Dan Luu
patreon
Linus's rants
🕑 Added 2018-01-22 19:40:43 +0000 UTC Linus has been venting his spleen on LKML again. The last rant I saw was about a set of patches intended to mitigate some of the damage from the recent discovered meltdown and spectre attacks. The rant is quite popular -- it’s been on every major tech news aggregator, and there are articles about it in most of the major online tech publications (zdnet, the register, the inquirer, etc.)If you read the comments, there’s the typical argument about how it’s not necessary to be so vitriolic vs. the argument about how it’s impossible to maintain quality without breaking a few eggs and how Linus is always right in his rants, which is why it’s valuable that he rants and also valuable to pass around his rants.
The funny thing is that if you read the thread, it appears that Linus’s rant is at least partially mistargeted and it seems that virtually no one who’s commenting about how Linus is correct has actually read the thread. Here’s a response from David Woodhouse:
On Sun, 2018-01-21 at 13:35 -0800, Linus Torvalds wrote:
> On Sun, Jan 21, 2018 at 12:28 PM, David Woodhouse wrote:
> > We do need the IBPB feature to complete the protection that retpoline
> > gives us â it's that or rebuild all of userspace with retpoline.
>
> BULLSHIT.
>
> Have you _looked_ at the patches you are talking about?ÂÂYou 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".ÂÂWe already have retpoline there, with less overhead.
You're looking at IBRS usage, not IBPB. They are different things.
Yes, the one you're looking at really *is* trying to protect the
kernel, and you're right that it's largely redundant with retpoline.
(Assuming we can live with the implications on Skylake, as I said.)
> 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.
That's what the *IBPB* patches do. Those were deliberately put first in
the series (and in fact that's where I stopped, when I posted).
Since Patreon doesn’t have a great way to do multilevel quotes, I haven’t quoted the response and am just going to note that this is the end of the quote.
If you read the followups to the rants that get passed around, you’ll find that it’s quite common for Linus to respond with a less vitriolic and more accurate comment after the storm has passed, but those comments don’t get passed around. My guess is that fewer than 1% of people who see a story about a rant will read enough of the thread to understand the context of the rant as well as read the follow-up responses. I suspect that Linus’s rants get passed around purely because people enjoy the spectacle -- if it was for the technical content, as is often claimed, people would wait to see if there’s a calmer and more accurate followup, but that’s not what happens. In the cases where there’s a more technically accurate or precise followup, it virtually never goes viral the way the angry rants go viral.
