Lines of Code
| Author | Changes | Lines of Code | Lines per Change |
|---|---|---|---|
| Totals | 59 (100.0%) | 369 (100.0%) | 6.2 |
| dav | 5 (8.5%) | 206 (55.8%) | 41.2 |
| ahumphrey | 19 (32.2%) | 135 (36.6%) | 7.1 |
| jas | 32 (54.2%) | 20 (5.4%) | 0.6 |
| allen | 3 (5.1%) | 8 (2.2%) | 2.6 |
Some cleanup and work toward refining and better understanding MPI timings and stats.
Pull ancient VAMPIR weeds.
2 lines of code changed in 4 files:
Remove antiquated TAU support.
1 lines of code changed in 2 files:
Revert c++-11 namespace changes.
1 lines of code changed in 2 files:
Explicit namespace usage as part of the transition to c++-11.
0 lines of code changed in 1 file:
* Fix compiler warning, signed and unsigned comparison.
* Report whether or not MPI message combining is turned on, even when <small_messages> is not found in <Scheduler> block.
* Parallel - report number of MPI processes, and threads per process (if any). The notion of "processor" is dated and now confusing. Update section in user guide to reflect this change.
2 lines of code changed in 2 files:
M Core/Malloc/Allocator.cc
- Purely cosmetic update - differentiate between two error messages that had same text.
M Core/Malloc/new.cc
M Core/Parallel/Parallel.cc
- Purely cosmetic white space/coding standard updates.
- Removed #if sgi code.
M CCA/Components/MPM/ImpMPM.cc
- Cosmetic update in preparation for XML memory fix update.
60 lines of code changed in 1 file:
change the isProc to isProc_macro to better label
0 lines of code changed in 2 files:
This commit is focused on clearing up & simplifying what's happening with task phases (delaying reductions, etc until late in a timestep to avoid out-of-order scheduling of reduction tasks across MPI ranks - deadlock) and MPI communicator duplication (used in conjunction with MPI collective calls due to the lack of support for message tags) in the threaded schedulers. Much of this cleanup and simplification will aid in finishing work on spatial task scheduling.
NOTE: These changes have been run through full nightly RT, so things should be OK...
* Fix detailed DebugStream output/reporting on task phases (threaded schedulers), e.g. when each task phase begins, what is executing, how many tasks, etc.
* Add many cerr and cout locks for threaded scheduler dbg streams. Detailed output is now readable with multiple threads.
* Massive cleanup in TaskGraph, DetailedTasks, ProcessorGroup and Parallel classes. Sort #includes, line up params/args, add comments, whitespace and { }s, misc formatting, etc.
* Some refactoring in threaded schedulers.
92 lines of code changed in 6 files:
Made the following changes:
1. Compress large xml files. Remove sus log.
2. Add Changelog.
3. Add instructions for adding a new component to the build system.
4. Update OSX build in install documentation.
5. Update petsc version in install documentation.
6. Change Petsc version.
7. Remove references to Teem in installation. Remove the build directory
after creating the tarball for the release.
8. Remove Tracker feature.
9. Remove AIX references. Remove Basis/. Remove Teem. Remove sci_defs
include files related to SCIRun.
10. Remove old build dependency versions.
11. Add wasatch_3P 1.6.0 branch.
12. Remove bench.config file, update excludes when building release tar ball.
13. Remove *.orig and *.release files.
14. Update copyright date. Update version to 1.6.
19 lines of code changed in 29 files:
A second pass on scheduler cleanup.
Much better reporting on task queue algorithm, use (or not) of combined MPI messages
Additional documentation on scheduler-related SCI_DEBUG flags.
More code formatting for readability, alphabetize #includes, white space, line up params/args, {}s, etc.
Remove stacks of old (and mostly bogus) debug code that was comented out.
38 lines of code changed in 5 files:
Created macro boolean for determining if processor 0 and thread 0
8 lines of code changed in 1 file:
These are code readability (cosmetic) fixes. A bunch of coding
standards, white space, untabbifying, adding needed {}, and variable
alignment fixes, etc. This includes naming class variables with trailing
'_' (or using 'd_' depending on the way other variables in the file
are named) so that we know they are class variables.
This should not change any answers.
M CCA/Components/Schedulers/CommRecMPI.cc
M CCA/Components/Schedulers/CommRecMPI.h
M CCA/Components/Schedulers/UnifiedScheduler.h
- d_id renamed to d_thread_id.
- A lot of these functions should be virtual... It is confusing as to what is supposed to be
inherited, what is overridden, etc. Someone needs to clean up the entire Scheduler class hierarchy.
M CCA/Components/Schedulers/MPIScheduler.h
M CCA/Components/Schedulers/UnifiedScheduler.cc
M Core/Parallel/PackBufferInfo.cc
M Core/Parallel/BufferInfo.cc
M Core/Parallel/PackBufferInfo.h
- One line functions in .h files traditionally are written on one line (so that the file reader
sees the important information (member function names) and is not immediately distracted by
the implementation).
M Core/Parallel/BufferInfo.h
146 lines of code changed in 4 files: