Rework configuration and use iteration/invocation #82
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
- will be replaced by an invocation/iteration override Signed-off-by: Stefan Marr <git@stefan-marr.de>
These used to be abort conditions, but are long removed again. Signed-off-by: Stefan Marr <git@stefan-marr.de>
Signed-off-by: Stefan Marr <git@stefan-marr.de>
…on_time Signed-off-by: Stefan Marr <git@stefan-marr.de>
Signed-off-by: Stefan Marr <git@stefan-marr.de>
We never really supported different actions beside benchmarking (profiling, …), and there are no current plans to do so. This can be easily achieved in other ways. For instance by defining a different VM that does profiling and uses a different gauge_adapter. Signed-off-by: Stefan Marr <git@stefan-marr.de>
Signed-off-by: Stefan Marr <git@stefan-marr.de>
- remove number_of_data_points - it is completely replaced by iterations/invocations - move warmup into EXP_RUN_DETAILS - move execute_exclusively into EXP_RUN_DETAILS - add EXP_VARIABLES to benchmark_type_map elements, and move the whole thing to resolve forward dependency - make some more attributes required: gauge_adapter, command, benchmarks - rename `benchmark` to `suites` - add defaults - some defaults can’t be defined in the schema, because it would interfere with our desired semantics of combining settings - improve descriptions - adapt conf files accordingly Signed-off-by: Stefan Marr <git@stefan-marr.de>
- it is only loading and checking the validity of the configuration Signed-off-by: Stefan Marr <git@stefan-marr.de>
This should probably still into an OS-specific package, but currently, only care for Unix-like things anyway. Signed-off-by: Stefan Marr <git@stefan-marr.de>
- make naming more consistent with high-level concepts: Benchmark instead of BenchmarkConfig Signed-off-by: Stefan Marr <git@stefan-marr.de>
This means, we have all top-level settings as overrides for the command line. Signed-off-by: Stefan Marr <git@stefan-marr.de>
The runs concept is gone. Signed-off-by: Stefan Marr <git@stefan-marr.de>
Signed-off-by: Stefan Marr <git@stefan-marr.de>
smarr
force-pushed
the
invocation-iterations
branch
from
cd8040d
to
777ee88
Compare
5 tasks
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.
This PR addresses some tasks for #68.
remove notion of number of data points to control experiments
remove naive quick run optinons
replace old config with max_invocation_time/min_iteration_time to keep terminology consistent
remove some old code
reimplement the processing of the configuration
The biggest change is to reimplement how the configuration file is processed and which elements can have with options. Documentation on these will follow, but the schema is now a good overview of what is possible.
Ordered by priority, the settings are now taken from:
This means, any benchmark specific setting will override the other settings. This can be a change of how configurations are interpreter.
I hope this does not reduce expressiveness, and assume that depending on the use case, it still gives enough degrees of freedom to encode the desired set of experiments, one may just need to move certain settings to another place.