SSD Drives Vulnerable to Attacks That Corrupt User Data

[Elixer]

I guess new firmware will be making the rounds in a few months.

pdf: https://pdfs.semanticscholar.org/b9bc/a3c9f531002854af48de121cdcc8e0520c7f.pdf

According to research published earlier this year, the programming logic powering MLCs is vulnerable to at least two types of attacks.
...
The first of these attacks, which they named a "program interference," takes place when an attacker manages to write data with a certain pattern to a target's SSD.
...
The second vulnerability researchers discovered in the programming logic of NAND flash memory chips is what they called a "read disturb."
...
Researchers have proposed mitigations in their researcher paper that could fix and counter the effects of both attacks.

https://www.bleepingcomputer.com/ne...vulnerable-to-attacks-that-corrupt-user-data/

 

[BarkingGhostar]

Interesting in that while a lot of folks use SSDs for their boot drives, I would imagine very few use them to store their data.

 

[XavierMace]

That depends if this is limited to single drive setups. More and more businesses are moving to all flash storage on the server side.

 

[VirtualLarry]

I didn't read the linked article yet, but is this the NAND equivalent of a "RowHammer" attack or something?

 

[Elixer]

I know people that have pretty big SSDs (1TB) and t

I didn't read the linked article yet, but is this the NAND equivalent of a "RowHammer" attack or something?

Pretty much.

Looks like the firmware needs to be even more complex than blindly accepting commands, which translates into needing more time to calculate everything which makes things slower.

 

[VirtualLarry]

Is this issue NAND-specific? Would be curious if Intel's Optane / 3DXpoint M.2 SSD form-factor devices are susceptible to this attack.

 

[coercitiv]

Is this issue NAND-specific? Would be curious if Intel's Optane / 3DXpoint M.2 SSD form-factor devices are susceptible to this attack.

I think not even all NAND is similarly susceptible to this kind of attack, since the kind of interference the attacks use scales with both process nodes and number of bits per cell. Need to read the paper though.

Thanks OP for the info, interesting stuff.