Theo De Raadt’s description of bugs in recent Intel processors has made it to Slashdot, where in the comments, Matthew Dillon went through each bug and listed his opinion on each. (two comment entries, starting here) In contrast, Linus Torvalds’ general response was much more subdued. (Thanks, Wiger Van Houten, for links)
2 Replies to “More on potential processor problems”
Comments are closed.
Hmm. Addressed in the Slashdot thread, but:
“AE21 – The execution disable bit is shared between cores. I’m not sure what this means but Intel seems to think that it compromises an anti-hacker feature. Sounds pretty serious.”
I read this the same way as the AC did, and imagine this could be an issue for OSes offering some sort of workaround for legacy software that breaks with it enabled. Trying to let just one ‘trusted’ process live insecurely for its timeslice would inadvertently do the same for whatever’s running on the other CPU.
Modest practical risk, but ugly (workaround: only use one core) if anyone ever wanted to implement the above.