From cde0441dc890ae401d7c6956a4f909aebe8ef5fd Mon Sep 17 00:00:00 2001 From: Sebastiaan van Stijn Date: Tue, 19 Sep 2023 16:38:35 +0200 Subject: [PATCH] vendor: github.com/klauspost/compress v1.16.5 full diff: https://github.com/klauspost/compress/compare/v1.16.3...v1.16.5 Signed-off-by: Sebastiaan van Stijn (cherry picked from commit 497b13c661f416aa9759db031188825df0ae9370) Signed-off-by: Sebastiaan van Stijn --- vendor.mod | 2 +- vendor.sum | 4 +- .../github.com/klauspost/compress/README.md | 11 + .../klauspost/compress/zstd/blockenc.go | 9 +- .../klauspost/compress/zstd/bytebuf.go | 2 +- .../klauspost/compress/zstd/decoder.go | 7 +- .../klauspost/compress/zstd/enc_best.go | 225 ++++++++---------- .../klauspost/compress/zstd/encoder.go | 78 +----- .../compress/zstd/encoder_options.go | 4 +- .../klauspost/compress/zstd/framedec.go | 35 +-- .../klauspost/compress/zstd/seqdec.go | 5 +- .../klauspost/compress/zstd/seqdec_amd64.go | 16 ++ .../klauspost/compress/zstd/seqdec_amd64.s | 124 ++++++++-- .../klauspost/compress/zstd/zstd.go | 4 +- vendor/modules.txt | 2 +- 15 files changed, 278 insertions(+), 250 deletions(-) diff --git a/vendor.mod b/vendor.mod index 0338823eca..8a883e1f19 100644 --- a/vendor.mod +++ b/vendor.mod @@ -57,7 +57,7 @@ require ( github.com/golang/protobuf v1.5.2 // indirect github.com/gorilla/mux v1.8.0 // indirect github.com/inconshreveable/mousetrap v1.1.0 // indirect - github.com/klauspost/compress v1.16.3 // indirect + github.com/klauspost/compress v1.16.5 // indirect github.com/matttproud/golang_protobuf_extensions v1.0.4 // indirect github.com/miekg/pkcs11 v1.1.1 // indirect github.com/moby/sys/symlink v0.2.0 // indirect diff --git a/vendor.sum b/vendor.sum index 5de667e3f2..7a368153d0 100644 --- a/vendor.sum +++ b/vendor.sum @@ -243,8 +243,8 @@ github.com/julienschmidt/httprouter v1.3.0/go.mod h1:JR6WtHb+2LUe8TCKY3cZOxFyyO8 github.com/kisielk/errcheck v1.2.0/go.mod h1:/BMXB+zMLi60iA8Vv6Ksmxu/1UDYcXs4uQLJ+jE2L00= github.com/kisielk/errcheck v1.5.0/go.mod h1:pFxgyoBC7bSaBwPgfKdkLd5X25qrDl4LWUI2bnpBCr8= github.com/kisielk/gotool v1.0.0/go.mod h1:XhKaO+MFFWcvkIS/tQcRk01m1F5IRFswLeQ+oQHNcck= -github.com/klauspost/compress v1.16.3 h1:XuJt9zzcnaz6a16/OU53ZjWp/v7/42WcR5t2a0PcNQY= -github.com/klauspost/compress v1.16.3/go.mod h1:ntbaceVETuRiXiv4DpjP66DpAtAGkEQskQzEyD//IeE= +github.com/klauspost/compress v1.16.5 h1:IFV2oUNUzZaz+XyusxpLzpzS8Pt5rh0Z16For/djlyI= +github.com/klauspost/compress v1.16.5/go.mod h1:ntbaceVETuRiXiv4DpjP66DpAtAGkEQskQzEyD//IeE= github.com/konsorten/go-windows-terminal-sequences v1.0.1/go.mod h1:T0+1ngSBFLxvqU3pZ+m/2kptfBszLMUkC4ZK/EgS/cQ= github.com/konsorten/go-windows-terminal-sequences v1.0.3/go.mod h1:T0+1ngSBFLxvqU3pZ+m/2kptfBszLMUkC4ZK/EgS/cQ= github.com/kr/logfmt v0.0.0-20140226030751-b84e30acd515/go.mod h1:+0opPa2QZZtGFBFZlji/RkVcI2GknAs/DXo4wKdlNEc= diff --git a/vendor/github.com/klauspost/compress/README.md b/vendor/github.com/klauspost/compress/README.md index c2c7252fe1..efab55e655 100644 --- a/vendor/github.com/klauspost/compress/README.md +++ b/vendor/github.com/klauspost/compress/README.md @@ -16,6 +16,15 @@ This package provides various compression algorithms. # changelog +* Apr 5, 2023 - [v1.16.4](https://github.com/klauspost/compress/releases/tag/v1.16.4) + * zstd: Improve zstd best efficiency by @greatroar and @klauspost in https://github.com/klauspost/compress/pull/784 + * zstd: Respect WithAllLitEntropyCompression https://github.com/klauspost/compress/pull/792 + * zstd: Fix amd64 not always detecting corrupt data https://github.com/klauspost/compress/pull/785 + * zstd: Various minor improvements by @greatroar in https://github.com/klauspost/compress/pull/788 https://github.com/klauspost/compress/pull/794 https://github.com/klauspost/compress/pull/795 + * s2: Fix huge block overflow https://github.com/klauspost/compress/pull/779 + * s2: Allow CustomEncoder fallback https://github.com/klauspost/compress/pull/780 + * gzhttp: Suppport ResponseWriter Unwrap() in gzhttp handler by @jgimenez in https://github.com/klauspost/compress/pull/799 + * Mar 13, 2023 - [v1.16.1](https://github.com/klauspost/compress/releases/tag/v1.16.1) * zstd: Speed up + improve best encoder by @greatroar in https://github.com/klauspost/compress/pull/776 * gzhttp: Add optional [BREACH mitigation](https://github.com/klauspost/compress/tree/master/gzhttp#breach-mitigation). https://github.com/klauspost/compress/pull/762 https://github.com/klauspost/compress/pull/768 https://github.com/klauspost/compress/pull/769 https://github.com/klauspost/compress/pull/770 https://github.com/klauspost/compress/pull/767 @@ -615,6 +624,8 @@ Here are other packages of good quality and pure Go (no cgo wrappers or autoconv * [github.com/pierrec/lz4](https://github.com/pierrec/lz4) - strong multithreaded LZ4 compression. * [github.com/cosnicolaou/pbzip2](https://github.com/cosnicolaou/pbzip2) - multithreaded bzip2 decompression. * [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. # license diff --git a/vendor/github.com/klauspost/compress/zstd/blockenc.go b/vendor/github.com/klauspost/compress/zstd/blockenc.go index 12e8f6f0b6..fd4a36f730 100644 --- a/vendor/github.com/klauspost/compress/zstd/blockenc.go +++ b/vendor/github.com/klauspost/compress/zstd/blockenc.go @@ -473,7 +473,7 @@ func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error { return b.encodeLits(b.literals, rawAllLits) } // We want some difference to at least account for the headers. - saved := b.size - len(b.literals) - (b.size >> 5) + saved := b.size - len(b.literals) - (b.size >> 6) if saved < 16 { if org == nil { return errIncompressible @@ -779,10 +779,13 @@ func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error { } b.output = wr.out + // Maybe even add a bigger margin. if len(b.output)-3-bhOffset >= b.size { - // Maybe even add a bigger margin. + // Discard and encode as raw block. + b.output = b.encodeRawTo(b.output[:bhOffset], org) + b.popOffsets() b.litEnc.Reuse = huff0.ReusePolicyNone - return errIncompressible + return nil } // Size is output minus block header. diff --git a/vendor/github.com/klauspost/compress/zstd/bytebuf.go b/vendor/github.com/klauspost/compress/zstd/bytebuf.go index 512ffe5b95..55a388553d 100644 --- a/vendor/github.com/klauspost/compress/zstd/bytebuf.go +++ b/vendor/github.com/klauspost/compress/zstd/bytebuf.go @@ -109,7 +109,7 @@ func (r *readerWrapper) readBig(n int, dst []byte) ([]byte, error) { } func (r *readerWrapper) readByte() (byte, error) { - n2, err := r.r.Read(r.tmp[:1]) + n2, err := io.ReadFull(r.r, r.tmp[:1]) if err != nil { if err == io.EOF { err = io.ErrUnexpectedEOF diff --git a/vendor/github.com/klauspost/compress/zstd/decoder.go b/vendor/github.com/klauspost/compress/zstd/decoder.go index 7113e69ee3..f04aaa21eb 100644 --- a/vendor/github.com/klauspost/compress/zstd/decoder.go +++ b/vendor/github.com/klauspost/compress/zstd/decoder.go @@ -455,12 +455,7 @@ func (d *Decoder) nextBlock(blocking bool) (ok bool) { } if len(next.b) > 0 { - n, err := d.current.crc.Write(next.b) - if err == nil { - if n != len(next.b) { - d.current.err = io.ErrShortWrite - } - } + d.current.crc.Write(next.b) } if next.err == nil && next.d != nil && next.d.hasCRC { got := uint32(d.current.crc.Sum64()) diff --git a/vendor/github.com/klauspost/compress/zstd/enc_best.go b/vendor/github.com/klauspost/compress/zstd/enc_best.go index 07f657d36e..9819d41453 100644 --- a/vendor/github.com/klauspost/compress/zstd/enc_best.go +++ b/vendor/github.com/klauspost/compress/zstd/enc_best.go @@ -34,7 +34,7 @@ type match struct { est int32 } -const highScore = 25000 +const highScore = maxMatchLen * 8 // estBits will estimate output bits from predefined tables. func (m *match) estBits(bitsPerByte int32) { @@ -159,7 +159,6 @@ func (e *bestFastEncoder) Encode(blk *blockEnc, src []byte) { // nextEmit is where in src the next emitLiteral should start from. nextEmit := s - cv := load6432(src, s) // Relative offsets offset1 := int32(blk.recentOffsets[0]) @@ -173,7 +172,6 @@ func (e *bestFastEncoder) Encode(blk *blockEnc, src []byte) { blk.literals = append(blk.literals, src[nextEmit:until]...) s.litLen = uint32(until - nextEmit) } - _ = addLiterals if debugEncoder { println("recent offsets:", blk.recentOffsets) @@ -188,7 +186,9 @@ encodeLoop: panic("offset0 was 0") } - const goodEnough = 100 + const goodEnough = 250 + + cv := load6432(src, s) nextHashL := hashLen(cv, bestLongTableBits, bestLongLen) nextHashS := hashLen(cv, bestShortTableBits, bestShortLen) @@ -201,11 +201,45 @@ encodeLoop: return } if debugAsserts { + if offset <= 0 { + panic(offset) + } 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)) } } - cand := match{offset: offset, s: s, length: 4 + e.matchlen(s+4, offset+4, src), rep: rep} + // Try to quick reject if we already have a long match. + if m.length > 16 { + left := len(src) - int(m.s+m.length) + // If we are too close to the end, keep as is. + if left <= 0 { + return + } + checkLen := m.length - (s - m.s) - 8 + if left > 2 && checkLen > 4 { + // Check 4 bytes, 4 bytes from the end of the current match. + a := load3232(src, offset+checkLen) + b := load3232(src, s+checkLen) + if a != b { + return + } + } + } + l := 4 + e.matchlen(s+4, offset+4, src) + if rep < 0 { + // Extend candidate match backwards as far as possible. + tMin := s - e.maxMatchOff + if tMin < 0 { + tMin = 0 + } + for offset > tMin && s > nextEmit && src[offset-1] == src[s-1] && l < maxMatchLength { + s-- + offset-- + l++ + } + } + + cand := match{offset: offset, s: s, length: l, rep: rep} cand.estBits(bitsPerByte) if m.est >= highScore || cand.est-m.est+(cand.s-m.s)*bitsPerByte>>10 < 0 { *m = cand @@ -219,17 +253,29 @@ encodeLoop: improve(&best, candidateS.prev-e.cur, s, uint32(cv), -1) if canRepeat && best.length < goodEnough { - cv32 := uint32(cv >> 8) - spp := s + 1 - improve(&best, spp-offset1, spp, cv32, 1) - improve(&best, spp-offset2, spp, cv32, 2) - improve(&best, spp-offset3, spp, cv32, 3) - if best.length > 0 { - cv32 = uint32(cv >> 24) - spp += 2 + if s == nextEmit { + // Check repeats straight after a match. + improve(&best, s-offset2, s, uint32(cv), 1|4) + improve(&best, s-offset3, s, uint32(cv), 2|4) + if offset1 > 1 { + improve(&best, s-(offset1-1), s, uint32(cv), 3|4) + } + } + + // If either no match or a non-repeat match, check at + 1 + if best.rep <= 0 { + cv32 := uint32(cv >> 8) + spp := s + 1 improve(&best, spp-offset1, spp, cv32, 1) improve(&best, spp-offset2, spp, cv32, 2) improve(&best, spp-offset3, spp, cv32, 3) + if best.rep < 0 { + cv32 = uint32(cv >> 24) + spp += 2 + improve(&best, spp-offset1, spp, cv32, 1) + improve(&best, spp-offset2, spp, cv32, 2) + improve(&best, spp-offset3, spp, cv32, 3) + } } } // Load next and check... @@ -244,41 +290,44 @@ encodeLoop: if s >= sLimit { break encodeLoop } - cv = load6432(src, s) continue } - s++ candidateS = e.table[hashLen(cv>>8, bestShortTableBits, bestShortLen)] - cv = load6432(src, s) - cv2 := load6432(src, s+1) + cv = load6432(src, s+1) + cv2 := load6432(src, s+2) candidateL = e.longTable[hashLen(cv, bestLongTableBits, bestLongLen)] candidateL2 := e.longTable[hashLen(cv2, bestLongTableBits, bestLongLen)] // Short at s+1 - improve(&best, candidateS.offset-e.cur, s, uint32(cv), -1) + improve(&best, candidateS.offset-e.cur, s+1, uint32(cv), -1) // Long at s+1, s+2 - improve(&best, candidateL.offset-e.cur, s, uint32(cv), -1) - improve(&best, candidateL.prev-e.cur, s, uint32(cv), -1) - improve(&best, candidateL2.offset-e.cur, s+1, uint32(cv2), -1) - improve(&best, candidateL2.prev-e.cur, s+1, uint32(cv2), -1) + improve(&best, candidateL.offset-e.cur, s+1, uint32(cv), -1) + improve(&best, candidateL.prev-e.cur, s+1, uint32(cv), -1) + improve(&best, candidateL2.offset-e.cur, s+2, uint32(cv2), -1) + improve(&best, candidateL2.prev-e.cur, s+2, uint32(cv2), -1) if false { // Short at s+3. // Too often worse... - improve(&best, e.table[hashLen(cv2>>8, bestShortTableBits, bestShortLen)].offset-e.cur, s+2, uint32(cv2>>8), -1) + improve(&best, e.table[hashLen(cv2>>8, bestShortTableBits, bestShortLen)].offset-e.cur, s+3, uint32(cv2>>8), -1) } - // See if we can find a better match by checking where the current best ends. - // Use that offset to see if we can find a better full match. - if sAt := best.s + best.length; sAt < sLimit { - nextHashL := hashLen(load6432(src, sAt), bestLongTableBits, bestLongLen) - candidateEnd := e.longTable[nextHashL] - // Start check at a fixed offset to allow for a few mismatches. - // For this compression level 2 yields the best results. - const skipBeginning = 2 - if pos := candidateEnd.offset - e.cur - best.length + skipBeginning; pos >= 0 { - improve(&best, pos, best.s+skipBeginning, load3232(src, best.s+skipBeginning), -1) - if pos := candidateEnd.prev - e.cur - best.length + skipBeginning; pos >= 0 { - improve(&best, pos, best.s+skipBeginning, load3232(src, best.s+skipBeginning), -1) + + // Start check at a fixed offset to allow for a few mismatches. + // For this compression level 2 yields the best results. + // We cannot do this if we have already indexed this position. + const skipBeginning = 2 + if best.s > s-skipBeginning { + // See if we can find a better match by checking where the current best ends. + // Use that offset to see if we can find a better full match. + if sAt := best.s + best.length; sAt < sLimit { + nextHashL := hashLen(load6432(src, sAt), bestLongTableBits, bestLongLen) + candidateEnd := e.longTable[nextHashL] + + if off := candidateEnd.offset - e.cur - best.length + skipBeginning; off >= 0 { + improve(&best, off, best.s+skipBeginning, load3232(src, best.s+skipBeginning), -1) + if off := candidateEnd.prev - e.cur - best.length + skipBeginning; off >= 0 { + improve(&best, off, best.s+skipBeginning, load3232(src, best.s+skipBeginning), -1) + } } } } @@ -292,51 +341,34 @@ encodeLoop: // We have a match, we can store the forward value if best.rep > 0 { - s = best.s var seq seq seq.matchLen = uint32(best.length - zstdMinMatch) - - // We might be able to match backwards. - // Extend as long as we can. - start := best.s - // We end the search early, so we don't risk 0 literals - // and have to do special offset treatment. - startLimit := nextEmit + 1 - - tMin := s - e.maxMatchOff - if tMin < 0 { - tMin = 0 + if debugAsserts && s <= nextEmit { + panic("s <= nextEmit") } - repIndex := best.offset - for repIndex > tMin && start > startLimit && src[repIndex-1] == src[start-1] && seq.matchLen < maxMatchLength-zstdMinMatch-1 { - repIndex-- - start-- - seq.matchLen++ - } - addLiterals(&seq, start) + addLiterals(&seq, best.s) - // rep 0 - seq.offset = uint32(best.rep) + // Repeat. If bit 4 is set, this is a non-lit repeat. + seq.offset = uint32(best.rep & 3) if debugSequences { println("repeat sequence", seq, "next s:", s) } blk.sequences = append(blk.sequences, seq) - // Index match start+1 (long) -> s - 1 - index0 := s + // Index old s + 1 -> s - 1 + index0 := s + 1 s = best.s + best.length nextEmit = s if s >= sLimit { if debugEncoder { println("repeat ended", s, best.length) - } break encodeLoop } // Index skipped... off := index0 + e.cur - for index0 < s-1 { + for index0 < s { cv0 := load6432(src, index0) h0 := hashLen(cv0, bestLongTableBits, bestLongLen) h1 := hashLen(cv0, bestShortTableBits, bestShortLen) @@ -346,17 +378,19 @@ encodeLoop: index0++ } switch best.rep { - case 2: + case 2, 4 | 1: offset1, offset2 = offset2, offset1 - case 3: + case 3, 4 | 2: offset1, offset2, offset3 = offset3, offset1, offset2 + case 4 | 3: + offset1, offset2, offset3 = offset1-1, offset1, offset2 } - cv = load6432(src, s) continue } // A 4-byte match has been found. Update recent offsets. // We'll later see if more than 4 bytes. + index0 := s + 1 s = best.s t := best.offset offset1, offset2, offset3 = s-t, offset1, offset2 @@ -369,22 +403,9 @@ encodeLoop: panic("invalid offset") } - // Extend the n-byte match as long as possible. - l := best.length - - // Extend backwards - tMin := s - e.maxMatchOff - if tMin < 0 { - tMin = 0 - } - for t > tMin && s > nextEmit && src[t-1] == src[s-1] && l < maxMatchLength { - s-- - t-- - l++ - } - // Write our sequence var seq seq + l := best.length seq.litLen = uint32(s - nextEmit) seq.matchLen = uint32(l - zstdMinMatch) if seq.litLen > 0 { @@ -401,10 +422,8 @@ encodeLoop: break encodeLoop } - // Index match start+1 (long) -> s - 1 - index0 := s - l + 1 - // every entry - for index0 < s-1 { + // Index old s + 1 -> s - 1 + for index0 < s { cv0 := load6432(src, index0) h0 := hashLen(cv0, bestLongTableBits, bestLongLen) h1 := hashLen(cv0, bestShortTableBits, bestShortLen) @@ -413,50 +432,6 @@ encodeLoop: e.table[h1] = prevEntry{offset: off, prev: e.table[h1].offset} index0++ } - - cv = load6432(src, s) - if !canRepeat { - continue - } - - // Check offset 2 - for { - o2 := s - offset2 - if load3232(src, o2) != uint32(cv) { - // Do regular search - break - } - - // Store this, since we have it. - nextHashS := hashLen(cv, bestShortTableBits, bestShortLen) - nextHashL := hashLen(cv, bestLongTableBits, bestLongLen) - - // We have at least 4 byte match. - // No need to check backwards. We come straight from a match - l := 4 + e.matchlen(s+4, o2+4, src) - - e.longTable[nextHashL] = prevEntry{offset: s + e.cur, prev: e.longTable[nextHashL].offset} - e.table[nextHashS] = prevEntry{offset: s + e.cur, prev: e.table[nextHashS].offset} - seq.matchLen = uint32(l) - zstdMinMatch - seq.litLen = 0 - - // Since litlen is always 0, this is offset 1. - seq.offset = 1 - s += l - nextEmit = s - if debugSequences { - println("sequence", seq, "next s:", s) - } - blk.sequences = append(blk.sequences, seq) - - // Swap offset 1 and 2. - offset1, offset2 = offset2, offset1 - if s >= sLimit { - // Finished - break encodeLoop - } - cv = load6432(src, s) - } } if int(nextEmit) < len(src) { diff --git a/vendor/github.com/klauspost/compress/zstd/encoder.go b/vendor/github.com/klauspost/compress/zstd/encoder.go index 65c6c36dc1..4de0aed0d0 100644 --- a/vendor/github.com/klauspost/compress/zstd/encoder.go +++ b/vendor/github.com/klauspost/compress/zstd/encoder.go @@ -277,23 +277,9 @@ func (e *Encoder) nextBlock(final bool) error { s.eofWritten = true } - err := errIncompressible - // If we got the exact same number of literals as input, - // assume the literals cannot be compressed. - if len(src) != len(blk.literals) || len(src) != e.o.blockSize { - err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) - } - switch err { - case errIncompressible: - if debugEncoder { - println("Storing incompressible block as raw") - } - blk.encodeRaw(src) - // In fast mode, we do not transfer offsets, so we don't have to deal with changing the. - case nil: - default: - s.err = err - return err + s.err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) + if s.err != nil { + return s.err } _, s.err = s.w.Write(blk.output) s.nWritten += int64(len(blk.output)) @@ -343,22 +329,8 @@ func (e *Encoder) nextBlock(final bool) error { } s.wWg.Done() }() - err := errIncompressible - // If we got the exact same number of literals as input, - // assume the literals cannot be