Using Kexec Allows Starting Linux In PlayStation 4

jones_supa writes: Team fail0verflow, the hacker group who made Sony PlayStation 4, has introduced another method to start Linux in the game console. Instead of the previous exploit which was based on a security hole in an old PS4 firmware version, the new trick allows a kexec call to start Linux through Orbis OS (the FreeBSD-based system software of PS4). The code can be found in GitHub. Maybe this will lead to more and better PlayStation clusters.

Comments

What?

[bondsbw]

Team fail0verflow, the hacker group who made Sony PlayStation 4,

Highlighted for the accuracy of this summary.

Re:What?

[jones_supa]

Facepalm. It seems that both I and Timothy are sleeping on the wheel. :D

Anyway, I would also add to the summary that you need some another way to actually make the kexec call. Over at PSXHAX there was posted a new BadIRET exploit last Wednesday. Maybe it works.

Re:What?

[davester666]

It was nice of them to turn over production of it to Sony.

Okay, but

[Anonymous Coward]

Launching Linux from FreeBSD is also known as "downgrade". Just sayin'

Re:Okay, but

[KiloByte]

Launching Linux from FreeBSD is also known as "downgrade". Just sayin'

Only if you'd run systemd.

PS3 clusters were *briefly* interesting...

[Junta]

The cell processor was very briefly an interesting beast at the time it came out. It represented surprisingly good bang for the buck when the PS3 released. No console hardware before or since has been 'ahead of its time' enough to offset the inherent limitations of a home entertainment device.

Unfortunately, while it had tremendous capability to run certain traditional HPC jobs, it wasn't that good a match for what game developers needed most...

The current crop is particularly less compelling, since they were basically midrange PC at the time of launch.

Re:PS3 clusters were *briefly* interesting...

[Dutch Gun]

Unfortunately, while it had tremendous capability to run certain traditional HPC jobs, it wasn't that good a match for what game developers needed most...

Which is???

The cell broadband chip was originally designed to be used for multimedia processing. It had amazing vector processing capabilities, but videogame code doesn't have a lot of cases where you're crunching large amounts of data using simple algorithms, as with video decoding. As it turns out, typical game code tends to have lots of branches, special cases, one offs, etc (physics, AI, pathfinding, skeletal animation, etc). I imagine this is also why the major manufacturers returned to a more traditional x86-64 architecture instead of the PPC. The PowerPC had crappy out-of-order processing and branch prediction, and suffers rather badly when faced with branch-heavy code.

In other words, most videogames require processors that can deal efficiently with branching code and lots of small, irregular data sets via caching. That's pretty much the exact opposite of what the cell is good at. There were some some devs who went through some fairly heroic efforts to squeeze all they could out of the cell, but in the end, all that effort netted about the same results as another couple of general-purpose CPU cores that took almost no coding effort. Most game developers I know disliked the PS3, because to get the same performance as on the Xbox 360 required about 10x the effort.

I'm impressed

[NotDrWho]

Team fail0verflow, the hacker group who made Sony PlayStation 4

Well if they built it in the first place, no wonder they're able to hack it.

Why bother

[Anonymous Coward]

If freebsd is underneath the whole PS4 system, then why not just use that instead of booting into linux? Seems a little excessive.

Re:Why bother

[marcansoft]

Because it's not FreeBSD. Just because Sony based their kernel on FreeBSD doesn't mean it has a FreeBSD userland, nor does it mean you can just slap on a FreeBSD userland and make it work.

You'd have to port FreeBSD all over again - and it turns out that Linux has better off-the-shelf support for the PS4 hardware than FreeBSD does. The only reason Sony didn't use Linux is because of the license, not because it isn't easier to make work on this hardware.

How long

[Dunbal]

Downloading Steam games to play PC games under linux on a PS4... use an xbox controller to play that game and the universe might just end.

in before "update"

[Gravis Zero]

it should only be a day or so before Sony removes this ability for "security" reasons. fuck Sony. yes, fuck Microsoft more but still, fuck Sony.

Re:in before "update"

[marcansoft]

How exactly do you alter the kernel to stop you from running kernel code when you can already run kernel code? I'd like to hear about this magical technology that Sony has invented.

Try better reading comprehension next time. This is just code. It's not a way to run code. Therefore, Sony can't do anything about it, because there's nothing to be done. Sony can't magically make code stop being code. That's like saying Microsoft is capable of making Linux stop working on an (open) machine you choose to install Windows on.

PowerPC was common in gaming...

[Kenshin]

The Xbox 360 had a PowerPC-based chip as well, named Xenon. The original devkits for it were PowerMac G5s.

The GameCube, Wii, and Wii U were also all based around PowerPC.

Re:PowerPC was common in gaming...

[Dutch Gun]

Correct - I didn't mean to imply otherwise, if that's what it sounded like. The reason Xbox 360 fared better was because of it's symmetric three-core/six-thread configuration, which was much easier to program than the PS3's asymmetric single PPE + seven SPE configuration. But all the PPC-based chips had the same issue with the CPUs stalling quite a bit in normal gameplay code - there's really very little you can do about that as a developer, as you can only simplify or re-architect your code so much to help the CPU along.