In the interest of openness, I've decided to release the exploit. Hopefully, this will ignite the PS3 scene, and you will organize and figure out how to use this to do practical things, like the iPhone when jailbreaks were first released. I have a life to get back to and can't keep working on this all day and night.
Please document your findings on the psDevWiki. They have been a great resource so far, and with the power this exploit gives, opens tons of new stuff to document. I'd like to see the missing HV calls filled in, nice memory maps, the boot chain better documented, and progress on a 3D GPU driver. And of course, the search for a software exploit.
This is the coveted PS3 exploit, gives full memory access and therefore ring 0 access from OtherOS. Enjoy your hypervisor dumps. This is known to work with version 2.4.2 only, but I imagine it works on all current versions. Maybe later I'll write up how it works
Compile and run the kernel module.
When the "PRESS THE BUTTON IN THE MIDDLE OF THIS" comes on, pulse the line circled in the picture low for ~40ns.
Try this multiple times, I rigged an FPGA button to send the pulse.
Sometimes it kernel panics, sometimes it lv1 panics, but sometimes you get the exploit!!
If the module exits, you are now exploited.
This adds two new HV calls,
u64 lv1_peek(16)(u64 address)
void lv1_poke(20)(u64 address, u64 data)
which allow any access to real memory.
The PS3 is hacked, its your job to figure out something useful to do with it.
How it works:
geohot: well actually it's pretty simple
geohot: i allocate a piece of memory
geohot: using map_htab and write_htab, you can figure out the real address of the memory
geohot: which is a big win, and something the hv shouldn't allow
geohot: i fill the htab with tons of entries pointing to that piece of memory
geohot: and since i allocated it, i can map it read/write
geohot: then, i deallocate the memory
geohot: all those entries are set to invalid
geohot: well while it's setting entries invalid, i glitch the memory control bus
geohot: the cache writeback misses the memory
geohot: and i have entries allowing r/w to a piece of memory the hypervisor thinks is deallocated
geohot: then i create a virtual segment with the htab overlapping that piece of memory i have
geohot: write an entry into the virtual segment htab allowing r/w to the main segment htab
geohot: switch to virtual segment
geohot: write to main segment htab a r/w mapping of itself
geohot: switch back
geohot: and would work if memory were encrypted or had ECC
geohot: the way i actually glitch the memory bus is really funny
geohot: i have a button on my FPGA board
geohot: that pulses low for 40ns
geohot: i set up the htab with the tons of entries
geohot: and spam press the button
geohot: right after i send the deallocate call