Re: Simplify load_unaligned_zeropad() (was Re: [GIT PULL] Ceph updates for 5.20-rc1)

[Mike Rapoport]

On Sun, Aug 14, 2022 at 08:43:09PM -0700, Linus Torvalds wrote:
> On Sun, Aug 14, 2022 at 3:59 PM Linus Torvalds
> <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
> >
> > If TDX has problems with it, then TDX needs to be fixed. And it's
> > simple enough - just make sure you have a guard page between any
> > kernel RAM mapping and whatever odd crazy page.
> 
> .. thinking about this more, I thought we had already done that in the
> memory initialization code - ie make sure that we always leave a gap
> between any page we mark and any IO memory after it.
> 
> But it's possible that I'm confused with the IO window allocation
> code, which does the reverse (ie actively try to avoid starting
> allocations close to the end-of-RAM because there is often
> undocumented stolen memory there)
> 
> I'd much rather lose one page from the page allocator at the end of a
> RAM region than lose the ability to do string word operations.
> 
> Of course, it's also entirely possible that even if my memory about us
> already trying to do that is right (which it might not be), we might
> also have lost that whole thing over time, since we've had a lot of
> updates to the bootmem/memblock setup.

We don't create gaps in the memory initialization code, everything that's
E820_TYPE_RAM goes in the end to the page allocator with exception of
partial pages. And this didn't change in the last few years.
 
>                Linus

-- 
Sincerely yours,
Mike.