You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I think this is the penultimate question on everyone's minds right now.
Okay, so sandsifter did the heavy lifting; but that just firehoses random instructions at the CPU and looks for interesting results. That sounds like it would generate a nontrivial amount of noise - not so much so as to make the whole process overly tedious or simply unviable, but enough that skill and experience would be paramount to figure out what's worth following up on and what can be safely discarded.
But sandsifter just deals in discovery, and (in this case) to a very rudimentary extent - sandsifter was designed to find oddness in x86 CPUs, not entirely new processor architectures!
I am very interested to find out how you went from "hmm, that's weird" to pinpointing/establishing the very existence of the coprocessor, to identifying 21 of its instructions - and the x86 wrappings!
Also, in the same way a fighter pilot might share how to start up and fly an F-18, would you mind sharing how work on the assembler could be furthered by anyone with a VIA C3 who's interested in playing with this beyond going "huh, it's vulnerable"?
Please tell us that research papers and/or in-depth blog posts are in the pipeline. :)
And thanks, too. This is really awesome. ME, eat your heart out.
The text was updated successfully, but these errors were encountered:
In this video Mr Domas details his methodology and answers some of your questions above: https://youtu.be/_eSAF_qT_FY
It has been pointed out elsewhere that the DEC(sic) has been documented elsewhere in leaked proprietary documents. An online search for via C3 ALTINST should yield some useful results.
Not to detract in any way from Mr Domas' research, the video presentation detailing his methodology is simply amazing.
I think this is the penultimate question on everyone's minds right now.
Okay, so sandsifter did the heavy lifting; but that just firehoses random instructions at the CPU and looks for interesting results. That sounds like it would generate a nontrivial amount of noise - not so much so as to make the whole process overly tedious or simply unviable, but enough that skill and experience would be paramount to figure out what's worth following up on and what can be safely discarded.
But sandsifter just deals in discovery, and (in this case) to a very rudimentary extent - sandsifter was designed to find oddness in x86 CPUs, not entirely new processor architectures!
I am very interested to find out how you went from "hmm, that's weird" to pinpointing/establishing the very existence of the coprocessor, to identifying 21 of its instructions - and the x86 wrappings!
Also, in the same way a fighter pilot might share how to start up and fly an F-18, would you mind sharing how work on the assembler could be furthered by anyone with a VIA C3 who's interested in playing with this beyond going "huh, it's vulnerable"?
Please tell us that research papers and/or in-depth blog posts are in the pipeline. :)
And thanks, too. This is really awesome. ME, eat your heart out.
The text was updated successfully, but these errors were encountered: