Re: Fixing excessive shootdowns during FFS read (kern/53124) on x86 - emap, direct map

[Chuck Silvers <chuq%chuq.com@localhost>]

On Sun, May 13, 2018 at 02:59:50PM +0200, Jaromr Dole?ek wrote:
> 2018-05-12 21:24 GMT+02:00 Chuck Silvers <chuq%chuq.com@localhost>:
> > the problem with keeping the pages locked (ie. PG_BUSY) while accessing
> > the user address space is that it can lead to deadlock if the page
> > we are accessing via the kernel mapping is the same page that we are
> > accessing via the user mapping.
> 
> The pages marked PG_BUSY for the UBC transfer are part of page cache.
> I sincerely hope it's impossible to have any user mapping for it.

the whole point of UBC is that the pages caching the data used by read/write
are the same pages used by mmap.  that's how we avoid the double-caching
and cache coherency problems that the pre-UBC scheme had.
so yes, these pages can have user mappings at the same time as
they are mapped in the kernel.

-Chuck