aleesa-bot/aleesa-irc-go
0
Fork 0
Code Activity

Update dependencies

Browse Source
This commit is contained in:
eleksir
2025-10-12 13:55:23 +03:00
parent 897a2035d9
commit 3ad60b9c69
437 changed files with 26255 additions and 43246 deletions
Download Patch File Download Diff File Expand all files Collapse all files

28
vendor/github.com/klauspost/compress/.goreleaser.yml generated vendored
View File

@@ -1,8 +1,9 @@
version: 2
# This is an example goreleaser.yaml file with some sane defaults.
# Make sure to check the documentation at http://goreleaser.com
before:
hooks:
- ./gen.sh
- go install mvdan.cc/garble@v0.9.3
builds:
-
@@ -31,6 +32,7 @@ builds:
- mips64le
goarm:
- 7
gobinary: garble
-
id: "s2d"
binary: s2d
@@ -57,6 +59,7 @@ builds:
- mips64le
goarm:
- 7
gobinary: garble
-
id: "s2sx"
binary: s2sx
@@ -84,11 +87,21 @@ builds:
- mips64le
goarm:
- 7
gobinary: garble
archives:
-
id: s2-binaries
name_template: "s2-{{ .Os }}_{{ .Arch }}{{ if .Arm }}v{{ .Arm }}{{ end }}"
name_template: "s2-{{ .Os }}_{{ .Arch }}_{{ .Version }}"
replacements:
aix: AIX
darwin: OSX
linux: Linux
windows: Windows
386: i386
amd64: x86_64
freebsd: FreeBSD
netbsd: NetBSD
format_overrides:
- goos: windows
format: zip
@@ -99,7 +112,7 @@ archives:
checksum:
name_template: 'checksums.txt'
snapshot:
version_template: "{{ .Tag }}-next"
name_template: "{{ .Tag }}-next"
changelog:
sort: asc
filters:
@@ -112,7 +125,7 @@ changelog:
nfpms:
-
file_name_template: "s2_package__{{ .Os }}_{{ .Arch }}{{ if .Arm }}v{{ .Arm }}{{ end }}"
file_name_template: "s2_package_{{ .Version }}_{{ .Os }}_{{ .Arch }}"
vendor: Klaus Post
homepage: https://github.com/klauspost/compress
maintainer: Klaus Post <klauspost@gmail.com>
@@ -121,3 +134,8 @@ nfpms:
formats:
- deb
- rpm
replacements:
darwin: Darwin
linux: Linux
freebsd: FreeBSD
amd64: x86_64

259
vendor/github.com/klauspost/compress/README.md generated vendored
View File

@@ -14,143 +14,10 @@ This package provides various compression algorithms.
[![Go](https://github.com/klauspost/compress/actions/workflows/go.yml/badge.svg)](https://github.com/klauspost/compress/actions/workflows/go.yml)
[![Sourcegraph Badge](https://sourcegraph.com/github.com/klauspost/compress/-/badge.svg)](https://sourcegraph.com/github.com/klauspost/compress?badge)
# package usage
Use `go get github.com/klauspost/compress@latest` to add it to your project.
This package will support the current Go version and 2 versions back.
* Use the `nounsafe` tag to disable all use of the "unsafe" package.
* Use the `noasm` tag to disable all assembly across packages.
Use the links above for more information on each.
# changelog
* Feb 19th, 2025 - [1.18.0](https://github.com/klauspost/compress/releases/tag/v1.18.0)
* Add unsafe little endian loaders https://github.com/klauspost/compress/pull/1036
* fix: check `r.err != nil` but return a nil value error `err` by @alingse in https://github.com/klauspost/compress/pull/1028
* flate: Simplify L4-6 loading https://github.com/klauspost/compress/pull/1043
* flate: Simplify matchlen (remove asm) https://github.com/klauspost/compress/pull/1045
* s2: Improve small block compression speed w/o asm https://github.com/klauspost/compress/pull/1048
* flate: Fix matchlen L5+L6 https://github.com/klauspost/compress/pull/1049
* flate: Cleanup & reduce casts https://github.com/klauspost/compress/pull/1050
* Oct 11th, 2024 - [1.17.11](https://github.com/klauspost/compress/releases/tag/v1.17.11)
* zstd: Fix extra CRC written with multiple Close calls https://github.com/klauspost/compress/pull/1017
* s2: Don't use stack for index tables https://github.com/klauspost/compress/pull/1014
* gzhttp: No content-type on no body response code by @juliens in https://github.com/klauspost/compress/pull/1011
* gzhttp: Do not set the content-type when response has no body by @kevinpollet in https://github.com/klauspost/compress/pull/1013
* Sep 23rd, 2024 - [1.17.10](https://github.com/klauspost/compress/releases/tag/v1.17.10)
* gzhttp: Add TransportAlwaysDecompress option. https://github.com/klauspost/compress/pull/978
* gzhttp: Add supported decompress request body by @mirecl in https://github.com/klauspost/compress/pull/1002
* s2: Add EncodeBuffer buffer recycling callback https://github.com/klauspost/compress/pull/982
* zstd: Improve memory usage on small streaming encodes https://github.com/klauspost/compress/pull/1007
* flate: read data written with partial flush by @vajexal in https://github.com/klauspost/compress/pull/996
* Jun 12th, 2024 - [1.17.9](https://github.com/klauspost/compress/releases/tag/v1.17.9)
* s2: Reduce ReadFrom temporary allocations https://github.com/klauspost/compress/pull/949
* flate, zstd: Shave some bytes off amd64 matchLen by @greatroar in https://github.com/klauspost/compress/pull/963
* Upgrade zip/zlib to 1.22.4 upstream https://github.com/klauspost/compress/pull/970 https://github.com/klauspost/compress/pull/971
* zstd: BuildDict fails with RLE table https://github.com/klauspost/compress/pull/951
* Apr 9th, 2024 - [1.17.8](https://github.com/klauspost/compress/releases/tag/v1.17.8)
* zstd: Reject blocks where reserved values are not 0 https://github.com/klauspost/compress/pull/885
* zstd: Add RLE detection+encoding https://github.com/klauspost/compress/pull/938
* Feb 21st, 2024 - [1.17.7](https://github.com/klauspost/compress/releases/tag/v1.17.7)
* s2: Add AsyncFlush method: Complete the block without flushing by @Jille in https://github.com/klauspost/compress/pull/927
* s2: Fix literal+repeat exceeds dst crash https://github.com/klauspost/compress/pull/930
* Feb 5th, 2024 - [1.17.6](https://github.com/klauspost/compress/releases/tag/v1.17.6)
* zstd: Fix incorrect repeat coding in best mode https://github.com/klauspost/compress/pull/923
* s2: Fix DecodeConcurrent deadlock on errors https://github.com/klauspost/compress/pull/925
* Jan 26th, 2024 - [v1.17.5](https://github.com/klauspost/compress/releases/tag/v1.17.5)
* flate: Fix reset with dictionary on custom window encodes https://github.com/klauspost/compress/pull/912
* zstd: Add Frame header encoding and stripping https://github.com/klauspost/compress/pull/908
* zstd: Limit better/best default window to 8MB https://github.com/klauspost/compress/pull/913
* zstd: Speed improvements by @greatroar in https://github.com/klauspost/compress/pull/896 https://github.com/klauspost/compress/pull/910
* s2: Fix callbacks for skippable blocks and disallow 0xfe (Padding) by @Jille in https://github.com/klauspost/compress/pull/916 https://github.com/klauspost/compress/pull/917
https://github.com/klauspost/compress/pull/919 https://github.com/klauspost/compress/pull/918
* Dec 1st, 2023 - [v1.17.4](https://github.com/klauspost/compress/releases/tag/v1.17.4)
* huff0: Speed up symbol counting by @greatroar in https://github.com/klauspost/compress/pull/887
* huff0: Remove byteReader by @greatroar in https://github.com/klauspost/compress/pull/886
* gzhttp: Allow overriding decompression on transport https://github.com/klauspost/compress/pull/892
* gzhttp: Clamp compression level https://github.com/klauspost/compress/pull/890
* gzip: Error out if reserved bits are set https://github.com/klauspost/compress/pull/891
* Nov 15th, 2023 - [v1.17.3](https://github.com/klauspost/compress/releases/tag/v1.17.3)
* fse: Fix max header size https://github.com/klauspost/compress/pull/881
* zstd: Improve better/best compression https://github.com/klauspost/compress/pull/877
* gzhttp: Fix missing content type on Close https://github.com/klauspost/compress/pull/883
* Oct 22nd, 2023 - [v1.17.2](https://github.com/klauspost/compress/releases/tag/v1.17.2)
* zstd: Fix rare *CORRUPTION* output in "best" mode. See https://github.com/klauspost/compress/pull/876
* Oct 14th, 2023 - [v1.17.1](https://github.com/klauspost/compress/releases/tag/v1.17.1)
* s2: Fix S2 "best" dictionary wrong encoding https://github.com/klauspost/compress/pull/871
* flate: Reduce allocations in decompressor and minor code improvements by @fakefloordiv in https://github.com/klauspost/compress/pull/869
* s2: Fix EstimateBlockSize on 6&7 length input https://github.com/klauspost/compress/pull/867
* 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
<details>
<summary>See changes to v1.16.x</summary>
* 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
* gzhttp: Handle informational headers by @rtribotte in https://github.com/klauspost/compress/pull/815
* s2: Improve Better compression slightly https://github.com/klauspost/compress/pull/663
* Apr 16, 2023 - [v1.16.5](https://github.com/klauspost/compress/releases/tag/v1.16.5)
* zstd: readByte needs to use io.ReadFull by @jnoxon in https://github.com/klauspost/compress/pull/802
* gzip: Fix WriterTo after initial read https://github.com/klauspost/compress/pull/804
* 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: Support 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
* s2: Add Intel LZ4s converter https://github.com/klauspost/compress/pull/766
* zstd: Minor bug fixes https://github.com/klauspost/compress/pull/771 https://github.com/klauspost/compress/pull/772 https://github.com/klauspost/compress/pull/773
* huff0: Speed up compress1xDo by @greatroar in https://github.com/klauspost/compress/pull/774
* Feb 26, 2023 - [v1.16.0](https://github.com/klauspost/compress/releases/tag/v1.16.0)
* s2: Add [Dictionary](https://github.com/klauspost/compress/tree/master/s2#dictionaries) support. https://github.com/klauspost/compress/pull/685
* s2: Add Compression Size Estimate. https://github.com/klauspost/compress/pull/752
* s2: Add support for custom stream encoder. https://github.com/klauspost/compress/pull/755
* s2: Add LZ4 block converter. https://github.com/klauspost/compress/pull/748
* 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>
<details>
<summary>See changes to v1.15.x</summary>
* Jan 21st, 2023 (v1.15.15)
* deflate: Improve level 7-9 https://github.com/klauspost/compress/pull/739
* 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
* zstd: Various speed improvements by @greatroar https://github.com/klauspost/compress/pull/741 https://github.com/klauspost/compress/pull/734 https://github.com/klauspost/compress/pull/736 https://github.com/klauspost/compress/pull/744 https://github.com/klauspost/compress/pull/743 https://github.com/klauspost/compress/pull/745
* gzhttp: Add SuffixETag() and DropETag() options to prevent ETag collisions on compressed responses by @willbicks in https://github.com/klauspost/compress/pull/740
@@ -183,7 +50,7 @@ https://github.com/klauspost/compress/pull/919 https://github.com/klauspost/comp
* zstd: Add [WithDecodeAllCapLimit](https://pkg.go.dev/github.com/klauspost/compress@v1.15.10/zstd#WithDecodeAllCapLimit) https://github.com/klauspost/compress/pull/649
* Add Go 1.19 - deprecate Go 1.16 https://github.com/klauspost/compress/pull/651
* flate: Improve level 5+6 compression https://github.com/klauspost/compress/pull/656
* zstd: Improve "better" compression https://github.com/klauspost/compress/pull/657
* zstd: Improve "better" compresssion https://github.com/klauspost/compress/pull/657
* s2: Improve "best" compression https://github.com/klauspost/compress/pull/658
* s2: Improve "better" compression. https://github.com/klauspost/compress/pull/635
* s2: Slightly faster non-assembly decompression https://github.com/klauspost/compress/pull/646
@@ -193,7 +60,7 @@ https://github.com/klauspost/compress/pull/919 https://github.com/klauspost/comp
* zstd: Fix decoder crash on amd64 (no BMI) on invalid input https://github.com/klauspost/compress/pull/645
* zstd: Disable decoder extended memory copies (amd64) due to possible crashes https://github.com/klauspost/compress/pull/644
* zstd: Allow single segments up to "max decoded size" https://github.com/klauspost/compress/pull/643
* zstd: Allow single segments up to "max decoded size" by @klauspost in https://github.com/klauspost/compress/pull/643
* July 13, 2022 (v1.15.8)
@@ -235,7 +102,7 @@ https://github.com/klauspost/compress/pull/919 https://github.com/klauspost/comp
* zstd: Speed up when WithDecoderLowmem(false) https://github.com/klauspost/compress/pull/599
* zstd: faster next state update in BMI2 version of decode by @WojciechMula in https://github.com/klauspost/compress/pull/593
* huff0: Do not check max size when reading table. https://github.com/klauspost/compress/pull/586
* flate: Inplace hashing for level 7-9 https://github.com/klauspost/compress/pull/590
* flate: Inplace hashing for level 7-9 by @klauspost in https://github.com/klauspost/compress/pull/590
* May 11, 2022 (v1.15.4)
@@ -262,12 +129,12 @@ https://github.com/klauspost/compress/pull/919 https://github.com/klauspost/comp
* zstd: Add stricter block size checks in [#523](https://github.com/klauspost/compress/pull/523)
* Mar 3, 2022 (v1.15.0)
* zstd: Refactor decoder [#498](https://github.com/klauspost/compress/pull/498)
* zstd: Add stream encoding without goroutines [#505](https://github.com/klauspost/compress/pull/505)
* zstd: Refactor decoder by @klauspost in [#498](https://github.com/klauspost/compress/pull/498)
* zstd: Add stream encoding without goroutines by @klauspost in [#505](https://github.com/klauspost/compress/pull/505)
* huff0: Prevent single blocks exceeding 16 bits by @klauspost in[#507](https://github.com/klauspost/compress/pull/507)
* flate: Inline literal emission [#509](https://github.com/klauspost/compress/pull/509)
* gzhttp: Add zstd to transport [#400](https://github.com/klauspost/compress/pull/400)
* gzhttp: Make content-type optional [#510](https://github.com/klauspost/compress/pull/510)
* flate: Inline literal emission by @klauspost in [#509](https://github.com/klauspost/compress/pull/509)
* gzhttp: Add zstd to transport by @klauspost in [#400](https://github.com/klauspost/compress/pull/400)
* gzhttp: Make content-type optional by @klauspost in [#510](https://github.com/klauspost/compress/pull/510)
Both compression and decompression now supports "synchronous" stream operations. This means that whenever "concurrency" is set to 1, they will operate without spawning goroutines.
@@ -275,8 +142,6 @@ 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>
@@ -284,7 +149,7 @@ While the release has been extensively tested, it is recommended to testing when
* flate: Fix rare huffman only (-2) corruption. [#503](https://github.com/klauspost/compress/pull/503)
* zip: Update deprecated CreateHeaderRaw to correctly call CreateRaw by @saracen in [#502](https://github.com/klauspost/compress/pull/502)
* zip: don't read data descriptor early by @saracen in [#501](https://github.com/klauspost/compress/pull/501) #501
* huff0: Use static decompression buffer up to 30% faster [#499](https://github.com/klauspost/compress/pull/499) [#500](https://github.com/klauspost/compress/pull/500)
* huff0: Use static decompression buffer up to 30% faster by @klauspost in [#499](https://github.com/klauspost/compress/pull/499) [#500](https://github.com/klauspost/compress/pull/500)
* Feb 17, 2022 (v1.14.3)
* flate: Improve fastest levels compression speed ~10% more throughput. [#482](https://github.com/klauspost/compress/pull/482) [#489](https://github.com/klauspost/compress/pull/489) [#490](https://github.com/klauspost/compress/pull/490) [#491](https://github.com/klauspost/compress/pull/491) [#494](https://github.com/klauspost/compress/pull/494) [#478](https://github.com/klauspost/compress/pull/478)
@@ -386,7 +251,7 @@ While the release has been extensively tested, it is recommended to testing when
* s2: Fix binaries.
* Feb 25, 2021 (v1.11.8)
* s2: Fixed occasional out-of-bounds write on amd64. Upgrade recommended.
* s2: Fixed occational out-of-bounds write on amd64. Upgrade recommended.
* s2: Add AMD64 assembly for better mode. 25-50% faster. [#315](https://github.com/klauspost/compress/pull/315)
* s2: Less upfront decoder allocation. [#322](https://github.com/klauspost/compress/pull/322)
* zstd: Faster "compression" of incompressible data. [#314](https://github.com/klauspost/compress/pull/314)
@@ -565,7 +430,7 @@ While the release has been extensively tested, it is recommended to testing when
* Feb 19, 2016: Faster bit writer, level -2 is 15% faster, level 1 is 4% faster.
* Feb 19, 2016: Handle small payloads faster in level 1-3.
* Feb 19, 2016: Added faster level 2 + 3 compression modes.
* Feb 19, 2016: [Rebalanced compression levels](https://blog.klauspost.com/rebalancing-deflate-compression-levels/), so there is a more even progression in terms of compression. New default level is 5.
* Feb 19, 2016: [Rebalanced compression levels](https://blog.klauspost.com/rebalancing-deflate-compression-levels/), so there is a more even progresssion in terms of compression. New default level is 5.
* Feb 14, 2016: Snappy: Merge upstream changes.
* Feb 14, 2016: Snappy: Fix aggressive skipping.
* Feb 14, 2016: Snappy: Update benchmark.
@@ -591,14 +456,12 @@ While the release has been extensively tested, it is recommended to testing when
The packages are drop-in replacements for standard libraries. Simply replace the import path to use them:
Typical speed is about 2x of the standard library packages.
| old import | new import | Documentation |
|------------------|---------------------------------------|-------------------------------------------------------------------------|
| `compress/gzip` | `github.com/klauspost/compress/gzip` | [gzip](https://pkg.go.dev/github.com/klauspost/compress/gzip?tab=doc) |
| `compress/zlib` | `github.com/klauspost/compress/zlib` | [zlib](https://pkg.go.dev/github.com/klauspost/compress/zlib?tab=doc) |
| `archive/zip` | `github.com/klauspost/compress/zip` | [zip](https://pkg.go.dev/github.com/klauspost/compress/zip?tab=doc) |
| `compress/flate` | `github.com/klauspost/compress/flate` | [flate](https://pkg.go.dev/github.com/klauspost/compress/flate?tab=doc) |
| old import | new import | Documentation
|--------------------|-----------------------------------------|--------------------|
| `compress/gzip` | `github.com/klauspost/compress/gzip` | [gzip](https://pkg.go.dev/github.com/klauspost/compress/gzip?tab=doc)
| `compress/zlib` | `github.com/klauspost/compress/zlib` | [zlib](https://pkg.go.dev/github.com/klauspost/compress/zlib?tab=doc)
| `archive/zip` | `github.com/klauspost/compress/zip` | [zip](https://pkg.go.dev/github.com/klauspost/compress/zip?tab=doc)
| `compress/flate` | `github.com/klauspost/compress/flate` | [flate](https://pkg.go.dev/github.com/klauspost/compress/flate?tab=doc)
* Optimized [deflate](https://godoc.org/github.com/klauspost/compress/flate) packages which can be used as a dropin replacement for [gzip](https://godoc.org/github.com/klauspost/compress/gzip), [zip](https://godoc.org/github.com/klauspost/compress/zip) and [zlib](https://godoc.org/github.com/klauspost/compress/zlib).
@@ -614,8 +477,6 @@ the stateless compress described below.
For compression performance, see: [this spreadsheet](https://docs.google.com/spreadsheets/d/1nuNE2nPfuINCZJRMt6wFWhKpToF95I47XjSsc-1rbPQ/edit?usp=sharing).
To disable all assembly add `-tags=noasm`. This works across all packages.
# Stateless compression
This package offers stateless compression as a special option for gzip/deflate.
@@ -634,7 +495,7 @@ For direct deflate use, NewStatelessWriter and StatelessDeflate are available. S
A `bufio.Writer` can of course be used to control write sizes. For example, to use a 4KB buffer:
```go
```
// replace 'ioutil.Discard' with your output.
gzw, err := gzip.NewWriterLevel(ioutil.Discard, gzip.StatelessCompression)
if err != nil {
@@ -653,6 +514,84 @@ This will only use up to 4KB in memory when the writer is idle.
Compression is almost always worse than the fastest compression level
and each write will allocate (a little) memory.
# Performance Update 2018
It has been a while since we have been looking at the speed of this package compared to the standard library, so I thought I would re-do my tests and give some overall recommendations based on the current state. All benchmarks have been performed with Go 1.10 on my Desktop Intel(R) Core(TM) i7-2600 CPU @3.40GHz. Since I last ran the tests, I have gotten more RAM, which means tests with big files are no longer limited by my SSD.
The raw results are in my [updated spreadsheet](https://docs.google.com/spreadsheets/d/1nuNE2nPfuINCZJRMt6wFWhKpToF95I47XjSsc-1rbPQ/edit?usp=sharing). Due to cgo changes and upstream updates i could not get the cgo version of gzip to compile. Instead I included the [zstd](https://github.com/datadog/zstd) cgo implementation. If I get cgo gzip to work again, I might replace the results in the sheet.
The columns to take note of are: *MB/s* - the throughput. *Reduction* - the data size reduction in percent of the original. *Rel Speed* relative speed compared to the standard library at the same level. *Smaller* - how many percent smaller is the compressed output compared to stdlib. Negative means the output was bigger. *Loss* means the loss (or gain) in compression as a percentage difference of the input.
The `gzstd` (standard library gzip) and `gzkp` (this package gzip) only uses one CPU core. [`pgzip`](https://github.com/klauspost/pgzip), [`bgzf`](https://github.com/biogo/hts/tree/master/bgzf) uses all 4 cores. [`zstd`](https://github.com/DataDog/zstd) uses one core, and is a beast (but not Go, yet).
## Overall differences.
There appears to be a roughly 5-10% speed advantage over the standard library when comparing at similar compression levels.
The biggest difference you will see is the result of [re-balancing](https://blog.klauspost.com/rebalancing-deflate-compression-levels/) the compression levels. I wanted by library to give a smoother transition between the compression levels than the standard library.
This package attempts to provide a more smooth transition, where "1" is taking a lot of shortcuts, "5" is the reasonable trade-off and "9" is the "give me the best compression", and the values in between gives something reasonable in between. The standard library has big differences in levels 1-4, but levels 5-9 having no significant gains - often spending a lot more time than can be justified by the achieved compression.
There are links to all the test data in the [spreadsheet](https://docs.google.com/spreadsheets/d/1nuNE2nPfuINCZJRMt6wFWhKpToF95I47XjSsc-1rbPQ/edit?usp=sharing) in the top left field on each tab.
## Web Content
This test set aims to emulate typical use in a web server. The test-set is 4GB data in 53k files, and is a mixture of (mostly) HTML, JS, CSS.
Since level 1 and 9 are close to being the same code, they are quite close. But looking at the levels in-between the differences are quite big.
Looking at level 6, this package is 88% faster, but will output about 6% more data. For a web server, this means you can serve 88% more data, but have to pay for 6% more bandwidth. You can draw your own conclusions on what would be the most expensive for your case.
## Object files
This test is for typical data files stored on a server. In this case it is a collection of Go precompiled objects. They are very compressible.
The picture is similar to the web content, but with small differences since this is very compressible. Levels 2-3 offer good speed, but is sacrificing quite a bit of compression.
The standard library seems suboptimal on level 3 and 4 - offering both worse compression and speed than level 6 & 7 of this package respectively.
## Highly Compressible File
This is a JSON file with very high redundancy. The reduction starts at 95% on level 1, so in real life terms we are dealing with something like a highly redundant stream of data, etc.
It is definitely visible that we are dealing with specialized content here, so the results are very scattered. This package does not do very well at levels 1-4, but picks up significantly at level 5 and levels 7 and 8 offering great speed for the achieved compression.
So if you know you content is extremely compressible you might want to go slightly higher than the defaults. The standard library has a huge gap between levels 3 and 4 in terms of speed (2.75x slowdown), so it offers little "middle ground".
## Medium-High Compressible
This is a pretty common test corpus: [enwik9](http://mattmahoney.net/dc/textdata.html). It contains the first 10^9 bytes of the English Wikipedia dump on Mar. 3, 2006. This is a very good test of typical text based compression and more data heavy streams.
We see a similar picture here as in "Web Content". On equal levels some compression is sacrificed for more speed. Level 5 seems to be the best trade-off between speed and size, beating stdlib level 3 in both.
## Medium Compressible
I will combine two test sets, one [10GB file set](http://mattmahoney.net/dc/10gb.html) and a VM disk image (~8GB). Both contain different data types and represent a typical backup scenario.
The most notable thing is how quickly the standard library drops to very low compression speeds around level 5-6 without any big gains in compression. Since this type of data is fairly common, this does not seem like good behavior.
## Un-compressible Content
This is mainly a test of how good the algorithms are at detecting un-compressible input. The standard library only offers this feature with very conservative settings at level 1. Obviously there is no reason for the algorithms to try to compress input that cannot be compressed. The only downside is that it might skip some compressible data on false detections.
## Huffman only compression
This compression library adds a special compression level, named `HuffmanOnly`, which allows near linear time compression. This is done by completely disabling matching of previous data, and only reduce the number of bits to represent each character.
This means that often used characters, like 'e' and ' ' (space) in text use the fewest bits to represent, and rare characters like '¤' takes more bits to represent. For more information see [wikipedia](https://en.wikipedia.org/wiki/Huffman_coding) or this nice [video](https://youtu.be/ZdooBTdW5bM).
Since this type of compression has much less variance, the compression speed is mostly unaffected by the input data, and is usually more than *180MB/s* for a single core.
The downside is that the compression ratio is usually considerably worse than even the fastest conventional compression. The compression ratio can never be better than 8:1 (12.5%).
The linear time compression can be used as a "better than nothing" mode, where you cannot risk the encoder to slow down on some content. For comparison, the size of the "Twain" text is *233460 bytes* (+29% vs. level 1) and encode speed is 144MB/s (4.5x level 1). So in this case you trade a 30% size increase for a 4 times speedup.
For more information see my blog post on [Fast Linear Time Compression](http://blog.klauspost.com/constant-time-gzipzip-compression/).
This is implemented on Go 1.7 as "Huffman Only" mode, though not exposed for gzip.
# Other packages
@@ -661,10 +600,6 @@ 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.
* [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

