added optional scheduler blocking #376
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
RalphSteinhagen
force-pushed
the
optionalIoBlocking
branch
2 times, most recently
from
a7d03ab
to
9b5a7f7
Compare
|
Needed to push to fix some of my last-minute 'verschlimmbesserungen'. 😬 Did I mention I like unit-tests... saved my day. |
RalphSteinhagen
force-pushed
the
optionalIoBlocking
branch
2 times, most recently
from
ffee047
to
cc99060
Compare
RalphSteinhagen
added this to the CALL#4 - OP OpenDigitizer Integration/Deployment milestone
RalphSteinhagen
force-pushed
the
optionalIoBlocking
branch
from
5625636
to
b693af1
Compare
ivan-cukic
approved these changes
Jul 12, 2024
RalphSteinhagen
force-pushed
the
optionalIoBlocking
branch
2 times, most recently
from
4898b42
to
51e3540
Compare
... in case there is no data to be processed and/or no IO activity. * added missing processBulk(..) user return status handling * changed ExecutionPolicy to 'enum class' (type safety) -> needed some adjustments in other unrelated unit-tests * added msg-to-stream processing ratio * simplified SchedulerBase and streamlined single- and multi-threaded handling * added provisions for runtime graph modifications Signed-off-by: rstein <[email protected]> Signed-off-by: Ralph J. Steinhagen <[email protected]> Signed-off-by: rstein <[email protected]> Signed-off-by: Ralph J. Steinhagen <[email protected]>
Signed-off-by: Ralph J. Steinhagen <[email protected]>
Signed-off-by: Ralph J. Steinhagen <[email protected]>
Signed-off-by: rstein <[email protected]> Signed-off-by: Ralph J. Steinhagen <[email protected]>
... to have the BusyLoopBlock/Graph settle into a steady state during the initially spin-up/filling of buffers Signed-off-by: Ralph J. Steinhagen <[email protected]>
before message were at least once (often twice as often) called per processing of streaming data (which was/is a bit excessive) and now reduced to favour low-latency on sync/stream processing. Signed-off-by: Ralph J. Steinhagen <[email protected]>
Signed-off-by: Ralph J. Steinhagen <[email protected]>
RalphSteinhagen
force-pushed
the
optionalIoBlocking
branch
from
51e3540
to
3cab7bc
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
... in case there is no data to be processed and/or no IO activity.
qa_PerformanceMonitor benchmark (GCC14, no Tags, 2" runtime) before:
after (this PR);
There have been some performance gains but the 'return status handling' added some overhead. Not critical but perhaps some remaining potential. The limit seems to be ~500 MS/s on my machine.
N.B. I had to disable some expect-statements in qa_GraphMessages (annotated 'FIXME') due to some missing runtime graph modification implementation that I believe @ivan-cukic is working on parallel in these issues/branches: