Messages in this thread | | | Date | Mon, 20 Feb 2012 22:32:31 -0500 | From | Konrad Rzeszutek Wilk <> | Subject | Re: [PATCH] x86 fixes for 3.3 impacting distros (v1). |
| |
On Mon, Feb 20, 2012 at 08:01:43PM -0500, Steven Rostedt wrote: > On Fri, 2012-02-10 at 10:34 -0500, Konrad Rzeszutek Wilk wrote: > > 66 66 66 90 data32 data32 xchg %ax,%ax > > > > [the 66 66 .. is 'nop']. Looks good right? Well, it does work very well on Intel > > (used an i3 2100), but on AMD A8-3850 it hits a performance wall - that I found out > > is a result of CONFIG_FUNCTION_TRACER (too many nops??) being compiled in (but the tracer > > is set to the default 'nop'). If I disable that specific config option the numbers > > are the same as the baseline (with CONFIG_FUNCTION_TRACER disabled) on the AMD box. > > Interestingly enough I only see these on AMD machines - not on the Intel ones. > > All paravirt ops should be labeled with "notrace" so that function > tracer does not trace those functions. Have you annotated your new > paravirt ops with notrace?
No. I hadn't realized that flag existed until your email a couple of days ago - I hadn't had a chance to see if the notrace would solve this. But let me do that and get back on this.
| |