compressed. - if len(src) != len(blk.literals) || len(src) != e.o.blockSize { - err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) - } - switch err { - case errIncompressible: - if debugEncoder { - println("Storing incompressible block as raw") - } - blk.encodeRaw(src) - // In fast mode, we do not transfer offsets, so we don't have to deal with changing the. - case nil: - default: - s.writeErr = err + s.writeErr = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) + if s.writeErr != nil { return } _, s.writeErr = s.w.Write(blk.output) @@ -568,25 +540,15 @@ func (e *Encoder) EncodeAll(src, dst []byte) []byte { // If we got the exact same number of literals as input, // assume the literals cannot be compressed. - err := errIncompressible oldout := blk.output - if len(blk.literals) != len(src) || len(src) != e.o.blockSize { - // Output directly to dst - blk.output = dst - err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) - } + // Output directly to dst + blk.output = dst - switch err { - case errIncompressible: - if debugEncoder { - println("Storing incompressible block as raw") - } - dst = blk.encodeRawTo(dst, src) - case nil: - dst = blk.output - default: + err := blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy) + if err != nil { panic(err) } + dst = blk.output blk.output = oldout } else { enc.Reset(e.o.dict, false) @@ -605,25 +567,11 @@ func (e *Encoder) EncodeAll(src, dst []byte) []byte { if len(src) == 0 { blk.last = true } - err := errIncompressible - // If we got the exact same number of literals as input, - // assume the literals cannot be compressed. - if len(blk.literals) != len(todo) || len(todo) != e.o.blockSize { - err = blk.encode(todo, e.o.noEntropy, !e.o.allLitEntropy) - } - - switch err { - case errIncompressible: - if debugEncoder { - println("Storing incompressible block as raw") - } - dst = blk.encodeRawTo(dst, todo) - blk.popOffsets() - case nil: - dst = append(dst, blk.output...) - default: + err := blk.encode(todo, e.o.noEntropy, !e.o.allLitEntropy) + if err != nil { panic(err) } + dst = append(dst, blk.output...) blk.reset(nil) } } diff --git a/vendor/github.com/klauspost/compress/zstd/encoder_options.go b/vendor/github.com/klauspost/compress/zstd/encoder_options.go index 8e15be2f7f..50f70533b4 100644 --- a/vendor/github.com/klauspost/compress/zstd/encoder_options.go +++ b/vendor/github.com/klauspost/compress/zstd/encoder_options.go @@ -39,7 +39,7 @@ func (o *encoderOptions) setDefault() { blockSize: maxCompressedBlockSize, windowSize: 8 << 20, level: SpeedDefault, - allLitEntropy: true, + allLitEntropy: false, lowMem: false, } } @@ -238,7 +238,7 @@ func WithEncoderLevel(l EncoderLevel) EOption { } } if !o.customALEntropy { - o.allLitEntropy = l > SpeedFastest + o.allLitEntropy = l > SpeedDefault } return nil diff --git a/vendor/github.com/klauspost/compress/zstd/framedec.go b/vendor/github.com/klauspost/compress/zstd/framedec.go index d8e8a05bd7..cc0aa22745 100644 --- a/vendor/github.com/klauspost/compress/zstd/framedec.go +++ b/vendor/github.com/klauspost/compress/zstd/framedec.go @@ -293,13 +293,9 @@ func (d *frameDec) next(block *blockDec) error { return nil } -// checkCRC will check the checksum if the frame has one. +// checkCRC will check the checksum, assuming the frame has one. // Will return ErrCRCMismatch if crc check failed, otherwise nil. func (d *frameDec) checkCRC() error { - if !d.HasCheckSum { - return nil - } - // We can overwrite upper tmp now buf, err := d.rawInput.readSmall(4) if err != nil { @@ -307,10 +303,6 @@ func (d *frameDec) checkCRC() error { return err } - if d.o.ignoreChecksum { - return nil - } - want := binary.LittleEndian.Uint32(buf[:4]) got := uint32(d.crc.Sum64()) @@ -326,17 +318,13 @@ func (d *frameDec) checkCRC() error { return nil } -// consumeCRC reads the checksum data if the frame has one. +// consumeCRC skips over the checksum, assuming the frame has one. func (d *frameDec) consumeCRC() error { - if d.HasCheckSum { - _, err := d.rawInput.readSmall(4) - if err != nil { - println("CRC missing?", err) - return err - } + _, err := d.rawInput.readSmall(4) + if err != nil { + println("CRC missing?", err) } - - return nil + return err } // runDecoder will run the decoder for the remainder of the frame. @@ -415,15 +403,8 @@ func (d *frameDec) runDecoder(dst []byte, dec *blockDec) ([]byte, error) { if d.o.ignoreChecksum { err = d.consumeCRC() } else { - var n int - n, err = d.crc.Write(dst[crcStart:]) - if err == nil { - if n != len(dst)-crcStart { - err = io.ErrShortWrite - } else { - err = d.checkCRC() - } - } + d.crc.Write(dst[crcStart:]) + err = d.checkCRC() } } } diff --git a/vendor/github.com/klauspost/compress/zstd/seqdec.go b/vendor/github.com/klauspost/compress/zstd/seqdec.go index 27fdf90fbc..9405fcf101 