Skip to content

Performance optimizations for data reading/writing #942

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.

Sign up for GitHub

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

Merged
merged 2 commits into from
Sep 24, 2024
Merged

Performance optimizations for data reading/writing #942

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
f8ade8b
(135743158) Performance optimizations for data reading/writing
jmschonfeld Sep 23, 2024
043fca8
Fix Windows build
jmschonfeld Sep 24, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
97 changes: 67 additions & 30 deletions Benchmarks/Benchmarks/DataIO/BenchmarkDataIO.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -30,14 +30,6 @@ import Glibc
import Darwin
#endif

func testPath() -> URL {
#if compiler(>=6)
FileManager.default.temporaryDirectory.appending(path: "testfile-\(UUID().uuidString)", directoryHint: .notDirectory)
#else
FileManager.default.temporaryDirectory.appendingPathComponent("testfile-\(UUID().uuidString)")
#endif
}

func generateTestData(count: Int) -> Data {
let memory = malloc(count)!
let ptr = memory.bindMemory(to: UInt8.self, capacity: count)
Expand All @@ -51,18 +43,29 @@ func generateTestData(count: Int) -> Data {
return Data(bytesNoCopy: ptr, count: count, deallocator: .free)
}

func cleanup(at path: URL) {
try? FileManager.default.removeItem(at: path)
func cleanupTestPath() {
try? FileManager.default.removeItem(at: testPath)
// Ignore any errors
}

// 16 MB file, big enough to trigger things like chunking
let data = generateTestData(count: 1 << 24)
let readMe = testPath()
#if compiler(>=6)
let testPath = FileManager.default.temporaryDirectory.appending(path: "testfile-\(UUID().uuidString)", directoryHint: .notDirectory)
#else
let testPath = FileManager.default.temporaryDirectory.appendingPathComponent("testfile-\(UUID().uuidString)")
#endif
let nonExistentPath = URL(filePath: "/does-not-exist", directoryHint: .notDirectory)

let base64Data = generateTestData(count: 1024 * 1024)
let base64DataString = base64Data.base64EncodedString()

extension Benchmark.Configuration {
fileprivate static var cleanupTestPathConfig: Self {
.init(teardown: cleanupTestPath)
}
}

let benchmarks = {
Benchmark.defaultConfiguration.maxIterations = 1_000_000_000
Benchmark.defaultConfiguration.maxDuration = .seconds(3)
Expand All @@ -75,44 +78,78 @@ let benchmarks = {
Benchmark.defaultConfiguration.metrics = [.cpuTotal, .wallClock, .mallocCountTotal, .throughput]
#endif

Benchmark("read-write-emptyFile") { benchmark in
let path = testPath()
Benchmark("read-write-emptyFile", configuration: .cleanupTestPathConfig) { benchmark in
let data = Data()
try data.write(to: path)
let read = try Data(contentsOf: path, options: [])
cleanup(at: path)
try data.write(to: testPath)
let read = try Data(contentsOf: testPath, options: [])
}

Benchmark("write-regularFile") { benchmark in
let path = testPath()
try data.write(to: path)
cleanup(at: path)
Benchmark("write-regularFile", configuration: .cleanupTestPathConfig) { benchmark in
try data.write(to: testPath)
}

Benchmark("write-regularFile-atomic", configuration: .cleanupTestPathConfig) { benchmark in
try data.write(to: testPath, options: .atomic)
}

Benchmark("write-regularFile-alreadyExists",
configuration: .init(
setup: {
try! Data().write(to: testPath)
},
teardown: cleanupTestPath
)
) { benchmark in
try? data.write(to: testPath)
}

Benchmark("write-regularFile-alreadyExists-atomic",
configuration: .init(
setup: {
try! Data().write(to: testPath)
},
teardown: cleanupTestPath
)
) { benchmark in
try? data.write(to: testPath, options: .atomic)
}

Benchmark("read-regularFile",
configuration: .init(
setup: {
try! data.write(to: readMe)
try! data.write(to: testPath)
},
teardown: {
cleanup(at: readMe)
}
teardown: cleanupTestPath
)
) { benchmark in
blackHole(try Data(contentsOf: readMe))
blackHole(try Data(contentsOf: testPath))
}

Benchmark("read-nonExistentFile") { benchmark in
for _ in benchmark.scaledIterations {
blackHole(try? Data(contentsOf: nonExistentPath))
}
}

Benchmark("read-nonExistentFile-userInfo") { benchmark in
for _ in benchmark.scaledIterations {
do {
blackHole(try Data(contentsOf: nonExistentPath))
} catch {
blackHole((error as? CocoaError)?.userInfo["NSURLErrorKey"])
}
}
}

Benchmark("read-hugeFile",
configuration: .init(
setup: {
try! generateTestData(count: 1 << 30).write(to: readMe)
try! generateTestData(count: 1 << 30).write(to: testPath)
},
teardown: {
cleanup(at: readMe)
}
teardown: cleanupTestPath
)
) { benchmark in
blackHole(try Data(contentsOf: readMe))
blackHole(try Data(contentsOf: testPath))
}

