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

[Jaromír Doleček <jaromir.dolecek%gmail.com@localhost>]

2018-05-12 7:07 GMT+02:00 Michael van Elst <mlelstv%serpens.de@localhost>:
> On Fri, May 11, 2018 at 05:28:18PM +0200, Jaromír Dole?ek wrote:
>> I've implemented a tweak to read-ahead code to skip the full
>> read-ahead if last page of the range is already in cache, this
>> improved things a lot:
>
> That looks a bit like a hack. Can't you just call uvm_pagelookup()
> to determine if the last page is cached? Or if that's too UVM internal,
> call uvm_findpages() with UFP_NOALLOC.

I thought it would be less abstraction violation when done in
getpages. But actually using uvm_pagelookup() seems to be fine from
that standpoint. Furthermore, it avoids some page flag manipulation,
and makes the change much more localized, to only uvm_readahead.c. So
I'll go with that.

Thanks.

Jaromir