Merge pull request #2140 from containers/renovate/go-github.com/docker/docker-vulnerability

chore(deps): update module github.com/docker/docker to v24.0.7+incompatible [security]

go.mod

@@ -38,7 +38,7 @@ require (
 	github.com/cyphar/filepath-securejoin v0.2.4 // indirect
 	github.com/davecgh/go-spew v1.1.1 // indirect
 	github.com/distribution/reference v0.5.0 // indirect
-	github.com/docker/docker v24.0.6+incompatible // indirect
+	github.com/docker/docker v24.0.7+incompatible // indirect
 	github.com/docker/docker-credential-helpers v0.8.0 // indirect
 	github.com/docker/go-connections v0.4.0 // indirect
 	github.com/docker/go-units v0.5.0 // indirect
@@ -70,7 +70,7 @@ require (
 	github.com/inconshreveable/mousetrap v1.1.0 // indirect
 	github.com/josharian/intern v1.0.0 // indirect
 	github.com/json-iterator/go v1.1.12 // indirect
-	github.com/klauspost/compress v1.16.7 // indirect
+	github.com/klauspost/compress v1.17.2 // indirect
 	github.com/klauspost/pgzip v1.2.6 // indirect
 	github.com/letsencrypt/boulder v0.0.0-20230213213521-fdfea0d469b6 // indirect
 	github.com/mailru/easyjson v0.7.7 // indirect

go.sum

@@ -55,8 +55,8 @@ github.com/distribution/reference v0.5.0 h1:/FUIFXtfc/x2gpa5/VGfiGLuOIdYa1t65IKK
 github.com/distribution/reference v0.5.0/go.mod h1:BbU0aIcezP1/5jX/8MP0YiH4SdvB5Y4f/wlDRiLyi3E=
 github.com/docker/distribution v2.8.3+incompatible h1:AtKxIZ36LoNK51+Z6RpzLpddBirtxJnzDrHLEKxTAYk=
 github.com/docker/distribution v2.8.3+incompatible/go.mod h1:J2gT2udsDAN96Uj4KfcMRqY0/ypR+oyYUYmja8H+y+w=
-github.com/docker/docker v24.0.6+incompatible h1:hceabKCtUgDqPu+qm0NgsaXf28Ljf4/pWFL7xjWWDgE=
+github.com/docker/docker v24.0.7+incompatible h1:Wo6l37AuwP3JaMnZa226lzVXGA3F9Ig1seQen0cKYlM=
-github.com/docker/docker v24.0.6+incompatible/go.mod h1:eEKB0N0r5NX/I1kEveEz05bcu8tLC/8azJZsviup8Sk=
+github.com/docker/docker v24.0.7+incompatible/go.mod h1:eEKB0N0r5NX/I1kEveEz05bcu8tLC/8azJZsviup8Sk=
 github.com/docker/docker-credential-helpers v0.8.0 h1:YQFtbBQb4VrpoPxhFuzEBPQ9E16qz5SpHLS+uswaCp8=
 github.com/docker/docker-credential-helpers v0.8.0/go.mod h1:UGFXcuoQ5TxPiB54nHOZ32AWRqQdECoh/Mg0AlEYb40=
 github.com/docker/go-connections v0.4.0 h1:El9xVISelRB7BuFusrZozjnkIM5YnzCViNKohAFqRJQ=
@@ -217,8 +217,8 @@ github.com/kisielk/errcheck v1.5.0/go.mod h1:pFxgyoBC7bSaBwPgfKdkLd5X25qrDl4LWUI
 github.com/kisielk/gotool v1.0.0/go.mod h1:XhKaO+MFFWcvkIS/tQcRk01m1F5IRFswLeQ+oQHNcck=
 github.com/klauspost/compress v1.4.1/go.mod h1:RyIbtBH6LamlWaDj8nUwkbUhJ87Yi3uG0guNDohfE1A=
 github.com/klauspost/compress v1.13.6/go.mod h1:/3/Vjq9QcHkK5uEr5lBEmyoZ1iFhe47etQ6QUkpK6sk=
-github.com/klauspost/compress v1.16.7 h1:2mk3MPGNzKyxErAw8YaohYh69+pa4sIQSC0fPGCFR9I=
+github.com/klauspost/compress v1.17.2 h1:RlWWUY/Dr4fL8qk9YG7DTZ7PDgME2V4csBXA8L/ixi4=
-github.com/klauspost/compress v1.16.7/go.mod h1:ntbaceVETuRiXiv4DpjP66DpAtAGkEQskQzEyD//IeE=
+github.com/klauspost/compress v1.17.2/go.mod h1:ntbaceVETuRiXiv4DpjP66DpAtAGkEQskQzEyD//IeE=
 github.com/klauspost/cpuid v1.2.0/go.mod h1:Pj4uuM528wm8OyEC2QMXAi2YiTZ96dNQPGgoMS4s3ek=
 github.com/klauspost/pgzip v1.2.6 h1:8RXeL5crjEUFnR2/Sn6GJNWtSQ3Dk8pq4CL3jvdDyjU=
 github.com/klauspost/pgzip v1.2.6/go.mod h1:Ch1tH69qFZu15pkjo5kYi6mth2Zzwzt50oCQKQE9RUs=

vendor/github.com/docker/docker/api/types/versions/compare.go

@@ -16,11 +16,11 @@ func compare(v1, v2 string) int {
 		otherTab = strings.Split(v2, ".")
 	)
 
-	max := len(currTab)
+	maxVer := len(currTab)
-	if len(otherTab) > max {
+	if len(otherTab) > maxVer {
-		max = len(otherTab)
+		maxVer = len(otherTab)
 	}
-	for i := 0; i < max; i++ {
+	for i := 0; i < maxVer; i++ {
 		var currInt, otherInt int
 
 		if len(currTab) > i {

vendor/github.com/klauspost/compress/.goreleaser.yml

@@ -3,7 +3,7 @@
 before:
   hooks:
     - ./gen.sh
-    - go install mvdan.cc/garble@v0.9.3
+    - go install mvdan.cc/garble@v0.10.1
 
 builds:
   -
@@ -92,16 +92,7 @@ builds:
 archives:
   -
     id: s2-binaries
-    name_template: "s2-{{ .Os }}_{{ .Arch }}_{{ .Version }}"
+    name_template: "s2-{{ .Os }}_{{ .Arch }}{{ if .Arm }}v{{ .Arm }}{{ end }}"
-    replacements:
-      aix: AIX
-      darwin: OSX
-      linux: Linux
-      windows: Windows
-      386: i386
-      amd64: x86_64
-      freebsd: FreeBSD
-      netbsd: NetBSD
     format_overrides:
       - goos: windows
         format: zip
@@ -125,7 +116,7 @@ changelog:
 
 nfpms:
   -
-    file_name_template: "s2_package_{{ .Version }}_{{ .Os }}_{{ .Arch }}"
+    file_name_template: "s2_package__{{ .Os }}_{{ .Arch }}{{ if .Arm }}v{{ .Arm }}{{ end }}"
     vendor: Klaus Post
     homepage: https://github.com/klauspost/compress
     maintainer: Klaus Post <klauspost@gmail.com>
@@ -134,8 +125,3 @@ nfpms:
     formats:
       - deb
       - rpm
-    replacements:
-      darwin: Darwin
-      linux: Linux
-      freebsd: FreeBSD
-      amd64: x86_64

vendor/github.com/klauspost/compress/README.md

@@ -16,6 +16,18 @@ This package provides various compression algorithms.
 
 # changelog
 
+* Sept 19th, 2023 - [v1.17.0](https://github.com/klauspost/compress/releases/tag/v1.17.0)
+	* Add experimental dictionary builder  https://github.com/klauspost/compress/pull/853
+	* Add xerial snappy read/writer https://github.com/klauspost/compress/pull/838
+	* flate: Add limited window compression https://github.com/klauspost/compress/pull/843
+	* s2: Do 2 overlapping match checks https://github.com/klauspost/compress/pull/839
+	* flate: Add amd64 assembly matchlen https://github.com/klauspost/compress/pull/837
+	* gzip: Copy bufio.Reader on Reset by @thatguystone in https://github.com/klauspost/compress/pull/860
+   
+* July 1st, 2023 - [v1.16.7](https://github.com/klauspost/compress/releases/tag/v1.16.7)
+	* zstd: Fix default level first dictionary encode https://github.com/klauspost/compress/pull/829
+	* s2: add GetBufferCapacity() method by @GiedriusS in https://github.com/klauspost/compress/pull/832
+
 * June 13, 2023 - [v1.16.6](https://github.com/klauspost/compress/releases/tag/v1.16.6)
 	* zstd: correctly ignore WithEncoderPadding(1) by @ianlancetaylor in https://github.com/klauspost/compress/pull/806
 	* zstd: Add amd64 match length assembly https://github.com/klauspost/compress/pull/824
@@ -50,6 +62,9 @@ This package provides various compression algorithms.
 	* s2: Support io.ReaderAt in ReadSeeker. https://github.com/klauspost/compress/pull/747
 	* s2c/s2sx: Use concurrent decoding. https://github.com/klauspost/compress/pull/746
 
+<details>
+	<summary>See changes to v1.15.x</summary>
+	
 * Jan 21st, 2023 (v1.15.15)
 	* deflate: Improve level 7-9 by @klauspost in https://github.com/klauspost/compress/pull/739
 	* zstd: Add delta encoding support by @greatroar in https://github.com/klauspost/compress/pull/728
@@ -176,6 +191,8 @@ Stream decompression is now faster on asynchronous, since the goroutine allocati
 
 While the release has been extensively tested, it is recommended to testing when upgrading.
 
+</details>
+
 <details>
 	<summary>See changes to v1.14.x</summary>
 	
@@ -636,6 +653,8 @@ Here are other packages of good quality and pure Go (no cgo wrappers or autoconv
 * [github.com/dsnet/compress](https://github.com/dsnet/compress) - brotli decompression, bzip2 writer.
 * [github.com/ronanh/intcomp](https://github.com/ronanh/intcomp) - Integer compression.
 * [github.com/spenczar/fpc](https://github.com/spenczar/fpc) - Float compression.
+* [github.com/minio/zipindex](https://github.com/minio/zipindex) - External ZIP directory index.
+* [github.com/ybirader/pzip](https://github.com/ybirader/pzip) - Fast concurrent zip archiver and extractor.
 
