b- The files *.html and *.js included in the zip files should be copied/moved to htdocs    ​
    3. Put a FAT32 USB key in port closest to BD Drive (/dev_usb000).

    4. DOUBLE-CHECK your flash drive on XMB to make sure it shows up under Music, Photos, Videos, etc.

    5. Open the PS3 browser File Address window, write the IP address of your server (and the port if not 80) & press the Start button.

    6. The dumper will detect the firmware version of your console automatically & setup the code appropriately so there is only one version for both 4.81 & 4.82. Run until ps3 beeps & shutdown. The flash dump should be a 16MB file on your USB drive as dump.hex.​

  • Frequently Asked Questions & Additional Notes​


    Any Noob Friendly Guides for OFW to CFW? APRROVED BY TEAM PS3Xploit

    Will this jailbreak my SuperSlim?
    • NO, (PS3Xploit has strong possibility to eventually evolve into a HEN style exploit (that aspect will take some additional development and time, at this time PS3Xploit exploit has not evolved enough)


    Is my Model Compatible & Which Writer does my console need (NOR or NAND)?
    • supported.png

    Can I install a CFW before 4.82, such as Rebug 4.81 or an earlier CFW?
    • Yes, however you must Toggle QA Flag. Once the Token is activated you have the ability to then freely jump CFW versions. (see below for details)

    How do I Toggle QA Flag?
    • When on a CFW download & install >>> QA TOGGLER (Standalone), (Note: Will just show a black screen then reboot the PS3 and returns to xmb. A Restart is Required. After toggling QA, cfw syscalls will be disabled (meaning your CFW patches will be disabled until the next boot, so a reboot is required after the Toggler exits back to the XMB.) Additional Info about Q/A flag can be seen here ( & also @ PS3Devwiki)

    Should i use the "999 Downgrader" vs "Toggle QA" to install a different CFW?
    • No, installing the "999 downgrader" PUP can cause various issues like on a 3.56 minver console, it will brick the console, simply just Toggle the Q/A Flag and play it safe and is so simple to move from CFW versions (up and down from version to version).

    How do i know for sure if my PS3 Model is compatible ?
    • You must have a PS3 Console that has a Factory Firmware of 3.56 and below.
    • To check, its easy with this simple tool for OFW, Download minverchk.rar
    • Then place the .pup file on a FAT32 USB Flash Drive in a PS3/UPDATE folder (create path if needed)
    • Now on the PS3 XMB goto Setting ->> System Update >>> Update via Media Storage
    • Once shown on the list select the PUP and install, shortly after there will be a message showing the factory firmware the console was shipped,
    • For this we want 3.56 and below.
    • ANYTHING HIGHER THEN 3.56 IS NOT ABLE INSTALL A CFW. Sorry this will not work for your console, but there could be a HEN (Homebrew Enabler) possible for running homebrew, but additional research and time is needed for achievement, additional details can be read here.

    What is the basic purpose of the Writer & Dumper Tools Release?
    • The dumper is to get a backup of the nor chip
    • The writer is to jailbreak your console. (Adding a patch to OFW to allow CFW installation)

    Do I have to setup my own web server or can (has) someone host this?
    • For best results and security it's advised/recommended to setup a local web-server to execute the webkit exploit, The best Unofficial Host we have found is from developer RED and his page: http://redthetrainer.com/ps3/

    How to go from Ferrox 4.82 to Rebug 4.81?
    • .Question Raised Here , OR alternatively you can uses this UNOFFICIAL modified version of REBUG 4.81.2, that will install on 4.82 (without QA FLAG as it contains an edit to the syscon version) >>>> (View Tweet & Download Link)

    Where can i find the latest 4.82 CFW?

    Where can i find PS3 Homebrew?

    New to PS3 CFW Community (Have CFW now installed and want to know a bit more)?
    • Here is a thread being started in the forums, An Intro to CFW & PS3 Homebrew . Covers various basic of firmware types and some essential apps. The thread is a WIP so expect additional items to be added.

