Linus Torvalds 关于 Intel Meltdown/Spectre 补丁的评论

On Sun, Jan 21, 2018 at 12:28 PM, David Woodhouse <dwmw2@xxxxxxxxxxxxx> wrote:

> On Sun, 2018-01-21 at 11:34 -0800, Linus Torvalds wrote:
>> All of this is pure garbage.
>>
>> Is Intel really planning on making this shit architectural? Has anybody talked to them and told them they are f*cking insane?
>>
>> Please, any Intel engineers here – talk to your managers.
>
> If the alternative was a two-decade product recall and giving everyone free CPUs, I’m not sure it was entirely insane.

You seem to have bought into the cool-aid. Please add a healthy dose of critical thinking. Because this isn’t the kind of cool-aid that makes for a fun trip with pretty pictures. This is the kind that melts your brain.

> Certainly it’s a nasty hack, but hey ? the world was on fire and in the  end we didn’t have to just turn the datacentres off and go back to goat farming, so it’s not all bad.

It’s not that it’s a nasty hack. It’s much worse than that.

> As a hack for existing CPUs, it’s just about tolerable ? as long as it  can die entirely by the next generation.

That’s part of the big problem here. The speculation control cpuid stuff shows that Intel actually seems to plan on doing the right thing for meltdown (the main question being _when_). Which is not a huge surprise, since it should be easy to fix, and it’s a really honking big hole to drive through. Not doing the right thing for meltdown would be completely unacceptable.

So the IBRS garbage implies that Intel is _not_ planning on doing the right thing for the indirect branch speculation.

Honestly, that’s completely unacceptable too.

> So the part is I think is odd is the IBRS_ALL feature, where a future  CPU will advertise “I am able to be not broken” and then you have to set the IBRS bit once at boot time to *ask* it not to be broken. That part is weird, because it ought to have been treated like the RDCL_NO bit ? just “you don’t have to worry any more, it got better”.

It’s not “weird” at all. It’s very much part of the whole “this is complete garbage” issue.

The whole IBRS_ALL feature to me very clearly says “Intel is not serious about this, we’ll have a ugly hack that will be so expensive that we don’t want to enable it by default, because that would look bad in benchmarks”.

So instead they try to push the garbage down to us. And they are doing it entirely wrong, even from a technical standpoint.

I’m sure there is some lawyer there who says “we’ll have to go through motions to protect against a lawsuit”. But legal reasons do not make for good technology, or good patches that I should apply.

> We do need the IBPB feature to complete the protection that retpoline gives us ? it’s that or rebuild all of userspace with retpoline.

BULLSHIT.

Have you _looked_ at the patches you are talking about? You should have – several of them bear your name.

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.

So somebody isn’t telling the truth here. Somebody is pushing complete garbage for unclear reasons. Sorry for having to point that out.

If this was about flushing the BTB at actual context switches between different users, I’d believe you. But that’s not at all what the
patches do.

As it is, the patches are COMPLETE AND UTTER GARBAGE.

They do literally insane things. They do things that do not make sense. That makes all your arguments questionable and suspicious. The patches do things that are not sane.

WHAT THE F*CK IS GOING ON?

And that’s actually ignoring the much _worse_ issue, namely that the whole hardware interface is literally mis-designed by morons.

It’s mis-designed for two major reasons:

– the “the interface implies Intel will never fix it” reason.

See the difference between IBRS_ALL and RDCL_NO. One implies Intel will fix something. The other does not.

Do you really think that is acceptable?

– the “there is no performance indicator”.

The whole point of having cpuid and flags from the microarchitecture is that we can use those to make decisions.

But since we already know that the IBRS overhead is huge on existing hardware, all those hardware capability bits are just complete and utter garbage. Nobody sane will use them, since the cost is too damn high. So you end up having to look at “which CPU stepping is this” anyway.

I think we need something better than this garbage.

Linus

【LMAO】

via：https://www.reddit.com/r/Amd/comments/7s3rnr/linus_torvalds_on_current_meltdownspectre_patches

source：http://lkml.iu.edu/hypermail/linux/kernel/1801.2/04628.html