25
vendor/github.com/klauspost/compress/SECURITY.md generated vendored
View File

@@ -1,25 +0,0 @@
# Security Policy
## Supported Versions
Security updates are applied only to the latest release.
## Vulnerability Definition
A security vulnerability is a bug that with certain input triggers a crash or an infinite loop. Most calls will have varying execution time and only in rare cases will slow operation be considered a security vulnerability.
Corrupted output generally is not considered a security vulnerability, unless independent operations are able to affect each other. Note that not all functionality is re-entrant and safe to use concurrently.
Out-of-memory crashes only applies if the en/decoder uses an abnormal amount of memory, with appropriate options applied, to limit maximum window size, concurrency, etc. However, if you are in doubt you are welcome to file a security issue.
It is assumed that all callers are trusted, meaning internal data exposed through reflection or inspection of returned data structures is not considered a vulnerability.
Vulnerabilities resulting from compiler/assembler errors should be reported upstream. Depending on the severity this package may or may not implement a workaround.
## Reporting a Vulnerability
If you have discovered a security vulnerability in this project, please report it privately. **Do not disclose it as a public issue.** This gives us time to work with you to fix the issue before public exposure, reducing the chance that the exploit will be used before a patch is released.
Please disclose it at [security advisory](https://github.com/klauspost/compress/security/advisories/new). If possible please provide a minimal reproducer. If the issue only applies to a single platform, it would be helpful to provide access to that.
This project is maintained by a team of volunteers on a reasonable-effort basis. As such, vulnerabilities will be disclosed in a best effort base.

3
vendor/github.com/klauspost/compress/fse/bitwriter.go generated vendored
View File

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

5
vendor/github.com/klauspost/compress/fse/compress.go generated vendored
View File

@@ -199,8 +199,7 @@ func (s *Scratch) compress(src []byte) error {
c2.flush(s.actualTableLog)
c1.flush(s.actualTableLog)
s.bw.close()
return nil
return s.bw.close()
}
// writeCount will write the normalized histogram count to header.
@@ -212,7 +211,7 @@ func (s *Scratch) writeCount() error {
previous0 bool
charnum uint16
maxHeaderSize = ((int(s.symbolLen)*int(tableLog) + 4 + 2) >> 3) + 3
maxHeaderSize = ((int(s.symbolLen) * int(tableLog)) >> 3) + 3
// Write Table Size
bitStream = uint32(tableLog - minTablelog)

6
vendor/github.com/klauspost/compress/fse/decompress.go generated vendored
View File

@@ -15,7 +15,7 @@ const (
// It is possible, but by no way guaranteed that corrupt data will
// return an error.
// It is up to the caller to verify integrity of the returned data.
// Use a predefined Scratch to set maximum acceptable output size.
// Use a predefined Scrach to set maximum acceptable output size.
func Decompress(b []byte, s *Scratch) ([]byte, error) {
s, err := s.prepare(b)
if err != nil {
@@ -260,9 +260,7 @@ func (s *Scratch) buildDtable() error {
// If the buffer is over-read an error is returned.
func (s *Scratch) decompress() error {
br := &s.bits
if err := br.init(s.br.unread()); err != nil {
return err
}
br.init(s.br.unread())
var s1, s2 decoder
// Initialize and decode first state and symbol.

25
vendor/github.com/klauspost/compress/huff0/bitreader.go generated vendored
View File

@@ -6,11 +6,10 @@
package huff0
import (
"encoding/binary"
"errors"
"fmt"
"io"
"github.com/klauspost/compress/internal/le"
)
// bitReader reads a bitstream in reverse.
@@ -47,7 +46,7 @@ func (b *bitReaderBytes) init(in []byte) error {
return nil
}
// peekByteFast requires that at least one byte is requested every time.
// peekBitsFast requires that at least one bit is requested every time.
// There are no checks if the buffer is filled.
func (b *bitReaderBytes) peekByteFast() uint8 {
got := uint8(b.value >> 56)
@@ -67,7 +66,8 @@ func (b *bitReaderBytes) fillFast() {
}
// 2 bounds checks.
low := le.Load32(b.in, b.off-4)
v := b.in[b.off-4 : b.off]
low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24)
b.value |= uint64(low) << (b.bitsRead - 32)
b.bitsRead -= 32
b.off -= 4
@@ -76,7 +76,7 @@ func (b *bitReaderBytes) fillFast() {
// fillFastStart() assumes the bitReaderBytes is empty and there is at least 8 bytes to read.
func (b *bitReaderBytes) fillFastStart() {
// Do single re-slice to avoid bounds checks.
b.value = le.Load64(b.in, b.off-8)
b.value = binary.LittleEndian.Uint64(b.in[b.off-8:])
b.bitsRead = 0
b.off -= 8
}
@@ -86,8 +86,9 @@ func (b *bitReaderBytes) fill() {
if b.bitsRead < 32 {
return
}
if b.off >= 4 {
low := le.Load32(b.in, b.off-4)
if b.off > 4 {
v := b.in[b.off-4 : b.off]
low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24)
b.value |= uint64(low) << (b.bitsRead - 32)
b.bitsRead -= 32
b.off -= 4
@@ -174,7 +175,9 @@ func (b *bitReaderShifted) fillFast() {
return
}
low := le.Load32(b.in, b.off-4)
// 2 bounds checks.
v := b.in[b.off-4 : b.off]
low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24)
b.value |= uint64(low) << ((b.bitsRead - 32) & 63)
b.bitsRead -= 32
b.off -= 4
@@ -182,7 +185,8 @@ func (b *bitReaderShifted) fillFast() {
// fillFastStart() assumes the bitReaderShifted is empty and there is at least 8 bytes to read.
func (b *bitReaderShifted) fillFastStart() {
b.value = le.Load64(b.in, b.off-8)
// Do single re-slice to avoid bounds checks.
b.value = binary.LittleEndian.Uint64(b.in[b.off-8:])
b.bitsRead = 0
b.off -= 8
}
@@ -193,7 +197,8 @@ func (b *bitReaderShifted) fill() {
return
}
if b.off > 4 {
low := le.Load32(b.in, b.off-4)
v := b.in[b.off-4 : b.off]
low := (uint32(v[0])) | (uint32(v[1]) << 8) | (uint32(v[2]) << 16) | (uint32(v[3]) << 24)
b.value |= uint64(low) << ((b.bitsRead - 32) & 63)
b.bitsRead -= 32
b.off -= 4

27
vendor/github.com/klauspost/compress/huff0/bitwriter.go generated vendored
View File

