[PATCH 0/10] perf: Add support for PMU events in JSON format

Jiri Olsa jolsa at redhat.com
Thu May 28 21:42:06 AEST 2015


On Wed, May 27, 2015 at 02:23:19PM -0700, Sukadev Bhattiprolu wrote:

SNIP

> 
> At run time, perf identifies the specific events table, based on the model
> of the CPU perf is running on. Perf uses that table to create event aliases
> which would allow the user to specify the event as:
> 
> 	perf stat -e pm_1plus_ppc_cmpl sleep 1
> 
> Note:
> 	- All known events tables for the architecture are included in the
> 	  perf binary.
> 
> 	- Inconsistencies between the JSON files and the mapfile can result
> 	  in build failures in perf (although jevents try to recover from
> 	  some and continue the build by leaving out event aliases).
> 
> 	- For architectures that don't have any JSON files, an empty mapping
> 	  table is created and they should continue to build)
> 
> Thanks to input from Andi Kleen, Jiri Olsa, Namhyung Kim and Ingo Molnar.
> 
> These patches are available from
> 
> 	git at github.com:sukadev/linux.git #branch json-v11

could you please also pull in Andi's changes with x86 events?
it seems to be under his tree branch perf/builtin-json-2,
follwing commits:

b68aa7b0d4e4 perf, tools: Add Broadwell V9 event file
1873bf40be27 perf, tools: Add HaswellX V12 event file
dc9a37d4f96c perf, tools: Add Bonnell V1 event file
f8cce1c4abb6 perf, tools: Add Silvermont V8 event file
c92a43e0a6c9 perf, tools: Add Haswell V18 event file
67ac6d3d8e2f perf, tools: Add IvyBridge V14 event file
3777bd7a8bf2 perf, tools: Add IvyTown V16 event file
21f34ff8ddb3 perf, tools: Add Jaketown V17 event file
22b22b279559 perf, tools: Add SandyBridge V11 event file
542c99a4836d perf, tools: Add WestmereEP-DP V1 event file
b2f0315e50b3 perf, tools: Add WestmereEP-SP V1 event file
82601ad46a58 perf, tools: Add WestmereEX V1 event file
a52987f230e0 perf, tools: Add NehalemEP V1 event file
a8fedce09cc1 perf, tools: Add NehalemEX V1 event file

no need to post them ATM.. just to have all stuff
on one place will ease up testing

Andi, you'd be ok with that I guess ;-)

thanks,
jirka


More information about the Linuxppc-dev mailing list