On Mon, Apr 29, 2013 at 10:35 PM, Steve Kaufmann <s...@cray.com> wrote:
>
> There is already Ivy Bridge support in libpfm4 but it is only for model 58.
> We require Ivy Bridge EP model 62 support.
>
> Locally I've change the pfmlib_intel_ivb.c file:
>
> $ diff --git a/lib/pfmlib_intel_ivb.c b/lib/pfmlib_intel_ivb.c
> index 64c6ed7..dae1325 100644
> --- a/lib/pfmlib_intel_ivb.c
> +++ b/lib/pfmlib_intel_ivb.c
> @@ -41,6 +41,8 @@ pfm_ivb_detect(void *this)
> switch (pfm_intel_x86_cfg.model) {
> case 58: /* Ivy Bridge (Core i3/i5/i7 3xxx) */
> break;
> + case 62: /* Ivy Bridge EP */
> + break;
> default:
> return PFM_ERR_NOTSUPP;
> }
>
> to support this (instead of requiring LIBPFM_FORCE_PMU to be set). The
> assumption is the IVB EP counters will be a proper superset of IVB
> counters like SNB EP was to SNB.
>
> What we'd really like are official tables for IVB EP, sometime analgous
> to intel_snbep_events.h (e.g., intel_ivbep_events.h).
>
> Does this sound like a logical expectation?
>
Yes, it is. But as I said before. I cannot release the IVB-EP event table
until Intel releases the specs. I will do that immediately when this happens.
So for now, the best we can do is pretend it is equivalent to IVB.
> Steve
>
> On Mon, 29 Apr 2013, Vince Weaver wrote:
>
>> Date: Mon, 29 Apr 2013 14:30:52 -0500
>> From: Vince Weaver <vincent.wea...@maine.edu>
>> To: perfmon2-devel@lists.sourceforge.net
>> Subject: [perfmon2] Ivy Bridge Uncore support
>>
>>
>> Intel Vol 3b says that Ivy Bridge support is more or less the same as
>> Sandy Bridge.
>>
>> The below patch just detects IvyBridge as a SandyBridge uncore. It seems
>> to work.
>>
>> Is this the kind of fix you want? Or would you prefer the SandyBridge
>> table to be copied over as a specific IvyBridge file?
>>
>> Signed-off-by: Vince Weaver <vincent.wea...@maine.edu>
>>
>> diff --git a/lib/pfmlib_intel_snb_unc.c b/lib/pfmlib_intel_snb_unc.c
>> index f4bfaba..00d7b86 100644
>> --- a/lib/pfmlib_intel_snb_unc.c
>> +++ b/lib/pfmlib_intel_snb_unc.c
>> @@ -44,6 +44,7 @@ pfm_snb_unc_detect(void *this)
>>
>> switch (pfm_intel_x86_cfg.model) {
>> case 42: /* Sandy Bridge (Core i7 26xx, 25xx) */
>> + case 58: /* Ivy Bridge */
>> break;
>> default:
>> return PFM_ERR_NOTSUPP;
>>
>>
>> ------------------------------------------------------------------------------
>> Try New Relic Now & We'll Send You this Cool Shirt
>> New Relic is the only SaaS-based application performance monitoring service
>> that delivers powerful full stack analytics. Optimize and monitor your
>> browser, app, & servers with just a few lines of code. Try New Relic
>> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
>> _______________________________________________
>> perfmon2-devel mailing list
>> perfmon2-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
>>
>
>
> ------------------------------------------------------------------------------
> Try New Relic Now & We'll Send You this Cool Shirt
> New Relic is the only SaaS-based application performance monitoring service
> that delivers powerful full stack analytics. Optimize and monitor your
> browser, app, & servers with just a few lines of code. Try New Relic
> and get this awesome Nerd Life shirt! http://p.sf.net/sfu/newrelic_d2d_apr
> _______________________________________________
> perfmon2-devel mailing list
> perfmon2-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
------------------------------------------------------------------------------
Introducing AppDynamics Lite, a free troubleshooting tool for Java/.NET
Get 100% visibility into your production application - at no cost.
Code-level diagnostics for performance bottlenecks with <2% overhead
Download for free and get started troubleshooting in minutes.
http://p.sf.net/sfu/appdyn_d2d_ap1
_______________________________________________
perfmon2-devel mailing list
perfmon2-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