@@ -13,6 +13,14 @@ type bitWriter struct {
out []byte
}
// bitMask16 is bitmasks. Has extra to avoid bounds check.
var bitMask16 = [32]uint16{
0, 1, 3, 7, 0xF, 0x1F,
0x3F, 0x7F, 0xFF, 0x1FF, 0x3FF, 0x7FF,
0xFFF, 0x1FFF, 0x3FFF, 0x7FFF, 0xFFFF, 0xFFFF,
0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF, 0xFFFF,
0xFFFF, 0xFFFF} /* up to 16 bits */
// addBits16Clean will add up to 16 bits. value may not contain more set bits than indicated.
// It will not check if there is space for them, so the caller must ensure that it has flushed recently.
func (b *bitWriter) addBits16Clean(value uint16, bits uint8) {
@@ -52,22 +60,6 @@ func (b *bitWriter) encTwoSymbols(ct cTable, av, bv byte) {
b.nBits += encA.nBits + encB.nBits
}
// encFourSymbols adds up to 32 bits from four symbols.
// It will not check if there is space for them,
// so the caller must ensure that b has been flushed recently.
func (b *bitWriter) encFourSymbols(encA, encB, encC, encD cTableEntry) {
bitsA := encA.nBits
bitsB := bitsA + encB.nBits
bitsC := bitsB + encC.nBits
bitsD := bitsC + encD.nBits
combined := uint64(encA.val) |
(uint64(encB.val) << (bitsA & 63)) |
(uint64(encC.val) << (bitsB & 63)) |
(uint64(encD.val) << (bitsC & 63))
b.bitContainer |= combined << (b.nBits & 63)
b.nBits += bitsD
}
// flush32 will flush out, so there are at least 32 bits available for writing.
func (b *bitWriter) flush32() {
if b.nBits < 32 {
@@ -94,9 +86,10 @@ func (b *bitWriter) flushAlign() {
// close will write the alignment bit and write the final byte(s)
// to the output.
func (b *bitWriter) close() {
func (b *bitWriter) close() error {
// End mark
b.addBits16Clean(1, 1)
// flush until next byte.
b.flushAlign()
return nil
}

44
vendor/github.com/klauspost/compress/huff0/bytereader.go generated vendored Normal file
View File

@@ -0,0 +1,44 @@
// Copyright 2018 Klaus Post. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Based on work Copyright (c) 2013, Yann Collet, released under BSD License.
package huff0
// byteReader provides a byte reader that reads
// little endian values from a byte stream.
// The input stream is manually advanced.
// The reader performs no bounds checks.
type byteReader struct {
b []byte
off int
}
// init will initialize the reader and set the input.
func (b *byteReader) init(in []byte) {
b.b = in
b.off = 0
}
// Int32 returns a little endian int32 starting at current offset.
func (b byteReader) Int32() int32 {
v3 := int32(b.b[b.off+3])
v2 := int32(b.b[b.off+2])
v1 := int32(b.b[b.off+1])
v0 := int32(b.b[b.off])
return (v3 << 24) | (v2 << 16) | (v1 << 8) | v0
}
// Uint32 returns a little endian uint32 starting at current offset.
func (b byteReader) Uint32() uint32 {
v3 := uint32(b.b[b.off+3])
v2 := uint32(b.b[b.off+2])
v1 := uint32(b.b[b.off+1])
v0 := uint32(b.b[b.off])
return (v3 << 24) | (v2 << 16) | (v1 << 8) | v0
}
// remain will return the number of bytes remaining.
func (b byteReader) remain() int {
return len(b.b) - b.off
}

28
vendor/github.com/klauspost/compress/huff0/compress.go generated vendored
View File

@@ -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), nil
return s.compress1xDo(s.Out, src)
}
func (s *Scratch) compress1xDo(dst, src []byte) []byte {
func (s *Scratch) compress1xDo(dst, src []byte) ([]byte, error) {
var bw = bitWriter{out: dst}
// N is length divisible by 4.
@@ -248,7 +248,8 @@ func (s *Scratch) compress1xDo(dst, src []byte) []byte {
tmp := src[n : n+4]
// tmp should be len 4
bw.flush32()
bw.encFourSymbols(cTable[tmp[3]], cTable[tmp[2]], cTable[tmp[1]], cTable[tmp[0]])
bw.encTwoSymbols(cTable, tmp[3], tmp[2])
bw.encTwoSymbols(cTable, tmp[1], tmp[0])
}
} else {
for ; n >= 0; n -= 4 {
@@ -260,8 +261,8 @@ func (s *Scratch) compress1xDo(dst, src []byte) []byte {
bw.encTwoSymbols(cTable, tmp[1], tmp[0])
}
}
bw.close()
return bw.out
err := bw.close()
return bw.out, err
}
var sixZeros [6]byte
@@ -283,8 +284,12 @@ func (s *Scratch) compress4X(src []byte) ([]byte, error) {
}
src = src[len(toDo):]
var err error
idx := len(s.Out)
s.Out = s.compress1xDo(s.Out, toDo)
s.Out, err = 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
@@ -311,6 +316,7 @@ 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
@@ -321,12 +327,15 @@ func (s *Scratch) compress4Xp(src []byte) ([]byte, error) {
// Separate goroutine for each block.
go func(i int) {
s.tmpOut[i] = s.compress1xDo(s.tmpOut[i][:0], toDo)
s.tmpOut[i], errs[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
@@ -350,7 +359,6 @@ func (s *Scratch) compress4Xp(src []byte) ([]byte, error) {
// Does not update s.clearCount.
func (s *Scratch) countSimple(in []byte) (max int, reuse bool) {
reuse = true
_ = s.count // Assert that s != nil to speed up the following loop.
for _, v := range in {
s.count[v]++
}
@@ -416,7 +424,7 @@ func (s *Scratch) validateTable(c cTable) bool {
// minTableLog provides the minimum logSize to safely represent a distribution.
func (s *Scratch) minTableLog() uint8 {
minBitsSrc := highBit32(uint32(s.srcLen)) + 1
minBitsSrc := highBit32(uint32(s.br.remain())) + 1
minBitsSymbols := highBit32(uint32(s.symbolLen-1)) + 2
if minBitsSrc < minBitsSymbols {
return uint8(minBitsSrc)
@@ -428,7 +436,7 @@ func (s *Scratch) minTableLog() uint8 {
func (s *Scratch) optimalTableLog() {
tableLog := s.TableLog
minBits := s.minTableLog()
maxBitsSrc := uint8(highBit32(uint32(s.srcLen-1))) - 1
maxBitsSrc := uint8(highBit32(uint32(s.br.remain()-1))) - 1
if maxBitsSrc < tableLog {
// Accuracy can be reduced
tableLog = maxBitsSrc

6
vendor/github.com/klauspost/compress/huff0/decompress.go generated vendored
View File

@@ -253,7 +253,7 @@ func (d *Decoder) decompress1X8Bit(dst, src []byte) ([]byte, error) {
switch d.actualTableLog {
case 8:
const shift = 0
const shift = 8 - 8
for br.off >= 4 {
br.fillFast()
v := dt[uint8(br.value>>(56+shift))]
@@ -1136,7 +1136,7 @@ func (s *Scratch) matches(ct cTable, w io.Writer) {
errs++
}
if errs > 0 {
fmt.Fprintf(w, "%d errors in base, stopping\n", errs)
fmt.Fprintf(w, "%d errros in base, stopping\n", errs)
continue
}
// Ensure that all combinations are covered.
@@ -1152,7 +1152,7 @@ func (s *Scratch) matches(ct cTable, w io.Writer) {
errs++
}
if errs > 20 {
fmt.Fprintf(w, "%d errors, stopping\n", errs)
fmt.Fprintf(w, "%d errros, stopping\n", errs)
break
}
}

4
vendor/github.com/klauspost/compress/huff0/huff0.go generated vendored
View File

@@ -88,7 +88,7 @@ type Scratch struct {
// Decoders will return ErrMaxDecodedSizeExceeded is this limit is exceeded.
MaxDecodedSize int
srcLen int
br byteReader
// MaxSymbolValue will override the maximum symbol value of the next block.
MaxSymbolValue uint8
@@ -170,7 +170,7 @@ func (s *Scratch) prepare(in []byte) (*Scratch, error) {
if s.fse == nil {
s.fse = &fse.Scratch{}
}
s.srcLen = len(in)
s.br.init(in)
return s, nil
}

5
vendor/github.com/klauspost/compress/internal/le/le.go generated vendored
View File

@@ -1,5 +0,0 @@
package le
type Indexer interface {
int | int8 | int16 | int32 | int64 | uint | uint8 | uint16 | uint32 | uint64
}

42
vendor/github.com/klauspost/compress/internal/le/unsafe_disabled.go generated vendored
View File

@@ -1,42 +0,0 @@
//go:build !(amd64 || arm64 || ppc64le || riscv64) || nounsafe || purego || appengine
package le
import (
"encoding/binary"
)
// Load8 will load from b at index i.
func Load8[I Indexer](b []byte, i I) byte {
return b[i]
}
// Load16 will load from b at index i.
func Load16[I Indexer](b []byte, i I) uint16 {
return binary.LittleEndian.Uint16(b[i:])
}
// Load32 will load from b at index i.
func Load32[I Indexer](b []byte, i I) uint32 {
return binary.LittleEndian.Uint32(b[i:])
}
// Load64 will load from b at index i.
func Load64[I Indexer](b []byte, i I) uint64 {
return binary.LittleEndian.Uint64(b[i:])
}
// Store16 will store v at b.
func Store16(b []byte, v uint16) {
binary.LittleEndian.PutUint16(b, v)
}
// Store32 will store v at b.
func Store32(b []byte, v uint32) {
binary.LittleEndian.PutUint32(b, v)
}
// Store64 will store v at b.
func Store64(b []byte, v uint64) {
binary.LittleEndian.PutUint64(b, v)
}

55
vendor/github.com/klauspost/compress/internal/le/unsafe_enabled.go generated vendored
View File

@@ -1,55 +0,0 @@
// We enable 64 bit LE platforms:
//go:build (amd64 || arm64 || ppc64le || riscv64) && !nounsafe && !purego && !appengine
package le
import (
"unsafe"
)
// Load8 will load from b at index i.
func Load8[I Indexer](b []byte, i I) byte {
//return binary.LittleEndian.Uint16(b[i:])
//return *(*uint16)(unsafe.Pointer(&b[i]))
return *(*byte)(unsafe.Add(unsafe.Pointer(unsafe.SliceData(b)), i))
}
// Load16 will load from b at index i.
func Load16[I Indexer](b []byte, i I) uint16 {
//return binary.LittleEndian.Uint16(b[i:])
//return *(*uint16)(unsafe.Pointer(&b[i]))
return *(*uint16)(unsafe.Add(unsafe.Pointer(unsafe.SliceData(b)), i))
}
// Load32 will load from b at index i.
func Load32[I Indexer](b []byte, i I) uint32 {
//return binary.LittleEndian.Uint32(b[i:])
//return *(*uint32)(unsafe.Pointer(&b[i]))
return *(*uint32)(unsafe.Add(unsafe.Pointer(unsafe.SliceData(b)), i))
}
// Load64 will load from b at index i.
func Load64[I Indexer](b []byte, i I) uint64 {
//return binary.LittleEndian.Uint64(b[i:])
//return *(*uint64)(unsafe.Pointer(&b[i]))
return *(*uint64)(unsafe.Add(unsafe.Pointer(unsafe.SliceData(b)), i))
}
// Store16 will store v at b.
func Store16(b []byte, v uint16) {
//binary.LittleEndian.PutUint16(b, v)
*(*uint16)(unsafe.Pointer(unsafe.SliceData(b))) = v
}
// Store32 will store v at b.
func Store32(b []byte, v uint32) {
//binary.LittleEndian.PutUint32(b, v)
*(*uint32)(unsafe.Pointer(unsafe.SliceData(b))) = v
}
// Store64 will store v at b.
func Store64(b []byte, v uint64) {
//binary.LittleEndian.PutUint64(b, v)
*(*uint64)(unsafe.Pointer(unsafe.SliceData(b))) = v
}

14
vendor/github.com/klauspost/compress/internal/snapref/encode_other.go generated vendored
View File

@@ -51,7 +51,7 @@ func emitCopy(dst []byte, offset, length int) int {
i := 0
// The maximum length for a single tagCopy1 or tagCopy2 op is 64 bytes. The
// threshold for this loop is a little higher (at 68 = 64 + 4), and the
// length emitted down below is a little lower (at 60 = 64 - 4), because
// length emitted down below is is a little lower (at 60 = 64 - 4), because
// it's shorter to encode a length 67 copy as a length 60 tagCopy2 followed
// by a length 7 tagCopy1 (which encodes as 3+2 bytes) than to encode it as
// a length 64 tagCopy2 followed by a length 3 tagCopy2 (which encodes as
@@ -87,6 +87,18 @@ func emitCopy(dst []byte, offset, length int) int {
return i + 2
}
// extendMatch returns the largest k such that k <= len(src) and that
// src[i:i+k-j] and src[j:k] have the same contents.
//
// It assumes that:
//
// 0 <= i && i < j && j <= len(src)
func extendMatch(src []byte, i, j int) int {
for ; j < len(src) && src[i] == src[j]; i, j = i+1, j+1 {
}
return j
}
func hash(u, shift uint32) uint32 {
return (u * 0x1e35a7bd) >> shift
}

3
vendor/github.com/klauspost/compress/s2sx.mod generated vendored
View File

@@ -1,3 +1,4 @@
module github.com/klauspost/compress
go 1.22
go 1.16

6
vendor/github.com/klauspost/compress/zstd/README.md generated vendored
View File

@@ -6,7 +6,7 @@ A high performance compression algorithm is implemented. For now focused on spee
This package provides [compression](#Compressor) to and [decompression](#Decompressor) of Zstandard content.
This package is pure Go. Use `noasm` and `nounsafe` to disable relevant features.
This package is pure Go and without use of "unsafe".
The `zstd` package is provided as open source software using a Go standard license.
@@ -259,7 +259,7 @@ nyc-taxi-data-10M.csv gzkp 1 3325605752 922273214 13929 227.68
## Decompressor
Status: STABLE - there may still be subtle bugs, but a wide variety of content has been tested.
Staus: STABLE - there may still be subtle bugs, but a wide variety of content has been tested.
This library is being continuously [fuzz-tested](https://github.com/klauspost/compress-fuzz),
kindly supplied by [fuzzit.dev](https://fuzzit.dev/).
@@ -304,7 +304,7 @@ import "github.com/klauspost/compress/zstd"
// Create a reader that caches decompressors.
// For this operation type we supply a nil Reader.
var decoder, _ = zstd.NewReader(nil, zstd.WithDecoderConcurrency(0))
var decoder, _ = zstd.NewReader(nil, WithDecoderConcurrency(0))
// Decompress a buffer. We don't supply a destination buffer,
// so it will be allocated by the decoder.