100644 --- a/vendor/github.com/klauspost/compress/zstd/seqdec.go +++ b/vendor/github.com/klauspost/compress/zstd/seqdec.go @@ -236,9 +236,12 @@ func (s *sequenceDecs) decodeSync(hist []byte) error { maxBlockSize = s.windowSize } + if debugDecoder { + println("decodeSync: decoding", seqs, "sequences", br.remain(), "bits remain on stream") + } for i := seqs - 1; i >= 0; i-- { if br.overread() { - printf("reading sequence %d, exceeded available data\n", seqs-i) + printf("reading sequence %d, exceeded available data. Overread by %d\n", seqs-i, -br.remain()) return io.ErrUnexpectedEOF } var ll, mo, ml int diff --git a/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.go b/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.go index 387a30e99d..8adabd8287 100644 --- a/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.go +++ b/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.go @@ -5,6 +5,7 @@ package zstd import ( "fmt" + "io" "github.com/klauspost/compress/internal/cpuinfo" ) @@ -134,6 +135,9 @@ func (s *sequenceDecs) decodeSyncSimple(hist []byte) (bool, error) { return true, fmt.Errorf("unexpected literal count, want %d bytes, but only %d is available", ctx.ll, ctx.litRemain+ctx.ll) + case errorOverread: + return true, io.ErrUnexpectedEOF + case errorNotEnoughSpace: size := ctx.outPosition + ctx.ll + ctx.ml if debugDecoder { @@ -202,6 +206,9 @@ const errorNotEnoughLiterals = 4 // error reported when capacity of `out` is too small const errorNotEnoughSpace = 5 +// error reported when bits are overread. +const errorOverread = 6 + // sequenceDecs_decode implements the main loop of sequenceDecs in x86 asm. // // Please refer to seqdec_generic.go for the reference implementation. @@ -247,6 +254,10 @@ func (s *sequenceDecs) decode(seqs []seqVals) error { litRemain: len(s.literals), } + if debugDecoder { + println("decode: decoding", len(seqs), "sequences", br.remain(), "bits remain on stream") + } + s.seqSize = 0 lte56bits := s.maxBits+s.offsets.fse.actualTableLog+s.matchLengths.fse.actualTableLog+s.litLengths.fse.actualTableLog <= 56 var errCode int @@ -277,6 +288,8 @@ func (s *sequenceDecs) decode(seqs []seqVals) error { case errorNotEnoughLiterals: ll := ctx.seqs[i].ll return fmt.Errorf("unexpected literal count, want %d bytes, but only %d is available", ll, ctx.litRemain+ll) + case errorOverread: + return io.ErrUnexpectedEOF } return fmt.Errorf("sequenceDecs_decode_amd64 returned erronous code %d", errCode) @@ -291,6 +304,9 @@ func (s *sequenceDecs) decode(seqs []seqVals) error { if s.seqSize > maxBlockSize { return fmt.Errorf("output bigger than max block size (%d)", maxBlockSize) } + if debugDecoder { + println("decode: ", br.remain(), "bits remain on stream. code:", errCode) + } err := br.close() if err != nil { printf("Closing sequences: %v, %+v\n", err, *br) diff --git a/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.s b/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.s index b94993a072..b6f4ba6fc5 100644 --- a/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.s +++ b/vendor/github.com/klauspost/compress/zstd/seqdec_amd64.s @@ -38,7 +38,7 @@ sequenceDecs_decode_amd64_main_loop: sequenceDecs_decode_amd64_fill_byte_by_byte: CMPQ SI, $0x00 - JLE sequenceDecs_decode_amd64_fill_end + JLE sequenceDecs_decode_amd64_fill_check_overread CMPQ BX, $0x07 JLE sequenceDecs_decode_amd64_fill_end SHLQ $0x08, DX @@ -49,6 +49,10 @@ sequenceDecs_decode_amd64_fill_byte_by_byte: ORQ AX, DX JMP sequenceDecs_decode_amd64_fill_byte_by_byte +sequenceDecs_decode_amd64_fill_check_overread: + CMPQ BX, $0x40 + JA error_overread + sequenceDecs_decode_amd64_fill_end: // Update offset MOVQ R9, AX @@ -105,7 +109,7 @@ sequenceDecs_decode_amd64_ml_update_zero: sequenceDecs_decode_amd64_fill_2_byte_by_byte: CMPQ SI, $0x00 - JLE sequenceDecs_decode_amd64_fill_2_end + JLE sequenceDecs_decode_amd64_fill_2_check_overread CMPQ BX, $0x07 JLE sequenceDecs_decode_amd64_fill_2_end SHLQ $0x08, DX @@ -116,6 +120,10 @@ sequenceDecs_decode_amd64_fill_2_byte_by_byte: ORQ AX, DX JMP sequenceDecs_decode_amd64_fill_2_byte_by_byte +sequenceDecs_decode_amd64_fill_2_check_overread: + CMPQ BX, $0x40 + JA error_overread + sequenceDecs_decode_amd64_fill_2_end: // Update literal length MOVQ DI, AX @@ -320,6 +328,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // func sequenceDecs_decode_56_amd64(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int // Requires: CMOV TEXT ·sequenceDecs_decode_56_amd64(SB), $8-32 @@ -356,7 +369,7 @@ sequenceDecs_decode_56_amd64_main_loop: sequenceDecs_decode_56_amd64_fill_byte_by_byte: CMPQ SI, $0x00 - JLE sequenceDecs_decode_56_amd64_fill_end + JLE sequenceDecs_decode_56_amd64_fill_check_overread CMPQ BX, $0x07 JLE sequenceDecs_decode_56_amd64_fill_end SHLQ $0x08, DX @@ -367,6 +380,10 @@ sequenceDecs_decode_56_amd64_fill_byte_by_byte: ORQ AX, DX JMP sequenceDecs_decode_56_amd64_fill_byte_by_byte +sequenceDecs_decode_56_amd64_fill_check_overread: + CMPQ BX, $0x40 + JA error_overread + sequenceDecs_decode_56_amd64_fill_end: // Update offset MOVQ R9, AX @@ -613,6 +630,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // func sequenceDecs_decode_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int // Requires: BMI, BMI2, CMOV TEXT ·sequenceDecs_decode_bmi2(SB), $8-32 @@ -649,7 +671,7 @@ sequenceDecs_decode_bmi2_main_loop: sequenceDecs_decode_bmi2_fill_byte_by_byte: CMPQ BX, $0x00 - JLE sequenceDecs_decode_bmi2_fill_end + JLE sequenceDecs_decode_bmi2_fill_check_overread CMPQ DX, $0x07 JLE sequenceDecs_decode_bmi2_fill_end SHLQ $0x08, AX @@ -660,6 +682,10 @@ sequenceDecs_decode_bmi2_fill_byte_by_byte: ORQ CX, AX JMP sequenceDecs_decode_bmi2_fill_byte_by_byte +sequenceDecs_decode_bmi2_fill_check_overread: + CMPQ DX, $0x40 + JA error_overread + sequenceDecs_decode_bmi2_fill_end: // Update offset MOVQ $0x00000808, CX @@ -700,7 +726,7 @@ sequenceDecs_decode_bmi2_fill_end: sequenceDecs_decode_bmi2_fill_2_byte_by_byte: CMPQ BX, $0x00 - JLE sequenceDecs_decode_bmi2_fill_2_end + JLE sequenceDecs_decode_bmi2_fill_2_check_overread CMPQ DX, $0x07 JLE sequenceDecs_decode_bmi2_fill_2_end SHLQ $0x08, AX @@ -711,6 +737,10 @@ sequenceDecs_decode_bmi2_fill_2_byte_by_byte: ORQ CX, AX JMP sequenceDecs_decode_bmi2_fill_2_byte_by_byte +sequenceDecs_decode_bmi2_fill_2_check_overread: + CMPQ DX, $0x40 + JA error_overread + sequenceDecs_decode_bmi2_fill_2_end: // Update literal length MOVQ $0x00000808, CX @@ -889,6 +919,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // func sequenceDecs_decode_56_bmi2(s *sequenceDecs, br *bitReader, ctx *decodeAsmContext) int // Requires: BMI, BMI2, CMOV TEXT ·sequenceDecs_decode_56_bmi2(SB), $8-32 @@ -925,7 +960,7 @@ sequenceDecs_decode_56_bmi2_main_loop: sequenceDecs_decode_56_bmi2_fill_byte_by_byte: CMPQ BX, $0x00 - JLE sequenceDecs_decode_56_bmi2_fill_end + JLE sequenceDecs_decode_56_bmi2_fill_check_overread CMPQ DX, $0x07 JLE sequenceDecs_decode_56_bmi2_fill_end SHLQ $0x08, AX @@ -936,6 +971,10 @@ sequenceDecs_decode_56_bmi2_fill_byte_by_byte: ORQ CX, AX JMP sequenceDecs_decode_56_bmi2_fill_byte_by_byte +sequenceDecs_decode_56_bmi2_fill_check_overread: + CMPQ DX, $0x40 + JA error_overread + sequenceDecs_decode_56_bmi2_fill_end: // Update offset MOVQ $0x00000808, CX @@ -1140,6 +1179,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // func sequenceDecs_executeSimple_amd64(ctx *executeAsmContext) bool // Requires: SSE TEXT ·sequenceDecs_executeSimple_amd64(SB), $8-9 @@ -1804,7 +1848,7 @@ sequenceDecs_decodeSync_amd64_main_loop: sequenceDecs_decodeSync_amd64_fill_byte_by_byte: CMPQ SI, $0x00 - JLE sequenceDecs_decodeSync_amd64_fill_end + JLE sequenceDecs_decodeSync_amd64_fill_check_overread CMPQ BX, $0x07 JLE sequenceDecs_decodeSync_amd64_fill_end SHLQ $0x08, DX @@ -1815,6 +1859,10 @@ sequenceDecs_decodeSync_amd64_fill_byte_by_byte: ORQ AX, DX JMP sequenceDecs_decodeSync_amd64_fill_byte_by_byte +sequenceDecs_decodeSync_amd64_fill_check_overread: + CMPQ BX, $0x40 + JA error_overread + sequenceDecs_decodeSync_amd64_fill_end: // Update offset MOVQ R9, AX @@ -1871,7 +1919,7 @@ sequenceDecs_decodeSync_amd64_ml_update_zero: sequenceDecs_decodeSync_amd64_fill_2_byte_by_byte: CMPQ SI, $0x00 - JLE sequenceDecs_decodeSync_amd64_fill_2_end + JLE sequenceDecs_decodeSync_amd64_fill_2_check_overread CMPQ BX, $0x07 JLE sequenceDecs_decodeSync_amd64_fill_2_end SHLQ $0x08, DX @@ -1882,6 +1930,10 @@ sequenceDecs_decodeSync_amd64_fill_2_byte_by_byte: ORQ AX, DX JMP sequenceDecs_decodeSync_amd64_fill_2_byte_by_byte +sequenceDecs_decodeSync_amd64_fill_2_check_overread: + CMPQ BX, $0x40 + JA error_overread + sequenceDecs_decodeSync_amd64_fill_2_end: // Update literal length MOVQ DI, AX @@ -2291,6 +2343,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // Return with not enough output space error error_not_enough_space: MOVQ ctx+16(FP), AX @@ -2356,7 +2413,7 @@ sequenceDecs_decodeSync_bmi2_main_loop: sequenceDecs_decodeSync_bmi2_fill_byte_by_byte: CMPQ BX, $0x00 - JLE sequenceDecs_decodeSync_bmi2_fill_end + JLE sequenceDecs_decodeSync_bmi2_fill_check_overread CMPQ DX, $0x07 JLE sequenceDecs_decodeSync_bmi2_fill_end SHLQ $0x08, AX @@ -2367,6 +2424,10 @@ sequenceDecs_decodeSync_bmi2_fill_byte_by_byte: ORQ CX, AX JMP sequenceDecs_decodeSync_bmi2_fill_byte_by_byte +sequenceDecs_decodeSync_bmi2_fill_check_overread: + CMPQ DX, $0x40 + JA error_overread + sequenceDecs_decodeSync_bmi2_fill_end: // Update offset MOVQ $0x00000808, CX @@ -2407,7 +2468,7 @@ sequenceDecs_decodeSync_bmi2_fill_end: sequenceDecs_decodeSync_bmi2_fill_2_byte_by_byte: CMPQ BX, $0x00 - JLE