Hacking Theory: Limited SysNAND Restore Count on N3DS

[3dsn00b]

This is just a theory. Maybe Nintendo has placed a finite counter on the SysNAND of N3DS's before it completely turns off and stops functioning.

As there have been cases where people have bricked their SysNAND's (myself included) due to restoring the SysNAND a lot of times.

There is another case I know of where the guy has his SysNAND bricked, but his EmuNAND still functions with A9LH.

Thoughts guys?

 

[Davidosky99]

This is just a theory. Maybe Nintendo has placed a finite counter on the SysNAND of N3DS's before it completely turns off and stops functioning.

As there have been cases where people have bricked their SysNAND's (myself included) due to restoring the SysNAND a lot of times.

There is another case I know of where the guy has his SysNAND bricked, but his EmuNAND still functions with A9LH.

Thoughts guys?

I think they aren't that bad. I mean they are kind of a pain blocking hax entrypoints.
They aren't like Samsung with their Knox counter

 

[a9lh-1user]

no

WOW you are very helpfull!

I thing that "some" people (dev's) have restored theyr NAND 100 times (and more) and it is still functional
But i may be wrong!
I guess that sometimes there are "bad" productions......maybe you got one.

 

[cearp]

stuff like efuses - like ms does.
that would be cruel, good idea, but surprising for nintendo to do.

 

[Urbanshadow]

I'm gonna try to be as polite as possible.

You do realize, besides SD and arguably bootrom, NAND is the only non-volatile storage of the system?

You see, when you restore a backup, you are atually "flashing" the NAND, all of it. Partitions in the NAND are many, and explained here.
Hence when you restore your nand you are rewriting all of this. If the system stored such counter it would be there. And you know what? Counter would be decreased by a restore.

So, either your backup was corrupt, or your system corrupted your sysnand somehow (NAND hardware failures or related failures).

In case the NAND hardware is context faulty, but the fault does not reside in the FIRM1 or FIRM0 partitions, a system like a9lh could perfectly work if the NAND is still accesible.
Since emunand loads and resides from/in SD, the case scenario you describe is completely possible.

 

[3dsn00b]

I thing that "some" people (dev's) have restored theyr NAND 100 times (and more) and it is still functional
But i may be wrong!

Is this on the N3DSXL? Which model? And it's confirmed?

stuff like efuses - like ms does.
that would be cruel, good idea, but surprising for nintendo to do.

I don't know, I just had the idea, as this is what I would do if I were Nintendo.

Also, anyone want to be a guinea pig and try on their N3DS?? FOR SCIENCE??

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

I'm gonna try to be as polite as possible.

You don't have to be polite. I can take it.

It was just a theory as stated. I wanted to see what others thought about this theory. Theories are usually wrong 99% of the time anyway.

 

[DjoeN]

I restored my N3DS XL sysnand more then 30 times since i have it (yes i like to mess around in sysnand, just for fun and mostly i f*ck*d it up), still working perfect.

 

[Davidosky99]

I restored my N3DS XL sysnand more then 30 times since i have it (yes i like to mess around in sysnand, just for fun and mostly i f*ck*d it up), still working perfect.

Same. If flashing my sysnand could potentially brick my 3ds I would have already bricked it. Like a long time ago

 

[3dsn00b]

I restored my N3DS XL sysnand more then 30 times since i have it (yes i like to mess around in sysnand, just for fun and mostly i f*ck*d it up), still working perfect.

Thanks for the reply. This pretty much confirms its fine to restore it multiple times.

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

Same. If flashing my sysnand could potentially brick my 3ds I would have already bricked it. Like a long time ago

Theory: DEAD!

 

[Queno138]

Just to ask, how is "flashing back nand" any difference from normal usage, like eshop usage and saving files.

In both cases data is written to nand.

Wouldn't this be anti-business:
The moment people use their n3ds too many times, it bricks.

It sounds like a Corrupted nand, or maybe the software that flashes the nand back started at wrong offset.

 

[duke_srg]

This is just a theory. Maybe Nintendo has placed a finite counter on the SysNAND of N3DS's before it completely turns off and stops functioning.
As there have been cases where people have bricked their SysNAND's (myself included) due to restoring the SysNAND a lot of times.
There is another case I know of where the guy has his SysNAND bricked, but his EmuNAND still functions with A9LH.
Thoughts guys?

The more you write to NAND the more is the possibility that reserved area for wear-leveling will be exhausted and flash fails. Restoring full NAND restore have times more writes compared to the normal system updated when only a small part is updated. So the probability of fail is number_of_writes * your_flash_chip_quality - just unlucky

 

[GothicIII]

And here it goes again... The master planned obsolescence conspiracy about nand chips. Oh yeah took it long enough.

No seriously the guy above me is absolute right. If you install an update the write operation is really small. It just replaces those parts where the firmware is really updated. The nand dump contains everything and rewriting it again and again will rise the risk in damaging sectors thus brick the device. It doesn't support TRIM because the NAND is just a flash chip and not a controller This would have solved the problem. EDIT: Im stupid no it wouldnt solve the problem when you rewrite the whole chip

This is also one of the reasons why I'm experimenting with emunand first before writing things to sysnand. If the SD fails, I can replace it. If the NAND fails I need to buy a new 3DS.