// MARK: base64
Expand Down
96 changes: 34 additions & 62 deletions Sources/FoundationEssentials/Data/Data+Writing.swift
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -125,12 +125,18 @@ private func writeToFileDescriptorWithProgress(_ fd: Int32, buffer: UnsafeRawBuf

private func cleanupTemporaryDirectory(at inPath: String?) {
guard let inPath else { return }
#if canImport(Darwin) || os(Linux)
// Since we expect the directory to be empty at this point, try rmdir which is much faster than Darwin's removefile(3) for known empty directories
if inPath.withFileSystemRepresentation({ $0.flatMap(rmdir) }) == 0 {
return
}
#endif
// Attempt to use FileManager, ignore error
try? FileManager.default.removeItem(atPath: inPath)
}

/// Caller is responsible for calling `close` on the `Int32` file descriptor.
private func createTemporaryFile(at destinationPath: String, inPath: PathOrURL, prefix: String, options: Data.WritingOptions) throws -> (Int32, String) {
private func createTemporaryFile(at destinationPath: String, inPath: PathOrURL, prefix: String, options: Data.WritingOptions, variant: String? = nil) throws -> (Int32, String) {
#if os(WASI)
// WASI does not have temp directories
throw CocoaError(.featureUnsupported)
Expand All @@ -154,12 +160,12 @@ private func createTemporaryFile(at destinationPath: String, inPath: PathOrURL,
// Furthermore, we can't compatibly switch to mkstemp() until we have the ability to set fchmod correctly, which requires the ability to query the current umask, which we don't have. (22033100)
#if os(Windows)
guard _mktemp_s(templateFileSystemRep, template.count + 1) == 0 else {
throw CocoaError.errorWithFilePath(inPath, errno: errno, reading: false)
throw CocoaError.errorWithFilePath(inPath, errno: errno, reading: false, variant: variant)
}
let flags: CInt = _O_BINARY | _O_CREAT | _O_EXCL | _O_RDWR
#else
guard mktemp(templateFileSystemRep) != nil else {
throw CocoaError.errorWithFilePath(inPath, errno: errno, reading: false)
throw CocoaError.errorWithFilePath(inPath, errno: errno, reading: false, variant: variant)
}
let flags: CInt = O_CREAT | O_EXCL | O_RDWR
#endif
Expand All @@ -172,7 +178,7 @@ private func createTemporaryFile(at destinationPath: String, inPath: PathOrURL,

// If the file exists, we repeat. Otherwise throw the error.
if errno != EEXIST {
throw CocoaError.errorWithFilePath(inPath, errno: errno, reading: false)
throw CocoaError.errorWithFilePath(inPath, errno: errno, reading: false, variant: variant)
}

// Try again
Expand All @@ -194,12 +200,25 @@ private func createTemporaryFile(at destinationPath: String, inPath: PathOrURL,

/// Returns `(file descriptor, temporary file path, temporary directory path)`
/// Caller is responsible for calling `close` on the `Int32` file descriptor and calling `cleanupTemporaryDirectory` on the temporary directory path. The temporary directory path may be nil, if it does not need to be cleaned up.
private func createProtectedTemporaryFile(at destinationPath: String, inPath: PathOrURL, options: Data.WritingOptions) throws -> (Int32, String, String?) {
private func createProtectedTemporaryFile(at destinationPath: String, inPath: PathOrURL, options: Data.WritingOptions, variant: String? = nil) throws -> (Int32, String, String?) {
#if FOUNDATION_FRAMEWORK
if _foundation_sandbox_check(getpid(), nil) != 0 {
// Convert the path back into a string
let url = URL(fileURLWithPath: destinationPath, isDirectory: false)
let temporaryDirectoryPath = try FileManager.default.url(for: .itemReplacementDirectory, in: .userDomainMask, appropriateFor: url, create: true).path(percentEncoded: false)
var temporaryDirectoryPath: String
do {
temporaryDirectoryPath = try FileManager.default.url(for: .itemReplacementDirectory, in: .userDomainMask, appropriateFor: url, create: true).path(percentEncoded: false)
} catch {
if let variant, let cocoaError = error as? CocoaError {
let code = cocoaError.code
var userInfo = cocoaError.userInfo
userInfo[NSUserStringVariantErrorKey] = variant

throw CocoaError(code, userInfo: userInfo)
} else {
throw error
}
}

let auxFile = temporaryDirectoryPath.appendingPathComponent(destinationPath.lastPathComponent)
return try auxFile.withFileSystemRepresentation { auxFileFileSystemRep in
Expand All @@ -212,14 +231,14 @@ private func createProtectedTemporaryFile(at destinationPath: String, inPath: Pa
} else {
let savedErrno = errno
cleanupTemporaryDirectory(at: temporaryDirectoryPath)
throw CocoaError.errorWithFilePath(inPath, errno: savedErrno, reading: false)
throw CocoaError.errorWithFilePath(inPath, errno: savedErrno, reading: false, variant: variant)
}
}
}
#endif

let temporaryDirectoryPath = destinationPath.deletingLastPathComponent()
let (fd, auxFile) = try createTemporaryFile(at: temporaryDirectoryPath, inPath: inPath, prefix: ".dat.nosync", options: options)
let (fd, auxFile) = try createTemporaryFile(at: temporaryDirectoryPath, inPath: inPath, prefix: ".dat.nosync", options: options, variant: variant)
return (fd, auxFile, nil)
}