 # license
 

vendor/github.com/klauspost/compress/flate/deflate.go

@@ -7,6 +7,7 @@ package flate
 
 import (
 	"encoding/binary"
+	"errors"
 	"fmt"
 	"io"
 	"math"
@@ -833,6 +834,12 @@ func (d *compressor) init(w io.Writer, level int) (err error) {
 		d.initDeflate()
 		d.fill = (*compressor).fillDeflate
 		d.step = (*compressor).deflateLazy
+	case -level >= MinCustomWindowSize && -level <= MaxCustomWindowSize:
+		d.w.logNewTablePenalty = 7
+		d.fast = &fastEncL5Window{maxOffset: int32(-level), cur: maxStoreBlockSize}
+		d.window = make([]byte, maxStoreBlockSize)
+		d.fill = (*compressor).fillBlock
+		d.step = (*compressor).storeFast
 	default:
 		return fmt.Errorf("flate: invalid compression level %d: want value in range [-2, 9]", level)
 	}
@@ -929,6 +936,28 @@ func NewWriterDict(w io.Writer, level int, dict []byte) (*Writer, error) {
 	return zw, err
 }
 
+// MinCustomWindowSize is the minimum window size that can be sent to NewWriterWindow.
+const MinCustomWindowSize = 32
+
+// MaxCustomWindowSize is the maximum custom window that can be sent to NewWriterWindow.
+const MaxCustomWindowSize = windowSize
+
+// NewWriterWindow returns a new Writer compressing data with a custom window size.
+// windowSize must be from MinCustomWindowSize to MaxCustomWindowSize.
+func NewWriterWindow(w io.Writer, windowSize int) (*Writer, error) {
+	if windowSize < MinCustomWindowSize {
+		return nil, errors.New("flate: requested window size less than MinWindowSize")
+	}
+	if windowSize > MaxCustomWindowSize {
+		return nil, errors.New("flate: requested window size bigger than MaxCustomWindowSize")
+	}
+	var dw Writer
+	if err := dw.d.init(w, -windowSize); err != nil {
+		return nil, err
+	}
+	return &dw, nil
+}
+
 // A Writer takes data written to it and writes the compressed
 // form of that data to an underlying writer (see NewWriter).
 type Writer struct {

vendor/github.com/klauspost/compress/flate/fast_encoder.go

@@ -8,7 +8,6 @@ package flate
 import (
 	"encoding/binary"
 	"fmt"
-	"math/bits"
 )
 
 type fastEnc interface {
@@ -192,25 +191,3 @@ func (e *fastGen) Reset() {
 	}
 	e.hist = e.hist[:0]
 }
-
-// matchLen returns the maximum length.
-// 'a' must be the shortest of the two.
-func matchLen(a, b []byte) int {
-	var checked int
-
-	for len(a) >= 8 {
-		if diff := binary.LittleEndian.Uint64(a) ^ binary.LittleEndian.Uint64(b); diff != 0 {
-			return checked + (bits.TrailingZeros64(diff) >> 3)
-		}
-		checked += 8
-		a = a[8:]
-		b = b[8:]
-	}
-	b = b[:len(a)]
-	for i := range a {
-		if a[i] != b[i] {
-			return i + checked
-		}
-	}
-	return len(a) + checked
-}

vendor/github.com/klauspost/compress/flate/inflate.go

@@ -120,8 +120,9 @@ func (h *huffmanDecoder) init(lengths []int) bool {
 	const sanity = false
 
 	if h.chunks == nil {
-		h.chunks = &[huffmanNumChunks]uint16{}
+		h.chunks = new([huffmanNumChunks]uint16)
 	}
+
 	if h.maxRead != 0 {
 		*h = huffmanDecoder{chunks: h.chunks, links: h.links}
 	}
@@ -175,6 +176,7 @@ func (h *huffmanDecoder) init(lengths []int) bool {
 	}
 
 	h.maxRead = min
+
 	chunks := h.chunks[:]
 	for i := range chunks {
 		chunks[i] = 0
@@ -202,8 +204,7 @@ func (h *huffmanDecoder) init(lengths []int) bool {
 			if cap(h.links[off]) < numLinks {
 				h.links[off] = make([]uint16, numLinks)
 			} else {
-				links := h.links[off][:0]
+				h.links[off] = h.links[off][:numLinks]
-				h.links[off] = links[:numLinks]
 			}
 		}
 	} else {
@@ -277,7 +278,7 @@ func (h *huffmanDecoder) init(lengths []int) bool {
 	return true
 }
 
-// The actual read interface needed by NewReader.
+// Reader is the actual read interface needed by NewReader.
 // If the passed in io.Reader does not also have ReadByte,
 // the NewReader will introduce its own buffering.
 type Reader interface {
@@ -285,6 +286,18 @@ type Reader interface {
 	io.ByteReader
 }
 
+type step uint8
+
+const (
+	copyData step = iota + 1
+	nextBlock
+	huffmanBytesBuffer
+	huffmanBytesReader
+	huffmanBufioReader
+	huffmanStringsReader
+	huffmanGenericReader
+)
+
 // Decompress state.
 type decompressor struct {
 	// Input source.
@@ -303,7 +316,7 @@ type decompressor struct {
 
 	// Next step in the decompression,
 	// and decompression state.
-	step      func(*decompressor)
+	step      step
 	stepState int
 	err       error
 	toRead    []byte
@@ -342,7 +355,7 @@ func (f *decompressor) nextBlock() {
 		// compressed, fixed Huffman tables
 		f.hl = &fixedHuffmanDecoder
 		f.hd = nil
-		f.huffmanBlockDecoder()()
+		f.huffmanBlockDecoder()
 		if debugDecode {
 			fmt.Println("predefinied huffman block")
 		}
@@ -353,7 +366,7 @@ func (f *decompressor) nextBlock() {
 		}
 		f.hl = &f.h1
 		f.hd = &f.h2
-		f.huffmanBlockDecoder()()
+		f.huffmanBlockDecoder()
 		if debugDecode {
 			fmt.Println("dynamic huffman block")
 		}
@@ -379,14 +392,16 @@ func (f *decompressor) Read(b []byte) (int, error) {
 		if f.err != nil {
 			return 0, f.err
 		}
-		f.step(f)
+
+		f.doStep()
+
 		if f.err != nil && len(f.toRead) == 0 {
 			f.toRead = f.dict.readFlush() // Flush what's left in case of error
 		}
 	}
 }
 
-// Support the io.WriteTo interface for io.Copy and friends.
+// WriteTo implements the io.WriteTo interface for io.Copy and friends.
 func (f *decompressor) WriteTo(w io.Writer) (int64, error) {
 	total := int64(0)
 	flushed := false
@@ -410,7 +425,7 @@ func (f *decompressor) WriteTo(w io.Writer) (int64, error) {
 			return total, f.err
 		}
 		if f.err == nil {
-			f.step(f)
+			f.doStep()
 		}
 		if len(f.toRead) == 0 && f.err != nil && !flushed {
 			f.toRead = f.dict.readFlush() // Flush what's left in case of error
@@ -631,7 +646,7 @@ func (f *decompressor) copyData() {
 
 	if f.dict.availWrite() == 0 || f.copyLen > 0 {
 		f.toRead = f.dict.readFlush()
-		f.step = (*decompressor).copyData
+		f.step = copyData
 		return
 	}
 	f.finishBlock()
@@ -644,7 +659,28 @@ func (f *decompressor) finishBlock() {
 		}
 		f.err = io.EOF
 	}
-	f.step = (*decompressor).nextBlock
+	f.step = nextBlock
+}
+
+func (f *decompressor) doStep() {
+	switch f.step {
+	case copyData:
+		f.copyData()
+	case nextBlock:
+		f.nextBlock()
+	case huffmanBytesBuffer:
+		f.huffmanBytesBuffer()
+	case huffmanBytesReader:
+		f.huffmanBytesReader()
+	case huffmanBufioReader:
+		f.huffmanBufioReader()
+	case huffmanStringsReader:
+		f.huffmanStringsReader()
+	case huffmanGenericReader:
+		f.huffmanGenericReader()
+	default:
+		panic("BUG: unexpected step state")
+	}
 }
 
 // noEOF returns err, unless err == io.EOF, in which case it returns io.ErrUnexpectedEOF.
@@ -747,7 +783,7 @@ func (f *decompressor) Reset(r io.Reader, dict []byte) error {
 		h1:       f.h1,
 		h2:       f.h2,
 		dict:     f.dict,
-		step:     (*decompressor).nextBlock,
+		step:     nextBlock,
 	}
 	f.dict.init(maxMatchOffset, dict)
 	return nil
@@ -768,7 +804,7 @@ func NewReader(r io.Reader) io.ReadCloser {
 	f.r = makeReader(r)
 	f.bits = new([maxNumLit + maxNumDist]int)
 	f.codebits = new([numCodes]int)
-	f.step = (*decompressor).nextBlock
+	f.step = nextBlock
 	f.dict.init(maxMatchOffset, nil)
 	return &f
 }
@@ -787,7 +823,7 @@ func NewReaderDict(r io.Reader, dict []byte) io.ReadCloser {
 	f.r = makeReader(r)
 	f.bits = new([maxNumLit + maxNumDist]int)
 	f.codebits = new([numCodes]int)
-	f.step = (*decompressor).nextBlock
+	f.step = nextBlock
 	f.dict.init(maxMatchOffset, dict)
 	return &f
 }

vendor/github.com/klauspost/compress/flate/inflate_gen.go

@@ -85,7 +85,7 @@ readLiteral:
 			dict.writeByte(byte(v))
 			if dict.availWrite() == 0 {
 				f.toRead = dict.readFlush()
-				f.step = (*decompressor).huffmanBytesBuffer
+				f.step = huffmanBytesBuffer
 				f.stepState = stateInit
 				f.b, f.nb = fb, fnb
 				return
@@ -251,7 +251,7 @@ copyHistory:
 
 		if dict.availWrite() == 0 || f.copyLen > 0 {
 			f.toRead = dict.readFlush()
-			f.step = (*decompressor).huffmanBytesBuffer // We need to continue this work
+			f.step = huffmanBytesBuffer // We need to continue this work
 			f.stepState = stateDict
 			f.b, f.nb = fb, fnb
 			return
@@ -336,7 +336,7 @@ readLiteral:
 			dict.writeByte(byte(v))
 			if dict.availWrite() == 0 {
 				f.toRead = dict.readFlush()
-				f.step = (*decompressor).huffmanBytesReader
+				f.step = huffmanBytesReader
 				f.stepState = stateInit
 				f.b, f.nb = fb, fnb
 				return
@@ -502,7 +502,7 @@ copyHistory:
 
