diff --git a/vendor/github.com/klauspost/compress/fse/bitreader.go b/vendor/github.com/klauspost/compress/fse/bitreader.go
index b9db204..f65eb39 100644
--- a/vendor/github.com/klauspost/compress/fse/bitreader.go
+++ b/vendor/github.com/klauspost/compress/fse/bitreader.go
@@ -6,6 +6,7 @@
 package fse
 
 import (
+	"encoding/binary"
 	"errors"
 	"io"
 )
@@ -34,8 +35,12 @@
 	}
 	b.bitsRead = 64
 	b.value = 0
-	b.fill()
-	b.fill()
+	if len(in) >= 8 {
+		b.fillFastStart()
+	} else {
+		b.fill()
+		b.fill()
+	}
 	b.bitsRead += 8 - uint8(highBits(uint32(v)))
 	return nil
 }
@@ -63,8 +68,9 @@
 	if b.bitsRead < 32 {
 		return
 	}
-	// Do single re-slice to avoid bounds checks.
-	v := b.in[b.off-4 : b.off]
+	// 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
@@ -77,7 +83,8 @@
 		return
 	}
 	if b.off > 4 {
-		v := b.in[b.off-4 : b.off]
+		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
@@ -91,9 +98,17 @@
 	}
 }
 
+// fillFastStart() assumes the bitreader is empty and there is at least 8 bytes to read.
+func (b *bitReader) fillFastStart() {
+	// Do single re-slice to avoid bounds checks.
+	b.value = binary.LittleEndian.Uint64(b.in[b.off-8:])
+	b.bitsRead = 0
+	b.off -= 8
+}
+
 // finished returns true if all bits have been read from the bit stream.
 func (b *bitReader) finished() bool {
-	return b.off == 0 && b.bitsRead >= 64
+	return b.bitsRead >= 64 && b.off == 0
 }
 
 // close the bitstream and returns an error if out-of-buffer reads occurred.
diff --git a/vendor/github.com/klauspost/compress/fse/bytereader.go b/vendor/github.com/klauspost/compress/fse/bytereader.go
index f228a46..abade2d 100644
--- a/vendor/github.com/klauspost/compress/fse/bytereader.go
+++ b/vendor/github.com/klauspost/compress/fse/bytereader.go
@@ -25,19 +25,10 @@
 	b.off += int(n)
 }
 
-// Int32 returns a little endian int32 starting at current offset.
-func (b byteReader) Int32() int32 {
-	b2 := b.b[b.off : b.off+4 : b.off+4]
-	v3 := int32(b2[3])
-	v2 := int32(b2[2])
-	v1 := int32(b2[1])
-	v0 := int32(b2[0])
-	return v0 | (v1 << 8) | (v2 << 16) | (v3 << 24)
-}
-
 // Uint32 returns a little endian uint32 starting at current offset.
 func (b byteReader) Uint32() uint32 {
-	b2 := b.b[b.off : b.off+4 : b.off+4]
+	b2 := b.b[b.off:]
+	b2 = b2[:4]
 	v3 := uint32(b2[3])
 	v2 := uint32(b2[2])
 	v1 := uint32(b2[1])
diff --git a/vendor/github.com/klauspost/compress/fse/compress.go b/vendor/github.com/klauspost/compress/fse/compress.go
index b69237c..6f34191 100644
--- a/vendor/github.com/klauspost/compress/fse/compress.go
+++ b/vendor/github.com/klauspost/compress/fse/compress.go
@@ -92,7 +92,6 @@
 	im := int32((nbBitsOut << 16) - first.deltaNbBits)
 	lu := (im >> nbBitsOut) + first.deltaFindState
 	c.state = c.stateTable[lu]
-	return
 }
 
 // encode the output symbol provided and write it to the bitstream.
@@ -301,7 +300,7 @@
 	out[outP+1] = byte(bitStream >> 8)
 	outP += (bitCount + 7) / 8
 
-	if uint16(charnum) > s.symbolLen {
+	if charnum > s.symbolLen {
 		return errors.New("internal error: charnum > s.symbolLen")
 	}
 	s.Out = out[:outP]
@@ -331,7 +330,7 @@
 func (s *Scratch) allocCtable() {
 	tableSize := 1 << s.actualTableLog
 	// get tableSymbol that is big enough.
-	if cap(s.ct.tableSymbol) < int(tableSize) {
+	if cap(s.ct.tableSymbol) < tableSize {
 		s.ct.tableSymbol = make([]byte, tableSize)
 	}
 	s.ct.tableSymbol = s.ct.tableSymbol[:tableSize]
@@ -565,8 +564,8 @@
 		distributed  uint32
 		total        = uint32(s.br.remain())
 		tableLog     = s.actualTableLog
-		lowThreshold = uint32(total >> tableLog)
-		lowOne       = uint32((total * 3) >> (tableLog + 1))
+		lowThreshold = total >> tableLog
+		lowOne       = (total * 3) >> (tableLog + 1)
 	)
 	for i, cnt := range s.count[:s.symbolLen] {
 		if cnt == 0 {
@@ -591,7 +590,7 @@
 
 	if (total / toDistribute) > lowOne {
 		// risk of rounding to zero
-		lowOne = uint32((total * 3) / (toDistribute * 2))
+		lowOne = (total * 3) / (toDistribute * 2)
 		for i, cnt := range s.count[:s.symbolLen] {
 			if (s.norm[i] == notYetAssigned) && (cnt <= lowOne) {
 				s.norm[i] = 1
diff --git a/vendor/github.com/klauspost/compress/fse/decompress.go b/vendor/github.com/klauspost/compress/fse/decompress.go
index 413ec3b..926f5f1 100644
--- a/vendor/github.com/klauspost/compress/fse/decompress.go
+++ b/vendor/github.com/klauspost/compress/fse/decompress.go
@@ -172,7 +172,7 @@
 // allocDtable will allocate decoding tables if they are not big enough.
 func (s *Scratch) allocDtable() {
 	tableSize := 1 << s.actualTableLog
-	if cap(s.decTable) < int(tableSize) {
+	if cap(s.decTable) < tableSize {
 		s.decTable = make([]decSymbol, tableSize)
 	}
 	s.decTable = s.decTable[:tableSize]
@@ -340,7 +340,7 @@
 func (d *decoder) init(in *bitReader, dt []decSymbol, tableLog uint8) {
 	d.dt = dt
 	d.br = in
-	d.state = uint16(in.getBits(tableLog))
+	d.state = in.getBits(tableLog)
 }
 
 // next returns the next symbol and sets the next state.
diff --git a/vendor/github.com/klauspost/compress/fse/fse.go b/vendor/github.com/klauspost/compress/fse/fse.go
index 075357b..535cbad 100644
--- a/vendor/github.com/klauspost/compress/fse/fse.go
+++ b/vendor/github.com/klauspost/compress/fse/fse.go
@@ -44,18 +44,14 @@
 // Scratch provides temporary storage for compression and decompression.
 type Scratch struct {
 	// Private
-	count          [maxSymbolValue + 1]uint32
-	norm           [maxSymbolValue + 1]int16
-	symbolLen      uint16 // Length of active part of the symbol table.
-	actualTableLog uint8  // Selected tablelog.
-	br             byteReader
-	bits           bitReader
-	bw             bitWriter
-	ct             cTable      // Compression tables.
-	decTable       []decSymbol // Decompression table.
-	zeroBits       bool        // no bits has prob > 50%.
-	clearCount     bool        // clear count
-	maxCount       int         // count of the most probable symbol
+	count    [maxSymbolValue + 1]uint32
+	norm     [maxSymbolValue + 1]int16
+	br       byteReader
+	bits     bitReader
+	bw       bitWriter
+	ct       cTable      // Compression tables.
+	decTable []decSymbol // Decompression table.
+	maxCount int         // count of the most probable symbol
 
 	// Per block parameters.
 	// These can be used to override compression parameters of the block.
@@ -68,17 +64,22 @@
 	// and allocation will be avoided.
 	Out []byte
 
-	// MaxSymbolValue will override the maximum symbol value of the next block.
-	MaxSymbolValue uint8
-
-	// TableLog will attempt to override the tablelog for the next block.
-	TableLog uint8
-
 	// DecompressLimit limits the maximum decoded size acceptable.
 	// If > 0 decompression will stop when approximately this many bytes
 	// has been decoded.
 	// If 0, maximum size will be 2GB.
 	DecompressLimit int
+
+	symbolLen      uint16 // Length of active part of the symbol table.
+	actualTableLog uint8  // Selected tablelog.
+	zeroBits       bool   // no bits has prob > 50%.
+	clearCount     bool   // clear count
+
+	// MaxSymbolValue will override the maximum symbol value of the next block.
+	MaxSymbolValue uint8
+
+	// TableLog will attempt to override the tablelog for the next block.
+	TableLog uint8
 }
 
 // Histogram allows to populate the histogram and skip that step in the compression,