lkml.org 
[lkml]   [2011]   [Oct]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 2/3] pci: Clamp pcie_set_readrq() when using "performance" settings
On Tue, Oct 04, 2011 at 07:30:39PM +0200, Benjamin Herrenschmidt wrote:
> Well, thing is, you -already- have the whole "performance" option which
> is what we are 'arguing' about upstream. Except that the implementation
> of it that you have in your tree now has very nasty bugs (ie it doesn't
> do what it's supposed to and really doesn't work).
>
> Patches 1 and 2 fix it to do what it's supposed to.

I'll agree with benh for now since, yes, the performance option is
completely broken in just blindly applying settings right now. It should
be revisted and made so that a safe performance setting is supported in
the future (possibly via suggestions from arch-specific quirks), since
having to configure this makes little sense for the vast majority of
users. Applying 3/3 first would also make the most sense.

-ben



\
 
 \ /
  Last update: 2011-10-04 19:43    [W:0.113 / U:0.184 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site