 		if dict.availWrite() == 0 || f.copyLen > 0 {
 			f.toRead = dict.readFlush()
-			f.step = (*decompressor).huffmanBytesReader // We need to continue this work
+			f.step = huffmanBytesReader // We need to continue this work
 			f.stepState = stateDict
 			f.b, f.nb = fb, fnb
 			return
@@ -587,7 +587,7 @@ readLiteral:
 			dict.writeByte(byte(v))
 			if dict.availWrite() == 0 {
 				f.toRead = dict.readFlush()
-				f.step = (*decompressor).huffmanBufioReader
+				f.step = huffmanBufioReader
 				f.stepState = stateInit
 				f.b, f.nb = fb, fnb
 				return
@@ -753,7 +753,7 @@ copyHistory:
 
 		if dict.availWrite() == 0 || f.copyLen > 0 {
 			f.toRead = dict.readFlush()
-			f.step = (*decompressor).huffmanBufioReader // We need to continue this work
+			f.step = huffmanBufioReader // We need to continue this work
 			f.stepState = stateDict
 			f.b, f.nb = fb, fnb
 			return
@@ -838,7 +838,7 @@ readLiteral:
 			dict.writeByte(byte(v))
 			if dict.availWrite() == 0 {
 				f.toRead = dict.readFlush()
-				f.step = (*decompressor).huffmanStringsReader
+				f.step = huffmanStringsReader
 				f.stepState = stateInit
 				f.b, f.nb = fb, fnb
 				return
@@ -1004,7 +1004,7 @@ copyHistory:
 
 		if dict.availWrite() == 0 || f.copyLen > 0 {
 			f.toRead = dict.readFlush()
-			f.step = (*decompressor).huffmanStringsReader // We need to continue this work
+			f.step = huffmanStringsReader // We need to continue this work
 			f.stepState = stateDict
 			f.b, f.nb = fb, fnb
 			return
@@ -1089,7 +1089,7 @@ readLiteral:
 			dict.writeByte(byte(v))
 			if dict.availWrite() == 0 {
 				f.toRead = dict.readFlush()
-				f.step = (*decompressor).huffmanGenericReader
+				f.step = huffmanGenericReader
 				f.stepState = stateInit
 				f.b, f.nb = fb, fnb
 				return
@@ -1255,7 +1255,7 @@ copyHistory:
 
 		if dict.availWrite() == 0 || f.copyLen > 0 {
 			f.toRead = dict.readFlush()
-			f.step = (*decompressor).huffmanGenericReader // We need to continue this work
+			f.step = huffmanGenericReader // We need to continue this work
 			f.stepState = stateDict
 			f.b, f.nb = fb, fnb
 			return
@@ -1265,19 +1265,19 @@ copyHistory:
 	// Not reached
 }
 
-func (f *decompressor) huffmanBlockDecoder() func() {
+func (f *decompressor) huffmanBlockDecoder() {
 	switch f.r.(type) {
 	case *bytes.Buffer:
-		return f.huffmanBytesBuffer
+		f.huffmanBytesBuffer()
 	case *bytes.Reader:
-		return f.huffmanBytesReader
+		f.huffmanBytesReader()
 	case *bufio.Reader:
-		return f.huffmanBufioReader
+		f.huffmanBufioReader()
 	case *strings.Reader:
-		return f.huffmanStringsReader
+		f.huffmanStringsReader()
 	case Reader:
-		return f.huffmanGenericReader
+		f.huffmanGenericReader()
 	default:
-		return f.huffmanGenericReader
+		f.huffmanGenericReader()
 	}
 }

