[PATCH] kmap_local: don't assume kmap PTEs are linear arrays in memory

[Ard Biesheuvel]

On Fri, 5 Nov 2021 at 15:50, Russell King (Oracle)
<linux at armlinux.org.uk> wrote:
>
> On Fri, Nov 05, 2021 at 03:21:00PM +0100, Ard Biesheuvel wrote:
> > On Tue, 26 Oct 2021 at 15:13, Ard Biesheuvel <ardb at kernel.org> wrote:
> > >
> > > The kmap_local conversion broke the ARM architecture, because the new
> > > code assumes that all PTEs used for creating kmaps form a linear array
> > > in memory, and uses array indexing to look up the kmap PTE belonging to
> > > a certain kmap index.
> > >
> > > On ARM, this cannot work, not only because the PTE pages may be
> > > non-adjacent in memory, but also because ARM/!LPAE interleaves hardware
> > > entries and extended entries (carrying software-only bits) in a way that
> > > is not compatible with array indexing.
> > >
> > > Fortunately, this only seems to affect configurations with more than 8
> > > CPUs, due to the way the per-CPU kmap slots are organized in memory.
> > >
> > > Work around this by permitting an architecture to set a Kconfig symbol
> > > that signifies that the kmap PTEs do not form a lineary array in memory,
> > > and so the only way to locate the appropriate one is to walk the page
> > > tables.
> > >
> > > Reported-by: Quanyang Wang <quanyang.wang at windriver.com>
> > > Signed-off-by: Ard Biesheuvel <ardb at kernel.org>
> > > ---
> >
> > Ping? Can we get this fixed please?
>
> Who are you expecting to apply it? It seems to be touching only core
> code, so I don't think it's up to me - not without some kind of review
> from mm guys.
>

I was hoping Thomas would respond, given that he introduced the
problem in the first place, and can either carry the fix himself, or
tell us whether he wants it fixed in a different way.