You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@datafusion.apache.org by gi...@apache.org on 2026/03/30 07:20:25 UTC
(datafusion) branch main updated: chore(benchmark): Fix/update compile profile benchmark (#21223)
This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/datafusion.git
The following commit(s) were added to refs/heads/main by this push:
new a0869e95ed chore(benchmark): Fix/update compile profile benchmark (#21223)
a0869e95ed is described below
commit a0869e95ed65b55bda79ce0d3067c43b4724b273
Author: Yongting You <20...@gmail.com>
AuthorDate: Mon Mar 30 15:19:59 2026 +0800
chore(benchmark): Fix/update compile profile benchmark (#21223)
## Which issue does this PR close?
<!--
We generally require a GitHub issue to be filed for all bug fixes and
enhancements and this helps us generate change logs for our releases.
You can link an issue to this PR using the GitHub syntax. For example
`Closes #123` indicates that this PR will close issue #123.
-->
- Closes #.
## Rationale for this change
<!--
Why are you proposing this change? If this is already explained clearly
in the issue then this section is not needed.
Explaining clearly why changes are proposed helps reviewers understand
your changes and offer better suggestions for fixes.
-->
The existing compile profile benchmark is not able to run due to the
benchmark binary change, and recently there are several new profiles
added. This PR updates the benchmark.
## What changes are included in this PR?
<!--
There is no need to duplicate the description in the issue here but it
is sometimes worth providing a summary of the individual changes in this
PR.
-->
## Are these changes tested?
<!--
We typically require tests for all PRs in order to:
1. Prevent the code from being accidentally broken by subsequent changes
2. Serve as another way to document the expected behavior of the code
If tests are not included in your PR, please explain why (for example,
are they covered by existing tests)?
-->
I have tested it locally
```
yongting@Yongtings-MacBook-Pro-2 ~/C/d/benchmarks (update-compile-prof…)> ./bench.sh run compile_profile
...
Summary
Profile Compile Run Size
---------------------------------------------------
dev 80.81s 11.26s 238.6MB
release 370.69s 1.01s 62.4MB
ci 7.51s 7.69s 194.5MB
ci-optimized 215.97s 1.13s 114.0MB
release-nonlto 15.95s 1.05s 128.0MB
profiling 953.66s 0.98s 75.9MB
```
Also, there is a ongoing work to run benchmarks in CI in case it break
again: https://github.com/apache/datafusion/issues/21165
## Are there any user-facing changes?
<!--
If there are user-facing changes then we may require documentation to be
updated before approving the PR.
-->
<!--
If there are any breaking changes to public APIs, please add the `api
change` label.
-->
---
benchmarks/README.md | 2 +-
benchmarks/compile_profile.py | 43 +++++++++++++++++++++++++++++++++----------
2 files changed, 34 insertions(+), 11 deletions(-)
diff --git a/benchmarks/README.md b/benchmarks/README.md
index 3aa4f4bb86..df602ea538 100644
--- a/benchmarks/README.md
+++ b/benchmarks/README.md
@@ -95,7 +95,7 @@ Generate the data required for the compile profile helper (TPC-H SF=1):
./bench.sh data compile_profile
```
-Run the benchmark across all default Cargo profiles (`dev`, `release`, `ci`, `release-nonlto`):
+Run the benchmark across all default Cargo profiles (`dev`, `release`, `ci`, `ci-optimized`, `release-nonlto`, `profiling`):
```shell
./bench.sh run compile_profile
diff --git a/benchmarks/compile_profile.py b/benchmarks/compile_profile.py
index ae51de9493..a85e15ddac 100644
--- a/benchmarks/compile_profile.py
+++ b/benchmarks/compile_profile.py
@@ -19,8 +19,10 @@
"""Compile profile benchmark runner for DataFusion.
-Builds the `tpch` benchmark binary with several Cargo profiles (e.g. `--release` or `--profile ci`), runs the full TPC-H suite against the Parquet data under `benchmarks/data/tpch_sf1`, and reports compile time, execution time, and resulting
-binary size.
+Builds the `dfbench` benchmark binary with several Cargo profiles
+(e.g. `--release` or `--profile ci`), runs the full TPC-H suite against
+the Parquet data under `benchmarks/data/tpch_sf1`, and reports compile
+time, execution time, and resulting binary size.
See `benchmarks/README.md` for usages.
"""
@@ -40,12 +42,15 @@ DEFAULT_DATA_DIR = REPO_ROOT / "benchmarks" / "data" / "tpch_sf1"
DEFAULT_ITERATIONS = 1
DEFAULT_FORMAT = "parquet"
DEFAULT_PARTITIONS: int | None = None
-TPCH_BINARY = "tpch.exe" if os.name == "nt" else "tpch"
+BENCHMARK_PACKAGE = "datafusion-benchmarks"
+BENCHMARK_BINARY = "dfbench.exe" if os.name == "nt" else "dfbench"
PROFILE_TARGET_DIR = {
"dev": "debug",
"release": "release",
"ci": "ci",
+ "ci-optimized": "ci-optimized",
"release-nonlto": "release-nonlto",
+ "profiling": "profiling",
}
@@ -62,7 +67,10 @@ def parse_args() -> argparse.Namespace:
"--profiles",
nargs="+",
default=list(PROFILE_TARGET_DIR.keys()),
- help="Cargo profiles to test (default: dev release ci release-nonlto)",
+ help=(
+ "Cargo profiles to test "
+ "(default: dev release ci ci-optimized release-nonlto profiling)"
+ ),
)
parser.add_argument(
"--data",
@@ -84,9 +92,25 @@ def timed_run(command: Iterable[str]) -> float:
def cargo_build(profile: str) -> float:
if profile == "dev":
- command = ["cargo", "build", "--bin", "tpch"]
+ command = [
+ "cargo",
+ "build",
+ "--package",
+ BENCHMARK_PACKAGE,
+ "--bin",
+ "dfbench",
+ ]
else:
- command = ["cargo", "build", "--profile", profile, "--bin", "tpch"]
+ command = [
+ "cargo",
+ "build",
+ "--profile",
+ profile,
+ "--package",
+ BENCHMARK_PACKAGE,
+ "--bin",
+ "dfbench",
+ ]
return timed_run(command)
@@ -102,14 +126,13 @@ def run_benchmark(profile: str, data_path: Path) -> float:
binary_dir = PROFILE_TARGET_DIR.get(profile)
if not binary_dir:
raise ValueError(f"unknown profile '{profile}'")
- binary_path = REPO_ROOT / "target" / binary_dir / TPCH_BINARY
+ binary_path = REPO_ROOT / "target" / binary_dir / BENCHMARK_BINARY
if not binary_path.exists():
raise FileNotFoundError(f"compiled binary not found at {binary_path}")
command = [
str(binary_path),
- "benchmark",
- "datafusion",
+ "tpch",
"--iterations",
str(DEFAULT_ITERATIONS),
"--path",
@@ -132,7 +155,7 @@ def run_benchmark(profile: str, data_path: Path) -> float:
def binary_size(profile: str) -> int:
binary_dir = PROFILE_TARGET_DIR[profile]
- binary_path = REPO_ROOT / "target" / binary_dir / TPCH_BINARY
+ binary_path = REPO_ROOT / "target" / binary_dir / BENCHMARK_BINARY
return binary_path.stat().st_size
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@datafusion.apache.org
For additional commands, e-mail: commits-help@datafusion.apache.org