vendor/github.com/klauspost/compress/flate/level5.go

@@ -308,3 +308,401 @@ emitRemainder:
 		emitLiteral(dst, src[nextEmit:])
 	}
 }
+
+// fastEncL5Window is a level 5 encoder,
+// but with a custom window size.
+type fastEncL5Window struct {
+	hist      []byte
+	cur       int32
+	maxOffset int32
+	table     [tableSize]tableEntry
+	bTable    [tableSize]tableEntryPrev
+}
+
+func (e *fastEncL5Window) Encode(dst *tokens, src []byte) {
+	const (
+		inputMargin            = 12 - 1
+		minNonLiteralBlockSize = 1 + 1 + inputMargin
+		hashShortBytes         = 4
+	)
+	maxMatchOffset := e.maxOffset
+	if debugDeflate && e.cur < 0 {
+		panic(fmt.Sprint("e.cur < 0: ", e.cur))
+	}
+
+	// Protect against e.cur wraparound.
+	for e.cur >= bufferReset {
+		if len(e.hist) == 0 {
+			for i := range e.table[:] {
+				e.table[i] = tableEntry{}
+			}
+			for i := range e.bTable[:] {
+				e.bTable[i] = tableEntryPrev{}
+			}
+			e.cur = maxMatchOffset
+			break
+		}
+		// Shift down everything in the table that isn't already too far away.
+		minOff := e.cur + int32(len(e.hist)) - maxMatchOffset
+		for i := range e.table[:] {
+			v := e.table[i].offset
+			if v <= minOff {
+				v = 0
+			} else {
+				v = v - e.cur + maxMatchOffset
+			}
+			e.table[i].offset = v
+		}
+		for i := range e.bTable[:] {
+			v := e.bTable[i]
+			if v.Cur.offset <= minOff {
+				v.Cur.offset = 0
+				v.Prev.offset = 0
+			} else {
+				v.Cur.offset = v.Cur.offset - e.cur + maxMatchOffset
+				if v.Prev.offset <= minOff {
+					v.Prev.offset = 0
+				} else {
+					v.Prev.offset = v.Prev.offset - e.cur + maxMatchOffset
+				}
+			}
+			e.bTable[i] = v
+		}
+		e.cur = maxMatchOffset
+	}
+
+	s := e.addBlock(src)
+
+	// This check isn't in the Snappy implementation, but there, the caller
+	// instead of the callee handles this case.
+	if len(src) < minNonLiteralBlockSize {
+		// We do not fill the token table.
+		// This will be picked up by caller.
+		dst.n = uint16(len(src))
+		return
+	}
+
+	// Override src
+	src = e.hist
+	nextEmit := s
+
+	// sLimit is when to stop looking for offset/length copies. The inputMargin
+	// lets us use a fast path for emitLiteral in the main loop, while we are
+	// looking for copies.
+	sLimit := int32(len(src) - inputMargin)
+
+	// nextEmit is where in src the next emitLiteral should start from.
+	cv := load6432(src, s)
+	for {
+		const skipLog = 6
+		const doEvery = 1
+
+		nextS := s
+		var l int32
+		var t int32
+		for {
+			nextHashS := hashLen(cv, tableBits, hashShortBytes)
+			nextHashL := hash7(cv, tableBits)
+
+			s = nextS
+			nextS = s + doEvery + (s-nextEmit)>>skipLog
+			if nextS > sLimit {
+				goto emitRemainder
+			}
+			// Fetch a short+long candidate
+			sCandidate := e.table[nextHashS]
+			lCandidate := e.bTable[nextHashL]
+			next := load6432(src, nextS)
+			entry := tableEntry{offset: s + e.cur}
+			e.table[nextHashS] = entry
+			eLong := &e.bTable[nextHashL]
+			eLong.Cur, eLong.Prev = entry, eLong.Cur
+
+			nextHashS = hashLen(next, tableBits, hashShortBytes)
+			nextHashL = hash7(next, tableBits)
+
+			t = lCandidate.Cur.offset - e.cur
+			if s-t < maxMatchOffset {
+				if uint32(cv) == load3232(src, lCandidate.Cur.offset-e.cur) {
+					// Store the next match
+					e.table[nextHashS] = tableEntry{offset: nextS + e.cur}
+					eLong := &e.bTable[nextHashL]
+					eLong.Cur, eLong.Prev = tableEntry{offset: nextS + e.cur}, eLong.Cur
+
+					t2 := lCandidate.Prev.offset - e.cur
+					if s-t2 < maxMatchOffset && uint32(cv) == load3232(src, lCandidate.Prev.offset-e.cur) {
+						l = e.matchlen(s+4, t+4, src) + 4
+						ml1 := e.matchlen(s+4, t2+4, src) + 4
+						if ml1 > l {
+							t = t2
+							l = ml1
+							break
+						}
+					}
+					break
+				}
+				t = lCandidate.Prev.offset - e.cur
+				if s-t < maxMatchOffset && uint32(cv) == load3232(src, lCandidate.Prev.offset-e.cur) {
+					// Store the next match
+					e.table[nextHashS] = tableEntry{offset: nextS + e.cur}
+					eLong := &e.bTable[nextHashL]
+					eLong.Cur, eLong.Prev = tableEntry{offset: nextS + e.cur}, eLong.Cur
+					break
+				}
+			}
+
+			t = sCandidate.offset - e.cur
+			if s-t < maxMatchOffset && uint32(cv) == load3232(src, sCandidate.offset-e.cur) {
+				// Found a 4 match...
+				l = e.matchlen(s+4, t+4, src) + 4
+				lCandidate = e.bTable[nextHashL]
+				// Store the next match
+
+				e.table[nextHashS] = tableEntry{offset: nextS + e.cur}
+				eLong := &e.bTable[nextHashL]
+				eLong.Cur, eLong.Prev = tableEntry{offset: nextS + e.cur}, eLong.Cur
+
+				// If the next long is a candidate, use that...
+				t2 := lCandidate.Cur.offset - e.cur
+				if nextS-t2 < maxMatchOffset {
+					if load3232(src, lCandidate.Cur.offset-e.cur) == uint32(next) {
+						ml := e.matchlen(nextS+4, t2+4, src) + 4
+						if ml > l {
+							t = t2
+							s = nextS
+							l = ml
+							break
+						}
+					}
+					// If the previous long is a candidate, use that...
+					t2 = lCandidate.Prev.offset - e.cur
+					if nextS-t2 < maxMatchOffset && load3232(src, lCandidate.Prev.offset-e.cur) == uint32(next) {
+						ml := e.matchlen(nextS+4, t2+4, src) + 4
+						if ml > l {
+							t = t2
+							s = nextS
+							l = ml
+							break
+						}
+					}
+				}
+				break
+			}
+			cv = next
+		}
+
+		// A 4-byte match has been found. We'll later see if more than 4 bytes
+		// match. But, prior to the match, src[nextEmit:s] are unmatched. Emit
+		// them as literal bytes.
+
+		if l == 0 {
+			// Extend the 4-byte match as long as possible.
+			l = e.matchlenLong(s+4, t+4, src) + 4
+		} else if l == maxMatchLength {
+			l += e.matchlenLong(s+l, t+l, src)
+		}
+
+		// Try to locate a better match by checking the end of best match...
+		if sAt := s + l; l < 30 && sAt < sLimit {
+			// Allow some bytes at the beginning to mismatch.
+			// Sweet spot is 2/3 bytes depending on input.
+			// 3 is only a little better when it is but sometimes a lot worse.
+			// The skipped bytes are tested in Extend backwards,
+			// and still picked up as part of the match if they do.
+			const skipBeginning = 2
+			eLong := e.bTable[hash7(load6432(src, sAt), tableBits)].Cur.offset
+			t2 := eLong - e.cur - l + skipBeginning
+			s2 := s + skipBeginning
+			off := s2 - t2
+			if t2 >= 0 && off < maxMatchOffset && off > 0 {
+				if l2 := e.matchlenLong(s2, t2, src); l2 > l {
+					t = t2
+					l = l2
+					s = s2
+				}
+			}
+		}
+
+		// Extend backwards
+		for t > 0 && s > nextEmit && src[t-1] == src[s-1] {
+			s--
+			t--
+			l++
+		}
+		if nextEmit < s {
+			if false {
+				emitLiteral(dst, src[nextEmit:s])
+			} else {
+				for _, v := range src[nextEmit:s] {
+					dst.tokens[dst.n] = token(v)
+					dst.litHist[v]++
+					dst.n++
+				}
+			}
+		}
+		if debugDeflate {
+			if t >= s {
+				panic(fmt.Sprintln("s-t", s, t))
+			}
+			if (s - t) > maxMatchOffset {
+				panic(fmt.Sprintln("mmo", s-t))
+			}
+			if l < baseMatchLength {
+				panic("bml")
+			}
+		}
+
+		dst.AddMatchLong(l, uint32(s-t-baseMatchOffset))
+		s += l
+		nextEmit = s
+		if nextS >= s {
+			s = nextS + 1
+		}
+
+		if s >= sLimit {
+			goto emitRemainder
+		}
+
+		// Store every 3rd hash in-between.
+		if true {
+			const hashEvery = 3
+			i := s - l + 1
+			if i < s-1 {
+				cv := load6432(src, i)
+				t := tableEntry{offset: i + e.cur}
+				e.table[hashLen(cv, tableBits, hashShortBytes)] = t
+				eLong := &e.bTable[hash7(cv, tableBits)]
+				eLong.Cur, eLong.Prev = t, eLong.Cur
+
+				// Do an long at i+1
+				cv >>= 8
+				t = tableEntry{offset: t.offset + 1}
+				eLong = &e.bTable[hash7(cv, tableBits)]
+				eLong.Cur, eLong.Prev = t, eLong.Cur
+
+				// We only have enough bits for a short entry at i+2
+				cv >>= 8
+				t = tableEntry{offset: t.offset + 1}
+				e.table[hashLen(cv, tableBits, hashShortBytes)] = t
+
+				// Skip one - otherwise we risk hitting 's'
+				i += 4
+				for ; i < s-1; i += hashEvery {
+					cv := load6432(src, i)
+					t := tableEntry{offset: i + e.cur}
+					t2 := tableEntry{offset: t.offset + 1}
+					eLong := &e.bTable[hash7(cv, tableBits)]
+					eLong.Cur, eLong.Prev = t, eLong.Cur
+					e.table[hashLen(cv>>8, tableBits, hashShortBytes)] = t2
+				}
+			}
+		}
+
+		// We could immediately start working at s now, but to improve
+		// compression we first update the hash table at s-1 and at s.
+		x := load6432(src, s-1)
+		o := e.cur + s - 1
+		prevHashS := hashLen(x, tableBits, hashShortBytes)
+		prevHashL := hash7(x, tableBits)
+		e.table[prevHashS] = tableEntry{offset: o}
+		eLong := &e.bTable[prevHashL]
+		eLong.Cur, eLong.Prev = tableEntry{offset: o}, eLong.Cur
+		cv = x >> 8
+	}
+
+emitRemainder:
+	if int(nextEmit) < len(src) {
+		// If nothing was added, don't encode literals.
+		if dst.n == 0 {
+			return
+		}
+
+		emitLiteral(dst, src[nextEmit:])
+	}
+}
+
+// Reset the encoding table.
+func (e *fastEncL5Window) Reset() {
+	// We keep the same allocs, since we are compressing the same block sizes.
+	if cap(e.hist) < allocHistory {
+		e.hist = make([]byte, 0, allocHistory)
+	}
+
+	// We offset current position so everything will be out of reach.
+	// If we are above the buffer reset it will be cleared anyway since len(hist) == 0.
+	if e.cur <= int32(bufferReset) {
+		e.cur += e.maxOffset + int32(len(e.hist))
+	}
+	e.hist = e.hist[:0]
+}
+
+func (e *fastEncL5Window) addBlock(src []byte) int32 {
+	// check if we have space already
+	maxMatchOffset := e.maxOffset
+
+	if len(e.hist)+len(src) > cap(e.hist) {
+		if cap(e.hist) == 0 {
+			e.hist = make([]byte, 0, allocHistory)
+		} else {
+			if cap(e.hist) < int(maxMatchOffset*2) {
+				panic("unexpected buffer size")
+			}
+			// Move down
+			offset := int32(len(e.hist)) - maxMatchOffset
+			copy(e.hist[0:maxMatchOffset], e.hist[offset:])
+			e.cur += offset
+			e.hist = e.hist[:maxMatchOffset]
+		}
+	}
+	s := int32(len(e.hist))
+	e.hist = append(e.hist, src...)
+	return s
+}
+
+// matchlen will return the match length between offsets and t in src.
+// The maximum length returned is maxMatchLength - 4.
+// It is assumed that s > t, that t >=0 and s < len(src).
+func (e *fastEncL5Window) matchlen(s, t int32, src []byte) int32 {
+	if debugDecode {
+		if t >= s {
+			panic(fmt.Sprint("t >=s:", t, s))
+		}
+		if int(s) >= len(src) {
+			panic(fmt.Sprint("s >= len(src):", s, len(src)))
+		}
+		if t < 0 {
+			panic(fmt.Sprint("t < 0:", t))
+		}
+		if s-t > e.maxOffset {
+			panic(fmt.Sprint(s, "-", t, "(", s-t, ") > maxMatchLength (", maxMatchOffset, ")"))
+		}
+	}
+	s1 := int(s) + maxMatchLength - 4
+	if s1 > len(src) {
+		s1 = len(src)
+	}
+
+	// Extend the match to be as long as possible.
+	return int32(matchLen(src[s:s1], src[t:]))
+}
+
+// matchlenLong will return the match length between offsets and t in src.
+// It is assumed that s > t, that t >=0 and s < len(src).
+func (e *fastEncL5Window) matchlenLong(s, t int32, src []byte) int32 {
+	if debugDeflate {
+		if t >= s {
+			panic(fmt.Sprint("t >=s:", t, s))
+		}
+		if int(s) >= len(src) {
+			panic(fmt.Sprint("s >= len(src):", s, len(src)))
+		}
+		if t < 0 {
+			panic(fmt.Sprint("t < 0:", t))
+		}
+		if s-t > e.maxOffset {
+			panic(fmt.Sprint(s, "-", t, "(", s-t, ") > maxMatchLength (", maxMatchOffset, ")"))
+		}
+	}
+	// Extend the match to be as long as possible.
+	return int32(matchLen(src[s:], src[t:]))
+}

vendor/github.com/klauspost/compress/flate/matchlen_amd64.go

@@ -0,0 +1,16 @@
+//go:build amd64 && !appengine && !noasm && gc
+// +build amd64,!appengine,!noasm,gc
+
+// Copyright 2019+ Klaus Post. All rights reserved.
+// License information can be found in the LICENSE file.
+
+package flate
+
+// matchLen returns how many bytes match in a and b
+//
+// It assumes that:
+//
+//	len(a) <= len(b) and len(a) > 0
+//
+//go:noescape
+func matchLen(a []byte, b []byte) int

vendor/github.com/klauspost/compress/flate/matchlen_amd64.s

@@ -0,0 +1,68 @@
+// Copied from S2 implementation.
+
+//go:build !appengine && !noasm && gc && !noasm
+
+#include "textflag.h"
+
+// func matchLen(a []byte, b []byte) int
+// Requires: BMI
+TEXT ·matchLen(SB), NOSPLIT, $0-56
+	MOVQ a_base+0(FP), AX
+	MOVQ b_base+24(FP), CX
+	MOVQ a_len+8(FP), DX
+
+	// matchLen
+	XORL SI, SI
+	CMPL DX, $0x08
+	JB   matchlen_match4_standalone
+
+matchlen_loopback_standalone:
+	MOVQ  (AX)(SI*1), BX
+	XORQ  (CX)(SI*1), BX
+	TESTQ BX, BX
+	JZ    matchlen_loop_standalone
+
+#ifdef GOAMD64_v3
+	TZCNTQ BX, BX
+#else
+	BSFQ BX, BX
+#endif
+	SARQ $0x03, BX
+	LEAL (SI)(BX*1), SI
+	JMP  gen_match_len_end
+
+matchlen_loop_standalone:
+	LEAL -8(DX), DX
+	LEAL 8(SI), SI
+	CMPL DX, $0x08
+	JAE  matchlen_loopback_standalone
+
+matchlen_match4_standalone:
+	CMPL DX, $0x04
+	JB   matchlen_match2_standalone
+	MOVL (AX)(SI*1), BX
+	CMPL (CX)(SI*1), BX
+	JNE  matchlen_match2_standalone
+	LEAL -4(DX), DX
+	LEAL 4(SI), SI
+
+matchlen_match2_standalone:
+	CMPL DX, $0x02
+	JB   matchlen_match1_standalone
+	MOVW (AX)(SI*1), BX
+	CMPW (CX)(SI*1), BX
+	JNE  matchlen_match1_standalone
+	LEAL -2(DX), DX
+	LEAL 2(SI), SI
+
+matchlen_match1_standalone:
+	CMPL DX, $0x01
+	JB   gen_match_len_end
+	MOVB (AX)(SI*1), BL
+	CMPB (CX)(SI*1), BL
+	JNE  gen_match_len_end
+	INCL SI
+
+gen_match_len_end:
+	MOVQ SI, ret+48(FP)
+	RET

vendor/github.com/klauspost/compress/flate/matchlen_generic.go

@@ -0,0 +1,33 @@
+//go:build !amd64 || appengine || !gc || noasm
+// +build !amd64 appengine !gc noasm
+
+// Copyright 2019+ Klaus Post. All rights reserved.
+// License information can be found in the LICENSE file.
+
+package flate
+
+import (
+	"encoding/binary"
+	"math/bits"
+)
+
+// matchLen returns the maximum common prefix length of a and b.
+// a must be the shortest of the two.
+func matchLen(a, b []byte) (n int) {
+	for ; len(a) >= 8 && len(b) >= 8; a, b = a[8:], b[8:] {
+		diff := binary.LittleEndian.Uint64(a) ^ binary.LittleEndian.Uint64(b)
+		if diff != 0 {
+			return n + bits.TrailingZeros64(diff)>>3
+		}
+		n += 8
+	}
+
+	for i := range a {
+		if a[i] != b[i] {
+			break
+		}
+		n++
+	}
+	return n
+
+}

vendor/github.com/klauspost/compress/fse/bitwriter.go

@@ -152,12 +152,11 @@ func (b *bitWriter) flushAlign() {
 
 // close will write the alignment bit and write the final byte(s)
 // to the output.
-func (b *bitWriter) close() error {
+func (b *bitWriter) close() {
 	// End mark
 	b.addBits16Clean(1, 1)
 	// flush until next byte.
 	b.flushAlign()
-	return nil
 }
 
 // reset and continue writing by appending to out.

vendor/github.com/klauspost/compress/fse/compress.go

@@ -199,7 +199,8 @@ func (s *Scratch) compress(src []byte) error {
 	c2.flush(s.actualTableLog)
 	c1.flush(s.actualTableLog)
 
-	return s.bw.close()
+	s.bw.close()
+	return nil
 }
 
 // writeCount will write the normalized histogram count to header.

vendor/github.com/klauspost/compress/huff0/bitwriter.go

@@ -94,10 +94,9 @@ func (b *bitWriter) flushAlign() {
 
 // close will write the alignment bit and write the final byte(s)
 // to the output.
-func (b *bitWriter) close() error {
+func (b *bitWriter) close() {
 	// End mark
 	b.addBits16Clean(1, 1)
 	// flush until next byte.
 	b.flushAlign()
-	return nil
 }

vendor/github.com/klauspost/compress/huff0/compress.go

@@ -227,10 +227,10 @@ func EstimateSizes(in []byte, s *Scratch) (tableSz, dataSz, reuseSz int, err err
 }
 
 func (s *Scratch) compress1X(src []byte) ([]byte, error) {
-	return s.compress1xDo(s.Out, src)
+	return s.compress1xDo(s.Out, src), nil
 }
 
-func (s *Scratch) compress1xDo(dst, src []byte) ([]byte, error) {
+func (s *Scratch) compress1xDo(dst, src []byte) []byte {
 	var bw = bitWriter{out: dst}
 
 	// N is length divisible by 4.
@@ -260,8 +260,8 @@ func (s *Scratch) compress1xDo(dst, src []byte) ([]byte, error) {
 			bw.encTwoSymbols(cTable, tmp[1], tmp[0])
 		}
 	}
-	err := bw.close()
+	bw.close()
-	return bw.out, err
+	return bw.out
 }
 
 var sixZeros [6]byte
@@ -283,12 +283,8 @@ func (s *Scratch) compress4X(src []byte) ([]byte, error) {
 		}
 		src = src[len(toDo):]
 
-		var err error
 		idx := len(s.Out)
-		s.Out, err = s.compress1xDo(s.Out, toDo)
+		s.Out = s.compress1xDo(s.Out, toDo)
-		if err != nil {
-			return nil, err
-		}
 		if len(s.Out)-idx > math.MaxUint16 {
 			// We cannot store the size in the jump table
 			return nil, ErrIncompressible
@@ -315,7 +311,6 @@ func (s *Scratch) compress4Xp(src []byte) ([]byte, error) {
 
 	segmentSize := (len(src) + 3) / 4
 	var wg sync.WaitGroup
-	var errs [4]error
 	wg.Add(4)
 	for i := 0; i < 4; i++ {
 		toDo := src
@@ -326,15 +321,12 @@ func (s *Scratch) compress4Xp(src []byte) ([]byte, error) {
 
 		// Separate goroutine for each block.
 		go func(i int) {
-			s.tmpOut[i], errs[i] = s.compress1xDo(s.tmpOut[i][:0], toDo)
+			s.tmpOut[i] = s.compress1xDo(s.tmpOut[i][:0], toDo)
 			wg.Done()
 		}(i)
 	}
 	wg.Wait()
 	for i := 0; i < 4; i++ {
-		if errs[i] != nil {
-			return nil, errs[i]
-		}
 		o := s.tmpOut[i]
 		if len(o) > math.MaxUint16 {
 			// We cannot store the size in the jump table

vendor/github.com/klauspost/compress/zstd/bitreader.go

@@ -17,7 +17,6 @@ import (
 // for aligning the input.
 type bitReader struct {
 	in       []byte
-	off      uint   // next byte to read is at in[off - 1]
 	value    uint64 // Maybe use [16]byte, but shifting is awkward.
 	bitsRead uint8
 }
@@ -28,7 +27,6 @@ func (b *bitReader) init(in []byte) error {
 		return errors.New("corrupt stream: too short")
 	}
 	b.in = in
-	b.off = uint(len(in))
 	// The highest bit of the last byte indicates where to start
 	v := in[len(in)-1]
 	if v == 0 {
@@ -69,21 +67,19 @@ func (b *bitReader) fillFast() {
 	if b.bitsRead < 32 {
 		return
 	}
-	// 2 bounds checks.
+	v := b.in[len(b.in)-4:]
-	v := b.in[b.off-4:]
+	b.in = b.in[:len(b.in)-4]
-	v = v[:4]
 	low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24)
 	b.value = (b.value << 32) | uint64(low)
 	b.bitsRead -= 32
-	b.off -= 4
 }
 
 // fillFastStart() assumes the bitreader is empty and there is at least 8 bytes to read.
 func (b *bitReader) fillFastStart() {
-	// Do single re-slice to avoid bounds checks.
+	v := b.in[len(b.in)-8:]
-	b.value = binary.LittleEndian.Uint64(b.in[b.off-8:])
+	b.in = b.in[:len(b.in)-8]
+	b.value = binary.LittleEndian.Uint64(v)
 	b.bitsRead = 0
-	b.off -= 8
 }
 
 // fill() will make sure at least 32 bits are available.
@@ -91,25 +87,25 @@ func (b *bitReader) fill() {
 	if b.bitsRead < 32 {
 		return
 	}
-	if b.off >= 4 {
+	if len(b.in) >= 4 {
-		v := b.in[b.off-4:]
+		v := b.in[len(b.in)-4:]
-		v = v[:4]
+		b.in = b.in[:len(b.in)-4]
 		low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24)
 		b.value = (b.value << 32) | uint64(low)
 		b.bitsRead -= 32
-		b.off -= 4
 		return
 	}
-	for b.off > 0 {
+
-		b.value = (b.value << 8) | uint64(b.in[b.off-1])
+	b.bitsRead -= uint8(8 * len(b.in))
-		b.bitsRead -= 8
+	for len(b.in) > 0 {
-		b.off--
+		b.value = (b.value << 8) | uint64(b.in[len(b.in)-1])
+		b.in = b.in[:len(b.in)-1]
 	}
 }
 
 // finished returns true if all bits have been read from the bit stream.
 func (b *bitReader) finished() bool {
-	return b.off == 0 && b.bitsRead >= 64
+	return len(b.in) == 0 && b.bitsRead >= 64
 }
 
 // overread returns true if more bits have been requested than is on the stream.
@@ -119,7 +115,7 @@ func (b *bitReader) overread() bool {
 
 // remain returns the number of bits remaining.
 func (b *bitReader) remain() uint {
-	return b.off*8 + 64 - uint(b.bitsRead)
+	return 8*uint(len(b.in)) + 64 - uint(b.bitsRead)
 }
 
 // close the bitstream and returns an error if out-of-buffer reads occurred.

vendor/github.com/klauspost/compress/zstd/bitwriter.go

@@ -97,12 +97,11 @@ func (b *bitWriter) flushAlign() {
 
 // close will write the alignment bit and write the final byte(s)
 // to the output.
-func (b *bitWriter) close() error {
+func (b *bitWriter) close() {
 	// End mark
 	b.addBits16Clean(1, 1)
 	// flush until next byte.
 	b.flushAlign()
-	return nil
 }
 
 // reset and continue writing by appending to out.

vendor/github.com/klauspost/compress/zstd/blockenc.go

@@ -361,14 +361,21 @@ func (b *blockEnc) encodeLits(lits []byte, raw bool) error {
 	if len(lits) >= 1024 {
 		// Use 4 Streams.
 		out, reUsed, err = huff0.Compress4X(lits, b.litEnc)
-	} else if len(lits) > 32 {
+	} else if len(lits) > 16 {
 		// Use 1 stream
 		single = true
 		out, reUsed, err = huff0.Compress1X(lits, b.litEnc)
 	} else {
 		err = huff0.ErrIncompressible
 	}
-
+	if err == nil && len(out)+5 > len(lits) {
+		// If we are close, we may still be worse or equal to raw.
+		var lh literalsHeader
+		lh.setSizes(len(out), len(lits), single)
+		if len(out)+lh.size() >= len(lits) {
+			err = huff0.ErrIncompressible
+		}
+	}
 	switch err {
 	case huff0.ErrIncompressible:
 		if debugEncoder {
@@ -503,7 +510,7 @@ func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error {
 	if len(b.literals) >= 1024 && !raw {
 		// Use 4 Streams.
 		out, reUsed, err = huff0.Compress4X(b.literals, b.litEnc)
-	} else if len(b.literals) > 32 && !raw {
+	} else if len(b.literals) > 16 && !raw {
 		// Use 1 stream
 		single = true
 		out, reUsed, err = huff0.Compress1X(b.literals, b.litEnc)
@@ -511,6 +518,17 @@ func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error {
 		err = huff0.ErrIncompressible
 	}
 
