Merge branch 'dev'

Merge pull request #48 from yincongcyincong/dev
Dev
2018-03-28 14:10:10 +08:00 · 2018-03-28 14:08:28 +08:00 · 2018-03-28 14:06:19 +08:00 · 2018-03-28 13:56:02 +08:00 · 2018-03-28 13:53:09 +08:00 · 2018-03-28 13:34:08 +08:00
374 changed files with 28242 additions and 829 deletions
--- a/35
+++ b/35
@ -1,4 +1,39 @@
 proxy更新日志
+v4.6
+1.sps,http(s),socks5,内网穿透都做了大量的超时优化处理,更加稳定.
+2.sps增加了强大的树形级联认证支持,可以轻松构建你的认证代理网络.
+3.手册增加了6.6对sps认证功能的介绍.
+
+
+v4.5
+1.优化了mux内网穿透连接管理逻辑,增强了稳定性.  
+2.mux内网穿透增加了tcp和kcp协议支持,之前是tls,现在支持三种协议tcp,tls,kcp.  
+3.keygen参数增加了用法: proxy keygen usage.  
+4.http(s)/socks5代理,tls增加了自签名证书支持.  
+5.建议升级.   
+v4.4
+1.增加了协议转换sps功能，代理协议转换使用的是sps子命令(socks+https的缩写)，
+sps本身不提供代理功能，只是接受代理请求"转换并转发"给已经存在的http(s)代理
+或者socks5代理；sps可以把已经存在的http(s)代理或者socks5代理转换为一个端口
+同时支持http(s)和socks5代理，而且http(s)代理支持正向代理和反向代理(SNI)，转
+换后的SOCKS5代理不支持UDP功能；另外对于已经存在的http(s)代理或者socks5代理，
+支持tls、tcp、kcp三种模式，支持链式连接，也就是可以多个sps结点层级连接构建
+加密通道。
+2.增加了对KCP传输参数的配置，多达17个参数可以自由的配置对kcp传输效率调优。
+3.内网穿透功能，server和client增加了--session-count参数，可以设置server每个
+监听端口到bridge打开的session数量，可以设置client到bridge打开的session数量，
+之前都是1个，现在性能提升N倍，N就是你自己设置的--session-count，这个参数很大
+程度上解决了多路复用的拥塞问题，v4.4开始默认10个。
+
+v4.3
+1.优化了参数keygen生成证书逻辑，避免证书出现特征。
+2.http(s)和socks代理增加了--dns-address和--dns-ttl参数。
+ 用于自己指定proxy访问域名的时候使用的dns（--dns-address）以及解析结果缓存时间（--dns-ttl）秒数，
+ 避免系统dns对proxy的干扰，另外缓存功能还能减少dns解析时间提高访问速度。
+3.优化了http代理的basic认证逻辑。
+提示：
+v4.3生成的证书不适用于v4.2及以下版本。
+
 v4.2
 1.优化了内网穿透,避免了client意外下线,导致链接信息残留的问题.
 2.http代理增加了SNI支持,现在http(s)代理模式支持反向代理,支持http(s)透明代理.
--- a/Godeps/Godeps.json
+++ b/Godeps/Godeps.json
@ -1,36 +1,80 @@
 {
-	"ImportPath": "proxy",
-	"GoVersion": "go1.8",
-	"GodepVersion": "v79",
+	"ImportPath": "snail007/proxy",
+	"GoVersion": "go1.9",
+	"GodepVersion": "v80",
 	"Packages": [
 		"./..."
 	],
 	"Deps": [
-		{
-			"ImportPath": "github.com/alecthomas/template",
-			"Rev": "a0175ee3bccc567396460bf5acd36800cb10c49c"
-		},
-		{
-			"ImportPath": "github.com/alecthomas/template/parse",
-			"Rev": "a0175ee3bccc567396460bf5acd36800cb10c49c"
-		},
-		{
-			"ImportPath": "github.com/alecthomas/units",
-			"Rev": "2efee857e7cfd4f3d0138cc3cbb1b4966962b93a"
-		},
 		{
 			"ImportPath": "github.com/golang/snappy",
 			"Rev": "553a641470496b2327abcac10b36396bd98e45c9"
 		},
+		{
+			"ImportPath": "github.com/miekg/dns",
+			"Comment": "v1.0.4-1-g40b5202",
+			"Rev": "40b520211179dbf7eaafaa7fe1ffaa1b7d929ee0"
+		},
+		{
+			"ImportPath": "github.com/xtaci/kcp-go",
+			"Comment": "v3.19-6-g21da33a",
+			"Rev": "21da33a6696d67c1bffb3c954366499d613097a6"
+		},
+		{
+			"ImportPath": "github.com/xtaci/smux",
+			"Comment": "v1.0.6",
+			"Rev": "ebec7ef2574b42a7088cd7751176483e0a27d458"
+		},
+		{
+			"ImportPath": "golang.org/x/crypto/pbkdf2",
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
+		},
+		{
+			"ImportPath": "golang.org/x/crypto/ssh",
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
+		},
+		{
+			"ImportPath": "golang.org/x/time/rate",
+			"Rev": "6dc17368e09b0e8634d71cac8168d853e869a0c7"
+		},
+		{
+			"ImportPath": "gopkg.in/alecthomas/kingpin.v2",
+			"Comment": "v2.2.5",
+			"Rev": "1087e65c9441605df944fb12c33f0fe7072d18ca"
+		},
+		{
+			"ImportPath": "golang.org/x/crypto/ed25519",
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
+		},
+		{
+			"ImportPath": "golang.org/x/net/ipv4",
+			"Rev": "5ccada7d0a7ba9aeb5d3aca8d3501b4c2a509fec"
+		},
+		{
+			"ImportPath": "golang.org/x/net/ipv6",
+			"Rev": "5ccada7d0a7ba9aeb5d3aca8d3501b4c2a509fec"
+		},
+		{
+			"ImportPath": "golang.org/x/crypto/ed25519/internal/edwards25519",
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
+		},
+		{
+			"ImportPath": "golang.org/x/net/bpf",
+			"Rev": "5ccada7d0a7ba9aeb5d3aca8d3501b4c2a509fec"
+		},
+		{
+			"ImportPath": "golang.org/x/net/internal/iana",
+			"Rev": "5ccada7d0a7ba9aeb5d3aca8d3501b4c2a509fec"
+		},
+		{
+			"ImportPath": "golang.org/x/net/internal/socket",
+			"Rev": "5ccada7d0a7ba9aeb5d3aca8d3501b4c2a509fec"
+		},
 		{
 			"ImportPath": "github.com/pkg/errors",
 			"Comment": "v0.8.0-6-g602255c",
 			"Rev": "602255cdb6deaf1523ea53ac30eae5554ba7bee9"
 		},
-		{
-			"ImportPath": "github.com/templexxx/cpufeat",
-			"Rev": "3794dfbfb04749f896b521032f69383f24c3687e"
-		},
 		{
 			"ImportPath": "github.com/templexxx/reedsolomon",
 			"Comment": "0.1.1-4-g7092926",
@ -46,88 +90,53 @@
 			"Comment": "v1.0.1-3-g9d99fac",
 			"Rev": "9d99face20b0dd300b7db50b3f69758de41c096a"
 		},
-		{
-			"ImportPath": "github.com/xtaci/kcp-go",
-			"Comment": "v3.19-6-g21da33a",
-			"Rev": "21da33a6696d67c1bffb3c954366499d613097a6"
-		},
-		{
-			"ImportPath": "github.com/xtaci/smux",
-			"Comment": "v1.0.6",
-			"Rev": "ebec7ef2574b42a7088cd7751176483e0a27d458"
-		},
 		{
 			"ImportPath": "golang.org/x/crypto/blowfish",
-			"Rev": "f899cbd3df85058aa20d1cf129473b18f2a2b49f"
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
 			"ImportPath": "golang.org/x/crypto/cast5",
-			"Rev": "86e16787bfd59cb4db9e278c51a95488c141a5d6"
-		},
-		{
-			"ImportPath": "golang.org/x/crypto/curve25519",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
-		},
-		{
-			"ImportPath": "golang.org/x/crypto/ed25519",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
-		},
-		{
-			"ImportPath": "golang.org/x/crypto/ed25519/internal/edwards25519",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
-		},
-		{
-			"ImportPath": "golang.org/x/crypto/pbkdf2",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
 			"ImportPath": "golang.org/x/crypto/salsa20",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
-		},
-		{
-			"ImportPath": "golang.org/x/crypto/salsa20/salsa",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
-		},
-		{
-			"ImportPath": "golang.org/x/crypto/ssh",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
 			"ImportPath": "golang.org/x/crypto/tea",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
 			"ImportPath": "golang.org/x/crypto/twofish",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
 			"ImportPath": "golang.org/x/crypto/xtea",
-			"Rev": "1843fabd21d7180cf65e36759986d00c13dbb0fd"
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
-			"ImportPath": "golang.org/x/net/bpf",
-			"Rev": "114479435b31b5077a087cc5303a45cb5d355dc4"
+			"ImportPath": "github.com/templexxx/cpufeat",
+			"Rev": "3794dfbfb04749f896b521032f69383f24c3687e"
 		},
 		{
-			"ImportPath": "golang.org/x/net/internal/iana",
-			"Rev": "114479435b31b5077a087cc5303a45cb5d355dc4"
+			"ImportPath": "golang.org/x/crypto/salsa20/salsa",
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
-			"ImportPath": "golang.org/x/net/internal/socket",
-			"Rev": "114479435b31b5077a087cc5303a45cb5d355dc4"
+			"ImportPath": "golang.org/x/crypto/curve25519",
+			"Rev": "0fcca4842a8d74bfddc2c96a073bd2a4d2a7a2e8"
 		},
 		{
-			"ImportPath": "golang.org/x/net/ipv4",
-			"Rev": "114479435b31b5077a087cc5303a45cb5d355dc4"
+			"ImportPath": "github.com/alecthomas/template",
+			"Rev": "a0175ee3bccc567396460bf5acd36800cb10c49c"
 		},
 		{
-			"ImportPath": "golang.org/x/time/rate",
-			"Rev": "8be79e1e0910c292df4e79c241bb7e8f7e725959"
+			"ImportPath": "github.com/alecthomas/units",
+			"Rev": "2efee857e7cfd4f3d0138cc3cbb1b4966962b93a"
 		},
 		{
-			"ImportPath": "gopkg.in/alecthomas/kingpin.v2",
-			"Comment": "v2.2.5",
-			"Rev": "1087e65c9441605df944fb12c33f0fe7072d18ca"
+			"ImportPath": "github.com/alecthomas/template/parse",
+			"Rev": "a0175ee3bccc567396460bf5acd36800cb10c49c"
 		}
 	]
 }
--- a/README.md
+++ b/README.md
@ -1,5 +1,6 @@
-<img src="https://github.com/snail007/goproxy/blob/master/docs/images/logo.jpg?raw=true" width="200"/>  
-Proxy is a high performance HTTP, HTTPS, HTTPS, websocket, TCP, UDP, Socks5 proxy server implemented by golang. It supports parent proxy,nat forward,TCP/UDP port forwarding, SSH transfer. you can expose a local server behind a NAT or firewall to the internet.  
+<img src="https://github.com/snail007/goproxy/blob/master/docs/images/logo.jpg?raw=true" width="200"/>
+Proxy is a high performance HTTP, HTTPS, HTTPS, websocket, TCP, UDP, Socks5 proxy server implemented by golang. It supports parent proxy,nat forward,TCP/UDP port forwarding, SSH transfer, TLS encrypted transmission, protocol conversion. you can expose a local server behind a NAT or firewall to the internet.  
+
  
 ---  
  
@ -19,6 +20,9 @@ Proxy is a high performance HTTP, HTTPS, HTTPS, websocket, TCP, UDP, Socks5 prox
 - SSH forwarding: HTTP (S), SOCKS5 proxy support SSH transfer, parent Linux server does not need any server, a local proxy can be happy to access the Internet.  
 - [KCP](https://github.com/xtaci/kcp-go) protocol is supported: HTTP (S), SOCKS5 proxy supports the KCP protocol which can transmit data, reduce latency, and improve the browsing experience.  
 - The integrated external API, HTTP (S): SOCKS5 proxy authentication can be integrated with the external HTTP API, which can easily control the user's access through the external system.  
+- Reverse proxy: goproxy supports directly parsing the domain to proxy monitor IP, and then proxy will help you to access the HTTP (S) site that you need to access.
+- Transparent proxy: with the iptables, goproxy can directly forward the 80 and 443 port's traffic to proxy in the gateway, and can realize the unaware intelligent router proxy.  
+- Protocol conversion: The existing HTTP (S) or SOCKS5 proxy can be converted to a proxy which support both HTTP (S) and SOCKS5 by one port, but the converted SOCKS5 proxy does not support the UDP function.Also support powerful cascading authentication.  
  
 ### Why need these?  
 - Because for some reason, we cannot access our services elsewhere. We can build a secure tunnel to access our services through multiple connected proxy nodes.  
@ -30,7 +34,12 @@ Proxy is a high performance HTTP, HTTPS, HTTPS, websocket, TCP, UDP, Socks5 prox
 - ...  

 
-This page is the v4.0-v4.1 manual, and the other version of the manual can be checked by the following link.  
+This page is the v4.6 manual, and the other version of the manual can be checked by the following link.  
+- [v4.5 manual](https://github.com/snail007/goproxy/tree/v4.5)
+- [v4.4 manual](https://github.com/snail007/goproxy/tree/v4.4)
+- [v4.3 manual](https://github.com/snail007/goproxy/tree/v4.3)
+- [v4.2 manual](https://github.com/snail007/goproxy/tree/v4.2)
+- [v4.0-4.1 manual](https://github.com/snail007/goproxy/tree/v4.1)
 - [v3.9 manual](https://github.com/snail007/goproxy/tree/v3.9)
 - [v3.8 manual](https://github.com/snail007/goproxy/tree/v3.8)
 - [v3.6-v3.7 manual](https://github.com/snail007/goproxy/tree/v3.6)
@ -43,13 +52,15 @@ This page is the v4.0-v4.1 manual, and the other version of the manual can be ch
 - [v2.x manual](https://github.com/snail007/goproxy/tree/v2.2)  

 ### How to find the organization?  
-[Click to join the communication organization](https://gitter.im/go-proxy/Lobby?utm_source=share-link&utm_medium=link&utm_campaign=share-link)  
+[Click to join the proxy group of gitter](https://gitter.im/go-proxy/Lobby?utm_source=share-link&utm_medium=link&utm_campaign=share-link)  
+[Click to join the proxy group of telegram](https://t.me/joinchat/GYHXghCDSBmkKZrvu4wIdQ)    
+

 ### Installation
 - [Quick installation](#quick-installation)
 - [Manual installation](#manual-installation)

-### First use must be read
+### First use must read
 - [Environmental Science](#environmental-science)
 - [Use configuration file](#use-configuration-file)
 - [Debug output](#debug-output)
@ -71,7 +82,10 @@ This page is the v4.0-v4.1 manual, and the other version of the manual can be ch
        - [1.7.1 The way of username and password](#171the-way-of-username-and-password)
        - [1.7.2 The way of username and key](#172the-way-of-username-and-key)
    - [1.8 KCP protocol transmission](#18kcp-protocol-transmission)
-    - [1.9 View help](#19view-help)
+    - [1.9 HTTP(S) reverse proxy](#19http-reverse-proxy)
+    - [1.10 HTTP(S) transparent proxy](#110http-transparent-proxy)
+    - [1.11 Custom DNS](#111custom-dns)
+    - [1.12 View help](#112view-help)
 - [2.TCP proxy](#2tcp-proxy)
    - [2.1 Common TCP first level proxy](#21common-tcp-first-level-proxy)
    - [2.2 Common TCP second level proxy](#22common-tcp-second-level-proxy)
@ -106,12 +120,25 @@ This page is the v4.0-v4.1 manual, and the other version of the manual can be ch
        - [5.6.2 The way of username and key](#562the-way-of-username-and-key)
    - [5.7 Authentication](#57authentication)
    - [5.8 KCP protocol transmission](#58kcp-protocol-transmission)
-    - [5.9 View help](#59view-help)
+    - [5.9 Custom DNS](#59custom-dns)
+    - [5.10 View help](#510view-help)
+- [6.Proxy protocol conversion](#6proxy-protocol-conversion)
+    - [6.1 Functional introduction](#61functional-introduction)
+    - [6.2 HTTP(S) to HTTP(S) + SOCKS5](#62http-to-http-socks5)
+    - [6.3 SOCKS5 to HTTP(S) + SOCKS5](#63socks5-to-http-socks5)
+    - [6.4 Chain style connection](#64chain-style-connection)
+    - [6.5 Listening on multiple ports](#65listening-on-multiple-ports)
+    - [6.6 Authentication](#66authentication)
+    - [6.7 View Help](#67view-help)
+- [7.KCP Configuration](#7kcp-configuration)
+    - [7.1 Configuration introduction](#71configuration-introduction)
+    - [7.2 Configuration details](#72configuration-details)
+

 ### Fast Start  
 tips:all operations require root permissions.   
 #### Quick installation
-#### **0. If your VPS is a linux64 system, you can complete the automatic installation and configuration by the following sentence.**  
+#### **0. If your VPS is linux64, you can complete the automatic installation and configuration by the following sentence.**  
 ```shell  
 curl -L https://raw.githubusercontent.com/snail007/goproxy/master/install_auto.sh | bash  
 ```  
@ -124,7 +151,7 @@ If the installation fails or your VPS is not a linux64 system, please follow the
 Download address: https://github.com/snail007/goproxy/releases  
 ```shell  
 cd /root/proxy/  
-wget https://github.com/snail007/goproxy/releases/download/v4.0/proxy-linux-amd64.tar.gz  
+wget https://github.com/snail007/goproxy/releases/download/v4.6/proxy-linux-amd64.tar.gz  
 ```  
 #### **2.Download the automatic installation script**  
 ```shell  
@ -137,14 +164,14 @@ chmod +x install.sh
 ## **First use must be read**  
  
 ### **Environmental Science**  
-The following tutorial, the default system is Linux, the program is proxy; all operations require root permissions.   
+The following tutorial defaults system is Linux, the program is proxy and all operations require root permissions.   
 If the system are windows, please use proxy.exe.  
  
 ### **Use configuration file**  
-The following tutorial is to introduce the use method by the command line parameters, or by reading the configuration file to get the parameters.  
+The following tutorial is to introduce the useage by the command line parameters, or by reading the configuration file to get the parameters.  
 The specific format is to specify a configuration file by the @ symbol, for example, ./proxy @configfile.txt.   
-configfile.txt's format: The first line is the subcommand name, and the second line begins one line: the long format of the parameter = the parameter value, there is no space and double quotes before and after.  
-The long format of the parameter's beginning is always --, the short format of the parameter's beginning is always -. If you don't know which short form parameter corresponds to the long format parameter, please look at the help command.  
+configfile.txt's format: The first line is the subcommand name, and the second line begins a new line: the long format of the parameter = the parameter value, there is no space and double quotes before and after.  
+The long format of the parameter's beginning is --, the short format of the parameter's beginning is -. If you don't know which short form corresponds to the long format, please look at the help command.  
 For example, the contents of configfile.txt are as follows:
 ```shell
 http
@ -183,6 +210,7 @@ Assuming that your VPS outer external network IP is 23.23.23.23, the following c

 ### **1.HTTP proxy**  
 #### **1.1.common HTTP proxy**  
+![1.1](/docs/images/1.1.jpg)
 `./proxy http -t tcp -p "0.0.0.0:38080"`  
  
 #### **1.2.Common HTTP second level proxy**  
@ -256,26 +284,93 @@ Local HTTP (S) proxy use 28080 port,excute:
 `./proxy http -T ssh -P "2.2.2.2:22" -u user -S user.key -t tcp -p ":28080"`  

 #### **1.8.KCP protocol transmission**  
-The KCP protocol requires a -B parameter to set a password which can encrypt and decrypt data.  
+The KCP protocol requires a --kcp-key parameter to set a password which can encrypt and decrypt data.  

 Http first level proxy(VPS,IP:22.22.22.22)  
-`./proxy http -t kcp -p ":38080" -B mypassword`  
+`./proxy http -t kcp -p ":38080" --kcp-key mypassword`  
  
 Http second level proxy(os is Linux)  
-`./proxy http -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" -B mypassword`  
+`./proxy http -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" --kcp-key mypassword`  
 Then access to the local 8080 port is access to the proxy's port 38080 on the VPS, and the data is transmitted through the KCP protocol.  
+#### **1.9.HTTP reverse proxy** 
+Proxy supports not only set up a proxy through in other software, to provide services for other software, but support the request directly to the website domain to proxy monitor IP when proxy monitors 80 and 443 ports, then proxy will automatically access to the HTTP proxy access website for you.  

-#### **1.9.view help**  
+How to use:  
+On the last level proxy computer, because proxy is disguised as all websites and the default port of HTTP is 80, HTTPS is 443, the proxy listens to 80 and 443 port. Parameters -p multiple addresses are separated by commas.  
+`./proxy http -t tcp -p :80,:443`    
+
+This command starts a proxy on the computer, and listens to 80 and 443 ports. It can be used as a common proxy and it can directly resolve the domain that needs proxy to the IP of the computer. 
+
+If a parent proxy exist, you can refer to the above tutorial to set up a parent. The way of use is exactly the same.  
+`./proxy http -t tcp -p :80,:443 -T tls -P "2.2.2.2:33080" -C proxy.crt -K proxy.key`   
+
+Notice:  
+The result of the DNS parsing of the server in which proxy is located can not affected by a custom parsing, if not, it is dead cycle.  
+  
+#### **1.10.HTTP transparent proxy** 
+The mode needs a certain network knowledge, if the related concepts don't understand, you must search it by yourself.  
+Assuming that proxy is now running on the router, the boot command is as follows:  
+`./proxy http -t tcp -p :33080 -T tls -P "2.2.2.2:33090" -C proxy.crt -K proxy.key`   
+
+Then the iptables rule is added, and the following rule is a reference rule:  
+```shell
+#IP of parent proxy:
+proxy_server_ip=2.2.2.2
+
+#Proxy that the router runs monitor the port:
+proxy_local_port=33080
+
+#The following don't need to be modified
+#create a new chain named PROXY
+iptables -t nat -N PROXY
+
+# Ignore your PROXY server's addresses
+# It's very IMPORTANT, just be careful.
+
+iptables -t nat -A PROXY -d $proxy_server_ip -j RETURN
+
+# Ignore LANs IP address
+iptables -t nat -A PROXY -d 0.0.0.0/8 -j RETURN
+iptables -t nat -A PROXY -d 10.0.0.0/8 -j RETURN
+iptables -t nat -A PROXY -d 127.0.0.0/8 -j RETURN
+iptables -t nat -A PROXY -d 169.254.0.0/16 -j RETURN
+iptables -t nat -A PROXY -d 172.16.0.0/12 -j RETURN
+iptables -t nat -A PROXY -d 192.168.0.0/16 -j RETURN
+iptables -t nat -A PROXY -d 224.0.0.0/4 -j RETURN
+iptables -t nat -A PROXY -d 240.0.0.0/4 -j RETURN
+
+# Anything to port 80 443 should be redirected to PROXY's local port
+iptables -t nat -A PROXY -p tcp --dport 80 -j REDIRECT --to-ports $proxy_local_port
+iptables -t nat -A PROXY -p tcp --dport 443 -j REDIRECT --to-ports $proxy_local_port
+
+# Apply the rules to nat client
+iptables -t nat -A PREROUTING -p tcp -j PROXY
+# Apply the rules to localhost
+iptables -t nat -A OUTPUT -p tcp -j PROXY
+```
+- Clearing the whole chain command is iptables -F chain name, such as iptables -t NAT -F PROXY
+- Deleting the specified chain that user defined command is iptables -X chain name, such as iptables -t NAT -X PROXY
+- Deleting the rules of the chain command is iptables -D chain name from the selected chain, such as  iptables -t nat -D PROXY -d 223.223.192.0/255.255.240.0 -j RETURN
+
+#### **1.11.Custom DNS** 
+--dns-address and --dns-ttl parameters can be used to specify DNS（--dns-address） when you use proxy to access to a domain.  
+they also can specify dns result cache time (--dns-ttl) which unit is second. they can avoid the interference of system DNS to proxy. cache can reduce DNS resolution time and increase access speed.  
+for example:  
+`./proxy http -p ":33080" --dns-address "8.8.8.8:53" --dns-ttl 300`  
+
+#### **1.12.view help**  
 `./proxy help http`  
  
 ### **2.TCP proxy**  
  
 #### **2.1.Common TCP first level proxy**  
+![2.1](/docs/images/2.1.png)
 Local execution:  
 `./proxy tcp -p ":33080" -T tcp -P "192.168.22.33:22" -L 0`  
 Then access to the local 33080 port is the 22 port of access to 192.168.22.33.  
  
 #### **2.2.Common TCP second level proxy**  
+![2.2](/docs/images/2.2.png)
 VPS(IP:22.22.22.33) execute:  
 `./proxy tcp -p ":33080" -T tcp -P "127.0.0.1:8080" -L 0`  
 local execution:  
@ -354,7 +449,7 @@ Then access to the local UDP:5353 port is access to the UDP:53 port of the 8.8.8
  
 ### **4.Nat forward**  
 #### **4.1、Principle explanation**  
-Nat forward, divided into two versions, "multi-link version" and "multiplexed version", generally like web services Which is not a long time to connect the service recommended "multi-link version", if you want to keep long Time connection, "multiplexed version" is recommended.
+Nat forward, is divided into two versions, "multi-link version" and "multiplexed version", generally like web services Which is not a long time to connect the service recommende "multi-link version", if you want to keep long Time connection, "multiplexed version" is recommended.
 1. Multilink version, the corresponding subcommand is tserver，tclient，tbridge。  
 1. Multiplexed version, the corresponding subcommand is server，client，bridge。  
 1. the parameters and use of Multilink version and multiplexed is exactly the same.  
@ -476,8 +571,8 @@ Procedure:
  
  4.7.1.PROTOCOL is tcp or udp.  
  for example: `-r "udp://:10053@:53" -r "tcp://:10800@:1080" -r ":8080@:80"`  
-  If the --udp parameter is specified, PROTOCOL is UDP by default, then `-r ": 8080@: 80"` is UDP.  
-  If the --udp parameter is not specified, PROTOCOL is TCP by default, then `-r ": 8080@: 80"` is TCP.  
+  If the --udp parameter is specified, PROTOCOL is UDP by default, then `-r ":8080@:80"` is UDP.  
+  If the --udp parameter is not specified, PROTOCOL is TCP by default, then `-r ":8080@:80"` is TCP.  
  
  4.7.2.CLIENT_KEY by default is 'default'.  
  for example: -r "udp://:10053@[test1]:53" -r "tcp://:10800@[test2]:1080" -r ":8080@:80"  
@ -489,7 +584,7 @@ Procedure:
 #### **4.8.view help**  
 `./proxy help bridge`  
 `./proxy help server`  
-`./proxy help server`  
+`./proxy help client`  
  
 ### **5.SOCKS5 proxy**  
 Tips: SOCKS5 proxy, support CONNECT, UDP protocol and don't support BIND and support username password authentication.  
@ -497,6 +592,7 @@ Tips: SOCKS5 proxy, support CONNECT, UDP protocol and don't support BIND and sup
 `./proxy socks -t tcp -p "0.0.0.0:38080"`  
   
 #### **5.2.Common SOCKS5 second level proxy**  
+![5.2](/docs/images/5.2.png)
 Using local port 8090, assume that the parent SOCKS5 proxy is `22.22.22.22:8080`  
 `./proxy socks -t tcp -p "0.0.0.0:8090" -T tcp -P "22.22.22.22:8080" `  
 We can also specify the black and white list files of the domain name, one line for one domain name. The matching rule is the most right-hand matching. For example, baidu.com is *.*.baidu.com, the domain name of the blacklist is directly accessed by the parent proxy, and the domain name of the white list does not access to the parent proxy.  
@ -550,7 +646,7 @@ You can also be placed in a file, which is a line, a ‘username: password’, a
 `./proxy socks -t tcp -p ":33080" -F auth-file.txt`  

 In addition, socks5 proxy also integrates external HTTP API authentication, we can specify a http url interface address through the --auth-url parameter,  
-Then when the user is connected, the proxy GET request this url, with the following four parameters, if the return HTTP status code 204, on behalf of the authentication is successful.  
+Then when the user is connected, the proxy request this url by get way, with the following four parameters, if the return HTTP status code 204, on behalf of the authentication is successful.  
 In other cases, the authentication fails.  
 for example:  
 `./proxy socks -t tcp -p ":33080" --auth-url "http://test.com/auth.php"`  
@ -564,26 +660,166 @@ ip: user's IP, for example: 192.168.1.200
 If there is no -a or -F or --auth-url parameters, it means to turn off the authentication.    

 #### **5.8.KCP protocol transmission**  
-The KCP protocol requires a -B parameter to set a password to encrypt and decrypt data.  
+The KCP protocol requires a --kcp-key parameter which can set a password to encrypt and decrypt data.  

 HTTP first level proxy(VPS,IP:22.22.22.22)  
-`./proxy socks -t kcp -p ":38080" -B mypassword`  
+`./proxy socks -t kcp -p ":38080" --kcp-key mypassword`  
  
 HTTP two level proxy(local os is Linux)  
-`./proxy socks -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" -B mypassword`  
-Then access to the local 8080 port is access to the proxy port 38080 on the VPS, and the data is transmitted through the KCP protocol.  
+`./proxy socks -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" --kcp-key mypassword`  
+Then access to the local 8080 port is access to the proxy port 38080 on the VPS, and the data is transmitted through the KCP protocol.

-#### **5.9.view help**  
+#### **5.9.Custom DNS** 
+--dns-address and --dns-ttl parameters can be used to specify DNS（--dns-address） when you use proxy to access to a domain.  
+they also can specify dns result cache time (--dns-ttl) which unit is second. they can avoid the interference of system DNS to proxy. cache can reduce DNS resolution time and increase access speed.  
+for example:  
+`./proxy socks -p ":33080" --dns-address "8.8.8.8:53" --dns-ttl 300`  
+
+
+#### **5.10.view help**  
 `./proxy help socks`  

+### **6.Proxy protocol conversion** 
+
+#### **6.1.Functional introduction** 
+The proxy protocol conversion use the SPS subcommand (abbreviation of socks+https), SPS itself does not provide the proxy function, just accept the proxy request and then converse protocol and forwarded to the existing HTTP (s) or Socks5 proxy. SPS can use existing HTTP (s) or Socks5 proxy converse to support HTTP (s) and Socks5 HTTP (s) proxy at the same time by one port, and proxy supports forward and reverse proxy (SNI), SOCKS5 proxy which is conversed does not support UDP. in addition to the existing HTTP or Socks5 proxy, which supports TLS, TCP, KCP three modes and chain-style connection. That is more than one SPS node connection can build encryption channel.
+
+#### **6.2.HTTP(S) to HTTP(S) + SOCKS5** 
+Suppose there is a common HTTP (s) proxy: 127.0.0.1:8080. Now we turn it into a common proxy that supports HTTP (s) and Socks5 at the same time. The local port after transformation is 18080.  
+command：  
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p :18080`
+
+Suppose that there is a TLS HTTP (s) proxy: 127.0.0.1:8080. Now we turn it into a common proxy that supports HTTP (s) and Socks5 at the same time. The local port after transformation is 18080, TLS needs certificate file.  
+command：  
+`./proxy sps -S http -T tls -P 127.0.0.1:8080 -t tcp -p :18080 -C proxy.crt -K proxy.key`   
+
+Suppose there is a KCP HTTP (s) proxy (password: demo123): 127.0.0.1:8080. Now we turn it into a common proxy that supports HTTP (s) and Socks5 at the same time. The local port after transformation is 18080.  
+command：  
+`./proxy sps -S http -T kcp -P 127.0.0.1:8080 -t tcp -p :18080 --kcp-key demo123`  
+
+#### **6.3.SOCKS5 to HTTP(S) + SOCKS5** 
+Suppose there is a common Socks5 proxy: 127.0.0.1:8080, now we turn it into a common proxy that supports HTTP (s) and Socks5 at the same time, and the local port after transformation is 18080.  
+command：  
+`./proxy sps -S socks -T tcp -P 127.0.0.1:8080 -t tcp -p :18080`
+
+Suppose there is a TLS Socks5 proxy: 127.0.0.1:8080. Now we turn it into a common proxy that support HTTP (s) and Socks5 at the same time. The local port after transformation is 18080, TLS needs certificate file.  
+command：  
+`./proxy sps -S socks -T tls -P 127.0.0.1:8080 -t tcp -p :18080 -C proxy.crt -K proxy.key`   
+
+Suppose there is a KCP Socks5 proxy (password: demo123): 127.0.0.1:8080, now we turn it into a common proxy that support HTTP (s) and Socks5 at the same time, and the local port after transformation is 18080.  
+command：  
+`./proxy sps -S socks -T kcp -P 127.0.0.1:8080 -t tcp -p :18080 --kcp-key demo123`  
+
+#### **6.4.Chain style connection** 
+It is mentioned above that multiple SPS nodes can be connected to build encrypted channels, assuming you have the following VPS and a PC.  
+vps01：2.2.2.2  
+vps02：3.3.3.3  
+Now we want to use PC and vps01 and vps02 to build an encrypted channel. In this example, TLS is used. KCP also supports encryption in addition to TLS. and accessing to local 18080 port on PC is accessing to the local 8080 ports of vps01.  
+First, on vps01 (2.2.2.2), we run a HTTP (s) proxy that only can be accessed locally,excute：  
+`./proxy -t tcp -p 127.0.0.1:8080`  
+
+Then run a SPS node on vps01 (2.2.2.2)，excute：  
+`./proxy -S http -T tcp -P 127.0.0.1:8080 -t tls -p :8081 -C proxy.crt -K proxy.key`  
+
+Then run a SPS node on vps02 (3.3.3.3)，excute：  
+`./proxy -S http -T tls -P 2.2.2.2:8081 -t tls -p :8082 -C proxy.crt -K proxy.key`  
+
+Then run a SPS node on the PC，excute：  
+`./proxy -S http -T tls -P 3.3.3.3:8082 -t tcp -p :18080 -C proxy.crt -K proxy.key`  
+
+finish。  
+
+#### **6.5.Listening on multiple ports**   
+In general, listening one port is enough, but if you need to monitor 80 and 443 ports at the same time as a reverse proxy, the -p parameter can support it.  
+The format is：`-p 0.0.0.0:80,0.0.0.0:443`, Multiple bindings are separated by a comma.  
+
+#### **6.6.Authentication** 
+SPS supports HTTP(s)\socks5 proxy authentication, which can concatenate authentication, there are four important information:  
+1:Users send authentication information`user-auth`。   
+2:Local authentication information set up`local-auth`。  
+3:Set the authentication information accessing to the father proxy`parent-auth`。  
+4:The final authentication information sent to the father proxy`auth-info-to-parent`。  
+The relationship between them is as follows:   
+
+| user-auth | local-auth | parent-auth | auth-info-to-paren 
+| ------ | ------ | ------ | ------  
+| yes/no  | yes    | yes   |  come from parent-auth  
+| yes/no  | no    |    yes    |   come from parent-auth  
+| yes/no  | yes     |     no  |   no  
+| no   | no    |   no    |   no  
+| yes    | no    |   no    |   come from user-auth  
+
+For SPS proxy we can have username and password to authenticate, and the authentication username and password can be specified on the command line    
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" -a "user1:pass1" -a "user2:pass2"`  
+if there are multiple users, repeat the -a parameters.  
+It can also be placed in a file, which is a line to a username: password, and then specified in -F parameter.  
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" -F auth-file.txt`  
+
+If the father proxy is authenticated, the lower level can set the authentication information through the -A parameters, such as:  
+father proxy:`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" -a "user1:pass1" -a "user2:pass2"`  
+local proxy:`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -A "user1:pass1" -t tcp -p ":33080" `  
+
+In addition, SPS proxy, local authentication is integrated with external HTTP API authentication, and we can specify a HTTP URL interface address through the --auth-url parameter,    
+Then, when there is a user connection, proxy will request this URL by GET way, with the following four parameters, and if the HTTP state code 204 is returned, the authentication is successful.  
+Other cases consider authentication failure.  
+for example:  
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" --auth-url "http://test.com/auth.php"`  
+When the user is connected, proxy will request this URL by GET way("http://test.com/auth.php"),  
+Four parameters with user, pass, IP, and target:  
+http://test.com/auth.php?user={USER}&pass={PASS}&ip={IP}&target={TARGET}  
+user:username   
+pass:password   
+ip:user's ip,for example:192.168.1.200   
+target: if the client is the HTTP (s) proxy request, this represents the complete URL of the request, and the other cases are empty.  
+
+If there is no -a or -F or --auth-url parameters, local authentication is closed.  
+If there is no -A parameter, the connection to the father proxy does not use authentication.  
+
+#### **6.7.view help** 
+`./proxy help sps` 
+
+### **7.KCP Configuration**   
+
+#### **7.1.Configuration introduction**   
+Many functions of the proxy support the KCP protocol, and all the functions that can use the KCP protocol support the configuration parameters introduced here.  
+So here is a unified introduction to the KCP configuration parameters.  
+
+#### **7.2.Configuration details**   
+The number of KCP configuration parameters is 17, you don't have to set up them. they all have the default value, if for the best effect,  
+You need to configure the parameters according to your own network conditions. Due to the complexity of KCP configuration, a certain network basic knowledge is required,  
+If you want to get a more detailed configuration and explanation of the KCP parameters, search for yourself. The command line name for each parameter, as well as the default and simple functions, are described as follows：  
+```
+--kcp-key="secrect"        pre-shared secret between client and server
+--kcp-method="aes"         encrypt/decrypt method, can be: aes, aes-128, aes-192, salsa20, blowfish, 
+                           twofish, cast5, 3des, tea, xtea, xor, sm4, none
+--kcp-mode="secrect"       profiles: fast3, fast2, fast, normal, manual
+--kcp-mtu=1350             set maximum transmission unit for UDP packets
+--kcp-sndwnd=1024          set send window size(num of packets)
+--kcp-rcvwnd=1024          set receive window size(num of packets)
+--kcp-ds=10                set reed-solomon erasure coding - datashard
+--kcp-ps=3                 set reed-solomon erasure coding - parityshard
+--kcp-dscp=0               set DSCP(6bit)
+--kcp-nocomp               disable compression
+--kcp-acknodelay           be carefull! flush ack immediately when a packet is received
+--kcp-nodelay=0            be carefull!
+--kcp-interval=50          be carefull!
+--kcp-resend=0             be carefull!
+--kcp-nc=0                 be carefull! no congestion
+--kcp-sockbuf=4194304      be carefull!
+--kcp-keepalive=10         be carefull!
+```
+
 ### TODO  
- Welcome adding group feedback...
+- Welcome joining group feedback...

 ### How to use the source code?   
-use command cd to enter your go SRC directory and then git clone https://github.com/snail007/goproxy.git and execute ./proxy.   
-Direct compilation: go build     
-execution: go run *.go    
-Utils is a toolkit, and service is a specific service class.  
+use command cd to enter your go SRC directory and then  
+mkdir snail007  
+cd snail007   
+execute `git clone https://github.com/snail007/goproxy.git ./proxy`     
+Direct compilation: `go build`    
+execution: `go run *.go`    
+`utils` is a toolkit, and `service` is a specific service class.  

 ### License  
 Proxy is licensed under GPLv3 license.  
--- a/README_ZH.md
+++ b/README_ZH.md
@ -1,6 +1,8 @@
 <img src="https://github.com/snail007/goproxy/blob/master/docs/images/logo.jpg?raw=true" width="200"/>  
-Proxy是golang实现的高性能http,https,websocket,tcp,udp,socks5代理服务器,支持正向代理、反向代理、透明代理、内网穿透、TCP/UDP端口映射、SSH中转，TLS加密传输。下载地址:https://github.com/snail007/goproxy/releases 官方QQ交流群:189618940  
-  
+Proxy是golang实现的高性能http,https,websocket,tcp,udp,socks5代理服务器,支持正向代理、反向代理、透明代理、内网穿透、TCP/UDP端口映射、SSH中转，TLS加密传输，协议转换。
+
+[点击下载](https://github.com/snail007/goproxy/releases) 官方QQ交流群:189618940  
+
 ---  
  
 [![stable](https://img.shields.io/badge/stable-stable-green.svg)](https://github.com/snail007/goproxy/) [![license](https://img.shields.io/github/license/snail007/goproxy.svg?style=plastic)]() [![download_count](https://img.shields.io/github/downloads/snail007/goproxy/total.svg?style=plastic)](https://github.com/snail007/goproxy/releases) [![download](https://img.shields.io/github/release/snail007/goproxy.svg?style=plastic)](https://github.com/snail007/goproxy/releases)  
@ -21,6 +23,7 @@ Proxy是golang实现的高性能http,https,websocket,tcp,udp,socks5代理服务
 - 集成外部API，HTTP(S),SOCKS5代理认证功能可以与外部HTTP API集成，可以方便的通过外部系统控制代理用户．  
 - 反向代理,支持直接把域名解析到proxy监听的ip,然后proxy就会帮你代理访问需要访问的HTTP(S)网站.
 - 透明HTTP(S)代理,配合iptables,在网关直接把出去的80,443方向的流量转发到proxy,就能实现无感知的智能路由器代理.  
+- 协议转换，可以把已经存在的HTTP(S)或SOCKS5代理转换为一个端口同时支持HTTP(S)和SOCKS5代理，转换后的SOCKS5代理不支持UDP功能,同时支持强大的级联认证功能。

 ### Why need these?  
 - 当由于某某原因,我们不能访问我们在其它地方的服务,我们可以通过多个相连的proxy节点建立起一个安全的隧道访问我们的服务.  
@ -32,7 +35,11 @@ Proxy是golang实现的高性能http,https,websocket,tcp,udp,socks5代理服务
 - ...  