    PSX-Place Member @lord3490 provided some extra FAQs, to which the PS3Xploit team also added information.
    ​
    Q: my console just froze and nothing's happening for over 10 minutes
    • A: turn off console, the exploit failed.
      • 1. Make sure you got the correct file on USB thumb drive and it's formatted to fat32.
      • 2. Use an ethernet cable connection instead of WiFi.
      • 3. Double check your local Web server configuration & logs to see if it is not simply failing to send files to the ps3. A http server could fail for many reasons, bad configuration, bad permissions, Windows firewall options, AV software, router settings etc....
      • 4. Clean browser cache
      • 5. Set ps3 browser homepage to about:blank (or to your exploit Web page)
      • 6. Restart Browser
      • 7. Try again

    Q: the console shut down and beeped when using the exploit, however I'm getting an error when trying to install cfw?
    • A: there are a couple of possible reasons for that:
      • 1. Did you make sure you flashed the correct file (nand/nor)? See q/a above
      • 2. Try different 4.82 CFWs and make sure the md5 is correct after copying to fat32 thumb drive.
      • 3. Try a different USB thumb drive or reformat it.
      • 4. Install OFW 4.82 two times in a row then apply the patch using the PS3Xploit flash writer & finally install a 4.82 CFW. That should always solve the issue.
      • 5. As a last resort, make a backup and format internal HDD (I just read that solved the problem for one user).

    Q: when will there be a CFW or a HEN type of hack for newer ps3 models?
    • CFW (Custom Firmware): Not Possible
    • HEN (Homebrew Enabler): you may be able to use homebrew (even backup managers) later on. The devs are working on it and they won't be faster or release it earlier because you ask. The PS Vita / PlayStation TV use a HEN exploit (HENkaku) to give you a idea on what a HEN is for those of you who have followed the Vita Scene.
    • Keep dropping by this forum and you won't miss it once it's there. You will here it first from psx-place.com the official home of the PS3Xploit Team

    Q: okay, I got a cfw installed. What do I do now?
    • A: Read. There is a lot of information on this forum. Use search function for specific topics and check out this thread >> An Intro to CFW & PS3 Homebrew to get started.

    Q: where can I download games?
    • A: From PSN, for anything else you may want to read the forums rules! psx-place.com
    --------------------------------------------------------​
    ​

Downloads:

dIMPORTANT UPDATE (MARCH-2019) v3.0.1 was released providing support for 4.84 HFW (a new hybrid firmware that restores PS3Xploit for 4.84 OFW user's (See Link to HFW)
d
Courtesy of Team PS3Xploit:
W (Javascript, Research & Testing)
[COLOR=#003366]@esc0rtd3w[/COLOR] (Debugging, Research & Testing)
[COLOR=#003366]@habib[/COLOR] (ROP & Debugging)
[COLOR=#003366]@bguerville[/COLOR] (ROP/Javascript & Debugging)​
 
Last edited:
Click to expand...
bguerville
nCadeRegal said:
Are you guys still working on the nand dumper? @bguerville @habib
Click to expand...
Not exactly...
The size of nand makes it a problem.
Each flash memory read is 2Mb max (limited by system) so, for instance, the nor dumper repeats the read operation 8 times with different parameters to get 16Mb in total. Originally I was hoping to use a sys_storage_read loop gadget, unfortunately there are only 4 available in system code, 1 applicable to the situation but unusable due to one tiny param hard-coded in it so using a loop is not feasible at this stage.

That would mean repeating the read operation 128 times to dump the full nand to memory. As one read uses 3 gadgets (consider a gadget as a portion of system code), that means repeating 3 gadgets in ROP 128 times which is a pita.
Furthermore it gets even more annoying because of the available RAM. Obviously it's impossible to dump the whole nand to RAM because there is less than 256 Mb available to us.. So that would mean having to dump 128Mb then save to file then do the next 128Mb & append to file.

Finally there is one last detail.
We have noticed that on NOR, reads bigger than 1Mb can fail on occasion, it explains that some dumps are full of of 00s as some users reported. I have therefore changed the read size to 1Mb, for 16 operations on NOR, that update will be released soon...
If the issue were the same on nand we would be talking about repeating 3 gadgets 256 times to read the entire nand without getting read fails.

So obviously you must understand why this has not been done yet. It's possible but totally unpractical to do in ROP.
Depending on our research & findings, a full nand dumper will eventually be provided, at a later stage, in one form or another.

There is an idps dumper available on nand & emmc so in practice the dumpers are not really essential. The flash writer cannot brick fully your console & any partial brick can be recovered from by patching the corrupted ros regions with 3Mb from ofw CoreOS.
 