Expand Down Expand Up @@ -310,25 +329,7 @@ private func writeToFileAux(path inPath: PathOrURL, buffer: UnsafeRawBufferPoint

#if os(Windows)
try inPath.path.withNTPathRepresentation { pwszPath in
var fd: CInt
var auxPath: String?
var temporaryDirectoryPath: String?

do {
(fd, auxPath, temporaryDirectoryPath) = try createProtectedTemporaryFile(at: inPath.path, inPath: inPath, options: options)
} catch {
if let cocoaError = error as? CocoaError {
// Extract code and userInfo, then re-create it with an additional userInfo key.
let code = cocoaError.code
var userInfo = cocoaError.userInfo
userInfo[NSUserStringVariantErrorKey] = "Folder"

throw CocoaError(code, userInfo: userInfo)
} else {
// These should all be CocoaErrors, but just in case we re-throw the original one here.
throw error
}
}
var (fd, auxPath, temporaryDirectoryPath) = try createProtectedTemporaryFile(at: inPath.path, inPath: inPath, options: options, variant: "Folder")

// Cleanup temporary directory
defer { cleanupTemporaryDirectory(at: temporaryDirectoryPath) }
Expand All @@ -344,10 +345,8 @@ private func writeToFileAux(path inPath: PathOrURL, buffer: UnsafeRawBufferPoint
do {
try write(buffer: buffer, toFileDescriptor: fd, path: inPath, parentProgress: callback)
} catch {
if let auxPath {
try auxPath.withNTPathRepresentation { pwszAuxPath in
_ = DeleteFileW(pwszAuxPath)
}
try auxPath.withNTPathRepresentation { pwszAuxPath in
_ = DeleteFileW(pwszAuxPath)
}

if callback?.isCancelled ?? false {
Expand All @@ -359,9 +358,6 @@ private func writeToFileAux(path inPath: PathOrURL, buffer: UnsafeRawBufferPoint

writeExtendedAttributes(fd: fd, attributes: attributes)

// We're done now
guard let auxPath else { return }

_close(fd)
fd = -1

Expand All @@ -379,10 +375,7 @@ private func writeToFileAux(path inPath: PathOrURL, buffer: UnsafeRawBufferPoint
throw CocoaError(.fileWriteInvalidFileName)
}

let fd: Int32
var mode: mode_t?
var temporaryDirectoryPath: String?
var auxPath: String?

#if FOUNDATION_FRAMEWORK
var newPath = inPath.path
Expand Down Expand Up @@ -410,21 +403,7 @@ private func writeToFileAux(path inPath: PathOrURL, buffer: UnsafeRawBufferPoint
let newPath = inPath.path
#endif

do {
(fd, auxPath, temporaryDirectoryPath) = try createProtectedTemporaryFile(at: newPath, inPath: inPath, options: options)
} catch {
if let cocoaError = error as? CocoaError {
// Extract code and userInfo, then re-create it with an additional userInfo key.
let code = cocoaError.code
var userInfo = cocoaError.userInfo
userInfo[NSUserStringVariantErrorKey] = "Folder"

throw CocoaError(code, userInfo: userInfo)
} else {
// These should all be CocoaErrors, but just in case we re-throw the original one here.
throw error
}
}
var (fd, auxPath, temporaryDirectoryPath) = try createProtectedTemporaryFile(at: newPath, inPath: inPath, options: options, variant: "Folder")

guard fd >= 0 else {
let savedErrno = errno
Expand All @@ -442,11 +421,9 @@ private func writeToFileAux(path inPath: PathOrURL, buffer: UnsafeRawBufferPoint
} catch {
let savedError = errno

if let auxPath {
auxPath.withFileSystemRepresentation { pathFileSystemRep in
guard let pathFileSystemRep else { return }
unlink(pathFileSystemRep)
}
auxPath.withFileSystemRepresentation { pathFileSystemRep in
guard let pathFileSystemRep else { return }
unlink(pathFileSystemRep)
}
cleanupTemporaryDirectory(at: temporaryDirectoryPath)

Expand All @@ -458,11 +435,6 @@ private func writeToFileAux(path inPath: PathOrURL, buffer: UnsafeRawBufferPoint
}

writeExtendedAttributes(fd: fd, attributes: attributes)

guard let auxPath else {
// We're done now
return
}

try auxPath.withFileSystemRepresentation { auxPathFileSystemRep in
guard let auxPathFileSystemRep else {
Expand Down
Loading