 
-本页是v4.2手册,其他版本手册请点击下面链接查看.  
+本页是v4.6手册,其他版本手册请点击下面链接查看. 
+- [v4.5手册](https://github.com/snail007/goproxy/tree/v4.5) 
+- [v4.4手册](https://github.com/snail007/goproxy/tree/v4.4) 
+- [v4.3手册](https://github.com/snail007/goproxy/tree/v4.3) 
+- [v4.2手册](https://github.com/snail007/goproxy/tree/v4.2) 
 - [v4.0-v4.1手册](https://github.com/snail007/goproxy/tree/v4.1)
 - [v3.9手册](https://github.com/snail007/goproxy/tree/v3.9)
 - [v3.8手册](https://github.com/snail007/goproxy/tree/v3.8)
@ -46,8 +53,8 @@ Proxy是golang实现的高性能http,https,websocket,tcp,udp,socks5代理服务
 - [v2.x手册](https://github.com/snail007/goproxy/tree/v2.2)  

 ### 怎么找到组织?  
-[点击加入交流组织](https://gitter.im/go-proxy/Lobby?utm_source=share-link&utm_medium=link&utm_campaign=share-link)  
-
+[点击加入交流组织gitter](https://gitter.im/go-proxy/Lobby?utm_source=share-link&utm_medium=link&utm_campaign=share-link)  
+[点击加入交流组织TG](https://t.me/joinchat/GYHXghCDSBmkKZrvu4wIdQ)  
 ### 安装 
 1. [快速安装](#自动安装)
 1. [手动安装](#手动安装)
@ -76,7 +83,8 @@ Proxy是golang实现的高性能http,https,websocket,tcp,udp,socks5代理服务
    - [1.8 KCP协议传输](#18kcp协议传输)
    - [1.9 HTTP(S)反向代理](#19-https反向代理)
    - [1.10 HTTP(S)透明代理](#110-https透明代理)
-    - [1.11 查看帮助](#111查看帮助)
+    - [1.11 自定义DNS](#111-自定义dns)
+    - [1.12 查看帮助](#112-查看帮助)
 - [2. TCP代理](#2tcp代理)
    - [2.1 普通一级TCP代理](#21普通一级tcp代理)
    - [2.2 普通二级TCP代理](#22普通二级tcp代理)
@ -111,7 +119,19 @@ Proxy是golang实现的高性能http,https,websocket,tcp,udp,socks5代理服务
        - [5.6.2 用户名和密钥的方式](#562-ssh用户名和密钥的方式)
    - [5.7 认证](#57认证)
    - [5.8 KCP协议传输](#58kcp协议传输)
-    - [5.9 查看帮助](#59查看帮助)
+    - [5.9 自定义DNS](#59自定义dns)
+    - [5.10 查看帮助](#510查看帮助)
+- [6. 代理协议转换](#6代理协议转换)
+    - [6.1 功能介绍](#61-功能介绍)
+    - [6.2 HTTP(S)转HTTP(S)+SOCKS5](#62-https转httpssocks5)
+    - [6.3 SOCKS5转HTTP(S)+SOCKS5](#63-socks5转httpssocks5)
+    - [6.4 链式连接](#64-链式连接)
+    - [6.5 监听多个端口](#65-监听多个端口)
+    - [6.6 认证功能](#66-认证功能)
+    - [6.7 查看帮助](#67-查看帮助)
+- [7. KCP配置](#7kcp配置)
+    - [7.1 配置介绍](#71-配置介绍)
+    - [7.2 详细配置](#72-详细配置)