It would be wise to implement block checking before writing. So e.g. 64KB chunks are read out from NAND and compared with the nand dump. If they're the same then they'll be skipped. This wouldn't save time and wouldn't take longer in writing the nand but it would increase the life span of the nand chip.

 

[duke_srg]

It would be wise to implement block checking before writing. So e.g. 64KB chunks are read out from NAND and compared with the nand dump. If they're the same then they'll be skipped. This wouldn't save time and wouldn't take longer in writing the nand but it would increase the life span of the nand chip.

Good idea, but we need the actual flash erase page size for each chip type to make it work properly.

 

[Deleted member 333767]

One would imagine the limit of how many times you can restore your sysNAND is governed by the P/E cycles (program-erase) your eMMC has left. That isn't a limitation placed by Nintendo, that's just the nature of the technology.

Shouldn't have to worry about that for a very long time, I've got an SSD for Win7 install in my PC, been going strong for 5 years now, no degradation that I can detect so far!

 

[GothicIII]

One would imagine the limit of how many times you can restore your sysNAND is governed by the P/E cycles (program-erase) your eMMC has left. That isn't a limitation placed by Nintendo, that's just the nature of the technology.

Shouldn't have to worry about that for a very long time, I've got an SSD for Win7 install in my PC, been going strong for 5 years now, no degradation that I can detect so far!

You really have no clue... Your SSD has a builtin controller which is responsible for sectors being flagged as "deleted" without writing to them directly. This is the TRIM command. It is essential for increasing life span for a SSD. Also there is another mechanism which reallocates broken sectors to other ones so the drive won't fail completely. This is the reason why it is suggested to have a blank, unused partition of 10% in size of your SSD because so there are enough sectors which can be reallocated+the general performance of the ssd improves because of the free space.

The NAND doesn't have any of these mechanisms and you write the nand completely from sector 0 until end. If you do this to your SSD I swear you even your drive will fail after around 1 month of fullspeed write cycles.
Background info: http://techreport.com/review/27909/the-ssd-endurance-experiment-theyre-all-dead

 

[democracy]

This is just a theory. Maybe Nintendo has placed a finite counter on the SysNAND of N3DS's before it completely turns off and stops functioning.

As there have been cases where people have bricked their SysNAND's (myself included) due to restoring the SysNAND a lot of times.

There is another case I know of where the guy has his SysNAND bricked, but his EmuNAND still functions with A9LH.

Thoughts guys?

I think its great news about the "There is another case I know of where the guy has his SysNAND bricked, but his EmuNAND still functions with A9LH."

Everyone should just move to updated sysnand, only ever backup and restore ctrnand and wait for that partition to die and when it does, emunand forever!

 

[duke_srg]

The NAND doesn't have any of these mechanisms and you write the nand completely from sector 0 until end. If you do this to your SSD I swear you even your drive will fail after around 1 month of fullspeed write cycles.

Actually wear leveling and reserved area should exist in eMMC used in 3DS.

 

[zoogie]

Nintendo's history has dictated that they're deathly afraid of lawsuits and bad publicity. Consider also that many of Nintendo's customers are adults who buy consoles for their children/teenagers and will get extremly angry if nintendo used extreme tactics like this. They would just see that their kids system stopped working and Nintendo's to blame. It's also a legal grey area to intentionally damage property to avenge a TOS wrongdoing. It would likely just backfire on them.

tl;dr Nintendo isn't Gateway.

 

[Dennis G]

This is just a theory. Maybe Nintendo has placed a finite counter on the SysNAND of N3DS's before it completely turns off and stops functioning.

As there have been cases where people have bricked their SysNAND's (myself included) due to restoring the SysNAND a lot of times.

There is another case I know of where the guy has his SysNAND bricked, but his EmuNAND still functions with A9LH.

Thoughts guys?

my sysnand broke after less then 5 restores with hard mod so if ther would be a counter everyone would be doomed since that's not the case i don't think so

 

[Deleted member 333767]

You really have no clue... Your SSD has a builtin controller which is responsible for sectors being flagged as "deleted" without writing to them directly. This is the TRIM command. It is essential for increasing life span for a SSD. Also there is another mechanism which reallocates broken sectors to other ones so the drive won't fail completely. This is the reason why it is suggested to have a blank, unused partition of 10% in size of your SSD because so there are enough sectors which can be reallocated+the general performance of the ssd improves because of the free space.

The NAND doesn't have any of these mechanisms and you write the nand completely from sector 0 until end. If you do this to your SSD I swear you even your drive will fail after around 1 month of fullspeed write cycles.
Background info: http://techreport.com/review/27909/the-ssd-endurance-experiment-theyre-all-dead

Right, someone who has converted their 3DS eMMC into a PC hard drive, they will have it fail within one month? Lol, if somebody does that, let me know! I highly doubt anyone will flash their 3DS NAND enough to cause degradation ever.