Last edited:
K
bguerville said:
Not exactly...
The size of nand makes it a problem.
Each flash memory read is 2Mb max (limited by system) so, for instance, the nor dumper repeats the read operation 8 times with different parameters to get 16Mb in total. Originally I was hoping to use a sys_storage_read loop gadget, unfortunately there are only 4 available in system code, 1 applicable to the situation but unusable due to one tiny param hard-coded in it so using a loop is not feasible at this stage.

That would mean repeating the read operation 128 times to dump the full nand to memory. As one read uses 3 gadgets (consider a gadget as a portion of system code), that means repeating 3 gadgets in ROP 128 times which is a pita.
Furthermore it gets even more annoying because of the available RAM. Obviously it's impossible to dump the whole nand to RAM because there is less than 256 Mb available to us.. So that would mean having to dump 128Mb then save to file then do the next 128Mb & append to file.

Finally there is one last detail.
We have noticed that on NOR, reads bigger than 1Mb can fail on occasion, it explains that some dumps are full of of 00s as some users reported. I have therefore changed the read size to 1Mb, for 16 operations on NOR, that update will be released soon...
If the issue were the same on nand we would be talking about repeating 3 gadgets 256 times to read the entire nand without getting read fails.

So obviously you must understand why this has not been done yet. It's possible but totally unpractical to do in ROP.
Depending on our research & findings, a full nand dumper will eventually be provided, at a later stage, in one form or another.

There is an idps dumper available on nand & emmc so in practice the dumpers are not really essential. The flash writer cannot brick fully your console & any partial brick can be recovered from by patching the corrupted ros regions with 3Mb from ofw CoreOS.
Click to expand...
Is the emmc the same situation as nand ?
 
S
k9mo said:
Is the emmc the same situation as nand ?
Click to expand...

Consider eMMC as a modified NAND to add hard disk support.

So if they cannot do it on NAND due to memory failure or gadget repetition problem, well they need to adapt it on eMMC, so it will have the same problem.

Consider it now as a second step. So if NAND is not done, eMMC won't be done for a long time.
 
K
bguerville said:
Well obviously if we are looking at dumping 12Gb of emmc, ROP isn't the best way to do it.
If push comes to shove it can be done on nand but emmc dumping would remain problematic until unsigned code can be run.
Click to expand...
thanks for explainning
One question though are you guys still working on the usb writer to hdd and if not could a person with good knowledge of ROP chaining be able to achieve it?
 
J
I've tried the NOR writer more than 10 times now, waiting up to 25 minutes twice, and each time after getting the SUCCESS message the ps3 fails to turn off on its own. I formatted the thumb drive, copied the flsh.hex file to the thumb drive and checked its' md5. Cleared cache and cookies from web browser, set the home page (I tried using blank as the homepage and using the nor html as the homepage), exited browser and opened it again before trying the exploit. I've tried two different thumb drives (an older 4gb and a newer 16gb one), tried using Red's website and tried hosting the site on my local network. The NOR dumper worked perfectly fine multiple times.

Can anyone let me know what I might be doing wrong? I'd really appreciate some help.

Edit: I should add that I'm trying this on CECH2503B with a minver of 3.50
 
habib
jumpychampak said:
I've tried the NOR writer more than 10 times now, waiting up to 25 minutes twice, and each time after getting the SUCCESS message the ps3 fails to turn off on its own. I formatted the thumb drive, copied the flsh.hex file to the thumb drive and checked its' md5. Cleared cache and cookies from web browser, set the home page (I tried using blank as the homepage and using the nor html as the homepage), exited browser and opened it again before trying the exploit. I've tried two different thumb drives (an older 4gb and a newer 16gb one), tried using Red's website and tried hosting the site on my local network. The NOR dumper worked perfectly fine multiple times.

Can anyone let me know what I might be doing wrong? I'd really appreciate some help.

Edit: I should add that I'm trying this on CECH2503B with a minver of 3.50
Click to expand...
Move your homepage directly to the exploit site.
Clear browser data
 
sandungas
jumpychampak said:
I've tried the NOR writer more than 10 times now, waiting up to 25 minutes twice, and each time after getting the SUCCESS message the ps3 fails to turn off on its own. I formatted the thumb drive, copied the flsh.hex file to the thumb drive and checked its' md5. Cleared cache and cookies from web browser, set the home page (I tried using blank as the homepage and using the nor html as the homepage), exited browser and opened it again before trying the exploit. I've tried two different thumb drives (an older 4gb and a newer 16gb one), tried using Red's website and tried hosting the site on my local network. The NOR dumper worked perfectly fine multiple times.