+	if err == nil && len(out)+5 > len(b.literals) {
+		// If we are close, we may still be worse or equal to raw.
+		var lh literalsHeader
+		lh.setSize(len(b.literals))
+		szRaw := lh.size()
+		lh.setSizes(len(out), len(b.literals), single)
+		szComp := lh.size()
+		if len(out)+szComp >= len(b.literals)+szRaw {
+			err = huff0.ErrIncompressible
+		}
+	}
 	switch err {
 	case huff0.ErrIncompressible:
 		lh.setType(literalsBlockRaw)
@@ -773,10 +791,7 @@ func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error {
 	ml.flush(mlEnc.actualTableLog)
 	of.flush(ofEnc.actualTableLog)
 	ll.flush(llEnc.actualTableLog)
-	err = wr.close()
+	wr.close()
-	if err != nil {
-		return err
-	}
 	b.output = wr.out
 
 	// Maybe even add a bigger margin.

vendor/github.com/klauspost/compress/zstd/dict.go

@@ -1,10 +1,13 @@
 package zstd
 
 import (
+	"bytes"
 	"encoding/binary"
 	"errors"
 	"fmt"
 	"io"
+	"math"
+	"sort"
 
 	"github.com/klauspost/compress/huff0"
 )
@@ -14,7 +17,6 @@ type dict struct {
 
 	litEnc              *huff0.Scratch
 	llDec, ofDec, mlDec sequenceDec
-	//llEnc, ofEnc, mlEnc []*fseEncoder
 	offsets             [3]int
 	content             []byte
 }
@@ -159,3 +161,374 @@ func InspectDictionary(b []byte) (interface {
 	d, err := loadDict(b)
 	return d, err
 }
+
+type BuildDictOptions struct {
+	// Dictionary ID.
+	ID uint32
+
+	// Content to use to create dictionary tables.
+	Contents [][]byte
+
+	// History to use for all blocks.
+	History []byte
+
+	// Offsets to use.
+	Offsets [3]int
+
+	// CompatV155 will make the dictionary compatible with Zstd v1.5.5 and earlier.
+	// See https://github.com/facebook/zstd/issues/3724
+	CompatV155 bool
+
+	// Use the specified encoder level.
+	// The dictionary will be built using the specified encoder level,
+	// which will reflect speed and make the dictionary tailored for that level.
+	// If not set SpeedBestCompression will be used.
+	Level EncoderLevel
+
+	// DebugOut will write stats and other details here if set.
+	DebugOut io.Writer
+}
+
+func BuildDict(o BuildDictOptions) ([]byte, error) {
+	initPredefined()
+	hist := o.History
+	contents := o.Contents
+	debug := o.DebugOut != nil
+	println := func(args ...interface{}) {
+		if o.DebugOut != nil {
+			fmt.Fprintln(o.DebugOut, args...)
+		}
+	}
+	printf := func(s string, args ...interface{}) {
+		if o.DebugOut != nil {
+			fmt.Fprintf(o.DebugOut, s, args...)
+		}
+	}
+	print := func(args ...interface{}) {
+		if o.DebugOut != nil {
+			fmt.Fprint(o.DebugOut, args...)
+		}
+	}
+
+	if int64(len(hist)) > dictMaxLength {
+		return nil, fmt.Errorf("dictionary of size %d > %d", len(hist), int64(dictMaxLength))
+	}
+	if len(hist) < 8 {
+		return nil, fmt.Errorf("dictionary of size %d < %d", len(hist), 8)
+	}
+	if len(contents) == 0 {
+		return nil, errors.New("no content provided")
+	}
+	d := dict{
+		id:      o.ID,
+		litEnc:  nil,
+		llDec:   sequenceDec{},
+		ofDec:   sequenceDec{},
+		mlDec:   sequenceDec{},
+		offsets: o.Offsets,
+		content: hist,
+	}
+	block := blockEnc{lowMem: false}
+	block.init()
+	enc := encoder(&bestFastEncoder{fastBase: fastBase{maxMatchOff: int32(maxMatchLen), bufferReset: math.MaxInt32 - int32(maxMatchLen*2), lowMem: false}})
+	if o.Level != 0 {
+		eOpts := encoderOptions{
+			level:      o.Level,
+			blockSize:  maxMatchLen,
+			windowSize: maxMatchLen,
+			dict:       &d,
+			lowMem:     false,
+		}
+		enc = eOpts.encoder()
+	} else {
+		o.Level = SpeedBestCompression
+	}
+	var (
+		remain [256]int
+		ll     [256]int
+		ml     [256]int
+		of     [256]int
+	)
+	addValues := func(dst *[256]int, src []byte) {
+		for _, v := range src {
+			dst[v]++
+		}
+	}
+	addHist := func(dst *[256]int, src *[256]uint32) {
+		for i, v := range src {
+			dst[i] += int(v)
+		}
+	}
+	seqs := 0
+	nUsed := 0
+	litTotal := 0
+	newOffsets := make(map[uint32]int, 1000)
+	for _, b := range contents {
+		block.reset(nil)
+		if len(b) < 8 {
+			continue
+		}
+		nUsed++
+		enc.Reset(&d, true)
+		enc.Encode(&block, b)
+		addValues(&remain, block.literals)
+		litTotal += len(block.literals)
+		seqs += len(block.sequences)
+		block.genCodes()
+		addHist(&ll, block.coders.llEnc.Histogram())
+		addHist(&ml, block.coders.mlEnc.Histogram())
+		addHist(&of, block.coders.ofEnc.Histogram())
+		for i, seq := range block.sequences {
+			if i > 3 {
+				break
+			}
+			offset := seq.offset
+			if offset == 0 {
+				continue
+			}
+			if offset > 3 {
+				newOffsets[offset-3]++
+			} else {
+				newOffsets[uint32(o.Offsets[offset-1])]++
+			}
+		}
+	}
+	// Find most used offsets.
+	var sortedOffsets []uint32
+	for k := range newOffsets {
+		sortedOffsets = append(sortedOffsets, k)
+	}
+	sort.Slice(sortedOffsets, func(i, j int) bool {
+		a, b := sortedOffsets[i], sortedOffsets[j]
+		if a == b {
+			// Prefer the longer offset
+			return sortedOffsets[i] > sortedOffsets[j]
+		}
+		return newOffsets[sortedOffsets[i]] > newOffsets[sortedOffsets[j]]
+	})
+	if len(sortedOffsets) > 3 {
+		if debug {
+			print("Offsets:")
+			for i, v := range sortedOffsets {
+				if i > 20 {
+					break
+				}
+				printf("[%d: %d],", v, newOffsets[v])
+			}
+			println("")
+		}
+
+		sortedOffsets = sortedOffsets[:3]
+	}
+	for i, v := range sortedOffsets {
+		o.Offsets[i] = int(v)
+	}
+	if debug {
+		println("New repeat offsets", o.Offsets)
+	}
+
+	if nUsed == 0 || seqs == 0 {
+		return nil, fmt.Errorf("%d blocks, %d sequences found", nUsed, seqs)
+	}
+	if debug {
+		println("Sequences:", seqs, "Blocks:", nUsed, "Literals:", litTotal)
+	}
+	if seqs/nUsed < 512 {
+		// Use 512 as minimum.
+		nUsed = seqs / 512
+	}
+	copyHist := func(dst *fseEncoder, src *[256]int) ([]byte, error) {
+		hist := dst.Histogram()
+		var maxSym uint8
+		var maxCount int
+		var fakeLength int
+		for i, v := range src {
+			if v > 0 {
+				v = v / nUsed
+				if v == 0 {
+					v = 1
+				}
+			}
+			if v > maxCount {
+				maxCount = v
+			}
+			if v != 0 {
+				maxSym = uint8(i)
+			}
+			fakeLength += v
+			hist[i] = uint32(v)
+		}
+		dst.HistogramFinished(maxSym, maxCount)
+		dst.reUsed = false
+		dst.useRLE = false
+		err := dst.normalizeCount(fakeLength)
+		if err != nil {
+			return nil, err
+		}
+		if debug {
+			println("RAW:", dst.count[:maxSym+1], "NORM:", dst.norm[:maxSym+1], "LEN:", fakeLength)
+		}
+		return dst.writeCount(nil)
+	}
+	if debug {
+		print("Literal lengths: ")
+	}
+	llTable, err := copyHist(block.coders.llEnc, &ll)
+	if err != nil {
+		return nil, err
+	}
+	if debug {
+		print("Match lengths: ")
+	}
+	mlTable, err := copyHist(block.coders.mlEnc, &ml)
+	if err != nil {
+		return nil, err
+	}
+	if debug {
+		print("Offsets: ")
+	}
+	ofTable, err := copyHist(block.coders.ofEnc, &of)
+	if err != nil {
+		return nil, err
+	}
+
+	// Literal table
+	avgSize := litTotal
+	if avgSize > huff0.BlockSizeMax/2 {
+		avgSize = huff0.BlockSizeMax / 2
+	}
+	huffBuff := make([]byte, 0, avgSize)
+	// Target size
+	div := litTotal / avgSize
+	if div < 1 {
+		div = 1
+	}
+	if debug {
+		println("Huffman weights:")
+	}
+	for i, n := range remain[:] {
+		if n > 0 {
+			n = n / div
+			// Allow all entries to be represented.
+			if n == 0 {
+				n = 1
+			}
+			huffBuff = append(huffBuff, bytes.Repeat([]byte{byte(i)}, n)...)
+			if debug {
+				printf("[%d: %d], ", i, n)
+			}
+		}
+	}
+	if o.CompatV155 && remain[255]/div == 0 {
+		huffBuff = append(huffBuff, 255)
+	}
+	scratch := &huff0.Scratch{TableLog: 11}
+	for tries := 0; tries < 255; tries++ {
+		scratch = &huff0.Scratch{TableLog: 11}
+		_, _, err = huff0.Compress1X(huffBuff, scratch)
+		if err == nil {
+			break
+		}
+		if debug {
+			printf("Try %d: Huffman error: %v\n", tries+1, err)
+		}
+		huffBuff = huffBuff[:0]
+		if tries == 250 {
+			if debug {
+				println("Huffman: Bailing out with predefined table")
+			}
+
+			// Bail out.... Just generate something
+			huffBuff = append(huffBuff, bytes.Repeat([]byte{255}, 10000)...)
+			for i := 0; i < 128; i++ {
+				huffBuff = append(huffBuff, byte(i))
+			}
+			continue
+		}
+		if errors.Is(err, huff0.ErrIncompressible) {
+			// Try truncating least common.
+			for i, n := range remain[:] {
+				if n > 0 {
+					n = n / (div * (i + 1))
+					if n > 0 {
+						huffBuff = append(huffBuff, bytes.Repeat([]byte{byte(i)}, n)...)
+					}
+				}
+			}
+			if o.CompatV155 && len(huffBuff) > 0 && huffBuff[len(huffBuff)-1] != 255 {
+				huffBuff = append(huffBuff, 255)
+			}
+			if len(huffBuff) == 0 {
+				huffBuff = append(huffBuff, 0, 255)
+			}
+		}
+		if errors.Is(err, huff0.ErrUseRLE) {
+			for i, n := range remain[:] {
+				n = n / (div * (i + 1))
+				// Allow all entries to be represented.
+				if n == 0 {
+					n = 1
+				}
+				huffBuff = append(huffBuff, bytes.Repeat([]byte{byte(i)}, n)...)
+			}
+		}
+	}
+
+	var out bytes.Buffer
+	out.Write([]byte(dictMagic))
+	out.Write(binary.LittleEndian.AppendUint32(nil, o.ID))
+	out.Write(scratch.OutTable)
+	if debug {
+		println("huff table:", len(scratch.OutTable), "bytes")
+		println("of table:", len(ofTable), "bytes")
+		println("ml table:", len(mlTable), "bytes")
+		println("ll table:", len(llTable), "bytes")
+	}
+	out.Write(ofTable)
+	out.Write(mlTable)
+	out.Write(llTable)
+	out.Write(binary.LittleEndian.AppendUint32(nil, uint32(o.Offsets[0])))
+	out.Write(binary.LittleEndian.AppendUint32(nil, uint32(o.Offsets[1])))
+	out.Write(binary.LittleEndian.AppendUint32(nil, uint32(o.Offsets[2])))
+	out.Write(hist)
+	if debug {
+		_, err := loadDict(out.Bytes())
+		if err != nil {
+			panic(err)
+		}
+		i, err := InspectDictionary(out.Bytes())
+		if err != nil {
+			panic(err)
+		}
+		println("ID:", i.ID())
+		println("Content size:", i.ContentSize())
+		println("Encoder:", i.LitEncoder() != nil)
+		println("Offsets:", i.Offsets())
+		var totalSize int
+		for _, b := range contents {
+			totalSize += len(b)
+		}
+
+		encWith := func(opts ...EOption) int {
+			enc, err := NewWriter(nil, opts...)
+			if err != nil {
+				panic(err)
+			}
+			defer enc.Close()
+			var dst []byte
+			var totalSize int
+			for _, b := range contents {
+				dst = enc.EncodeAll(b, dst[:0])
+				totalSize += len(dst)
+			}
+			return totalSize
+		}
+		plain := encWith(WithEncoderLevel(o.Level))
+		withDict := encWith(WithEncoderLevel(o.Level), WithEncoderDict(out.Bytes()))
+		println("Input size:", totalSize)
+		println("Plain Compressed:", plain)
+		println("Dict Compressed:", withDict)
+		println("Saved:", plain-withDict, (plain-withDict)/len(contents), "bytes per input (rounded down)")
+	}
+	return out.Bytes(), nil
+}

vendor/github.com/klauspost/compress/zstd/enc_best.go

@@ -197,12 +197,13 @@ encodeLoop:
 
 		// Set m to a match at offset if it looks like that will improve compression.
 		improve := func(m *match, offset int32, s int32, first uint32, rep int32) {
-			if s-offset >= e.maxMatchOff || load3232(src, offset) != first {
+			delta := s - offset
+			if delta >= e.maxMatchOff || delta <= 0 || load3232(src, offset) != first {
 				return
 			}
 			if debugAsserts {
-				if offset <= 0 {
+				if offset >= s {
-					panic(offset)
+					panic(fmt.Sprintf("offset: %d - s:%d - rep: %d - cur :%d - max: %d", offset, s, rep, e.cur, e.maxMatchOff))
 				}
 				if !bytes.Equal(src[s:s+4], src[offset:offset+4]) {
 					panic(fmt.Sprintf("first match mismatch: %v != %v, first: %08x", src[s:s+4], src[offset:offset+4], first))
@@ -343,8 +344,8 @@ encodeLoop:
 		if best.rep > 0 {
 			var seq seq
 			seq.matchLen = uint32(best.length - zstdMinMatch)
-			if debugAsserts && s <= nextEmit {
+			if debugAsserts && s < nextEmit {
-				panic("s <= nextEmit")
+				panic("s < nextEmit")
 			}
 			addLiterals(&seq, best.s)
 

vendor/github.com/klauspost/compress/zstd/encoder.go

@@ -227,10 +227,7 @@ func (e *Encoder) nextBlock(final bool) error {
 			DictID:        e.o.dict.ID(),
 		}
 
-		dst, err := fh.appendTo(tmp[:0])
+		dst := fh.appendTo(tmp[:0])
-		if err != nil {
-			return err
-		}
 		s.headerWritten = true
 		s.wWg.Wait()
 		var n2 int
@@ -483,7 +480,7 @@ func (e *Encoder) EncodeAll(src, dst []byte) []byte {
 				Checksum: false,
 				DictID:   0,
 			}
-			dst, _ = fh.appendTo(dst)
+			dst = fh.appendTo(dst)
 
 			// Write raw block as last one only.
 			var blk blockHeader
@@ -518,10 +515,7 @@ func (e *Encoder) EncodeAll(src, dst []byte) []byte {
 	if len(dst) == 0 && cap(dst) == 0 && len(src) < 1<<20 && !e.o.lowMem {
 		dst = make([]byte, 0, len(src))
 	}
-	dst, err := fh.appendTo(dst)
+	dst = fh.appendTo(dst)
-	if err != nil {
-		panic(err)
-	}
 
 	// If we can do everything in one block, prefer that.
 	if len(src) <= e.o.blockSize {
@@ -581,6 +575,7 @@ func (e *Encoder) EncodeAll(src, dst []byte) []byte {
 	// Add padding with content from crypto/rand.Reader
 	if e.o.pad > 0 {
 		add := calcSkippableFrame(int64(len(dst)), int64(e.o.pad))
+		var err error
 		dst, err = skippableFrame(dst, add, rand.Reader)
 		if err != nil {
 			panic(err)

vendor/github.com/klauspost/compress/zstd/frameenc.go

@@ -22,7 +22,7 @@ type frameHeader struct {
 
 const maxHeaderSize = 14
 
-func (f frameHeader) appendTo(dst []byte) ([]byte, error) {
+func (f frameHeader) appendTo(dst []byte) []byte {
 	dst = append(dst, frameMagic...)
 	var fhd uint8
 	if f.Checksum {
@@ -88,7 +88,7 @@ func (f frameHeader) appendTo(dst []byte) ([]byte, error) {
 	default:
 		panic("invalid fcs")
 	}
-	return dst, nil
+	return dst
 }
 
 const skippableFrameHeader = 4 + 4

vendor/github.com/klauspost/compress/zstd/seqdec.go

@@ -245,7 +245,7 @@ func (s *sequenceDecs) decodeSync(hist []byte) error {
 			return io.ErrUnexpectedEOF
 		}
 		var ll, mo, ml int
-		if br.off > 4+((maxOffsetBits+16+16)>>3) {
+		if len(br.in) > 4+((maxOffsetBits+16+16)>>3) {
 			// inlined function:
 			// ll, mo, ml = s.nextFast(br, llState, mlState, ofState)
 
@@ -452,18 +452,13 @@ func (s *sequenceDecs) next(br *bitReader, llState, mlState, ofState decSymbol)
 
 	// extra bits are stored in reverse order.
 	br.fill()
-	if s.maxBits <= 32 {
-		mo += br.getBits(moB)
-		ml += br.getBits(mlB)
-		ll += br.getBits(llB)
-	} else {
 	mo += br.getBits(moB)
+	if s.maxBits > 32 {
 		br.fill()
+	}
 	// matchlength+literal length, max 32 bits
 	ml += br.getBits(mlB)
 	ll += br.getBits(llB)
-
-	}
 	mo = s.adjustOffset(mo, ll, moB)
 	return
 }

vendor/github.com/klauspost/compress/zstd/seqdec_amd64.s

@@ -5,11 +5,11 @@
 // func sequenceDecs_decode_amd64(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int
 // Requires: CMOV
 TEXT ·sequenceDecs_decode_amd64(SB), $8-32
-	MOVQ    br+8(FP), AX
+	MOVQ    br+8(FP), CX
-	MOVQ    32(AX), DX
+	MOVQ    24(CX), DX
-	MOVBQZX 40(AX), BX
+	MOVBQZX 32(CX), BX
-	MOVQ    24(AX), SI
+	MOVQ    (CX), AX
-	MOVQ    (AX), AX
+	MOVQ    8(CX), SI
 	ADDQ    SI, AX
 	MOVQ    AX, (SP)
 	MOVQ    ctx+16(FP), AX
@@ -301,9 +301,9 @@ sequenceDecs_decode_amd64_match_len_ofs_ok:
 	MOVQ R12, 152(AX)
 	MOVQ R13, 160(AX)
 	MOVQ br+8(FP), AX
-	MOVQ DX, 32(AX)
+	MOVQ DX, 24(AX)
-	MOVB BL, 40(AX)
+	MOVB BL, 32(AX)
-	MOVQ SI, 24(AX)
+	MOVQ SI, 8(AX)
 
 	// Return success
 	MOVQ $0x00000000, ret+24(FP)
@@ -336,11 +336,11 @@ error_overread:
 // func sequenceDecs_decode_56_amd64(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int
 // Requires: CMOV
 TEXT ·sequenceDecs_decode_56_amd64(SB), $8-32
-	MOVQ    br+8(FP), AX
+	MOVQ    br+8(FP), CX
-	MOVQ    32(AX), DX
+	MOVQ    24(CX), DX
-	MOVBQZX 40(AX), BX
+	MOVBQZX 32(CX), BX
-	MOVQ    24(AX), SI
+	MOVQ    (CX), AX
-	MOVQ    (AX), AX
+	MOVQ    8(CX), SI
 	ADDQ    SI, AX
 	MOVQ    AX, (SP)
 	MOVQ    ctx+16(FP), AX
@@ -603,9 +603,9 @@ sequenceDecs_decode_56_amd64_match_len_ofs_ok:
 	MOVQ R12, 152(AX)
 	MOVQ R13, 160(AX)
 	MOVQ br+8(FP), AX
-	MOVQ DX, 32(AX)
+	MOVQ DX, 24(AX)
-	MOVB BL, 40(AX)
+	MOVB BL, 32(AX)
-	MOVQ SI, 24(AX)
+	MOVQ SI, 8(AX)
 
 	// Return success
 	MOVQ $0x00000000, ret+24(FP)
@@ -638,11 +638,11 @@ error_overread:
 // func sequenceDecs_decode_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int
 // Requires: BMI, BMI2, CMOV
 TEXT ·sequenceDecs_decode_bmi2(SB), $8-32
-	MOVQ    br+8(FP), CX
+	MOVQ    br+8(FP), BX
-	MOVQ    32(CX), AX
+	MOVQ    24(BX), AX
-	MOVBQZX 40(CX), DX
+	MOVBQZX 32(BX), DX
-	MOVQ    24(CX), BX
+	MOVQ    (BX), CX
-	MOVQ    (CX), CX
+	MOVQ    8(BX), BX
 	ADDQ    BX, CX
 	MOVQ    CX, (SP)
 	MOVQ    ctx+16(FP), CX
@@ -892,9 +892,9 @@ sequenceDecs_decode_bmi2_match_len_ofs_ok:
 	MOVQ R11, 152(CX)
 	MOVQ R12, 160(CX)
 	MOVQ br+8(FP), CX
-	MOVQ AX, 32(CX)
+	MOVQ AX, 24(CX)
-	MOVB DL, 40(CX)
+	MOVB DL, 32(CX)
-	MOVQ BX, 24(CX)
+	MOVQ BX, 8(CX)
 
 	// Return success
 	MOVQ $0x00000000, ret+24(FP)
@@ -927,11 +927,11 @@ error_overread:
 // func sequenceDecs_decode_56_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int
 // Requires: BMI, BMI2, CMOV
 TEXT ·sequenceDecs_decode_56_bmi2(SB), $8-32
-	MOVQ    br+8(FP), CX
+	MOVQ    br+8(FP), BX
-	MOVQ    32(CX), AX
+	MOVQ    24(BX), AX
-	MOVBQZX 40(CX), DX
+	MOVBQZX 32(BX), DX
-	MOVQ    24(CX), BX
+	MOVQ    (BX), CX
-	MOVQ    (CX), CX
+	MOVQ    8(BX), BX
 	ADDQ    BX, CX
 	MOVQ    CX, (SP)
 	MOVQ    ctx+16(FP), CX
@@ -1152,9 +1152,9 @@ sequenceDecs_decode_56_bmi2_match_len_ofs_ok:
 	MOVQ R11, 152(CX)
 	MOVQ R12, 160(CX)
 	MOVQ br+8(FP), CX
-	MOVQ AX, 32(CX)
+	MOVQ AX, 24(CX)
-	MOVB DL, 40(CX)
+	MOVB DL, 32(CX)
-	MOVQ BX, 24(CX)
+	MOVQ BX, 8(CX)
 
 	// Return success
 	MOVQ $0x00000000, ret+24(FP)
@@ -1797,11 +1797,11 @@ empty_seqs:
 // func sequenceDecs_decodeSync_amd64(s *sequenceDecs, br *bitReader, ctx *decodeSyncAsmContext) int
 // Requires: CMOV, SSE
 TEXT ·sequenceDecs_decodeSync_amd64(SB), $64-32
-	MOVQ    br+8(FP), AX
+	MOVQ    br+8(FP), CX
-	MOVQ    32(AX), DX
+	MOVQ    24(CX), DX
-	MOVBQZX 40(AX), BX
+	MOVBQZX 32(CX), BX
-	MOVQ    24(AX), SI
+	MOVQ    (CX), AX
-	MOVQ    (AX), AX
+	MOVQ    8(CX), SI
 	ADDQ    SI, AX
 	MOVQ    AX, (SP)
 	MOVQ    ctx+16(FP), AX
@@ -2295,9 +2295,9 @@ handle_loop:
 
 loop_finished:
 	MOVQ br+8(FP), AX
-	MOVQ DX, 32(AX)
+	MOVQ DX, 24(AX)
-	MOVB BL, 40(AX)
+	MOVB BL, 32(AX)
-	MOVQ SI, 24(AX)
+	MOVQ SI, 8(AX)
 
 	// Update the context
 	MOVQ ctx+16(FP), AX
@@ -2362,11 +2362,11 @@ error_not_enough_space:
 // func sequenceDecs_decodeSync_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeSyncAsmContext) int
 // Requires: BMI, BMI2, CMOV, SSE
 TEXT ·sequenceDecs_decodeSync_bmi2(SB), $64-32
-	MOVQ    br+8(FP), CX
+	MOVQ    br+8(FP), BX
-	MOVQ    32(CX), AX
+	MOVQ    24(BX), AX
-	MOVBQZX 40(CX), DX
+	MOVBQZX 32(BX), DX
-	MOVQ    24(CX), BX
+	MOVQ    (BX), CX
-	MOVQ    (CX), CX
+	MOVQ    8(BX), BX
 	ADDQ    BX, CX
 	MOVQ    CX, (SP)
 	MOVQ    ctx+16(FP), CX
@@ -2818,9 +2818,9 @@ handle_loop:
 
 loop_finished:
 	MOVQ br+8(FP), CX
-	MOVQ AX, 32(CX)
+	MOVQ AX, 24(CX)
-	MOVB DL, 40(CX)
+	MOVB DL, 32(CX)
-	MOVQ BX, 24(CX)
+	MOVQ BX, 8(CX)
 
 	// Update the context
 	MOVQ ctx+16(FP), AX
@@ -2885,11 +2885,11 @@ error_not_enough_space:
 // func sequenceDecs_decodeSync_safe_amd64(s *sequenceDecs, br *bitReader, ctx *decodeSyncAsmContext) int
 // Requires: CMOV, SSE
 TEXT ·sequenceDecs_decodeSync_safe_amd64(SB), $64-32
-	MOVQ    br+8(FP), AX
+	MOVQ    br+8(FP), CX
-	MOVQ    32(AX), DX
+	MOVQ    24(CX), DX
-	MOVBQZX 40(AX), BX
+	MOVBQZX 32(CX), BX
-	MOVQ    24(AX), SI
+	MOVQ    (CX), AX
-	MOVQ    (AX), AX
+	MOVQ    8(CX), SI
 	ADDQ    SI, AX
 	MOVQ    AX, (SP)
 	MOVQ    ctx+16(FP), AX
@@ -3485,9 +3485,9 @@ handle_loop:
 
 loop_finished:
 	MOVQ br+8(FP), AX
-	MOVQ DX, 32(AX)
+	MOVQ DX, 24(AX)
-	MOVB BL, 40(AX)
+	MOVB BL, 32(AX)
-	MOVQ SI, 24(AX)
+	MOVQ SI, 8(AX)
 
 	// Update the context
 	MOVQ ctx+16(FP), AX
@@ -3552,11 +3552,11 @@ error_not_enough_space:
 // func sequenceDecs_decodeSync_safe_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeSyncAsmContext) int
 // Requires: BMI, BMI2, CMOV, SSE
 TEXT ·sequenceDecs_decodeSync_safe_bmi2(SB), $64-32
-	MOVQ    br+8(FP), CX
+	MOVQ    br+8(FP), BX
-	MOVQ    32(CX), AX
+	MOVQ    24(BX), AX
-	MOVBQZX 40(CX), DX
+	MOVBQZX 32(BX), DX
-	MOVQ    24(CX), BX
+	MOVQ    (BX), CX
-	MOVQ    (CX), CX
+	MOVQ    8(BX), BX
 	ADDQ    BX, CX
 	MOVQ    CX, (SP)
 	MOVQ    ctx+16(FP), CX
@@ -4110,9 +4110,9 @@ handle_loop:
 
 loop_finished:
 	MOVQ br+8(FP), CX
-	MOVQ AX, 32(CX)
+	MOVQ AX, 24(CX)
-	MOVB DL, 40(CX)
+	MOVB DL, 32(CX)
-	MOVQ BX, 24(CX)
+	MOVQ BX, 8(CX)
 
 	// Update the context
 	MOVQ ctx+16(FP), AX

vendor/github.com/klauspost/compress/zstd/seqdec_generic.go

@@ -29,7 +29,7 @@ func (s *sequenceDecs) decode(seqs []seqVals) error {
 	}
 	for i := range seqs {
 		var ll, mo, ml int
-		if br.off > 4+((maxOffsetBits+16+16)>>3) {
+		if len(br.in) > 4+((maxOffsetBits+16+16)>>3) {
 			// inlined function:
 			// ll, mo, ml = s.nextFast(br, llState, mlState, ofState)
 

vendor/github.com/klauspost/compress/zstd/snappy.go

@@ -95,10 +95,9 @@ func (r *SnappyConverter) Convert(in io.Reader, w io.Writer) (int64, error) {
 	var written int64
 	var readHeader bool
 	{
-		var header []byte
+		header := frameHeader{WindowSize: snappyMaxBlockSize}.appendTo(r.buf[:0])
-		var n int
-		header, r.err = frameHeader{WindowSize: snappyMaxBlockSize}.appendTo(r.buf[:0])
 
+		var n int
 		n, r.err = w.Write(header)
 		if r.err != nil {
 			return written, r.err

vendor/modules.txt

@@ -231,7 +231,7 @@ github.com/docker/distribution/reference
 github.com/docker/distribution/registry/api/errcode
 github.com/docker/distribution/registry/api/v2
 github.com/docker/distribution/registry/client/auth/challenge
-# github.com/docker/docker v24.0.6+incompatible
+# github.com/docker/docker v24.0.7+incompatible
 ## explicit
 github.com/docker/docker/api
 github.com/docker/docker/api/types
@@ -371,7 +371,7 @@ github.com/josharian/intern
 # github.com/json-iterator/go v1.1.12
 ## explicit; go 1.12
 github.com/json-iterator/go
-# github.com/klauspost/compress v1.16.7
+# github.com/klauspost/compress v1.17.2
 ## explicit; go 1.18
 github.com/klauspost/compress
 github.com/klauspost/compress/flate