43
vendor/github.com/klauspost/compress/zstd/bitreader.go generated vendored
View File

@@ -5,12 +5,11 @@
package zstd
import (
"encoding/binary"
"errors"
"fmt"
"io"
"math/bits"
"github.com/klauspost/compress/internal/le"
)
// bitReader reads a bitstream in reverse.
@@ -18,8 +17,8 @@ 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.
cursor int // offset where next read should end
bitsRead uint8
}
@@ -29,12 +28,12 @@ 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 {
return errors.New("corrupt stream, did not find end of stream")
}
b.cursor = len(in)
b.bitsRead = 64
b.value = 0
if len(in) >= 8 {
@@ -70,16 +69,21 @@ func (b *bitReader) fillFast() {
if b.bitsRead < 32 {
return
}
b.cursor -= 4
b.value = (b.value << 32) | uint64(le.Load32(b.in, b.cursor))
// 2 bounds checks.
v := b.in[b.off-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() {
b.cursor -= 8
b.value = le.Load64(b.in, b.cursor)
// Do single re-slice to avoid bounds checks.
b.value = binary.LittleEndian.Uint64(b.in[b.off-8:])
b.bitsRead = 0
b.off -= 8
}
// fill() will make sure at least 32 bits are available.
@@ -87,23 +91,25 @@ func (b *bitReader) fill() {
if b.bitsRead < 32 {
return
}
if b.cursor >= 4 {
b.cursor -= 4
b.value = (b.value << 32) | uint64(le.Load32(b.in, b.cursor))
if b.off >= 4 {
v := b.in[b.off-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
return
}
b.bitsRead -= uint8(8 * b.cursor)
for b.cursor > 0 {
b.cursor -= 1
b.value = (b.value << 8) | uint64(b.in[b.cursor])
for b.off > 0 {
b.value = (b.value << 8) | uint64(b.in[b.off-1])
b.bitsRead -= 8
b.off--
}
}
// finished returns true if all bits have been read from the bit stream.
func (b *bitReader) finished() bool {
return b.cursor == 0 && b.bitsRead >= 64
return b.off == 0 && b.bitsRead >= 64
}
// overread returns true if more bits have been requested than is on the stream.
@@ -113,14 +119,13 @@ func (b *bitReader) overread() bool {
// remain returns the number of bits remaining.
func (b *bitReader) remain() uint {
return 8*uint(b.cursor) + 64 - uint(b.bitsRead)
return b.off*8 + 64 - uint(b.bitsRead)
}
// close the bitstream and returns an error if out-of-buffer reads occurred.
func (b *bitReader) close() error {
// Release reference.
b.in = nil
b.cursor = 0
if !b.finished() {
return fmt.Errorf("%d extra bits on block, should be 0", b.remain())
}

3
vendor/github.com/klauspost/compress/zstd/bitwriter.go generated vendored
View File

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

32
vendor/github.com/klauspost/compress/zstd/blockdec.go generated vendored
View File

@@ -5,10 +5,13 @@
package zstd
import (
"bytes"
"encoding/binary"
"errors"
"fmt"
"hash/crc32"
"io"
"os"
"path/filepath"
"sync"
"github.com/klauspost/compress/huff0"
@@ -439,9 +442,6 @@ func (b *blockDec) decodeLiterals(in []byte, hist *history) (remain []byte, err
}
}
var err error
if debugDecoder {
println("huff table input:", len(literals), "CRC:", crc32.ChecksumIEEE(literals))
}
huff, literals, err = huff0.ReadTable(literals, huff)
if err != nil {
println("reading huffman table:", err)
@@ -550,9 +550,6 @@ func (b *blockDec) prepareSequences(in []byte, hist *history) (err error) {
if debugDecoder {
printf("Compression modes: 0b%b", compMode)
}
if compMode&3 != 0 {
return errors.New("corrupt block: reserved bits not zero")
}
for i := uint(0); i < 3; i++ {
mode := seqCompMode((compMode >> (6 - i*2)) & 3)
if debugDecoder {
@@ -591,12 +588,10 @@ func (b *blockDec) prepareSequences(in []byte, hist *history) (err error) {
}
seq.fse.setRLE(symb)
if debugDecoder {
printf("RLE set to 0x%x, code: %v", symb, v)
printf("RLE set to %+v, code: %v", symb, v)
}
case compModeFSE:
if debugDecoder {
println("Reading table for", tableIndex(i))
}
println("Reading table for", tableIndex(i))
if seq.fse == nil || seq.fse.preDefined {
seq.fse = fseDecoderPool.Get().(*fseDecoder)
}
@@ -644,6 +639,21 @@ func (b *blockDec) prepareSequences(in []byte, hist *history) (err error) {
println("initializing sequences:", err)
return err
}
// Extract blocks...
if false && hist.dict == nil {
fatalErr := func(err error) {
if err != nil {
panic(err)
}
}
fn := fmt.Sprintf("n-%d-lits-%d-prev-%d-%d-%d-win-%d.blk", hist.decoders.nSeqs, len(hist.decoders.literals), hist.recentOffsets[0], hist.recentOffsets[1], hist.recentOffsets[2], hist.windowSize)
var buf bytes.Buffer
fatalErr(binary.Write(&buf, binary.LittleEndian, hist.decoders.litLengths.fse))
fatalErr(binary.Write(&buf, binary.LittleEndian, hist.decoders.matchLengths.fse))
fatalErr(binary.Write(&buf, binary.LittleEndian, hist.decoders.offsets.fse))
buf.Write(in)
os.WriteFile(filepath.Join("testdata", "seqs", fn), buf.Bytes(), os.ModePerm)
}
return nil
}

85
vendor/github.com/klauspost/compress/zstd/blockenc.go generated vendored
View File

@@ -9,7 +9,6 @@ import (
"fmt"
"math"
"math/bits"
"slices"
"github.com/klauspost/compress/huff0"
)
@@ -362,21 +361,14 @@ 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) > 16 {
} else if len(lits) > 32 {
// 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 {
@@ -428,16 +420,6 @@ func (b *blockEnc) encodeLits(lits []byte, raw bool) error {
return nil
}
// encodeRLE will encode an RLE block.
func (b *blockEnc) encodeRLE(val byte, length uint32) {
var bh blockHeader
bh.setLast(b.last)
bh.setSize(length)
bh.setType(blockTypeRLE)
b.output = bh.appendTo(b.output)
b.output = append(b.output, val)
}
// fuzzFseEncoder can be used to fuzz the FSE encoder.
func fuzzFseEncoder(data []byte) int {
if len(data) > maxSequences || len(data) < 2 {
@@ -458,7 +440,16 @@ func fuzzFseEncoder(data []byte) int {
// All 0
return 0
}
cnt := int(slices.Max(hist[:maxSym]))
maxCount := func(a []uint32) int {
var max uint32
for _, v := range a {
if v > max {
max = v
}
}
return int(max)
}
cnt := maxCount(hist[:maxSym])
if cnt == len(data) {
// RLE
return 0
@@ -481,18 +472,8 @@ func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error {
if len(b.sequences) == 0 {
return b.encodeLits(b.literals, rawAllLits)
}
if len(b.sequences) == 1 && len(org) > 0 && len(b.literals) <= 1 {
// Check common RLE cases.
seq := b.sequences[0]
if seq.litLen == uint32(len(b.literals)) && seq.offset-3 == 1 {
// Offset == 1 and 0 or 1 literals.
b.encodeRLE(org[0], b.sequences[0].matchLen+zstdMinMatch+seq.litLen)
return nil
}
}
// We want some difference to at least account for the headers.
saved := b.size - len(b.literals) - (b.size >> 6)
saved := b.size - len(b.literals) - (b.size >> 5)
if saved < 16 {
if org == nil {
return errIncompressible
@@ -522,7 +503,7 @@ 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) > 16 && !raw {
} else if len(b.literals) > 32 && !raw {
// Use 1 stream
single = true
out, reUsed, err = huff0.Compress1X(b.literals, b.litEnc)
@@ -530,17 +511,6 @@ 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)
@@ -803,16 +773,16 @@ func (b *blockEnc) encode(org []byte, raw, rawAllLits bool) error {
ml.flush(mlEnc.actualTableLog)
of.flush(ofEnc.actualTableLog)
ll.flush(llEnc.actualTableLog)
wr.close()
err = wr.close()
if err != nil {
return err
}
b.output = wr.out
// Maybe even add a bigger margin.
if len(b.output)-3-bhOffset >= b.size {
// Discard and encode as raw block.
b.output = b.encodeRawTo(b.output[:bhOffset], org)
b.popOffsets()
// Maybe even add a bigger margin.
b.litEnc.Reuse = huff0.ReusePolicyNone
return nil
return errIncompressible
}
// Size is output minus block header.
@@ -876,6 +846,15 @@ func (b *blockEnc) genCodes() {
}
}
}
maxCount := func(a []uint32) int {
var max uint32
for _, v := range a {
if v > max {
max = v
}
}
return int(max)
}
if debugAsserts && mlMax > maxMatchLengthSymbol {
panic(fmt.Errorf("mlMax > maxMatchLengthSymbol (%d)", mlMax))
}
@@ -886,7 +865,7 @@ func (b *blockEnc) genCodes() {
panic(fmt.Errorf("llMax > maxLiteralLengthSymbol (%d)", llMax))
}
b.coders.mlEnc.HistogramFinished(mlMax, int(slices.Max(mlH[:mlMax+1])))
b.coders.ofEnc.HistogramFinished(ofMax, int(slices.Max(ofH[:ofMax+1])))
b.coders.llEnc.HistogramFinished(llMax, int(slices.Max(llH[:llMax+1])))
b.coders.mlEnc.HistogramFinished(mlMax, maxCount(mlH[:mlMax+1]))
b.coders.ofEnc.HistogramFinished(ofMax, maxCount(ofH[:ofMax+1]))
b.coders.llEnc.HistogramFinished(llMax, maxCount(llH[:llMax+1]))
}

4
vendor/github.com/klauspost/compress/zstd/bytebuf.go generated vendored
View File

@@ -54,7 +54,7 @@ func (b *byteBuf) readBig(n int, dst []byte) ([]byte, error) {
func (b *byteBuf) readByte() (byte, error) {
bb := *b
if len(bb) < 1 {
return 0, io.ErrUnexpectedEOF
return 0, nil
}
r := bb[0]
*b = bb[1:]
@@ -109,7 +109,7 @@ func (r *readerWrapper) readBig(n int, dst []byte) ([]byte, error) {
}
func (r *readerWrapper) readByte() (byte, error) {
n2, err := io.ReadFull(r.r, r.tmp[:1])
n2, err := r.r.Read(r.tmp[:1])
if err != nil {
if err == io.EOF {
err = io.ErrUnexpectedEOF

56
vendor/github.com/klauspost/compress/zstd/decodeheader.go generated vendored
View File

@@ -95,54 +95,42 @@ type Header struct {
// If there isn't enough input, io.ErrUnexpectedEOF is returned.
// The FirstBlock.OK will indicate if enough information was available to decode the first block header.
func (h *Header) Decode(in []byte) error {
_, err := h.DecodeAndStrip(in)
return err
}
// DecodeAndStrip will decode the header from the beginning of the stream
// and on success return the remaining bytes.
// This will decode the frame header and the first block header if enough bytes are provided.
// It is recommended to provide at least HeaderMaxSize bytes.
// If the frame header cannot be read an error will be returned.
// If there isn't enough input, io.ErrUnexpectedEOF is returned.
// The FirstBlock.OK will indicate if enough information was available to decode the first block header.
func (h *Header) DecodeAndStrip(in []byte) (remain []byte, err error) {
*h = Header{}
if len(in) < 4 {
return nil, io.ErrUnexpectedEOF
return io.ErrUnexpectedEOF
}
h.HeaderSize += 4
b, in := in[:4], in[4:]
if string(b) != frameMagic {
if string(b[1:4]) != skippableFrameMagic || b[0]&0xf0 != 0x50 {
return nil, ErrMagicMismatch
return ErrMagicMismatch
}
if len(in) < 4 {
return nil, io.ErrUnexpectedEOF
return io.ErrUnexpectedEOF
}
h.HeaderSize += 4
h.Skippable = true
h.SkippableID = int(b[0] & 0xf)
h.SkippableSize = binary.LittleEndian.Uint32(in)
return in[4:], nil
return nil
}
// Read Window_Descriptor
// https://github.com/facebook/zstd/blob/dev/doc/zstd_compression_format.md#window_descriptor
if len(in) < 1 {
return nil, io.ErrUnexpectedEOF
return io.ErrUnexpectedEOF
}
fhd, in := in[0], in[1:]
h.HeaderSize++
h.SingleSegment = fhd&(1<<5) != 0
h.HasCheckSum = fhd&(1<<2) != 0
if fhd&(1<<3) != 0 {
return nil, errors.New("reserved bit set on frame header")
return errors.New("reserved bit set on frame header")
}
if !h.SingleSegment {
if len(in) < 1 {
return nil, io.ErrUnexpectedEOF
return io.ErrUnexpectedEOF
}
var wd byte
wd, in = in[0], in[1:]
@@ -160,7 +148,7 @@ func (h *Header) DecodeAndStrip(in []byte) (remain []byte, err error) {
size = 4
}
if len(in) < int(size) {
return nil, io.ErrUnexpectedEOF
return io.ErrUnexpectedEOF
}
b, in = in[:size], in[size:]
h.HeaderSize += int(size)
@@ -190,7 +178,7 @@ func (h *Header) DecodeAndStrip(in []byte) (remain []byte, err error) {
if fcsSize > 0 {
h.HasFCS = true
if len(in) < fcsSize {
return nil, io.ErrUnexpectedEOF
return io.ErrUnexpectedEOF
}
b, in = in[:fcsSize], in[fcsSize:]
h.HeaderSize += int(fcsSize)
@@ -211,7 +199,7 @@ func (h *Header) DecodeAndStrip(in []byte) (remain []byte, err error) {
// Frame Header done, we will not fail from now on.
if len(in) < 3 {
return in, nil
return nil
}
tmp := in[:3]
bh := uint32(tmp[0]) | (uint32(tmp[1]) << 8) | (uint32(tmp[2]) << 16)
@@ -221,7 +209,7 @@ func (h *Header) DecodeAndStrip(in []byte) (remain []byte, err error) {
cSize := int(bh >> 3)
switch blockType {
case blockTypeReserved:
return in, nil
return nil
case blockTypeRLE:
h.FirstBlock.Compressed = true
h.FirstBlock.DecompressedSize = cSize
@@ -237,25 +225,5 @@ func (h *Header) DecodeAndStrip(in []byte) (remain []byte, err error) {
}
h.FirstBlock.OK = true
return in, nil
}
// AppendTo will append the encoded header to the dst slice.
// There is no error checking performed on the header values.
func (h *Header) AppendTo(dst []byte) ([]byte, error) {
if h.Skippable {
magic := [4]byte{0x50, 0x2a, 0x4d, 0x18}
magic[0] |= byte(h.SkippableID & 0xf)
dst = append(dst, magic[:]...)
f := h.SkippableSize
return append(dst, uint8(f), uint8(f>>8), uint8(f>>16), uint8(f>>24)), nil
}
f := frameHeader{
ContentSize: h.FrameContentSize,
WindowSize: uint32(h.WindowSize),
SingleSegment: h.SingleSegment,
Checksum: h.HasCheckSum,
DictID: h.DictionaryID,
}
return f.appendTo(dst), nil
return nil
}

12
vendor/github.com/klauspost/compress/zstd/decoder.go generated vendored
View File

@@ -82,7 +82,7 @@ var (
// can run multiple concurrent stateless decodes. It is even possible to
// use stateless decodes while a stream is being decoded.
//
// The Reset function can be used to initiate a new stream, which will considerably
// The Reset function can be used to initiate a new stream, which is will considerably
// reduce the allocations normally caused by NewReader.
func NewReader(r io.Reader, opts ...DOption) (*Decoder, error) {
initPredefined()
@@ -123,7 +123,7 @@ func NewReader(r io.Reader, opts ...DOption) (*Decoder, error) {
}
// Read bytes from the decompressed stream into p.
// Returns the number of bytes read and any error that occurred.
// Returns the number of bytes written and any error that occurred.
// When the stream is done, io.EOF will be returned.
func (d *Decoder) Read(p []byte) (int, error) {
var n int
@@ -323,7 +323,6 @@ func (d *Decoder) DecodeAll(input, dst []byte) ([]byte, error) {
frame.bBuf = nil
if frame.history.decoders.br != nil {
frame.history.decoders.br.in = nil
frame.history.decoders.br.cursor = 0
}
d.decoders <- block
}()
@@ -456,7 +455,12 @@ func (d *Decoder) nextBlock(blocking bool) (ok bool) {
}
if len(next.b) > 0 {
d.current.crc.Write(next.b)
n, err := d.current.crc.Write(next.b)
if err == nil {
if n != len(next.b) {
d.current.err = io.ErrShortWrite
}
}
}
if next.err == nil && next.d != nil && next.d.hasCRC {
got := uint32(d.current.crc.Sum64())

2
vendor/github.com/klauspost/compress/zstd/decoder_options.go generated vendored
View File

@@ -107,7 +107,7 @@ func WithDecoderDicts(dicts ...[]byte) DOption {
}
}
// WithDecoderDictRaw registers a dictionary that may be used by the decoder.
// WithEncoderDictRaw registers a dictionary that may be used by the decoder.
// The slice content can be arbitrary data.
func WithDecoderDictRaw(id uint32, content []byte) DOption {
return func(o *decoderOptions) error {

410
vendor/github.com/klauspost/compress/zstd/dict.go generated vendored
View File

@@ -1,13 +1,10 @@
package zstd
import (
"bytes"
"encoding/binary"
"errors"
"fmt"
"io"
"math"
"sort"
"github.com/klauspost/compress/huff0"
)
@@ -17,8 +14,9 @@ type dict struct {
litEnc *huff0.Scratch
llDec, ofDec, mlDec sequenceDec
offsets [3]int
content []byte
//llEnc, ofEnc, mlEnc []*fseEncoder
offsets [3]int
content []byte
}
const dictMagic = "\x37\xa4\x30\xec"
@@ -161,405 +159,3 @@ 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)
if len(block.sequences) == 0 {
continue
}
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 int(offset) >= len(o.History) {
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
if nUsed == 0 {
nUsed = 1
}
}
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)
}
// Ensure we aren't trying to represent RLE.
if maxCount == fakeLength {
for i := range hist {
if uint8(i) == maxSym {
fakeLength++
maxSym++
hist[i+1] = 1
if maxSym > 1 {
break
}
}
if hist[0] == 0 {
fakeLength++
hist[i] = 1
if maxSym > 1 {
break
}
}
}
}
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
}

3
vendor/github.com/klauspost/compress/zstd/enc_base.go generated vendored
View File