Can anyone let me know what I might be doing wrong? I'd really appreciate some help.

Edit: I should add that I'm trying this on CECH2503B with a minver of 3.50
Click to expand...
The dumpers are nice for pre-testing and seting up the enviroment, but have you validated them ?, it seems there has been reports of dumps with "empty" sections, maybe thats caused by problematic USB sticks

One thing i mentioned before (as a paranoid meassure to know if the USB is compatible) is to install an official firmware from it (if you are on 4.82 OFW... you can install 4.82 OFW on top)... if that works the USB should be fine... i guess
There has been people that had to try several different USB sticks until they found one working
 
sandungas
k9mo said:
thanks for explainning
One question though are you guys still working on the usb writer to hdd and if not could a person with good knowledge of ROP chaining be able to achieve it?
Click to expand...
In plain words... the unofficial running code triggered by the webkit exploit is "trapped" in the web browser ram area (dedicated to javascript iirc ?)
Inside that enviroment everything needs to be made with ROP and is very restrictive and small
So there are two development roads they had to decide to choose... one is to try to build other functions with ROP that requires lot of work and the complexity and size could take them to a road end and waste time in something that is not going to work as well as expected... and the other development road is to try to "escape" from the web browser area

They are focusing the efforts in the second road because is the good one. If at some point they achieves it there will be more privileges, more ram to use, and other functions will be easyer to build (using real compiled custom code, instead of ROP recycled functions)
 
K
sandungas said:
In plain words... the unofficial running code triggered by the webkit exploit is "trapped" in the web browser ram area (dedicated to javascript iirc ?)
Inside that enviroment everything needs to be made with ROP and is very restrictive and small
So there are two development roads they had to decide to choose... one is to try to build other functions with ROP that requires lot of work and the complexity and size could take them to a road end and waste time in something that is not going to work as well as expected... and the other development road is to try to "escape" from the web browser area

They are focusing the efforts in the second road because is the good one. If at some point they achieves it there will be more privileges, more ram to use, and other functions will be easyer to build (using real compiled custom code, instead of ROP recycled functions)
Click to expand...
Great explanation thanks bro
 
bguerville
sandungas said:
The dumpers are nice for pre-testing and seting up the enviroment, but have you validated them ?, it seems there has been reports of dumps with "empty" sections, maybe thats caused by problematic USB sticks

One thing i mentioned before (as a paranoid meassure to know if the USB is compatible) is to install an official firmware from it (if you are on 4.82 OFW... you can install 4.82 OFW on top)... if that works the USB should be fine... i guess
There has been people that had to try several different USB sticks until they found one working
Click to expand...
Incomplete dumps are not caused by usb issues. The potential read failures can happen when we try to read more than 1Mb at a time from the flash memory.
The dumpers currently read 2Mb chunks & as a result can get that kind of failure. Usually when one read fails so do the others & the whole dump is filled within 00s.

I already fixed this problem last week actually.
I also fixed the occasional freeze issue.
:)

An update for all dumpers (and for the flash writer) will be posted in the coming days...
 
sandungas
bguerville said:
Incomplete dumps are not caused by usb issues. The potential read failures can happen when we try to read more than 1Mb at a time from the flash memory.
The dumpers currently read 2Mb chunks & as a result can get that kind of failure. Usually when one read fails so do the others & the whole dump is filled within 00s.

I already fixed this problem last week actually.
I also fixed the occasional freeze issue.
:)

An update for all dumpers (and for the flash writer) will be posted in the coming days...
Click to expand...
Ohh, ok, now i reminded i did read your message the other day where you explained it, i like how you care about details btw :)

And btw, when i did read your message, one of the first things that came to mind is if you had more time probably you would have noticed it in the internal beta tests before release
But with all that problems with the leak and hurry stuff nobody can blame you, is completly understandable, and is not happening in a critical process so no harm
 
You must log in or register to reply here.

Similar threads

Featured content

Trending content

Latest posts

Back
Top