Big Vulnerability Found in Surface Pro 3

[piemadd]

So turns out, if you short the pins in a USB port on a SurfacePro 3, the entire thing will crash. I don't know if anybody else found this, but, OK.

[79wjd]

There's an even bigger vulnerability that lies at the driver/software level......there are so many random bugs, glitches, and hitches, that if I didn't have as much self control as I did I would've thrown it out the window several dozen times over by now.

[Drak3]

Might be a failsafe to try to protect the device from things like USB killer.

[Lumi]

I get excited when I hear the term vulnerability then I read this garbage.

[BlueChinchillaEatingDorito]

Now try that with your desktop.

[Windows7ge]

14 minutes ago, Drak3 said:

Might be a failsafe to try to protect the device from things like USB killer.

I doubt this. Just because the software stops operating there's still physical connections inside the system. Plus a USB killer doesn't short the GND & +5V pins. It charges a bank of capacitors off it and then discharges them through the data + & - pins. The SurfacePro 3 would still get damaged.

[DocSwag]

I once stuck a piece of metal in the USB port of my MBP. It shut off and I had to unplug and replug the battery to get it to turn on again.

 

Might just be a failsafe to protect against shorts.

[Drak3]

20 minutes ago, Windows7ge said:

I doubt this. Just because the software stops operating there's still physical connections inside the system. Plus a USB killer doesn't short the GND & +5V pins. It charges a bank of capacitors off it and then discharges them through the data + & - pins. The SurfacePro 3 would still get damaged.

The first version operated on GND and 5V. Second version operates on Data.

 

And I wouldn't exactly toss this in the unlikely bin. I had an Acer laptop from the Vista era that did this. Error log reported it as well.

[Windows7ge]

2 minutes ago, Drak3 said:

The first version operated on GND and 5V. Second version operates on Data.

 

And I wouldn't exactly toss this in the unlikely bin. I had an Acer laptop from the Vista era that did this. Error log reported it as well.

Fair enough. At the same time though OP didn't specify which pins are being shorted. Shorting 5V to a data pin or 5V to GND, or data + to GND. Depending on what is actually crashing the computer I just think it's more likely to be a fault rather than a feature. Especially if its had a reputation of previous issues.

[Drak3]

Just now, Windows7ge said:

Fair enough. At the same time though OP didn't specify which pins are being shorted. Shorting 5V to a data pin or 5V to GND, or data + to GND. Depending on what is actually crashing the computer I just think it's more likely to be a fault rather than a feature. Especially if its had a reputation of previous issues.

Well, we've got 3 people, running 3 different devices, that had the same thing happen. Different hardware platforms, and software being ran. Either a common fault that's in most devices, or a safeguard.

[Windows7ge]

1 minute ago, Drak3 said:

Well, we've got 3 people, running 3 different devices, that had the same thing happen. Different hardware platforms, and software being ran. Either a common fault that's in most devices, or a safeguard.

I think it's in the latest macbook where the USB C connectors are designed in such a way where if a USB killer is used on it it only damages the port itself but the rest of the system will continue to function without issue.

A feature where the OS crashes when pins get shorted is a pretty piss poor feature especially getting back to my original argument where this won't protect the hardware. The high voltage discharge will still rampage through the system and probably kill the motherboard if not some other components with it.

[Drak3]

Just now, Windows7ge said:

A feature where the OS crashes when pins get shorted is a pretty piss poor feature especially getting back to my original argument where this won't protect the hardware.

Likely not the OS itself, but firmware or hardware. As Doc's MBP wouldn't start without removing the battery first, same with my old Acer, it's likely that there is a trip that doesn't reset so long as it is powered. It'll also protect hardware from realistic shorts well enough that it's worth implementing.

2 minutes ago, Windows7ge said:

The high voltage discharge will still rampage through the system and probably kill the motherboard if not some other components with it.

Only if there is no failsafe, or you design something explicitly to kill a system and circumvent failsafes, like USB Killer 2.0.

[Windows7ge]

1 minute ago, Drak3 said:

Only if there is no failsafe, or you design something explicitly to kill a system and circumvent failsafes, like USB Killer 2.0.

With news like that if I find a thumb drive it makes me want to crack it open and check the PCB before I plug it into anything.

[GoodBytes]

I just tried, and my Surface Pro 3 (running Windows 10) doesn't have this problem.

I plug a USB cable, where the end is cut, and connected orange/red cable with the ground, I also tried connecting all wires together in a bundle... beside hearing small shock sound, system is fine, and USB port still work after plugging a USB Flash drive.

[Drak3]

1 minute ago, GoodBytes said:

I just tried, and my Surface Pro 3 (running Windows 10) doesn't have this problem.

 

Now, we just need one more person willing to potentially kill their SP3's USB port to verify if your machine or OP's machine is the fluke.

[GoodBytes]

9 minutes ago, Drak3 said:

Now, we just need one more person willing to potentially kill their SP3's USB port to verify if your machine or OP's machine is the fluke.

I don't expect the SP3 to be problem free... it is, at the end of the day the 3rd computer that Microsoft ever made. This is compared to countless systems and years of experience from all other OEMs. And heck, at every release of Surface product, it has all sorts of driver and firmware issues (always never before occurred issues). Microsoft has yet to master computer hardware testing.

 

I am just glad that they fixed the UEFI/firmware upgrade process issue of the Surface Pro 2, where in the end, once Microsoft got it working, they went in a "let's not touch this anymore.. it works", like it is walking on eggs. So, the Surface line is in this strange position. Fantastic hardware, great design, incredible how much performance they manage to pack in, while keeping decently good battery life... and despite the premium price, the firmware, drivers usually need many many months before all works fine.

[Drak3]

1 minute ago, GoodBytes said:

I don't expect the SP3 to be problem free... it is, at the end of the day the 3rd computer that Microsoft ever made. This is compared to countless systems and years of experience from all other OEMs. And heck, at every release of Surface product, it has all sorts of driver and firmware issues (always never before occurred issues). Microsoft has yet to master computer hardware testing.

 

I am just glad that they fixed the UEFI/firmware upgrade process issue of the Surface Pro 2, where in the end, once Microsoft got it work, they went in a "let's not touch this anymore.. it works", like it is walking on eggs.

Being 100% honest, the Surface lineup has been as dependable as the more reliable lineups of companies like Dell and Apple, with more issues being PEBKAC or third party softwares breaking something from everything I've seen.

[Skanky Sylveon]

2 hours ago, Drak3 said:

Might be a failsafe to try to protect the device from things like USB killer.

Interesting, the question is whether it will also protect the device from hammers.

A lot more people have those.

[79wjd]

48 minutes ago, Drak3 said:

Being 100% honest, the Surface lineup has been as dependable as the more reliable lineups of companies like Dell and Apple, with more issues being PEBKAC or third party softwares breaking something from everything I've seen.

Tell that to my Surface that has nothing but One Note and Chrome Installed on it when One Note decides to act up, when it decides to drop wifi every five minutes, when it fails to recognize the keyboard is attached, when it fails to open the on screen keyboard when you click a text box, or when it loses connection to the pen.

[piemadd]

9 hours ago, Drak3 said:

Now, we just need one more person willing to potentially kill their SP3's USB port to verify if your machine or OP's machine is the fluke.

Not just my machine. I shared the information with a few friends who have the same machine, and it is the same.

[Drak3]

5 minutes ago, Implosivetech said:

Not just my machine. I shared the information with a few friends who have the same machine, and it is the same.

Goodbyte's machine is probably a fluke then.