This article might be handy for someone interviewing at that firm (Red Balloon) that sends you a "weird" hard drive as the interview CTF? I still have it sitting around but it arrived around finals season so I never really looked at it, but since they bothered to send a whole drive and SATA-USB adapter, it obviously must have something to do with the drive itself.
If someone had a ton of money, it would be funny to just send the thing to a data recovery lab, have them swap the platters onto an unmodified model and get a raw image of the data to work with. (Or maybe the key is hidden inside the drive firmware chip itself?)
i still have mine too! managed to talk to the microcontroller and dump its firmware, but didn't know enough about how to make it arbitrarily run code without worrying about ruining it all
There's also another very good series of articles about hacking the firmware of a HDD, with modifications of /etc/shadow hashed passwords: https://spritesmods.com/?art=hddhack
Came across it looking how to deal with multiple different samsung drives caught in bad states due to shitty firmware. My original salty post warning about vendor branded Samsung drives on eBay is here: https://news.ycombinator.com/item?id=37165189
Samsung has lost any credibility they had as a competent manufacturer years ago. Their other products are beyond junk (fridges, washing machines…), their customer service is abysmal (they managed to “repair” my mp3 player and smartphone by returning it even more broken than they got it, and I’ve seen how the company works from the inside when they bought a startup I was working at. I know many people with Samsung fridges failing after a few years (or having too little coolant in them so that they make loud popping sounds when running and Samsung saying you’re holding it wrong)
From these experiences, I’m going out of my way to never buy anything made by Samsung.
In my case, the drive suddenly only shows having 1GB of read only space available. The firmware version will be reported as “ERRORMOD” (meaning, error mode). There are no warning signs, it just happens.
All data is lost the moment you see ERRORMOD, there is no recovery of data that I am aware of. It is sometimes possible to clear the drive and recover function for the now untrustworthy drive: https://forums.servethehome.com/index.php?threads/pm9a3-firm...
It’s not the only way a drive can fail, but it’s the most immediately obvious one.
Other ways of the firmware failing result in no drive showing up at all, or data corruption. Physical failures can also happen, like breakage of the solder balls under the chips (which fixable enough to get data off it).
Since this is xb360, this is SATA rather than IDE, but in a similar vein I am really looking forward to my PicoIDE to play with adversarial hdd controllers in real systems.
The obfuscation hardware vendors do is so trivial, why do they even bother?
One of the current vendor provided consumer SSD firmware update utilities for Linux as a live-usb decrypts the firmware and writes it out to disk decrypted before uploading it, so simply using seccomp to fail a rmdir syscall nets you the decrypted version without having to reverse engineer any of the updater/decryption code.
I deleted my own negative rant about SSD manufacturers not opting in to lvfs/fwupd when drives have a high risk of bricking without firmware updates.
The fact that vendors still ship firmware with trivial obfuscation in 2026 is wild. I wonder how many data recovery shops already reverse-engineer these routinely but just don't publish.
One of my favorite things to do is update the firmware of devices. I know it is often ill-advised because if it is working fine, why risk something going wrong? But it’s kind of fun to imagine gaining tiny speed increments with optimizations. I like to do it on Fridays - Firmware Fridays - vacuum cleaners, hard drives, motherboards, ip cameras, Apple IIGS expansion cards, Bluetooth scales, and on and on.
For anybody involved with research of any nature, you don't need to be interested in HDDs or SSDs or even hacking hardware or software of any kind.
This says a lot right here:
>One of my initial ideas was to modify the HDD firmware to introduce a delay of a few hundred milliseconds when a specific sector is read from the drive, which would give enough time for the exploit to trigger successfully.
>As it would later turn out I found other ways to dial in my race condition attack and ended up not needing to modify the HDD firmware at all.
The result is a remarkable paper documenting outstanding milestones that is outstanding on its own, and was completely unintentional to begin with, and with subject matter that was also unintentional if not a completely unrelated subject than the direction that the initial ambition was leading toward.
If your research leaders or techniques don't allow for excursions like this, you'd probably be better off getting some.
That would be challenging. Its not a matter of replacing a quartz oscillator like in the eighties. HDDs run very integrated SOCs. Actual SATA, DMA state machines, everything is clocked from same crystal so you would have to know internal architecture very well (clock tree) and know actual internals of the PLL driving everything (registers/datasheet) to be able to reprogram/temporarily slow down just the CPU speed while maintaining SATA link.
23 comments
[ 3.0 ms ] story [ 47.0 ms ] threadIf someone had a ton of money, it would be funny to just send the thing to a data recovery lab, have them swap the platters onto an unmodified model and get a raw image of the data to work with. (Or maybe the key is hidden inside the drive firmware chip itself?)
* https://www.cbc.ca/news/science/nsa-hid-spying-software-in-h...
* https://www.wired.com/2015/02/nsa-firmware-hacking/
:)
Came across it looking how to deal with multiple different samsung drives caught in bad states due to shitty firmware. My original salty post warning about vendor branded Samsung drives on eBay is here: https://news.ycombinator.com/item?id=37165189
BTW thank you for raising this.
From these experiences, I’m going out of my way to never buy anything made by Samsung.
All data is lost the moment you see ERRORMOD, there is no recovery of data that I am aware of. It is sometimes possible to clear the drive and recover function for the now untrustworthy drive: https://forums.servethehome.com/index.php?threads/pm9a3-firm...
It’s not the only way a drive can fail, but it’s the most immediately obvious one.
Other ways of the firmware failing result in no drive showing up at all, or data corruption. Physical failures can also happen, like breakage of the solder balls under the chips (which fixable enough to get data off it).
One of the current vendor provided consumer SSD firmware update utilities for Linux as a live-usb decrypts the firmware and writes it out to disk decrypted before uploading it, so simply using seccomp to fail a rmdir syscall nets you the decrypted version without having to reverse engineer any of the updater/decryption code.
I deleted my own negative rant about SSD manufacturers not opting in to lvfs/fwupd when drives have a high risk of bricking without firmware updates.
This says a lot right here:
>One of my initial ideas was to modify the HDD firmware to introduce a delay of a few hundred milliseconds when a specific sector is read from the drive, which would give enough time for the exploit to trigger successfully.
>As it would later turn out I found other ways to dial in my race condition attack and ended up not needing to modify the HDD firmware at all.
The result is a remarkable paper documenting outstanding milestones that is outstanding on its own, and was completely unintentional to begin with, and with subject matter that was also unintentional if not a completely unrelated subject than the direction that the initial ambition was leading toward.
If your research leaders or techniques don't allow for excursions like this, you'd probably be better off getting some.