@@ -116,7 +116,7 @@ func (e *fastBase) matchlen(s, t int32, src []byte) int32 {
panic(err)
}
if t < 0 {
err := fmt.Sprintf("t (%d) < 0", t)
err := fmt.Sprintf("s (%d) < 0", s)
panic(err)
}
if s-t > e.maxMatchOff {
@@ -144,7 +144,6 @@ func (e *fastBase) resetBase(d *dict, singleBlock bool) {
} else {
e.crc.Reset()
}
e.blk.dictLitEnc = nil
if d != nil {
low := e.lowMem
if singleBlock {

328
vendor/github.com/klauspost/compress/zstd/enc_best.go generated vendored
View File

@@ -32,9 +32,10 @@ type match struct {
length int32
rep int32
est int32
_ [12]byte // Aligned size to cache line: 4+4+4+4+4 bytes + 12 bytes padding = 32 bytes
}
const highScore = maxMatchLen * 8
const highScore = 25000
// estBits will estimate output bits from predefined tables.
func (m *match) estBits(bitsPerByte int32) {
@@ -43,7 +44,7 @@ func (m *match) estBits(bitsPerByte int32) {
if m.rep < 0 {
ofc = ofCode(uint32(m.s-m.offset) + 3)
} else {
ofc = ofCode(uint32(m.rep) & 3)
ofc = ofCode(uint32(m.rep))
}
// Cost, excluding
ofTT, mlTT := fsePredefEnc[tableOffsets].ct.symbolTT[ofc], fsePredefEnc[tableMatchLengths].ct.symbolTT[mlc]
@@ -135,20 +136,8 @@ func (e *bestFastEncoder) Encode(blk *blockEnc, src []byte) {
break
}
// Add block to history
s := e.addBlock(src)
blk.size = len(src)
// Check RLE first
if len(src) > zstdMinMatch {
ml := matchLen(src[1:], src)
if ml == len(src)-1 {
blk.literals = append(blk.literals, src[0])
blk.sequences = append(blk.sequences, seq{litLen: 1, matchLen: uint32(len(src)-1) - zstdMinMatch, offset: 1 + 3})
return
}
}
if len(src) < minNonLiteralBlockSize {
blk.extraLits = len(src)
blk.literals = blk.literals[:len(src)]
@@ -171,6 +160,7 @@ 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])
@@ -184,6 +174,7 @@ 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)
@@ -198,104 +189,58 @@ encodeLoop:
panic("offset0 was 0")
}
const goodEnough = 250
cv := load6432(src, s)
bestOf := func(a, b *match) *match {
if a.est-b.est+(a.s-b.s)*bitsPerByte>>10 < 0 {
return a
}
return b
}
const goodEnough = 100
nextHashL := hashLen(cv, bestLongTableBits, bestLongLen)
nextHashS := hashLen(cv, bestShortTableBits, bestShortLen)
candidateL := e.longTable[nextHashL]
candidateS := e.table[nextHashS]
// 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) {
delta := s - offset
if delta >= e.maxMatchOff || delta <= 0 || load3232(src, offset) != first {
return
}
// 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 m.rep <= 0 {
// Extend candidate match backwards as far as possible.
// Do not extend repeats as we can assume they are optimal
// and offsets change if s == nextEmit.
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++
}
matchAt := func(offset int32, s int32, first uint32, rep int32) match {
if s-offset >= e.maxMatchOff || load3232(src, offset) != first {
return match{s: s, est: highScore}
}
if debugAsserts {
if offset >= s {
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+l], src[offset:offset+l]) {
panic(fmt.Sprintf("second match mismatch: %v != %v, first: %08x", src[s:s+4], src[offset:offset+4], first))
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: l, rep: rep}
cand.estBits(bitsPerByte)
if m.est >= highScore || cand.est-m.est+(cand.s-m.s)*bitsPerByte>>10 < 0 {
*m = cand
}
m := match{offset: offset, s: s, length: 4 + e.matchlen(s+4, offset+4, src), rep: rep}
m.estBits(bitsPerByte)
return m
}
best := match{s: s, est: highScore}
improve(&best, candidateL.offset-e.cur, s, uint32(cv), -1)
improve(&best, candidateL.prev-e.cur, s, uint32(cv), -1)
improve(&best, candidateS.offset-e.cur, s, uint32(cv), -1)
improve(&best, candidateS.prev-e.cur, s, uint32(cv), -1)
m1 := matchAt(candidateL.offset-e.cur, s, uint32(cv), -1)
m2 := matchAt(candidateL.prev-e.cur, s, uint32(cv), -1)
m3 := matchAt(candidateS.offset-e.cur, s, uint32(cv), -1)
m4 := matchAt(candidateS.prev-e.cur, s, uint32(cv), -1)
best := bestOf(bestOf(&m1, &m2), bestOf(&m3, &m4))
if canRepeat && best.length < goodEnough {
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)
}
cv32 := uint32(cv >> 8)
spp := s + 1
m1 := matchAt(spp-offset1, spp, cv32, 1)
m2 := matchAt(spp-offset2, spp, cv32, 2)
m3 := matchAt(spp-offset3, spp, cv32, 3)
best = bestOf(bestOf(best, &m1), bestOf(&m2, &m3))
if best.length > 0 {
cv32 = uint32(cv >> 24)
spp += 2
m1 := matchAt(spp-offset1, spp, cv32, 1)
m2 := matchAt(spp-offset2, spp, cv32, 2)
m3 := matchAt(spp-offset3, spp, cv32, 3)
best = bestOf(bestOf(best, &m1), bestOf(&m2, &m3))
}
}
// Load next and check...
e.longTable[nextHashL] = prevEntry{offset: s + e.cur, prev: candidateL.offset}
e.table[nextHashS] = prevEntry{offset: s + e.cur, prev: candidateS.offset}
index0 := s + 1
// Look far ahead, unless we have a really long match already...
if best.length < goodEnough {
@@ -305,89 +250,104 @@ encodeLoop:
if s >= sLimit {
break encodeLoop
}
cv = load6432(src, s)
continue
}
s++
candidateS = e.table[hashLen(cv>>8, bestShortTableBits, bestShortLen)]
cv = load6432(src, s+1)
cv2 := load6432(src, s+2)
cv = load6432(src, s)
cv2 := load6432(src, s+1)
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+1, uint32(cv), -1)
m1 := matchAt(candidateS.offset-e.cur, s, uint32(cv), -1)
// Long at s+1, s+2
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)
m2 := matchAt(candidateL.offset-e.cur, s, uint32(cv), -1)
m3 := matchAt(candidateL.prev-e.cur, s, uint32(cv), -1)
m4 := matchAt(candidateL2.offset-e.cur, s+1, uint32(cv2), -1)
m5 := matchAt(candidateL2.prev-e.cur, s+1, uint32(cv2), -1)
best = bestOf(bestOf(bestOf(best, &m1), &m2), bestOf(bestOf(&m3, &m4), &m5))
if false {
// Short at s+3.
// Too often worse...
improve(&best, e.table[hashLen(cv2>>8, bestShortTableBits, bestShortLen)].offset-e.cur, s+3, uint32(cv2>>8), -1)
m := matchAt(e.table[hashLen(cv2>>8, bestShortTableBits, bestShortLen)].offset-e.cur, s+2, uint32(cv2>>8), -1)
best = bestOf(best, &m)
}
// 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)
}
// 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 {
m := matchAt(pos, best.s+skipBeginning, load3232(src, best.s+skipBeginning), -1)
bestEnd := bestOf(best, &m)
if pos := candidateEnd.prev - e.cur - best.length + skipBeginning; pos >= 0 {
m := matchAt(pos, best.s+skipBeginning, load3232(src, best.s+skipBeginning), -1)
bestEnd = bestOf(bestEnd, &m)
}
best = bestEnd
}
}
}
if debugAsserts {
if best.offset >= best.s {
panic(fmt.Sprintf("best.offset > s: %d >= %d", best.offset, best.s))
}
if best.s < nextEmit {
panic(fmt.Sprintf("s %d < nextEmit %d", best.s, nextEmit))
}
if best.offset < s-e.maxMatchOff {
panic(fmt.Sprintf("best.offset < s-e.maxMatchOff: %d < %d", best.offset, s-e.maxMatchOff))
}
if !bytes.Equal(src[best.s:best.s+best.length], src[best.offset:best.offset+best.length]) {
panic(fmt.Sprintf("match mismatch: %v != %v", src[best.s:best.s+best.length], src[best.offset:best.offset+best.length]))
}
}
// We have a match, we can store the forward value
s = best.s
if best.rep > 0 {
s = best.s
var seq seq
seq.matchLen = uint32(best.length - zstdMinMatch)
addLiterals(&seq, best.s)
// Repeat. If bit 4 is set, this is a non-lit repeat.
seq.offset = uint32(best.rep & 3)
// 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
}
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)
// rep 0
seq.offset = uint32(best.rep)
if debugSequences {
println("repeat sequence", seq, "next s:", best.s, "off:", best.s-best.offset)
println("repeat sequence", seq, "next s:", s)
}
blk.sequences = append(blk.sequences, seq)
// Index old s + 1 -> s - 1
// Index match start+1 (long) -> s - 1
index0 := s
s = best.s + best.length
nextEmit = s
// Index skipped...
end := s
if s > sLimit+4 {
end = sLimit + 4
nextEmit = s
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, best.length)
}
break encodeLoop
}
// Index skipped...
off := index0 + e.cur
for index0 < end {
for index0 < s-1 {
cv0 := load6432(src, index0)
h0 := hashLen(cv0, bestLongTableBits, bestLongLen)
h1 := hashLen(cv0, bestShortTableBits, bestShortLen)
@@ -396,26 +356,19 @@ encodeLoop:
off++
index0++
}
switch best.rep {
case 2, 4 | 1:
case 2:
offset1, offset2 = offset2, offset1
case 3, 4 | 2:
case 3:
offset1, offset2, offset3 = offset3, offset1, offset2
case 4 | 3:
offset1, offset2, offset3 = offset1-1, offset1, offset2
}
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, best.length)
}
break encodeLoop
}
cv = load6432(src, s)
continue
}
// A 4-byte match has been found. Update recent offsets.
// We'll later see if more than 4 bytes.
s = best.s
t := best.offset
offset1, offset2, offset3 = s-t, offset1, offset2
@@ -427,9 +380,22 @@ 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 {
@@ -442,25 +408,65 @@ encodeLoop:
}
blk.sequences = append(blk.sequences, seq)
nextEmit = s
// Index old s + 1 -> s - 1 or sLimit
end := s
if s > sLimit-4 {
end = sLimit - 4
if s >= sLimit {
break encodeLoop
}
off := index0 + e.cur
for index0 < end {
// Index match start+1 (long) -> s - 1
index0 := s - l + 1
// every entry
for index0 < s-1 {
cv0 := load6432(src, index0)
h0 := hashLen(cv0, bestLongTableBits, bestLongLen)
h1 := hashLen(cv0, bestShortTableBits, bestShortLen)
off := index0 + e.cur
e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset}
e.table[h1] = prevEntry{offset: off, prev: e.table[h1].offset}
index0++
off++
}
if s >= sLimit {
break encodeLoop
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)
}
}

62
vendor/github.com/klauspost/compress/zstd/enc_better.go generated vendored
View File

@@ -102,20 +102,9 @@ func (e *betterFastEncoder) Encode(blk *blockEnc, src []byte) {
e.cur = e.maxMatchOff
break
}
// Add block to history
s := e.addBlock(src)
blk.size = len(src)
// Check RLE first
if len(src) > zstdMinMatch {
ml := matchLen(src[1:], src)
if ml == len(src)-1 {
blk.literals = append(blk.literals, src[0])
blk.sequences = append(blk.sequences, seq{litLen: 1, matchLen: uint32(len(src)-1) - zstdMinMatch, offset: 1 + 3})
return
}
}
if len(src) < minNonLiteralBlockSize {
blk.extraLits = len(src)
blk.literals = blk.literals[:len(src)]
@@ -156,7 +145,7 @@ encodeLoop:
var t int32
// We allow the encoder to optionally turn off repeat offsets across blocks
canRepeat := len(blk.sequences) > 2
var matched, index0 int32
var matched int32
for {
if debugAsserts && canRepeat && offset1 == 0 {
@@ -173,15 +162,14 @@ encodeLoop:
off := s + e.cur
e.longTable[nextHashL] = prevEntry{offset: off, prev: candidateL.offset}
e.table[nextHashS] = tableEntry{offset: off, val: uint32(cv)}
index0 = s + 1
if canRepeat {
if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) {
// Consider history as well.
var seq seq
length := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
seq.matchLen = uint32(length - zstdMinMatch)
seq.matchLen = uint32(lenght - zstdMinMatch)
// We might be able to match backwards.
// Extend as long as we can.
@@ -210,12 +198,12 @@ encodeLoop:
// Index match start+1 (long) -> s - 1
index0 := s + repOff
s += length + repOff
s += lenght + repOff
nextEmit = s
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, length)
println("repeat ended", s, lenght)
}
break encodeLoop
@@ -241,9 +229,9 @@ encodeLoop:
if false && repIndex >= 0 && load6432(src, repIndex) == load6432(src, s+repOff) {
// Consider history as well.
var seq seq
length := 8 + e.matchlen(s+8+repOff2, repIndex+8, src)
lenght := 8 + e.matchlen(s+8+repOff2, repIndex+8, src)
seq.matchLen = uint32(length - zstdMinMatch)
seq.matchLen = uint32(lenght - zstdMinMatch)
// We might be able to match backwards.
// Extend as long as we can.
@@ -270,11 +258,12 @@ encodeLoop:
}
blk.sequences = append(blk.sequences, seq)
s += length + repOff2
index0 := s + repOff2
s += lenght + repOff2
nextEmit = s
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, length)
println("repeat ended", s, lenght)
}
break encodeLoop
@@ -509,15 +498,15 @@ encodeLoop:
}
// Index match start+1 (long) -> s - 1
off := index0 + e.cur
index0 := s - l + 1
for index0 < s-1 {
cv0 := load6432(src, index0)
cv1 := cv0 >> 8
h0 := hashLen(cv0, betterLongTableBits, betterLongLen)
off := index0 + e.cur
e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset}
e.table[hashLen(cv1, betterShortTableBits, betterShortLen)] = tableEntry{offset: off + 1, val: uint32(cv1)}
index0 += 2
off += 2
}
cv = load6432(src, s)
@@ -683,7 +672,7 @@ encodeLoop:
var t int32
// We allow the encoder to optionally turn off repeat offsets across blocks
canRepeat := len(blk.sequences) > 2
var matched, index0 int32
var matched int32
for {
if debugAsserts && canRepeat && offset1 == 0 {
@@ -702,15 +691,14 @@ encodeLoop:
e.markLongShardDirty(nextHashL)
e.table[nextHashS] = tableEntry{offset: off, val: uint32(cv)}
e.markShortShardDirty(nextHashS)
index0 = s + 1
if canRepeat {
if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) {
// Consider history as well.
var seq seq
length := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
seq.matchLen = uint32(length - zstdMinMatch)
seq.matchLen = uint32(lenght - zstdMinMatch)
// We might be able to match backwards.
// Extend as long as we can.
@@ -738,12 +726,13 @@ encodeLoop:
blk.sequences = append(blk.sequences, seq)
// Index match start+1 (long) -> s - 1
s += length + repOff
index0 := s + repOff
s += lenght + repOff
nextEmit = s
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, length)
println("repeat ended", s, lenght)
}
break encodeLoop
@@ -772,9 +761,9 @@ encodeLoop:
if false && repIndex >= 0 && load6432(src, repIndex) == load6432(src, s+repOff) {
// Consider history as well.
var seq seq
length := 8 + e.matchlen(s+8+repOff2, repIndex+8, src)
lenght := 8 + e.matchlen(s+8+repOff2, repIndex+8, src)
seq.matchLen = uint32(length - zstdMinMatch)
seq.matchLen = uint32(lenght - zstdMinMatch)
// We might be able to match backwards.
// Extend as long as we can.
@@ -801,11 +790,12 @@ encodeLoop:
}
blk.sequences = append(blk.sequences, seq)
s += length + repOff2
index0 := s + repOff2
s += lenght + repOff2
nextEmit = s
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, length)
println("repeat ended", s, lenght)
}
break encodeLoop
@@ -1034,18 +1024,18 @@ encodeLoop:
}
// Index match start+1 (long) -> s - 1
off := index0 + e.cur
index0 := s - l + 1
for index0 < s-1 {
cv0 := load6432(src, index0)
cv1 := cv0 >> 8
h0 := hashLen(cv0, betterLongTableBits, betterLongLen)
off := index0 + e.cur
e.longTable[h0] = prevEntry{offset: off, prev: e.longTable[h0].offset}
e.markLongShardDirty(h0)
h1 := hashLen(cv1, betterShortTableBits, betterShortLen)
e.table[h1] = tableEntry{offset: off + 1, val: uint32(cv1)}
e.markShortShardDirty(h1)
index0 += 2
off += 2
}
cv = load6432(src, s)

18
vendor/github.com/klauspost/compress/zstd/enc_dfast.go generated vendored
View File

