On Mon, Jan 04, 2016 at 02:33:12PM -0800, Andy Lutomirski wrote:
> On Mon, Jan 4, 2016 at 12:26 PM, Marcelo Tosatti wrote:
> > On Sun, Dec 20, 2015 at 03:05:41AM -0800, Andy Lutomirski wrote:
> >> From: Andy Lutomirski
> >>
> >> The pvclock vdso code was too abstracted to understand easily and
>
On Mon, Jan 4, 2016 at 12:26 PM, Marcelo Tosatti wrote:
> On Sun, Dec 20, 2015 at 03:05:41AM -0800, Andy Lutomirski wrote:
>> From: Andy Lutomirski
>>
>> The pvclock vdso code was too abstracted to understand easily and
>> excessively paranoid. Simplify it for a huge speedup.
>>
>> This opens th
On Sun, Dec 20, 2015 at 03:05:41AM -0800, Andy Lutomirski wrote:
> From: Andy Lutomirski
>
> The pvclock vdso code was too abstracted to understand easily and
> excessively paranoid. Simplify it for a huge speedup.
>
> This opens the door for additional simplifications, as the vdso no
> longer
From: Andy Lutomirski
The pvclock vdso code was too abstracted to understand easily and
excessively paranoid. Simplify it for a huge speedup.
This opens the door for additional simplifications, as the vdso no
longer accesses the pvti for any vcpu other than vcpu 0.
Before, vclock_gettime using