Skip to content

Changelog

Thomas Roehl edited this page Jul 23, 2015 · 4 revisions

Release 4.0.1

  • likwid-bench: Iteration determination is done serially
  • likwid-bench: Manual selection of iterations possible
  • likwid-perfctr: Set cpuset to all CPUs not only the first
  • likwid-pin: Set cpuset to all CPUs not only the first
  • likwid-accuracy.py: Enhanced plotting functions, use only instrumented likwid-bench
  • likwid-accessD: Check for allowed register for PCI accesses
  • Add models HASWELL_M1 (0x45) and HASWELL_M2 (0x46) to likwid-powermeter and likwid-accessD
  • New test application using Cilk and Marker API
  • New test application using C++11 threads and Marker API
  • likwid-agent: gmetric version check for --group option and s/\s*/_/ in metric names
  • likwid-powermeter: Print RAPL domain name
  • Marker API: Initialize access already at likwid_markerInit()
  • Marker API: likwid_markerThreadInit() only pins if not already pinned

Release 4.0.0

  • Support for Intel Broadwell
  • Uncore support for all Uncore-aware architectures
    • Nehalem (EX)
    • Westmere (EX)
    • SandyBridge EP
    • IvyBridge EP
    • Haswell EP
  • Measure multiple event sets in a round-robin fashion (no multiplexing!)
  • Event options to filter the counter increments
  • Whole LIKWID functionality is exposed as API for C/C++ and Lua
  • New functions in the Marker API to switch event sets and get intermediate results
  • Topology code relies on hwloc. CPUID is still included but only as fallback
  • Most LIKWID applications are written in Lua (only exception likwid-bench)
  • Monitoring daemon likwid-agent with multiple output backends
  • More performance groups

Release 3.1

  • Full support for Intel Xeon Phi
  • Support for Intel IvyBridge-EP (partial Uncore support)
  • Support for Intel Haswell
  • Support for AMD Kabini architecture
  • Improved Intel Atom support
  • Adding expression based thread group syntax for all LIKWID tools
  • Add scatter option to thread groups
  • Add possibility to dump topology info to cfg file. This enables simpler usage of LIKWID in a restricted cpuset
  • Using the Marker API a variable number of threads may enter a region now
  • Simple lock file based control to synchronize Monitoring and Profiling in a cluster environment
  • Using custom event sets on Intel Processors fixed counters are automatically added and connected derived metrics runtime, clock and CPI are output
  • Oversubscription of processors with round robin placement is now default
  • Add event for TM/TM2 temperature sensors on Intel processors
  • Improved RDTSC timing measurement
  • Review of performance groups
  • Performance improvements in Marker API
  • Lots of bugfixes and improvements

Release 3.0

  • New application likwid-memsweeper to cleanup ccNUMA memory domains. This functionality is also integrated in likwid-pin.
  • Support for Intel SandyBridge Uncore (partial)
  • Support for Intel IvyBridge (only core part)
  • Initial support for Intel Xeon Phi (KNC)
  • Better support for AMD Interlagos
  • The OpenMP type is now detected automatically while pinning
  • Lots of bugfixes and improvements
  • Marker API works now for threaded code and accessDaemon
  • Uncore support for timeline mode
  • Convenient macro wrapper for Marker API
  • Data volume as new metric in all memory/cache groups
  • Updated Wiki documentation

Release 2.3

  • New application likwid-powermeter to measure Energy consumption on SandyBridge (RAPL) and query turbo mode steps.
  • RAPL counters are also integrated into likwid-perfctr
  • likwid-mpirun: Many improvements and initial support for mvapich2
  • likwid-mpirun has now integrated perfctr support
  • Much improved fortran interface for the marker API
  • Support for Intel SandyBridge
  • Support for AMD Interlagos
  • Initial support for NehalemEX and WestmereEX Uncore
  • Better default options for likwid-pin
  • Lots of Bugfixes and small improvements

Release 2.2

  • New application likwid-perfscope as frontend to timeline mode of likwid-perfctr
  • Initial support for OpenMPI in likwid-mpirun
  • Initial support for Intel SandyBridge (Core) in likwid-perfctr
  • Improvements in likwid-msrD
  • Native Fortran 90 interface for Marker API
  • Completetly rewrite of marker API. Simplified usage. Allows now inclusive and overlapping regions.
  • New output filter subsystem. Allows you to define your own output file formats and filters. Default comes with CSV and XML formats.

Release 2.1

  • likwid-perfctr supports output in files with placeholders for MPI settings
  • improved and extended performance groups in likwid-perfctr
  • msr daemon added for encapsulation of msr device file access (experimental)
  • add numa thread domains (M prefix) in affinity module
  • fix bug in likwid-bench code generation preventing correct loop generation
  • better dependency tracking in build system
  • Extended documentation
  • likwid-perfctr also includes now the pin functionality
  • Again switch back to cpumask sys interface for the cpu to node mapping
  • likwid-mpirun script enables simple pinning of pure MPI or hybrid OpenMP/MPI

Release 2.0

  • likwid-pin and likwid-perfctr supports logical pinning (node, socket, cache group)
  • likwid-perfctr supports multiple instance marker API runs on the same node
  • performance groups in likwid-perfctr are now generated from simple text files
  • likwid-topology also reports on NUMA information
  • Support for second fixed counter on Intel CPUS
  • likwid-pin can set NUMA mempolicy
  • statistical output in likwid-perfctr for threaded results
  • support for Nehalem EX core events in likwid-perfctr
  • likwid-perfctr allows lightweight continous measurements with daemon mode
  • silent mode switch for likwid-pin
Clone this wiki locally