 ### Fast Start  
 提示:所有操作需要root权限.  
@ -129,7 +149,7 @@ curl -L https://raw.githubusercontent.com/snail007/goproxy/master/install_auto.s
 下载地址:https://github.com/snail007/goproxy/releases  
 ```shell  
 cd /root/proxy/  
-wget https://github.com/snail007/goproxy/releases/download/v4.2/proxy-linux-amd64.tar.gz  
+wget https://github.com/snail007/goproxy/releases/download/v4.6/proxy-linux-amd64.tar.gz  
 ```  
 #### **2.下载自动安装脚本**  
 ```shell  
@ -170,7 +190,8 @@ http,tcp,udp代理过程会和上级通讯,为了安全我们采用加密通讯,
 在linux上并安装了openssl命令，可以直接通过下面的命令生成证书和key文件.  
 `./proxy keygen`  
 默认会在当前程序目录下面生成证书文件proxy.crt和key文件proxy.key。  
-  
+更多用法:`proxy keygen usage`。    
+
 ### **后台运行**
 默认执行proxy之后,如果要保持proxy运行,不能关闭命令行.  
 如果想在后台运行proxy,命令行可以关闭,只需要在命令最后加上--daemon参数即可.  
@ -264,13 +285,13 @@ target:用户访问的URL,比如:http://demo.com:80/1.html或https://www.baidu.c
 `./proxy http -T ssh -P "2.2.2.2:22" -u user -S user.key -t tcp -p ":28080"`  

 #### **1.8.KCP协议传输**  
-KCP协议需要-B参数设置一个密码用于加密解密数据  
+KCP协议需要--kcp-key参数设置一个密码用于加密解密数据  

 一级HTTP代理(VPS,IP:22.22.22.22)  
-`./proxy http -t kcp -p ":38080" -B mypassword`  
+`./proxy http -t kcp -p ":38080" --kcp-key mypassword`  
  
 二级HTTP代理(本地Linux)  
-`./proxy http -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" -B mypassword`  
+`./proxy http -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" --kcp-key mypassword`  
 那么访问本地的8080端口就是访问VPS上面的代理端口38080,数据通过kcp协议传输.  

 #### **1.9 HTTP(S)反向代理** 
@ -333,7 +354,13 @@ iptables -t nat -A OUTPUT -p tcp -j PROXY
 - 删除指定的用户自定义链 iptables -X 链名 比如 iptables -t nat -X PROXY
 - 从所选链中删除规则 iptables -D 链名 规则详情 比如 iptables -t nat -D PROXY -d 223.223.192.0/255.255.240.0 -j RETURN

-#### **1.9.查看帮助**  
+#### **1.11 自定义DNS** 
+--dns-address和--dns-ttl参数,用于自己指定proxy访问域名的时候使用的dns（--dns-address）  
+以及解析结果缓存时间（--dns-ttl）秒数，避免系统dns对proxy的干扰，另外缓存功能还能减少dns解析时间提高访问速度.    
+比如：  
+`./proxy http -p ":33080" --dns-address "8.8.8.8:53" --dns-ttl 300`  
+
+#### **1.12 查看帮助**  
 `./proxy help http`  
  
 ### **2.TCP代理**  
@ -363,16 +390,16 @@ VPS(IP:22.22.22.33)执行:
  
 #### **2.4.加密二级TCP代理**  
 VPS(IP:22.22.22.33)执行:  
-`./proxy tcp -t tcp -p ":33080" -T tcp -P "127.0.0.1:8080" -C proxy.crt -K proxy.key`  
+`./proxy tcp -t tls -p ":33080" -T tcp -P "127.0.0.1:8080" -C proxy.crt -K proxy.key`  
 本地执行:  
 `./proxy tcp -p ":23080" -T tls -P "22.22.22.33:33080" -C proxy.crt -K proxy.key`  
 那么访问本地23080端口就是通过加密TCP隧道访问22.22.22.33的8080端口.  
  
 #### **2.5.加密三级TCP代理**  
 一级TCP代理VPS_01,IP:22.22.22.22  
-`./proxy tcp -t tcp -p ":38080" -T tcp -P "66.66.66.66:8080" -C proxy.crt -K proxy.key`  
+`./proxy tcp -t tls -p ":38080" -T tcp -P "66.66.66.66:8080" -C proxy.crt -K proxy.key`  
 二级TCP代理VPS_02,IP:33.33.33.33  
-`./proxy tcp -t tcp -p ":28080" -T tls -P "22.22.22.22:38080" -C proxy.crt -K proxy.key`  
+`./proxy tcp -t tls -p ":28080" -T tls -P "22.22.22.22:38080" -C proxy.crt -K proxy.key`  
 三级TCP代理(本地)  
 `./proxy tcp -p ":8080" -T tls -P "33.33.33.33:28080" -C proxy.crt -K proxy.key`  
 那么访问本地8080端口就是通过加密TCP隧道访问66.66.66.66的8080端口.  
@ -405,16 +432,16 @@ VPS(IP:22.22.22.33)执行:
  
 #### **3.4.加密二级UDP代理**  
 VPS(IP:22.22.22.33)执行:  
-`./proxy tcp -t tcp -p ":33080" -T udp -P "8.8.8.8:53" -C proxy.crt -K proxy.key`  
+`./proxy tcp -t tls -p ":33080" -T udp -P "8.8.8.8:53" -C proxy.crt -K proxy.key`  
 本地执行:  
 `./proxy udp -p ":5353" -T tls -P "22.22.22.33:33080" -C proxy.crt -K proxy.key`  
 那么访问本地UDP:5353端口就是通过加密TCP隧道,通过VPS访问8.8.8.8的UDP:53端口.  
  
 #### **3.5.加密三级UDP代理**  
 一级TCP代理VPS_01,IP:22.22.22.22  
-`./proxy tcp -t tcp -p ":38080" -T udp -P "8.8.8.8:53" -C proxy.crt -K proxy.key`  
+`./proxy tcp -t tls -p ":38080" -T udp -P "8.8.8.8:53" -C proxy.crt -K proxy.key`  
 二级TCP代理VPS_02,IP:33.33.33.33  
-`./proxy tcp -t tcp -p ":28080" -T tls -P "22.22.22.22:38080" -C proxy.crt -K proxy.key`  
+`./proxy tcp -t tls -p ":28080" -T tls -P "22.22.22.22:38080" -C proxy.crt -K proxy.key`  
 三级TCP代理(本地)  
 `./proxy udp -p ":5353" -T tls -P "33.33.33.33:28080" -C proxy.crt -K proxy.key`  
 那么访问本地5353端口就是通过加密TCP隧道,通过VPS_01访问8.8.8.8的53端口.  
@ -562,7 +589,7 @@ server连接到bridge的时候,如果同时有多个client连接到同一个brid
 #### **4.8.查看帮助**  
 `./proxy help bridge`  
 `./proxy help server`  
-`./proxy help server`  
+`./proxy help client`  
  
 ### **5.SOCKS5代理**  
 提示:SOCKS5代理,支持CONNECT,UDP协议,不支持BIND,支持用户名密码认证.  
@ -573,7 +600,7 @@ server连接到bridge的时候,如果同时有多个client连接到同一个brid
 ![5.2](/docs/images/5.2.png)  
 使用本地端口8090,假设上级SOCKS5代理是`22.22.22.22:8080`  
 `./proxy socks -t tcp -p "0.0.0.0:8090" -T tcp -P "22.22.22.22:8080" `  
-我们还可以指定网站域名的黑白名单文件,一行一个域名,匹配规则是最右匹配,比如:baidu.com,匹配的是*.*.baidu.com,黑名单的域名域名直接走上级代理,白名单的域名不走上级代理.  
+我们还可以指定网站域名的黑白名单文件,一行一个域名,匹配规则是最右匹配,比如:baidu.com,匹配的是*.*.baidu.com,黑名单的域名域名直接走上级代理,白名单的域名不走上级代理;如果域名即在黑名单又在白名单中,那么黑名单起作用.  
 `./proxy socks -p "0.0.0.0:8090" -T tcp -P "22.22.22.22:8080"  -b blocked.txt -d direct.txt`  
  
 #### **5.3.SOCKS二级代理(加密)**  
@ -638,29 +665,173 @@ ip:用户的IP,比如:192.168.1.200
 如果没有-a或-F或--auth-url参数,就是关闭认证.    

 #### **5.8.KCP协议传输**  
-KCP协议需要-B参数设置一个密码用于加密解密数据  
+KCP协议需要--kcp-key参数设置一个密码用于加密解密数据  

 一级HTTP代理(VPS,IP:22.22.22.22)  
-`./proxy socks -t kcp -p ":38080" -B mypassword`  
+`./proxy socks -t kcp -p ":38080" --kcp-key mypassword`  
  
 二级HTTP代理(本地Linux)  
-`./proxy socks -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" -B mypassword`  
+`./proxy socks -t tcp -p ":8080" -T kcp -P "22.22.22.22:38080" --kcp-key mypassword`  
 那么访问本地的8080端口就是访问VPS上面的代理端口38080,数据通过kcp协议传输.  

-#### **5.9.查看帮助**  
+#### **5.9.自定义DNS** 
+--dns-address和--dns-ttl参数,用于自己指定proxy访问域名的时候使用的dns（--dns-address）  
+以及解析结果缓存时间（--dns-ttl）秒数，避免系统dns对proxy的干扰，另外缓存功能还能减少dns解析时间提高访问速度.    
+比如：  
+`./proxy socks -p ":33080" --dns-address "8.8.8.8:53" --dns-ttl 300`  
+
+#### **5.10.查看帮助**  
 `./proxy help socks`  

+### **6.代理协议转换** 
+
+#### **6.1 功能介绍** 
+代理协议转换使用的是sps子命令(socks+https的缩写)，sps本身不提供代理功能，只是接受代理请求"转换并转发"给已经存在的http(s)代理或者socks5代理；sps可以把已经存在的http(s)代理或者socks5代理转换为一个端口同时支持http(s)和socks5代理，而且http(s)代理支持正向代理和反向代理(SNI)，转换后的SOCKS5代理不支持UDP功能；另外对于已经存在的http(s)代理或者socks5代理，支持tls、tcp、kcp三种模式，支持链式连接，也就是可以多个sps结点层级连接构建加密通道。
+
+#### **6.2 HTTP(S)转HTTP(S)+SOCKS5** 
+假设已经存在一个普通的http(s)代理：127.0.0.1:8080,现在我们把它转为同时支持http(s)和socks5的普通代理,转换后的本地端口为18080。  
+命令如下：  
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p :18080`
+
+假设已经存在一个tls的http(s)代理：127.0.0.1:8080,现在我们把它转为同时支持http(s)和socks5的普通代理,转换后的本地端口为18080，tls需要证书文件。  
+命令如下：  
+`./proxy sps -S http -T tls -P 127.0.0.1:8080 -t tcp -p :18080 -C proxy.crt -K proxy.key`   
+
+假设已经存在一个kcp的http(s)代理（密码是：demo123）：127.0.0.1:8080,现在我们把它转为同时支持http(s)和socks5的普通代理,转换后的本地端口为18080。  
+命令如下：  
+`./proxy sps -S http -T kcp -P 127.0.0.1:8080 -t tcp -p :18080 --kcp-key demo123`  
+
+#### **6.3 SOCKS5转HTTP(S)+SOCKS5** 
+假设已经存在一个普通的socks5代理：127.0.0.1:8080,现在我们把它转为同时支持http(s)和socks5的普通代理,转换后的本地端口为18080。  
+命令如下：  
+`./proxy sps -S socks -T tcp -P 127.0.0.1:8080 -t tcp -p :18080`
+
+假设已经存在一个tls的socks5代理：127.0.0.1:8080,现在我们把它转为同时支持http(s)和socks5的普通代理,转换后的本地端口为18080，tls需要证书文件。  
+命令如下：  
+`./proxy sps -S socks -T tls -P 127.0.0.1:8080 -t tcp -p :18080 -C proxy.crt -K proxy.key`   
+
+假设已经存在一个kcp的socks5代理（密码是：demo123）：127.0.0.1:8080,现在我们把它转为同时支持http(s)和socks5的普通代理,转换后的本地端口为18080。  
+命令如下：  
+`./proxy sps -S socks -T kcp -P 127.0.0.1:8080 -t tcp -p :18080 --kcp-key demo123`  
+
+#### **6.4 链式连接** 
+上面提过多个sps结点可以层级连接构建加密通道，假设有如下vps和家里的pc电脑。  
+vps01：2.2.2.2  
+vps02：3.3.3.3  
+现在我们想利用pc和vps01和vps02构建一个加密通道，本例子用tls加密也可以用kcp，在pc上访问本地18080端口就是访问vps01的本地8080端口。  
+首先在vps01(2.2.2.2)上我们运行一个只有本地可以访问的http(s)代理,执行：  
+`./proxy http -t tcp -p 127.0.0.1:8080`  
+
+然后在vps01(2.2.2.2)上运行一个sps结点，执行：  
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tls -p :8081 -C proxy.crt -K proxy.key`  
+
+然后在vps02(3.3.3.3)上运行一个sps结点，执行：  
+`./proxy sps -S http -T tls -P 2.2.2.2:8081 -t tls -p :8082 -C proxy.crt -K proxy.key`  
+
+然后在pc上运行一个sps结点，执行：  
+`./proxy sps -S http -T tls -P 3.3.3.3:8082 -t tcp -p :18080 -C proxy.crt -K proxy.key`  
+
+完成。  
+
+#### **6.5 监听多个端口**   
+一般情况下监听一个端口就可以，不过如果作为反向代理需要同时监听80和443两个端口，那么-p参数是支持的，  
+格式是：`-p 0.0.0.0:80,0.0.0.0:443`，多个绑定用逗号分隔即可。  
+
+#### **6.6 认证功能**   
+sps支持http(s)\socks5代理认证,可以级联认证,有四个重要的信息:  
+1:用户发送认证信息`user-auth`。   
+2:设置的本地认证信息`local-auth`。  
+3:设置的连接上级使用的认证信息`parent-auth`。  
+4:最终发送给上级的认证信息`auth-info-to-parent`。  
+他们的情况关系如下:  
+
+| user-auth | local-auth | parent-auth | auth-info-to-paren 
+| ------ | ------ | ------ | ------  
+| 有/没有  | 有     |     有   |   来自parent-auth  
+| 有/没有  | 没有    |    有    |   来自parent-auth  
+| 有/没有  | 有     |     没有  |   无  
+| 没有   | 没有    |   没有    |   无  
+| 有    | 没有    |   没有    |   来自user-auth  
+
+对于sps代理我们可以进行用户名密码认证,认证的用户名和密码可以在命令行指定    
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" -a "user1:pass1" -a "user2:pass2"`  
+多个用户,重复-a参数即可.  
+也可以放在文件中,格式是一行一个"用户名:密码",然后用-F指定.  
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" -F auth-file.txt`  
+
+如果上级有认证,下级可以通过-A参数设置认证信息,比如:  
+上级:`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" -a "user1:pass1" -a "user2:pass2"`  
+下级:`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -A "user1:pass1" -t tcp -p ":33080" `  
+
+另外,sps代理,本地认证集成了外部HTTP API认证,我们可以通过--auth-url参数指定一个http url接口地址,    
+然后有用户连接的时候,proxy会GET方式请求这url,带上下面四个参数,如果返回HTTP状态码204,代表认证成功  
+其它情况认为认证失败.  
+比如:  
+`./proxy sps -S http -T tcp -P 127.0.0.1:8080 -t tcp -p ":33080" --auth-url "http://test.com/auth.php"`  
+用户连接的时候,proxy会GET方式请求这url("http://test.com/auth.php"),  
+带上user,pass,ip,target四个参数:  
+http://test.com/auth.php?user={USER}&pass={PASS}&ip={IP}&target={TARGET}  
+user:用户名   
+pass:密码   
+ip:用户的IP,比如:192.168.1.200   
+target:如果客户端是http(s)代理请求,这里代表的是请求的完整url,其它情况为空.  
+
+如果没有-a或-F或--auth-url参数,就是关闭本地认证.  
+如果没有-A参数,连接上级不使用认证.  
+
+#### **6.7 查看帮助** 
+`./proxy help sps`  
+
+### **7.KCP配置**   
+
+#### **7.1 配置介绍**   
+proxy的很多功能都支持kcp协议，凡是使用了kcp协议的功能都支持这里介绍的配置参数。  
+所以这里统一对KCP配置参数进行介绍。  
+
+#### **7.2 详细配置**   
+所有的KCP配置参数共有17个，你可以都不用设置，他们都有默认值，如果为了或者最好的效果，  
+就需要自己根据自己根据网络情况对参数进行配置。由于kcp配置很复杂需要一定的网络基础知识，  
+如果想获得kcp参数更详细的配置和解说，请自行搜索。每个参数的命令行名称以及默认值和简单的功能说明如下：  
+```
+--kcp-key="secrect"        pre-shared secret between client and server
+--kcp-method="aes"         encrypt/decrypt method, can be: aes, aes-128, aes-192, salsa20, blowfish, 
+                           twofish, cast5, 3des, tea, xtea, xor, sm4, none
+--kcp-mode="fast"       profiles: fast3, fast2, fast, normal, manual
+--kcp-mtu=1350             set maximum transmission unit for UDP packets
+--kcp-sndwnd=1024          set send window size(num of packets)
+--kcp-rcvwnd=1024          set receive window size(num of packets)
+--kcp-ds=10                set reed-solomon erasure coding - datashard
+--kcp-ps=3                 set reed-solomon erasure coding - parityshard
+--kcp-dscp=0               set DSCP(6bit)
+--kcp-nocomp               disable compression
+--kcp-acknodelay           be carefull! flush ack immediately when a packet is received
+--kcp-nodelay=0            be carefull!
+--kcp-interval=50          be carefull!
+--kcp-resend=0             be carefull!
+--kcp-nc=0                 be carefull! no congestion
+--kcp-sockbuf=4194304      be carefull!
+--kcp-keepalive=10         be carefull!
+```
+提示：  
+参数：--kcp-mode中的四种fast3, fast2, fast, normal模式，  
+相当于设置了下面四个参数：  
+normal：`--nodelay=0 --interval=40 --resend=2 --nc=1`  
+fast ：`--nodelay=0 --interval=30 --resend=2 --nc=1`  
+fast2：`--nodelay=1 --interval=20 --resend=2 --nc=1`  
+fast3：`--nodelay=1 --interval=10 --resend=2 --nc=1`  
+
 ### TODO  
 - http,socks代理多个上级负载均衡?
 - http(s)代理增加pac支持?
 - 欢迎加群反馈...

-### 如何使用源码? 
-建议go1.8,不保证>=1.9能用.   
-cd进入你的go src目录,然后git clone https://github.com/snail007/goproxy.git ./proxy 即可.   
-编译直接:go build     
-运行: go run *.go    
-utils是工具包,service是具体的每个服务类.   
+### 如何使用源码?   
+建议go1.8.5,不保证>=1.9能用.    
+cd进入你的go src目录,新建文件夹snail007,   
+cd进入snail007,然后git clone https://github.com/snail007/goproxy.git ./proxy 即可.     
+编译直接:go build       
+运行: go run *.go      
+utils是工具包,service是具体的每个服务类.     

 ### License  
 Proxy is licensed under GPLv3 license.  
@ -673,4 +844,4 @@ QQ交流群:189618940
 <img src="https://github.com/snail007/goproxy/blob/master/docs/images/alipay.jpg?raw=true" width="200"/>  
 <img src="https://github.com/snail007/goproxy/blob/master/docs/images/wxpay.jpg?raw=true" width="200"/>  

-  
+  
--- a/config.go
+++ b/config.go
@ -2,14 +2,18 @@ package main

 import (
 	"bufio"
+	"crypto/sha1"
 	"fmt"
 	"log"
 	"os"
 	"os/exec"
-	"proxy/services"
-	"proxy/utils"
+	"snail007/proxy/services"
+	"snail007/proxy/services/kcpcfg"
+	"snail007/proxy/utils"
 	"time"

+	kcp "github.com/xtaci/kcp-go"
+	"golang.org/x/crypto/pbkdf2"
 	kingpin "gopkg.in/alecthomas/kingpin.v2"
 )

@ -39,6 +43,8 @@ func initConfig() (err error) {
 	muxBridgeArgs := services.MuxBridgeArgs{}
 	udpArgs := services.UDPArgs{}
 	socksArgs := services.SocksArgs{}
+	spsArgs := services.SPSArgs{}
+	kcpArgs := kcpcfg.KCPConfigArgs{}
 	//build srvice args
 	app = kingpin.New("proxy", "happy with proxy")
 	app.Author("snail").Version(APP_VERSION)
@ -46,10 +52,28 @@ func initConfig() (err error) {
 	daemon := app.Flag("daemon", "run proxy in background").Default("false").Bool()
 	forever := app.Flag("forever", "run proxy in forever,fail and retry").Default("false").Bool()
 	logfile := app.Flag("log", "log file path").Default("").String()
+	kcpArgs.Key = app.Flag("kcp-key", "pre-shared secret between client and server").Default("secrect").String()
+	kcpArgs.Crypt = app.Flag("kcp-method", "encrypt/decrypt method, can be: aes, aes-128, aes-192, salsa20, blowfish, twofish, cast5, 3des, tea, xtea, xor, sm4, none").Default("aes").Enum("aes", "aes-128", "aes-192", "salsa20", "blowfish", "twofish", "cast5", "3des", "tea", "xtea", "xor", "sm4", "none")
+	kcpArgs.Mode = app.Flag("kcp-mode", "profiles: fast3, fast2, fast, normal, manual").Default("fast").Enum("fast3", "fast2", "fast", "normal", "manual")
+	kcpArgs.MTU = app.Flag("kcp-mtu", "set maximum transmission unit for UDP packets").Default("1350").Int()
+	kcpArgs.SndWnd = app.Flag("kcp-sndwnd", "set send window size(num of packets)").Default("1024").Int()
+	kcpArgs.RcvWnd = app.Flag("kcp-rcvwnd", "set receive window size(num of packets)").Default("1024").Int()
+	kcpArgs.DataShard = app.Flag("kcp-ds", "set reed-solomon erasure coding - datashard").Default("10").Int()
+	kcpArgs.ParityShard = app.Flag("kcp-ps", "set reed-solomon erasure coding - parityshard").Default("3").Int()
+	kcpArgs.DSCP = app.Flag("kcp-dscp", "set DSCP(6bit)").Default("0").Int()
+	kcpArgs.NoComp = app.Flag("kcp-nocomp", "disable compression").Default("false").Bool()
+	kcpArgs.AckNodelay = app.Flag("kcp-acknodelay", "be carefull! flush ack immediately when a packet is received").Default("true").Bool()
+	kcpArgs.NoDelay = app.Flag("kcp-nodelay", "be carefull!").Default("0").Int()
+	kcpArgs.Interval = app.Flag("kcp-interval", "be carefull!").Default("50").Int()
+	kcpArgs.Resend = app.Flag("kcp-resend", "be carefull!").Default("0").Int()
+	kcpArgs.NoCongestion = app.Flag("kcp-nc", "be carefull! no congestion").Default("0").Int()
+	kcpArgs.SockBuf = app.Flag("kcp-sockbuf", "be carefull!").Default("4194304").Int()
+	kcpArgs.KeepAlive = app.Flag("kcp-keepalive", "be carefull!").Default("10").Int()

 	//########http#########
 	http := app.Command("http", "proxy on http mode")
 	httpArgs.Parent = http.Flag("parent", "parent address, such as: \"23.32.32.19:28008\"").Default("").Short('P').String()
+	httpArgs.CaCertFile = http.Flag("ca", "ca cert file for tls").Default("").String()
 	httpArgs.CertFile = http.Flag("cert", "cert file for tls").Short('C').Default("proxy.crt").String()
 	httpArgs.KeyFile = http.Flag("key", "key file for tls").Short('K').Default("proxy.key").String()
 	httpArgs.LocalType = http.Flag("local-type", "local protocol type <tls|tcp|kcp>").Default("tcp").Short('t').Enum("tls", "tcp", "kcp")
@ -69,13 +93,13 @@ func initConfig() (err error) {
 	httpArgs.SSHKeyFile = http.Flag("ssh-key", "private key file for ssh").Short('S').Default("").String()
 	httpArgs.SSHKeyFileSalt = http.Flag("ssh-keysalt", "salt of ssh private key").Short('s').Default("").String()
 	httpArgs.SSHPassword = http.Flag("ssh-password", "password for ssh").Short('A').Default("").String()
-	httpArgs.KCPKey = http.Flag("kcp-key", "key for kcp encrypt/decrypt data").Short('B').Default("encrypt").String()
-	httpArgs.KCPMethod = http.Flag("kcp-method", "kcp encrypt/decrypt method").Short('M').Default("3des").String()
 	httpArgs.LocalIPS = http.Flag("local bind ips", "if your host behind a nat,set your public ip here avoid dead loop").Short('g').Strings()
 	httpArgs.AuthURL = http.Flag("auth-url", "http basic auth username and password will send to this url,response http code equal to 'auth-code' means ok,others means fail.").Default("").String()
 	httpArgs.AuthURLTimeout = http.Flag("auth-timeout", "access 'auth-url' timeout milliseconds").Default("3000").Int()
 	httpArgs.AuthURLOkCode = http.Flag("auth-code", "access 'auth-url' success http code").Default("204").Int()
 	httpArgs.AuthURLRetry = http.Flag("auth-retry", "access 'auth-url' fail and retry count").Default("1").Int()
+	httpArgs.DNSAddress = http.Flag("dns-address", "if set this, proxy will use this dns for resolve doamin").Short('q').Default("").String()
+	httpArgs.DNSTTL = http.Flag("dns-ttl", "caching seconds of dns query result").Short('e').Default("300").Int()

 	//########tcp#########
 	tcp := app.Command("tcp", "proxy on tcp mode")
@ -88,8 +112,6 @@ func initConfig() (err error) {
 	tcpArgs.PoolSize = tcp.Flag("pool-size", "conn pool size , which connect to parent proxy, zero: means turn off pool").Short('L').Default("0").Int()
 	tcpArgs.CheckParentInterval = tcp.Flag("check-parent-interval", "check if proxy is okay every interval seconds,zero: means no check").Short('I').Default("3").Int()
 	tcpArgs.Local = tcp.Flag("local", "local ip:port to listen").Short('p').Default(":33080").String()
-	tcpArgs.KCPKey = tcp.Flag("kcp-key", "key for kcp encrypt/decrypt data").Short('B').Default("encrypt").String()
-	tcpArgs.KCPMethod = tcp.Flag("kcp-method", "kcp encrypt/decrypt method").Short('M').Default("3des").String()

 	//########udp#########
 	udp := app.Command("udp", "proxy on udp mode")
@ -105,29 +127,34 @@ func initConfig() (err error) {
 	//########mux-server#########
 	muxServer := app.Command("server", "proxy on mux server mode")
 	muxServerArgs.Parent = muxServer.Flag("parent", "parent address, such as: \"23.32.32.19:28008\"").Default("").Short('P').String()
+	muxServerArgs.ParentType = muxServer.Flag("parent-type", "parent protocol type <tls|tcp|kcp>").Default("tls").Short('T').Enum("tls", "tcp", "kcp")
 	muxServerArgs.CertFile = muxServer.Flag("cert", "cert file for tls").Short('C').Default("proxy.crt").String()
 	muxServerArgs.KeyFile = muxServer.Flag("key", "key file for tls").Short('K').Default("proxy.key").String()
-	muxServerArgs.Timeout = muxServer.Flag("timeout", "tcp timeout with milliseconds").Short('t').Default("2000").Int()
+	muxServerArgs.Timeout = muxServer.Flag("timeout", "tcp timeout with milliseconds").Short('i').Default("2000").Int()
 	muxServerArgs.IsUDP = muxServer.Flag("udp", "proxy on udp mux server mode").Default("false").Bool()
 	muxServerArgs.Key = muxServer.Flag("k", "client key").Default("default").String()
 	muxServerArgs.Route = muxServer.Flag("route", "local route to client's network, such as: PROTOCOL://LOCAL_IP:LOCAL_PORT@[CLIENT_KEY]CLIENT_LOCAL_HOST:CLIENT_LOCAL_PORT").Short('r').Default("").Strings()
-	muxServerArgs.IsCompress = muxServer.Flag("c", "compress data when tcp mode").Default("false").Bool()
+	muxServerArgs.IsCompress = muxServer.Flag("c", "compress data when tcp|tls mode").Default("false").Bool()
+	muxServerArgs.SessionCount = muxServer.Flag("session-count", "session count which connect to bridge").Short('n').Default("10").Int()

 	//########mux-client#########
 	muxClient := app.Command("client", "proxy on mux client mode")
 	muxClientArgs.Parent = muxClient.Flag("parent", "parent address, such as: \"23.32.32.19:28008\"").Default("").Short('P').String()
+	muxClientArgs.ParentType = muxClient.Flag("parent-type", "parent protocol type <tls|tcp|kcp>").Default("tls").Short('T').Enum("tls", "tcp", "kcp")
 	muxClientArgs.CertFile = muxClient.Flag("cert", "cert file for tls").Short('C').Default("proxy.crt").String()
 	muxClientArgs.KeyFile = muxClient.Flag("key", "key file for tls").Short('K').Default("proxy.key").String()
-	muxClientArgs.Timeout = muxClient.Flag("timeout", "tcp timeout with milliseconds").Short('t').Default("2000").Int()
+	muxClientArgs.Timeout = muxClient.Flag("timeout", "tcp timeout with milliseconds").Short('i').Default("2000").Int()
 	muxClientArgs.Key = muxClient.Flag("k", "key same with server").Default("default").String()
-	muxClientArgs.IsCompress = muxClient.Flag("c", "compress data when tcp mode").Default("false").Bool()
+	muxClientArgs.IsCompress = muxClient.Flag("c", "compress data when tcp|tls mode").Default("false").Bool()
+	muxClientArgs.SessionCount = muxClient.Flag("session-count", "session count which connect to bridge").Short('n').Default("10").Int()

 	//########mux-bridge#########
 	muxBridge := app.Command("bridge", "proxy on mux bridge mode")
 	muxBridgeArgs.CertFile = muxBridge.Flag("cert", "cert file for tls").Short('C').Default("proxy.crt").String()
 	muxBridgeArgs.KeyFile = muxBridge.Flag("key", "key file for tls").Short('K').Default("proxy.key").String()
-	muxBridgeArgs.Timeout = muxBridge.Flag("timeout", "tcp timeout with milliseconds").Short('t').Default("2000").Int()
+	muxBridgeArgs.Timeout = muxBridge.Flag("timeout", "tcp timeout with milliseconds").Short('i').Default("2000").Int()
 	muxBridgeArgs.Local = muxBridge.Flag("local", "local ip:port to listen").Short('p').Default(":33080").String()
+	muxBridgeArgs.LocalType = muxBridge.Flag("local-type", "local protocol type <tls|tcp|kcp>").Default("tls").Short('t').Enum("tls", "tcp", "kcp")

 	//########tunnel-server#########
 	tunnelServer := app.Command("tserver", "proxy on tunnel server mode")
@ -163,6 +190,7 @@ func initConfig() (err error) {
 	socksArgs.UDPParent = socks.Flag("udp-parent", "udp parent address, such as: \"23.32.32.19:33090\"").Default("").Short('X').String()
 	socksArgs.UDPLocal = socks.Flag("udp-local", "udp local ip:port to listen").Short('x').Default(":33090").String()
 	socksArgs.CertFile = socks.Flag("cert", "cert file for tls").Short('C').Default("proxy.crt").String()
+	socksArgs.CaCertFile = socks.Flag("ca", "ca cert file for tls").Default("").String()
 	socksArgs.KeyFile = socks.Flag("key", "key file for tls").Short('K').Default("proxy.key").String()
 	socksArgs.SSHUser = socks.Flag("ssh-user", "user for ssh").Short('u').Default("").String()
 	socksArgs.SSHKeyFile = socks.Flag("ssh-key", "private key file for ssh").Short('S').Default("").String()
@ -175,16 +203,89 @@ func initConfig() (err error) {
 	socksArgs.Direct = socks.Flag("direct", "direct domain file , one domain each line").Default("direct").Short('d').String()
 	socksArgs.AuthFile = socks.Flag("auth-file", "http basic auth file,\"username:password\" each line in file").Short('F').String()
 	socksArgs.Auth = socks.Flag("auth", "socks auth username and password, mutiple user repeat -a ,such as: -a user1:pass1 -a user2:pass2").Short('a').Strings()
-	socksArgs.KCPKey = socks.Flag("kcp-key", "key for kcp encrypt/decrypt data").Short('B').Default("encrypt").String()
-	socksArgs.KCPMethod = socks.Flag("kcp-method", "kcp encrypt/decrypt method").Short('M').Default("3des").String()
 	socksArgs.LocalIPS = socks.Flag("local bind ips", "if your host behind a nat,set your public ip here avoid dead loop").Short('g').Strings()
 	socksArgs.AuthURL = socks.Flag("auth-url", "auth username and password will send to this url,response http code equal to 'auth-code' means ok,others means fail.").Default("").String()
 	socksArgs.AuthURLTimeout = socks.Flag("auth-timeout", "access 'auth-url' timeout milliseconds").Default("3000").Int()
 	socksArgs.AuthURLOkCode = socks.Flag("auth-code", "access 'auth-url' success http code").Default("204").Int()
 	socksArgs.AuthURLRetry = socks.Flag("auth-retry", "access 'auth-url' fail and retry count").Default("0").Int()
-
+	socksArgs.DNSAddress = socks.Flag("dns-address", "if set this, proxy will use this dns for resolve doamin").Short('q').Default("").String()
+	socksArgs.DNSTTL = socks.Flag("dns-ttl", "caching seconds of dns query result").Short('e').Default("300").Int()
+	//########socks+http(s)#########
+	sps := app.Command("sps", "proxy on socks+http(s) mode")
+	spsArgs.Parent = sps.Flag("parent", "parent address, such as: \"23.32.32.19:28008\"").Default("").Short('P').String()
+	spsArgs.CertFile = sps.Flag("cert", "cert file for tls").Short('C').Default("proxy.crt").String()
+	spsArgs.KeyFile = sps.Flag("key", "key file for tls").Short('K').Default("proxy.key").String()
+	spsArgs.CaCertFile = sps.Flag("ca", "ca cert file for tls").Default("").String()
+	spsArgs.Timeout = sps.Flag("timeout", "tcp timeout milliseconds when connect to real server or parent proxy").Short('i').Default("2000").Int()
+	spsArgs.ParentType = sps.Flag("parent-type", "parent protocol type <tls|tcp|kcp>").Short('T').Enum("tls", "tcp", "kcp")
+	spsArgs.LocalType = sps.Flag("local-type", "local protocol type <tls|tcp|kcp>").Default("tcp").Short('t').Enum("tls", "tcp", "kcp")
+	spsArgs.Local = sps.Flag("local", "local ip:port to listen,multiple address use comma split,such as: 0.0.0.0:80,0.0.0.0:443").Short('p').Default(":33080").String()
+	spsArgs.ParentServiceType = sps.Flag("parent-service-type", "parent service type <http|socks>").Short('S').Enum("http", "socks")
+	spsArgs.DNSAddress = sps.Flag("dns-address", "if set this, proxy will use this dns for resolve doamin").Short('q').Default("").String()
+	spsArgs.DNSTTL = sps.Flag("dns-ttl", "caching seconds of dns query result").Short('e').Default("300").Int()
+	spsArgs.AuthFile = sps.Flag("auth-file", "http basic auth file,\"username:password\" each line in file").Short('F').String()
+	spsArgs.Auth = sps.Flag("auth", "socks auth username and password, mutiple user repeat -a ,such as: -a user1:pass1 -a user2:pass2").Short('a').Strings()
+	spsArgs.LocalIPS = sps.Flag("local bind ips", "if your host behind a nat,set your public ip here avoid dead loop").Short('g').Strings()
+	spsArgs.AuthURL = sps.Flag("auth-url", "auth username and password will send to this url,response http code equal to 'auth-code' means ok,others means fail.").Default("").String()
+	spsArgs.AuthURLTimeout = sps.Flag("auth-timeout", "access 'auth-url' timeout milliseconds").Default("3000").Int()
+	spsArgs.AuthURLOkCode = sps.Flag("auth-code", "access 'auth-url' success http code").Default("204").Int()
+	spsArgs.AuthURLRetry = sps.Flag("auth-retry", "access 'auth-url' fail and retry count").Default("0").Int()
+	spsArgs.ParentAuth = sps.Flag("parent-auth", "parent socks auth username and password, such as: -A user1:pass1").Short('A').String()
 	//parse args
 	serviceName := kingpin.MustParse(app.Parse(os.Args[1:]))
+
+	//set kcp config
+
+	switch *kcpArgs.Mode {
+	case "normal":
+		*kcpArgs.NoDelay, *kcpArgs.Interval, *kcpArgs.Resend, *kcpArgs.NoCongestion = 0, 40, 2, 1
+	case "fast":
+		*kcpArgs.NoDelay, *kcpArgs.Interval, *kcpArgs.Resend, *kcpArgs.NoCongestion = 0, 30, 2, 1
+	case "fast2":
+		*kcpArgs.NoDelay, *kcpArgs.Interval, *kcpArgs.Resend, *kcpArgs.NoCongestion = 1, 20, 2, 1
+	case "fast3":
+		*kcpArgs.NoDelay, *kcpArgs.Interval, *kcpArgs.Resend, *kcpArgs.NoCongestion = 1, 10, 2, 1
+	}
+	pass := pbkdf2.Key([]byte(*kcpArgs.Key), []byte("snail007-goproxy"), 4096, 32, sha1.New)
+
+	switch *kcpArgs.Crypt {
+	case "sm4":
+		kcpArgs.Block, _ = kcp.NewSM4BlockCrypt(pass[:16])
+	case "tea":
+		kcpArgs.Block, _ = kcp.NewTEABlockCrypt(pass[:16])
+	case "xor":
+		kcpArgs.Block, _ = kcp.NewSimpleXORBlockCrypt(pass)
+	case "none":
+		kcpArgs.Block, _ = kcp.NewNoneBlockCrypt(pass)
+	case "aes-128":
+		kcpArgs.Block, _ = kcp.NewAESBlockCrypt(pass[:16])
+	case "aes-192":
+		kcpArgs.Block, _ = kcp.NewAESBlockCrypt(pass[:24])
+	case "blowfish":
+		kcpArgs.Block, _ = kcp.NewBlowfishBlockCrypt(pass)
+	case "twofish":
+		kcpArgs.Block, _ = kcp.NewTwofishBlockCrypt(pass)
+	case "cast5":
+		kcpArgs.Block, _ = kcp.NewCast5BlockCrypt(pass[:16])
+	case "3des":
+		kcpArgs.Block, _ = kcp.NewTripleDESBlockCrypt(pass[:24])
+	case "xtea":
+		kcpArgs.Block, _ = kcp.NewXTEABlockCrypt(pass[:16])
+	case "salsa20":
+		kcpArgs.Block, _ = kcp.NewSalsa20BlockCrypt(pass)
+	default:
+		*kcpArgs.Crypt = "aes"
+		kcpArgs.Block, _ = kcp.NewAESBlockCrypt(pass)
+	}
+	//attach kcp config
+	tcpArgs.KCP = kcpArgs
+	httpArgs.KCP = kcpArgs
+	socksArgs.KCP = kcpArgs
+	spsArgs.KCP = kcpArgs
+	muxBridgeArgs.KCP = kcpArgs
+	muxServerArgs.KCP = kcpArgs
+	muxClientArgs.KCP = kcpArgs
+
 	flags := log.Ldate
 	if *debug {
 		flags |= log.Lshortfile | log.Lmicroseconds
@ -281,6 +382,7 @@ func initConfig() (err error) {
 	services.Regist("client", services.NewMuxClient(), muxClientArgs)
 	services.Regist("bridge", services.NewMuxBridge(), muxBridgeArgs)
 	services.Regist("socks", services.NewSocks(), socksArgs)
+	services.Regist("sps", services.NewSPS(), spsArgs)
 	service, err = services.Run(serviceName)
 	if err != nil {
 		log.Fatalf("run service [%s] fail, ERR:%s", serviceName, err)
--- a/install_auto.sh
+++ b/install_auto.sh
@ -5,7 +5,7 @@ if [ -e /tmp/proxy ]; then
 fi
 mkdir /tmp/proxy
 cd /tmp/proxy
-wget https://github.com/snail007/goproxy/releases/download/v4.2/proxy-linux-amd64.tar.gz
+wget https://github.com/snail007/goproxy/releases/download/v4.6/proxy-linux-amd64.tar.gz

 # #install proxy
 tar zxvf proxy-linux-amd64.tar.gz
--- a/main.go
+++ b/main.go
@ -4,11 +4,11 @@ import (
 	"log"
 	"os"
 	"os/signal"
-	"proxy/services"
+	"snail007/proxy/services"
 	"syscall"
 )

-const APP_VERSION = "4.2"
+const APP_VERSION = "4.6"

 func main() {
 	err := initConfig()
--- a/release.sh
+++ b/release.sh
@ -1,5 +1,5 @@
 #!/bin/bash
-VER="4.2"
+VER="4.6"
 RELEASE="release-${VER}"
 rm -rf .cert
 mkdir .cert
--- a/services/args.go
+++ b/services/args.go
@ -1,6 +1,10 @@
 package services

-import "golang.org/x/crypto/ssh"
+import (
+	"snail007/proxy/services/kcpcfg"
+
+	"golang.org/x/crypto/ssh"
+)

 // tcp := app.Command("tcp", "proxy on tcp mode")
 // t := tcp.Flag("tcp-timeout", "tcp timeout milliseconds when connect to real server or parent proxy").Default("2000").Int()
@ -21,39 +25,46 @@ const (
 )

 type MuxServerArgs struct {
-	Parent     *string
-	CertFile   *string
-	KeyFile    *string
-	CertBytes  []byte
-	KeyBytes   []byte
-	Local      *string
-	IsUDP      *bool
-	Key        *string
-	Remote     *string
-	Timeout    *int
-	Route      *[]string
-	Mgr        *MuxServerManager
-	IsCompress *bool
+	Parent       *string
+	ParentType   *string
+	CertFile     *string
+	KeyFile      *string
+	CertBytes    []byte
+	KeyBytes     []byte
+	Local        *string
+	IsUDP        *bool
+	Key          *string
+	Remote       *string
+	Timeout      *int
+	Route        *[]string
+	Mgr          *MuxServerManager
+	IsCompress   *bool
+	SessionCount *int
+	KCP          kcpcfg.KCPConfigArgs
 }
 type MuxClientArgs struct {
-	Parent     *string
-	CertFile   *string
-	KeyFile    *string
-	CertBytes  []byte
-	KeyBytes   []byte
-	Key        *string
-	Timeout    *int
-	IsCompress *bool
+	Parent       *string
+	ParentType   *string
+	CertFile     *string
+	KeyFile      *string
+	CertBytes    []byte
+	KeyBytes     []byte
+	Key          *string
+	Timeout      *int
+	IsCompress   *bool
+	SessionCount *int
+	KCP          kcpcfg.KCPConfigArgs
 }
 type MuxBridgeArgs struct {
-	Parent     *string
 	CertFile   *string
 	KeyFile    *string
 	CertBytes  []byte
 	KeyBytes   []byte
 	Local      *string
+	LocalType  *string
 	Timeout    *int
 	IsCompress *bool
+	KCP        kcpcfg.KCPConfigArgs
 }
 type TunnelServerArgs struct {
 	Parent    *string
@ -102,14 +113,15 @@ type TCPArgs struct {
 	Timeout             *int
 	PoolSize            *int
 	CheckParentInterval *int
-	KCPMethod           *string
-	KCPKey              *string
+	KCP                 kcpcfg.KCPConfigArgs
 }

 type HTTPArgs struct {
 	Parent              *string
 	CertFile            *string
 	KeyFile             *string
+	CaCertFile          *string
+	CaCertBytes         []byte
 	CertBytes           []byte
 	KeyBytes            []byte
 	Local               *string
@ -135,9 +147,10 @@ type HTTPArgs struct {
 	SSHUser             *string
 	SSHKeyBytes         []byte
 	SSHAuthMethod       ssh.AuthMethod
-	KCPMethod           *string
-	KCPKey              *string
+	KCP                 kcpcfg.KCPConfigArgs
 	LocalIPS            *[]string
+	DNSAddress          *string
+	DNSTTL              *int
 }
 type UDPArgs struct {
 	Parent              *string
@ -158,6 +171,8 @@ type SocksArgs struct {
 	LocalType      *string
 	CertFile       *string
 	KeyFile        *string
+	CaCertFile     *string
+	CaCertBytes    []byte
 	CertBytes      []byte
 	KeyBytes       []byte
 	SSHKeyFile     *string
@ -177,13 +192,50 @@ type SocksArgs struct {
 	AuthURLOkCode  *int
 	AuthURLTimeout *int
 	AuthURLRetry   *int
-	KCPMethod      *string
-	KCPKey         *string
+	KCP            kcpcfg.KCPConfigArgs
 	UDPParent      *string
 	UDPLocal       *string
 	LocalIPS       *[]string
+	DNSAddress     *string
+	DNSTTL         *int
+}
+type SPSArgs struct {
+	Parent            *string
+	CertFile          *string
+	KeyFile           *string
+	CaCertFile        *string
+	CaCertBytes       []byte
+	CertBytes         []byte
+	KeyBytes          []byte
+	Local             *string
+	ParentType        *string
+	LocalType         *string
+	Timeout           *int
+	KCP               kcpcfg.KCPConfigArgs
+	ParentServiceType *string
+	DNSAddress        *string
+	DNSTTL            *int
+	AuthFile          *string
+	Auth              *[]string
+	AuthURL           *string
+	AuthURLOkCode     *int
+	AuthURLTimeout    *int
+	AuthURLRetry      *int
+	LocalIPS          *[]string
+	ParentAuth        *string
 }

+func (a *SPSArgs) Protocol() string {
+	switch *a.LocalType {
+	case TYPE_TLS:
+		return TYPE_TLS
+	case TYPE_TCP:
+		return TYPE_TCP
+	case TYPE_KCP:
+		return TYPE_KCP
+	}
+	return "unknown"
+}
 func (a *TCPArgs) Protocol() string {
 	switch *a.LocalType {
 	case TYPE_TLS:
--- a/services/http.go
+++ b/services/http.go
@ -6,8 +6,8 @@ import (
 	"io/ioutil"
 	"log"
 	"net"
-	"proxy/utils"
 	"runtime/debug"
+	"snail007/proxy/utils"
 	"strconv"
 	"strings"
 	"time"
@ -16,12 +16,13 @@ import (
 )

 type HTTP struct {
-	outPool   utils.OutPool
-	cfg       HTTPArgs
-	checker   utils.Checker
-	basicAuth utils.BasicAuth
-	sshClient *ssh.Client
-	lockChn   chan bool
+	outPool        utils.OutPool
+	cfg            HTTPArgs
+	checker        utils.Checker
+	basicAuth      utils.BasicAuth
+	sshClient      *ssh.Client
+	lockChn        chan bool
+	domainResolver utils.DomainResolver
 }

 func NewHTTP() Service {
@ -36,10 +37,16 @@ func NewHTTP() Service {
 func (s *HTTP) CheckArgs() {
 	var err error
 	if *s.cfg.Parent != "" && *s.cfg.ParentType == "" {
-		log.Fatalf("parent type unkown,use -T <tls|tcp|ssh>")
+		log.Fatalf("parent type unkown,use -T <tls|tcp|ssh|kcp>")
 	}
 	if *s.cfg.ParentType == "tls" || *s.cfg.LocalType == "tls" {
 		s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+		if *s.cfg.CaCertFile != "" {
+			s.cfg.CaCertBytes, err = ioutil.ReadFile(*s.cfg.CaCertFile)
+			if err != nil {
+				log.Fatalf("read ca file error,ERR:%s", err)
+			}
+		}
 	}
 	if *s.cfg.ParentType == "ssh" {
 		if *s.cfg.SSHUser == "" {
@ -74,6 +81,9 @@ func (s *HTTP) InitService() {
 	if *s.cfg.Parent != "" {
 		s.checker = utils.NewChecker(*s.cfg.HTTPTimeout, int64(*s.cfg.Interval), *s.cfg.Blocked, *s.cfg.Direct)
 	}
+	if *s.cfg.DNSAddress != "" {
+		(*s).domainResolver = utils.NewDomainResolver(*s.cfg.DNSAddress, *s.cfg.DNSTTL)
+	}
 	if *s.cfg.ParentType == "ssh" {
 		err := s.ConnectSSH()
 		if err != nil {
@ -82,9 +92,11 @@ func (s *HTTP) InitService() {
 		go func() {
 			//循环检查ssh网络连通性
 			for {
-				conn, err := utils.ConnectHost(*s.cfg.Parent, *s.cfg.Timeout*2)
+				conn, err := utils.ConnectHost(s.Resolve(*s.cfg.Parent), *s.cfg.Timeout*2)
 				if err == nil {
+					conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 					_, err = conn.Write([]byte{0})
+					conn.SetDeadline(time.Time{})
 				}
 				if err != nil {
 					if s.sshClient != nil {
@ -124,9 +136,9 @@ func (s *HTTP) Start(args interface{}) (err error) {
 			if *s.cfg.LocalType == TYPE_TCP {
 				err = sc.ListenTCP(s.callback)
 			} else if *s.cfg.LocalType == TYPE_TLS {
-				err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, s.callback)
+				err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, s.cfg.CaCertBytes, s.callback)
 			} else if *s.cfg.LocalType == TYPE_KCP {
-				err = sc.ListenKCP(*s.cfg.KCPMethod, *s.cfg.KCPKey, s.callback)
+				err = sc.ListenKCP(s.cfg.KCP, s.callback)
 			}
 			if err != nil {
 				return
@ -166,9 +178,10 @@ func (s *HTTP) callback(inConn net.Conn) {
 		} else if *s.cfg.Always {
 			useProxy = true
 		} else {
-			s.checker.Add(address)
+			k := s.Resolve(address)
+			s.checker.Add(k)
 			//var n, m uint
-			useProxy, _, _ = s.checker.IsBlocked(req.Host)
+			useProxy, _, _ = s.checker.IsBlocked(k)
 			//log.Printf("blocked ? : %v, %s , fail:%d ,success:%d", useProxy, address, n, m)
 		}
 	}
@ -204,14 +217,14 @@ func (s *HTTP) OutToTCP(useProxy bool, address string, inConn *net.Conn, req *ut
 			if *s.cfg.ParentType == "ssh" {
 				outConn, err = s.getSSHConn(address)
 			} else {
-				//log.Printf("%v", s.outPool)
+				// log.Printf("%v", s.outPool)
 				_outConn, err = s.outPool.Pool.Get()
 				if err == nil {
 					outConn = _outConn.(net.Conn)
 				}
 			}
 		} else {
-			outConn, err = utils.ConnectHost(address, *s.cfg.Timeout)
+			outConn, err = utils.ConnectHost(s.Resolve(address), *s.cfg.Timeout)
 		}
 		tryCount++
 		if err == nil || tryCount > maxTryCount {
@ -226,16 +239,16 @@ func (s *HTTP) OutToTCP(useProxy bool, address string, inConn *net.Conn, req *ut
 		utils.CloseConn(inConn)
 		return
 	}
-
 	outAddr := outConn.RemoteAddr().String()
 	//outLocalAddr := outConn.LocalAddr().String()
-
 	if req.IsHTTPS() && (!useProxy || *s.cfg.ParentType == "ssh") {
 		//https无上级或者上级非代理,proxy需要响应connect请求,并直连目标
 		err = req.HTTPSReply()
 	} else {
 		//https或者http,上级是代理,proxy需要转发
+		outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		_, err = outConn.Write(req.HeadBuf)
+		outConn.SetDeadline(time.Time{})
 		if err != nil {
 			log.Printf("write to %s , err:%s", *s.cfg.Parent, err)
 			utils.CloseConn(inConn)
@ -304,7 +317,7 @@ func (s *HTTP) ConnectSSH() (err error) {
 	if s.sshClient != nil {
 		s.sshClient.Close()
 	}
-	s.sshClient, err = ssh.Dial("tcp", *s.cfg.Parent, &config)
+	s.sshClient, err = ssh.Dial("tcp", s.Resolve(*s.cfg.Parent), &config)
 	<-s.lockChn
 	return
 }
@ -315,10 +328,9 @@ func (s *HTTP) InitOutConnPool() {
 		s.outPool = utils.NewOutPool(
 			*s.cfg.CheckParentInterval,
 			*s.cfg.ParentType,
-			*s.cfg.KCPMethod,
-			*s.cfg.KCPKey,
-			s.cfg.CertBytes, s.cfg.KeyBytes,
-			*s.cfg.Parent,
+			s.cfg.KCP,
+			s.cfg.CertBytes, s.cfg.KeyBytes, s.cfg.CaCertBytes,
+			s.Resolve(*s.cfg.Parent),
 			*s.cfg.Timeout,
 			*s.cfg.PoolSize,
 			*s.cfg.PoolSize*2,
@ -326,7 +338,11 @@ func (s *HTTP) InitOutConnPool() {
 	}
 }
 func (s *HTTP) InitBasicAuth() (err error) {
-	s.basicAuth = utils.NewBasicAuth()
+	if *s.cfg.DNSAddress != "" {
+		s.basicAuth = utils.NewBasicAuth(&(*s).domainResolver)
+	} else {
+		s.basicAuth = utils.NewBasicAuth(nil)
+	}
 	if *s.cfg.AuthURL != "" {
 		s.basicAuth.SetAuthURL(*s.cfg.AuthURL, *s.cfg.AuthURLOkCode, *s.cfg.AuthURLTimeout, *s.cfg.AuthURLRetry)
 		log.Printf("auth from %s", *s.cfg.AuthURL)
@ -360,7 +376,11 @@ func (s *HTTP) IsDeadLoop(inLocalAddr string, host string) bool {
 	}
 	if inPort == outPort {
 		var outIPs []net.IP
-		outIPs, err = net.LookupIP(outDomain)
+		if *s.cfg.DNSAddress != "" {
+			outIPs = []net.IP{net.ParseIP(s.Resolve(outDomain))}
+		} else {
+			outIPs, err = net.LookupIP(outDomain)
+		}
 		if err == nil {
 			for _, ip := range outIPs {
 				if ip.String() == inIP {
@ -384,3 +404,13 @@ func (s *HTTP) IsDeadLoop(inLocalAddr string, host string) bool {
 	}
 	return false
 }
+func (s *HTTP) Resolve(address string) string {
+	if *s.cfg.DNSAddress == "" {
+		return address
+	}
+	ip, err := s.domainResolver.Resolve(address)
+	if err != nil {
+		log.Printf("dns error %s , ERR:%s", address, err)
+	}
+	return ip
+}
--- a/services/kcpcfg/args.go
+++ b/services/kcpcfg/args.go
@ -0,0 +1,24 @@
+package kcpcfg
+
+import kcp "github.com/xtaci/kcp-go"
+
+type KCPConfigArgs struct {
+	Key          *string
+	Crypt        *string
+	Mode         *string
+	MTU          *int
+	SndWnd       *int
+	RcvWnd       *int
+	DataShard    *int
+	ParityShard  *int
+	DSCP         *int
+	NoComp       *bool
+	AckNodelay   *bool
+	NoDelay      *int
+	Interval     *int
+	Resend       *int
+	NoCongestion *int
+	SockBuf      *int
+	KeepAlive    *int
+	Block        kcp.BlockCrypt
+}
--- a/services/mux_bridge.go
+++ b/services/mux_bridge.go
@ -4,9 +4,12 @@ import (
 	"bufio"
 	"io"
 	"log"
+	"math/rand"
 	"net"
-	"proxy/utils"
+	"snail007/proxy/utils"
 	"strconv"
+	"strings"
+	"sync"
 	"time"

 	"github.com/xtaci/smux"
@ -15,13 +18,18 @@ import (
 type MuxBridge struct {
 	cfg                MuxBridgeArgs
 	clientControlConns utils.ConcurrentMap
+	router             utils.ClientKeyRouter
+	l                  *sync.Mutex
 }

 func NewMuxBridge() Service {
