[parisc-linux] Re: Problems with kernel mmap (failing tst-mmap-eofsync in glibc on parisc)

[David S. Miller]

On Fri, 22 Aug 2003 18:41:03 +0100
Matthew Wilcox <willy@debian.org> wrote:

> Uhm.  So what happens when the user has stored into the page and now
> the kernel wants to read from it?  There's still data in the cache for
> the user mapping that's non-coherent with the kernel mapping.

I see.  This causes the page cache read flush_dcache_page() call
not to trigger.

I was very confused by the fact that this bug was explained by
saying that "the shared mmap list that flush_dcache_page() checks".

So the idea is that VM_SHARED should be set based upon whether
we mmap() the thing writable _not_ whether the open() was done
with write permission enabled.

Yes, I agree with that.