Commit 80932db8 authored by Misbah Mubarak's avatar Misbah Mubarak

Updating release notes, making debug_cols option configurable

parent 39ef9691
0.6.0 (July 03, 2017)
general:
========
C++ models can now be built and integrated with CODES. The new dragonfly model
has been implemented in C++.
support has been added for building CODES with CoRTex -- a library for translating collectives
to point to point operations. For details see:
https://xgitlab.cels.anl.gov/codes/codes/wikis/codes-cortex-workload
CODES wiki with details on model development, networking, MPI simulation and
more has been added at:
https://xgitlab.cels.anl.gov/codes/codes/wikis/home
Test suite has been extended -- tests for DUMPI trace replay have been added.
Unused variable warnings have been fixed.
networks:
==========
dragonfly network model based on Cray XC topology has been added. The model can
use the network configurations of Theta and Edison systems. Custom network
configurations can be generated using C scripts. For details see:
https://xgitlab.cels.anl.gov/codes/codes/wikis/codes-dragonfly
fat tree network model with support for adaptive and static routing has been
added. The model can support both full and pruned fat tree configurations. For
details, refer to the wiki:
https://xgitlab.cels.anl.gov/codes/codes/wikis/codes-fattree
memory consumption has been reduced by doing lazy allocation of hash memory.
MPI trace replay:
==============
MPI rendezvous protocol can now be replayed in addition to the eager protocol.
The transition point for switching between the two protocols is configurable.
Background network communication using uniform random workload can now be
generated. The traffic generation gets automatically shut off when the main workload
finishes.
Collectives can now be translated into point to point using the CoRTex library.
Performance of MPI_AllReduce is reported when debug_cols option is enabled.
Aggregate and average performance of different message size is reported when
message tracking is enabled (this option is for debugging, works in sequential
mode only).
0.5.2 (July 13, 2016)
Summer of CODES was another huge success! This release was created during the
......
......@@ -34,7 +34,7 @@ static int msg_size_hash_compare(
void *key, struct qhash_head *link);
/* NOTE: Message tracking works in sequential mode only! */
int debug_cols = 0;
static int debug_cols = 0;
int enable_msg_tracking = 0;
tw_lpid TRACK_LP = -1;
......@@ -2287,6 +2287,7 @@ const tw_optdef app_opt [] =
TWOPT_UINT("num_net_traces", num_net_traces, "number of network traces"),
TWOPT_UINT("eager_threshold", EAGER_THRESHOLD, "the transition point for eager/rendezvous protocols (Default 8192)"),
TWOPT_UINT("disable_compute", disable_delay, "disable compute simulation"),
TWOPT_UINT("debug_cols", debug_cols, "completion time of collective operations (currently MPI_AllReduce)"),
TWOPT_UINT("enable_mpi_debug", enable_debug, "enable debugging of MPI sim layer (works with sync=1 only)"),
TWOPT_UINT("sampling_interval", sampling_interval, "sampling interval for MPI operations"),
TWOPT_UINT("enable_sampling", enable_sampling, "enable sampling (only works in sequential mode)"),
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment