1938 remove unneeded calls to mpi wtime in trace and lb #1998

stmcgovern · 2022-10-18T16:40:17Z

Closes #1938.
~~Pass the current time from the scheduler to LB and Trace contexts.~~
This has been modified to allow the scheduler to hold a current_sched_time_ and then the appropriate runnable contexts (here LB and Trace) can query their start time from the scheduler. This change was primarily the result of not wanting to force all runnable contexts to take a time (even in contexts where that is inappropriate). Still targets the issue by reducing calls to MPI_Wtime (although increased dereferencing of the theSched())

github-actions · 2022-10-18T17:35:09Z

Pipelines results

PR tests (gcc-12, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (gcc-5, ubuntu, mpich)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-3.9, ubuntu, mpich)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-5.0, ubuntu, mpich)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (gcc-10, ubuntu, openmpi, no LB)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (gcc-6, ubuntu, mpich)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-9, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-13, alpine, mpich)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (gcc-9, ubuntu, mpich, zoltan)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (gcc-8, ubuntu, mpich, address sanitizer)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-11, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-12, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (intel icpx, ubuntu, mpich)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-13, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (nvidia cuda 10.1, ubuntu, mpich)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (nvidia cuda 11.0, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
FAILED: tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o 
/usr/bin/ccache /nvcc_wrapper/build/nvcc_wrapper -DJSON_USE_IMPLICIT_CONVERSIONS=1 -DVT_NO_COLOR_ENABLED -I/vt/tests/unit -I/vt/lib/CLI -I/vt/lib/json/include -I/vt/lib/brotli/c/include -Irelease -I/vt/src -isystem /vt/tests/extern/googletest/googletest/include -isystem /vt/tests/extern/googletest/googletest -isystem /vt/lib/fmt/include -isystem /vt/lib/EngFormat-Cpp/include -isystem /build/checkpoint/install/include -isystem /build/detector/install/include -Wno-deprecated-gpu-targets -O3 -DNDEBUG -fdiagnostics-color=always -Wall -pedantic -Wshadow -Wno-unknown-pragmas -Wsign-compare -ftemplate-backtrace-limit=100 -Werror -fPIC -std=c++1z -MD -MT tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o -MF tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o.d -o tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o -c tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
/build/checkpoint/install/include/checkpoint/container/tuple_serialize.h(86): error%0D%0A%0D%0A%0D%0A ==> And there is more. Read log. <==

Build log

PR tests (clang-14, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (clang-10, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (gcc-11, ubuntu, mpich, json schema test)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (intel icpc, ubuntu, mpich)

Build for d46c6d1 (2023-02-15 01:01:05 UTC)

intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.
intel-cc: remark #10441: The Intel(R) C++ Compiler Classic (ICC) is deprecated and will be removed from product release in the second half of 2023. The Intel(R) oneAPI DPC++/C++ Compiler (ICX) is the recommended compiler moving forward. Please transition to use this compiler. Use '-diag-disable=10441' to disable this message.-%0D%0A%0D%0A%0D%0A ==> And there is more. Read log. <==

Build log

PR tests (gcc-7, ubuntu, mpich, trace runtime, LB)

Build for 5046d72 (2022-11-17 20:24:02 UTC)

Compilation - successful

Testing - passed

Build log

PR tests (nvidia cuda 11.2, ubuntu, mpich)

Build for ( UTC)

nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
FAILED: tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o 
/usr/bin/ccache /nvcc_wrapper/build/nvcc_wrapper -DJSON_USE_IMPLICIT_CONVERSIONS=1 -DVT_NO_COLOR_ENABLED -I/vt/tests/unit -I/vt/lib/CLI -I/vt/lib/json/include -I/vt/lib/brotli/c/include -Irelease -I/vt/src -isystem /vt/tests/extern/googletest/googletest/include -isystem /vt/tests/extern/googletest/googletest -isystem /vt/lib/fmt/include -isystem /vt/lib/EngFormat-Cpp/include -isystem /build/checkpoint/install/include -isystem /build/detector/install/include -Wno-deprecated-gpu-targets -O3 -DNDEBUG -fdiagnostics-color=always -Wall -pedantic -Wshadow -Wno-unknown-pragmas -Wsign-compare -ftemplate-backtrace-limit=100 -Werror -fPIC -std=c++1z -MD -MT tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o -MF tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o.d -o tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx.o -c tests/CMakeFiles/epoch_nompi.dir/Unity/unity_0_cxx.cxx
nvcc_wrapper does not accept standard flags -std=c++1z since partial standard flags and standards after C++17 are not supported. nvcc_wrapper will use -std=c++14 instead. It is undefined behavior to use this flag. This should only be occurring during CMake configuration.
/build/checkpoint/install/include/checkpoint/container/tuple_serialize.h(86): error: namespace "std" has no member "launder"

/build/checkpoint/install/include/checkpoint/container/tuple_serialize.h(86): error: type name is not allowed

/vt/src/vt/runnable/runnable.h(238): warning: constexpr if statements are a C++17 feature

/vt/src/vt/runnable/runnable.h(238): error: namespace "std" has no member "is_void_v"

/vt/src/vt/runnable/runnable.h(238): error: namespace "std" has no member "invoke_result_t"

/vt/src/vt/runnable/runnable.h(238): error: type name is not allowed-=-%0D%0A%0D%0A%0D%0A ==> And there is more. Read log. <==

Build log

codecov · 2022-10-25T00:27:47Z

Codecov Report

Merging #1998 (9fe7f0a) into develop (54ad63d) will decrease coverage by 1.87%.
The diff coverage is 66.66%.

❗ Current head 9fe7f0a differs from pull request most recent head d46c6d1. Consider uploading reports for the commit d46c6d1 to get more accurate results

@@             Coverage Diff             @@
##           develop    #1998      +/-   ##
===========================================
- Coverage    84.86%   82.99%   -1.87%     
===========================================
  Files          720      731      +11     
  Lines        25674    25851     +177     
===========================================
- Hits         21788    21456     -332     
- Misses        3886     4395     +509

Impacted Files	Coverage Δ
src/vt/elm/elm_lb_data.h	`100.00% <ø> (ø)`
src/vt/parameterization/parameterization.h	`100.00% <ø> (ø)`
src/vt/scheduler/scheduler.cc	`67.87% <0.00%> (-0.84%)`	⬇️
src/vt/trace/trace.cc	`49.71% <ø> (ø)`
src/vt/trace/trace.h	`100.00% <ø> (ø)`
src/vt/rdma/state/rdma_state.cc	`37.06% <50.00%> (ø)`
src/vt/context/runnable_context/lb_data.cc	`70.00% <100.00%> (+1.57%)`	⬆️
src/vt/context/runnable_context/trace.cc	`71.42% <100.00%> (+1.42%)`	⬆️
src/vt/elm/elm_lb_data.cc	`87.91% <100.00%> (-0.14%)`	⬇️
src/vt/runnable/invoke.h	`100.00% <100.00%> (ø)`
... and 134 more

src/vt/event/event.h

src/vt/runnable/runnable.cc

src/vt/runnable/runnable.h

src/vt/vrt/collection/balance/baselb/baselb.cc

src/vt/vrt/collection/manager.cc

PhilMiller · 2022-10-25T01:26:24Z

I would approve the first 3 commits of this PR as resolving 1938 (modulo my comments about default parameters). The latter two commits seem to be less motivated, and less clearly correct.

stmcgovern · 2022-10-25T16:22:45Z

Thanks for your comments @PhilMiller. I'll drop the last 2 commits and see about the default arguments.

lifflander

Looks good!

PhilMiller

Frankly, I was much happier with the approach of just passing a time down the call stack.

This PR now adds a bunch of calls to timing::getCurrentTime, though I haven't checked whether they're in specifically cold paths.

This change in approach may have been discussed, but that discussion is supposed to be memorialized in comments here. I'll note that the PR description no longer quite matches the code.

Let's talk about this in the meeting this week.

PhilMiller · 2022-11-29T18:18:15Z

src/vt/scheduler/scheduler.cc

@@ -258,7 +258,8 @@ void Scheduler::runSchedulerOnceImpl(bool msg_only) {
    }
  } else if (work_queue_.empty()) {
    if (curRT->needsCurrentTime()) {
-      runProgress(msg_only, timing::getCurrentTime());
+      current_sched_time_ = timing::getCurrentTime();
+      runProgress(msg_only, current_sched_time_);


This may not be worth changing, but since runProgress is a sibling member, it doesn't need to take current_sched_time_ as an argument, since it can access it directly.

stmcgovern · 2022-11-29T20:06:58Z

Frankly, I was much happier with the approach of just passing a time down the call stack.

This PR now adds a bunch of calls to timing::getCurrentTime, though I haven't checked whether they're in specifically cold paths.

This change in approach may have been discussed, but that discussion is supposed to be memorialized in comments here. I'll note that the PR description no longer quite matches the code.

Let's talk about this in the meeting this week.

Thanks @PhilMiller. Yes, the idea was to just pass time down the call stack. After removing the default arguments (per your first review), then a time argument needed to be passed through to any runnable (void RunnableNew::run(). Aside from complicating the RunnableNew interface, @lifflander remarked that not all runnables need a time. Then the change of tactic to have the scheduler store a time and allow the appropriate runnable contexts (in particular LB and trace) to ask for a start time was adopted.

theTrace()->beginProcessing called timing::getCurrentTime by default. Removing this implicit use required getting the current time explicitly at certain call sites.

I'll update the PR description to reflect these changes.

PhilMiller · 2022-11-29T21:51:38Z

src/vt/runnable/invoke.h

@@ -140,7 +140,7 @@ static trace::TraceProcessingTag BeginProcessingInvokeEvent() {
  const auto trace_event = theTrace()->messageCreation(trace_id, 0);
  const auto from_node = theContext()->getNode();

-  return theTrace()->beginProcessing(trace_id, 0, trace_event, from_node);
+  return theTrace()->beginProcessing(trace_id, 0, trace_event, from_node, timing::getCurrentTime());


This may actually illustrate a case for why I like the 'pass the time as an argument to contexts' approach better - if we want to add LB attribution for methods called through invoke, we would want to share this timer call for the start time, rather than have it reference the scheduler's recorded time, or need another new interface for the divergent case.

PhilMiller · 2023-02-07T23:23:49Z

src/vt/scheduler/base_unit.cc

@@ -62,6 +62,7 @@ void BaseUnit::execute() {
    #endif
  } else if (work_) {
    work_();
+    theSched()->setRecentTimeToStale();


Should this be outside the if / else if, or duplicated in both arms? Why shouldn't the time be updated after we've run stuff in a thread?

Yeah, it's in the run function (runnable.cc line 183). Should we just pull it out to here (outside the if/else)?

Left as is, since the runnable branch will also need to get a finish time.

PhilMiller · 2023-02-08T01:09:07Z

I think I'm generally ok with this, and just need to see the couple comments addressed.

This does need to be run through the same fine-grain overhead benchmarks we used to identify time calls as a concern to begin with. I'd like to be confident that added calls to theSched() aren't losing us whatever ground we're gaining, or even a large fraction of it. If they are, then I'd push for accepting that every context is going to take a time argument, useful or not, and shifting the logic for when to check the clock and what time to pass down the stack more thoroughly into the scheduler.

…ableNew

…d adjusting for CI

PhilMiller · 2023-02-14T18:50:18Z

Some runs of the microbenchmark were getting ~30% reduction in time, directly attributable to reduced time spent in PMPI_Wtime. So, I think we're good on achieving the goal here.

stmcgovern · 2023-02-15T00:00:39Z

Improvements to the tests will be pursued in VT #2017

stmcgovern linked an issue Oct 18, 2022 that may be closed by this pull request

Remove unneeded calls to MPI_Wtime in Trace and LB #1938

Closed

nlslatt mentioned this pull request Oct 18, 2022

Meeting Agenda [do not close] #925

Open

stmcgovern marked this pull request as ready for review October 24, 2022 23:32

stmcgovern force-pushed the 1938-remove-unneeded-calls-to-mpi_wtime-in-trace-and-lb branch from 57775d0 to 6ef8a04 Compare October 24, 2022 23:34

PhilMiller reviewed Oct 25, 2022

View reviewed changes

src/vt/event/event.h Outdated Show resolved Hide resolved

PhilMiller reviewed Oct 25, 2022

View reviewed changes

src/vt/runnable/runnable.cc Outdated Show resolved Hide resolved

PhilMiller reviewed Oct 25, 2022

View reviewed changes

src/vt/runnable/runnable.h Outdated Show resolved Hide resolved

PhilMiller reviewed Oct 25, 2022

View reviewed changes

src/vt/vrt/collection/balance/baselb/baselb.cc Outdated Show resolved Hide resolved

PhilMiller reviewed Oct 25, 2022

View reviewed changes

src/vt/vrt/collection/manager.cc Outdated Show resolved Hide resolved

PhilMiller mentioned this pull request Oct 25, 2022

Rename begin/end to something like start/finish #2002

Closed

stmcgovern force-pushed the 1938-remove-unneeded-calls-to-mpi_wtime-in-trace-and-lb branch 2 times, most recently from 7f463fa to 31b91f3 Compare November 15, 2022 23:55

stmcgovern requested review from lifflander, cz4rs, JacobDomagala, nlslatt, nmm0, PhilMiller and thearusable November 17, 2022 20:20

stmcgovern force-pushed the 1938-remove-unneeded-calls-to-mpi_wtime-in-trace-and-lb branch from 31b91f3 to 5046d72 Compare November 17, 2022 20:24

stmcgovern force-pushed the 1938-remove-unneeded-calls-to-mpi_wtime-in-trace-and-lb branch from 5046d72 to 5880acc Compare November 29, 2022 00:44

lifflander previously approved these changes Nov 29, 2022

View reviewed changes

PhilMiller previously requested changes Nov 29, 2022

View reviewed changes

PhilMiller reviewed Nov 29, 2022

View reviewed changes

PhilMiller reviewed Feb 7, 2023

View reviewed changes

stmcgovern and others added 18 commits February 13, 2023 23:52

#1938: pass time as an argument to runnable context LBData

9b0f56f

#1938: pass time as argument to start stop in ElementLBData

664304a

#1938: fix serializer list - add special_progress_

b90b118

#1938: pass time as argument in RunnableNew member functions

bbe8629

#1938: add recent_time_ and get function to scheduler

d55db73

#1938: use a flag to decide when to update recent time

09af68a

#1938: use recent time in runProgress

29db02e

#1938: pass start_time down in RunnableNew::run and end_time

87baa47

#1938: set recent time to stale when work is done by a lambda

70fd5dc

#1938: adjust whitespace

24d6ec1

#1938: change trace's beginProcessing interface and uses

a76fd5f

#1938: change interface and calls to endProcessing

164f013

#1938: adjust runnable context for trace to use passed in time

1f75321

#1938: runnable: add interface for testing to get the underlying Runn…

1585b13

…ableNew

#1938: tests: write test for LB data timings

9d78f6f

#1938: remove whitespace

e343c02

#1938: runLambda uses time conservatively

81249b2

#1938: add another test for cumulative elapsed time - values will nee…

44f3132

…d adjusting for CI

stmcgovern force-pushed the 1938-remove-unneeded-calls-to-mpi_wtime-in-trace-and-lb branch from e6bb911 to 44f3132 Compare February 14, 2023 16:53

#1938: fix license

af8b217

#1938: tests: set comparison margin and clean up test

a4ea3a8

stmcgovern marked this pull request as ready for review February 14, 2023 23:20

lifflander approved these changes Feb 14, 2023

View reviewed changes

#1938: add lb vt_check and loosen requirements for tests

d46c6d1

lifflander merged commit 893ca59 into develop Feb 15, 2023

nmm0 mentioned this pull request May 11, 2023

invalid trace timings #2147

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

1938 remove unneeded calls to mpi wtime in trace and lb #1998

1938 remove unneeded calls to mpi wtime in trace and lb #1998

stmcgovern commented Oct 18, 2022 •

edited

Loading

github-actions bot commented Oct 18, 2022 •

edited

Loading

codecov bot commented Oct 25, 2022 •

edited

Loading

PhilMiller commented Oct 25, 2022

stmcgovern commented Oct 25, 2022

lifflander left a comment

PhilMiller left a comment

PhilMiller Nov 29, 2022

stmcgovern commented Nov 29, 2022

PhilMiller Nov 29, 2022

PhilMiller Feb 7, 2023

stmcgovern Feb 9, 2023

stmcgovern Feb 14, 2023

PhilMiller commented Feb 8, 2023

PhilMiller commented Feb 14, 2023

stmcgovern commented Feb 15, 2023

1938 remove unneeded calls to mpi wtime in trace and lb #1998

1938 remove unneeded calls to mpi wtime in trace and lb #1998

Conversation

stmcgovern commented Oct 18, 2022 • edited Loading

github-actions bot commented Oct 18, 2022 • edited Loading

Pipelines results

codecov bot commented Oct 25, 2022 • edited Loading

Codecov Report

PhilMiller commented Oct 25, 2022

stmcgovern commented Oct 25, 2022

lifflander left a comment

Choose a reason for hiding this comment

PhilMiller left a comment

Choose a reason for hiding this comment

PhilMiller Nov 29, 2022

Choose a reason for hiding this comment

stmcgovern commented Nov 29, 2022

PhilMiller Nov 29, 2022

Choose a reason for hiding this comment

PhilMiller Feb 7, 2023

Choose a reason for hiding this comment

stmcgovern Feb 9, 2023

Choose a reason for hiding this comment

stmcgovern Feb 14, 2023

Choose a reason for hiding this comment

PhilMiller commented Feb 8, 2023

PhilMiller commented Feb 14, 2023

stmcgovern commented Feb 15, 2023

stmcgovern commented Oct 18, 2022 •

edited

Loading

github-actions bot commented Oct 18, 2022 •

edited

Loading

codecov bot commented Oct 25, 2022 •

edited

Loading