January 23, 2018

OH NO: Linus Torvalds declares Intel fix for Meltdown/Spectre ‘COMPLETE AND UTTER GARBAGE.’. “As a potential line of inquiry, he suggests: ‘Has anybody talked to them and told them they are f*cking insane?'”

These and other kind epithets are awarded by Torvalds in a public email chain between him and David Woodhouse, an engineer at Amazon in the U.K., regarding Intel’s solution as relating to the Linux kernel. The issue is (as far as I can tell as someone far out of their depth) a clumsy and, Torvalds argues, “insane” implementation of a fix that essentially does nothing while also doing a bunch of unnecessary things.

The fix needs to address Meltdown (which primarily affects Intel chips), but instead of just doing so across the board, it makes the whole fix something the user or administrator has to opt into at boot. Why even ask, if this is such a huge vulnerability? And why do it at such a low level when future CPUs will supposedly not require it, at which point the choice would be at best unnecessary and at worst misleading or lead to performance issues?

Meanwhile, a bunch of other things are added in the same patch that Torvalds points out are redundant with existing solutions, for instance adding protections against an exploit already mitigated by Google Project Zero’s “retpoline” technique.

Why do this? Torvalds speculates that a major part of Intel’s technique, in this case “Indirect Branch Restricted Speculation” or IBRS, is so inefficient that to roll it out universally would result in widespread performance hits. So instead, it made the main Meltdown fix optional and added the redundant stuff to make the patch look more comprehensive.

RELATED: Intel asks customers to halt patching for chip bug, citing flaw.

That appears to be something of an understatement.