@@ -138,9 +138,9 @@ encodeLoop:
if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) {
// Consider history as well.
var seq seq
length := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
seq.matchLen = uint32(length - zstdMinMatch)
seq.matchLen = uint32(lenght - zstdMinMatch)
// We might be able to match backwards.
// Extend as long as we can.
@@ -166,11 +166,11 @@ encodeLoop:
println("repeat sequence", seq, "next s:", s)
}
blk.sequences = append(blk.sequences, seq)
s += length + repOff
s += lenght + repOff
nextEmit = s
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, length)
println("repeat ended", s, lenght)
}
break encodeLoop
@@ -798,9 +798,9 @@ encodeLoop:
if repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>(repOff*8)) {
// Consider history as well.
var seq seq
length := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
lenght := 4 + e.matchlen(s+4+repOff, repIndex+4, src)
seq.matchLen = uint32(length - zstdMinMatch)
seq.matchLen = uint32(lenght - zstdMinMatch)
// We might be able to match backwards.
// Extend as long as we can.
@@ -826,11 +826,11 @@ encodeLoop:
println("repeat sequence", seq, "next s:", s)
}
blk.sequences = append(blk.sequences, seq)
s += length + repOff
s += lenght + repOff
nextEmit = s
if s >= sLimit {
if debugEncoder {
println("repeat ended", s, length)
println("repeat ended", s, lenght)
}
break encodeLoop
@@ -1084,7 +1084,7 @@ func (e *doubleFastEncoderDict) Reset(d *dict, singleBlock bool) {
}
}
e.lastDictID = d.id
allDirty = true
e.allDirty = true
}
// Reset table to initial state
e.cur = e.maxMatchOff

17
vendor/github.com/klauspost/compress/zstd/enc_fast.go generated vendored
View File

@@ -133,7 +133,8 @@ encodeLoop:
if canRepeat && repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>16) {
// Consider history as well.
var seq seq
length := 4 + e.matchlen(s+6, repIndex+4, src)
var length int32
length = 4 + e.matchlen(s+6, repIndex+4, src)
seq.matchLen = uint32(length - zstdMinMatch)
// We might be able to match backwards.
@@ -644,7 +645,8 @@ encodeLoop:
if canRepeat && repIndex >= 0 && load3232(src, repIndex) == uint32(cv>>16) {
// Consider history as well.
var seq seq
length := 4 + e.matchlen(s+6, repIndex+4, src)
var length int32
length = 4 + e.matchlen(s+6, repIndex+4, src)
seq.matchLen = uint32(length - zstdMinMatch)
@@ -829,12 +831,13 @@ func (e *fastEncoderDict) Reset(d *dict, singleBlock bool) {
}
if true {
end := e.maxMatchOff + int32(len(d.content)) - 8
for i := e.maxMatchOff; i < end; i += 2 {
for i := e.maxMatchOff; i < end; i += 3 {
const hashLog = tableBits
cv := load6432(d.content, i-e.maxMatchOff)
nextHash := hashLen(cv, hashLog, tableFastHashLen) // 0 -> 6
nextHash1 := hashLen(cv>>8, hashLog, tableFastHashLen) // 1 -> 7
nextHash := hashLen(cv, hashLog, tableFastHashLen) // 0 -> 5
nextHash1 := hashLen(cv>>8, hashLog, tableFastHashLen) // 1 -> 6
nextHash2 := hashLen(cv>>16, hashLog, tableFastHashLen) // 2 -> 7
e.dictTable[nextHash] = tableEntry{
val: uint32(cv),
offset: i,
@@ -843,6 +846,10 @@ func (e *fastEncoderDict) Reset(d *dict, singleBlock bool) {
val: uint32(cv >> 8),
offset: i + 1,
}
e.dictTable[nextHash2] = tableEntry{
val: uint32(cv >> 16),
offset: i + 2,
}
}
}
e.lastDictID = d.id

136
vendor/github.com/klauspost/compress/zstd/encoder.go generated vendored
View File

@@ -6,7 +6,6 @@ package zstd
import (
"crypto/rand"
"errors"
"fmt"
"io"
"math"
@@ -150,9 +149,6 @@ func (e *Encoder) ResetContentSize(w io.Writer, size int64) {
// and write CRC if requested.
func (e *Encoder) Write(p []byte) (n int, err error) {
s := &e.state
if s.eofWritten {
return 0, ErrEncoderClosed
}
for len(p) > 0 {
if len(p)+len(s.filling) < e.o.blockSize {
if e.o.crc {
@@ -206,7 +202,7 @@ func (e *Encoder) nextBlock(final bool) error {
return nil
}
if final && len(s.filling) > 0 {
s.current = e.encodeAll(s.encoder, s.filling, s.current[:0])
s.current = e.EncodeAll(s.filling, s.current[:0])
var n2 int
n2, s.err = s.w.Write(s.current)
if s.err != nil {
@@ -231,7 +227,10 @@ func (e *Encoder) nextBlock(final bool) error {
DictID: e.o.dict.ID(),
}
dst := fh.appendTo(tmp[:0])
dst, err := fh.appendTo(tmp[:0])
if err != nil {
return err
}
s.headerWritten = true
s.wWg.Wait()
var n2 int
@@ -278,9 +277,23 @@ func (e *Encoder) nextBlock(final bool) error {
s.eofWritten = true
}
s.err = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy)
if s.err != nil {
return s.err
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 = s.w.Write(blk.output)
s.nWritten += int64(len(blk.output))
@@ -292,9 +305,6 @@ func (e *Encoder) nextBlock(final bool) error {
s.filling, s.current, s.previous = s.previous[:0], s.filling, s.current
s.nInput += int64(len(s.current))
s.wg.Add(1)
if final {
s.eofWritten = true
}
go func(src []byte) {
if debugEncoder {
println("Adding block,", len(src), "bytes, final:", final)
@@ -310,6 +320,9 @@ func (e *Encoder) nextBlock(final bool) error {
blk := enc.Block()
enc.Encode(blk, src)
blk.last = final
if final {
s.eofWritten = true
}
// Wait for pending writes.
s.wWg.Wait()
if s.writeErr != nil {
@@ -330,8 +343,22 @@ func (e *Encoder) nextBlock(final bool) error {
}
s.wWg.Done()
}()
s.writeErr = blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy)
if s.writeErr != nil {
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
return
}
_, s.writeErr = s.w.Write(blk.output)
@@ -405,20 +432,12 @@ func (e *Encoder) Flush() error {
if len(s.filling) > 0 {
err := e.nextBlock(false)
if err != nil {
// Ignore Flush after Close.
if errors.Is(s.err, ErrEncoderClosed) {
return nil
}
return err
}
}
s.wg.Wait()
s.wWg.Wait()
if s.err != nil {
// Ignore Flush after Close.
if errors.Is(s.err, ErrEncoderClosed) {
return nil
}
return s.err
}
return s.writeErr
@@ -434,9 +453,6 @@ func (e *Encoder) Close() error {
}
err := e.nextBlock(true)
if err != nil {
if errors.Is(s.err, ErrEncoderClosed) {
return nil
}
return err
}
if s.frameContentSize > 0 {
@@ -474,11 +490,6 @@ func (e *Encoder) Close() error {
}
_, s.err = s.w.Write(frame)
}
if s.err == nil {
s.err = ErrEncoderClosed
return nil
}
return s.err
}
@@ -489,15 +500,6 @@ func (e *Encoder) Close() error {
// Data compressed with EncodeAll can be decoded with the Decoder,
// using either a stream or DecodeAll.
func (e *Encoder) EncodeAll(src, dst []byte) []byte {
e.init.Do(e.initialize)
enc := <-e.encoders
defer func() {
e.encoders <- enc
}()
return e.encodeAll(enc, src, dst)
}
func (e *Encoder) encodeAll(enc encoder, src, dst []byte) []byte {
if len(src) == 0 {
if e.o.fullZero {
// Add frame header.
@@ -509,7 +511,7 @@ func (e *Encoder) encodeAll(enc encoder, 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
@@ -520,7 +522,13 @@ func (e *Encoder) encodeAll(enc encoder, src, dst []byte) []byte {
}
return dst
}
e.init.Do(e.initialize)
enc := <-e.encoders
defer func() {
// Release encoder reference to last block.
// If a non-single block is needed the encoder will reset again.
e.encoders <- enc
}()
// Use single segments when above minimum window and below window size.
single := len(src) <= e.o.windowSize && len(src) > MinWindowSize
if e.o.single != nil {
@@ -538,7 +546,10 @@ func (e *Encoder) encodeAll(enc encoder, 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 = fh.appendTo(dst)
dst, err := fh.appendTo(dst)
if err != nil {
panic(err)
}
// If we can do everything in one block, prefer that.
if len(src) <= e.o.blockSize {
@@ -557,15 +568,25 @@ func (e *Encoder) encodeAll(enc encoder, 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
// Output directly to dst
blk.output = dst
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)
}
err := blk.encode(src, e.o.noEntropy, !e.o.allLitEntropy)
if err != nil {
switch err {
case errIncompressible:
if debugEncoder {
println("Storing incompressible block as raw")
}
dst = blk.encodeRawTo(dst, src)
case nil:
dst = blk.output
default:
panic(err)
}
dst = blk.output
blk.output = oldout
} else {
enc.Reset(e.o.dict, false)
@@ -584,11 +605,25 @@ func (e *Encoder) encodeAll(enc encoder, src, dst []byte) []byte {
if len(src) == 0 {
blk.last = true
}
err := blk.encode(todo, e.o.noEntropy, !e.o.allLitEntropy)
if err != nil {
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:
panic(err)
}
dst = append(dst, blk.output...)
blk.reset(nil)
}
}
@@ -598,7 +633,6 @@ func (e *Encoder) encodeAll(enc encoder, 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)

12
vendor/github.com/klauspost/compress/zstd/encoder_options.go generated vendored
View File

@@ -39,7 +39,7 @@ func (o *encoderOptions) setDefault() {
blockSize: maxCompressedBlockSize,
windowSize: 8 << 20,
level: SpeedDefault,
allLitEntropy: false,
allLitEntropy: true,
lowMem: false,
}
}
@@ -94,7 +94,7 @@ func WithEncoderConcurrency(n int) EOption {
// The value must be a power of two between MinWindowSize and MaxWindowSize.
// A larger value will enable better compression but allocate more memory and,
// for above-default values, take considerably longer.
// The default value is determined by the compression level and max 8MB.
// The default value is determined by the compression level.
func WithWindowSize(n int) EOption {
return func(o *encoderOptions) error {
switch {
@@ -129,7 +129,7 @@ func WithEncoderPadding(n int) EOption {
}
// No need to waste our time.
if n == 1 {
n = 0
o.pad = 0
}
if n > 1<<30 {
return fmt.Errorf("padding must less than 1GB (1<<30 bytes) ")
@@ -232,13 +232,13 @@ func WithEncoderLevel(l EncoderLevel) EOption {
case SpeedDefault:
o.windowSize = 8 << 20
case SpeedBetterCompression:
o.windowSize = 8 << 20
o.windowSize = 16 << 20
case SpeedBestCompression:
o.windowSize = 8 << 20
o.windowSize = 32 << 20
}
}
if !o.customALEntropy {
o.allLitEntropy = l > SpeedDefault
o.allLitEntropy = l > SpeedFastest
}
return nil

47
vendor/github.com/klauspost/compress/zstd/framedec.go generated vendored
View File

@@ -73,20 +73,20 @@ func (d *frameDec) reset(br byteBuffer) error {
switch err {
case io.EOF, io.ErrUnexpectedEOF:
return io.EOF
case nil:
signature[0] = b[0]
default:
return err
case nil:
signature[0] = b[0]
}
// Read the rest, don't allow io.ErrUnexpectedEOF
b, err = br.readSmall(3)
switch err {
case io.EOF:
return io.EOF
case nil:
copy(signature[1:], b)
default:
return err
case nil:
copy(signature[1:], b)
}
if string(signature[1:4]) != skippableFrameMagic || signature[0]&0xf0 != 0x50 {
@@ -146,9 +146,7 @@ func (d *frameDec) reset(br byteBuffer) error {
}
return err
}
if debugDecoder {
printf("raw: %x, mantissa: %d, exponent: %d\n", wd, wd&7, wd>>3)
}
printf("raw: %x, mantissa: %d, exponent: %d\n", wd, wd&7, wd>>3)
windowLog := 10 + (wd >> 3)
windowBase := uint64(1) << windowLog
windowAdd := (windowBase / 8) * uint64(wd&0x7)
@@ -295,9 +293,13 @@ func (d *frameDec) next(block *blockDec) error {
return nil
}
// checkCRC will check the checksum, assuming the frame has one.
// checkCRC will check the checksum if 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 {
@@ -305,6 +307,10 @@ func (d *frameDec) checkCRC() error {
return err
}
if d.o.ignoreChecksum {
return nil
}
want := binary.LittleEndian.Uint32(buf[:4])
got := uint32(d.crc.Sum64())
@@ -320,13 +326,17 @@ func (d *frameDec) checkCRC() error {
return nil
}
// consumeCRC skips over the checksum, assuming the frame has one.
// consumeCRC reads the checksum data if the frame has one.
func (d *frameDec) consumeCRC() error {
_, err := d.rawInput.readSmall(4)
if err != nil {
println("CRC missing?", err)
if d.HasCheckSum {
_, err := d.rawInput.readSmall(4)
if err != nil {
println("CRC missing?", err)
return err
}
}
return err
return nil
}
// runDecoder will run the decoder for the remainder of the frame.
@@ -405,8 +415,15 @@ func (d *frameDec) runDecoder(dst []byte, dec *blockDec) ([]byte, error) {
if d.o.ignoreChecksum {
err = d.consumeCRC()
} else {
d.crc.Write(dst[crcStart:])
err = d.checkCRC()
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()
}
}
}
}
}

6
vendor/github.com/klauspost/compress/zstd/frameenc.go generated vendored
View File

@@ -22,7 +22,7 @@ type frameHeader struct {
const maxHeaderSize = 14
func (f frameHeader) appendTo(dst []byte) []byte {
func (f frameHeader) appendTo(dst []byte) ([]byte, error) {
dst = append(dst, frameMagic...)
var fhd uint8
if f.Checksum {
@@ -76,7 +76,7 @@ func (f frameHeader) appendTo(dst []byte) []byte {
if f.SingleSegment {
dst = append(dst, uint8(f.ContentSize))
}
// Unless SingleSegment is set, framessizes < 256 are not stored.
// Unless SingleSegment is set, framessizes < 256 are nto stored.
case 1:
f.ContentSize -= 256
dst = append(dst, uint8(f.ContentSize), uint8(f.ContentSize>>8))
@@ -88,7 +88,7 @@ func (f frameHeader) appendTo(dst []byte) []byte {
default:
panic("invalid fcs")
}
return dst
return dst, nil
}
const skippableFrameHeader = 4 + 4

11
vendor/github.com/klauspost/compress/zstd/fse_decoder_generic.go generated vendored
View File

@@ -20,9 +20,10 @@ func (s *fseDecoder) buildDtable() error {
if v == -1 {
s.dt[highThreshold].setAddBits(uint8(i))
highThreshold--
v = 1
symbolNext[i] = 1
} else {
symbolNext[i] = uint16(v)
}
symbolNext[i] = uint16(v)
}
}
@@ -34,12 +35,10 @@ func (s *fseDecoder) buildDtable() error {
for ss, v := range s.norm[:s.symbolLen] {
for i := 0; i < int(v); i++ {
s.dt[position].setAddBits(uint8(ss))
for {
position = (position + step) & tableMask
for position > highThreshold {
// lowprob area
position = (position + step) & tableMask
if position <= highThreshold {
break
}
}
}
}

4
vendor/github.com/klauspost/compress/zstd/internal/xxhash/xxhash_arm64.s generated vendored
View File

@@ -162,12 +162,12 @@ finalize:
MOVD h, ret+24(FP)
RET
// func writeBlocks(s *Digest, b []byte) int
// func writeBlocks(d *Digest, b []byte) int
TEXT ·writeBlocks(SB), NOSPLIT|NOFRAME, $0-40
LDP ·primes+0(SB), (prime1, prime2)
// Load state. Assume v[1-4] are stored contiguously.
MOVD s+0(FP), digest
MOVD d+0(FP), digest
LDP 0(digest), (v1, v2)
LDP 16(digest), (v3, v4)

16
vendor/github.com/klauspost/compress/zstd/matchlen_amd64.go generated vendored
View File

@@ -1,16 +0,0 @@
//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 zstd
// 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

66
vendor/github.com/klauspost/compress/zstd/matchlen_amd64.s generated vendored
View File

