News for hackers. Stuff that matters.

A comment on New flaw in Intel chips lets attackers slip their own data into secure enclave. Parent comment

Re:New flaw in Intel chips lets attackers slip their own dat... by akiselev on Tuesday March 10, 2020 @02:00PM (#22537818)
It's probably a bit of both. Intel started implementing speculative execution decades ago but the first exploits like Specter/Meltdown weren't published until a few years ago so that alone is strong evidence that there might be ancient attack vectors that we haven't even considered yet.

On the other hand, interest in this topic among the public has definitely grown since branding departments started getting their hands on the exploits and Bloomberg published that sensationalist Micro story so it's a self reinforcing cycle.

[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by WrtCdEvrydy on Tuesday March 10, 2020 @02:09PM (#22537911)
Interestingly, there was a 1998 paper about speculative execution that basically outlined all of Spectre/Meltdown.
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by akga on Tuesday March 10, 2020 @02:16PM (#22537970)
Sounds interesting. Do you have a link to it?
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by Kurakuan on Tuesday March 10, 2020 @04:27PM (#22539332)
I'm not sure if there was another paper in '98, but this one in 1995 may be the one being referenced: https://web.archive.org/web/20180506083456/https://pdfs.sema...
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by cpach on Tuesday March 10, 2020 @02:57PM (#22538399)
And 'cperciva found some weird stuff long before Spectre. He reported to Intel and their reply was basically “meh”.
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by cperciva on Tuesday March 10, 2020 @03:15PM (#22538575)
Intel is a very different company in 2020 than they were in 2005.
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by saagarjha on Tuesday March 10, 2020 @04:05PM (#22539080)
Are they better, or worse?
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by cpach on Tuesday March 10, 2020 @03:39PM (#22538801)
Alright! What are the most significant changes they have made?
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by cperciva on Tuesday March 10, 2020 @05:05PM (#22539673)
They've gone from "try to get someone fired to get them to shut up" to "offer millions of dollars of bounties to researchers". It's a completely different attitude.
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by ngneer on Tuesday March 10, 2020 @06:25PM (#22540497)
Right, but side channels in Intel products were known. It is unreasonable to expect the literature to account for every single aspect of your design. Had the side channel threat been taken more seriously at the time, it would presumably have led to more robust designs.
[ Parent | on HN ]
Re:New flaw in Intel chips lets attackers slip their own dat... by kevin_thibedeau on Tuesday March 10, 2020 @03:25PM (#22538679)
> sensationalist Micro story

Considering it was published with the intent to "move markets" I would consider it a fraudulent Micro story.

[ Parent | on HN ]