Hacker News Reader

Hacker News

Comment by mm-vorticesoft | Hacker News Reader

Full thread

mm-vorticesoft

Linus was right

View on HN

Well he is known for calling a spade a spade. No matter how blunt that may be.

except for when he is wrong.

Why he gets a pass for being the 'nice guy' and DeRaadt gets the bad rep is still a mystery to me

I thought he was angry about the mitigation being disabled by default, not being unstable.

Nah, it was more than that: The patches do things like add the garbage MSR writes to the kernel entry/exit points. That's insane. That says "we're trying to protect the kernel". We already have retpoline there, with less overhead.

That was about patches to the linux kernel, not the microcode patches.

Yes, but as far as I know Linus has made no comment on the microcode patches, so mm-vorticesoft is probably referring to the Spectre patches in general.

The microcode patches are binary blobs against a proprietary and secret ISA, how can anyone comment on their quality?

When I read it I believed that Linus was implying that the suggested mitigation was so insane that it seemed like Intel MIGHT be hiding how broken they believed their hardware was with such over-the-top reactions. As well as indirectly asking if they believed the currently accepted mitigation method (retpoline) was considered ineffective.

His overall point was a bewilderment at the incompetent and non-sensical patches that were being given as "fixes" in this issue. Linus was pointing out a particular instance of that, but this news and other behaviour from Intel seems to indicate this is part of an endemic, cultural, administrative issue inside the company.

Aren't we talking about two different things? Linux vs Windows kernel?

OP is probably referencing the 'bullshit patches from Intel' comment from Linus about the patches they were sent, and that Microsoft might have been sent similar obfuscatory patches.

Given that the patches are CPU micro-code delivered by OS drivers, AFAIK, the actual OS won't make much difference.

"Decades old trap/fault software is being replaced by 10-20 operating systems, and there are going to be mistakes made."

- Theo de Raadt

Linus was talking about Linux patches, not these microcode patches. They've been known broken for at least a week

About what? Would you have a link?

Linus Torvalds: “Somebody is pushing complete garbage for unclear reasons.” http://lkml.iu.edu/hypermail/linux/kernel/1801.2/04628.html

Oh, this is the same issue?

It's the same bug, same company pushing patches, but we don't know if it's the same reason.

It's not the same company - David Woodhouse works for Amazon. He used to work for Intel but not for a year or so.

It's also not the same reason. Linus doesn't like the mitigation in the kernel, disagreeing on how Intel intends to implement it. This article is about unstable microcode patches that Intel retracted, and that retraction has been discussed on here a few times. The article is just exceedingly bad at describing the actual issue. It also doesn't help that the kernel mitigation depends on new flags introduced by the faulty microcode update, but the update being faulty is orthogonal to Linus' opinion.

Like always.

He generally is. His rants are almost always spot-on and pure gold. :)

What are you talking about?

Reply on news.ycombinator.com