-
Notifications
You must be signed in to change notification settings - Fork 41
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add optimized ReadAll function #229
Conversation
0b21e08
to
e3c4851
Compare
// specific language governing permissions and limitations | ||
// under the License. | ||
|
||
package iobuf |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Honestly, I'm terrible at naming things. I'm accepting suggestions for a package name or a better place to put this function.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Since we're already fixing linter issues, would be nice to replace how we create temporary directories in tests. Up to you to decide.
I'm fine with the current state of the PR too, given io.ReadAll
was left there on purpose.
This adds an optimized io.ReadAll that uses a bytes.Buffer + io.Copy internally, improving the buffer growth ratio over the runtime append and taking advange of io.WriterTo when available. The only scenario where this optimizes version is slower than io.ReadAll is when reading < 512 bytes of a reader that does not implement io.WriterTo. For now I only updated tests to use the new function as updating the code requires more careful consideration. While at it, migrate from the deprecated ioutil.ReadAll to io.ReadAll. goos: linux goarch: amd64 pkg: github.com/elastic/elastic-agent-libs/iobuf cpu: Intel(R) Xeon(R) CPU E5-2697A v4 @ 2.60GHz BenchmarkReadAll/size_32B/io.ReadAll/WriterTo-32 7479658 162.4 ns/op 512 B/op 1 allocs/op BenchmarkReadAll/size_32B/io.ReadAll/Reader-32 7000012 169.3 ns/op 512 B/op 1 allocs/op BenchmarkReadAll/size_32B/ReadAll/WriterTo-32 10630838 112.3 ns/op 112 B/op 2 allocs/op BenchmarkReadAll/size_32B/ReadAll/Reader-32 2111246 567.3 ns/op 1584 B/op 3 allocs/op BenchmarkReadAll/size_64B/io.ReadAll/WriterTo-32 7154652 163.2 ns/op 512 B/op 1 allocs/op BenchmarkReadAll/size_64B/io.ReadAll/Reader-32 7223264 166.1 ns/op 512 B/op 1 allocs/op BenchmarkReadAll/size_64B/ReadAll/WriterTo-32 10400562 113.5 ns/op 112 B/op 2 allocs/op BenchmarkReadAll/size_64B/ReadAll/Reader-32 2129949 558.7 ns/op 1584 B/op 3 allocs/op BenchmarkReadAll/size_512B/io.ReadAll/WriterTo-32 2843871 419.1 ns/op 1408 B/op 2 allocs/op BenchmarkReadAll/size_512B/io.ReadAll/Reader-32 2871580 413.1 ns/op 1408 B/op 2 allocs/op BenchmarkReadAll/size_512B/ReadAll/WriterTo-32 4976233 241.5 ns/op 560 B/op 2 allocs/op BenchmarkReadAll/size_512B/ReadAll/Reader-32 2183186 552.9 ns/op 1584 B/op 3 allocs/op BenchmarkReadAll/size_10KB/io.ReadAll/WriterTo-32 142633 8235 ns/op 46080 B/op 10 allocs/op BenchmarkReadAll/size_10KB/io.ReadAll/Reader-32 148326 8229 ns/op 46080 B/op 10 allocs/op BenchmarkReadAll/size_10KB/ReadAll/WriterTo-32 574903 2210 ns/op 10288 B/op 2 allocs/op BenchmarkReadAll/size_10KB/ReadAll/Reader-32 147210 7995 ns/op 32304 B/op 7 allocs/op BenchmarkReadAll/size_100KB/io.ReadAll/WriterTo-32 13171 90853 ns/op 514304 B/op 18 allocs/op BenchmarkReadAll/size_100KB/io.ReadAll/Reader-32 12892 91787 ns/op 514304 B/op 18 allocs/op BenchmarkReadAll/size_100KB/ReadAll/WriterTo-32 51472 22420 ns/op 106544 B/op 2 allocs/op BenchmarkReadAll/size_100KB/ReadAll/Reader-32 21568 55070 ns/op 261680 B/op 10 allocs/op BenchmarkReadAll/size_1MB/io.ReadAll/WriterTo-32 1220 983276 ns/op 5241098 B/op 27 allocs/op BenchmarkReadAll/size_1MB/io.ReadAll/Reader-32 1089 990818 ns/op 5241100 B/op 27 allocs/op BenchmarkReadAll/size_1MB/ReadAll/WriterTo-32 4153 294507 ns/op 1048627 B/op 2 allocs/op BenchmarkReadAll/size_1MB/ReadAll/Reader-32 1195 944781 ns/op 4193852 B/op 14 allocs/op
7c1a9ba
to
4d5d1da
Compare
💚 Build Succeeded
History
|
What does this PR do?
This adds an optimized io.ReadAll that uses a bytes.Buffer + io.Copy
internally, improving the buffer growth ratio over the runtime append
and taking advange of io.WriterTo when available.
The only scenario where this optimized version is slower than io.ReadAll is
when reading < 512 bytes of a reader that does not implement io.WriterTo.
For now I only updated tests to use the new function as updating the code
requires more careful consideration.
While at it, migrate from the deprecated ioutil package to io.
Checklist
Related issues