sequenceDecs_decodeSync_bmi2_fill_2_end + JLE sequenceDecs_decodeSync_bmi2_fill_2_check_overread CMPQ DX, $0x07 JLE sequenceDecs_decodeSync_bmi2_fill_2_end SHLQ $0x08, AX @@ -2418,6 +2479,10 @@ sequenceDecs_decodeSync_bmi2_fill_2_byte_by_byte: ORQ CX, AX JMP sequenceDecs_decodeSync_bmi2_fill_2_byte_by_byte +sequenceDecs_decodeSync_bmi2_fill_2_check_overread: + CMPQ DX, $0x40 + JA error_overread + sequenceDecs_decodeSync_bmi2_fill_2_end: // Update literal length MOVQ $0x00000808, CX @@ -2801,6 +2866,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // Return with not enough output space error error_not_enough_space: MOVQ ctx+16(FP), AX @@ -2866,7 +2936,7 @@ sequenceDecs_decodeSync_safe_amd64_main_loop: sequenceDecs_decodeSync_safe_amd64_fill_byte_by_byte: CMPQ SI, $0x00 - JLE sequenceDecs_decodeSync_safe_amd64_fill_end + JLE sequenceDecs_decodeSync_safe_amd64_fill_check_overread CMPQ BX, $0x07 JLE sequenceDecs_decodeSync_safe_amd64_fill_end SHLQ $0x08, DX @@ -2877,6 +2947,10 @@ sequenceDecs_decodeSync_safe_amd64_fill_byte_by_byte: ORQ AX, DX JMP sequenceDecs_decodeSync_safe_amd64_fill_byte_by_byte +sequenceDecs_decodeSync_safe_amd64_fill_check_overread: + CMPQ BX, $0x40 + JA error_overread + sequenceDecs_decodeSync_safe_amd64_fill_end: // Update offset MOVQ R9, AX @@ -2933,7 +3007,7 @@ sequenceDecs_decodeSync_safe_amd64_ml_update_zero: sequenceDecs_decodeSync_safe_amd64_fill_2_byte_by_byte: CMPQ SI, $0x00 - JLE sequenceDecs_decodeSync_safe_amd64_fill_2_end + JLE sequenceDecs_decodeSync_safe_amd64_fill_2_check_overread CMPQ BX, $0x07 JLE sequenceDecs_decodeSync_safe_amd64_fill_2_end SHLQ $0x08, DX @@ -2944,6 +3018,10 @@ sequenceDecs_decodeSync_safe_amd64_fill_2_byte_by_byte: ORQ AX, DX JMP sequenceDecs_decodeSync_safe_amd64_fill_2_byte_by_byte +sequenceDecs_decodeSync_safe_amd64_fill_2_check_overread: + CMPQ BX, $0x40 + JA error_overread + sequenceDecs_decodeSync_safe_amd64_fill_2_end: // Update literal length MOVQ DI, AX @@ -3455,6 +3533,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // Return with not enough output space error error_not_enough_space: MOVQ ctx+16(FP), AX @@ -3520,7 +3603,7 @@ sequenceDecs_decodeSync_safe_bmi2_main_loop: sequenceDecs_decodeSync_safe_bmi2_fill_byte_by_byte: CMPQ BX, $0x00 - JLE sequenceDecs_decodeSync_safe_bmi2_fill_end + JLE sequenceDecs_decodeSync_safe_bmi2_fill_check_overread CMPQ DX, $0x07 JLE sequenceDecs_decodeSync_safe_bmi2_fill_end SHLQ $0x08, AX @@ -3531,6 +3614,10 @@ sequenceDecs_decodeSync_safe_bmi2_fill_byte_by_byte: ORQ CX, AX JMP sequenceDecs_decodeSync_safe_bmi2_fill_byte_by_byte +sequenceDecs_decodeSync_safe_bmi2_fill_check_overread: + CMPQ DX, $0x40 + JA error_overread + sequenceDecs_decodeSync_safe_bmi2_fill_end: // Update offset MOVQ $0x00000808, CX @@ -3571,7 +3658,7 @@ sequenceDecs_decodeSync_safe_bmi2_fill_end: sequenceDecs_decodeSync_safe_bmi2_fill_2_byte_by_byte: CMPQ BX, $0x00 - JLE sequenceDecs_decodeSync_safe_bmi2_fill_2_end + JLE sequenceDecs_decodeSync_safe_bmi2_fill_2_check_overread CMPQ DX, $0x07 JLE sequenceDecs_decodeSync_safe_bmi2_fill_2_end SHLQ $0x08, AX @@ -3582,6 +3669,10 @@ sequenceDecs_decodeSync_safe_bmi2_fill_2_byte_by_byte: ORQ CX, AX JMP sequenceDecs_decodeSync_safe_bmi2_fill_2_byte_by_byte +sequenceDecs_decodeSync_safe_bmi2_fill_2_check_overread: + CMPQ DX, $0x40 + JA error_overread + sequenceDecs_decodeSync_safe_bmi2_fill_2_end: // Update literal length MOVQ $0x00000808, CX @@ -4067,6 +4158,11 @@ error_not_enough_literals: MOVQ $0x00000004, ret+24(FP) RET + // Return with overread error +error_overread: + MOVQ $0x00000006, ret+24(FP) + RET + // Return with not enough output space error error_not_enough_space: MOVQ ctx+16(FP), AX diff --git a/vendor/github.com/klauspost/compress/zstd/zstd.go b/vendor/github.com/klauspost/compress/zstd/zstd.go index 5ffa82f5ac..89396673d9 100644 --- a/vendor/github.com/klauspost/compress/zstd/zstd.go +++ b/vendor/github.com/klauspost/compress/zstd/zstd.go @@ -128,11 +128,11 @@ func matchLen(a, b []byte) (n int) { } func load3232(b []byte, i int32) uint32 { - return binary.LittleEndian.Uint32(b[i:]) + return binary.LittleEndian.Uint32(b[:len(b):len(b)][i:]) } func load6432(b []byte, i int32) uint64 { - return binary.LittleEndian.Uint64(b[i:]) + return binary.LittleEndian.Uint64(b[:len(b):len(b)][i:]) } type byter interface { diff --git a/vendor/modules.txt b/vendor/modules.txt index 8b2c48fb36..39c1a262f6 100644 --- a/vendor/modules.txt +++ b/vendor/modules.txt @@ -138,7 +138,7 @@ github.com/imdario/mergo # github.com/inconshreveable/mousetrap v1.1.0 ## explicit; go 1.18 github.com/inconshreveable/mousetrap -# github.com/klauspost/compress v1.16.3 +# github.com/klauspost/compress v1.16.5 ## explicit; go 1.18 github.com/klauspost/compress github.com/klauspost/compress/fse