@@ -1,66 +0,0 @@
// Copied from S2 implementation.
//go:build !appengine && !noasm && gc && !noasm
#include "textflag.h"
// func matchLen(a []byte, b []byte) int
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
JZ matchlen_loop_standalone
#ifdef GOAMD64_v3
TZCNTQ BX, BX
#else
BSFQ BX, BX
#endif
SHRL $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

38
vendor/github.com/klauspost/compress/zstd/matchlen_generic.go generated vendored
View File

@@ -1,38 +0,0 @@
//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 zstd
import (
"math/bits"
"github.com/klauspost/compress/internal/le"
)
// 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) {
left := len(a)
for left >= 8 {
diff := le.Load64(a, n) ^ le.Load64(b, n)
if diff != 0 {
return n + bits.TrailingZeros64(diff)>>3
}
n += 8
left -= 8
}
a = a[n:]
b = b[n:]
for i := range a {
if a[i] != b[i] {
break
}
n++
}
return n
}

28
vendor/github.com/klauspost/compress/zstd/seqdec.go generated vendored
View File

@@ -236,16 +236,13 @@ 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. Overread by %d\n", seqs-i, -br.remain())
printf("reading sequence %d, exceeded available data\n", seqs-i)
return io.ErrUnexpectedEOF
}
var ll, mo, ml int
if br.cursor > 4+((maxOffsetBits+16+16)>>3) {
if br.off > 4+((maxOffsetBits+16+16)>>3) {
// inlined function:
// ll, mo, ml = s.nextFast(br, llState, mlState, ofState)
@@ -317,6 +314,9 @@ func (s *sequenceDecs) decodeSync(hist []byte) error {
}
size := ll + ml + len(out)
if size-startSize > maxBlockSize {
if size-startSize == 424242 {
panic("here")
}
return fmt.Errorf("output bigger than max block size (%d)", maxBlockSize)
}
if size > cap(out) {
@@ -427,7 +427,8 @@ func (s *sequenceDecs) decodeSync(hist []byte) error {
}
}
if size := len(s.literals) + len(out) - startSize; size > maxBlockSize {
// Check if space for literals
if size := len(s.literals) + len(s.out) - startSize; size > maxBlockSize {
return fmt.Errorf("output bigger than max block size (%d)", maxBlockSize)
}
@@ -452,13 +453,18 @@ func (s *sequenceDecs) next(br *bitReader, llState, mlState, ofState decSymbol)
// extra bits are stored in reverse order.
br.fill()
mo += br.getBits(moB)
if s.maxBits > 32 {
if s.maxBits <= 32 {
mo += br.getBits(moB)
ml += br.getBits(mlB)
ll += br.getBits(llB)
} else {
mo += br.getBits(moB)
br.fill()
// matchlength+literal length, max 32 bits
ml += br.getBits(mlB)
ll += br.getBits(llB)
}
// matchlength+literal length, max 32 bits
ml += br.getBits(mlB)
ll += br.getBits(llB)
mo = s.adjustOffset(mo, ll, moB)
return
}

21
vendor/github.com/klauspost/compress/zstd/seqdec_amd64.go generated vendored
View File

@@ -5,7 +5,6 @@ package zstd
import (
"fmt"
"io"
"github.com/klauspost/compress/internal/cpuinfo"
)
@@ -135,9 +134,6 @@ 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 {
@@ -146,12 +142,13 @@ func (s *sequenceDecs) decodeSyncSimple(hist []byte) (bool, error) {
return true, fmt.Errorf("output bigger than max block size (%d)", maxBlockSize)
default:
return true, fmt.Errorf("sequenceDecs_decode returned erroneous code %d", errCode)
return true, fmt.Errorf("sequenceDecs_decode returned erronous code %d", errCode)
}
s.seqSize += ctx.litRemain
if s.seqSize > maxBlockSize {
return true, fmt.Errorf("output bigger than max block size (%d)", maxBlockSize)
}
err := br.close()
if err != nil {
@@ -206,9 +203,6 @@ 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.
@@ -254,10 +248,6 @@ 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
@@ -288,11 +278,9 @@ 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 erroneous code %d", errCode)
return fmt.Errorf("sequenceDecs_decode_amd64 returned erronous code %d", errCode)
}
if ctx.litRemain < 0 {
@@ -304,9 +292,6 @@ 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)

380
vendor/github.com/klauspost/compress/zstd/seqdec_amd64.s generated vendored
View File

@@ -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), CX
MOVQ 24(CX), DX
MOVBQZX 40(CX), BX
MOVQ (CX), AX
MOVQ 32(CX), SI
MOVQ br+8(FP), AX
MOVQ 32(AX), DX
MOVBQZX 40(AX), BX
MOVQ 24(AX), SI
MOVQ (AX), AX
ADDQ SI, AX
MOVQ AX, (SP)
MOVQ ctx+16(FP), AX
@@ -38,7 +38,7 @@ sequenceDecs_decode_amd64_main_loop:
sequenceDecs_decode_amd64_fill_byte_by_byte:
CMPQ SI, $0x00
JLE sequenceDecs_decode_amd64_fill_check_overread
JLE sequenceDecs_decode_amd64_fill_end
CMPQ BX, $0x07
JLE sequenceDecs_decode_amd64_fill_end
SHLQ $0x08, DX
@@ -49,10 +49,6 @@ 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
@@ -109,7 +105,7 @@ sequenceDecs_decode_amd64_ml_update_zero:
sequenceDecs_decode_amd64_fill_2_byte_by_byte:
CMPQ SI, $0x00
JLE sequenceDecs_decode_amd64_fill_2_check_overread
JLE sequenceDecs_decode_amd64_fill_2_end
CMPQ BX, $0x07
JLE sequenceDecs_decode_amd64_fill_2_end
SHLQ $0x08, DX
@@ -120,10 +116,6 @@ 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
@@ -157,7 +149,8 @@ sequenceDecs_decode_amd64_ll_update_zero:
// Update Literal Length State
MOVBQZX DI, R14
SHRL $0x10, DI
SHRQ $0x10, DI
MOVWQZX DI, DI
LEAQ (BX)(R14*1), CX
MOVQ DX, R15
MOVQ CX, BX
@@ -176,7 +169,8 @@ sequenceDecs_decode_amd64_ll_update_zero:
// Update Match Length State
MOVBQZX R8, R14
SHRL $0x10, R8
SHRQ $0x10, R8
MOVWQZX R8, R8
LEAQ (BX)(R14*1), CX
MOVQ DX, R15
MOVQ CX, BX
@@ -195,7 +189,8 @@ sequenceDecs_decode_amd64_ll_update_zero:
// Update Offset State
MOVBQZX R9, R14
SHRL $0x10, R9
SHRQ $0x10, R9
MOVWQZX R9, R9
LEAQ (BX)(R14*1), CX
MOVQ DX, R15
MOVQ CX, BX
@@ -298,9 +293,9 @@ sequenceDecs_decode_amd64_match_len_ofs_ok:
MOVQ R12, 152(AX)
MOVQ R13, 160(AX)
MOVQ br+8(FP), AX
MOVQ DX, 24(AX)
MOVQ DX, 32(AX)
MOVB BL, 40(AX)
MOVQ SI, 32(AX)
MOVQ SI, 24(AX)
// Return success
MOVQ $0x00000000, ret+24(FP)
@@ -325,19 +320,14 @@ 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
MOVQ br+8(FP), CX
MOVQ 24(CX), DX
MOVBQZX 40(CX), BX
MOVQ (CX), AX
MOVQ 32(CX), SI
MOVQ br+8(FP), AX
MOVQ 32(AX), DX
MOVBQZX 40(AX), BX
MOVQ 24(AX), SI
MOVQ (AX), AX
ADDQ SI, AX
MOVQ AX, (SP)
MOVQ ctx+16(FP), AX
@@ -366,7 +356,7 @@ sequenceDecs_decode_56_amd64_main_loop:
sequenceDecs_decode_56_amd64_fill_byte_by_byte:
CMPQ SI, $0x00
JLE sequenceDecs_decode_56_amd64_fill_check_overread
JLE sequenceDecs_decode_56_amd64_fill_end
CMPQ BX, $0x07
JLE sequenceDecs_decode_56_amd64_fill_end
SHLQ $0x08, DX
@@ -377,10 +367,6 @@ 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
@@ -456,7 +442,8 @@ sequenceDecs_decode_56_amd64_ll_update_zero:
// Update Literal Length State
MOVBQZX DI, R14
SHRL $0x10, DI
SHRQ $0x10, DI
MOVWQZX DI, DI
LEAQ (BX)(R14*1), CX
MOVQ DX, R15
MOVQ CX, BX
@@ -475,7 +462,8 @@ sequenceDecs_decode_56_amd64_ll_update_zero:
// Update Match Length State
MOVBQZX R8, R14
SHRL $0x10, R8
SHRQ $0x10, R8
MOVWQZX R8, R8
LEAQ (BX)(R14*1), CX
MOVQ DX, R15
MOVQ CX, BX
@@ -494,7 +482,8 @@ sequenceDecs_decode_56_amd64_ll_update_zero:
// Update Offset State
MOVBQZX R9, R14
SHRL $0x10, R9
SHRQ $0x10, R9
MOVWQZX R9, R9
LEAQ (BX)(R14*1), CX
MOVQ DX, R15
MOVQ CX, BX
@@ -597,9 +586,9 @@ sequenceDecs_decode_56_amd64_match_len_ofs_ok:
MOVQ R12, 152(AX)
MOVQ R13, 160(AX)
MOVQ br+8(FP), AX
MOVQ DX, 24(AX)
MOVQ DX, 32(AX)
MOVB BL, 40(AX)
MOVQ SI, 32(AX)
MOVQ SI, 24(AX)
// Return success
MOVQ $0x00000000, ret+24(FP)
@@ -624,19 +613,14 @@ 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
MOVQ br+8(FP), BX
MOVQ 24(BX), AX
MOVBQZX 40(BX), DX
MOVQ (BX), CX
MOVQ 32(BX), BX
MOVQ br+8(FP), CX
MOVQ 32(CX), AX
MOVBQZX 40(CX), DX
MOVQ 24(CX), BX
MOVQ (CX), CX
ADDQ BX, CX
MOVQ CX, (SP)
MOVQ ctx+16(FP), CX
@@ -665,7 +649,7 @@ sequenceDecs_decode_bmi2_main_loop:
sequenceDecs_decode_bmi2_fill_byte_by_byte:
CMPQ BX, $0x00
JLE sequenceDecs_decode_bmi2_fill_check_overread
JLE sequenceDecs_decode_bmi2_fill_end
CMPQ DX, $0x07
JLE sequenceDecs_decode_bmi2_fill_end
SHLQ $0x08, AX
@@ -676,10 +660,6 @@ 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
@@ -720,7 +700,7 @@ sequenceDecs_decode_bmi2_fill_end:
sequenceDecs_decode_bmi2_fill_2_byte_by_byte:
CMPQ BX, $0x00
JLE sequenceDecs_decode_bmi2_fill_2_check_overread
JLE sequenceDecs_decode_bmi2_fill_2_end
CMPQ DX, $0x07
JLE sequenceDecs_decode_bmi2_fill_2_end
SHLQ $0x08, AX
@@ -731,10 +711,6 @@ 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
@@ -766,10 +742,11 @@ sequenceDecs_decode_bmi2_fill_2_end:
BZHIQ R14, R15, R15
// Update Offset State
BZHIQ R8, R15, CX
SHRXQ R8, R15, R15
SHRL $0x10, R8
ADDQ CX, R8
BZHIQ R8, R15, CX
SHRXQ R8, R15, R15
MOVQ $0x00001010, R14
BEXTRQ R14, R8, R8
ADDQ CX, R8
// Load ctx.ofTable
MOVQ ctx+16(FP), CX
@@ -777,10 +754,11 @@ sequenceDecs_decode_bmi2_fill_2_end:
MOVQ (CX)(R8*8), R8
// Update Match Length State
BZHIQ DI, R15, CX
SHRXQ DI, R15, R15
SHRL $0x10, DI
ADDQ CX, DI
BZHIQ DI, R15, CX
SHRXQ DI, R15, R15
MOVQ $0x00001010, R14
BEXTRQ R14, DI, DI
ADDQ CX, DI
// Load ctx.mlTable
MOVQ ctx+16(FP), CX
@@ -788,9 +766,10 @@ sequenceDecs_decode_bmi2_fill_2_end:
MOVQ (CX)(DI*8), DI
// Update Literal Length State
BZHIQ SI, R15, CX
SHRL $0x10, SI
ADDQ CX, SI
BZHIQ SI, R15, CX
MOVQ $0x00001010, R14
BEXTRQ R14, SI, SI
ADDQ CX, SI
// Load ctx.llTable
MOVQ ctx+16(FP), CX
@@ -883,9 +862,9 @@ sequenceDecs_decode_bmi2_match_len_ofs_ok:
MOVQ R11, 152(CX)
MOVQ R12, 160(CX)
MOVQ br+8(FP), CX
MOVQ AX, 24(CX)
MOVQ AX, 32(CX)
MOVB DL, 40(CX)
MOVQ BX, 32(CX)
MOVQ BX, 24(CX)
// Return success
MOVQ $0x00000000, ret+24(FP)
@@ -910,19 +889,14 @@ 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
MOVQ br+8(FP), BX
MOVQ 24(BX), AX
MOVBQZX 40(BX), DX
MOVQ (BX), CX
MOVQ 32(BX), BX
MOVQ br+8(FP), CX
MOVQ 32(CX), AX
MOVBQZX 40(CX), DX
MOVQ 24(CX), BX
MOVQ (CX), CX
ADDQ BX, CX
MOVQ CX, (SP)
MOVQ ctx+16(FP), CX
@@ -951,7 +925,7 @@ sequenceDecs_decode_56_bmi2_main_loop:
sequenceDecs_decode_56_bmi2_fill_byte_by_byte:
CMPQ BX, $0x00
JLE sequenceDecs_decode_56_bmi2_fill_check_overread
JLE sequenceDecs_decode_56_bmi2_fill_end
CMPQ DX, $0x07
JLE sequenceDecs_decode_56_bmi2_fill_end
SHLQ $0x08, AX
@@ -962,10 +936,6 @@ 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
@@ -1023,10 +993,11 @@ sequenceDecs_decode_56_bmi2_fill_end:
BZHIQ R14, R15, R15
// Update Offset State
BZHIQ R8, R15, CX
SHRXQ R8, R15, R15
SHRL $0x10, R8
ADDQ CX, R8
BZHIQ R8, R15, CX
SHRXQ R8, R15, R15
MOVQ $0x00001010, R14
BEXTRQ R14, R8, R8
ADDQ CX, R8
// Load ctx.ofTable
MOVQ ctx+16(FP), CX
@@ -1034,10 +1005,11 @@ sequenceDecs_decode_56_bmi2_fill_end:
MOVQ (CX)(R8*8), R8
// Update Match Length State
BZHIQ DI, R15, CX
SHRXQ DI, R15, R15
SHRL $0x10, DI
ADDQ CX, DI
BZHIQ DI, R15, CX
SHRXQ DI, R15, R15
MOVQ $0x00001010, R14
BEXTRQ R14, DI, DI
ADDQ CX, DI
// Load ctx.mlTable
MOVQ ctx+16(FP), CX
@@ -1045,9 +1017,10 @@ sequenceDecs_decode_56_bmi2_fill_end:
MOVQ (CX)(DI*8), DI
// Update Literal Length State
BZHIQ SI, R15, CX
SHRL $0x10, SI
ADDQ CX, SI
BZHIQ SI, R15, CX
MOVQ $0x00001010, R14
BEXTRQ R14, SI, SI
ADDQ CX, SI
// Load ctx.llTable
MOVQ ctx+16(FP), CX
@@ -1140,9 +1113,9 @@ sequenceDecs_decode_56_bmi2_match_len_ofs_ok:
MOVQ R11, 152(CX)
MOVQ R12, 160(CX)
MOVQ br+8(FP), CX
MOVQ AX, 24(CX)
MOVQ AX, 32(CX)
MOVB DL, 40(CX)
MOVQ BX, 32(CX)
MOVQ BX, 24(CX)
// Return success
MOVQ $0x00000000, ret+24(FP)
@@ -1167,11 +1140,6 @@ 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
@@ -1785,11 +1753,11 @@ 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), CX
MOVQ 24(CX), DX
MOVBQZX 40(CX), BX
MOVQ (CX), AX
MOVQ 32(CX), SI
MOVQ br+8(FP), AX
MOVQ 32(AX), DX
MOVBQZX 40(AX), BX
MOVQ 24(AX), SI
MOVQ (AX), AX
ADDQ SI, AX
MOVQ AX, (SP)
MOVQ ctx+16(FP), AX
@@ -1814,7 +1782,7 @@ TEXT ·sequenceDecs_decodeSync_amd64(SB), $64-32
MOVQ 40(SP), AX
ADDQ AX, 48(SP)
// Calculate pointer to s.out[cap(s.out)] (a past-end pointer)
// Calculate poiter to s.out[cap(s.out)] (a past-end pointer)
ADDQ R10, 32(SP)
// outBase += outPosition
@@ -1836,7 +1804,7 @@ sequenceDecs_decodeSync_amd64_main_loop:
sequenceDecs_decodeSync_amd64_fill_byte_by_byte:
CMPQ SI, $0x00
JLE sequenceDecs_decodeSync_amd64_fill_check_overread
JLE sequenceDecs_decodeSync_amd64_fill_end
CMPQ BX, $0x07
JLE sequenceDecs_decodeSync_amd64_fill_end
SHLQ $0x08, DX
@@ -1847,10 +1815,6 @@ 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
@@ -1907,7 +1871,7 @@ sequenceDecs_decodeSync_amd64_ml_update_zero:
sequenceDecs_decodeSync_amd64_fill_2_byte_by_byte:
CMPQ SI, $0x00
JLE sequenceDecs_decodeSync_amd64_fill_2_check_overread
JLE sequenceDecs_decodeSync_amd64_fill_2_end
CMPQ BX, $0x07
JLE sequenceDecs_decodeSync_amd64_fill_2_end
SHLQ $0x08, DX
@@ -1918,10 +1882,6 @@ 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
@@ -1955,7 +1915,8 @@ sequenceDecs_decodeSync_amd64_ll_update_zero:
// Update Literal Length State
MOVBQZX DI, R13
SHRL $0x10, DI
SHRQ $0x10, DI
MOVWQZX DI, DI
LEAQ (BX)(R13*1), CX
MOVQ DX, R14
MOVQ CX, BX
@@ -1974,7 +1935,8 @@ sequenceDecs_decodeSync_amd64_ll_update_zero:
// Update Match Length State
MOVBQZX R8, R13
SHRL $0x10, R8
SHRQ $0x10, R8
MOVWQZX R8, R8
LEAQ (BX)(R13*1), CX
MOVQ DX, R14
MOVQ CX, BX
@@ -1993,7 +1955,8 @@ sequenceDecs_decodeSync_amd64_ll_update_zero:
// Update Offset State
MOVBQZX R9, R13
SHRL $0x10, R9
SHRQ $0x10, R9
MOVWQZX R9, R9
LEAQ (BX)(R13*1), CX
MOVQ DX, R14
MOVQ CX, BX
@@ -2280,9 +2243,9 @@ handle_loop:
loop_finished:
MOVQ br+8(FP), AX
MOVQ DX, 24(AX)
MOVQ DX, 32(AX)
MOVB BL, 40(AX)
MOVQ SI, 32(AX)
MOVQ SI, 24(AX)
// Update the context
MOVQ ctx+16(FP), AX
@@ -2328,11 +2291,6 @@ 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
@@ -2347,11 +2305,11 @@ 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), BX
MOVQ 24(BX), AX
MOVBQZX 40(BX), DX
MOVQ (BX), CX
MOVQ 32(BX), BX
MOVQ br+8(FP), CX
MOVQ 32(CX), AX
MOVBQZX 40(CX), DX
MOVQ 24(CX), BX
MOVQ (CX), CX
ADDQ BX, CX
MOVQ CX, (SP)
MOVQ ctx+16(FP), CX
@@ -2376,7 +2334,7 @@ TEXT ·sequenceDecs_decodeSync_bmi2(SB), $64-32
MOVQ 40(SP), CX
ADDQ CX, 48(SP)
// Calculate pointer to s.out[cap(s.out)] (a past-end pointer)
// Calculate poiter to s.out[cap(s.out)] (a past-end pointer)
ADDQ R9, 32(SP)
// outBase += outPosition
@@ -2398,7 +2356,7 @@ sequenceDecs_decodeSync_bmi2_main_loop:
sequenceDecs_decodeSync_bmi2_fill_byte_by_byte:
CMPQ BX, $0x00
JLE sequenceDecs_decodeSync_bmi2_fill_check_overread
JLE sequenceDecs_decodeSync_bmi2_fill_end
CMPQ DX, $0x07
JLE sequenceDecs_decodeSync_bmi2_fill_end
SHLQ $0x08, AX
@@ -2409,10 +2367,6 @@ 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
@@ -2453,7 +2407,7 @@ sequenceDecs_decodeSync_bmi2_fill_end:
sequenceDecs_decodeSync_bmi2_fill_2_byte_by_byte:
CMPQ BX, $0x00
JLE sequenceDecs_decodeSync_bmi2_fill_2_check_overread
JLE sequenceDecs_decodeSync_bmi2_fill_2_end
CMPQ DX, $0x07
JLE sequenceDecs_decodeSync_bmi2_fill_2_end
SHLQ $0x08, AX
@@ -2464,10 +2418,6 @@ 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
@@ -2499,10 +2449,11 @@ sequenceDecs_decodeSync_bmi2_fill_2_end:
BZHIQ R13, R14, R14
// Update Offset State
BZHIQ R8, R14, CX
SHRXQ R8, R14, R14
SHRL $0x10, R8
ADDQ CX, R8
BZHIQ R8, R14, CX
SHRXQ R8, R14, R14
MOVQ $0x00001010, R13
BEXTRQ R13, R8, R8
ADDQ CX, R8
// Load ctx.ofTable
MOVQ ctx+16(FP), CX
@@ -2510,10 +2461,11 @@ sequenceDecs_decodeSync_bmi2_fill_2_end:
MOVQ (CX)(R8*8), R8
// Update Match Length State
BZHIQ DI, R14, CX
SHRXQ DI, R14, R14
SHRL $0x10, DI
ADDQ CX, DI
BZHIQ DI, R14, CX
SHRXQ DI, R14, R14
MOVQ $0x00001010, R13
BEXTRQ R13, DI, DI
ADDQ CX, DI
// Load ctx.mlTable
MOVQ ctx+16(FP), CX
@@ -2521,9 +2473,10 @@ sequenceDecs_decodeSync_bmi2_fill_2_end:
MOVQ (CX)(DI*8), DI
// Update Literal Length State
BZHIQ SI, R14, CX
SHRL $0x10, SI
ADDQ CX, SI
BZHIQ SI, R14, CX
MOVQ $0x00001010, R13
BEXTRQ R13, SI, SI
ADDQ CX, SI
// Load ctx.llTable
MOVQ ctx+16(FP), CX
@@ -2800,9 +2753,9 @@ handle_loop:
loop_finished:
MOVQ br+8(FP), CX
MOVQ AX, 24(CX)
MOVQ AX, 32(CX)
MOVB DL, 40(CX)
MOVQ BX, 32(CX)
MOVQ BX, 24(CX)
// Update the context
MOVQ ctx+16(FP), AX
@@ -2848,11 +2801,6 @@ 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
@@ -2867,11 +2815,11 @@ 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), CX
MOVQ 24(CX), DX
MOVBQZX 40(CX), BX
MOVQ (CX), AX
MOVQ 32(CX), SI
MOVQ br+8(FP), AX
MOVQ 32(AX), DX
MOVBQZX 40(AX), BX
MOVQ 24(AX), SI
MOVQ (AX), AX
ADDQ SI, AX
MOVQ AX, (SP)
MOVQ ctx+16(FP), AX
@@ -2896,7 +2844,7 @@ TEXT ·sequenceDecs_decodeSync_safe_amd64(SB), $64-32
MOVQ 40(SP), AX
ADDQ AX, 48(SP)
// Calculate pointer to s.out[cap(s.out)] (a past-end pointer)
// Calculate poiter to s.out[cap(s.out)] (a past-end pointer)
ADDQ R10, 32(SP)
// outBase += outPosition
@@ -2918,7 +2866,7 @@ sequenceDecs_decodeSync_safe_amd64_main_loop:
sequenceDecs_decodeSync_safe_amd64_fill_byte_by_byte:
CMPQ SI, $0x00
JLE sequenceDecs_decodeSync_safe_amd64_fill_check_overread
JLE sequenceDecs_decodeSync_safe_amd64_fill_end
CMPQ BX, $0x07
JLE sequenceDecs_decodeSync_safe_amd64_fill_end
SHLQ $0x08, DX
@@ -2929,10 +2877,6 @@ 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
@@ -2989,7 +2933,7 @@ 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_check_overread
JLE sequenceDecs_decodeSync_safe_amd64_fill_2_end
CMPQ BX, $0x07
JLE sequenceDecs_decodeSync_safe_amd64_fill_2_end
SHLQ $0x08, DX
@@ -3000,10 +2944,6 @@ 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
@@ -3037,7 +2977,8 @@ sequenceDecs_decodeSync_safe_amd64_ll_update_zero:
// Update Literal Length State
MOVBQZX DI, R13
SHRL $0x10, DI
SHRQ $0x10, DI
MOVWQZX DI, DI
LEAQ (BX)(R13*1), CX
MOVQ DX, R14
MOVQ CX, BX
@@ -3056,7 +2997,8 @@ sequenceDecs_decodeSync_safe_amd64_ll_update_zero:
// Update Match Length State
MOVBQZX R8, R13
SHRL $0x10, R8
SHRQ $0x10, R8
MOVWQZX R8, R8
LEAQ (BX)(R13*1), CX
MOVQ DX, R14
MOVQ CX, BX
@@ -3075,7 +3017,8 @@ sequenceDecs_decodeSync_safe_amd64_ll_update_zero:
// Update Offset State
MOVBQZX R9, R13
SHRL $0x10, R9
SHRQ $0x10, R9
MOVWQZX R9, R9
LEAQ (BX)(R13*1), CX
MOVQ DX, R14
MOVQ CX, BX
@@ -3464,9 +3407,9 @@ handle_loop:
loop_finished:
MOVQ br+8(FP), AX
MOVQ DX, 24(AX)
MOVQ DX, 32(AX)
MOVB BL, 40(AX)
MOVQ SI, 32(AX)
MOVQ SI, 24(AX)
// Update the context
MOVQ ctx+16(FP), AX
@@ -3512,11 +3455,6 @@ 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
@@ -3531,11 +3469,11 @@ 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), BX
MOVQ 24(BX), AX
MOVBQZX 40(BX), DX
MOVQ (BX), CX
MOVQ 32(BX), BX
MOVQ br+8(FP), CX
MOVQ 32(CX), AX
MOVBQZX 40(CX), DX
MOVQ 24(CX), BX
MOVQ (CX), CX
ADDQ BX, CX
MOVQ CX, (SP)
MOVQ ctx+16(FP), CX
@@ -3560,7 +3498,7 @@ TEXT ·sequenceDecs_decodeSync_safe_bmi2(SB), $64-32
MOVQ 40(SP), CX
ADDQ CX, 48(SP)
// Calculate pointer to s.out[cap(s.out)] (a past-end pointer)
// Calculate poiter to s.out[cap(s.out)] (a past-end pointer)
ADDQ R9, 32(SP)
// outBase += outPosition
@@ -3582,7 +3520,7 @@ sequenceDecs_decodeSync_safe_bmi2_main_loop:
sequenceDecs_decodeSync_safe_bmi2_fill_byte_by_byte:
CMPQ BX, $0x00
JLE sequenceDecs_decodeSync_safe_bmi2_fill_check_overread
JLE sequenceDecs_decodeSync_safe_bmi2_fill_end
CMPQ DX, $0x07
JLE sequenceDecs_decodeSync_safe_bmi2_fill_end
SHLQ $0x08, AX
@@ -3593,10 +3531,6 @@ 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
@@ -3637,7 +3571,7 @@ 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_check_overread
JLE sequenceDecs_decodeSync_safe_bmi2_fill_2_end
CMPQ DX, $0x07
JLE sequenceDecs_decodeSync_safe_bmi2_fill_2_end
SHLQ $0x08, AX
@@ -3648,10 +3582,6 @@ 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
@@ -3683,10 +3613,11 @@ sequenceDecs_decodeSync_safe_bmi2_fill_2_end:
BZHIQ R13, R14, R14
// Update Offset State
BZHIQ R8, R14, CX
SHRXQ R8, R14, R14
SHRL $0x10, R8
ADDQ CX, R8
BZHIQ R8, R14, CX
SHRXQ R8, R14, R14
MOVQ $0x00001010, R13
BEXTRQ R13, R8, R8
ADDQ CX, R8
// Load ctx.ofTable
MOVQ ctx+16(FP), CX
@@ -3694,10 +3625,11 @@ sequenceDecs_decodeSync_safe_bmi2_fill_2_end:
MOVQ (CX)(R8*8), R8
// Update Match Length State
BZHIQ DI, R14, CX
SHRXQ DI, R14, R14
SHRL $0x10, DI
ADDQ CX, DI
BZHIQ DI, R14, CX
SHRXQ DI, R14, R14
MOVQ $0x00001010, R13
BEXTRQ R13, DI, DI
ADDQ CX, DI
// Load ctx.mlTable
MOVQ ctx+16(FP), CX
@@ -3705,9 +3637,10 @@ sequenceDecs_decodeSync_safe_bmi2_fill_2_end:
MOVQ (CX)(DI*8), DI
// Update Literal Length State
BZHIQ SI, R14, CX
SHRL $0x10, SI
ADDQ CX, SI
BZHIQ SI, R14, CX
MOVQ $0x00001010, R13
BEXTRQ R13, SI, SI
ADDQ CX, SI
// Load ctx.llTable
MOVQ ctx+16(FP), CX
@@ -4086,9 +4019,9 @@ handle_loop:
loop_finished:
MOVQ br+8(FP), CX
MOVQ AX, 24(CX)
MOVQ AX, 32(CX)
MOVB DL, 40(CX)
MOVQ BX, 32(CX)
MOVQ BX, 24(CX)
// Update the context
MOVQ ctx+16(FP), AX
@@ -4134,11 +4067,6 @@ 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

2
vendor/github.com/klauspost/compress/zstd/seqdec_generic.go generated vendored
View File

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

2
vendor/github.com/klauspost/compress/zstd/seqenc.go generated vendored
View File

@@ -69,6 +69,7 @@ var llBitsTable = [maxLLCode + 1]byte{
func llCode(litLength uint32) uint8 {
const llDeltaCode = 19
if litLength <= 63 {
// Compiler insists on bounds check (Go 1.12)
return llCodeTable[litLength&63]
}
return uint8(highBit(litLength)) + llDeltaCode
@@ -101,6 +102,7 @@ var mlBitsTable = [maxMLCode + 1]byte{
func mlCode(mlBase uint32) uint8 {
const mlDeltaCode = 36
if mlBase <= 127 {
// Compiler insists on bounds check (Go 1.12)
return mlCodeTable[mlBase&127]
}
return uint8(highBit(mlBase)) + mlDeltaCode

9
vendor/github.com/klauspost/compress/zstd/snappy.go generated vendored
View File

@@ -95,9 +95,10 @@ func (r *SnappyConverter) Convert(in io.Reader, w io.Writer) (int64, error) {
var written int64
var readHeader bool
{
header := frameHeader{WindowSize: snappyMaxBlockSize}.appendTo(r.buf[:0])
var header []byte
var n int
header, r.err = frameHeader{WindowSize: snappyMaxBlockSize}.appendTo(r.buf[:0])
n, r.err = w.Write(header)
if r.err != nil {
return written, r.err
@@ -197,7 +198,7 @@ func (r *SnappyConverter) Convert(in io.Reader, w io.Writer) (int64, error) {
n, r.err = w.Write(r.block.output)
if r.err != nil {
return written, r.err
return written, err
}
written += int64(n)
continue
@@ -239,7 +240,7 @@ func (r *SnappyConverter) Convert(in io.Reader, w io.Writer) (int64, error) {
}
n, r.err = w.Write(r.block.output)
if r.err != nil {
return written, r.err
return written, err
}
written += int64(n)
continue

33
vendor/github.com/klauspost/compress/zstd/zstd.go generated vendored
View File

@@ -5,11 +5,11 @@ package zstd
import (
"bytes"
"encoding/binary"
"errors"
"log"
"math"
"github.com/klauspost/compress/internal/le"
"math/bits"
)
// enable debug printing
@@ -89,10 +89,6 @@ var (
// Close has been called.
ErrDecoderClosed = errors.New("decoder used after Close")
// ErrEncoderClosed will be returned if the Encoder was used after
// Close has been called.
ErrEncoderClosed = errors.New("encoder used after Close")
// ErrDecoderNilInput is returned when a nil Reader was provided
// and an operation other than Reset/DecodeAll/Close was attempted.
ErrDecoderNilInput = errors.New("nil input provided as reader")
@@ -110,12 +106,33 @@ func printf(format string, a ...interface{}) {
}
}
// 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
}
func load3232(b []byte, i int32) uint32 {
return le.Load32(b, i)
return binary.LittleEndian.Uint32(b[i:])
}
func load6432(b []byte, i int32) uint64 {
return le.Load64(b, i)
return binary.LittleEndian.Uint64(b[i:])
}
type byter interface {