1908 1909 1910 performance optimizations 1 #1916
Conversation
stmcgovern
force-pushed
the
1908-1909-1910-performance-optimizations-1
branch
from
dcf5234 to
b189a58
Compare
Pipelines resultsPR tests (gcc-12, ubuntu, mpich) Build for 5a7e810 PR tests (gcc-5, ubuntu, mpich) Build for 5a7e810 PR tests (clang-3.9, ubuntu, mpich) Build for 5a7e810 PR tests (gcc-10, ubuntu, openmpi, no LB) Build for 5a7e810 PR tests (clang-5.0, ubuntu, mpich) Build for 5a7e810 PR tests (gcc-6, ubuntu, mpich) Build for 5a7e810 PR tests (gcc-9, ubuntu, mpich, zoltan) Build for 5a7e810 PR tests (clang-9, ubuntu, mpich) Build for 5a7e810 PR tests (clang-13, alpine, mpich) Build for 5a7e810 PR tests (clang-11, ubuntu, mpich) Build for 5a7e810 PR tests (gcc-8, ubuntu, mpich, address sanitizer) Build for 5a7e810 PR tests (nvidia cuda 11.0, ubuntu, mpich) Build for 5a7e810 PR tests (clang-12, ubuntu, mpich) Build for 5a7e810 PR tests (nvidia cuda 10.1, ubuntu, mpich) Build for 5a7e810 PR tests (clang-14, ubuntu, mpich) Build for 5a7e810 PR tests (intel icpx, ubuntu, mpich) Build for 5a7e810 PR tests (clang-10, ubuntu, mpich) Build for 5a7e810 PR tests (clang-13, ubuntu, mpich) Build for 5a7e810 PR tests (gcc-11, ubuntu, mpich) Build for 5a7e810 PR tests (intel icpc, ubuntu, mpich) Build for 5a7e810 |
stmcgovern
force-pushed
the
1908-1909-1910-performance-optimizations-1
branch
from
b189a58 to
0bad323
Compare
stmcgovern
force-pushed
the
1908-1909-1910-performance-optimizations-1
branch
from
0bad323 to
5a7e810
Compare
Codecov Report
@@ Coverage Diff @@
## develop #1916 +/- ##
========================================
Coverage 84.39% 84.40%
========================================
Files 761 761
Lines 26866 26870 +4
========================================
+ Hits 22673 22679 +6
+ Misses 4193 4191 -2
|
stmcgovern
requested review from
nlslatt,
lifflander,
cz4rs,
JacobDomagala,
jstrzebonski,
nmm0,
PhilMiller and
thearusable
| @@ -91,6 +95,7 @@ | |||
| vt::config::CtxEnum::ctx, \ | |||
| vt::config::ModeEnum::inmode \ | |||
| >::apply(vt_debug_argument_option(cat) or force, __VA_ARGS__) | |||
| #endif | |||
There was a problem hiding this comment.
Though it's small, I think it's preferable to put this in a separate PR, for easier traceability.
| @@ -152,6 +152,10 @@ ActiveMessenger::ActiveMessenger() | |||
| }; | |||
| } | |||
|
|
|||
| void ActiveMessenger::initialize() { | |||
| comm_ = theContext()->getComm(); | |||
| } | |||
There was a problem hiding this comment.
Did we resolve the question of whether the communicator will always be initialized by the time this method gets called?
| @@ -83,6 +83,10 @@ | |||
| #define vt_debug_argument_option(opt) \ | |||
| ::vt::debug::preConfig()->vt_debug_ ## opt | |||
|
|
|||
| #if vt_check_enabled(production_build) | |||
| #define vt_debug_print_impl(force, inconfig, inmode, cat, ctx, ...) \ | |||
| vt_force_use(__VA_ARGS__) | |||
There was a problem hiding this comment.
| vt_force_use(__VA_ARGS__) | |
| do { if (false) vt_force_use(__VA_ARGS__); } while (false); |
Closes #1908.
Closes #1909.
Closes #1910.
First PR of multiple for fine-grained performance enhancements.