-	return &MuxBridge{
+	b := &MuxBridge{
 		cfg:                MuxBridgeArgs{},
 		clientControlConns: utils.NewConcurrentMap(),
+		l:                  &sync.Mutex{},
 	}
+	b.router = utils.NewClientKeyRouter(&b.clientControlConns, 50000)
+	return b
 }

 func (s *MuxBridge) InitService() {
@ -31,7 +39,9 @@ func (s *MuxBridge) CheckArgs() {
 	if *s.cfg.CertFile == "" || *s.cfg.KeyFile == "" {
 		log.Fatalf("cert and key file required")
 	}
-	s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+	if *s.cfg.LocalType == TYPE_TLS {
+		s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+	}
 }
 func (s *MuxBridge) StopService() {

@ -43,74 +53,110 @@ func (s *MuxBridge) Start(args interface{}) (err error) {
 	host, port, _ := net.SplitHostPort(*s.cfg.Local)
 	p, _ := strconv.Atoi(port)
 	sc := utils.NewServerChannel(host, p)
-
-	err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, func(inConn net.Conn) {
-		reader := bufio.NewReader(inConn)
-
-		var err error
-		var connType uint8
-		var key string
-		err = utils.ReadPacket(reader, &connType, &key)
-		if err != nil {
-			log.Printf("read error,ERR:%s", err)
-			return
-		}
-		switch connType {
-		case CONN_SERVER:
-			var serverID string
-			err = utils.ReadPacketData(reader, &serverID)
-			if err != nil {
-				log.Printf("read error,ERR:%s", err)
-				return
-			}
-			log.Printf("server connection %s %s connected", serverID, key)
-			session, err := smux.Server(inConn, nil)
-			if err != nil {
-				utils.CloseConn(&inConn)
-				log.Printf("server session error,ERR:%s", err)
-				return
-			}
-			for {
-				stream, err := session.AcceptStream()
-				if err != nil {
-					session.Close()
-					utils.CloseConn(&inConn)
-					return
-				}
-				go s.callback(stream, serverID, key)
-			}
-		case CONN_CLIENT:
-
-			log.Printf("client connection %s connected", key)
-			session, err := smux.Client(inConn, nil)
-			if err != nil {
-				utils.CloseConn(&inConn)
-				log.Printf("client session error,ERR:%s", err)
-				return
-			}
-			s.clientControlConns.Set(key, session)
-			go func() {
-				for {
-					if session.IsClosed() {
-						s.clientControlConns.Remove(key)
-						break
-					}
-					time.Sleep(time.Second * 5)
-				}
-			}()
-			//log.Printf("set client session,key: %s", key)
-		}
-
-	})
+	if *s.cfg.LocalType == TYPE_TCP {
+		err = sc.ListenTCP(s.handler)
+	} else if *s.cfg.LocalType == TYPE_TLS {
+		err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, nil, s.handler)
+	} else if *s.cfg.LocalType == TYPE_KCP {
+		err = sc.ListenKCP(s.cfg.KCP, s.handler)
+	}
 	if err != nil {
 		return
 	}
-	log.Printf("proxy on mux bridge mode %s", (*sc.Listener).Addr())
+	log.Printf("%s bridge on %s", *s.cfg.LocalType, (*sc.Listener).Addr())
 	return
 }
 func (s *MuxBridge) Clean() {
 	s.StopService()
 }
+func (s *MuxBridge) handler(inConn net.Conn) {
+	reader := bufio.NewReader(inConn)
+
+	var err error
+	var connType uint8
+	var key string
+	inConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
+	err = utils.ReadPacket(reader, &connType, &key)
+	inConn.SetDeadline(time.Time{})
+	if err != nil {
+		log.Printf("read error,ERR:%s", err)
+		return
+	}
+	switch connType {
+	case CONN_SERVER:
+		var serverID string
+		inConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
+		err = utils.ReadPacketData(reader, &serverID)
+		inConn.SetDeadline(time.Time{})
+		if err != nil {
+			log.Printf("read error,ERR:%s", err)
+			return
+		}
+		log.Printf("server connection %s %s connected", serverID, key)
+		session, err := smux.Server(inConn, nil)
+		if err != nil {
+			utils.CloseConn(&inConn)
+			log.Printf("server session error,ERR:%s", err)
+			return
+		}
+		for {
+			stream, err := session.AcceptStream()
+			if err != nil {
+				session.Close()
+				utils.CloseConn(&inConn)
+				return
+			}
+			go s.callback(stream, serverID, key)
+		}
+	case CONN_CLIENT:
+		log.Printf("client connection %s connected", key)
+		session, err := smux.Client(inConn, nil)
+		if err != nil {
+			utils.CloseConn(&inConn)
+			log.Printf("client session error,ERR:%s", err)
+			return
+		}
+		keyInfo := strings.Split(key, "-")
+		if len(keyInfo) != 2 {
+			utils.CloseConn(&inConn)
+			log.Printf("client key format error,key:%s", key)
+			return
+		}
+		groupKey := keyInfo[0]
+		index := keyInfo[1]
+		s.l.Lock()
+		defer s.l.Unlock()
+		if !s.clientControlConns.Has(groupKey) {
+			item := utils.NewConcurrentMap()
+			s.clientControlConns.Set(groupKey, &item)
+		}
+		_group, _ := s.clientControlConns.Get(groupKey)
+		group := _group.(*utils.ConcurrentMap)
+		if v, ok := group.Get(index); ok {
+			v.(*smux.Session).Close()
+		}
+		group.Set(index, session)
+		// s.clientControlConns.Set(key, session)
+		go func() {
+			for {
+				if session.IsClosed() {
+					s.l.Lock()
+					defer s.l.Unlock()
+					if sess, ok := group.Get(index); ok && sess.(*smux.Session).IsClosed() {
+						group.Remove(index)
+					}
+					if group.IsEmpty() {
+						s.clientControlConns.Remove(groupKey)
+					}
+					break
+				}
+				time.Sleep(time.Second * 5)
+			}
+		}()
+		//log.Printf("set client session,key: %s", key)
+	}
+
+}
 func (s *MuxBridge) callback(inConn net.Conn, serverID, key string) {
 	try := 20
 	for {
@ -118,19 +164,38 @@ func (s *MuxBridge) callback(inConn net.Conn, serverID, key string) {
 		if try == 0 {
 			break
 		}
-		session, ok := s.clientControlConns.Get(key)
+		if key == "*" {
+			key = s.router.GetKey()
+		}
+		_group, ok := s.clientControlConns.Get(key)
 		if !ok {
-			log.Printf("client %s session not exists for server stream %s", key, serverID)
+			log.Printf("client %s session not exists for server stream %s, retrying...", key, serverID)
 			time.Sleep(time.Second * 3)
 			continue
 		}
+		group := _group.(*utils.ConcurrentMap)
+		keys := group.Keys()
+		keysLen := len(keys)
+		i := 0
+		if keysLen > 0 {
+			i = rand.Intn(keysLen)
+		} else {
+			log.Printf("client %s session empty for server stream %s, retrying...", key, serverID)
+			time.Sleep(time.Second * 3)
+			continue
+		}
+		index := keys[i]
+		log.Printf("select client : %s-%s", key, index)
+		session, _ := group.Get(index)
+		session.(*smux.Session).SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		stream, err := session.(*smux.Session).OpenStream()
+		session.(*smux.Session).SetDeadline(time.Time{})
 		if err != nil {
 			log.Printf("%s client session open stream %s fail, err: %s, retrying...", key, serverID, err)
 			time.Sleep(time.Second * 3)
 			continue
 		} else {
-			log.Printf("%s server %s stream created", key, serverID)
+			log.Printf("stream %s -> %s created", serverID, key)
 			die1 := make(chan bool, 1)
 			die2 := make(chan bool, 1)
 			go func() {
--- a/services/mux_client.go
+++ b/services/mux_client.go
@ -2,10 +2,11 @@ package services

 import (
 	"crypto/tls"
+	"fmt"
 	"io"
 	"log"
 	"net"
-	"proxy/utils"
+	"snail007/proxy/utils"
 	"time"

 	"github.com/golang/snappy"
@ -35,7 +36,9 @@ func (s *MuxClient) CheckArgs() {
 	if *s.cfg.CertFile == "" || *s.cfg.KeyFile == "" {
 		log.Fatalf("cert and key file required")
 	}
-	s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+	if *s.cfg.ParentType == "tls" {
+		s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+	}
 }
 func (s *MuxClient) StopService() {

@ -44,67 +47,106 @@ func (s *MuxClient) Start(args interface{}) (err error) {
 	s.cfg = args.(MuxClientArgs)
 	s.CheckArgs()
 	s.InitService()
-	log.Printf("proxy on mux client mode, compress %v", *s.cfg.IsCompress)
-	for {
-		var _conn tls.Conn
-		_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes)
-		if err != nil {
-			log.Printf("connection err: %s, retrying...", err)
-			time.Sleep(time.Second * 3)
-			continue
-		}
-		conn := net.Conn(&_conn)
-		_, err = conn.Write(utils.BuildPacket(CONN_CLIENT, *s.cfg.Key))
-		if err != nil {
-			conn.Close()
-			log.Printf("connection err: %s, retrying...", err)
-			time.Sleep(time.Second * 3)
-			continue
-		}
-		session, err := smux.Server(conn, nil)
-		if err != nil {
-			log.Printf("session err: %s, retrying...", err)
-			conn.Close()
-			time.Sleep(time.Second * 3)
-			continue
-		}
-		for {
-			stream, err := session.AcceptStream()
-			if err != nil {
-				log.Printf("accept stream err: %s, retrying...", err)
-				session.Close()
-				time.Sleep(time.Second * 3)
-				break
-			}
-			go func() {
-				var ID, clientLocalAddr, serverID string
-				err = utils.ReadPacketData(stream, &ID, &clientLocalAddr, &serverID)
-				if err != nil {
-					log.Printf("read stream signal err: %s", err)
-					stream.Close()
-					return
-				}
-				log.Printf("signal revecived,server %s stream %s %s", serverID, ID, clientLocalAddr)
-				protocol := clientLocalAddr[:3]
-				localAddr := clientLocalAddr[4:]
-				if protocol == "udp" {
-					s.ServeUDP(stream, localAddr, ID)
-				} else {
-					s.ServeConn(stream, localAddr, ID)
+	log.Printf("client started")
+	count := 1
+	if *s.cfg.SessionCount > 0 {
+		count = *s.cfg.SessionCount
+	}
+	for i := 1; i <= count; i++ {
+		log.Printf("session worker[%d] started", i)
+		go func(i int) {
+			defer func() {
+				e := recover()
+				if e != nil {
+					log.Printf("session worker crashed: %s", e)
 				}
 			}()
-		}
-	}
+			for {
+				conn, err := s.getParentConn()
+				if err != nil {
+					log.Printf("connection err: %s, retrying...", err)
+					time.Sleep(time.Second * 3)
+					continue
+				}
+				conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
+				_, err = conn.Write(utils.BuildPacket(CONN_CLIENT, fmt.Sprintf("%s-%d", *s.cfg.Key, i)))
+				conn.SetDeadline(time.Time{})
+				if err != nil {
+					conn.Close()
+					log.Printf("connection err: %s, retrying...", err)
+					time.Sleep(time.Second * 3)
+					continue
+				}
+				session, err := smux.Server(conn, nil)
+				if err != nil {
+					log.Printf("session err: %s, retrying...", err)
+					conn.Close()
+					time.Sleep(time.Second * 3)
+					continue
+				}
+				for {
+					stream, err := session.AcceptStream()
+					if err != nil {
+						log.Printf("accept stream err: %s, retrying...", err)
+						session.Close()
+						time.Sleep(time.Second * 3)
+						break
+					}
+					go func() {
+						defer func() {
+							e := recover()
+							if e != nil {
+								log.Printf("stream handler crashed: %s", e)
+							}
+						}()
+						var ID, clientLocalAddr, serverID string
+						stream.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
+						err = utils.ReadPacketData(stream, &ID, &clientLocalAddr, &serverID)
+						stream.SetDeadline(time.Time{})
+						if err != nil {
+							log.Printf("read stream signal err: %s", err)
+							stream.Close()
+							return
+						}
+						log.Printf("worker[%d] signal revecived,server %s stream %s %s", i, serverID, ID, clientLocalAddr)
+						protocol := clientLocalAddr[:3]
+						localAddr := clientLocalAddr[4:]
+						if protocol == "udp" {
+							s.ServeUDP(stream, localAddr, ID)
+						} else {
+							s.ServeConn(stream, localAddr, ID)
+						}
+					}()
+				}
+			}

+		}(i)
+	}
+	return
 }
 func (s *MuxClient) Clean() {
 	s.StopService()
 }
-
+func (s *MuxClient) getParentConn() (conn net.Conn, err error) {
+	if *s.cfg.ParentType == "tls" {
+		var _conn tls.Conn
+		_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes, nil)
+		if err == nil {
+			conn = net.Conn(&_conn)
+		}
+	} else if *s.cfg.ParentType == "kcp" {
+		conn, err = utils.ConnectKCPHost(*s.cfg.Parent, s.cfg.KCP)
+	} else {
+		conn, err = utils.ConnectHost(*s.cfg.Parent, *s.cfg.Timeout)
+	}
+	return
+}
 func (s *MuxClient) ServeUDP(inConn *smux.Stream, localAddr, ID string) {

 	for {
+		inConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		srcAddr, body, err := utils.ReadUDPPacket(inConn)
+		inConn.SetDeadline(time.Time{})
 		if err != nil {
 			log.Printf("udp packet revecived fail, err: %s", err)
 			log.Printf("connection %s released", ID)
@ -133,13 +175,16 @@ func (s *MuxClient) processUDPPacket(inConn *smux.Stream, srcAddr, localAddr str
 	}
 	conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 	_, err = conn.Write(body)
+	conn.SetDeadline(time.Time{})
 	if err != nil {
 		log.Printf("send udp packet to %s fail,ERR:%s", dstAddr.String(), err)
 		return
 	}
 	//log.Printf("send udp packet to %s success", dstAddr.String())
 	buf := make([]byte, 1024)
+	conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 	length, _, err := conn.ReadFromUDP(buf)
+	conn.SetDeadline(time.Time{})
 	if err != nil {
 		log.Printf("read udp response from %s fail ,ERR:%s", dstAddr.String(), err)
 		return
@ -147,7 +192,9 @@ func (s *MuxClient) processUDPPacket(inConn *smux.Stream, srcAddr, localAddr str
 	respBody := buf[0:length]
 	//log.Printf("revecived udp packet from %s , %v", dstAddr.String(), respBody)
 	bs := utils.UDPPacket(srcAddr, respBody)
+	(*inConn).SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 	_, err = (*inConn).Write(bs)
+	(*inConn).SetDeadline(time.Time{})
 	if err != nil {
 		log.Printf("send udp response fail ,ERR:%s", err)
 		inConn.Close()
--- a/services/mux_server.go
+++ b/services/mux_server.go
@ -5,9 +5,10 @@ import (
 	"fmt"
 	"io"
 	"log"
+	"math/rand"
 	"net"
-	"proxy/utils"
 	"runtime/debug"
+	"snail007/proxy/utils"
 	"strconv"
 	"strings"
 	"time"
@ -17,11 +18,11 @@ import (
 )

 type MuxServer struct {
-	cfg     MuxServerArgs
-	udpChn  chan MuxUDPItem
-	sc      utils.ServerChannel
-	session *smux.Session
-	lockChn chan bool
+	cfg      MuxServerArgs
+	udpChn   chan MuxUDPItem
+	sc       utils.ServerChannel
+	sessions utils.ConcurrentMap
+	lockChn  chan bool
 }

 type MuxServerManager struct {
@ -42,7 +43,7 @@ func (s *MuxServerManager) Start(args interface{}) (err error) {
 	s.cfg = args.(MuxServerArgs)
 	s.CheckArgs()
 	if *s.cfg.Parent != "" {
-		log.Printf("use tls parent %s", *s.cfg.Parent)
+		log.Printf("use %s parent %s", *s.cfg.ParentType, *s.cfg.Parent)
 	} else {
 		log.Fatalf("parent required")
 	}
@ -74,18 +75,21 @@ func (s *MuxServerManager) Start(args interface{}) (err error) {
 			remote = fmt.Sprintf("127.0.0.1%s", remote)
 		}
 		err = server.Start(MuxServerArgs{
-			CertBytes:  s.cfg.CertBytes,
-			KeyBytes:   s.cfg.KeyBytes,
-			Parent:     s.cfg.Parent,
-			CertFile:   s.cfg.CertFile,
-			KeyFile:    s.cfg.KeyFile,
-			Local:      &local,
-			IsUDP:      &IsUDP,
-			Remote:     &remote,
-			Key:        &KEY,
-			Timeout:    s.cfg.Timeout,
-			Mgr:        s,
-			IsCompress: s.cfg.IsCompress,
+			CertBytes:    s.cfg.CertBytes,
+			KeyBytes:     s.cfg.KeyBytes,
+			Parent:       s.cfg.Parent,
+			CertFile:     s.cfg.CertFile,
+			KeyFile:      s.cfg.KeyFile,
+			Local:        &local,
+			IsUDP:        &IsUDP,
+			Remote:       &remote,
+			Key:          &KEY,
+			Timeout:      s.cfg.Timeout,
+			Mgr:          s,
+			IsCompress:   s.cfg.IsCompress,
+			SessionCount: s.cfg.SessionCount,
+			KCP:          s.cfg.KCP,
+			ParentType:   s.cfg.ParentType,
 		})

 		if err != nil {
@ -103,16 +107,19 @@ func (s *MuxServerManager) CheckArgs() {
 	if *s.cfg.CertFile == "" || *s.cfg.KeyFile == "" {
 		log.Fatalf("cert and key file required")
 	}
-	s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+	if *s.cfg.ParentType == "tls" {
+		s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+	}
 }
 func (s *MuxServerManager) InitService() {
 }

 func NewMuxServer() Service {
 	return &MuxServer{
-		cfg:     MuxServerArgs{},
-		udpChn:  make(chan MuxUDPItem, 50000),
-		lockChn: make(chan bool, 1),
+		cfg:      MuxServerArgs{},
+		udpChn:   make(chan MuxUDPItem, 50000),
+		lockChn:  make(chan bool, 1),
+		sessions: utils.NewConcurrentMap(),
 	}
 }

@ -149,7 +156,7 @@ func (s *MuxServer) Start(args interface{}) (err error) {
 		if err != nil {
 			return
 		}
-		log.Printf("proxy on udp mux server mode %s", (*s.sc.UDPListener).LocalAddr())
+		log.Printf("server on %s", (*s.sc.UDPListener).LocalAddr())
 	} else {
 		err = s.sc.ListenTCP(func(inConn net.Conn) {
 			defer func() {
@ -198,7 +205,7 @@ func (s *MuxServer) Start(args interface{}) (err error) {
 		if err != nil {
 			return
 		}
-		log.Printf("proxy on mux server mode %s, compress %v", (*s.sc.Listener).Addr(), *s.cfg.IsCompress)
+		log.Printf("server on %s", (*s.sc.Listener).Addr())
 	}
 	return
 }
@ -206,7 +213,11 @@ func (s *MuxServer) Clean() {

 }
 func (s *MuxServer) GetOutConn() (outConn net.Conn, ID string, err error) {
-	outConn, err = s.GetConn()
+	i := 1
+	if *s.cfg.SessionCount > 0 {
+		i = rand.Intn(*s.cfg.SessionCount)
+	}
+	outConn, err = s.GetConn(fmt.Sprintf("%d", i))
 	if err != nil {
 		log.Printf("connection err: %s", err)
 		return
@ -216,7 +227,9 @@ func (s *MuxServer) GetOutConn() (outConn net.Conn, ID string, err error) {
 		remoteAddr = "udp:" + *s.cfg.Remote
 	}
 	ID = utils.Uniqueid()
+	outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 	_, err = outConn.Write(utils.BuildPacketData(ID, remoteAddr, s.cfg.Mgr.serverID))
+	outConn.SetDeadline(time.Time{})
 	if err != nil {
 		log.Printf("write stream data err: %s ,retrying...", err)
 		utils.CloseConn(&outConn)
@ -224,7 +237,7 @@ func (s *MuxServer) GetOutConn() (outConn net.Conn, ID string, err error) {
 	}
 	return
 }
-func (s *MuxServer) GetConn() (conn net.Conn, err error) {
+func (s *MuxServer) GetConn(index string) (conn net.Conn, err error) {
 	select {
 	case s.lockChn <- true:
 	default:
@ -234,34 +247,60 @@ func (s *MuxServer) GetConn() (conn net.Conn, err error) {
 	defer func() {
 		<-s.lockChn
 	}()
-	if s.session == nil {
-		var _conn tls.Conn
-		_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes)
+	var session *smux.Session
+	_session, ok := s.sessions.Get(index)
+	if !ok {
+		var c net.Conn
+		c, err = s.getParentConn()
 		if err != nil {
-			s.session = nil
 			return
 		}
-		c := net.Conn(&_conn)
+		c.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		_, err = c.Write(utils.BuildPacket(CONN_SERVER, *s.cfg.Key, s.cfg.Mgr.serverID))
+		c.SetDeadline(time.Time{})
 		if err != nil {
 			c.Close()
-			s.session = nil
 			return
 		}
 		if err == nil {
-			s.session, err = smux.Client(c, nil)
+			session, err = smux.Client(c, nil)
 			if err != nil {
-				s.session = nil
 				return
 			}
 		}
+		s.sessions.Set(index, session)
+		log.Printf("session[%s] created", index)
+		go func() {
+			for {
+				if session.IsClosed() {
+					s.sessions.Remove(index)
+					break
+				}
+				time.Sleep(time.Second * 5)
+			}
+		}()
+	} else {
+		session = _session.(*smux.Session)
 	}
-	conn, err = s.session.OpenStream()
+	conn, err = session.OpenStream()
 	if err != nil {
-		s.session.Close()
-		s.session = nil
+		session.Close()
+		s.sessions.Remove(index)
+	}
+	return
+}
+func (s *MuxServer) getParentConn() (conn net.Conn, err error) {
+	if *s.cfg.ParentType == "tls" {
+		var _conn tls.Conn
+		_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes, nil)
+		if err == nil {
+			conn = net.Conn(&_conn)
+		}
+	} else if *s.cfg.ParentType == "kcp" {
+		conn, err = utils.ConnectKCPHost(*s.cfg.Parent, s.cfg.KCP)
+	} else {
+		conn, err = utils.ConnectHost(*s.cfg.Parent, *s.cfg.Timeout)
 	}
-
 	return
 }
 func (s *MuxServer) UDPConnDeamon() {
@ -292,7 +331,9 @@ func (s *MuxServer) UDPConnDeamon() {
 								// outConn.Close()
 							}()
 							for {
+								outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 								srcAddrFromConn, body, err := utils.ReadUDPPacket(outConn)
+								outConn.SetDeadline(time.Time{})
 								if err != nil {
 									log.Printf("parse revecived udp packet fail, err: %s ,%v", err, body)
 									log.Printf("UDP deamon connection %s exited", ID)
@ -306,7 +347,9 @@ func (s *MuxServer) UDPConnDeamon() {
 								}
 								port, _ := strconv.Atoi(_srcAddr[1])
 								dstAddr := &net.UDPAddr{IP: net.ParseIP(_srcAddr[0]), Port: port}
+								s.sc.UDPListener.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 								_, err = s.sc.UDPListener.WriteToUDP(body, dstAddr)
+								s.sc.UDPListener.SetDeadline(time.Time{})
 								if err != nil {
 									log.Printf("udp response to local %s fail,ERR:%s", srcAddrFromConn, err)
 									continue
--- a/services/socks.go
+++ b/services/socks.go
@ -6,10 +6,10 @@ import (
 	"io/ioutil"
 	"log"
 	"net"
-	"proxy/utils"
-	"proxy/utils/aes"
-	"proxy/utils/socks"
 	"runtime/debug"
+	"snail007/proxy/utils"
+	"snail007/proxy/utils/aes"
+	"snail007/proxy/utils/socks"
 	"strings"
 	"time"

@ -17,12 +17,13 @@ import (
 )

 type Socks struct {
-	cfg       SocksArgs
-	checker   utils.Checker
-	basicAuth utils.BasicAuth
-	sshClient *ssh.Client
-	lockChn   chan bool
-	udpSC     utils.ServerChannel
+	cfg            SocksArgs
+	checker        utils.Checker
+	basicAuth      utils.BasicAuth
+	sshClient      *ssh.Client
+	lockChn        chan bool
+	udpSC          utils.ServerChannel
+	domainResolver utils.DomainResolver
 }

 func NewSocks() Service {
@ -36,16 +37,28 @@ func NewSocks() Service {

 func (s *Socks) CheckArgs() {
 	var err error
-	if *s.cfg.LocalType == "tls" {
+	if *s.cfg.LocalType == "tls" || (*s.cfg.Parent != "" && *s.cfg.ParentType == "tls") {
 		s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+		if *s.cfg.CaCertFile != "" {
+			s.cfg.CaCertBytes, err = ioutil.ReadFile(*s.cfg.CaCertFile)
+			if err != nil {
+				log.Fatalf("read ca file error,ERR:%s", err)
+			}
+		}
 	}
 	if *s.cfg.Parent != "" {
 		if *s.cfg.ParentType == "" {
-			log.Fatalf("parent type unkown,use -T <tls|tcp|ssh>")
-		}
-		if *s.cfg.ParentType == "tls" {
-			s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+			log.Fatalf("parent type unkown,use -T <tls|tcp|ssh|kcp>")
 		}
+		// if *s.cfg.ParentType == "tls" {
+		// 	s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+		// 	if *s.cfg.CaCertFile != "" {
+		// 		s.cfg.CaCertBytes, err = ioutil.ReadFile(*s.cfg.CaCertFile)
+		// 		if err != nil {
+		// 			log.Fatalf("read ca file error,ERR:%s", err)
+		// 		}
+		// 	}
+		// }
 		if *s.cfg.ParentType == "ssh" {
 			if *s.cfg.SSHUser == "" {
 				log.Fatalf("ssh user required")
@ -77,6 +90,9 @@ func (s *Socks) CheckArgs() {
 }
 func (s *Socks) InitService() {
 	s.InitBasicAuth()
+	if *s.cfg.DNSAddress != "" {
+		(*s).domainResolver = utils.NewDomainResolver(*s.cfg.DNSAddress, *s.cfg.DNSTTL)
+	}
 	s.checker = utils.NewChecker(*s.cfg.Timeout, int64(*s.cfg.Interval), *s.cfg.Blocked, *s.cfg.Direct)
 	if *s.cfg.ParentType == "ssh" {
 		err := s.ConnectSSH()
@ -86,9 +102,11 @@ func (s *Socks) InitService() {
 		go func() {
 			//循环检查ssh网络连通性
 			for {
-				conn, err := utils.ConnectHost(*s.cfg.Parent, *s.cfg.Timeout*2)
+				conn, err := utils.ConnectHost(s.Resolve(*s.cfg.Parent), *s.cfg.Timeout*2)
 				if err == nil {
+					conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 					_, err = conn.Write([]byte{0})
+					conn.SetDeadline(time.Time{})
 				}
 				if err != nil {
 					if s.sshClient != nil {
@ -106,7 +124,6 @@ func (s *Socks) InitService() {
 	if *s.cfg.ParentType == "ssh" {
 		log.Println("warn: socks udp not suppored for ssh")
 	} else {
-
 		s.udpSC = utils.NewServerChannelHost(*s.cfg.UDPLocal)
 		err := s.udpSC.ListenUDP(s.udpCallback)
 		if err != nil {
@ -135,9 +152,9 @@ func (s *Socks) Start(args interface{}) (err error) {
 	if *s.cfg.LocalType == TYPE_TCP {
 		err = sc.ListenTCP(s.socksConnCallback)
 	} else if *s.cfg.LocalType == TYPE_TLS {
-		err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, s.socksConnCallback)
+		err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, nil, s.socksConnCallback)
 	} else if *s.cfg.LocalType == TYPE_KCP {
-		err = sc.ListenKCP(*s.cfg.KCPMethod, *s.cfg.KCPKey, s.socksConnCallback)
+		err = sc.ListenKCP(s.cfg.KCP, s.socksConnCallback)
 	}
 	if err != nil {
 		return
@ -188,7 +205,7 @@ func (s *Socks) udpCallback(b []byte, localAddr, srcAddr *net.UDPAddr) {
 		if parent == "" {
 			parent = *s.cfg.Parent
 		}
-		dstAddr, err := net.ResolveUDPAddr("udp", parent)
+		dstAddr, err := net.ResolveUDPAddr("udp", s.Resolve(parent))
 		if err != nil {
 			log.Printf("can't resolve address: %s", err)
 			return
@ -201,6 +218,7 @@ func (s *Socks) udpCallback(b []byte, localAddr, srcAddr *net.UDPAddr) {
 		}
 		conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout*5)))
 		_, err = conn.Write(rawB)
+		conn.SetDeadline(time.Time{})
 		log.Printf("udp request:%v", len(rawB))
 		if err != nil {
 			log.Printf("send udp packet to %s fail,ERR:%s", dstAddr.String(), err)
@ -210,7 +228,9 @@ func (s *Socks) udpCallback(b []byte, localAddr, srcAddr *net.UDPAddr) {

 		//log.Printf("send udp packet to %s success", dstAddr.String())
 		buf := make([]byte, 10*1024)
+		conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		length, _, err := conn.ReadFromUDP(buf)
+		conn.SetDeadline(time.Time{})
 		if err != nil {
 			log.Printf("read udp response from %s fail ,ERR:%s", dstAddr.String(), err)
 			conn.Close()
@ -235,16 +255,20 @@ func (s *Socks) udpCallback(b []byte, localAddr, srcAddr *net.UDPAddr) {
 				conn.Close()
 				return
 			}
+			s.udpSC.UDPListener.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 			s.udpSC.UDPListener.WriteToUDP(d, srcAddr)
+			s.udpSC.UDPListener.SetDeadline(time.Time{})
 			log.Printf("udp reply:%v", len(d))
 		} else {
+			s.udpSC.UDPListener.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 			s.udpSC.UDPListener.WriteToUDP(respBody, srcAddr)
+			s.udpSC.UDPListener.SetDeadline(time.Time{})
 			log.Printf("udp reply:%v", len(respBody))
 		}

 	} else {
 		//本地代理
-		dstAddr, err := net.ResolveUDPAddr("udp", net.JoinHostPort(p.Host(), p.Port()))
+		dstAddr, err := net.ResolveUDPAddr("udp", net.JoinHostPort(s.Resolve(p.Host()), p.Port()))
 		if err != nil {
 			log.Printf("can't resolve address: %s", err)
 			return
@ -257,6 +281,7 @@ func (s *Socks) udpCallback(b []byte, localAddr, srcAddr *net.UDPAddr) {
 		}
 		conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout*3)))
 		_, err = conn.Write(p.Data())
+		conn.SetDeadline(time.Time{})
 		log.Printf("udp send:%v", len(p.Data()))
 		if err != nil {
 			log.Printf("send udp packet to %s fail,ERR:%s", dstAddr.String(), err)
@ -265,7 +290,10 @@ func (s *Socks) udpCallback(b []byte, localAddr, srcAddr *net.UDPAddr) {
 		}
 		//log.Printf("send udp packet to %s success", dstAddr.String())
 		buf := make([]byte, 10*1024)
+		conn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		length, _, err := conn.ReadFromUDP(buf)
+		conn.SetDeadline(time.Time{})
+
 		if err != nil {
 			log.Printf("read udp response from %s fail ,ERR:%s", dstAddr.String(), err)
 			conn.Close()
@ -282,9 +310,13 @@ func (s *Socks) udpCallback(b []byte, localAddr, srcAddr *net.UDPAddr) {
 				conn.Close()
 				return
 			}
+			s.udpSC.UDPListener.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 			s.udpSC.UDPListener.WriteToUDP(d, srcAddr)
+			s.udpSC.UDPListener.SetDeadline(time.Time{})
 		} else {
+			s.udpSC.UDPListener.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 			s.udpSC.UDPListener.WriteToUDP(respPacket, srcAddr)
+			s.udpSC.UDPListener.SetDeadline(time.Time{})
 		}
 		log.Printf("udp reply:%v", len(respPacket))
 	}
@ -356,9 +388,15 @@ func (s *Socks) socksConnCallback(inConn net.Conn) {
 		//auth
 		_addr := strings.Split(inConn.RemoteAddr().String(), ":")
 		if s.basicAuth.CheckUserPass(user, pass, _addr[0], "") {
+			inConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 			inConn.Write([]byte{0x01, 0x00})
+			inConn.SetDeadline(time.Time{})
+
 		} else {
+			inConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 			inConn.Write([]byte{0x01, 0x01})
+			inConn.SetDeadline(time.Time{})
+
 			utils.CloseConn(&inConn)
 			return
 		}
@ -421,16 +459,17 @@ func (s *Socks) proxyTCP(inConn *net.Conn, methodReq socks.MethodsRequest, reque
 				if utils.IsIternalIP(host) {
 					useProxy = false
 				} else {
-					s.checker.Add(request.Addr())
-					useProxy, _, _ = s.checker.IsBlocked(request.Addr())
+					k := s.Resolve(request.Addr())
+					s.checker.Add(k)
+					useProxy, _, _ = s.checker.IsBlocked(k)
 				}
 				if useProxy {
 					outConn, err = s.getOutConn(methodReq.Bytes(), request.Bytes(), request.Addr())
 				} else {
-					outConn, err = utils.ConnectHost(request.Addr(), *s.cfg.Timeout)
+					outConn, err = utils.ConnectHost(s.Resolve(request.Addr()), *s.cfg.Timeout)
 				}
 			} else {
-				outConn, err = utils.ConnectHost(request.Addr(), *s.cfg.Timeout)
+				outConn, err = utils.ConnectHost(s.Resolve(request.Addr()), *s.cfg.Timeout)
 				useProxy = false
 			}
 		}
@ -467,12 +506,12 @@ func (s *Socks) getOutConn(methodBytes, reqBytes []byte, host string) (outConn n
 	case "tcp":
 		if *s.cfg.ParentType == "tls" {
 			var _outConn tls.Conn
-			_outConn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes)
+			_outConn, err = utils.TlsConnectHost(s.Resolve(*s.cfg.Parent), *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes, nil)
 			outConn = net.Conn(&_outConn)
 		} else if *s.cfg.ParentType == "kcp" {
-			outConn, err = utils.ConnectKCPHost(*s.cfg.Parent, *s.cfg.KCPMethod, *s.cfg.KCPKey)
+			outConn, err = utils.ConnectKCPHost(s.Resolve(*s.cfg.Parent), s.cfg.KCP)
 		} else {
-			outConn, err = utils.ConnectHost(*s.cfg.Parent, *s.cfg.Timeout)
+			outConn, err = utils.ConnectHost(s.Resolve(*s.cfg.Parent), *s.cfg.Timeout)
 		}
 		if err != nil {
 			err = fmt.Errorf("connect fail,%s", err)
@ -480,25 +519,32 @@ func (s *Socks) getOutConn(methodBytes, reqBytes []byte, host string) (outConn n
 		}
 		var buf = make([]byte, 1024)
 		//var n int
+		outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		_, err = outConn.Write(methodBytes)
+		outConn.SetDeadline(time.Time{})
 		if err != nil {
 			err = fmt.Errorf("write method fail,%s", err)
 			return
 		}
+		outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		_, err = outConn.Read(buf)
+		outConn.SetDeadline(time.Time{})
 		if err != nil {
 			err = fmt.Errorf("read method reply fail,%s", err)
 			return
 		}
 		//resp := buf[:n]
 		//log.Printf("resp:%v", resp)
-
+		outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		_, err = outConn.Write(reqBytes)
+		outConn.SetDeadline(time.Time{})
 		if err != nil {
 			err = fmt.Errorf("write req detail fail,%s", err)
 			return
 		}
+		outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
 		_, err = outConn.Read(buf)
+		outConn.SetDeadline(time.Time{})
 		if err != nil {
 			err = fmt.Errorf("read req reply fail,%s", err)
 			return
@ -562,12 +608,16 @@ func (s *Socks) ConnectSSH() (err error) {
 	if s.sshClient != nil {
 		s.sshClient.Close()
 	}
-	s.sshClient, err = ssh.Dial("tcp", *s.cfg.Parent, &config)
+	s.sshClient, err = ssh.Dial("tcp", s.Resolve(*s.cfg.Parent), &config)
 	<-s.lockChn
 	return
 }
 func (s *Socks) InitBasicAuth() (err error) {
-	s.basicAuth = utils.NewBasicAuth()
+	if *s.cfg.DNSAddress != "" {
+		s.basicAuth = utils.NewBasicAuth(&(*s).domainResolver)
+	} else {
+		s.basicAuth = utils.NewBasicAuth(nil)
+	}
 	if *s.cfg.AuthURL != "" {
 		s.basicAuth.SetAuthURL(*s.cfg.AuthURL, *s.cfg.AuthURLOkCode, *s.cfg.AuthURLTimeout, *s.cfg.AuthURLRetry)
 		log.Printf("auth from %s", *s.cfg.AuthURL)
@ -601,7 +651,11 @@ func (s *Socks) IsDeadLoop(inLocalAddr string, host string) bool {
 	}
 	if inPort == outPort {
 		var outIPs []net.IP
-		outIPs, err = net.LookupIP(outDomain)
+		if *s.cfg.DNSAddress != "" {
+			outIPs = []net.IP{net.ParseIP(s.Resolve(outDomain))}
+		} else {
+			outIPs, err = net.LookupIP(outDomain)
+		}
 		if err == nil {
 			for _, ip := range outIPs {
 				if ip.String() == inIP {
@ -625,3 +679,13 @@ func (s *Socks) IsDeadLoop(inLocalAddr string, host string) bool {
 	}
 	return false
 }
+func (s *Socks) Resolve(address string) string {
+	if *s.cfg.DNSAddress == "" {
+		return address
+	}
+	ip, err := s.domainResolver.Resolve(address)
+	if err != nil {
+		log.Printf("dns error %s , ERR:%s", address, err)
+	}
+	return ip
+}
--- a/services/sps.go
+++ b/services/sps.go
@ -0,0 +1,363 @@
+package services
+
+import (
+	"bytes"
+	"encoding/base64"
+	"errors"
+	"fmt"
+	"io/ioutil"
+	"log"
+	"net"
+	"runtime/debug"
+	"snail007/proxy/utils"
+	"snail007/proxy/utils/socks"
+	"strconv"
+	"strings"
+	"time"
+)
+
+type SPS struct {
+	outPool        utils.OutPool
+	cfg            SPSArgs
+	domainResolver utils.DomainResolver
+	basicAuth      utils.BasicAuth
+}
+
+func NewSPS() Service {
+	return &SPS{
+		outPool:   utils.OutPool{},
+		cfg:       SPSArgs{},
+		basicAuth: utils.BasicAuth{},
+	}
+}
+func (s *SPS) CheckArgs() {
+	if *s.cfg.Parent == "" {
+		log.Fatalf("parent required for %s %s", s.cfg.Protocol(), *s.cfg.Local)
+	}
+	if *s.cfg.ParentType == "" {
+		log.Fatalf("parent type unkown,use -T <tls|tcp|kcp>")
+	}
+	if *s.cfg.ParentType == TYPE_TLS || *s.cfg.LocalType == TYPE_TLS {
+		s.cfg.CertBytes, s.cfg.KeyBytes = utils.TlsBytes(*s.cfg.CertFile, *s.cfg.KeyFile)
+		if *s.cfg.CaCertFile != "" {
+			var err error
+			s.cfg.CaCertBytes, err = ioutil.ReadFile(*s.cfg.CaCertFile)
+			if err != nil {
+				log.Fatalf("read ca file error,ERR:%s", err)
+			}
+		}
+	}
+}
+func (s *SPS) InitService() {
+	s.InitOutConnPool()
+	if *s.cfg.DNSAddress != "" {
+		(*s).domainResolver = utils.NewDomainResolver(*s.cfg.DNSAddress, *s.cfg.DNSTTL)
+	}
+	s.InitBasicAuth()
+}
+func (s *SPS) InitOutConnPool() {
+	if *s.cfg.ParentType == TYPE_TLS || *s.cfg.ParentType == TYPE_TCP || *s.cfg.ParentType == TYPE_KCP {
+		//dur int, isTLS bool, certBytes, keyBytes []byte,
+		//parent string, timeout int, InitialCap int, MaxCap int
+		s.outPool = utils.NewOutPool(
+			0,
+			*s.cfg.ParentType,
+			s.cfg.KCP,
+			s.cfg.CertBytes, s.cfg.KeyBytes, nil,
+			*s.cfg.Parent,
+			*s.cfg.Timeout,
+			0,
+			0,
+		)
+	}
+}
+
+func (s *SPS) StopService() {
+	if s.outPool.Pool != nil {
+		s.outPool.Pool.ReleaseAll()
+	}
+}
+func (s *SPS) Start(args interface{}) (err error) {
+	s.cfg = args.(SPSArgs)
+	s.CheckArgs()
+	log.Printf("use %s %s parent %s", *s.cfg.ParentType, *s.cfg.ParentServiceType, *s.cfg.Parent)
+	s.InitService()
+
+	for _, addr := range strings.Split(*s.cfg.Local, ",") {
+		if addr != "" {
+			host, port, _ := net.SplitHostPort(*s.cfg.Local)
+			p, _ := strconv.Atoi(port)
+			sc := utils.NewServerChannel(host, p)
+			if *s.cfg.LocalType == TYPE_TCP {
+				err = sc.ListenTCP(s.callback)
+			} else if *s.cfg.LocalType == TYPE_TLS {
+				err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, nil, s.callback)
+			} else if *s.cfg.LocalType == TYPE_KCP {
+				err = sc.ListenKCP(s.cfg.KCP, s.callback)
+			}
+			if err != nil {
+				return
+			}
+			log.Printf("%s http(s)+socks proxy on %s", s.cfg.Protocol(), (*sc.Listener).Addr())
+		}
+	}
+	return
+}
+
+func (s *SPS) Clean() {
+	s.StopService()
+}
+func (s *SPS) callback(inConn net.Conn) {
+	defer func() {
+		if err := recover(); err != nil {
+			log.Printf("%s conn handler crashed with err : %s \nstack: %s", s.cfg.Protocol(), err, string(debug.Stack()))
+		}
+	}()
+	var err error
+	switch *s.cfg.ParentType {
+	case TYPE_KCP:
+		fallthrough
+	case TYPE_TCP:
+		fallthrough
+	case TYPE_TLS:
+		err = s.OutToTCP(&inConn)
+	default:
+		err = fmt.Errorf("unkown parent type %s", *s.cfg.ParentType)
+	}
+	if err != nil {
+		log.Printf("connect to %s parent %s fail, ERR:%s from %s", *s.cfg.ParentType, *s.cfg.Parent, err, inConn.RemoteAddr())
+		utils.CloseConn(&inConn)
+	}
+}
+func (s *SPS) OutToTCP(inConn *net.Conn) (err error) {
+	buf := make([]byte, 1024)
+	n, err := (*inConn).Read(buf)
+	header := buf[:n]
+	if err != nil {
+		log.Printf("ERR:%s", err)
+		utils.CloseConn(inConn)
+		return
+	}
+	address := ""
+	var auth socks.Auth
+	var forwardBytes []byte
+	//fmt.Printf("%v", header)
+	if header[0] == socks.VERSION_V5 {
+		//socks5 server
+		var serverConn *socks.ServerConn
+		if s.IsBasicAuth() {
+			serverConn = socks.NewServerConn(inConn, time.Millisecond*time.Duration(*s.cfg.Timeout), &s.basicAuth, "", header)
+		} else {
+			serverConn = socks.NewServerConn(inConn, time.Millisecond*time.Duration(*s.cfg.Timeout), nil, "", header)
+		}
+		if err = serverConn.Handshake(); err != nil {
+			return
+		}
+		address = serverConn.Target()
+		auth = serverConn.AuthData()
+	} else if bytes.IndexByte(header, '\n') != -1 {
+		//http
+		var request utils.HTTPRequest
+		(*inConn).SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
+		if s.IsBasicAuth() {
+			request, err = utils.NewHTTPRequest(inConn, 1024, true, &s.basicAuth, header)
+		} else {
+			request, err = utils.NewHTTPRequest(inConn, 1024, false, nil, header)
+		}
+		(*inConn).SetDeadline(time.Time{})
+		if err != nil {
+			log.Printf("new http request fail,ERR: %s", err)
+			utils.CloseConn(inConn)
+			return
+		}
+		if len(header) >= 7 && strings.ToLower(string(header[:7])) == "connect" {
+			//https
+			request.HTTPSReply()
+			//log.Printf("https reply: %s", request.Host)
+		} else {
+			forwardBytes = request.HeadBuf
+		}
+		address = request.Host
+		var userpass string
+		if s.IsBasicAuth() {
+			userpass, err = request.GetAuthDataStr()
+			if err != nil {
+				return
+			}
+			userpassA := strings.Split(userpass, ":")
+			if len(userpassA) == 2 {
+				auth = socks.Auth{User: userpassA[0], Password: userpassA[1]}
+			}
+		}
+	} else {
+		log.Printf("unknown request from: %s,%s", (*inConn).RemoteAddr(), string(header))
+		utils.CloseConn(inConn)
+		err = errors.New("unknown request")
+		return
+	}
+	//connect to parent
+	var outConn net.Conn
+	var _outConn interface{}
+	_outConn, err = s.outPool.Pool.Get()
+	if err == nil {
+		outConn = _outConn.(net.Conn)
+	}
+	if err != nil {
+		log.Printf("connect to %s , err:%s", *s.cfg.Parent, err)
+		utils.CloseConn(inConn)
+		return
+	}
+
+	//ask parent for connect to target address
+	if *s.cfg.ParentServiceType == "http" {
+		//http parent
+		pb := new(bytes.Buffer)
+		pb.Write([]byte(fmt.Sprintf("CONNECT %s HTTP/1.1\r\nProxy-Connection: Keep-Alive\r\n", address)))
+		//Proxy-Authorization:\r\n
+		u := ""
+		if *s.cfg.ParentAuth != "" {
+			a := strings.Split(*s.cfg.ParentAuth, ":")
+			if len(a) != 2 {
+				err = fmt.Errorf("parent auth data format error")
+				return
+			}
+			u = fmt.Sprintf("%s:%s", a[0], a[1])
+		} else {
+			if !s.IsBasicAuth() && auth.Password != "" && auth.User != "" {
+				u = fmt.Sprintf("%s:%s", auth.User, auth.Password)
+			}
+		}
+		if u != "" {
+			pb.Write([]byte(fmt.Sprintf("Proxy-Authorization:Basic %s\r\n", base64.StdEncoding.EncodeToString([]byte(u)))))
+		}
+		outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
+		_, err = outConn.Write(pb.Bytes())
+		outConn.SetDeadline(time.Time{})
+		if err != nil {
+			log.Printf("write CONNECT to %s , err:%s", *s.cfg.Parent, err)
+			utils.CloseConn(inConn)
+			utils.CloseConn(&outConn)
+			return
+		}
+		reply := make([]byte, 100)
+		outConn.SetDeadline(time.Now().Add(time.Millisecond * time.Duration(*s.cfg.Timeout)))
+		_, err = outConn.Read(reply)
+		outConn.SetDeadline(time.Time{})
+		if err != nil {
+			log.Printf("read reply from %s , err:%s", *s.cfg.Parent, err)
+			utils.CloseConn(inConn)
+			utils.CloseConn(&outConn)
+			return
+		}
+		//log.Printf("reply: %s", string(reply[:n]))
+	} else {
+		log.Printf("connect %s", address)
+		//socks client
+		var clientConn *socks.ClientConn
+		if *s.cfg.ParentAuth != "" {
+			a := strings.Split(*s.cfg.ParentAuth, ":")
+			if len(a) != 2 {
+				err = fmt.Errorf("parent auth data format error")
+				return
+			}
+			clientConn = socks.NewClientConn(&outConn, "tcp", address, time.Millisecond*time.Duration(*s.cfg.Timeout), &socks.Auth{User: a[0], Password: a[1]}, header)
+		} else {
+			if !s.IsBasicAuth() && auth.Password != "" && auth.User != "" {
+				clientConn = socks.NewClientConn(&outConn, "tcp", address, time.Millisecond*time.Duration(*s.cfg.Timeout), &auth, header)
+			} else {
+				clientConn = socks.NewClientConn(&outConn, "tcp", address, time.Millisecond*time.Duration(*s.cfg.Timeout), nil, header)
+			}
+		}
+		if err = clientConn.Handshake(); err != nil {
+			return
+		}
+	}
+	//forward client data to target,if necessary.
+	if len(forwardBytes) > 0 {
+		outConn.Write(forwardBytes)
+	}
+	//bind
+	inAddr := (*inConn).RemoteAddr().String()
+	outAddr := outConn.RemoteAddr().String()
+	utils.IoBind((*inConn), outConn, func(err interface{}) {
+		log.Printf("conn %s - %s released", inAddr, outAddr)
+	})
+	log.Printf("conn %s - %s connected", inAddr, outAddr)
+	return
+}
+func (s *SPS) InitBasicAuth() (err error) {
+	if *s.cfg.DNSAddress != "" {
+		s.basicAuth = utils.NewBasicAuth(&(*s).domainResolver)
+	} else {
+		s.basicAuth = utils.NewBasicAuth(nil)
+	}
+	if *s.cfg.AuthURL != "" {
+		s.basicAuth.SetAuthURL(*s.cfg.AuthURL, *s.cfg.AuthURLOkCode, *s.cfg.AuthURLTimeout, *s.cfg.AuthURLRetry)
+		log.Printf("auth from %s", *s.cfg.AuthURL)
+	}
+	if *s.cfg.AuthFile != "" {
+		var n = 0
+		n, err = s.basicAuth.AddFromFile(*s.cfg.AuthFile)
+		if err != nil {
+			err = fmt.Errorf("auth-file ERR:%s", err)
+			return
+		}
+		log.Printf("auth data added from file %d , total:%d", n, s.basicAuth.Total())
+	}
+	if len(*s.cfg.Auth) > 0 {
+		n := s.basicAuth.Add(*s.cfg.Auth)
+		log.Printf("auth data added %d, total:%d", n, s.basicAuth.Total())
+	}
+	return
+}
+func (s *SPS) IsBasicAuth() bool {
+	return *s.cfg.AuthFile != "" || len(*s.cfg.Auth) > 0 || *s.cfg.AuthURL != ""
+}
+func (s *SPS) buildRequest(address string) (buf []byte, err error) {
+	host, portStr, err := net.SplitHostPort(address)
+	if err != nil {
+		return nil, err
+	}
+
+	port, err := strconv.Atoi(portStr)
+	if err != nil {
+		err = errors.New("proxy: failed to parse port number: " + portStr)
+		return
+	}
+	if port < 1 || port > 0xffff {
+		err = errors.New("proxy: port number out of range: " + portStr)
+		return
+	}
+	buf = buf[:0]
+	buf = append(buf, 0x05, 0x01, 0 /* reserved */)
+
+	if ip := net.ParseIP(host); ip != nil {
+		if ip4 := ip.To4(); ip4 != nil {
+			buf = append(buf, 0x01)
+			ip = ip4
+		} else {
+			buf = append(buf, 0x04)
+		}
+		buf = append(buf, ip...)
+	} else {
+		if len(host) > 255 {
+			err = errors.New("proxy: destination host name too long: " + host)
+			return
+		}
+		buf = append(buf, 0x03)
+		buf = append(buf, byte(len(host)))
+		buf = append(buf, host...)
+	}
+	buf = append(buf, byte(port>>8), byte(port))
+	return
+}
+func (s *SPS) Resolve(address string) string {
+	if *s.cfg.DNSAddress == "" {
+		return address
+	}
+	ip, err := s.domainResolver.Resolve(address)
+	if err != nil {
+		log.Printf("dns error %s , ERR:%s", address, err)
+	}
+	return ip
+}
--- a/services/tcp.go
+++ b/services/tcp.go
@ -6,8 +6,8 @@ import (
 	"io"
 	"log"
 	"net"
-	"proxy/utils"
 	"runtime/debug"
+	"snail007/proxy/utils"
 	"time"

 	"strconv"
@ -56,9 +56,9 @@ func (s *TCP) Start(args interface{}) (err error) {
 	if *s.cfg.LocalType == TYPE_TCP {
 		err = sc.ListenTCP(s.callback)
 	} else if *s.cfg.LocalType == TYPE_TLS {
-		err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, s.callback)
+		err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, nil, s.callback)
 	} else if *s.cfg.LocalType == TYPE_KCP {
-		err = sc.ListenKCP(*s.cfg.KCPMethod, *s.cfg.KCPKey, s.callback)
+		err = sc.ListenKCP(s.cfg.KCP, s.callback)
 	}
 	if err != nil {
 		return
@ -171,9 +171,8 @@ func (s *TCP) InitOutConnPool() {
 		s.outPool = utils.NewOutPool(
 			*s.cfg.CheckParentInterval,
 			*s.cfg.ParentType,
-			*s.cfg.KCPMethod,
-			*s.cfg.KCPKey,
-			s.cfg.CertBytes, s.cfg.KeyBytes,
+			s.cfg.KCP,
+			s.cfg.CertBytes, s.cfg.KeyBytes, nil,
 			*s.cfg.Parent,
 			*s.cfg.Timeout,
 			*s.cfg.PoolSize,
--- a/services/tunnel_bridge.go
+++ b/services/tunnel_bridge.go
@ -4,7 +4,7 @@ import (
 	"bufio"
 	"log"
 	"net"
-	"proxy/utils"
+	"snail007/proxy/utils"
 	"strconv"
 	"time"
 )
@ -51,7 +51,7 @@ func (s *TunnelBridge) Start(args interface{}) (err error) {
 	p, _ := strconv.Atoi(port)
 	sc := utils.NewServerChannel(host, p)

-	err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, func(inConn net.Conn) {
+	err = sc.ListenTls(s.cfg.CertBytes, s.cfg.KeyBytes, nil, func(inConn net.Conn) {
 		//log.Printf("connection from %s ", inConn.RemoteAddr())

 		reader := bufio.NewReader(inConn)
--- a/services/tunnel_client.go
+++ b/services/tunnel_client.go
@ -6,7 +6,7 @@ import (
 	"io"
 	"log"
 	"net"
-	"proxy/utils"
+	"snail007/proxy/utils"
 	"time"
 )

@ -161,7 +161,7 @@ func (s *TunnelClient) GetInConn(typ uint8, data ...string) (outConn net.Conn, e
 }
 func (s *TunnelClient) GetConn() (conn net.Conn, err error) {
 	var _conn tls.Conn
-	_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes)
+	_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes, nil)
 	if err == nil {
 		conn = net.Conn(&_conn)
 	}
--- a/services/tunnel_server.go
+++ b/services/tunnel_server.go
@ -6,8 +6,8 @@ import (
 	"io"
 	"log"
 	"net"
-	"proxy/utils"
 	"runtime/debug"
+	"snail007/proxy/utils"
 	"strconv"
 	"strings"
 	"time"
@ -174,7 +174,7 @@ func (s *TunnelServerManager) GetOutConn(typ uint8) (outConn net.Conn, ID string
 }
 func (s *TunnelServerManager) GetConn() (conn net.Conn, err error) {
 	var _conn tls.Conn
-	_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes)
+	_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes, nil)
 	if err == nil {
 		conn = net.Conn(&_conn)
 	}
@ -280,7 +280,7 @@ func (s *TunnelServer) GetOutConn(typ uint8) (outConn net.Conn, ID string, err e
 }
 func (s *TunnelServer) GetConn() (conn net.Conn, err error) {
 	var _conn tls.Conn
-	_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes)
+	_conn, err = utils.TlsConnectHost(*s.cfg.Parent, *s.cfg.Timeout, s.cfg.CertBytes, s.cfg.KeyBytes, nil)
 	if err == nil {
 		conn = net.Conn(&_conn)
 	}
--- a/services/udp.go
+++ b/services/udp.go
@ -7,8 +7,9 @@ import (
 	"io"
 	"log"
 	"net"
-	"proxy/utils"
 	"runtime/debug"
+	"snail007/proxy/services/kcpcfg"
+	"snail007/proxy/utils"
 	"strconv"
 	"strings"
 	"time"
@ -208,8 +209,8 @@ func (s *UDP) InitOutConnPool() {
 		s.outPool = utils.NewOutPool(
 			*s.cfg.CheckParentInterval,
 			*s.cfg.ParentType,
-			"", "",
-			s.cfg.CertBytes, s.cfg.KeyBytes,
+			kcpcfg.KCPConfigArgs{},
+			s.cfg.CertBytes, s.cfg.KeyBytes, nil,
 			*s.cfg.Parent,
 			*s.cfg.Timeout,
 			*s.cfg.PoolSize,
--- a/utils/functions.go
+++ b/utils/functions.go
@ -7,19 +7,22 @@ import (
 	"crypto/tls"
 	"crypto/x509"
 	"encoding/binary"
+	"encoding/pem"
 	"errors"
 	"fmt"
 	"io"
 	"io/ioutil"
 	"log"
+	"math/rand"
 	"net"
 	"net/http"
 	"os"
 	"os/exec"
+	"snail007/proxy/services/kcpcfg"

 	"golang.org/x/crypto/pbkdf2"

-	"proxy/utils/id"
+	"snail007/proxy/utils/id"
 	"strconv"
 	"strings"
 	"time"
@ -83,14 +86,14 @@ func ioCopy(dst io.ReadWriter, src io.ReadWriter) (err error) {
 		}
 	}
 }
-func TlsConnectHost(host string, timeout int, certBytes, keyBytes []byte) (conn tls.Conn, err error) {
+func TlsConnectHost(host string, timeout int, certBytes, keyBytes, caCertBytes []byte) (conn tls.Conn, err error) {
 	h := strings.Split(host, ":")
 	port, _ := strconv.Atoi(h[1])
-	return TlsConnect(h[0], port, timeout, certBytes, keyBytes)
+	return TlsConnect(h[0], port, timeout, certBytes, keyBytes, caCertBytes)
 }

-func TlsConnect(host string, port, timeout int, certBytes, keyBytes []byte) (conn tls.Conn, err error) {
-	conf, err := getRequestTlsConfig(certBytes, keyBytes)
+func TlsConnect(host string, port, timeout int, certBytes, keyBytes, caCertBytes []byte) (conn tls.Conn, err error) {
+	conf, err := getRequestTlsConfig(certBytes, keyBytes, caCertBytes)
 	if err != nil {
 		return
 	}
@ -100,22 +103,49 @@ func TlsConnect(host string, port, timeout int, certBytes, keyBytes []byte) (con
 	}
 	return *tls.Client(_conn, conf), err
 }
-func getRequestTlsConfig(certBytes, keyBytes []byte) (conf *tls.Config, err error) {
+func getRequestTlsConfig(certBytes, keyBytes, caCertBytes []byte) (conf *tls.Config, err error) {
+
 	var cert tls.Certificate
 	cert, err = tls.X509KeyPair(certBytes, keyBytes)
 	if err != nil {
 		return
 	}
 	serverCertPool := x509.NewCertPool()
-	ok := serverCertPool.AppendCertsFromPEM(certBytes)
+	caBytes := certBytes
+	if caCertBytes != nil {
+		caBytes = caCertBytes
+
+	}
+	ok := serverCertPool.AppendCertsFromPEM(caBytes)
 	if !ok {
 		err = errors.New("failed to parse root certificate")
 	}
+	block, _ := pem.Decode(caBytes)
+	if block == nil {
+		panic("failed to parse certificate PEM")
+	}
+	x509Cert, _ := x509.ParseCertificate(block.Bytes)
+	if x509Cert == nil {
+		panic("failed to parse block")
+	}
 	conf = &tls.Config{
 		RootCAs:            serverCertPool,
 		Certificates:       []tls.Certificate{cert},
-		ServerName:         "proxy",
-		InsecureSkipVerify: false,
+		InsecureSkipVerify: true,
+		ServerName:         x509Cert.Subject.CommonName,
+		VerifyPeerCertificate: func(rawCerts [][]byte, verifiedChains [][]*x509.Certificate) error {
+			opts := x509.VerifyOptions{
+				Roots: serverCertPool,
+			}
+			for _, rawCert := range rawCerts {
+				cert, _ := x509.ParseCertificate(rawCert)
+				_, err := cert.Verify(opts)
+				if err != nil {
+					return err
+				}
+			}
+			return nil
+		},
 	}
 	return
 }
@ -124,31 +154,41 @@ func ConnectHost(hostAndPort string, timeout int) (conn net.Conn, err error) {
 	conn, err = net.DialTimeout("tcp", hostAndPort, time.Duration(timeout)*time.Millisecond)
 	return
 }
-func ConnectKCPHost(hostAndPort, method, key string) (conn net.Conn, err error) {
-	kcpconn, err := kcp.DialWithOptions(hostAndPort, GetKCPBlock(method, key), 10, 3)
+func ConnectKCPHost(hostAndPort string, config kcpcfg.KCPConfigArgs) (conn net.Conn, err error) {
+	kcpconn, err := kcp.DialWithOptions(hostAndPort, config.Block, *config.DataShard, *config.ParityShard)
 	if err != nil {
 		return
 	}
-	kcpconn.SetNoDelay(1, 10, 2, 1)
-	kcpconn.SetWindowSize(1024, 1024)
-	kcpconn.SetMtu(1400)
-	kcpconn.SetACKNoDelay(false)
-	return kcpconn, err
+	kcpconn.SetStreamMode(true)
+	kcpconn.SetWriteDelay(true)
+	kcpconn.SetNoDelay(*config.NoDelay, *config.Interval, *config.Resend, *config.NoCongestion)
+	kcpconn.SetMtu(*config.MTU)
+	kcpconn.SetWindowSize(*config.SndWnd, *config.RcvWnd)
+	kcpconn.SetACKNoDelay(*config.AckNodelay)
+	if *config.NoComp {
+		return kcpconn, err
+	}
+	return NewCompStream(kcpconn), err
 }
-func ListenTls(ip string, port int, certBytes, keyBytes []byte) (ln *net.Listener, err error) {
+
+func ListenTls(ip string, port int, certBytes, keyBytes, caCertBytes []byte) (ln *net.Listener, err error) {
+
 	var cert tls.Certificate
 	cert, err = tls.X509KeyPair(certBytes, keyBytes)
 	if err != nil {
 		return
 	}
 	clientCertPool := x509.NewCertPool()
-	ok := clientCertPool.AppendCertsFromPEM(certBytes)
+	caBytes := certBytes
+	if caCertBytes != nil {
+		caBytes = caCertBytes
+	}
+	ok := clientCertPool.AppendCertsFromPEM(caBytes)
 	if !ok {
 		err = errors.New("failed to parse root certificate")
 	}
 	config := &tls.Config{
 		ClientCAs:    clientCertPool,
-		ServerName:   "proxy",
 		Certificates: []tls.Certificate{cert},
 		ClientAuth:   tls.RequireAndVerifyClientCert,
 	}
@ -193,20 +233,95 @@ func CloseConn(conn *net.Conn) {
 	}
 }
 func Keygen() (err error) {
-	cmd := exec.Command("sh", "-c", "openssl genrsa -out proxy.key 2048")
-	out, err := cmd.CombinedOutput()
-	if err != nil {
-		log.Printf("err:%s", err)
-		return
+	CList := []string{"AD", "AE", "AF", "AG", "AI", "AL", "AM", "AO", "AR", "AT", "AU", "AZ", "BB", "BD", "BE", "BF", "BG", "BH", "BI", "BJ", "BL", "BM", "BN", "BO", "BR", "BS", "BW", "BY", "BZ", "CA", "CF", "CG", "CH", "CK", "CL", "CM", "CN", "CO", "CR", "CS", "CU", "CY", "CZ", "DE", "DJ", "DK", "DO", "DZ", "EC", "EE", "EG", "ES", "ET", "FI", "FJ", "FR", "GA", "GB", "GD", "GE", "GF", "GH", "GI", "GM", "GN", "GR", "GT", "GU", "GY", "HK", "HN", "HT", "HU", "ID", "IE", "IL", "IN", "IQ", "IR", "IS", "IT", "JM", "JO", "JP", "KE", "KG", "KH", "KP", "KR", "KT", "KW", "KZ", "LA", "LB", "LC", "LI", "LK", "LR", "LS", "LT", "LU", "LV", "LY", "MA", "MC", "MD", "MG", "ML", "MM", "MN", "MO", "MS", "MT", "MU", "MV", "MW", "MX", "MY", "MZ", "NA", "NE", "NG", "NI", "NL", "NO", "NP", "NR", "NZ", "OM", "PA", "PE", "PF", "PG", "PH", "PK", "PL", "PR", "PT", "PY", "QA", "RO", "RU", "SA", "SB", "SC", "SD", "SE", "SG", "SI", "SK", "SL", "SM", "SN", "SO", "SR", "ST", "SV", "SY", "SZ", "TD", "TG", "TH", "TJ", "TM", "TN", "TO", "TR", "TT", "TW", "TZ", "UA", "UG", "US", "UY", "UZ", "VC", "VE", "VN", "YE", "YU", "ZA", "ZM", "ZR", "ZW"}
+	domainSubfixList := []string{".com", ".edu", ".gov", ".int", ".mil", ".net", ".org", ".biz", ".info", ".pro", ".name", ".museum", ".coop", ".aero", ".xxx", ".idv", ".ac", ".ad", ".ae", ".af", ".ag", ".ai", ".al", ".am", ".an", ".ao", ".aq", ".ar", ".as", ".at", ".au", ".aw", ".az", ".ba", ".bb", ".bd", ".be", ".bf", ".bg", ".bh", ".bi", ".bj", ".bm", ".bn", ".bo", ".br", ".bs", ".bt", ".bv", ".bw", ".by", ".bz", ".ca", ".cc", ".cd", ".cf", ".cg", ".ch", ".ci", ".ck", ".cl", ".cm", ".cn", ".co", ".cr", ".cu", ".cv", ".cx", ".cy", ".cz", ".de", ".dj", ".dk", ".dm", ".do", ".dz", ".ec", ".ee", ".eg", ".eh", ".er", ".es", ".et", ".eu", ".fi", ".fj", ".fk", ".fm", ".fo", ".fr", ".ga", ".gd", ".ge", ".gf", ".gg", ".gh", ".gi", ".gl", ".gm", ".gn", ".gp", ".gq", ".gr", ".gs", ".gt", ".gu", ".gw", ".gy", ".hk", ".hm", ".hn", ".hr", ".ht", ".hu", ".id", ".ie", ".il", ".im", ".in", ".io", ".iq", ".ir", ".is", ".it", ".je", ".jm", ".jo", ".jp", ".ke", ".kg", ".kh", ".ki", ".km", ".kn", ".kp", ".kr", ".kw", ".ky", ".kz", ".la", ".lb", ".lc", ".li", ".lk", ".lr", ".ls", ".lt", ".lu", ".lv", ".ly", ".ma", ".mc", ".md", ".mg", ".mh", ".mk", ".ml", ".mm", ".mn", ".mo", ".mp", ".mq", ".mr", ".ms", ".mt", ".mu", ".mv", ".mw", ".mx", ".my", ".mz", ".na", ".nc", ".ne", ".nf", ".ng", ".ni", ".nl", ".no", ".np", ".nr", ".nu", ".nz", ".om", ".pa", ".pe", ".pf", ".pg", ".ph", ".pk", ".pl", ".pm", ".pn", ".pr", ".ps", ".pt", ".pw", ".py", ".qa", ".re", ".ro", ".ru", ".rw", ".sa", ".sb", ".sc", ".sd", ".se", ".sg", ".sh", ".si", ".sj", ".sk", ".sl", ".sm", ".sn", ".so", ".sr", ".st", ".sv", ".sy", ".sz", ".tc", ".td", ".tf", ".tg", ".th", ".tj", ".tk", ".tl", ".tm", ".tn", ".to", ".tp", ".tr", ".tt", ".tv", ".tw", ".tz", ".ua", ".ug", ".uk", ".um", ".us", ".uy", ".uz", ".va", ".vc", ".ve", ".vg", ".vi", ".vn", ".vu", ".wf", ".ws", ".ye", ".yt", ".yu", ".yr", ".za", ".zm", ".zw"}
+	C := CList[int(RandInt(4))%len(CList)]
+	ST := RandString(int(RandInt(4) % 10))
+	O := RandString(int(RandInt(4) % 10))
+	CN := strings.ToLower(RandString(int(RandInt(4)%10)) + domainSubfixList[int(RandInt(4))%len(domainSubfixList)])
+	//log.Printf("C: %s, ST: %s, O: %s, CN: %s", C, ST, O, CN)
+	var out []byte
+	if len(os.Args) == 3 && os.Args[2] == "ca" {
+		cmd := exec.Command("sh", "-c", "openssl genrsa -out ca.key 2048")
+		out, err = cmd.CombinedOutput()
+		if err != nil {
+			log.Printf("err:%s", err)
+			return
+		}
+		fmt.Println(string(out))
+
+		cmdStr := fmt.Sprintf("openssl req -new -key ca.key -x509 -days 36500 -out ca.crt -subj /C=%s/ST=%s/O=%s/CN=%s", C, ST, O, "*."+CN)
+		cmd = exec.Command("sh", "-c", cmdStr)
+		out, err = cmd.CombinedOutput()
+		if err != nil {
+			log.Printf("err:%s", err)
+			return
+		}
+		fmt.Println(string(out))
+	} else if len(os.Args) == 5 && os.Args[2] == "ca" && os.Args[3] != "" && os.Args[4] != "" {
+		certBytes, _ := ioutil.ReadFile("ca.crt")
+		block, _ := pem.Decode(certBytes)
+		if block == nil || certBytes == nil {
+			panic("failed to parse ca certificate PEM")
+		}
+		x509Cert, _ := x509.ParseCertificate(block.Bytes)
+		if x509Cert == nil {
+			panic("failed to parse block")
+		}
+		name := os.Args[3]
+		days := os.Args[4]
+		cmd := exec.Command("sh", "-c", "openssl genrsa -out "+name+".key 2048")
+		out, err = cmd.CombinedOutput()
+		if err != nil {
+			log.Printf("err:%s", err)
+			return
+		}
+		fmt.Println(string(out))
+
+		cmdStr := fmt.Sprintf("openssl req -new -key %s.key -out %s.csr -subj /C=%s/ST=%s/O=%s/CN=%s", name, name, C, ST, O, CN)
+		fmt.Printf("%s", cmdStr)
+		cmd = exec.Command("sh", "-c", cmdStr)
+		out, err = cmd.CombinedOutput()
+		if err != nil {
+			log.Printf("err:%s", err)
+			return
+		}
+		fmt.Println(string(out))
+
+		cmdStr = fmt.Sprintf("openssl x509 -req -days %s -in %s.csr -CA ca.crt -CAkey ca.key -CAcreateserial -out %s.crt", days, name, name)
+		fmt.Printf("%s", cmdStr)
+		cmd = exec.Command("sh", "-c", cmdStr)
+		out, err = cmd.CombinedOutput()
+		if err != nil {
+			log.Printf("err:%s", err)
+			return
+		}
+
+		fmt.Println(string(out))
+	} else if len(os.Args) == 3 && os.Args[2] == "usage" {
+		fmt.Println(`proxy keygen //generate proxy.crt and proxy.key
+proxy keygen ca //generate ca.crt and ca.key
+proxy keygen ca client0 30 //generate client0.crt client0.key and use ca.crt sign it with 30 days
+	`)
+	} else if len(os.Args) == 2 {
+		cmd := exec.Command("sh", "-c", "openssl genrsa -out proxy.key 2048")
+		out, err = cmd.CombinedOutput()
+		if err != nil {
+			log.Printf("err:%s", err)
+			return
+		}
+		fmt.Println(string(out))
+
+		cmdStr := fmt.Sprintf("openssl req -new -key proxy.key -x509 -days 36500 -out proxy.crt -subj /C=%s/ST=%s/O=%s/CN=%s", C, ST, O, CN)
+		cmd = exec.Command("sh", "-c", cmdStr)
+		out, err = cmd.CombinedOutput()
+		if err != nil {
+			log.Printf("err:%s", err)
+			return
+		}
+		fmt.Println(string(out))
 	}
-	fmt.Println(string(out))
-	cmd = exec.Command("sh", "-c", `openssl req -new -key proxy.key -x509 -days 3650 -out proxy.crt -subj /C=CN/ST=BJ/O="Localhost Ltd"/CN=proxy`)
-	out, err = cmd.CombinedOutput()
-	if err != nil {
-		log.Printf("err:%s", err)
-		return
-	}
-	fmt.Println(string(out))
+
 	return
 }
 func GetAllInterfaceAddr() ([]net.IP, error) {
@ -306,6 +421,29 @@ func Uniqueid() string {
 	// s := fmt.Sprintf("%d", src.Int63())
 	// return s[len(s)-5:len(s)-1] + fmt.Sprintf("%d", uint64(time.Now().UnixNano()))[8:]
 }
+func RandString(strlen int) string {
+	codes := "QWERTYUIOPLKJHGFDSAZXCVBNMabcdefghijklmnopqrstuvwxyz0123456789"
+	codeLen := len(codes)
+	data := make([]byte, strlen)
+	rand.Seed(time.Now().UnixNano() + rand.Int63() + rand.Int63() + rand.Int63() + rand.Int63())
+	for i := 0; i < strlen; i++ {
+		idx := rand.Intn(codeLen)
+		data[i] = byte(codes[idx])
+	}
+	return string(data)
+}
+func RandInt(strLen int) int64 {
+	codes := "123456789"
+	codeLen := len(codes)
+	data := make([]byte, strLen)
+	rand.Seed(time.Now().UnixNano() + rand.Int63() + rand.Int63() + rand.Int63() + rand.Int63())
+	for i := 0; i < strLen; i++ {
+		idx := rand.Intn(codeLen)
+		data[i] = byte(codes[idx])
+	}
+	i, _ := strconv.ParseInt(string(data), 10, 64)
+	return i
+}
 func ReadData(r io.Reader) (data string, err error) {
 	var len uint16
 	err = binary.Read(r, binary.LittleEndian, &len)
@ -431,7 +569,7 @@ func GetKCPBlock(method, key string) (block kcp.BlockCrypt) {
 	}
 	return
 }
-func HttpGet(URL string, timeout int) (body []byte, code int, err error) {
+func HttpGet(URL string, timeout int, host ...string) (body []byte, code int, err error) {
 	var tr *http.Transport
 	var client *http.Client
 	conf := &tls.Config{
@ -445,7 +583,16 @@ func HttpGet(URL string, timeout int) (body []byte, code int, err error) {
 		client = &http.Client{Timeout: time.Millisecond * time.Duration(timeout), Transport: tr}
 	}
 	defer tr.CloseIdleConnections()
-	resp, err := client.Get(URL)
+
+	//resp, err := client.Get(URL)
+	req, err := http.NewRequest("GET", URL, nil)
+	if err != nil {
+		return
+	}
+	if len(host) == 1 && host[0] != "" {
+		req.Host = host[0]
+	}
+	resp, err := client.Do(req)
 	if err != nil {
 		return
 	}
--- a/utils/serve-channel.go
+++ b/utils/serve-channel.go
@ -5,6 +5,7 @@ import (
 	"log"
 	"net"
 	"runtime/debug"
+	"snail007/proxy/services/kcpcfg"
 	"strconv"

 	kcp "github.com/xtaci/kcp-go"
@ -41,8 +42,8 @@ func NewServerChannelHost(host string) ServerChannel {
 func (sc *ServerChannel) SetErrAcceptHandler(fn func(err error)) {
 	sc.errAcceptHandler = fn
 }
-func (sc *ServerChannel) ListenTls(certBytes, keyBytes []byte, fn func(conn net.Conn)) (err error) {
-	sc.Listener, err = ListenTls(sc.ip, sc.port, certBytes, keyBytes)
+func (sc *ServerChannel) ListenTls(certBytes, keyBytes, caCertBytes []byte, fn func(conn net.Conn)) (err error) {
+	sc.Listener, err = ListenTls(sc.ip, sc.port, certBytes, keyBytes, caCertBytes)
 	if err == nil {
 		go func() {
 			defer func() {
@ -138,11 +139,23 @@ func (sc *ServerChannel) ListenUDP(fn func(packet []byte, localAddr, srcAddr *ne
 	}
 	return
 }
-func (sc *ServerChannel) ListenKCP(method, key string, fn func(conn net.Conn)) (err error) {
-	var l net.Listener
-	l, err = kcp.ListenWithOptions(fmt.Sprintf("%s:%d", sc.ip, sc.port), GetKCPBlock(method, key), 10, 3)
+func (sc *ServerChannel) ListenKCP(config kcpcfg.KCPConfigArgs, fn func(conn net.Conn)) (err error) {
+	lis, err := kcp.ListenWithOptions(fmt.Sprintf("%s:%d", sc.ip, sc.port), config.Block, *config.DataShard, *config.ParityShard)
 	if err == nil {
-		sc.Listener = &l
+		if err = lis.SetDSCP(*config.DSCP); err != nil {
+			log.Println("SetDSCP:", err)
+			return
+		}
+		if err = lis.SetReadBuffer(*config.SockBuf); err != nil {
+			log.Println("SetReadBuffer:", err)
+			return
+		}
+		if err = lis.SetWriteBuffer(*config.SockBuf); err != nil {
+			log.Println("SetWriteBuffer:", err)
+			return
+		}
+		sc.Listener = new(net.Listener)
+		*sc.Listener = lis
 		go func() {
 			defer func() {
 				if e := recover(); e != nil {
@ -150,8 +163,8 @@ func (sc *ServerChannel) ListenKCP(method, key string, fn func(conn net.Conn)) (
 				}
 			}()
 			for {
-				var conn net.Conn
-				conn, err = (*sc.Listener).Accept()
+				//var conn net.Conn
+				conn, err := lis.AcceptKCP()
 				if err == nil {
 					go func() {
 						defer func() {
@ -159,7 +172,18 @@ func (sc *ServerChannel) ListenKCP(method, key string, fn func(conn net.Conn)) (
 								log.Printf("kcp connection handler crashed , err : %s , \ntrace:%s", e, string(debug.Stack()))
 							}
 						}()
-						fn(conn)
+						conn.SetStreamMode(true)
+						conn.SetWriteDelay(true)
+						conn.SetNoDelay(*config.NoDelay, *config.Interval, *config.Resend, *config.NoCongestion)
+						conn.SetMtu(*config.MTU)
+						conn.SetWindowSize(*config.SndWnd, *config.RcvWnd)
+						conn.SetACKNoDelay(*config.AckNodelay)
+						if *config.NoComp {
+							fn(conn)
+						} else {
+							cconn := NewCompStream(conn)
+							fn(cconn)
+						}
 					}()
 				} else {
 					sc.errAcceptHandler(err)
--- a/utils/socks/client.go
+++ b/utils/socks/client.go
@ -0,0 +1,253 @@
+package socks
+
+import (
+	"encoding/binary"
+	"errors"
+	"fmt"
+	"io"
+	"net"
+	"strconv"
+	"time"
+)
+
+var socks5Errors = []string{
+	"",
+	"general failure",
+	"connection forbidden",
+	"network unreachable",
+	"host unreachable",
+	"connection refused",
+	"TTL expired",
+	"command not supported",
+	"address type not supported",
+}
+
+type Auth struct {
+	User, Password string
+}
+type ClientConn struct {
+	user     string
+	password string
+	conn     *net.Conn
+	header   []byte
+	timeout  time.Duration
+	addr     string
+	network  string
+	udpAddr  string
+}
+
+// SOCKS5 returns a Dialer that makes SOCKSv5 connections to the given address
+// with an optional username and password. See RFC 1928 and RFC 1929.
+// target must be a canonical address with a host and port.
+// network : tcp udp
+func NewClientConn(conn *net.Conn, network, target string, timeout time.Duration, auth *Auth, header []byte) *ClientConn {
+	s := &ClientConn{
+		conn:    conn,
+		network: network,
+		timeout: timeout,
+	}
+	if auth != nil {
+		s.user = auth.User
+		s.password = auth.Password
+	}
+	if header != nil && len(header) > 0 {
+		s.header = header
+	}
+	if network == "udp" && target == "" {
+		target = "0.0.0.0:1"
+	}
+	s.addr = target
+	return s
+}
+
+// connect takes an existing connection to a socks5 proxy server,
+// and commands the server to extend that connection to target,
+// which must be a canonical address with a host and port.
+func (s *ClientConn) Handshake() error {
+	host, portStr, err := net.SplitHostPort(s.addr)
+	if err != nil {
+		return err
+	}
+	port, err := strconv.Atoi(portStr)
+	if err != nil {
+		return errors.New("proxy: failed to parse port number: " + portStr)
+	}
+	if port < 1 || port > 0xffff {
+		return errors.New("proxy: port number out of range: " + portStr)
+	}
+
+	if err := s.handshake(host); err != nil {
+		return err
+	}
+	buf := []byte{}
+	if s.network == "tcp" {
+		buf = append(buf, VERSION_V5, CMD_CONNECT, 0 /* reserved */)
+
+	} else {
+		buf = append(buf, VERSION_V5, CMD_ASSOCIATE, 0 /* reserved */)
+	}
+	if ip := net.ParseIP(host); ip != nil {
+		if ip4 := ip.To4(); ip4 != nil {
+			buf = append(buf, ATYP_IPV4)
+			ip = ip4
+		} else {
+			buf = append(buf, ATYP_IPV6)
+		}
+		buf = append(buf, ip...)
+	} else {
+		if len(host) > 255 {
+			return errors.New("proxy: destination host name too long: " + host)
+		}
+		buf = append(buf, ATYP_DOMAIN)
+		buf = append(buf, byte(len(host)))
+		buf = append(buf, host...)
+	}
+	buf = append(buf, byte(port>>8), byte(port))
+	(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+	if _, err := (*s.conn).Write(buf); err != nil {
+		return errors.New("proxy: failed to write connect request to SOCKS5 proxy at " + s.addr + ": " + err.Error())
+	}
+	(*s.conn).SetDeadline(time.Time{})
+	(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+	if _, err := io.ReadFull((*s.conn), buf[:4]); err != nil {
+		return errors.New("proxy: failed to read connect reply from SOCKS5 proxy at " + s.addr + ": " + err.Error())
+	}
+	(*s.conn).SetDeadline(time.Time{})
+	failure := "unknown error"
+	if int(buf[1]) < len(socks5Errors) {
+		failure = socks5Errors[buf[1]]
+	}
+
+	if len(failure) > 0 {
+		return errors.New("proxy: SOCKS5 proxy at " + s.addr + " failed to connect: " + failure)
+	}
+
+	bytesToDiscard := 0
+	switch buf[3] {
+	case ATYP_IPV4:
+		bytesToDiscard = net.IPv4len
+	case ATYP_IPV6:
+		bytesToDiscard = net.IPv6len
+	case ATYP_DOMAIN:
+		(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+		_, err := io.ReadFull((*s.conn), buf[:1])
+		(*s.conn).SetDeadline(time.Time{})
+		if err != nil {
+			return errors.New("proxy: failed to read domain length from SOCKS5 proxy at " + s.addr + ": " + err.Error())
+		}
+		bytesToDiscard = int(buf[0])
+	default:
+		return errors.New("proxy: got unknown address type " + strconv.Itoa(int(buf[3])) + " from SOCKS5 proxy at " + s.addr)
+	}
+
+	if cap(buf) < bytesToDiscard {
+		buf = make([]byte, bytesToDiscard)
+	} else {
+		buf = buf[:bytesToDiscard]
+	}
+	(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+	if _, err := io.ReadFull((*s.conn), buf); err != nil {
+		return errors.New("proxy: failed to read address from SOCKS5 proxy at " + s.addr + ": " + err.Error())
+	}
+	(*s.conn).SetDeadline(time.Time{})
+	var ip net.IP
+	ip = buf
+	ipStr := ""
+	if bytesToDiscard == net.IPv4len || bytesToDiscard == net.IPv6len {
+		if ipv4 := ip.To4(); ipv4 != nil {
+			ipStr = ipv4.String()
+		} else {
+			ipStr = ip.To16().String()
+		}
+	}
+	//log.Printf("%v", ipStr)
+	// Also need to discard the port number
+	(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+	if _, err := io.ReadFull((*s.conn), buf[:2]); err != nil {
+		return errors.New("proxy: failed to read port from SOCKS5 proxy at " + s.addr + ": " + err.Error())
+	}
+	p := binary.BigEndian.Uint16([]byte{buf[0], buf[1]})
+	//log.Printf("%v", p)
+	s.udpAddr = net.JoinHostPort(ipStr, fmt.Sprintf("%d", p))
+	//log.Printf("%v", s.udpAddr)
+	(*s.conn).SetDeadline(time.Time{})
+	return nil
+}
+func (s *ClientConn) SendUDP(data []byte, addr string) (respData []byte, err error) {
+
+	c, err := net.DialTimeout("udp", s.udpAddr, s.timeout)
+	if err != nil {
+		return
+	}
+	conn := c.(*net.UDPConn)
+
+	p := NewPacketUDP()
+	p.Build(addr, data)
+	conn.SetDeadline(time.Now().Add(s.timeout))
+	conn.Write(p.Bytes())
+	conn.SetDeadline(time.Time{})
+
+	buf := make([]byte, 1024)
+	conn.SetDeadline(time.Now().Add(s.timeout))
+	n, _, err := conn.ReadFrom(buf)
+	conn.SetDeadline(time.Time{})
+	if err != nil {
+		return
+	}
+	respData = buf[:n]
+	return
+}
+func (s *ClientConn) handshake(host string) error {
+
+	// the size here is just an estimate
+	buf := make([]byte, 0, 6+len(host))
+
+	buf = append(buf, VERSION_V5)
+	if len(s.user) > 0 && len(s.user) < 256 && len(s.password) < 256 {
+		buf = append(buf, 2 /* num auth methods */, Method_NO_AUTH, Method_USER_PASS)
+	} else {
+		buf = append(buf, 1 /* num auth methods */, Method_NO_AUTH)
+	}
+	(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+	if _, err := (*s.conn).Write(buf); err != nil {
+		return errors.New("proxy: failed to write greeting to SOCKS5 proxy at " + s.addr + ": " + err.Error())
+	}
+	(*s.conn).SetDeadline(time.Time{})
+
+	(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+	if _, err := io.ReadFull((*s.conn), buf[:2]); err != nil {
+		return errors.New("proxy: failed to read greeting from SOCKS5 proxy at " + s.addr + ": " + err.Error())
+	}
+	(*s.conn).SetDeadline(time.Time{})
+
+	if buf[0] != 5 {
+		return errors.New("proxy: SOCKS5 proxy at " + s.addr + " has unexpected version " + strconv.Itoa(int(buf[0])))
+	}
+	if buf[1] == 0xff {
+		return errors.New("proxy: SOCKS5 proxy at " + s.addr + " requires authentication")
+	}
+
+	// See RFC 1929
+	if buf[1] == Method_USER_PASS {
+		buf = buf[:0]
+		buf = append(buf, 1 /* password protocol version */)
+		buf = append(buf, uint8(len(s.user)))
+		buf = append(buf, s.user...)
+		buf = append(buf, uint8(len(s.password)))
+		buf = append(buf, s.password...)
+		(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+		if _, err := (*s.conn).Write(buf); err != nil {
+			return errors.New("proxy: failed to write authentication request to SOCKS5 proxy at " + s.addr + ": " + err.Error())
+		}
+		(*s.conn).SetDeadline(time.Time{})
+		(*s.conn).SetDeadline(time.Now().Add(s.timeout))
+		if _, err := io.ReadFull((*s.conn), buf[:2]); err != nil {
+			return errors.New("proxy: failed to read authentication reply from SOCKS5 proxy at " + s.addr + ": " + err.Error())
+		}
+		(*s.conn).SetDeadline(time.Time{})
+		if buf[1] != 0 {
+			return errors.New("proxy: SOCKS5 proxy at " + s.addr + " rejected username/password")
+		}
+	}
+	return nil
+}
--- a/utils/socks/server.go
+++ b/utils/socks/server.go
@ -0,0 +1,226 @@
+package socks
+
+import (
+	"fmt"
+	"net"
+	"snail007/proxy/utils"
+	"strings"
+	"time"
+)
+
+const (
+	Method_NO_AUTH         = uint8(0x00)
+	Method_GSSAPI          = uint8(0x01)
+	Method_USER_PASS       = uint8(0x02)
+	Method_IANA            = uint8(0x7F)
+	Method_RESVERVE        = uint8(0x80)
+	Method_NONE_ACCEPTABLE = uint8(0xFF)
+	VERSION_V5             = uint8(0x05)
+	CMD_CONNECT            = uint8(0x01)
+	CMD_BIND               = uint8(0x02)
+	CMD_ASSOCIATE          = uint8(0x03)
+	ATYP_IPV4              = uint8(0x01)
+	ATYP_DOMAIN            = uint8(0x03)
+	ATYP_IPV6              = uint8(0x04)
+	REP_SUCCESS            = uint8(0x00)
+	REP_REQ_FAIL           = uint8(0x01)
+	REP_RULE_FORBIDDEN     = uint8(0x02)
+	REP_NETWOR_UNREACHABLE = uint8(0x03)
+	REP_HOST_UNREACHABLE   = uint8(0x04)
+	REP_CONNECTION_REFUSED = uint8(0x05)
+	REP_TTL_TIMEOUT        = uint8(0x06)
+	REP_CMD_UNSUPPORTED    = uint8(0x07)
+	REP_ATYP_UNSUPPORTED   = uint8(0x08)
+	REP_UNKNOWN            = uint8(0x09)
+	RSV                    = uint8(0x00)
+)
+
+var (
+	ZERO_IP   = []byte{0x00, 0x00, 0x00, 0x00}
+	ZERO_PORT = []byte{0x00, 0x00}
+)
+
+type ServerConn struct {
+	target   string
+	user     string
+	password string
+	conn     *net.Conn
+	timeout  time.Duration
+	auth     *utils.BasicAuth
+	header   []byte
+	ver      uint8
+	//method
+	methodsCount uint8
+	methods      []uint8
+	method       uint8
+	//request
+	cmd         uint8
+	reserve     uint8
+	addressType uint8
+	dstAddr     string
+	dstPort     string
+	dstHost     string
+	udpAddress  string
+}
+
+func NewServerConn(conn *net.Conn, timeout time.Duration, auth *utils.BasicAuth, udpAddress string, header []byte) *ServerConn {
+	if udpAddress == "" {
+		udpAddress = "0.0.0.0:16666"
+	}
+	s := &ServerConn{
+		conn:       conn,
+		timeout:    timeout,
+		auth:       auth,
+		header:     header,
+		ver:        VERSION_V5,
+		udpAddress: udpAddress,
+	}
+	return s
+
+}
+func (s *ServerConn) Close() {
+	utils.CloseConn(s.conn)
+}
+func (s *ServerConn) AuthData() Auth {
+	return Auth{s.user, s.password}
+}
+func (s *ServerConn) Method() uint8 {
+	return s.method
+}
+func (s *ServerConn) Target() string {
+	return s.target
+}
+func (s *ServerConn) Handshake() (err error) {
+	remoteAddr := (*s.conn).RemoteAddr()
+	//协商开始
+	//method select request
+	var methodReq MethodsRequest
+	(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+
+	methodReq, e := NewMethodsRequest((*s.conn), s.header)
+	(*s.conn).SetReadDeadline(time.Time{})
+	if e != nil {
+		(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+		methodReq.Reply(Method_NONE_ACCEPTABLE)
+		(*s.conn).SetReadDeadline(time.Time{})
+		err = fmt.Errorf("new methods request fail,ERR: %s", e)
+		return
+	}
+	//log.Printf("%v,s.auth == %v && methodReq.Select(Method_NO_AUTH) %v", methodReq.methods, s.auth, methodReq.Select(Method_NO_AUTH))
+	if s.auth == nil && methodReq.Select(Method_NO_AUTH) && !methodReq.Select(Method_USER_PASS) {
+		// if !methodReq.Select(Method_NO_AUTH) {
+		// 	(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+		// 	methodReq.Reply(Method_NONE_ACCEPTABLE)
+		// 	(*s.conn).SetReadDeadline(time.Time{})
+		// 	err = fmt.Errorf("none method found : Method_NO_AUTH")
+		// 	return
+		// }
+		s.method = Method_NO_AUTH
+		//method select reply
+		(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+		err = methodReq.Reply(Method_NO_AUTH)
+		(*s.conn).SetReadDeadline(time.Time{})
+		if err != nil {
+			err = fmt.Errorf("reply answer data fail,ERR: %s", err)
+			return
+		}
+		// err = fmt.Errorf("% x", methodReq.Bytes())
+	} else {
+		//auth
+		if !methodReq.Select(Method_USER_PASS) {
+			(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+			methodReq.Reply(Method_NONE_ACCEPTABLE)
+			(*s.conn).SetReadDeadline(time.Time{})
+			err = fmt.Errorf("none method found : Method_USER_PASS")
+			return
+		}
+		s.method = Method_USER_PASS
+		//method reply need auth
+		(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+		err = methodReq.Reply(Method_USER_PASS)
+		(*s.conn).SetReadDeadline(time.Time{})
+		if err != nil {
+			err = fmt.Errorf("reply answer data fail,ERR: %s", err)
+			return
+		}
+		//read auth
+		buf := make([]byte, 500)
+		var n int
+		(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+		n, err = (*s.conn).Read(buf)
+		(*s.conn).SetReadDeadline(time.Time{})
+		if err != nil {
+			err = fmt.Errorf("read auth info fail,ERR: %s", err)
+			return
+		}
+		r := buf[:n]
+		s.user = string(r[2 : r[1]+2])
+		s.password = string(r[2+r[1]+1:])
+		//err = fmt.Errorf("user:%s,pass:%s", user, pass)
+		//auth
+		_addr := strings.Split(remoteAddr.String(), ":")
+		if s.auth == nil || s.auth.CheckUserPass(s.user, s.password, _addr[0], "") {
+			(*s.conn).SetDeadline(time.Now().Add(time.Millisecond * time.Duration(s.timeout)))
+			_, err = (*s.conn).Write([]byte{0x01, 0x00})
+			(*s.conn).SetDeadline(time.Time{})
+			if err != nil {
+				err = fmt.Errorf("answer auth success to %s fail,ERR: %s", remoteAddr, err)
+				return
+			}
+		} else {
+			(*s.conn).SetDeadline(time.Now().Add(time.Millisecond * time.Duration(s.timeout)))
+			_, err = (*s.conn).Write([]byte{0x01, 0x01})
+			(*s.conn).SetDeadline(time.Time{})
+			if err != nil {
+				err = fmt.Errorf("answer auth fail to %s fail,ERR: %s", remoteAddr, err)
+				return
+			}
+			err = fmt.Errorf("auth fail from %s", remoteAddr)
+			return
+		}
+	}
+	//request detail
+	(*s.conn).SetReadDeadline(time.Now().Add(time.Second * s.timeout))
+	request, e := NewRequest(*s.conn)
+	(*s.conn).SetReadDeadline(time.Time{})
+	if e != nil {
+		err = fmt.Errorf("read request data fail,ERR: %s", e)
+		return
+	}
+	//协商结束
+
+	switch request.CMD() {
+	case CMD_BIND:
+		err = request.TCPReply(REP_UNKNOWN)
+		if err != nil {
+			err = fmt.Errorf("TCPReply REP_UNKNOWN to %s fail,ERR: %s", remoteAddr, err)
+			return
+		}
+		err = fmt.Errorf("cmd bind not supported, form: %s", remoteAddr)
+		return
+	case CMD_CONNECT:
+		err = request.TCPReply(REP_SUCCESS)
+		if err != nil {
+			err = fmt.Errorf("TCPReply REP_SUCCESS to %s fail,ERR: %s", remoteAddr, err)
+			return
+		}
+	case CMD_ASSOCIATE:
+		err = request.UDPReply(REP_SUCCESS, s.udpAddress)
+		if err != nil {
+			err = fmt.Errorf("UDPReply REP_SUCCESS to %s fail,ERR: %s", remoteAddr, err)
+			return
+		}
+	}
+
+	//fill socks info
+	s.target = request.Addr()
+	s.methodsCount = methodReq.MethodsCount()
+	s.methods = methodReq.Methods()
+	s.cmd = request.CMD()
+	s.reserve = request.reserve
+	s.addressType = request.addressType
+	s.dstAddr = request.dstAddr
+	s.dstHost = request.dstHost
+	s.dstPort = request.dstPort
+	return
+}
--- a/utils/socks/structs.go
+++ b/utils/socks/structs.go
@ -3,44 +3,13 @@ package socks
 import (
 	"bytes"
 	"encoding/binary"
+	"errors"
 	"fmt"
 	"io"
 	"net"
 	"strconv"
 )

-const (
-	Method_NO_AUTH         = uint8(0x00)
-	Method_GSSAPI          = uint8(0x01)
-	Method_USER_PASS       = uint8(0x02)
-	Method_IANA            = uint8(0x7F)
-	Method_RESVERVE        = uint8(0x80)
-	Method_NONE_ACCEPTABLE = uint8(0xFF)
-	VERSION_V5             = uint8(0x05)
-	CMD_CONNECT            = uint8(0x01)
-	CMD_BIND               = uint8(0x02)
-	CMD_ASSOCIATE          = uint8(0x03)
-	ATYP_IPV4              = uint8(0x01)
-	ATYP_DOMAIN            = uint8(0x03)
-	ATYP_IPV6              = uint8(0x04)
-	REP_SUCCESS            = uint8(0x00)
-	REP_REQ_FAIL           = uint8(0x01)
-	REP_RULE_FORBIDDEN     = uint8(0x02)
-	REP_NETWOR_UNREACHABLE = uint8(0x03)
-	REP_HOST_UNREACHABLE   = uint8(0x04)
-	REP_CONNECTION_REFUSED = uint8(0x05)
-	REP_TTL_TIMEOUT        = uint8(0x06)
-	REP_CMD_UNSUPPORTED    = uint8(0x07)
-	REP_ATYP_UNSUPPORTED   = uint8(0x08)
-	REP_UNKNOWN            = uint8(0x09)
-	RSV                    = uint8(0x00)
-)
-
-var (
-	ZERO_IP   = []byte{0x00, 0x00, 0x00, 0x00}
-	ZERO_PORT = []byte{0x00, 0x00}
-)
-
 type Request struct {
 	ver         uint8
 	cmd         uint8
@ -53,20 +22,24 @@ type Request struct {
 	rw          io.ReadWriter
 }

-func NewRequest(rw io.ReadWriter) (req Request, err interface{}) {
-	var b [1024]byte
+func NewRequest(rw io.ReadWriter, header ...[]byte) (req Request, err interface{}) {
+	var b = make([]byte, 1024)
 	var n int
 	req = Request{rw: rw}
-	n, err = rw.Read(b[:])
-	if err != nil {
-		err = fmt.Errorf("read req data fail,ERR: %s", err)
-		return
+	if header != nil && len(header) == 1 && len(header[0]) > 1 {
+		b = header[0]
+		n = len(header[0])
+	} else {
+		n, err = rw.Read(b[:])
+		if err != nil {
+			err = fmt.Errorf("read req data fail,ERR: %s", err)
+			return
+		}
 	}
 	req.ver = uint8(b[0])
 	req.cmd = uint8(b[1])
 	req.reserve = uint8(b[2])
 	req.addressType = uint8(b[3])
-
 	if b[0] != 0x5 {
 		err = fmt.Errorf("sosck version supported")
 		req.TCPReply(REP_REQ_FAIL)
@ -124,7 +97,7 @@ func (s *Request) NewReply(rep uint8, addr string) []byte {
 		ipv6[4], ipv6[5], ipv6[6], ipv6[7],
 		ipv6[8], ipv6[9], ipv6[10], ipv6[11],
 	)
-	if ipv6 != nil && "0000000000255255" != zeroiIPv6 {
+	if ipb == nil && ipv6 != nil && "0000000000255255" != zeroiIPv6 {
 		atyp = ATYP_IPV6
 		ipb = ip.To16()
 	}
@ -150,7 +123,7 @@ type MethodsRequest struct {
 	rw           *io.ReadWriter
 }

-func NewMethodsRequest(r io.ReadWriter) (s MethodsRequest, err interface{}) {
+func NewMethodsRequest(r io.ReadWriter, header ...[]byte) (s MethodsRequest, err interface{}) {
 	defer func() {
 		if err == nil {
 			err = recover()
@ -160,9 +133,14 @@ func NewMethodsRequest(r io.ReadWriter) (s MethodsRequest, err interface{}) {
 	s.rw = &r
 	var buf = make([]byte, 300)
 	var n int
-	n, err = r.Read(buf)
-	if err != nil {
-		return
+	if header != nil && len(header) == 1 && len(header[0]) > 1 {
+		buf = header[0]
+		n = len(header[0])
+	} else {
+		n, err = r.Read(buf)
+		if err != nil {
+			return
+		}
 	}
 	if buf[0] != 0x05 {
 		err = fmt.Errorf("socks version not supported")
@ -172,7 +150,6 @@ func NewMethodsRequest(r io.ReadWriter) (s MethodsRequest, err interface{}) {
 		err = fmt.Errorf("socks methods data length error")
 		return
 	}
-
 	s.ver = buf[0]
 	s.methodsCount = buf[1]
 	s.methods = buf[2:n]
@ -185,6 +162,9 @@ func (s *MethodsRequest) Version() uint8 {
 func (s *MethodsRequest) MethodsCount() uint8 {
 	return s.methodsCount
 }
+func (s *MethodsRequest) Methods() []uint8 {
+	return s.methods
+}
 func (s *MethodsRequest) Select(method uint8) bool {
 	for _, m := range s.methods {
 		if m == method {
@ -201,17 +181,6 @@ func (s *MethodsRequest) Bytes() []byte {
 	return s.bytes
 }

-type UDPPacket struct {
-	rsv     uint16
-	frag    uint8
-	atype   uint8
-	dstHost string
-	dstPort string
-	data    []byte
-	header  []byte
-	bytes   []byte
-}
-
 func ParseUDPPacket(b []byte) (p UDPPacket, err error) {
 	p = UDPPacket{}
 	p.frag = uint8(b[2])
@ -239,6 +208,18 @@ func ParseUDPPacket(b []byte) (p UDPPacket, err error) {
 	p.header = b[:portIndex+2]
 	return
 }
+
+type UDPPacket struct {
+	rsv     uint16
+	frag    uint8
+	atype   uint8
+	dstHost string
+	dstPort string
+	data    []byte
+	header  []byte
+	bytes   []byte
+}
+
 func (s *UDPPacket) Header() []byte {
 	return s.header
 }
@ -258,3 +239,104 @@ func (s *UDPPacket) Port() string {
 func (s *UDPPacket) Data() []byte {
 	return s.data
 }
+
+type PacketUDP struct {
+	rsv     uint16
+	frag    uint8
+	atype   uint8
+	dstHost string
+	dstPort string
+	data    []byte
+}
+
+func NewPacketUDP() (p PacketUDP) {
+	return PacketUDP{}
+}
+func (p *PacketUDP) Build(destAddr string, data []byte) (err error) {
+	host, port, err := net.SplitHostPort(destAddr)
+	if err != nil {
+		return
+	}
+	p.rsv = 0
+	p.frag = 0
+	p.dstHost = host
+	p.dstPort = port
+	p.atype = ATYP_IPV4
+	if ip := net.ParseIP(host); ip != nil {
+		if ip4 := ip.To4(); ip4 != nil {
+			p.atype = ATYP_IPV4
+			ip = ip4
+		} else {
+			p.atype = ATYP_IPV6
+		}
+	} else {
+		if len(host) > 255 {
+			err = errors.New("proxy: destination host name too long: " + host)
+			return
+		}
+		p.atype = ATYP_DOMAIN
+	}
+	p.data = data
+
+	return
+}
+func (p *PacketUDP) Parse(b []byte) (err error) {
+	p.frag = uint8(b[2])
+	if p.frag != 0 {
+		err = fmt.Errorf("FRAG only support for 0 , %v ,%v", p.frag, b[:4])
+		return
+	}
+	portIndex := 0
+	p.atype = b[3]
+	switch p.atype {
+	case ATYP_IPV4: //IP V4
+		p.dstHost = net.IPv4(b[4], b[5], b[6], b[7]).String()
+		portIndex = 8
+	case ATYP_DOMAIN: //域名
+		domainLen := uint8(b[4])
+		p.dstHost = string(b[5 : 5+domainLen]) //b[4]表示域名的长度
+		portIndex = int(5 + domainLen)
+	case ATYP_IPV6: //IP V6
+		p.dstHost = net.IP{b[4], b[5], b[6], b[7], b[8], b[9], b[10], b[11], b[12], b[13], b[14], b[15], b[16], b[17], b[18], b[19]}.String()
+		portIndex = 20
+	}
+	p.dstPort = strconv.Itoa(int(b[portIndex])<<8 | int(b[portIndex+1]))
+	p.data = b[portIndex+2:]
+	return
+}
+func (p *PacketUDP) Header() []byte {
+	header := new(bytes.Buffer)
+	header.Write([]byte{0x00, 0x00, p.frag, p.atype})
+	if p.atype == ATYP_IPV4 {
+		ip := net.ParseIP(p.dstHost)
+		header.Write(ip.To4())
+	} else if p.atype == ATYP_IPV6 {
+		ip := net.ParseIP(p.dstHost)
+		header.Write(ip.To16())
+	} else if p.atype == ATYP_DOMAIN {
+		hBytes := []byte(p.dstHost)
+		header.WriteByte(byte(len(hBytes)))
+		header.Write(hBytes)
+	}
+	port, _ := strconv.ParseUint(p.dstPort, 10, 64)
+	portBytes := new(bytes.Buffer)
+	binary.Write(portBytes, binary.BigEndian, port)
+	header.Write(portBytes.Bytes()[portBytes.Len()-2:])
+	return header.Bytes()
+}
+func (p *PacketUDP) Bytes() []byte {
+	packBytes := new(bytes.Buffer)
+	packBytes.Write(p.Header())
+	packBytes.Write(p.data)
+	return packBytes.Bytes()
+}
+func (p *PacketUDP) Host() string {
+	return p.dstHost
+}
+
+func (p *PacketUDP) Port() string {
+	return p.dstPort
+}
+func (p *PacketUDP) Data() []byte {
+	return p.data
+}
--- a/utils/structs.go
+++ b/utils/structs.go
@ -5,16 +5,21 @@ import (
 	"crypto/tls"
 	"encoding/base64"
 	"encoding/binary"
+	"errors"
 	"fmt"
 	"io"
 	"io/ioutil"
 	"log"
 	"net"
 	"net/url"
-	"proxy/utils/sni"
+	"snail007/proxy/services/kcpcfg"
+	"snail007/proxy/utils/sni"
 	"strings"
 	"sync"
 	"time"
+
+	"github.com/golang/snappy"
+	"github.com/miekg/dns"
 )

 type Checker struct {
@ -52,7 +57,10 @@ func NewChecker(timeout int, interval int64, blockedFile, directFile string) Che
 	if !ch.directMap.IsEmpty() {
 		log.Printf("direct file loaded , domains : %d", ch.directMap.Count())
 	}
-	ch.start()
+	if interval > 0 {
+		ch.start()
+	}
+
 	return ch
 }

@ -171,11 +179,13 @@ type BasicAuth struct {
 	authOkCode  int
 	authTimeout int
 	authRetry   int
+	dns         *DomainResolver
 }

-func NewBasicAuth() BasicAuth {
+func NewBasicAuth(dns *DomainResolver) BasicAuth {
 	return BasicAuth{
 		data: NewConcurrentMap(),
+		dns:  dns,
 	}
 }
 func (ba *BasicAuth) SetAuthURL(URL string, code, timeout, retry int) {
@ -239,18 +249,27 @@ func (ba *BasicAuth) checkFromURL(userpass, ip, target string) (err error) {
 	if len(u) != 2 {
 		return
 	}
+
 	URL := ba.authURL
 	if strings.Contains(URL, "?") {
 		URL += "&"
 	} else {
 		URL += "?"
 	}
-	URL += fmt.Sprintf("user=%s&pass=%s&ip=%s&target=%s", u[0], u[1], ip, target)
+	URL += fmt.Sprintf("user=%s&pass=%s&ip=%s&target=%s", u[0], u[1], ip, url.QueryEscape(target))
+	getURL := URL
+	var domain string
+	if ba.dns != nil {
+		_url, _ := url.Parse(ba.authURL)
+		domain = _url.Host
+		domainIP := ba.dns.MustResolve(domain)
+		getURL = strings.Replace(URL, domain, domainIP, 1)
+	}
 	var code int
 	var tryCount = 0
 	var body []byte
 	for tryCount <= ba.authRetry {
-		body, code, err = HttpGet(URL, ba.authTimeout)
+		body, code, err = HttpGet(getURL, ba.authTimeout, domain)
 		if err == nil && code == ba.authOkCode {
 			break
 		} else if err != nil {
@ -292,21 +311,27 @@ type HTTPRequest struct {
 	basicAuth   *BasicAuth
 }

-func NewHTTPRequest(inConn *net.Conn, bufSize int, isBasicAuth bool, basicAuth *BasicAuth) (req HTTPRequest, err error) {
+func NewHTTPRequest(inConn *net.Conn, bufSize int, isBasicAuth bool, basicAuth *BasicAuth, header ...[]byte) (req HTTPRequest, err error) {
 	buf := make([]byte, bufSize)
-	len := 0
+	n := 0
 	req = HTTPRequest{
 		conn: inConn,
 	}
-	len, err = (*inConn).Read(buf[:])
-	if err != nil {
-		if err != io.EOF {
-			err = fmt.Errorf("http decoder read err:%s", err)
+	if header != nil && len(header) == 1 && len(header[0]) > 1 {
+		buf = header[0]
+		n = len(header[0])
+	} else {
+		n, err = (*inConn).Read(buf[:])
+		if err != nil {
+			if err != io.EOF {
+				err = fmt.Errorf("http decoder read err:%s", err)
+			}
+			CloseConn(inConn)
+			return
 		}
-		CloseConn(inConn)
-		return
 	}
-	req.HeadBuf = buf[:len]
+
+	req.HeadBuf = buf[:n]
 	//fmt.Println(string(req.HeadBuf))
 	//try sni
 	serverName, err0 := sni.ServerNameFromBytes(req.HeadBuf)
@ -348,16 +373,14 @@ func (req *HTTPRequest) HTTP() (err error) {
 			return
 		}
 	}
-	req.URL, err = req.getHTTPURL()
-	if err == nil {
-		var u *url.URL
-		u, err = url.Parse(req.URL)
-		if err != nil {
-			return
-		}
-		req.Host = u.Host
-		req.addPortIfNot()
+	req.URL = req.getHTTPURL()
+	var u *url.URL
+	u, err = url.Parse(req.URL)
+	if err != nil {
+		return
 	}
+	req.Host = u.Host
+	req.addPortIfNot()
 	return
 }
 func (req *HTTPRequest) HTTPS() (err error) {
@ -369,7 +392,6 @@ func (req *HTTPRequest) HTTPS() (err error) {
 	}
 	req.Host = req.hostOrURL
 	req.addPortIfNot()
-	//_, err = fmt.Fprint(*req.conn, "HTTP/1.1 200 Connection established\r\n\r\n")
 	return
 }
 func (req *HTTPRequest) HTTPSReply() (err error) {
@ -380,26 +402,18 @@ func (req *HTTPRequest) IsHTTPS() bool {
 	return req.Method == "CONNECT"
 }

-func (req *HTTPRequest) BasicAuth() (err error) {
+func (req *HTTPRequest) GetAuthDataStr() (basicInfo string, err error) {
+	// log.Printf("request :%s", string(req.HeadBuf))
+	authorization := req.getHeader("Proxy-Authorization")

-	//log.Printf("request :%s", string(b[:n]))authorization
-	isProxyAuthorization := false
-	authorization, err := req.getHeader("Authorization")
-	if err != nil {
-		fmt.Fprint((*req.conn), "HTTP/1.1 401 Unauthorized\r\nWWW-Authenticate: Basic realm=\"\"\r\n\r\nUnauthorized")
+	authorization = strings.Trim(authorization, " \r\n\t")
+	if authorization == "" {
+		fmt.Fprintf((*req.conn), "HTTP/1.1 %s Unauthorized\r\nWWW-Authenticate: Basic realm=\"\"\r\n\r\nUnauthorized", "407")
 		CloseConn(req.conn)
+		err = errors.New("require auth header data")
 		return
 	}
-	if authorization == "" {
-		authorization, err = req.getHeader("Proxy-Authorization")
-		if err != nil {
-			fmt.Fprint((*req.conn), "HTTP/1.1 407 Unauthorized\r\nWWW-Authenticate: Basic realm=\"\"\r\n\r\nUnauthorized")
-			CloseConn(req.conn)
-			return
-		}
-		isProxyAuthorization = true
-	}
-	//log.Printf("Authorization:%s", authorization)
+	//log.Printf("Authorization:%authorization = req.getHeader("Authorization")
 	basic := strings.Fields(authorization)
 	if len(basic) != 2 {
 		err = fmt.Errorf("authorization data error,ERR:%s", authorization)
@ -412,39 +426,43 @@ func (req *HTTPRequest) BasicAuth() (err error) {
 		CloseConn(req.conn)
 		return
 	}
+	basicInfo = string(user)
+	return
+}
+func (req *HTTPRequest) BasicAuth() (err error) {
 	addr := strings.Split((*req.conn).RemoteAddr().String(), ":")
 	URL := ""
 	if req.IsHTTPS() {
 		URL = "https://" + req.Host
 	} else {
-		URL, _ = req.getHTTPURL()
+		URL = req.getHTTPURL()
+	}
+	user, err := req.GetAuthDataStr()
+	if err != nil {
+		return
 	}
 	authOk := (*req.basicAuth).Check(string(user), addr[0], URL)
 	//log.Printf("auth %s,%v", string(user), authOk)
 	if !authOk {
-		code := "401"
-		if isProxyAuthorization {
-			code = "407"
-		}
-		fmt.Fprintf((*req.conn), "HTTP/1.1 %s Unauthorized\r\n\r\nUnauthorized", code)
+		fmt.Fprintf((*req.conn), "HTTP/1.1 %s Unauthorized\r\n\r\nUnauthorized", "407")
 		CloseConn(req.conn)
 		err = fmt.Errorf("basic auth fail")
 		return
 	}
 	return
 }
-func (req *HTTPRequest) getHTTPURL() (URL string, err error) {
+func (req *HTTPRequest) getHTTPURL() (URL string) {
 	if !strings.HasPrefix(req.hostOrURL, "/") {
-		return req.hostOrURL, nil
+		return req.hostOrURL
 	}
-	_host, err := req.getHeader("host")
-	if err != nil {
+	_host := req.getHeader("host")
+	if _host == "" {
 		return
 	}
 	URL = fmt.Sprintf("http://%s%s", _host, req.hostOrURL)
 	return
 }
-func (req *HTTPRequest) getHeader(key string) (val string, err error) {
+func (req *HTTPRequest) getHeader(key string) (val string) {
 	key = strings.ToUpper(key)
 	lines := strings.Split(string(req.HeadBuf), "\r\n")
 	//log.Println(lines)
@ -477,27 +495,27 @@ func (req *HTTPRequest) addPortIfNot() (newHost string) {
 }

 type OutPool struct {
-	Pool      ConnPool
-	dur       int
-	typ       string
-	certBytes []byte
-	keyBytes  []byte
-	kcpMethod string
-	kcpKey    string
-	address   string
-	timeout   int
+	Pool        ConnPool
+	dur         int
+	typ         string
+	certBytes   []byte
+	keyBytes    []byte
+	caCertBytes []byte
+	kcp         kcpcfg.KCPConfigArgs
+	address     string
+	timeout     int
 }

-func NewOutPool(dur int, typ, kcpMethod, kcpKey string, certBytes, keyBytes []byte, address string, timeout int, InitialCap int, MaxCap int) (op OutPool) {
+func NewOutPool(dur int, typ string, kcp kcpcfg.KCPConfigArgs, certBytes, keyBytes, caCertBytes []byte, address string, timeout int, InitialCap int, MaxCap int) (op OutPool) {
 	op = OutPool{
-		dur:       dur,
-		typ:       typ,
-		certBytes: certBytes,
-		keyBytes:  keyBytes,
-		kcpMethod: kcpMethod,
-		kcpKey:    kcpKey,
-		address:   address,
-		timeout:   timeout,
+		dur:         dur,
+		typ:         typ,
+		certBytes:   certBytes,
+		keyBytes:    keyBytes,
+		caCertBytes: caCertBytes,
+		kcp:         kcp,
+		address:     address,
+		timeout:     timeout,
 	}
 	var err error
 	op.Pool, err = NewConnPool(poolConfig{
@ -531,12 +549,12 @@ func NewOutPool(dur int, typ, kcpMethod, kcpKey string, certBytes, keyBytes []by
 func (op *OutPool) getConn() (conn interface{}, err error) {
 	if op.typ == "tls" {
 		var _conn tls.Conn
-		_conn, err = TlsConnectHost(op.address, op.timeout, op.certBytes, op.keyBytes)
+		_conn, err = TlsConnectHost(op.address, op.timeout, op.certBytes, op.keyBytes, op.caCertBytes)
 		if err == nil {
 			conn = net.Conn(&_conn)
 		}
 	} else if op.typ == "kcp" {
-		conn, err = ConnectKCPHost(op.address, op.kcpMethod, op.kcpKey)
+		conn, err = ConnectKCPHost(op.address, op.kcp)
 	} else {
 		conn, err = ConnectHost(op.address, op.timeout)
 	}
@ -773,3 +791,184 @@ func (cm *ConnManager) RemoveAll() {
 		cm.Remove(k)
 	}
 }
+
+type ClientKeyRouter struct {
+	keyChan chan string
+	ctrl    *ConcurrentMap
+	lock    *sync.Mutex
+}
+
+func NewClientKeyRouter(ctrl *ConcurrentMap, size int) ClientKeyRouter {
+	return ClientKeyRouter{
+		keyChan: make(chan string, size),
+		ctrl:    ctrl,
+		lock:    &sync.Mutex{},
+	}
+}
+func (c *ClientKeyRouter) GetKey() string {
+	defer c.lock.Unlock()
+	c.lock.Lock()
+	if len(c.keyChan) == 0 {
+	EXIT:
+		for _, k := range c.ctrl.Keys() {
+			select {
+			case c.keyChan <- k:
+			default:
+				goto EXIT
+			}
+		}
+	}
+	for {
+		if len(c.keyChan) == 0 {
+			return "*"
+		}
+		select {
+		case key := <-c.keyChan:
+			if c.ctrl.Has(key) {
+				return key
+			}
+		default:
+			return "*"
+		}
+	}
+
+}
+
+type DomainResolver struct {
+	ttl         int
+	dnsAddrress string
+	data        ConcurrentMap
+}
+type DomainResolverItem struct {
+	ip        string
+	domain    string
+	expiredAt int64
+}
+
+func NewDomainResolver(dnsAddrress string, ttl int) DomainResolver {
+
+	return DomainResolver{
+		ttl:         ttl,
+		dnsAddrress: dnsAddrress,
+		data:        NewConcurrentMap(),
+	}
+}
+func (a *DomainResolver) MustResolve(address string) (ip string) {
+	ip, _ = a.Resolve(address)
+	return
+}
+func (a *DomainResolver) Resolve(address string) (ip string, err error) {
+	domain := address
+	port := ""
+	fromCache := "false"
+	defer func() {
+		if port != "" {
+			ip = net.JoinHostPort(ip, port)
+		}
+		log.Printf("dns:%s->%s,cache:%s", address, ip, fromCache)
+		//a.PrintData()
+	}()
+	if strings.Contains(domain, ":") {
+		domain, port, err = net.SplitHostPort(domain)
+		if err != nil {
+			return
+		}
+	}
+	if net.ParseIP(domain) != nil {
+		ip = domain
+		fromCache = "ip ignore"
+		return
+	}
+	item, ok := a.data.Get(domain)
+	if ok {
+		//log.Println("find ", domain)
+		if (*item.(*DomainResolverItem)).expiredAt > time.Now().Unix() {
+			ip = (*item.(*DomainResolverItem)).ip
+			fromCache = "true"
+			//log.Println("from cache ", domain)
+			return
+		}
+	} else {
+		item = &DomainResolverItem{
+			domain: domain,
+		}
+
+	}
+	c := new(dns.Client)
+	c.DialTimeout = time.Millisecond * 5000
+	c.ReadTimeout = time.Millisecond * 5000
+	c.WriteTimeout = time.Millisecond * 5000
+	m := new(dns.Msg)
+	m.SetQuestion(dns.Fqdn(domain), dns.TypeA)
+	m.RecursionDesired = true
+	r, _, err := c.Exchange(m, a.dnsAddrress)
+	if r == nil {
+		return
+	}
+	if r.Rcode != dns.RcodeSuccess {
+		err = fmt.Errorf(" *** invalid answer name %s after A query for %s", domain, a.dnsAddrress)
+		return
+	}
+	for _, answer := range r.Answer {
+		if answer.Header().Rrtype == dns.TypeA {
+			info := strings.Fields(answer.String())
+			if len(info) >= 5 {
+				ip = info[4]
+				_item := item.(*DomainResolverItem)
+				(*_item).expiredAt = time.Now().Unix() + int64(a.ttl)
+				(*_item).ip = ip
+				a.data.Set(domain, item)
+				return
+			}
+		}
+	}
+	return
+}
+func (a *DomainResolver) PrintData() {
+	for k, item := range a.data.Items() {
+		d := item.(*DomainResolverItem)
+		fmt.Printf("%s:ip[%s],domain[%s],expired at[%d]\n", k, (*d).ip, (*d).domain, (*d).expiredAt)
+	}
+}
+func NewCompStream(conn net.Conn) *CompStream {
+	c := new(CompStream)
+	c.conn = conn
+	c.w = snappy.NewBufferedWriter(conn)
+	c.r = snappy.NewReader(conn)
+	return c
+}
+
+type CompStream struct {
+	conn net.Conn
+	w    *snappy.Writer
+	r    *snappy.Reader
+}
+
+func (c *CompStream) Read(p []byte) (n int, err error) {
+	return c.r.Read(p)
+}
+
+func (c *CompStream) Write(p []byte) (n int, err error) {
+	n, err = c.w.Write(p)
+	err = c.w.Flush()
+	return n, err
+}
+
+func (c *CompStream) Close() error {
+	return c.conn.Close()
+}
+func (c *CompStream) LocalAddr() net.Addr {
+	return c.conn.LocalAddr()
+}
+func (c *CompStream) RemoteAddr() net.Addr {
+	return c.conn.RemoteAddr()
+}
+func (c *CompStream) SetDeadline(t time.Time) error {
+	return c.conn.SetDeadline(t)
+}
+func (c *CompStream) SetReadDeadline(t time.Time) error {
+	return c.conn.SetReadDeadline(t)
+}
+func (c *CompStream) SetWriteDeadline(t time.Time) error {
+	return c.conn.SetWriteDeadline(t)
+}
--- a/vendor/github.com/miekg/dns/AUTHORS
+++ b/vendor/github.com/miekg/dns/AUTHORS
@ -0,0 +1 @@
+Miek Gieben <miek@miek.nl>
--- a/vendor/github.com/miekg/dns/CONTRIBUTORS
+++ b/vendor/github.com/miekg/dns/CONTRIBUTORS
@ -0,0 +1,10 @@
+Alex A. Skinner
+Andrew Tunnell-Jones
+Ask Bjørn Hansen
+Dave Cheney
+Dusty Wilson
+Marek Majkowski
+Peter van Dijk
+Omri Bahumi
+Alex Sergeyev
+James Hartig
--- a/vendor/github.com/miekg/dns/COPYRIGHT
+++ b/vendor/github.com/miekg/dns/COPYRIGHT
@ -0,0 +1,9 @@
+Copyright 2009 The Go Authors. All rights reserved. Use of this source code
+is governed by a BSD-style license that can be found in the LICENSE file.
+Extensions of the original work are copyright (c) 2011 Miek Gieben
+
+Copyright 2011 Miek Gieben. All rights reserved. Use of this source code is
+governed by a BSD-style license that can be found in the LICENSE file.
+
+Copyright 2014 CloudFlare. All rights reserved. Use of this source code is
+governed by a BSD-style license that can be found in the LICENSE file.
--- a/vendor/github.com/miekg/dns/Gopkg.lock
+++ b/vendor/github.com/miekg/dns/Gopkg.lock
@ -0,0 +1,21 @@
+# This file is autogenerated, do not edit; changes may be undone by the next 'dep ensure'.
+
+
+[[projects]]
+  branch = "master"
+  name = "golang.org/x/crypto"
+  packages = ["ed25519","ed25519/internal/edwards25519"]
+  revision = "b080dc9a8c480b08e698fb1219160d598526310f"
+
+[[projects]]
+  branch = "master"
+  name = "golang.org/x/net"
+  packages = ["bpf","internal/iana","internal/socket","ipv4","ipv6"]
+  revision = "894f8ed5849b15b810ae41e9590a0d05395bba27"
+
+[solve-meta]
+  analyzer-name = "dep"
+  analyzer-version = 1
+  inputs-digest = "c4abc38abaeeeeb9be92455c9c02cae32841122b8982aaa067ef25bb8e86ff9d"
+  solver-name = "gps-cdcl"
+  solver-version = 1
--- a/vendor/github.com/miekg/dns/Gopkg.toml
+++ b/vendor/github.com/miekg/dns/Gopkg.toml
@ -0,0 +1,26 @@
+
+# Gopkg.toml example
+#
+# Refer to https://github.com/golang/dep/blob/master/docs/Gopkg.toml.md
+# for detailed Gopkg.toml documentation.
+#
+# required = ["github.com/user/thing/cmd/thing"]
+# ignored = ["github.com/user/project/pkgX", "bitbucket.org/user/project/pkgA/pkgY"]
+#
+# [[constraint]]
+#   name = "github.com/user/project"
+#   version = "1.0.0"
+#
+# [[constraint]]
+#   name = "github.com/user/project2"
+#   branch = "dev"
+#   source = "github.com/myfork/project2"
+#
+# [[override]]
+#  name = "github.com/x/y"
+#  version = "2.4.0"
+
+
+[[constraint]]
+  branch = "master"
+  name = "golang.org/x/crypto"
--- a/vendor/github.com/miekg/dns/LICENSE
+++ b/vendor/github.com/miekg/dns/LICENSE
@ -0,0 +1,32 @@
+Extensions of the original work are copyright (c) 2011 Miek Gieben
+
+As this is fork of the official Go code the same license applies:
+
+Copyright (c) 2009 The Go Authors. All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are
+met:
+
+   * Redistributions of source code must retain the above copyright
+notice, this list of conditions and the following disclaimer.
+   * Redistributions in binary form must reproduce the above
+copyright notice, this list of conditions and the following disclaimer
+in the documentation and/or other materials provided with the
+distribution.
+   * Neither the name of Google Inc. nor the names of its
+contributors may be used to endorse or promote products derived from
+this software without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
--- a/vendor/github.com/miekg/dns/Makefile.fuzz
+++ b/vendor/github.com/miekg/dns/Makefile.fuzz
@ -0,0 +1,33 @@
+# Makefile for fuzzing
+#
+# Use go-fuzz and needs the tools installed.
+# See https://blog.cloudflare.com/dns-parser-meet-go-fuzzer/
+#
+# Installing go-fuzz:
+# $ make -f Makefile.fuzz get
+# Installs:
+# * github.com/dvyukov/go-fuzz/go-fuzz
+# * get github.com/dvyukov/go-fuzz/go-fuzz-build
+
+all: build
+
+.PHONY: build
+build:
+	go-fuzz-build -tags fuzz github.com/miekg/dns
+
+.PHONY: build-newrr
+build-newrr:
+	go-fuzz-build -func FuzzNewRR -tags fuzz github.com/miekg/dns
+
+.PHONY: fuzz
+fuzz:
+	go-fuzz -bin=dns-fuzz.zip -workdir=fuzz
+
+.PHONY: get
+get:
+	go get github.com/dvyukov/go-fuzz/go-fuzz
+	go get github.com/dvyukov/go-fuzz/go-fuzz-build
+
+.PHONY: clean
+clean:
+	rm *-fuzz.zip
--- a/vendor/github.com/miekg/dns/Makefile.release
+++ b/vendor/github.com/miekg/dns/Makefile.release
@ -0,0 +1,52 @@
+# Makefile for releasing.
+#
+# The release is controlled from version.go. The version found there is
+# used to tag the git repo, we're not building any artifects so there is nothing
+# to upload to github.
+#
+# * Up the version in version.go
+# * Run: make -f Makefile.release release
+#   * will *commit* your change with 'Release $VERSION'
+#   * push to github
+#
+
+define GO
+//+build ignore
+
+package main
+
+import (
+	"fmt"
+
+	"github.com/miekg/dns"
+)
+
+func main() {
+	fmt.Println(dns.Version.String())
+}
+endef
+
+$(file > version_release.go,$(GO))
+VERSION:=$(shell go run version_release.go)
+TAG="v$(VERSION)"
+
+all:
+	@echo Use the \'release\' target to start a release $(VERSION)
+	rm -f version_release.go
+
+.PHONY: release
+release: commit push
+	@echo Released $(VERSION)
+	rm -f version_release.go
+
+.PHONY: commit
+commit:
+	@echo Committing release $(VERSION)
+	git commit -am"Release $(VERSION)"
+	git tag $(TAG)
+
+.PHONY: push
+push:
+	@echo Pushing release $(VERSION) to master
+	git push --tags
+	git push
--- a/vendor/github.com/miekg/dns/README.md
+++ b/vendor/github.com/miekg/dns/README.md
@ -0,0 +1,168 @@
+[![Build Status](https://travis-ci.org/miekg/dns.svg?branch=master)](https://travis-ci.org/miekg/dns)
+[![Code Coverage](https://img.shields.io/codecov/c/github/miekg/dns/master.svg)](https://codecov.io/github/miekg/dns?branch=master)
+[![Go Report Card](https://goreportcard.com/badge/github.com/miekg/dns)](https://goreportcard.com/report/miekg/dns)
+[![](https://godoc.org/github.com/miekg/dns?status.svg)](https://godoc.org/github.com/miekg/dns)
+
+# Alternative (more granular) approach to a DNS library
+
+> Less is more.
+
+Complete and usable DNS library. All widely used Resource Records are supported, including the
+DNSSEC types. It follows a lean and mean philosophy. If there is stuff you should know as a DNS
+programmer there isn't a convenience function for it. Server side and client side programming is
+supported, i.e. you can build servers and resolvers with it.
+
+We try to keep the "master" branch as sane as possible and at the bleeding edge of standards,
+avoiding breaking changes wherever reasonable. We support the last two versions of Go.
+
+# Goals
+
+* KISS;
+* Fast;
+* Small API. If it's easy to code in Go, don't make a function for it.
+
+# Users
+
+A not-so-up-to-date-list-that-may-be-actually-current:
+
+* https://github.com/coredns/coredns
+* https://cloudflare.com
+* https://github.com/abh/geodns
+* http://www.statdns.com/
+* http://www.dnsinspect.com/
+* https://github.com/chuangbo/jianbing-dictionary-dns
+* http://www.dns-lg.com/
+* https://github.com/fcambus/rrda
+* https://github.com/kenshinx/godns
+* https://github.com/skynetservices/skydns
+* https://github.com/hashicorp/consul
+* https://github.com/DevelopersPL/godnsagent
+* https://github.com/duedil-ltd/discodns
+* https://github.com/StalkR/dns-reverse-proxy
+* https://github.com/tianon/rawdns
+* https://mesosphere.github.io/mesos-dns/
+* https://pulse.turbobytes.com/
+* https://play.google.com/store/apps/details?id=com.turbobytes.dig
+* https://github.com/fcambus/statzone
+* https://github.com/benschw/dns-clb-go
+* https://github.com/corny/dnscheck for http://public-dns.info/
+* https://namesmith.io
+* https://github.com/miekg/unbound
+* https://github.com/miekg/exdns
+* https://dnslookup.org
+* https://github.com/looterz/grimd
+* https://github.com/phamhongviet/serf-dns
+* https://github.com/mehrdadrad/mylg
+* https://github.com/bamarni/dockness
+* https://github.com/fffaraz/microdns
+* http://kelda.io
+* https://github.com/ipdcode/hades (JD.COM)
+* https://github.com/StackExchange/dnscontrol/
+* https://www.dnsperf.com/
+* https://dnssectest.net/
+* https://dns.apebits.com
+* https://github.com/oif/apex
+* https://github.com/jedisct1/dnscrypt-proxy
+* https://github.com/jedisct1/rpdns
+
+Send pull request if you want to be listed here.
+
+# Features
+
+* UDP/TCP queries, IPv4 and IPv6;
+* RFC 1035 zone file parsing ($INCLUDE, $ORIGIN, $TTL and $GENERATE (for all record types) are supported;
+* Fast:
+    * Reply speed around ~ 80K qps (faster hardware results in more qps);
+    * Parsing RRs ~ 100K RR/s, that's 5M records in about 50 seconds;
+* Server side programming (mimicking the net/http package);
+* Client side programming;
+* DNSSEC: signing, validating and key generation for DSA, RSA, ECDSA and Ed25519;
+* EDNS0, NSID, Cookies;
+* AXFR/IXFR;
+* TSIG, SIG(0);
+* DNS over TLS: optional encrypted connection between client and server;
+* DNS name compression;
+* Depends only on the standard library.
+
+Have fun!
+
+Miek Gieben  -  2010-2012  -  <miek@miek.nl>
+
+# Building
+
+Building is done with the `go` tool. If you have setup your GOPATH correctly, the following should
+work:
+
+    go get github.com/miekg/dns
+    go build github.com/miekg/dns
+
+## Examples
+
+A short "how to use the API" is at the beginning of doc.go (this also will show
+when you call `godoc github.com/miekg/dns`).
+
+Example programs can be found in the `github.com/miekg/exdns` repository.
+
+## Supported RFCs
+
+*all of them*
+
+* 103{4,5} - DNS standard
+* 1348 - NSAP record (removed the record)
+* 1982 - Serial Arithmetic
+* 1876 - LOC record
+* 1995 - IXFR
+* 1996 - DNS notify
+* 2136 - DNS Update (dynamic updates)
+* 2181 - RRset definition - there is no RRset type though, just []RR
+* 2537 - RSAMD5 DNS keys
+* 2065 - DNSSEC (updated in later RFCs)
+* 2671 - EDNS record
+* 2782 - SRV record
+* 2845 - TSIG record
+* 2915 - NAPTR record
+* 2929 - DNS IANA Considerations
+* 3110 - RSASHA1 DNS keys
+* 3225 - DO bit (DNSSEC OK)
+* 340{1,2,3} - NAPTR record
+* 3445 - Limiting the scope of (DNS)KEY
+* 3597 - Unknown RRs
+* 403{3,4,5} - DNSSEC + validation functions
+* 4255 - SSHFP record
+* 4343 - Case insensitivity
+* 4408 - SPF record
+* 4509 - SHA256 Hash in DS
+* 4592 - Wildcards in the DNS
+* 4635 - HMAC SHA TSIG
+* 4701 - DHCID
+* 4892 - id.server
+* 5001 - NSID
+* 5155 - NSEC3 record
+* 5205 - HIP record
+* 5702 - SHA2 in the DNS
+* 5936 - AXFR
+* 5966 - TCP implementation recommendations
+* 6605 - ECDSA
+* 6725 - IANA Registry Update
+* 6742 - ILNP DNS
+* 6840 - Clarifications and Implementation Notes for DNS Security
+* 6844 - CAA record
+* 6891 - EDNS0 update
+* 6895 - DNS IANA considerations
+* 6975 - Algorithm Understanding in DNSSEC
+* 7043 - EUI48/EUI64 records
+* 7314 - DNS (EDNS) EXPIRE Option
+* 7477 - CSYNC RR
+* 7828 - edns-tcp-keepalive EDNS0 Option
+* 7553 - URI record
+* 7858 - DNS over TLS: Initiation and Performance Considerations
+* 7871 - EDNS0 Client Subnet
+* 7873 - Domain Name System (DNS) Cookies (draft-ietf-dnsop-cookies)
+* 8080 - EdDSA for DNSSEC
+
+## Loosely based upon
+
+* `ldns`
+* `NSD`
+* `Net::DNS`
+* `GRONG`
--- a/vendor/github.com/miekg/dns/client.go
+++ b/vendor/github.com/miekg/dns/client.go
@ -0,0 +1,506 @@
+package dns
+
+// A client implementation.
+
+import (
+	"bytes"
+	"context"
+	"crypto/tls"
+	"encoding/binary"
+	"io"
+	"net"
+	"strings"
+	"time"
+)
+
+const dnsTimeout time.Duration = 2 * time.Second
+const tcpIdleTimeout time.Duration = 8 * time.Second
+
+// A Conn represents a connection to a DNS server.
+type Conn struct {
+	net.Conn                         // a net.Conn holding the connection
+	UDPSize        uint16            // minimum receive buffer for UDP messages
+	TsigSecret     map[string]string // secret(s) for Tsig map[<zonename>]<base64 secret>, zonename must be in canonical form (lowercase, fqdn, see RFC 4034 Section 6.2)
+	rtt            time.Duration
+	t              time.Time
+	tsigRequestMAC string
+}
+
+// A Client defines parameters for a DNS client.
+type Client struct {
+	Net       string      // if "tcp" or "tcp-tls" (DNS over TLS) a TCP query will be initiated, otherwise an UDP one (default is "" for UDP)
+	UDPSize   uint16      // minimum receive buffer for UDP messages
+	TLSConfig *tls.Config // TLS connection configuration
+	Dialer    *net.Dialer // a net.Dialer used to set local address, timeouts and more
+	// Timeout is a cumulative timeout for dial, write and read, defaults to 0 (disabled) - overrides DialTimeout, ReadTimeout,
+	// WriteTimeout when non-zero. Can be overridden with net.Dialer.Timeout (see Client.ExchangeWithDialer and
+	// Client.Dialer) or context.Context.Deadline (see the deprecated ExchangeContext)
+	Timeout        time.Duration
+	DialTimeout    time.Duration     // net.DialTimeout, defaults to 2 seconds, or net.Dialer.Timeout if expiring earlier - overridden by Timeout when that value is non-zero
+	ReadTimeout    time.Duration     // net.Conn.SetReadTimeout value for connections, defaults to 2 seconds - overridden by Timeout when that value is non-zero
+	WriteTimeout   time.Duration     // net.Conn.SetWriteTimeout value for connections, defaults to 2 seconds - overridden by Timeout when that value is non-zero
+	TsigSecret     map[string]string // secret(s) for Tsig map[<zonename>]<base64 secret>, zonename must be in canonical form (lowercase, fqdn, see RFC 4034 Section 6.2)
+	SingleInflight bool              // if true suppress multiple outstanding queries for the same Qname, Qtype and Qclass
+	group          singleflight
+}
+
+// Exchange performs a synchronous UDP query. It sends the message m to the address
+// contained in a and waits for a reply. Exchange does not retry a failed query, nor
+// will it fall back to TCP in case of truncation.
+// See client.Exchange for more information on setting larger buffer sizes.
+func Exchange(m *Msg, a string) (r *Msg, err error) {
+	client := Client{Net: "udp"}
+	r, _, err = client.Exchange(m, a)
+	return r, err
+}
+
+func (c *Client) dialTimeout() time.Duration {
+	if c.Timeout != 0 {
+		return c.Timeout
+	}
+	if c.DialTimeout != 0 {
+		return c.DialTimeout
+	}
+	return dnsTimeout
+}
+
+func (c *Client) readTimeout() time.Duration {
+	if c.ReadTimeout != 0 {
+		return c.ReadTimeout
+	}
+	return dnsTimeout
+}
+
+func (c *Client) writeTimeout() time.Duration {
+	if c.WriteTimeout != 0 {
+		return c.WriteTimeout
+	}
+	return dnsTimeout
+}
+
+// Dial connects to the address on the named network.
+func (c *Client) Dial(address string) (conn *Conn, err error) {
+	// create a new dialer with the appropriate timeout
+	var d net.Dialer
+	if c.Dialer == nil {
+		d = net.Dialer{}
+	} else {
+		d = net.Dialer(*c.Dialer)
+	}
+	d.Timeout = c.getTimeoutForRequest(c.writeTimeout())
+
+	network := "udp"
+	useTLS := false
+
+	switch c.Net {
+	case "tcp-tls":
+		network = "tcp"
+		useTLS = true
+	case "tcp4-tls":
+		network = "tcp4"
+		useTLS = true
+	case "tcp6-tls":
+		network = "tcp6"
+		useTLS = true
+	default:
+		if c.Net != "" {
+			network = c.Net
+		}
+	}
+
+	conn = new(Conn)
+	if useTLS {
+		conn.Conn, err = tls.DialWithDialer(&d, network, address, c.TLSConfig)
+	} else {
+		conn.Conn, err = d.Dial(network, address)
+	}
+	if err != nil {
+		return nil, err
+	}
+	return conn, nil
+}
+
+// Exchange performs a synchronous query. It sends the message m to the address
+// contained in a and waits for a reply. Basic use pattern with a *dns.Client:
+//
+//	c := new(dns.Client)
+//	in, rtt, err := c.Exchange(message, "127.0.0.1:53")
+//
+// Exchange does not retry a failed query, nor will it fall back to TCP in
+// case of truncation.
+// It is up to the caller to create a message that allows for larger responses to be
+// returned. Specifically this means adding an EDNS0 OPT RR that will advertise a larger
+// buffer, see SetEdns0. Messages without an OPT RR will fallback to the historic limit
+// of 512 bytes
+// To specify a local address or a timeout, the caller has to set the `Client.Dialer`
+// attribute appropriately
+func (c *Client) Exchange(m *Msg, address string) (r *Msg, rtt time.Duration, err error) {
+	if !c.SingleInflight {
+		return c.exchange(m, address)
+	}
+
+	t := "nop"
+	if t1, ok := TypeToString[m.Question[0].Qtype]; ok {
+		t = t1
+	}
+	cl := "nop"
+	if cl1, ok := ClassToString[m.Question[0].Qclass]; ok {
+		cl = cl1
+	}
+	r, rtt, err, shared := c.group.Do(m.Question[0].Name+t+cl, func() (*Msg, time.Duration, error) {
+		return c.exchange(m, address)
+	})
+	if r != nil && shared {
+		r = r.Copy()
+	}
+	return r, rtt, err
+}
+
+func (c *Client) exchange(m *Msg, a string) (r *Msg, rtt time.Duration, err error) {
+	var co *Conn
+
+	co, err = c.Dial(a)
+
+	if err != nil {
+		return nil, 0, err
+	}
+	defer co.Close()
+
+	opt := m.IsEdns0()
+	// If EDNS0 is used use that for size.
+	if opt != nil && opt.UDPSize() >= MinMsgSize {
+		co.UDPSize = opt.UDPSize()
+	}
+	// Otherwise use the client's configured UDP size.
+	if opt == nil && c.UDPSize >= MinMsgSize {
+		co.UDPSize = c.UDPSize
+	}
+
+	co.TsigSecret = c.TsigSecret
+	// write with the appropriate write timeout
+	co.SetWriteDeadline(time.Now().Add(c.getTimeoutForRequest(c.writeTimeout())))
+	if err = co.WriteMsg(m); err != nil {
+		return nil, 0, err
+	}
+
+	co.SetReadDeadline(time.Now().Add(c.getTimeoutForRequest(c.readTimeout())))
+	r, err = co.ReadMsg()
+	if err == nil && r.Id != m.Id {
+		err = ErrId
+	}
+	return r, co.rtt, err
+}
+
+// ReadMsg reads a message from the connection co.
+// If the received message contains a TSIG record the transaction signature
+// is verified. This method always tries to return the message, however if an
+// error is returned there are no guarantees that the returned message is a
+// valid representation of the packet read.
+func (co *Conn) ReadMsg() (*Msg, error) {
+	p, err := co.ReadMsgHeader(nil)
+	if err != nil {
+		return nil, err
+	}
+
+	m := new(Msg)
+	if err := m.Unpack(p); err != nil {
+		// If an error was returned, we still want to allow the user to use
+		// the message, but naively they can just check err if they don't want
+		// to use an erroneous message
+		return m, err
+	}
+	if t := m.IsTsig(); t != nil {
+		if _, ok := co.TsigSecret[t.Hdr.Name]; !ok {
+			return m, ErrSecret
+		}
+		// Need to work on the original message p, as that was used to calculate the tsig.
+		err = TsigVerify(p, co.TsigSecret[t.Hdr.Name], co.tsigRequestMAC, false)
+	}
+	return m, err
+}
+
+// ReadMsgHeader reads a DNS message, parses and populates hdr (when hdr is not nil).
+// Returns message as a byte slice to be parsed with Msg.Unpack later on.
+// Note that error handling on the message body is not possible as only the header is parsed.
+func (co *Conn) ReadMsgHeader(hdr *Header) ([]byte, error) {
+	var (
+		p   []byte
+		n   int
+		err error
+	)
+
+	switch t := co.Conn.(type) {
+	case *net.TCPConn, *tls.Conn:
+		r := t.(io.Reader)
+
+		// First two bytes specify the length of the entire message.
+		l, err := tcpMsgLen(r)
+		if err != nil {
+			return nil, err
+		}
+		p = make([]byte, l)
+		n, err = tcpRead(r, p)
+		co.rtt = time.Since(co.t)
+	default:
+		if co.UDPSize > MinMsgSize {
+			p = make([]byte, co.UDPSize)
+		} else {
+			p = make([]byte, MinMsgSize)
+		}
+		n, err = co.Read(p)
+		co.rtt = time.Since(co.t)
+	}
+
+	if err != nil {
+		return nil, err
+	} else if n < headerSize {
+		return nil, ErrShortRead
+	}
+
+	p = p[:n]
+	if hdr != nil {
+		dh, _, err := unpackMsgHdr(p, 0)
+		if err != nil {
+			return nil, err
+		}
+		*hdr = dh
+	}
+	return p, err
+}
+
+// tcpMsgLen is a helper func to read first two bytes of stream as uint16 packet length.
+func tcpMsgLen(t io.Reader) (int, error) {
+	p := []byte{0, 0}
+	n, err := t.Read(p)
+	if err != nil {
+		return 0, err
+	}
+
+	// As seen with my local router/switch, returns 1 byte on the above read,
+	// resulting a a ShortRead. Just write it out (instead of loop) and read the
+	// other byte.
+	if n == 1 {
+		n1, err := t.Read(p[1:])
+		if err != nil {
+			return 0, err
+		}
+		n += n1
+	}
+
+	if n != 2 {
+		return 0, ErrShortRead
+	}
+	l := binary.BigEndian.Uint16(p)
+	if l == 0 {
+		return 0, ErrShortRead
+	}
+	return int(l), nil
+}
+
+// tcpRead calls TCPConn.Read enough times to fill allocated buffer.
+func tcpRead(t io.Reader, p []byte) (int, error) {
+	n, err := t.Read(p)
+	if err != nil {
+		return n, err
+	}
+	for n < len(p) {
+		j, err := t.Read(p[n:])
+		if err != nil {
+			return n, err
+		}
+		n += j
+	}
+	return n, err
+}
+
+// Read implements the net.Conn read method.
+func (co *Conn) Read(p []byte) (n int, err error) {
+	if co.Conn == nil {
+		return 0, ErrConnEmpty
+	}
+	if len(p) < 2 {
+		return 0, io.ErrShortBuffer
+	}
+	switch t := co.Conn.(type) {
+	case *net.TCPConn, *tls.Conn:
+		r := t.(io.Reader)
+
+		l, err := tcpMsgLen(r)
+		if err != nil {
+			return 0, err
+		}
+		if l > len(p) {
+			return int(l), io.ErrShortBuffer
+		}
+		return tcpRead(r, p[:l])
+	}
+	// UDP connection
+	n, err = co.Conn.Read(p)
+	if err != nil {
+		return n, err
+	}
+	return n, err
+}
+
+// WriteMsg sends a message through the connection co.
+// If the message m contains a TSIG record the transaction
+// signature is calculated.
+func (co *Conn) WriteMsg(m *Msg) (err error) {
+	var out []byte
+	if t := m.IsTsig(); t != nil {
+		mac := ""
+		if _, ok := co.TsigSecret[t.Hdr.Name]; !ok {
+			return ErrSecret
+		}
+		out, mac, err = TsigGenerate(m, co.TsigSecret[t.Hdr.Name], co.tsigRequestMAC, false)
+		// Set for the next read, although only used in zone transfers
+		co.tsigRequestMAC = mac
+	} else {
+		out, err = m.Pack()
+	}
+	if err != nil {
+		return err
+	}
+	co.t = time.Now()
+	if _, err = co.Write(out); err != nil {
+		return err
+	}
+	return nil
+}
+
+// Write implements the net.Conn Write method.
+func (co *Conn) Write(p []byte) (n int, err error) {
+	switch t := co.Conn.(type) {
+	case *net.TCPConn, *tls.Conn:
+		w := t.(io.Writer)
+
+		lp := len(p)
+		if lp < 2 {
+			return 0, io.ErrShortBuffer
+		}
+		if lp > MaxMsgSize {
+			return 0, &Error{err: "message too large"}
+		}
+		l := make([]byte, 2, lp+2)
+		binary.BigEndian.PutUint16(l, uint16(lp))
+		p = append(l, p...)
+		n, err := io.Copy(w, bytes.NewReader(p))
+		return int(n), err
+	}
+	n, err = co.Conn.Write(p)
+	return n, err
+}
+
+// Return the appropriate timeout for a specific request
+func (c *Client) getTimeoutForRequest(timeout time.Duration) time.Duration {
+	var requestTimeout time.Duration
+	if c.Timeout != 0 {
+		requestTimeout = c.Timeout
+	} else {
+		requestTimeout = timeout
+	}
+	// net.Dialer.Timeout has priority if smaller than the timeouts computed so
+	// far
+	if c.Dialer != nil && c.Dialer.Timeout != 0 {
+		if c.Dialer.Timeout < requestTimeout {
+			requestTimeout = c.Dialer.Timeout
+		}
+	}
+	return requestTimeout
+}
+
+// Dial connects to the address on the named network.
+func Dial(network, address string) (conn *Conn, err error) {
+	conn = new(Conn)
+	conn.Conn, err = net.Dial(network, address)
+	if err != nil {
+		return nil, err
+	}
+	return conn, nil
+}
+
+// ExchangeContext performs a synchronous UDP query, like Exchange. It
+// additionally obeys deadlines from the passed Context.
+func ExchangeContext(ctx context.Context, m *Msg, a string) (r *Msg, err error) {
+	client := Client{Net: "udp"}
+	r, _, err = client.ExchangeContext(ctx, m, a)
+	// ignorint rtt to leave the original ExchangeContext API unchanged, but
+	// this function will go away
+	return r, err
+}
+
+// ExchangeConn performs a synchronous query. It sends the message m via the connection
+// c and waits for a reply. The connection c is not closed by ExchangeConn.
+// This function is going away, but can easily be mimicked:
+//
+//	co := &dns.Conn{Conn: c} // c is your net.Conn
+//	co.WriteMsg(m)
+//	in, _  := co.ReadMsg()
+//	co.Close()
+//
+func ExchangeConn(c net.Conn, m *Msg) (r *Msg, err error) {
+	println("dns: ExchangeConn: this function is deprecated")
+	co := new(Conn)
+	co.Conn = c
+	if err = co.WriteMsg(m); err != nil {
+		return nil, err
+	}
+	r, err = co.ReadMsg()
+	if err == nil && r.Id != m.Id {
+		err = ErrId
+	}
+	return r, err
+}
+
+// DialTimeout acts like Dial but takes a timeout.
+func DialTimeout(network, address string, timeout time.Duration) (conn *Conn, err error) {
+	client := Client{Net: network, Dialer: &net.Dialer{Timeout: timeout}}
+	conn, err = client.Dial(address)
+	if err != nil {
+		return nil, err
+	}
+	return conn, nil
+}
+
+// DialWithTLS connects to the address on the named network with TLS.
+func DialWithTLS(network, address string, tlsConfig *tls.Config) (conn *Conn, err error) {
+	if !strings.HasSuffix(network, "-tls") {
+		network += "-tls"
+	}
+	client := Client{Net: network, TLSConfig: tlsConfig}
+	conn, err = client.Dial(address)
+
+	if err != nil {
+		return nil, err
+	}
+	return conn, nil
+}
+
+// DialTimeoutWithTLS acts like DialWithTLS but takes a timeout.
+func DialTimeoutWithTLS(network, address string, tlsConfig *tls.Config, timeout time.Duration) (conn *Conn, err error) {
+	if !strings.HasSuffix(network, "-tls") {
+		network += "-tls"
+	}
+	client := Client{Net: network, Dialer: &net.Dialer{Timeout: timeout}, TLSConfig: tlsConfig}
+	conn, err = client.Dial(address)
+	if err != nil {
+		return nil, err
+	}
+	return conn, nil
+}
+
+// ExchangeContext acts like Exchange, but honors the deadline on the provided
+// context, if present. If there is both a context deadline and a configured
+// timeout on the client, the earliest of the two takes effect.
+func (c *Client) ExchangeContext(ctx context.Context, m *Msg, a string) (r *Msg, rtt time.Duration, err error) {
+	var timeout time.Duration
+	if deadline, ok := ctx.Deadline(); !ok {
+		timeout = 0
+	} else {
+		timeout = deadline.Sub(time.Now())
+	}
+	// not passing the context to the underlying calls, as the API does not support
+	// context. For timeouts you should set up Client.Dialer and call Client.Exchange.
+	c.Dialer = &net.Dialer{Timeout: timeout}
+	return c.Exchange(m, a)
+}
--- a/vendor/github.com/miekg/dns/clientconfig.go
+++ b/vendor/github.com/miekg/dns/clientconfig.go
@ -0,0 +1,139 @@
+package dns
+
+import (
+	"bufio"
+	"io"
+	"os"
+	"strconv"
+	"strings"
+)
+
+// ClientConfig wraps the contents of the /etc/resolv.conf file.
+type ClientConfig struct {
+	Servers  []string // servers to use
+	Search   []string // suffixes to append to local name
+	Port     string   // what port to use
+	Ndots    int      // number of dots in name to trigger absolute lookup
+	Timeout  int      // seconds before giving up on packet
+	Attempts int      // lost packets before giving up on server, not used in the package dns
+}
+
+// ClientConfigFromFile parses a resolv.conf(5) like file and returns
+// a *ClientConfig.
+func ClientConfigFromFile(resolvconf string) (*ClientConfig, error) {
+	file, err := os.Open(resolvconf)
+	if err != nil {
+		return nil, err
+	}
+	defer file.Close()
+	return ClientConfigFromReader(file)
+}
+
+// ClientConfigFromReader works like ClientConfigFromFile but takes an io.Reader as argument
+func ClientConfigFromReader(resolvconf io.Reader) (*ClientConfig, error) {
+	c := new(ClientConfig)
+	scanner := bufio.NewScanner(resolvconf)
+	c.Servers = make([]string, 0)
+	c.Search = make([]string, 0)
+	c.Port = "53"
+	c.Ndots = 1
+	c.Timeout = 5
+	c.Attempts = 2
+
+	for scanner.Scan() {
+		if err := scanner.Err(); err != nil {
+			return nil, err
+		}
+		line := scanner.Text()
+		f := strings.Fields(line)
+		if len(f) < 1 {
+			continue
+		}
+		switch f[0] {
+		case "nameserver": // add one name server
+			if len(f) > 1 {
+				// One more check: make sure server name is
+				// just an IP address.  Otherwise we need DNS
+				// to look it up.
+				name := f[1]
+				c.Servers = append(c.Servers, name)
+			}
+
+		case "domain": // set search path to just this domain
+			if len(f) > 1 {
+				c.Search = make([]string, 1)
+				c.Search[0] = f[1]
+			} else {
+				c.Search = make([]string, 0)
+			}
+
+		case "search": // set search path to given servers
+			c.Search = make([]string, len(f)-1)
+			for i := 0; i < len(c.Search); i++ {
+				c.Search[i] = f[i+1]
+			}
+
+		case "options": // magic options
+			for i := 1; i < len(f); i++ {
+				s := f[i]
+				switch {
+				case len(s) >= 6 && s[:6] == "ndots:":
+					n, _ := strconv.Atoi(s[6:])
+					if n < 0 {
+						n = 0
+					} else if n > 15 {
+						n = 15
+					}
+					c.Ndots = n
+				case len(s) >= 8 && s[:8] == "timeout:":
+					n, _ := strconv.Atoi(s[8:])
+					if n < 1 {
+						n = 1
+					}
+					c.Timeout = n
+				case len(s) >= 8 && s[:9] == "attempts:":
+					n, _ := strconv.Atoi(s[9:])
+					if n < 1 {
+						n = 1
+					}
+					c.Attempts = n
+				case s == "rotate":
+					/* not imp */
+				}
+			}
+		}
+	}
+	return c, nil
+}
+
+// NameList returns all of the names that should be queried based on the
+// config. It is based off of go's net/dns name building, but it does not
+// check the length of the resulting names.
+func (c *ClientConfig) NameList(name string) []string {
+	// if this domain is already fully qualified, no append needed.
+	if IsFqdn(name) {
+		return []string{name}
+	}
+
+	// Check to see if the name has more labels than Ndots. Do this before making
+	// the domain fully qualified.
+	hasNdots := CountLabel(name) > c.Ndots
+	// Make the domain fully qualified.
+	name = Fqdn(name)
+
+	// Make a list of names based off search.
+	names := []string{}
+
+	// If name has enough dots, try that first.
+	if hasNdots {
+		names = append(names, name)
+	}
+	for _, s := range c.Search {
+		names = append(names, Fqdn(name+s))
+	}
+	// If we didn't have enough dots, try after suffixes.
+	if !hasNdots {
+		names = append(names, name)
+	}
+	return names
+}
--- a/vendor/github.com/miekg/dns/compress_generate.go
+++ b/vendor/github.com/miekg/dns/compress_generate.go
@ -0,0 +1,188 @@
+//+build ignore
+
+// compression_generate.go is meant to run with go generate. It will use
+// go/{importer,types} to track down all the RR struct types. Then for each type
+// it will look to see if there are (compressible) names, if so it will add that
+// type to compressionLenHelperType and comressionLenSearchType which "fake" the
+// compression so that Len() is fast.
+package main
+
+import (
+	"bytes"
+	"fmt"
+	"go/format"
+	"go/importer"
+	"go/types"
+	"log"
+	"os"
+)
+
+var packageHdr = `
+// Code generated by "go run compress_generate.go"; DO NOT EDIT.
+
+package dns
+
+`
+
+// getTypeStruct will take a type and the package scope, and return the
+// (innermost) struct if the type is considered a RR type (currently defined as
+// those structs beginning with a RR_Header, could be redefined as implementing
+// the RR interface). The bool return value indicates if embedded structs were
+// resolved.
+func getTypeStruct(t types.Type, scope *types.Scope) (*types.Struct, bool) {
+	st, ok := t.Underlying().(*types.Struct)
+	if !ok {
+		return nil, false
+	}
+	if st.Field(0).Type() == scope.Lookup("RR_Header").Type() {
+		return st, false
+	}
+	if st.Field(0).Anonymous() {
+		st, _ := getTypeStruct(st.Field(0).Type(), scope)
+		return st, true
+	}
+	return nil, false
+}
+
+func main() {
+	// Import and type-check the package
+	pkg, err := importer.Default().Import("github.com/miekg/dns")
+	fatalIfErr(err)
+	scope := pkg.Scope()
+
+	var domainTypes []string  // Types that have a domain name in them (either compressible or not).
+	var cdomainTypes []string // Types that have a compressible domain name in them (subset of domainType)
+Names:
+	for _, name := range scope.Names() {
+		o := scope.Lookup(name)
+		if o == nil || !o.Exported() {
+			continue
+		}
+		st, _ := getTypeStruct(o.Type(), scope)
+		if st == nil {
+			continue
+		}
+		if name == "PrivateRR" {
+			continue
+		}
+
+		if scope.Lookup("Type"+o.Name()) == nil && o.Name() != "RFC3597" {
+			log.Fatalf("Constant Type%s does not exist.", o.Name())
+		}
+
+		for i := 1; i < st.NumFields(); i++ {
+			if _, ok := st.Field(i).Type().(*types.Slice); ok {
+				if st.Tag(i) == `dns:"domain-name"` {
+					domainTypes = append(domainTypes, o.Name())
+					continue Names
+				}
+				if st.Tag(i) == `dns:"cdomain-name"` {
+					cdomainTypes = append(cdomainTypes, o.Name())
+					domainTypes = append(domainTypes, o.Name())
+					continue Names
+				}
+				continue
+			}
+
+			switch {
+			case st.Tag(i) == `dns:"domain-name"`:
+				domainTypes = append(domainTypes, o.Name())
+				continue Names
+			case st.Tag(i) == `dns:"cdomain-name"`:
+				cdomainTypes = append(cdomainTypes, o.Name())
+				domainTypes = append(domainTypes, o.Name())
+				continue Names
+			}
+		}
+	}
+
+	b := &bytes.Buffer{}
+	b.WriteString(packageHdr)
+
+	// compressionLenHelperType - all types that have domain-name/cdomain-name can be used for compressing names
+
+	fmt.Fprint(b, "func compressionLenHelperType(c map[string]int, r RR) {\n")
+	fmt.Fprint(b, "switch x := r.(type) {\n")
+	for _, name := range domainTypes {
+		o := scope.Lookup(name)
+		st, _ := getTypeStruct(o.Type(), scope)
+
+		fmt.Fprintf(b, "case *%s:\n", name)
+		for i := 1; i < st.NumFields(); i++ {
+			out := func(s string) { fmt.Fprintf(b, "compressionLenHelper(c, x.%s)\n", st.Field(i).Name()) }
+
+			if _, ok := st.Field(i).Type().(*types.Slice); ok {
+				switch st.Tag(i) {
+				case `dns:"domain-name"`:
+					fallthrough
+				case `dns:"cdomain-name"`:
+					// For HIP we need to slice over the elements in this slice.
+					fmt.Fprintf(b, `for i := range x.%s {
+						compressionLenHelper(c, x.%s[i])
+					}
+`, st.Field(i).Name(), st.Field(i).Name())
+				}
+				continue
+			}
+
+			switch {
+			case st.Tag(i) == `dns:"cdomain-name"`:
+				fallthrough
+			case st.Tag(i) == `dns:"domain-name"`:
+				out(st.Field(i).Name())
+			}
+		}
+	}
+	fmt.Fprintln(b, "}\n}\n\n")
+
+	// compressionLenSearchType - search cdomain-tags types for compressible names.
+
+	fmt.Fprint(b, "func compressionLenSearchType(c map[string]int, r RR) (int, bool) {\n")
+	fmt.Fprint(b, "switch x := r.(type) {\n")
+	for _, name := range cdomainTypes {
+		o := scope.Lookup(name)
+		st, _ := getTypeStruct(o.Type(), scope)
+
+		fmt.Fprintf(b, "case *%s:\n", name)
+		j := 1
+		for i := 1; i < st.NumFields(); i++ {
+			out := func(s string, j int) {
+				fmt.Fprintf(b, "k%d, ok%d := compressionLenSearch(c, x.%s)\n", j, j, st.Field(i).Name())
+			}
+
+			// There are no slice types with names that can be compressed.
+
+			switch {
+			case st.Tag(i) == `dns:"cdomain-name"`:
+				out(st.Field(i).Name(), j)
+				j++
+			}
+		}
+		k := "k1"
+		ok := "ok1"
+		for i := 2; i < j; i++ {
+			k += fmt.Sprintf(" + k%d", i)
+			ok += fmt.Sprintf(" && ok%d", i)
+		}
+		fmt.Fprintf(b, "return %s, %s\n", k, ok)
+	}
+	fmt.Fprintln(b, "}\nreturn 0, false\n}\n\n")
+
+	// gofmt
+	res, err := format.Source(b.Bytes())
+	if err != nil {
+		b.WriteTo(os.Stderr)
+		log.Fatal(err)
+	}
+
+	f, err := os.Create("zcompress.go")
+	fatalIfErr(err)
+	defer f.Close()
+	f.Write(res)
+}
+
+func fatalIfErr(err error) {
+	if err != nil {
+		log.Fatal(err)
+	}
+}
--- a/vendor/github.com/miekg/dns/dane.go
+++ b/vendor/github.com/miekg/dns/dane.go
@ -0,0 +1,43 @@
+package dns
+
+import (
+	"crypto/sha256"
+	"crypto/sha512"
+	"crypto/x509"
+	"encoding/hex"
+	"errors"
+)
+
+// CertificateToDANE converts a certificate to a hex string as used in the TLSA or SMIMEA records.
+func CertificateToDANE(selector, matchingType uint8, cert *x509.Certificate) (string, error) {
+	switch matchingType {
+	case 0:
+		switch selector {
+		case 0:
+			return hex.EncodeToString(cert.Raw), nil
+		case 1:
+			return hex.EncodeToString(cert.RawSubjectPublicKeyInfo), nil
+		}
+	case 1:
+		h := sha256.New()
+		switch selector {
+		case 0:
+			h.Write(cert.Raw)
+			return hex.EncodeToString(h.Sum(nil)), nil
+		case 1:
+			h.Write(cert.RawSubjectPublicKeyInfo)
+			return hex.EncodeToString(h.Sum(nil)), nil
+		}
+	case 2:
+		h := sha512.New()
+		switch selector {
+		case 0:
+			h.Write(cert.Raw)
+			return hex.EncodeToString(h.Sum(nil)), nil
+		case 1:
+			h.Write(cert.RawSubjectPublicKeyInfo)
+			return hex.EncodeToString(h.Sum(nil)), nil
+		}
+	}
+	return "", errors.New("dns: bad MatchingType or Selector")
+}
--- a/vendor/github.com/miekg/dns/defaults.go
+++ b/vendor/github.com/miekg/dns/defaults.go
@ -0,0 +1,288 @@
+package dns
+
+import (
+	"errors"
+	"net"
+	"strconv"
+)
+
+const hexDigit = "0123456789abcdef"
+
+// Everything is assumed in ClassINET.
+
+// SetReply creates a reply message from a request message.
+func (dns *Msg) SetReply(request *Msg) *Msg {
+	dns.Id = request.Id
+	dns.Response = true
+	dns.Opcode = request.Opcode
+	if dns.Opcode == OpcodeQuery {
+		dns.RecursionDesired = request.RecursionDesired // Copy rd bit
+		dns.CheckingDisabled = request.CheckingDisabled // Copy cd bit
+	}
+	dns.Rcode = RcodeSuccess
+	if len(request.Question) > 0 {
+		dns.Question = make([]Question, 1)
+		dns.Question[0] = request.Question[0]
+	}
+	return dns
+}
+
+// SetQuestion creates a question message, it sets the Question
+// section, generates an Id and sets the RecursionDesired (RD)
+// bit to true.
+func (dns *Msg) SetQuestion(z string, t uint16) *Msg {
+	dns.Id = Id()
+	dns.RecursionDesired = true
+	dns.Question = make([]Question, 1)
+	dns.Question[0] = Question{z, t, ClassINET}
+	return dns
+}
+
+// SetNotify creates a notify message, it sets the Question
+// section, generates an Id and sets the Authoritative (AA)
+// bit to true.
+func (dns *Msg) SetNotify(z string) *Msg {
+	dns.Opcode = OpcodeNotify
+	dns.Authoritative = true
+	dns.Id = Id()
+	dns.Question = make([]Question, 1)
+	dns.Question[0] = Question{z, TypeSOA, ClassINET}
+	return dns
+}
+
+// SetRcode creates an error message suitable for the request.
+func (dns *Msg) SetRcode(request *Msg, rcode int) *Msg {
+	dns.SetReply(request)
+	dns.Rcode = rcode
+	return dns
+}
+
+// SetRcodeFormatError creates a message with FormError set.
+func (dns *Msg) SetRcodeFormatError(request *Msg) *Msg {
+	dns.Rcode = RcodeFormatError
+	dns.Opcode = OpcodeQuery
+	dns.Response = true
+	dns.Authoritative = false
+	dns.Id = request.Id
+	return dns
+}
+
+// SetUpdate makes the message a dynamic update message. It
+// sets the ZONE section to: z, TypeSOA, ClassINET.
+func (dns *Msg) SetUpdate(z string) *Msg {
+	dns.Id = Id()
+	dns.Response = false
+	dns.Opcode = OpcodeUpdate
+	dns.Compress = false // BIND9 cannot handle compression
+	dns.Question = make([]Question, 1)
+	dns.Question[0] = Question{z, TypeSOA, ClassINET}
+	return dns
+}
+
+// SetIxfr creates message for requesting an IXFR.
+func (dns *Msg) SetIxfr(z string, serial uint32, ns, mbox string) *Msg {
+	dns.Id = Id()
+	dns.Question = make([]Question, 1)
+	dns.Ns = make([]RR, 1)
+	s := new(SOA)
+	s.Hdr = RR_Header{z, TypeSOA, ClassINET, defaultTtl, 0}
+	s.Serial = serial
+	s.Ns = ns
+	s.Mbox = mbox
+	dns.Question[0] = Question{z, TypeIXFR, ClassINET}
+	dns.Ns[0] = s
+	return dns
+}
+
+// SetAxfr creates message for requesting an AXFR.
+func (dns *Msg) SetAxfr(z string) *Msg {
+	dns.Id = Id()
+	dns.Question = make([]Question, 1)
+	dns.Question[0] = Question{z, TypeAXFR, ClassINET}
+	return dns
+}
+
+// SetTsig appends a TSIG RR to the message.
+// This is only a skeleton TSIG RR that is added as the last RR in the
+// additional section. The Tsig is calculated when the message is being send.
+func (dns *Msg) SetTsig(z, algo string, fudge uint16, timesigned int64) *Msg {
+	t := new(TSIG)
+	t.Hdr = RR_Header{z, TypeTSIG, ClassANY, 0, 0}
+	t.Algorithm = algo
+	t.Fudge = fudge
+	t.TimeSigned = uint64(timesigned)
+	t.OrigId = dns.Id
+	dns.Extra = append(dns.Extra, t)
+	return dns
+}
+
+// SetEdns0 appends a EDNS0 OPT RR to the message.
+// TSIG should always the last RR in a message.
+func (dns *Msg) SetEdns0(udpsize uint16, do bool) *Msg {
+	e := new(OPT)
+	e.Hdr.Name = "."
+	e.Hdr.Rrtype = TypeOPT
+	e.SetUDPSize(udpsize)
+	if do {
+		e.SetDo()
+	}
+	dns.Extra = append(dns.Extra, e)
+	return dns
+}
+
+// IsTsig checks if the message has a TSIG record as the last record
+// in the additional section. It returns the TSIG record found or nil.
+func (dns *Msg) IsTsig() *TSIG {
+	if len(dns.Extra) > 0 {
+		if dns.Extra[len(dns.Extra)-1].Header().Rrtype == TypeTSIG {
+			return dns.Extra[len(dns.Extra)-1].(*TSIG)
+		}
+	}
+	return nil
+}
+
+// IsEdns0 checks if the message has a EDNS0 (OPT) record, any EDNS0
+// record in the additional section will do. It returns the OPT record
+// found or nil.
+func (dns *Msg) IsEdns0() *OPT {
+	// EDNS0 is at the end of the additional section, start there.
+	// We might want to change this to *only* look at the last two
+	// records. So we see TSIG and/or OPT - this a slightly bigger
+	// change though.
+	for i := len(dns.Extra) - 1; i >= 0; i-- {
+		if dns.Extra[i].Header().Rrtype == TypeOPT {
+			return dns.Extra[i].(*OPT)
+		}
+	}
+	return nil
+}
+
+// IsDomainName checks if s is a valid domain name, it returns the number of
+// labels and true, when a domain name is valid.  Note that non fully qualified
+// domain name is considered valid, in this case the last label is counted in
+// the number of labels.  When false is returned the number of labels is not
+// defined.  Also note that this function is extremely liberal; almost any
+// string is a valid domain name as the DNS is 8 bit protocol. It checks if each
+// label fits in 63 characters, but there is no length check for the entire
+// string s. I.e.  a domain name longer than 255 characters is considered valid.
+func IsDomainName(s string) (labels int, ok bool) {
+	_, labels, err := packDomainName(s, nil, 0, nil, false)
+	return labels, err == nil
+}
+
+// IsSubDomain checks if child is indeed a child of the parent. If child and parent
+// are the same domain true is returned as well.
+func IsSubDomain(parent, child string) bool {
+	// Entire child is contained in parent
+	return CompareDomainName(parent, child) == CountLabel(parent)
+}
+
+// IsMsg sanity checks buf and returns an error if it isn't a valid DNS packet.
+// The checking is performed on the binary payload.
+func IsMsg(buf []byte) error {
+	// Header
+	if len(buf) < 12 {
+		return errors.New("dns: bad message header")
+	}
+	// Header: Opcode
+	// TODO(miek): more checks here, e.g. check all header bits.
+	return nil
+}
+
+// IsFqdn checks if a domain name is fully qualified.
+func IsFqdn(s string) bool {
+	l := len(s)
+	if l == 0 {
+		return false
+	}
+	return s[l-1] == '.'
+}
+
+// IsRRset checks if a set of RRs is a valid RRset as defined by RFC 2181.
+// This means the RRs need to have the same type, name, and class. Returns true
+// if the RR set is valid, otherwise false.
+func IsRRset(rrset []RR) bool {
+	if len(rrset) == 0 {
+		return false
+	}
+	if len(rrset) == 1 {
+		return true
+	}
+	rrHeader := rrset[0].Header()
+	rrType := rrHeader.Rrtype
+	rrClass := rrHeader.Class
+	rrName := rrHeader.Name
+
+	for _, rr := range rrset[1:] {
+		curRRHeader := rr.Header()
+		if curRRHeader.Rrtype != rrType || curRRHeader.Class != rrClass || curRRHeader.Name != rrName {
+			// Mismatch between the records, so this is not a valid rrset for
+			//signing/verifying
+			return false
+		}
+	}
+
+	return true
+}
+
+// Fqdn return the fully qualified domain name from s.
+// If s is already fully qualified, it behaves as the identity function.
+func Fqdn(s string) string {
+	if IsFqdn(s) {
+		return s
+	}
+	return s + "."
+}
+
+// Copied from the official Go code.
+
+// ReverseAddr returns the in-addr.arpa. or ip6.arpa. hostname of the IP
+// address suitable for reverse DNS (PTR) record lookups or an error if it fails
+// to parse the IP address.
+func ReverseAddr(addr string) (arpa string, err error) {
+	ip := net.ParseIP(addr)
+	if ip == nil {
+		return "", &Error{err: "unrecognized address: " + addr}
+	}
+	if ip.To4() != nil {
+		return strconv.Itoa(int(ip[15])) + "." + strconv.Itoa(int(ip[14])) + "." + strconv.Itoa(int(ip[13])) + "." +
+			strconv.Itoa(int(ip[12])) + ".in-addr.arpa.", nil
+	}
+	// Must be IPv6
+	buf := make([]byte, 0, len(ip)*4+len("ip6.arpa."))
+	// Add it, in reverse, to the buffer
+	for i := len(ip) - 1; i >= 0; i-- {
+		v := ip[i]
+		buf = append(buf, hexDigit[v&0xF])
+		buf = append(buf, '.')
+		buf = append(buf, hexDigit[v>>4])
+		buf = append(buf, '.')
+	}
+	// Append "ip6.arpa." and return (buf already has the final .)
+	buf = append(buf, "ip6.arpa."...)
+	return string(buf), nil
+}
+
+// String returns the string representation for the type t.
+func (t Type) String() string {
+	if t1, ok := TypeToString[uint16(t)]; ok {
+		return t1
+	}
+	return "TYPE" + strconv.Itoa(int(t))
+}
+
+// String returns the string representation for the class c.
+func (c Class) String() string {
+	if s, ok := ClassToString[uint16(c)]; ok {
+		// Only emit mnemonics when they are unambiguous, specically ANY is in both.
+		if _, ok := StringToType[s]; !ok {
+			return s
+		}
+	}
+	return "CLASS" + strconv.Itoa(int(c))
+}
+
+// String returns the string representation for the name n.
+func (n Name) String() string {
+	return sprintName(string(n))
+}
--- a/vendor/github.com/miekg/dns/dns.go
+++ b/vendor/github.com/miekg/dns/dns.go
@ -0,0 +1,107 @@
+package dns
+
+import "strconv"
+
+const (
+	year68     = 1 << 31 // For RFC1982 (Serial Arithmetic) calculations in 32 bits.
+	defaultTtl = 3600    // Default internal TTL.
+
+	// DefaultMsgSize is the standard default for messages larger than 512 bytes.
+	DefaultMsgSize = 4096
+	// MinMsgSize is the minimal size of a DNS packet.
+	MinMsgSize = 512
+	// MaxMsgSize is the largest possible DNS packet.
+	MaxMsgSize = 65535
+)
+
+// Error represents a DNS error.
+type Error struct{ err string }
+
+func (e *Error) Error() string {
+	if e == nil {
+		return "dns: <nil>"
+	}
+	return "dns: " + e.err
+}
+
+// An RR represents a resource record.
+type RR interface {
+	// Header returns the header of an resource record. The header contains
+	// everything up to the rdata.
+	Header() *RR_Header
+	// String returns the text representation of the resource record.
+	String() string
+
+	// copy returns a copy of the RR
+	copy() RR
+	// len returns the length (in octets) of the uncompressed RR in wire format.
+	len() int
+	// pack packs an RR into wire format.
+	pack([]byte, int, map[string]int, bool) (int, error)
+}
+
+// RR_Header is the header all DNS resource records share.
+type RR_Header struct {
+	Name     string `dns:"cdomain-name"`
+	Rrtype   uint16
+	Class    uint16
+	Ttl      uint32
+	Rdlength uint16 // Length of data after header.
+}
+
+// Header returns itself. This is here to make RR_Header implements the RR interface.
+func (h *RR_Header) Header() *RR_Header { return h }
+
+// Just to implement the RR interface.
+func (h *RR_Header) copy() RR { return nil }
+
+func (h *RR_Header) copyHeader() *RR_Header {
+	r := new(RR_Header)
+	r.Name = h.Name
+	r.Rrtype = h.Rrtype
+	r.Class = h.Class
+	r.Ttl = h.Ttl
+	r.Rdlength = h.Rdlength
+	return r
+}
+
+func (h *RR_Header) String() string {
+	var s string
+
+	if h.Rrtype == TypeOPT {
+		s = ";"
+		// and maybe other things
+	}
+
+	s += sprintName(h.Name) + "\t"
+	s += strconv.FormatInt(int64(h.Ttl), 10) + "\t"
+	s += Class(h.Class).String() + "\t"
+	s += Type(h.Rrtype).String() + "\t"
+	return s
+}
+
+func (h *RR_Header) len() int {
+	l := len(h.Name) + 1
+	l += 10 // rrtype(2) + class(2) + ttl(4) + rdlength(2)
+	return l
+}
+
+// ToRFC3597 converts a known RR to the unknown RR representation from RFC 3597.
+func (rr *RFC3597) ToRFC3597(r RR) error {
+	buf := make([]byte, r.len()*2)
+	off, err := PackRR(r, buf, 0, nil, false)
+	if err != nil {
+		return err
+	}
+	buf = buf[:off]
+	if int(r.Header().Rdlength) > off {
+		return ErrBuf
+	}
+
+	rfc3597, _, err := unpackRFC3597(*r.Header(), buf, off-int(r.Header().Rdlength))
+	if err != nil {
+		return err
+	}
+	*rr = *rfc3597.(*RFC3597)
+	return nil
+}
--- a/vendor/github.com/miekg/dns/dnssec.go
+++ b/vendor/github.com/miekg/dns/dnssec.go
@ -0,0 +1,784 @@
+package dns
+
+import (
+	"bytes"
+	"crypto"
+	"crypto/dsa"
+	"crypto/ecdsa"
+	"crypto/elliptic"
+	_ "crypto/md5"
+	"crypto/rand"
+	"crypto/rsa"
+	_ "crypto/sha1"
+	_ "crypto/sha256"
+	_ "crypto/sha512"
+	"encoding/asn1"
+	"encoding/binary"
+	"encoding/hex"
+	"math/big"
+	"sort"
+	"strings"
+	"time"
+
+	"golang.org/x/crypto/ed25519"
+)
+
+// DNSSEC encryption algorithm codes.
+const (
+	_ uint8 = iota
+	RSAMD5
+	DH
+	DSA
+	_ // Skip 4, RFC 6725, section 2.1
+	RSASHA1
+	DSANSEC3SHA1
+	RSASHA1NSEC3SHA1
+	RSASHA256
+	_ // Skip 9, RFC 6725, section 2.1
+	RSASHA512
+	_ // Skip 11, RFC 6725, section 2.1
+	ECCGOST
+	ECDSAP256SHA256
+	ECDSAP384SHA384
+	ED25519
+	ED448
+	INDIRECT   uint8 = 252
+	PRIVATEDNS uint8 = 253 // Private (experimental keys)
+	PRIVATEOID uint8 = 254
+)
+
+// AlgorithmToString is a map of algorithm IDs to algorithm names.
+var AlgorithmToString = map[uint8]string{
+	RSAMD5:           "RSAMD5",
+	DH:               "DH",
+	DSA:              "DSA",
+	RSASHA1:          "RSASHA1",
+	DSANSEC3SHA1:     "DSA-NSEC3-SHA1",
+	RSASHA1NSEC3SHA1: "RSASHA1-NSEC3-SHA1",
+	RSASHA256:        "RSASHA256",
+	RSASHA512:        "RSASHA512",
+	ECCGOST:          "ECC-GOST",
+	ECDSAP256SHA256:  "ECDSAP256SHA256",
+	ECDSAP384SHA384:  "ECDSAP384SHA384",
+	ED25519:          "ED25519",
+	ED448:            "ED448",
+	INDIRECT:         "INDIRECT",
+	PRIVATEDNS:       "PRIVATEDNS",
+	PRIVATEOID:       "PRIVATEOID",
+}
+
+// StringToAlgorithm is the reverse of AlgorithmToString.
+var StringToAlgorithm = reverseInt8(AlgorithmToString)
+
+// AlgorithmToHash is a map of algorithm crypto hash IDs to crypto.Hash's.
+var AlgorithmToHash = map[uint8]crypto.Hash{
+	RSAMD5:           crypto.MD5, // Deprecated in RFC 6725
+	RSASHA1:          crypto.SHA1,
+	RSASHA1NSEC3SHA1: crypto.SHA1,
+	RSASHA256:        crypto.SHA256,
+	ECDSAP256SHA256:  crypto.SHA256,
+	ECDSAP384SHA384:  crypto.SHA384,
+	RSASHA512:        crypto.SHA512,
+	ED25519:          crypto.Hash(0),
+}
+
+// DNSSEC hashing algorithm codes.
+const (
+	_      uint8 = iota
+	SHA1         // RFC 4034
+	SHA256       // RFC 4509
+	GOST94       // RFC 5933
+	SHA384       // Experimental
+	SHA512       // Experimental
+)
+
+// HashToString is a map of hash IDs to names.
+var HashToString = map[uint8]string{
+	SHA1:   "SHA1",
+	SHA256: "SHA256",
+	GOST94: "GOST94",
+	SHA384: "SHA384",
+	SHA512: "SHA512",
+}
+
+// StringToHash is a map of names to hash IDs.
+var StringToHash = reverseInt8(HashToString)
+
+// DNSKEY flag values.
+const (
+	SEP    = 1
+	REVOKE = 1 << 7
+	ZONE   = 1 << 8
+)
+
+// The RRSIG needs to be converted to wireformat with some of the rdata (the signature) missing.
+type rrsigWireFmt struct {
+	TypeCovered uint16
+	Algorithm   uint8
+	Labels      uint8
+	OrigTtl     uint32
+	Expiration  uint32
+	Inception   uint32
+	KeyTag      uint16
+	SignerName  string `dns:"domain-name"`
+	/* No Signature */
+}
+
+// Used for converting DNSKEY's rdata to wirefmt.
+type dnskeyWireFmt struct {
+	Flags     uint16
+	Protocol  uint8
+	Algorithm uint8
+	PublicKey string `dns:"base64"`
+	/* Nothing is left out */
+}
+
+func divRoundUp(a, b int) int {
+	return (a + b - 1) / b
+}
+
+// KeyTag calculates the keytag (or key-id) of the DNSKEY.
+func (k *DNSKEY) KeyTag() uint16 {
+	if k == nil {
+		return 0
+	}
+	var keytag int
+	switch k.Algorithm {
+	case RSAMD5:
+		// Look at the bottom two bytes of the modules, which the last
+		// item in the pubkey. We could do this faster by looking directly
+		// at the base64 values. But I'm lazy.
+		modulus, _ := fromBase64([]byte(k.PublicKey))
+		if len(modulus) > 1 {
+			x := binary.BigEndian.Uint16(modulus[len(modulus)-2:])
+			keytag = int(x)
+		}
+	default:
+		keywire := new(dnskeyWireFmt)
+		keywire.Flags = k.Flags
+		keywire.Protocol = k.Protocol
+		keywire.Algorithm = k.Algorithm
+		keywire.PublicKey = k.PublicKey
+		wire := make([]byte, DefaultMsgSize)
+		n, err := packKeyWire(keywire, wire)
+		if err != nil {
+			return 0
+		}
+		wire = wire[:n]
+		for i, v := range wire {
+			if i&1 != 0 {
+				keytag += int(v) // must be larger than uint32
+			} else {
+				keytag += int(v) << 8
+			}
+		}
+		keytag += (keytag >> 16) & 0xFFFF
+		keytag &= 0xFFFF
+	}
+	return uint16(keytag)
+}
+
+// ToDS converts a DNSKEY record to a DS record.
+func (k *DNSKEY) ToDS(h uint8) *DS {
+	if k == nil {
+		return nil
+	}
+	ds := new(DS)
+	ds.Hdr.Name = k.Hdr.Name
+	ds.Hdr.Class = k.Hdr.Class
+	ds.Hdr.Rrtype = TypeDS
+	ds.Hdr.Ttl = k.Hdr.Ttl
+	ds.Algorithm = k.Algorithm
+	ds.DigestType = h
+	ds.KeyTag = k.KeyTag()
+
+	keywire := new(dnskeyWireFmt)
+	keywire.Flags = k.Flags
+	keywire.Protocol = k.Protocol
+	keywire.Algorithm = k.Algorithm
+	keywire.PublicKey = k.PublicKey
+	wire := make([]byte, DefaultMsgSize)
+	n, err := packKeyWire(keywire, wire)
+	if err != nil {
+		return nil
+	}
+	wire = wire[:n]
+
+	owner := make([]byte, 255)
+	off, err1 := PackDomainName(strings.ToLower(k.Hdr.Name), owner, 0, nil, false)
+	if err1 != nil {
+		return nil
+	}
+	owner = owner[:off]
+	// RFC4034:
+	// digest = digest_algorithm( DNSKEY owner name | DNSKEY RDATA);
+	// "|" denotes concatenation
+	// DNSKEY RDATA = Flags | Protocol | Algorithm | Public Key.
+
+	var hash crypto.Hash
+	switch h {
+	case SHA1:
+		hash = crypto.SHA1
+	case SHA256:
+		hash = crypto.SHA256
+	case SHA384:
+		hash = crypto.SHA384
+	case SHA512:
+		hash = crypto.SHA512
+	default:
+		return nil
+	}
+
+	s := hash.New()
+	s.Write(owner)
+	s.Write(wire)
+	ds.Digest = hex.EncodeToString(s.Sum(nil))
+	return ds
+}
+
+// ToCDNSKEY converts a DNSKEY record to a CDNSKEY record.
+func (k *DNSKEY) ToCDNSKEY() *CDNSKEY {
+	c := &CDNSKEY{DNSKEY: *k}
+	c.Hdr = *k.Hdr.copyHeader()
+	c.Hdr.Rrtype = TypeCDNSKEY
+	return c
+}
+
+// ToCDS converts a DS record to a CDS record.
+func (d *DS) ToCDS() *CDS {
+	c := &CDS{DS: *d}
+	c.Hdr = *d.Hdr.copyHeader()
+	c.Hdr.Rrtype = TypeCDS
+	return c
+}
+
+// Sign signs an RRSet. The signature needs to be filled in with the values:
+// Inception, Expiration, KeyTag, SignerName and Algorithm.  The rest is copied
+// from the RRset. Sign returns a non-nill error when the signing went OK.
+// There is no check if RRSet is a proper (RFC 2181) RRSet.  If OrigTTL is non
+// zero, it is used as-is, otherwise the TTL of the RRset is used as the
+// OrigTTL.
+func (rr *RRSIG) Sign(k crypto.Signer, rrset []RR) error {
+	if k == nil {
+		return ErrPrivKey
+	}
+	// s.Inception and s.Expiration may be 0 (rollover etc.), the rest must be set
+	if rr.KeyTag == 0 || len(rr.SignerName) == 0 || rr.Algorithm == 0 {
+		return ErrKey
+	}
+
+	rr.Hdr.Rrtype = TypeRRSIG
+	rr.Hdr.Name = rrset[0].Header().Name
+	rr.Hdr.Class = rrset[0].Header().Class
+	if rr.OrigTtl == 0 { // If set don't override
+		rr.OrigTtl = rrset[0].Header().Ttl
+	}
+	rr.TypeCovered = rrset[0].Header().Rrtype
+	rr.Labels = uint8(CountLabel(rrset[0].Header().Name))
+
+	if strings.HasPrefix(rrset[0].Header().Name, "*") {
+		rr.Labels-- // wildcard, remove from label count
+	}
+
+	sigwire := new(rrsigWireFmt)
+	sigwire.TypeCovered = rr.TypeCovered
+	sigwire.Algorithm = rr.Algorithm
+	sigwire.Labels = rr.Labels
+	sigwire.OrigTtl = rr.OrigTtl
+	sigwire.Expiration = rr.Expiration
+	sigwire.Inception = rr.Inception
+	sigwire.KeyTag = rr.KeyTag
+	// For signing, lowercase this name
+	sigwire.SignerName = strings.ToLower(rr.SignerName)
+
+	// Create the desired binary blob
+	signdata := make([]byte, DefaultMsgSize)
+	n, err := packSigWire(sigwire, signdata)
+	if err != nil {
+		return err
+	}
+	signdata = signdata[:n]
+	wire, err := rawSignatureData(rrset, rr)
+	if err != nil {
+		return err
+	}
+
+	hash, ok := AlgorithmToHash[rr.Algorithm]
+	if !ok {
+		return ErrAlg
+	}
+
+	switch rr.Algorithm {
+	case ED25519:
+		// ed25519 signs the raw message and performs hashing internally.
+		// All other supported signature schemes operate over the pre-hashed
+		// message, and thus ed25519 must be handled separately here.
+		//
+		// The raw message is passed directly into sign and crypto.Hash(0) is
+		// used to signal to the crypto.Signer that the data has not been hashed.
+		signature, err := sign(k, append(signdata, wire...), crypto.Hash(0), rr.Algorithm)
+		if err != nil {
+			return err
+		}
+
+		rr.Signature = toBase64(signature)
+	default:
+		h := hash.New()
+		h.Write(signdata)
+		h.Write(wire)
+
+		signature, err := sign(k, h.Sum(nil), hash, rr.Algorithm)
+		if err != nil {
+			return err
+		}
+
+		rr.Signature = toBase64(signature)
+	}
+
+	return nil
+}
+
+func sign(k crypto.Signer, hashed []byte, hash crypto.Hash, alg uint8) ([]byte, error) {
+	signature, err := k.Sign(rand.Reader, hashed, hash)
+	if err != nil {
+		return nil, err
+	}
+
+	switch alg {
+	case RSASHA1, RSASHA1NSEC3SHA1, RSASHA256, RSASHA512:
+		return signature, nil
+
+	case ECDSAP256SHA256, ECDSAP384SHA384:
+		ecdsaSignature := &struct {
+			R, S *big.Int
+		}{}
+		if _, err := asn1.Unmarshal(signature, ecdsaSignature); err != nil {
+			return nil, err
+		}
+
+		var intlen int
+		switch alg {
+		case ECDSAP256SHA256:
+			intlen = 32
+		case ECDSAP384SHA384:
+			intlen = 48
+		}
+
+		signature := intToBytes(ecdsaSignature.R, intlen)
+		signature = append(signature, intToBytes(ecdsaSignature.S, intlen)...)
+		return signature, nil
+
+	// There is no defined interface for what a DSA backed crypto.Signer returns
+	case DSA, DSANSEC3SHA1:
+		// 	t := divRoundUp(divRoundUp(p.PublicKey.Y.BitLen(), 8)-64, 8)
+		// 	signature := []byte{byte(t)}
+		// 	signature = append(signature, intToBytes(r1, 20)...)
+		// 	signature = append(signature, intToBytes(s1, 20)...)
+		// 	rr.Signature = signature
+
+	case ED25519:
+		return signature, nil
+	}
+
+	return nil, ErrAlg
+}
+
+// Verify validates an RRSet with the signature and key. This is only the
+// cryptographic test, the signature validity period must be checked separately.
+// This function copies the rdata of some RRs (to lowercase domain names) for the validation to work.
+func (rr *RRSIG) Verify(k *DNSKEY, rrset []RR) error {
+	// First the easy checks
+	if !IsRRset(rrset) {
+		return ErrRRset
+	}
+	if rr.KeyTag != k.KeyTag() {
+		return ErrKey
+	}
+	if rr.Hdr.Class != k.Hdr.Class {
+		return ErrKey
+	}
+	if rr.Algorithm != k.Algorithm {
+		return ErrKey
+	}
+	if strings.ToLower(rr.SignerName) != strings.ToLower(k.Hdr.Name) {
+		return ErrKey
+	}
+	if k.Protocol != 3 {
+		return ErrKey
+	}
+
+	// IsRRset checked that we have at least one RR and that the RRs in
+	// the set have consistent type, class, and name. Also check that type and
+	// class matches the RRSIG record.
+	if rrset[0].Header().Class != rr.Hdr.Class {
+		return ErrRRset
+	}
+	if rrset[0].Header().Rrtype != rr.TypeCovered {
+		return ErrRRset
+	}
+
+	// RFC 4035 5.3.2.  Reconstructing the Signed Data
+	// Copy the sig, except the rrsig data
+	sigwire := new(rrsigWireFmt)
+	sigwire.TypeCovered = rr.TypeCovered
+	sigwire.Algorithm = rr.Algorithm
+	sigwire.Labels = rr.Labels
+	sigwire.OrigTtl = rr.OrigTtl
+	sigwire.Expiration = rr.Expiration
+	sigwire.Inception = rr.Inception
+	sigwire.KeyTag = rr.KeyTag
+	sigwire.SignerName = strings.ToLower(rr.SignerName)
+	// Create the desired binary blob
+	signeddata := make([]byte, DefaultMsgSize)
+	n, err := packSigWire(sigwire, signeddata)
+	if err != nil {
+		return err
+	}
+	signeddata = signeddata[:n]
+	wire, err := rawSignatureData(rrset, rr)
+	if err != nil {
+		return err
+	}
+
+	sigbuf := rr.sigBuf()           // Get the binary signature data
+	if rr.Algorithm == PRIVATEDNS { // PRIVATEOID
+		// TODO(miek)
+		// remove the domain name and assume its ours?
+	}
+
+	hash, ok := AlgorithmToHash[rr.Algorithm]
+	if !ok {
+		return ErrAlg
+	}
+
+	switch rr.Algorithm {
+	case RSASHA1, RSASHA1NSEC3SHA1, RSASHA256, RSASHA512, RSAMD5:
+		// TODO(mg): this can be done quicker, ie. cache the pubkey data somewhere??
+		pubkey := k.publicKeyRSA() // Get the key
+		if pubkey == nil {
+			return ErrKey
+		}
+
+		h := hash.New()
+		h.Write(signeddata)
+		h.Write(wire)
+		return rsa.VerifyPKCS1v15(pubkey, hash, h.Sum(nil), sigbuf)
+
+	case ECDSAP256SHA256, ECDSAP384SHA384:
+		pubkey := k.publicKeyECDSA()
+		if pubkey == nil {
+			return ErrKey
+		}
+
+		// Split sigbuf into the r and s coordinates
+		r := new(big.Int).SetBytes(sigbuf[:len(sigbuf)/2])
+		s := new(big.Int).SetBytes(sigbuf[len(sigbuf)/2:])
+
+		h := hash.New()
+		h.Write(signeddata)
+		h.Write(wire)
+		if ecdsa.Verify(pubkey, h.Sum(nil), r, s) {
+			return nil
+		}
+		return ErrSig
+
+	case ED25519:
+		pubkey := k.publicKeyED25519()
+		if pubkey == nil {
+			return ErrKey
+		}
+
+		if ed25519.Verify(pubkey, append(signeddata, wire...), sigbuf) {
+			return nil
+		}
+		return ErrSig
+
+	default:
+		return ErrAlg
+	}
+}
+
+// ValidityPeriod uses RFC1982 serial arithmetic to calculate
+// if a signature period is valid. If t is the zero time, the
+// current time is taken other t is. Returns true if the signature
+// is valid at the given time, otherwise returns false.
+func (rr *RRSIG) ValidityPeriod(t time.Time) bool {
+	var utc int64
+	if t.IsZero() {
+		utc = time.Now().UTC().Unix()
+	} else {
+		utc = t.UTC().Unix()
+	}
+	modi := (int64(rr.Inception) - utc) / year68
+	mode := (int64(rr.Expiration) - utc) / year68
+	ti := int64(rr.Inception) + (modi * year68)
+	te := int64(rr.Expiration) + (mode * year68)
+	return ti <= utc && utc <= te
+}
+
+// Return the signatures base64 encodedig sigdata as a byte slice.
+func (rr *RRSIG) sigBuf() []byte {
+	sigbuf, err := fromBase64([]byte(rr.Signature))
+	if err != nil {
+		return nil
+	}
+	return sigbuf
+}
+
+// publicKeyRSA returns the RSA public key from a DNSKEY record.
+func (k *DNSKEY) publicKeyRSA() *rsa.PublicKey {
+	keybuf, err := fromBase64([]byte(k.PublicKey))
+	if err != nil {
+		return nil
+	}
+
+	// RFC 2537/3110, section 2. RSA Public KEY Resource Records
+	// Length is in the 0th byte, unless its zero, then it
+	// it in bytes 1 and 2 and its a 16 bit number
+	explen := uint16(keybuf[0])
+	keyoff := 1
+	if explen == 0 {
+		explen = uint16(keybuf[1])<<8 | uint16(keybuf[2])
+		keyoff = 3
+	}
+	pubkey := new(rsa.PublicKey)
+
+	pubkey.N = big.NewInt(0)
+	shift := uint64((explen - 1) * 8)
+	expo := uint64(0)
+	for i := int(explen - 1); i > 0; i-- {
+		expo += uint64(keybuf[keyoff+i]) << shift
+		shift -= 8
+	}
+	// Remainder
+	expo += uint64(keybuf[keyoff])
+	if expo > (2<<31)+1 {
+		// Larger expo than supported.
+		// println("dns: F5 primes (or larger) are not supported")
+		return nil
+	}
+	pubkey.E = int(expo)
+
+	pubkey.N.SetBytes(keybuf[keyoff+int(explen):])
+	return pubkey
+}
+
+// publicKeyECDSA returns the Curve public key from the DNSKEY record.
+func (k *DNSKEY) publicKeyECDSA() *ecdsa.PublicKey {
+	keybuf, err := fromBase64([]byte(k.PublicKey))
+	if err != nil {
+		return nil
+	}
+	pubkey := new(ecdsa.PublicKey)
+	switch k.Algorithm {
+	case ECDSAP256SHA256:
+		pubkey.Curve = elliptic.P256()
+		if len(keybuf) != 64 {
+			// wrongly encoded key
+			return nil
+		}
+	case ECDSAP384SHA384:
+		pubkey.Curve = elliptic.P384()
+		if len(keybuf) != 96 {
+			// Wrongly encoded key
+			return nil
+		}
+	}
+	pubkey.X = big.NewInt(0)
+	pubkey.X.SetBytes(keybuf[:len(keybuf)/2])
+	pubkey.Y = big.NewInt(0)
+	pubkey.Y.SetBytes(keybuf[len(keybuf)/2:])
+	return pubkey
+}
+
+func (k *DNSKEY) publicKeyDSA() *dsa.PublicKey {
+	keybuf, err := fromBase64([]byte(k.PublicKey))
+	if err != nil {
+		return nil
+	}
+	if len(keybuf) < 22 {
+		return nil
+	}
+	t, keybuf := int(keybuf[0]), keybuf[1:]
+	size := 64 + t*8
+	q, keybuf := keybuf[:20], keybuf[20:]
+	if len(keybuf) != 3*size {
+		return nil
+	}
+	p, keybuf := keybuf[:size], keybuf[size:]
+	g, y := keybuf[:size], keybuf[size:]
+	pubkey := new(dsa.PublicKey)
+	pubkey.Parameters.Q = big.NewInt(0).SetBytes(q)
+	pubkey.Parameters.P = big.NewInt(0).SetBytes(p)
+	pubkey.Parameters.G = big.NewInt(0).SetBytes(g)
+	pubkey.Y = big.NewInt(0).SetBytes(y)
+	return pubkey
+}
+
+func (k *DNSKEY) publicKeyED25519() ed25519.PublicKey {
+	keybuf, err := fromBase64([]byte(k.PublicKey))
+	if err != nil {
+		return nil
+	}
+	if len(keybuf) != ed25519.PublicKeySize {
+		return nil
+	}
+	return keybuf
+}
+
+type wireSlice [][]byte
+
+func (p wireSlice) Len() int      { return len(p) }
+func (p wireSlice) Swap(i, j int) { p[i], p[j] = p[j], p[i] }
+func (p wireSlice) Less(i, j int) bool {
+	_, ioff, _ := UnpackDomainName(p[i], 0)
+	_, joff, _ := UnpackDomainName(p[j], 0)
+	return bytes.Compare(p[i][ioff+10:], p[j][joff+10:]) < 0
+}
+
+// Return the raw signature data.
+func rawSignatureData(rrset []RR, s *RRSIG) (buf []byte, err error) {
+	wires := make(wireSlice, len(rrset))
+	for i, r := range rrset {
+		r1 := r.copy()
+		r1.Header().Ttl = s.OrigTtl
+		labels := SplitDomainName(r1.Header().Name)
+		// 6.2. Canonical RR Form. (4) - wildcards
+		if len(labels) > int(s.Labels) {
+			// Wildcard
+			r1.Header().Name = "*." + strings.Join(labels[len(labels)-int(s.Labels):], ".") + "."
+		}
+		// RFC 4034: 6.2.  Canonical RR Form. (2) - domain name to lowercase
+		r1.Header().Name = strings.ToLower(r1.Header().Name)
+		// 6.2. Canonical RR Form. (3) - domain rdata to lowercase.
+		//   NS, MD, MF, CNAME, SOA, MB, MG, MR, PTR,
+		//   HINFO, MINFO, MX, RP, AFSDB, RT, SIG, PX, NXT, NAPTR, KX,
+		//   SRV, DNAME, A6
+		//
+		// RFC 6840 - Clarifications and Implementation Notes for DNS Security (DNSSEC):
+		//	Section 6.2 of [RFC4034] also erroneously lists HINFO as a record
+		//	that needs conversion to lowercase, and twice at that.  Since HINFO
+		//	records contain no domain names, they are not subject to case
+		//	conversion.
+		switch x := r1.(type) {
+		case *NS:
+			x.Ns = strings.ToLower(x.Ns)
+		case *MD:
+			x.Md = strings.ToLower(x.Md)
+		case *MF:
+			x.Mf = strings.ToLower(x.Mf)
+		case *CNAME:
+			x.Target = strings.ToLower(x.Target)
+		case *SOA:
+			x.Ns = strings.ToLower(x.Ns)
+			x.Mbox = strings.ToLower(x.Mbox)
+		case *MB:
+			x.Mb = strings.ToLower(x.Mb)
+		case *MG:
+			x.Mg = strings.ToLower(x.Mg)
+		case *MR:
+			x.Mr = strings.ToLower(x.Mr)
+		case *PTR:
+			x.Ptr = strings.ToLower(x.Ptr)
+		case *MINFO:
+			x.Rmail = strings.ToLower(x.Rmail)
+			x.Email = strings.ToLower(x.Email)
+		case *MX:
+			x.Mx = strings.ToLower(x.Mx)
+		case *RP:
+			x.Mbox = strings.ToLower(x.Mbox)
+			x.Txt = strings.ToLower(x.Txt)
+		case *AFSDB:
+			x.Hostname = strings.ToLower(x.Hostname)
+		case *RT:
+			x.Host = strings.ToLower(x.Host)
+		case *SIG:
+			x.SignerName = strings.ToLower(x.SignerName)
+		case *PX:
+			x.Map822 = strings.ToLower(x.Map822)
+			x.Mapx400 = strings.ToLower(x.Mapx400)
+		case *NAPTR:
+			x.Replacement = strings.ToLower(x.Replacement)
+		case *KX:
+			x.Exchanger = strings.ToLower(x.Exchanger)
+		case *SRV:
+			x.Target = strings.ToLower(x.Target)
+		case *DNAME:
+			x.Target = strings.ToLower(x.Target)
+		}
+		// 6.2. Canonical RR Form. (5) - origTTL
+		wire := make([]byte, r1.len()+1) // +1 to be safe(r)
+		off, err1 := PackRR(r1, wire, 0, nil, false)
+		if err1 != nil {
+			return nil, err1
+		}
+		wire = wire[:off]
+		wires[i] = wire
+	}
+	sort.Sort(wires)
+	for i, wire := range wires {
+		if i > 0 && bytes.Equal(wire, wires[i-1]) {
+			continue
+		}
+		buf = append(buf, wire...)
+	}
+	return buf, nil
+}
+
+func packSigWire(sw *rrsigWireFmt, msg []byte) (int, error) {
+	// copied from zmsg.go RRSIG packing
+	off, err := packUint16(sw.TypeCovered, msg, 0)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint8(sw.Algorithm, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint8(sw.Labels, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint32(sw.OrigTtl, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint32(sw.Expiration, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint32(sw.Inception, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint16(sw.KeyTag, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = PackDomainName(sw.SignerName, msg, off, nil, false)
+	if err != nil {
+		return off, err
+	}
+	return off, nil
+}
+
+func packKeyWire(dw *dnskeyWireFmt, msg []byte) (int, error) {
+	// copied from zmsg.go DNSKEY packing
+	off, err := packUint16(dw.Flags, msg, 0)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint8(dw.Protocol, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint8(dw.Algorithm, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packStringBase64(dw.PublicKey, msg, off)
+	if err != nil {
+		return off, err
+	}
+	return off, nil
+}
--- a/vendor/github.com/miekg/dns/dnssec_keygen.go
+++ b/vendor/github.com/miekg/dns/dnssec_keygen.go
@ -0,0 +1,178 @@
+package dns
+
+import (
+	"crypto"
+	"crypto/dsa"
+	"crypto/ecdsa"
+	"crypto/elliptic"
+	"crypto/rand"
+	"crypto/rsa"
+	"math/big"
+
+	"golang.org/x/crypto/ed25519"
+)
+
+// Generate generates a DNSKEY of the given bit size.
+// The public part is put inside the DNSKEY record.
+// The Algorithm in the key must be set as this will define
+// what kind of DNSKEY will be generated.
+// The ECDSA algorithms imply a fixed keysize, in that case
+// bits should be set to the size of the algorithm.
+func (k *DNSKEY) Generate(bits int) (crypto.PrivateKey, error) {
+	switch k.Algorithm {
+	case DSA, DSANSEC3SHA1:
+		if bits != 1024 {
+			return nil, ErrKeySize
+		}
+	case RSAMD5, RSASHA1, RSASHA256, RSASHA1NSEC3SHA1:
+		if bits < 512 || bits > 4096 {
+			return nil, ErrKeySize
+		}
+	case RSASHA512:
+		if bits < 1024 || bits > 4096 {
+			return nil, ErrKeySize
+		}
+	case ECDSAP256SHA256:
+		if bits != 256 {
+			return nil, ErrKeySize
+		}
+	case ECDSAP384SHA384:
+		if bits != 384 {
+			return nil, ErrKeySize
+		}
+	case ED25519:
+		if bits != 256 {
+			return nil, ErrKeySize
+		}
+	}
+
+	switch k.Algorithm {
+	case DSA, DSANSEC3SHA1:
+		params := new(dsa.Parameters)
+		if err := dsa.GenerateParameters(params, rand.Reader, dsa.L1024N160); err != nil {
+			return nil, err
+		}
+		priv := new(dsa.PrivateKey)
+		priv.PublicKey.Parameters = *params
+		err := dsa.GenerateKey(priv, rand.Reader)
+		if err != nil {
+			return nil, err
+		}
+		k.setPublicKeyDSA(params.Q, params.P, params.G, priv.PublicKey.Y)
+		return priv, nil
+	case RSAMD5, RSASHA1, RSASHA256, RSASHA512, RSASHA1NSEC3SHA1:
+		priv, err := rsa.GenerateKey(rand.Reader, bits)
+		if err != nil {
+			return nil, err
+		}
+		k.setPublicKeyRSA(priv.PublicKey.E, priv.PublicKey.N)
+		return priv, nil
+	case ECDSAP256SHA256, ECDSAP384SHA384:
+		var c elliptic.Curve
+		switch k.Algorithm {
+		case ECDSAP256SHA256:
+			c = elliptic.P256()
+		case ECDSAP384SHA384:
+			c = elliptic.P384()
+		}
+		priv, err := ecdsa.GenerateKey(c, rand.Reader)
+		if err != nil {
+			return nil, err
+		}
+		k.setPublicKeyECDSA(priv.PublicKey.X, priv.PublicKey.Y)
+		return priv, nil
+	case ED25519:
+		pub, priv, err := ed25519.GenerateKey(rand.Reader)
+		if err != nil {
+			return nil, err
+		}
+		k.setPublicKeyED25519(pub)
+		return priv, nil
+	default:
+		return nil, ErrAlg
+	}
+}
+
+// Set the public key (the value E and N)
+func (k *DNSKEY) setPublicKeyRSA(_E int, _N *big.Int) bool {
+	if _E == 0 || _N == nil {
+		return false
+	}
+	buf := exponentToBuf(_E)
+	buf = append(buf, _N.Bytes()...)
+	k.PublicKey = toBase64(buf)
+	return true
+}
+
+// Set the public key for Elliptic Curves
+func (k *DNSKEY) setPublicKeyECDSA(_X, _Y *big.Int) bool {
+	if _X == nil || _Y == nil {
+		return false
+	}
+	var intlen int
+	switch k.Algorithm {
+	case ECDSAP256SHA256:
+		intlen = 32
+	case ECDSAP384SHA384:
+		intlen = 48
+	}
+	k.PublicKey = toBase64(curveToBuf(_X, _Y, intlen))
+	return true
+}
+
+// Set the public key for DSA
+func (k *DNSKEY) setPublicKeyDSA(_Q, _P, _G, _Y *big.Int) bool {
+	if _Q == nil || _P == nil || _G == nil || _Y == nil {
+		return false
+	}
+	buf := dsaToBuf(_Q, _P, _G, _Y)
+	k.PublicKey = toBase64(buf)
+	return true
+}
+
+// Set the public key for Ed25519
+func (k *DNSKEY) setPublicKeyED25519(_K ed25519.PublicKey) bool {
+	if _K == nil {
+		return false
+	}
+	k.PublicKey = toBase64(_K)
+	return true
+}
+
+// Set the public key (the values E and N) for RSA
+// RFC 3110: Section 2. RSA Public KEY Resource Records
+func exponentToBuf(_E int) []byte {
+	var buf []byte
+	i := big.NewInt(int64(_E)).Bytes()
+	if len(i) < 256 {
+		buf = make([]byte, 1, 1+len(i))
+		buf[0] = uint8(len(i))
+	} else {
+		buf = make([]byte, 3, 3+len(i))
+		buf[0] = 0
+		buf[1] = uint8(len(i) >> 8)
+		buf[2] = uint8(len(i))
+	}
+	buf = append(buf, i...)
+	return buf
+}
+
+// Set the public key for X and Y for Curve. The two
+// values are just concatenated.
+func curveToBuf(_X, _Y *big.Int, intlen int) []byte {
+	buf := intToBytes(_X, intlen)
+	buf = append(buf, intToBytes(_Y, intlen)...)
+	return buf
+}
+
+// Set the public key for X and Y for Curve. The two
+// values are just concatenated.
+func dsaToBuf(_Q, _P, _G, _Y *big.Int) []byte {
+	t := divRoundUp(divRoundUp(_G.BitLen(), 8)-64, 8)
+	buf := []byte{byte(t)}
+	buf = append(buf, intToBytes(_Q, 20)...)
+	buf = append(buf, intToBytes(_P, 64+t*8)...)
+	buf = append(buf, intToBytes(_G, 64+t*8)...)
+	buf = append(buf, intToBytes(_Y, 64+t*8)...)
+	return buf
+}
--- a/vendor/github.com/miekg/dns/dnssec_keyscan.go
+++ b/vendor/github.com/miekg/dns/dnssec_keyscan.go
@ -0,0 +1,297 @@
+package dns
+
+import (
+	"bytes"
+	"crypto"
+	"crypto/dsa"
+	"crypto/ecdsa"
+	"crypto/rsa"
+	"io"
+	"math/big"
+	"strconv"
+	"strings"
+
+	"golang.org/x/crypto/ed25519"
+)
+
+// NewPrivateKey returns a PrivateKey by parsing the string s.
+// s should be in the same form of the BIND private key files.
+func (k *DNSKEY) NewPrivateKey(s string) (crypto.PrivateKey, error) {
+	if s == "" || s[len(s)-1] != '\n' { // We need a closing newline
+		return k.ReadPrivateKey(strings.NewReader(s+"\n"), "")
+	}
+	return k.ReadPrivateKey(strings.NewReader(s), "")
+}
+
+// ReadPrivateKey reads a private key from the io.Reader q. The string file is
+// only used in error reporting.
+// The public key must be known, because some cryptographic algorithms embed
+// the public inside the privatekey.
+func (k *DNSKEY) ReadPrivateKey(q io.Reader, file string) (crypto.PrivateKey, error) {
+	m, err := parseKey(q, file)
+	if m == nil {
+		return nil, err
+	}
+	if _, ok := m["private-key-format"]; !ok {
+		return nil, ErrPrivKey
+	}
+	if m["private-key-format"] != "v1.2" && m["private-key-format"] != "v1.3" {
+		return nil, ErrPrivKey
+	}
+	// TODO(mg): check if the pubkey matches the private key
+	algo, err := strconv.ParseUint(strings.SplitN(m["algorithm"], " ", 2)[0], 10, 8)
+	if err != nil {
+		return nil, ErrPrivKey
+	}
+	switch uint8(algo) {
+	case DSA:
+		priv, err := readPrivateKeyDSA(m)
+		if err != nil {
+			return nil, err
+		}
+		pub := k.publicKeyDSA()
+		if pub == nil {
+			return nil, ErrKey
+		}
+		priv.PublicKey = *pub
+		return priv, nil
+	case RSAMD5:
+		fallthrough
+	case RSASHA1:
+		fallthrough
+	case RSASHA1NSEC3SHA1:
+		fallthrough
+	case RSASHA256:
+		fallthrough
+	case RSASHA512:
+		priv, err := readPrivateKeyRSA(m)
+		if err != nil {
+			return nil, err
+		}
+		pub := k.publicKeyRSA()
+		if pub == nil {
+			return nil, ErrKey
+		}
+		priv.PublicKey = *pub
+		return priv, nil
+	case ECCGOST:
+		return nil, ErrPrivKey
+	case ECDSAP256SHA256:
+		fallthrough
+	case ECDSAP384SHA384:
+		priv, err := readPrivateKeyECDSA(m)
+		if err != nil {
+			return nil, err
+		}
+		pub := k.publicKeyECDSA()
+		if pub == nil {
+			return nil, ErrKey
+		}
+		priv.PublicKey = *pub
+		return priv, nil
+	case ED25519:
+		return readPrivateKeyED25519(m)
+	default:
+		return nil, ErrPrivKey
+	}
+}
+
+// Read a private key (file) string and create a public key. Return the private key.
+func readPrivateKeyRSA(m map[string]string) (*rsa.PrivateKey, error) {
+	p := new(rsa.PrivateKey)
+	p.Primes = []*big.Int{nil, nil}
+	for k, v := range m {
+		switch k {
+		case "modulus", "publicexponent", "privateexponent", "prime1", "prime2":
+			v1, err := fromBase64([]byte(v))
+			if err != nil {
+				return nil, err
+			}
+			switch k {
+			case "modulus":
+				p.PublicKey.N = big.NewInt(0)
+				p.PublicKey.N.SetBytes(v1)
+			case "publicexponent":
+				i := big.NewInt(0)
+				i.SetBytes(v1)
+				p.PublicKey.E = int(i.Int64()) // int64 should be large enough
+			case "privateexponent":
+				p.D = big.NewInt(0)
+				p.D.SetBytes(v1)
+			case "prime1":
+				p.Primes[0] = big.NewInt(0)
+				p.Primes[0].SetBytes(v1)
+			case "prime2":
+				p.Primes[1] = big.NewInt(0)
+				p.Primes[1].SetBytes(v1)
+			}
+		case "exponent1", "exponent2", "coefficient":
+			// not used in Go (yet)
+		case "created", "publish", "activate":
+			// not used in Go (yet)
+		}
+	}
+	return p, nil
+}
+
+func readPrivateKeyDSA(m map[string]string) (*dsa.PrivateKey, error) {
+	p := new(dsa.PrivateKey)
+	p.X = big.NewInt(0)
+	for k, v := range m {
+		switch k {
+		case "private_value(x)":
+			v1, err := fromBase64([]byte(v))
+			if err != nil {
+				return nil, err
+			}
+			p.X.SetBytes(v1)
+		case "created", "publish", "activate":
+			/* not used in Go (yet) */
+		}
+	}
+	return p, nil
+}
+
+func readPrivateKeyECDSA(m map[string]string) (*ecdsa.PrivateKey, error) {
+	p := new(ecdsa.PrivateKey)
+	p.D = big.NewInt(0)
+	// TODO: validate that the required flags are present
+	for k, v := range m {
+		switch k {
+		case "privatekey":
+			v1, err := fromBase64([]byte(v))
+			if err != nil {
+				return nil, err
+			}
+			p.D.SetBytes(v1)
+		case "created", "publish", "activate":
+			/* not used in Go (yet) */
+		}
+	}
+	return p, nil
+}
+
+func readPrivateKeyED25519(m map[string]string) (ed25519.PrivateKey, error) {
+	var p ed25519.PrivateKey
+	// TODO: validate that the required flags are present
+	for k, v := range m {
+		switch k {
+		case "privatekey":
+			p1, err := fromBase64([]byte(v))
+			if err != nil {
+				return nil, err
+			}
+			if len(p1) != 32 {
+				return nil, ErrPrivKey
+			}
+			// RFC 8080 and Golang's x/crypto/ed25519 differ as to how the
+			// private keys are represented. RFC 8080 specifies that private
+			// keys be stored solely as the seed value (p1 above) while the
+			// ed25519 package represents them as the seed value concatenated
+			// to the public key, which is derived from the seed value.
+			//
+			// ed25519.GenerateKey reads exactly 32 bytes from the passed in
+			// io.Reader and uses them as the seed. It also derives the
+			// public key and produces a compatible private key.
+			_, p, err = ed25519.GenerateKey(bytes.NewReader(p1))
+			if err != nil {
+				return nil, err
+			}
+		case "created", "publish", "activate":
+			/* not used in Go (yet) */
+		}
+	}
+	return p, nil
+}
+
+// parseKey reads a private key from r. It returns a map[string]string,
+// with the key-value pairs, or an error when the file is not correct.
+func parseKey(r io.Reader, file string) (map[string]string, error) {
+	s, cancel := scanInit(r)
+	m := make(map[string]string)
+	c := make(chan lex)
+	k := ""
+	defer func() {
+		cancel()
+		// zlexer can send up to two tokens, the next one and possibly 1 remainders.
+		// Do a non-blocking read.
+		_, ok := <-c
+		_, ok = <-c
+		if !ok {
+			// too bad
+		}
+	}()
+	// Start the lexer
+	go klexer(s, c)
+	for l := range c {
+		// It should alternate
+		switch l.value {
+		case zKey:
+			k = l.token
+		case zValue:
+			if k == "" {
+				return nil, &ParseError{file, "no private key seen", l}
+			}
+			//println("Setting", strings.ToLower(k), "to", l.token, "b")
+			m[strings.ToLower(k)] = l.token
+			k = ""
+		}
+	}
+	return m, nil
+}
+
+// klexer scans the sourcefile and returns tokens on the channel c.
+func klexer(s *scan, c chan lex) {
+	var l lex
+	str := "" // Hold the current read text
+	commt := false
+	key := true
+	x, err := s.tokenText()
+	defer close(c)
+	for err == nil {
+		l.column = s.position.Column
+		l.line = s.position.Line
+		switch x {
+		case ':':
+			if commt {
+				break
+			}
+			l.token = str
+			if key {
+				l.value = zKey
+				c <- l
+				// Next token is a space, eat it
+				s.tokenText()
+				key = false
+				str = ""
+			} else {
+				l.value = zValue
+			}
+		case ';':
+			commt = true
+		case '\n':
+			if commt {
+				// Reset a comment
+				commt = false
+			}
+			l.value = zValue
+			l.token = str
+			c <- l
+			str = ""
+			commt = false
+			key = true
+		default:
+			if commt {
+				break
+			}
+			str += string(x)
+		}
+		x, err = s.tokenText()
+	}
+	if len(str) > 0 {
+		// Send remainder
+		l.token = str
+		l.value = zValue
+		c <- l
+	}
+}
--- a/vendor/github.com/miekg/dns/dnssec_privkey.go
+++ b/vendor/github.com/miekg/dns/dnssec_privkey.go
@ -0,0 +1,93 @@
+package dns
+
+import (
+	"crypto"
+	"crypto/dsa"
+	"crypto/ecdsa"
+	"crypto/rsa"
+	"math/big"
+	"strconv"
+
+	"golang.org/x/crypto/ed25519"
+)
+
+const format = "Private-key-format: v1.3\n"
+
+// PrivateKeyString converts a PrivateKey to a string. This string has the same
+// format as the private-key-file of BIND9 (Private-key-format: v1.3).
+// It needs some info from the key (the algorithm), so its a method of the DNSKEY
+// It supports rsa.PrivateKey, ecdsa.PrivateKey and dsa.PrivateKey
+func (r *DNSKEY) PrivateKeyString(p crypto.PrivateKey) string {
+	algorithm := strconv.Itoa(int(r.Algorithm))
+	algorithm += " (" + AlgorithmToString[r.Algorithm] + ")"
+
+	switch p := p.(type) {
+	case *rsa.PrivateKey:
+		modulus := toBase64(p.PublicKey.N.Bytes())
+		e := big.NewInt(int64(p.PublicKey.E))
+		publicExponent := toBase64(e.Bytes())
+		privateExponent := toBase64(p.D.Bytes())
+		prime1 := toBase64(p.Primes[0].Bytes())
+		prime2 := toBase64(p.Primes[1].Bytes())
+		// Calculate Exponent1/2 and Coefficient as per: http://en.wikipedia.org/wiki/RSA#Using_the_Chinese_remainder_algorithm
+		// and from: http://code.google.com/p/go/issues/detail?id=987
+		one := big.NewInt(1)
+		p1 := big.NewInt(0).Sub(p.Primes[0], one)
+		q1 := big.NewInt(0).Sub(p.Primes[1], one)
+		exp1 := big.NewInt(0).Mod(p.D, p1)
+		exp2 := big.NewInt(0).Mod(p.D, q1)
+		coeff := big.NewInt(0).ModInverse(p.Primes[1], p.Primes[0])
+
+		exponent1 := toBase64(exp1.Bytes())
+		exponent2 := toBase64(exp2.Bytes())
+		coefficient := toBase64(coeff.Bytes())
+
+		return format +
+			"Algorithm: " + algorithm + "\n" +
+			"Modulus: " + modulus + "\n" +
+			"PublicExponent: " + publicExponent + "\n" +
+			"PrivateExponent: " + privateExponent + "\n" +
+			"Prime1: " + prime1 + "\n" +
+			"Prime2: " + prime2 + "\n" +
+			"Exponent1: " + exponent1 + "\n" +
+			"Exponent2: " + exponent2 + "\n" +
+			"Coefficient: " + coefficient + "\n"
+
+	case *ecdsa.PrivateKey:
+		var intlen int
+		switch r.Algorithm {
+		case ECDSAP256SHA256:
+			intlen = 32
+		case ECDSAP384SHA384:
+			intlen = 48
+		}
+		private := toBase64(intToBytes(p.D, intlen))
+		return format +
+			"Algorithm: " + algorithm + "\n" +
+			"PrivateKey: " + private + "\n"
+
+	case *dsa.PrivateKey:
+		T := divRoundUp(divRoundUp(p.PublicKey.Parameters.G.BitLen(), 8)-64, 8)
+		prime := toBase64(intToBytes(p.PublicKey.Parameters.P, 64+T*8))
+		subprime := toBase64(intToBytes(p.PublicKey.Parameters.Q, 20))
+		base := toBase64(intToBytes(p.PublicKey.Parameters.G, 64+T*8))
+		priv := toBase64(intToBytes(p.X, 20))
+		pub := toBase64(intToBytes(p.PublicKey.Y, 64+T*8))
+		return format +
+			"Algorithm: " + algorithm + "\n" +
+			"Prime(p): " + prime + "\n" +
+			"Subprime(q): " + subprime + "\n" +
+			"Base(g): " + base + "\n" +
+			"Private_value(x): " + priv + "\n" +
+			"Public_value(y): " + pub + "\n"
+
+	case ed25519.PrivateKey:
+		private := toBase64(p[:32])
+		return format +
+			"Algorithm: " + algorithm + "\n" +
+			"PrivateKey: " + private + "\n"
+
+	default:
+		return ""
+	}
+}
--- a/vendor/github.com/miekg/dns/doc.go
+++ b/vendor/github.com/miekg/dns/doc.go
@ -0,0 +1,272 @@
+/*
+Package dns implements a full featured interface to the Domain Name System.
+Server- and client-side programming is supported.
+The package allows complete control over what is sent out to the DNS. The package
+API follows the less-is-more principle, by presenting a small, clean interface.
+
+The package dns supports (asynchronous) querying/replying, incoming/outgoing zone transfers,
+TSIG, EDNS0, dynamic updates, notifies and DNSSEC validation/signing.
+Note that domain names MUST be fully qualified, before sending them, unqualified
+names in a message will result in a packing failure.
+
+Resource records are native types. They are not stored in wire format.
+Basic usage pattern for creating a new resource record:
+
+     r := new(dns.MX)
+     r.Hdr = dns.RR_Header{Name: "miek.nl.", Rrtype: dns.TypeMX,
+     Class: dns.ClassINET, Ttl: 3600}
+     r.Preference = 10
+     r.Mx = "mx.miek.nl."
+
+Or directly from a string:
+
+     mx, err := dns.NewRR("miek.nl. 3600 IN MX 10 mx.miek.nl.")
+
+Or when the default origin (.) and TTL (3600) and class (IN) suit you:
+
+     mx, err := dns.NewRR("miek.nl MX 10 mx.miek.nl")
+
+Or even:
+
+     mx, err := dns.NewRR("$ORIGIN nl.\nmiek 1H IN MX 10 mx.miek")
+
+In the DNS messages are exchanged, these messages contain resource
+records (sets). Use pattern for creating a message:
+
+     m := new(dns.Msg)
+     m.SetQuestion("miek.nl.", dns.TypeMX)
+
+Or when not certain if the domain name is fully qualified:
+
+	m.SetQuestion(dns.Fqdn("miek.nl"), dns.TypeMX)
+
+The message m is now a message with the question section set to ask
+the MX records for the miek.nl. zone.
+
+The following is slightly more verbose, but more flexible:
+
+     m1 := new(dns.Msg)
+     m1.Id = dns.Id()
+     m1.RecursionDesired = true
+     m1.Question = make([]dns.Question, 1)
+     m1.Question[0] = dns.Question{"miek.nl.", dns.TypeMX, dns.ClassINET}
+
+After creating a message it can be sent.
+Basic use pattern for synchronous querying the DNS at a
+server configured on 127.0.0.1 and port 53:
+
+     c := new(dns.Client)
+     in, rtt, err := c.Exchange(m1, "127.0.0.1:53")
+
+Suppressing multiple outstanding queries (with the same question, type and
+class) is as easy as setting:
+
+	c.SingleInflight = true
+
+More advanced options are available using a net.Dialer and the corresponding API.
+For example it is possible to set a timeout, or to specify a source IP address
+and port to use for the connection:
+
+	c := new(dns.Client)
+	laddr := net.UDPAddr{
+		IP: net.ParseIP("[::1]"),
+		Port: 12345,
+		Zone: "",
+	}
+	d := net.Dialer{
+		Timeout: 200 * time.Millisecond,
+		LocalAddr: &laddr,
+	}
+	in, rtt, err := c.ExchangeWithDialer(&d, m1, "8.8.8.8:53")
+
+If these "advanced" features are not needed, a simple UDP query can be sent,
+with:
+
+	in, err := dns.Exchange(m1, "127.0.0.1:53")
+
+When this functions returns you will get dns message. A dns message consists
+out of four sections.
+The question section: in.Question, the answer section: in.Answer,
+the authority section: in.Ns and the additional section: in.Extra.
+
+Each of these sections (except the Question section) contain a []RR. Basic
+use pattern for accessing the rdata of a TXT RR as the first RR in
+the Answer section:
+
+	if t, ok := in.Answer[0].(*dns.TXT); ok {
+		// do something with t.Txt
+	}
+
+Domain Name and TXT Character String Representations
+
+Both domain names and TXT character strings are converted to presentation
+form both when unpacked and when converted to strings.
+
+For TXT character strings, tabs, carriage returns and line feeds will be
+converted to \t, \r and \n respectively. Back slashes and quotations marks
+will be escaped. Bytes below 32 and above 127 will be converted to \DDD
+form.
+
+For domain names, in addition to the above rules brackets, periods,
+spaces, semicolons and the at symbol are escaped.
+
+DNSSEC
+
+DNSSEC (DNS Security Extension) adds a layer of security to the DNS. It
+uses public key cryptography to sign resource records. The
+public keys are stored in DNSKEY records and the signatures in RRSIG records.
+
+Requesting DNSSEC information for a zone is done by adding the DO (DNSSEC OK) bit
+to a request.
+
+     m := new(dns.Msg)
+     m.SetEdns0(4096, true)
+
+Signature generation, signature verification and key generation are all supported.
+
+DYNAMIC UPDATES
+
+Dynamic updates reuses the DNS message format, but renames three of
+the sections. Question is Zone, Answer is Prerequisite, Authority is
+Update, only the Additional is not renamed. See RFC 2136 for the gory details.
+
+You can set a rather complex set of rules for the existence of absence of
+certain resource records or names in a zone to specify if resource records
+should be added or removed. The table from RFC 2136 supplemented with the Go
+DNS function shows which functions exist to specify the prerequisites.
+
+ 3.2.4 - Table Of Metavalues Used In Prerequisite Section
+
+  CLASS    TYPE     RDATA    Meaning                    Function
+  --------------------------------------------------------------
+  ANY      ANY      empty    Name is in use             dns.NameUsed
+  ANY      rrset    empty    RRset exists (value indep) dns.RRsetUsed
+  NONE     ANY      empty    Name is not in use         dns.NameNotUsed
+  NONE     rrset    empty    RRset does not exist       dns.RRsetNotUsed
+  zone     rrset    rr       RRset exists (value dep)   dns.Used
+
+The prerequisite section can also be left empty.
+If you have decided on the prerequisites you can tell what RRs should
+be added or deleted. The next table shows the options you have and
+what functions to call.
+
+ 3.4.2.6 - Table Of Metavalues Used In Update Section
+
+  CLASS    TYPE     RDATA    Meaning                     Function
+  ---------------------------------------------------------------
+  ANY      ANY      empty    Delete all RRsets from name dns.RemoveName
+  ANY      rrset    empty    Delete an RRset             dns.RemoveRRset
+  NONE     rrset    rr       Delete an RR from RRset     dns.Remove
+  zone     rrset    rr       Add to an RRset             dns.Insert
+
+TRANSACTION SIGNATURE
+
+An TSIG or transaction signature adds a HMAC TSIG record to each message sent.
+The supported algorithms include: HmacMD5, HmacSHA1, HmacSHA256 and HmacSHA512.
+
+Basic use pattern when querying with a TSIG name "axfr." (note that these key names
+must be fully qualified - as they are domain names) and the base64 secret
+"so6ZGir4GPAqINNh9U5c3A==":
+
+If an incoming message contains a TSIG record it MUST be the last record in
+the additional section (RFC2845 3.2).  This means that you should make the
+call to SetTsig last, right before executing the query.  If you make any
+changes to the RRset after calling SetTsig() the signature will be incorrect.
+
+	c := new(dns.Client)
+	c.TsigSecret = map[string]string{"axfr.": "so6ZGir4GPAqINNh9U5c3A=="}
+	m := new(dns.Msg)
+	m.SetQuestion("miek.nl.", dns.TypeMX)
+	m.SetTsig("axfr.", dns.HmacMD5, 300, time.Now().Unix())
+	...
+	// When sending the TSIG RR is calculated and filled in before sending
+
+When requesting an zone transfer (almost all TSIG usage is when requesting zone transfers), with
+TSIG, this is the basic use pattern. In this example we request an AXFR for
+miek.nl. with TSIG key named "axfr." and secret "so6ZGir4GPAqINNh9U5c3A=="
+and using the server 176.58.119.54:
+
+	t := new(dns.Transfer)
+	m := new(dns.Msg)
+	t.TsigSecret = map[string]string{"axfr.": "so6ZGir4GPAqINNh9U5c3A=="}
+	m.SetAxfr("miek.nl.")
+	m.SetTsig("axfr.", dns.HmacMD5, 300, time.Now().Unix())
+	c, err := t.In(m, "176.58.119.54:53")
+	for r := range c { ... }
+
+You can now read the records from the transfer as they come in. Each envelope is checked with TSIG.
+If something is not correct an error is returned.
+
+Basic use pattern validating and replying to a message that has TSIG set.
+
+	server := &dns.Server{Addr: ":53", Net: "udp"}
+	server.TsigSecret = map[string]string{"axfr.": "so6ZGir4GPAqINNh9U5c3A=="}
+	go server.ListenAndServe()
+	dns.HandleFunc(".", handleRequest)
+
+	func handleRequest(w dns.ResponseWriter, r *dns.Msg) {
+		m := new(dns.Msg)
+		m.SetReply(r)
+		if r.IsTsig() != nil {
+			if w.TsigStatus() == nil {
+				// *Msg r has an TSIG record and it was validated
+				m.SetTsig("axfr.", dns.HmacMD5, 300, time.Now().Unix())
+			} else {
+				// *Msg r has an TSIG records and it was not valided
+			}
+		}
+		w.WriteMsg(m)
+	}
+
+PRIVATE RRS
+
+RFC 6895 sets aside a range of type codes for private use. This range
+is 65,280 - 65,534 (0xFF00 - 0xFFFE). When experimenting with new Resource Records these
+can be used, before requesting an official type code from IANA.
+
+see http://miek.nl/2014/September/21/idn-and-private-rr-in-go-dns/ for more
+information.
+
+EDNS0
+
+EDNS0 is an extension mechanism for the DNS defined in RFC 2671 and updated
+by RFC 6891. It defines an new RR type, the OPT RR, which is then completely
+abused.
+Basic use pattern for creating an (empty) OPT RR:
+
+	o := new(dns.OPT)
+	o.Hdr.Name = "." // MUST be the root zone, per definition.
+	o.Hdr.Rrtype = dns.TypeOPT
+
+The rdata of an OPT RR consists out of a slice of EDNS0 (RFC 6891)
+interfaces. Currently only a few have been standardized: EDNS0_NSID
+(RFC 5001) and EDNS0_SUBNET (draft-vandergaast-edns-client-subnet-02). Note
+that these options may be combined in an OPT RR.
+Basic use pattern for a server to check if (and which) options are set:
+
+	// o is a dns.OPT
+	for _, s := range o.Option {
+		switch e := s.(type) {
+		case *dns.EDNS0_NSID:
+			// do stuff with e.Nsid
+		case *dns.EDNS0_SUBNET:
+			// access e.Family, e.Address, etc.
+		}
+	}
+
+SIG(0)
+
+From RFC 2931:
+
+    SIG(0) provides protection for DNS transactions and requests ....
+    ... protection for glue records, DNS requests, protection for message headers
+    on requests and responses, and protection of the overall integrity of a response.
+
+It works like TSIG, except that SIG(0) uses public key cryptography, instead of the shared
+secret approach in TSIG.
+Supported algorithms: DSA, ECDSAP256SHA256, ECDSAP384SHA384, RSASHA1, RSASHA256 and
+RSASHA512.
+
+Signing subsequent messages in multi-message sessions is not implemented.
+*/
+package dns
--- a/vendor/github.com/miekg/dns/edns.go
+++ b/vendor/github.com/miekg/dns/edns.go
@ -0,0 +1,627 @@
+package dns
+
+import (
+	"encoding/binary"
+	"encoding/hex"
+	"errors"
+	"fmt"
+	"net"
+	"strconv"
+)
+
+// EDNS0 Option codes.
+const (
+	EDNS0LLQ          = 0x1     // long lived queries: http://tools.ietf.org/html/draft-sekar-dns-llq-01
+	EDNS0UL           = 0x2     // update lease draft: http://files.dns-sd.org/draft-sekar-dns-ul.txt
+	EDNS0NSID         = 0x3     // nsid (See RFC 5001)
+	EDNS0DAU          = 0x5     // DNSSEC Algorithm Understood
+	EDNS0DHU          = 0x6     // DS Hash Understood
+	EDNS0N3U          = 0x7     // NSEC3 Hash Understood
+	EDNS0SUBNET       = 0x8     // client-subnet (See RFC 7871)
+	EDNS0EXPIRE       = 0x9     // EDNS0 expire
+	EDNS0COOKIE       = 0xa     // EDNS0 Cookie
+	EDNS0TCPKEEPALIVE = 0xb     // EDNS0 tcp keep alive (See RFC 7828)
+	EDNS0PADDING      = 0xc     // EDNS0 padding (See RFC 7830)
+	EDNS0LOCALSTART   = 0xFDE9  // Beginning of range reserved for local/experimental use (See RFC 6891)
+	EDNS0LOCALEND     = 0xFFFE  // End of range reserved for local/experimental use (See RFC 6891)
+	_DO               = 1 << 15 // DNSSEC OK
+)
+
+// OPT is the EDNS0 RR appended to messages to convey extra (meta) information.
+// See RFC 6891.
+type OPT struct {
+	Hdr    RR_Header
+	Option []EDNS0 `dns:"opt"`
+}
+
+func (rr *OPT) String() string {
+	s := "\n;; OPT PSEUDOSECTION:\n; EDNS: version " + strconv.Itoa(int(rr.Version())) + "; "
+	if rr.Do() {
+		s += "flags: do; "
+	} else {
+		s += "flags: ; "
+	}
+	s += "udp: " + strconv.Itoa(int(rr.UDPSize()))
+
+	for _, o := range rr.Option {
+		switch o.(type) {
+		case *EDNS0_NSID:
+			s += "\n; NSID: " + o.String()
+			h, e := o.pack()
+			var r string
+			if e == nil {
+				for _, c := range h {
+					r += "(" + string(c) + ")"
+				}
+				s += "  " + r
+			}
+		case *EDNS0_SUBNET:
+			s += "\n; SUBNET: " + o.String()
+		case *EDNS0_COOKIE:
+			s += "\n; COOKIE: " + o.String()
+		case *EDNS0_UL:
+			s += "\n; UPDATE LEASE: " + o.String()
+		case *EDNS0_LLQ:
+			s += "\n; LONG LIVED QUERIES: " + o.String()
+		case *EDNS0_DAU:
+			s += "\n; DNSSEC ALGORITHM UNDERSTOOD: " + o.String()
+		case *EDNS0_DHU:
+			s += "\n; DS HASH UNDERSTOOD: " + o.String()
+		case *EDNS0_N3U:
+			s += "\n; NSEC3 HASH UNDERSTOOD: " + o.String()
+		case *EDNS0_LOCAL:
+			s += "\n; LOCAL OPT: " + o.String()
+		case *EDNS0_PADDING:
+			s += "\n; PADDING: " + o.String()
+		}
+	}
+	return s
+}
+
+func (rr *OPT) len() int {
+	l := rr.Hdr.len()
+	for i := 0; i < len(rr.Option); i++ {
+		l += 4 // Account for 2-byte option code and 2-byte option length.
+		lo, _ := rr.Option[i].pack()
+		l += len(lo)
+	}
+	return l
+}
+
+// return the old value -> delete SetVersion?
+
+// Version returns the EDNS version used. Only zero is defined.
+func (rr *OPT) Version() uint8 {
+	return uint8((rr.Hdr.Ttl & 0x00FF0000) >> 16)
+}
+
+// SetVersion sets the version of EDNS. This is usually zero.
+func (rr *OPT) SetVersion(v uint8) {
+	rr.Hdr.Ttl = rr.Hdr.Ttl&0xFF00FFFF | (uint32(v) << 16)
+}
+
+// ExtendedRcode returns the EDNS extended RCODE field (the upper 8 bits of the TTL).
+func (rr *OPT) ExtendedRcode() int {
+	return int((rr.Hdr.Ttl & 0xFF000000) >> 24)
+}
+
+// SetExtendedRcode sets the EDNS extended RCODE field.
+func (rr *OPT) SetExtendedRcode(v uint8) {
+	rr.Hdr.Ttl = rr.Hdr.Ttl&0x00FFFFFF | (uint32(v) << 24)
+}
+
+// UDPSize returns the UDP buffer size.
+func (rr *OPT) UDPSize() uint16 {
+	return rr.Hdr.Class
+}
+
+// SetUDPSize sets the UDP buffer size.
+func (rr *OPT) SetUDPSize(size uint16) {
+	rr.Hdr.Class = size
+}
+
+// Do returns the value of the DO (DNSSEC OK) bit.
+func (rr *OPT) Do() bool {
+	return rr.Hdr.Ttl&_DO == _DO
+}
+
+// SetDo sets the DO (DNSSEC OK) bit.
+// If we pass an argument, set the DO bit to that value.
+// It is possible to pass 2 or more arguments. Any arguments after the 1st is silently ignored.
+func (rr *OPT) SetDo(do ...bool) {
+	if len(do) == 1 {
+		if do[0] {
+			rr.Hdr.Ttl |= _DO
+		} else {
+			rr.Hdr.Ttl &^= _DO
+		}
+	} else {
+		rr.Hdr.Ttl |= _DO
+	}
+}
+
+// EDNS0 defines an EDNS0 Option. An OPT RR can have multiple options appended to it.
+type EDNS0 interface {
+	// Option returns the option code for the option.
+	Option() uint16
+	// pack returns the bytes of the option data.
+	pack() ([]byte, error)
+	// unpack sets the data as found in the buffer. Is also sets
+	// the length of the slice as the length of the option data.
+	unpack([]byte) error
+	// String returns the string representation of the option.
+	String() string
+}
+
+// EDNS0_NSID option is used to retrieve a nameserver
+// identifier. When sending a request Nsid must be set to the empty string
+// The identifier is an opaque string encoded as hex.
+// Basic use pattern for creating an nsid option:
+//
+//	o := new(dns.OPT)
+//	o.Hdr.Name = "."
+//	o.Hdr.Rrtype = dns.TypeOPT
+//	e := new(dns.EDNS0_NSID)
+//	e.Code = dns.EDNS0NSID
+//	e.Nsid = "AA"
+//	o.Option = append(o.Option, e)
+type EDNS0_NSID struct {
+	Code uint16 // Always EDNS0NSID
+	Nsid string // This string needs to be hex encoded
+}
+
+func (e *EDNS0_NSID) pack() ([]byte, error) {
+	h, err := hex.DecodeString(e.Nsid)
+	if err != nil {
+		return nil, err
+	}
+	return h, nil
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_NSID) Option() uint16        { return EDNS0NSID } // Option returns the option code.
+func (e *EDNS0_NSID) unpack(b []byte) error { e.Nsid = hex.EncodeToString(b); return nil }
+func (e *EDNS0_NSID) String() string        { return string(e.Nsid) }
+
+// EDNS0_SUBNET is the subnet option that is used to give the remote nameserver
+// an idea of where the client lives. See RFC 7871. It can then give back a different
+// answer depending on the location or network topology.
+// Basic use pattern for creating an subnet option:
+//
+//	o := new(dns.OPT)
+//	o.Hdr.Name = "."
+//	o.Hdr.Rrtype = dns.TypeOPT
+//	e := new(dns.EDNS0_SUBNET)
+//	e.Code = dns.EDNS0SUBNET
+//	e.Family = 1	// 1 for IPv4 source address, 2 for IPv6
+//	e.SourceNetmask = 32	// 32 for IPV4, 128 for IPv6
+//	e.SourceScope = 0
+//	e.Address = net.ParseIP("127.0.0.1").To4()	// for IPv4
+//	// e.Address = net.ParseIP("2001:7b8:32a::2")	// for IPV6
+//	o.Option = append(o.Option, e)
+//
+// This code will parse all the available bits when unpacking (up to optlen).
+// When packing it will apply SourceNetmask. If you need more advanced logic,
+// patches welcome and good luck.
+type EDNS0_SUBNET struct {
+	Code          uint16 // Always EDNS0SUBNET
+	Family        uint16 // 1 for IP, 2 for IP6
+	SourceNetmask uint8
+	SourceScope   uint8
+	Address       net.IP
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_SUBNET) Option() uint16 { return EDNS0SUBNET }
+
+func (e *EDNS0_SUBNET) pack() ([]byte, error) {
+	b := make([]byte, 4)
+	binary.BigEndian.PutUint16(b[0:], e.Family)
+	b[2] = e.SourceNetmask
+	b[3] = e.SourceScope
+	switch e.Family {
+	case 0:
+		// "dig" sets AddressFamily to 0 if SourceNetmask is also 0
+		// We might don't need to complain either
+		if e.SourceNetmask != 0 {
+			return nil, errors.New("dns: bad address family")
+		}
+	case 1:
+		if e.SourceNetmask > net.IPv4len*8 {
+			return nil, errors.New("dns: bad netmask")
+		}
+		if len(e.Address.To4()) != net.IPv4len {
+			return nil, errors.New("dns: bad address")
+		}
+		ip := e.Address.To4().Mask(net.CIDRMask(int(e.SourceNetmask), net.IPv4len*8))
+		needLength := (e.SourceNetmask + 8 - 1) / 8 // division rounding up
+		b = append(b, ip[:needLength]...)
+	case 2:
+		if e.SourceNetmask > net.IPv6len*8 {
+			return nil, errors.New("dns: bad netmask")
+		}
+		if len(e.Address) != net.IPv6len {
+			return nil, errors.New("dns: bad address")
+		}
+		ip := e.Address.Mask(net.CIDRMask(int(e.SourceNetmask), net.IPv6len*8))
+		needLength := (e.SourceNetmask + 8 - 1) / 8 // division rounding up
+		b = append(b, ip[:needLength]...)
+	default:
+		return nil, errors.New("dns: bad address family")
+	}
+	return b, nil
+}
+
+func (e *EDNS0_SUBNET) unpack(b []byte) error {
+	if len(b) < 4 {
+		return ErrBuf
+	}
+	e.Family = binary.BigEndian.Uint16(b)
+	e.SourceNetmask = b[2]
+	e.SourceScope = b[3]
+	switch e.Family {
+	case 0:
+		// "dig" sets AddressFamily to 0 if SourceNetmask is also 0
+		// It's okay to accept such a packet
+		if e.SourceNetmask != 0 {
+			return errors.New("dns: bad address family")
+		}
+		e.Address = net.IPv4(0, 0, 0, 0)
+	case 1:
+		if e.SourceNetmask > net.IPv4len*8 || e.SourceScope > net.IPv4len*8 {
+			return errors.New("dns: bad netmask")
+		}
+		addr := make([]byte, net.IPv4len)
+		for i := 0; i < net.IPv4len && 4+i < len(b); i++ {
+			addr[i] = b[4+i]
+		}
+		e.Address = net.IPv4(addr[0], addr[1], addr[2], addr[3])
+	case 2:
+		if e.SourceNetmask > net.IPv6len*8 || e.SourceScope > net.IPv6len*8 {
+			return errors.New("dns: bad netmask")
+		}
+		addr := make([]byte, net.IPv6len)
+		for i := 0; i < net.IPv6len && 4+i < len(b); i++ {
+			addr[i] = b[4+i]
+		}
+		e.Address = net.IP{addr[0], addr[1], addr[2], addr[3], addr[4],
+			addr[5], addr[6], addr[7], addr[8], addr[9], addr[10],
+			addr[11], addr[12], addr[13], addr[14], addr[15]}
+	default:
+		return errors.New("dns: bad address family")
+	}
+	return nil
+}
+
+func (e *EDNS0_SUBNET) String() (s string) {
+	if e.Address == nil {
+		s = "<nil>"
+	} else if e.Address.To4() != nil {
+		s = e.Address.String()
+	} else {
+		s = "[" + e.Address.String() + "]"
+	}
+	s += "/" + strconv.Itoa(int(e.SourceNetmask)) + "/" + strconv.Itoa(int(e.SourceScope))
+	return
+}
+
+// The EDNS0_COOKIE option is used to add a DNS Cookie to a message.
+//
+//	o := new(dns.OPT)
+//	o.Hdr.Name = "."
+//	o.Hdr.Rrtype = dns.TypeOPT
+//	e := new(dns.EDNS0_COOKIE)
+//	e.Code = dns.EDNS0COOKIE
+//	e.Cookie = "24a5ac.."
+//	o.Option = append(o.Option, e)
+//
+// The Cookie field consists out of a client cookie (RFC 7873 Section 4), that is
+// always 8 bytes. It may then optionally be followed by the server cookie. The server
+// cookie is of variable length, 8 to a maximum of 32 bytes. In other words:
+//
+//	cCookie := o.Cookie[:16]
+//	sCookie := o.Cookie[16:]
+//
+// There is no guarantee that the Cookie string has a specific length.
+type EDNS0_COOKIE struct {
+	Code   uint16 // Always EDNS0COOKIE
+	Cookie string // Hex-encoded cookie data
+}
+
+func (e *EDNS0_COOKIE) pack() ([]byte, error) {
+	h, err := hex.DecodeString(e.Cookie)
+	if err != nil {
+		return nil, err
+	}
+	return h, nil
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_COOKIE) Option() uint16        { return EDNS0COOKIE }
+func (e *EDNS0_COOKIE) unpack(b []byte) error { e.Cookie = hex.EncodeToString(b); return nil }
+func (e *EDNS0_COOKIE) String() string        { return e.Cookie }
+
+// The EDNS0_UL (Update Lease) (draft RFC) option is used to tell the server to set
+// an expiration on an update RR. This is helpful for clients that cannot clean
+// up after themselves. This is a draft RFC and more information can be found at
+// http://files.dns-sd.org/draft-sekar-dns-ul.txt
+//
+//	o := new(dns.OPT)
+//	o.Hdr.Name = "."
+//	o.Hdr.Rrtype = dns.TypeOPT
+//	e := new(dns.EDNS0_UL)
+//	e.Code = dns.EDNS0UL
+//	e.Lease = 120 // in seconds
+//	o.Option = append(o.Option, e)
+type EDNS0_UL struct {
+	Code  uint16 // Always EDNS0UL
+	Lease uint32
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_UL) Option() uint16 { return EDNS0UL }
+func (e *EDNS0_UL) String() string { return strconv.FormatUint(uint64(e.Lease), 10) }
+
+// Copied: http://golang.org/src/pkg/net/dnsmsg.go
+func (e *EDNS0_UL) pack() ([]byte, error) {
+	b := make([]byte, 4)
+	binary.BigEndian.PutUint32(b, e.Lease)
+	return b, nil
+}
+
+func (e *EDNS0_UL) unpack(b []byte) error {
+	if len(b) < 4 {
+		return ErrBuf
+	}
+	e.Lease = binary.BigEndian.Uint32(b)
+	return nil
+}
+
+// EDNS0_LLQ stands for Long Lived Queries: http://tools.ietf.org/html/draft-sekar-dns-llq-01
+// Implemented for completeness, as the EDNS0 type code is assigned.
+type EDNS0_LLQ struct {
+	Code      uint16 // Always EDNS0LLQ
+	Version   uint16
+	Opcode    uint16
+	Error     uint16
+	Id        uint64
+	LeaseLife uint32
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_LLQ) Option() uint16 { return EDNS0LLQ }
+
+func (e *EDNS0_LLQ) pack() ([]byte, error) {
+	b := make([]byte, 18)
+	binary.BigEndian.PutUint16(b[0:], e.Version)
+	binary.BigEndian.PutUint16(b[2:], e.Opcode)
+	binary.BigEndian.PutUint16(b[4:], e.Error)
+	binary.BigEndian.PutUint64(b[6:], e.Id)
+	binary.BigEndian.PutUint32(b[14:], e.LeaseLife)
+	return b, nil
+}
+
+func (e *EDNS0_LLQ) unpack(b []byte) error {
+	if len(b) < 18 {
+		return ErrBuf
+	}
+	e.Version = binary.BigEndian.Uint16(b[0:])
+	e.Opcode = binary.BigEndian.Uint16(b[2:])
+	e.Error = binary.BigEndian.Uint16(b[4:])
+	e.Id = binary.BigEndian.Uint64(b[6:])
+	e.LeaseLife = binary.BigEndian.Uint32(b[14:])
+	return nil
+}
+
+func (e *EDNS0_LLQ) String() string {
+	s := strconv.FormatUint(uint64(e.Version), 10) + " " + strconv.FormatUint(uint64(e.Opcode), 10) +
+		" " + strconv.FormatUint(uint64(e.Error), 10) + " " + strconv.FormatUint(uint64(e.Id), 10) +
+		" " + strconv.FormatUint(uint64(e.LeaseLife), 10)
+	return s
+}
+
+// EDNS0_DUA implements the EDNS0 "DNSSEC Algorithm Understood" option. See RFC 6975.
+type EDNS0_DAU struct {
+	Code    uint16 // Always EDNS0DAU
+	AlgCode []uint8
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_DAU) Option() uint16        { return EDNS0DAU }
+func (e *EDNS0_DAU) pack() ([]byte, error) { return e.AlgCode, nil }
+func (e *EDNS0_DAU) unpack(b []byte) error { e.AlgCode = b; return nil }
+
+func (e *EDNS0_DAU) String() string {
+	s := ""
+	for i := 0; i < len(e.AlgCode); i++ {
+		if a, ok := AlgorithmToString[e.AlgCode[i]]; ok {
+			s += " " + a
+		} else {
+			s += " " + strconv.Itoa(int(e.AlgCode[i]))
+		}
+	}
+	return s
+}
+
+// EDNS0_DHU implements the EDNS0 "DS Hash Understood" option. See RFC 6975.
+type EDNS0_DHU struct {
+	Code    uint16 // Always EDNS0DHU
+	AlgCode []uint8
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_DHU) Option() uint16        { return EDNS0DHU }
+func (e *EDNS0_DHU) pack() ([]byte, error) { return e.AlgCode, nil }
+func (e *EDNS0_DHU) unpack(b []byte) error { e.AlgCode = b; return nil }
+
+func (e *EDNS0_DHU) String() string {
+	s := ""
+	for i := 0; i < len(e.AlgCode); i++ {
+		if a, ok := HashToString[e.AlgCode[i]]; ok {
+			s += " " + a
+		} else {
+			s += " " + strconv.Itoa(int(e.AlgCode[i]))
+		}
+	}
+	return s
+}
+
+// EDNS0_N3U implements the EDNS0 "NSEC3 Hash Understood" option. See RFC 6975.
+type EDNS0_N3U struct {
+	Code    uint16 // Always EDNS0N3U
+	AlgCode []uint8
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_N3U) Option() uint16        { return EDNS0N3U }
+func (e *EDNS0_N3U) pack() ([]byte, error) { return e.AlgCode, nil }
+func (e *EDNS0_N3U) unpack(b []byte) error { e.AlgCode = b; return nil }
+
+func (e *EDNS0_N3U) String() string {
+	// Re-use the hash map
+	s := ""
+	for i := 0; i < len(e.AlgCode); i++ {
+		if a, ok := HashToString[e.AlgCode[i]]; ok {
+			s += " " + a
+		} else {
+			s += " " + strconv.Itoa(int(e.AlgCode[i]))
+		}
+	}
+	return s
+}
+
+// EDNS0_EXPIRE implementes the EDNS0 option as described in RFC 7314.
+type EDNS0_EXPIRE struct {
+	Code   uint16 // Always EDNS0EXPIRE
+	Expire uint32
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_EXPIRE) Option() uint16 { return EDNS0EXPIRE }
+func (e *EDNS0_EXPIRE) String() string { return strconv.FormatUint(uint64(e.Expire), 10) }
+
+func (e *EDNS0_EXPIRE) pack() ([]byte, error) {
+	b := make([]byte, 4)
+	b[0] = byte(e.Expire >> 24)
+	b[1] = byte(e.Expire >> 16)
+	b[2] = byte(e.Expire >> 8)
+	b[3] = byte(e.Expire)
+	return b, nil
+}
+
+func (e *EDNS0_EXPIRE) unpack(b []byte) error {
+	if len(b) < 4 {
+		return ErrBuf
+	}
+	e.Expire = binary.BigEndian.Uint32(b)
+	return nil
+}
+
+// The EDNS0_LOCAL option is used for local/experimental purposes. The option
+// code is recommended to be within the range [EDNS0LOCALSTART, EDNS0LOCALEND]
+// (RFC6891), although any unassigned code can actually be used.  The content of
+// the option is made available in Data, unaltered.
+// Basic use pattern for creating a local option:
+//
+//	o := new(dns.OPT)
+//	o.Hdr.Name = "."
+//	o.Hdr.Rrtype = dns.TypeOPT
+//	e := new(dns.EDNS0_LOCAL)
+//	e.Code = dns.EDNS0LOCALSTART
+//	e.Data = []byte{72, 82, 74}
+//	o.Option = append(o.Option, e)
+type EDNS0_LOCAL struct {
+	Code uint16
+	Data []byte
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_LOCAL) Option() uint16 { return e.Code }
+func (e *EDNS0_LOCAL) String() string {
+	return strconv.FormatInt(int64(e.Code), 10) + ":0x" + hex.EncodeToString(e.Data)
+}
+
+func (e *EDNS0_LOCAL) pack() ([]byte, error) {
+	b := make([]byte, len(e.Data))
+	copied := copy(b, e.Data)
+	if copied != len(e.Data) {
+		return nil, ErrBuf
+	}
+	return b, nil
+}
+
+func (e *EDNS0_LOCAL) unpack(b []byte) error {
+	e.Data = make([]byte, len(b))
+	copied := copy(e.Data, b)
+	if copied != len(b) {
+		return ErrBuf
+	}
+	return nil
+}
+
+// EDNS0_TCP_KEEPALIVE is an EDNS0 option that instructs the server to keep
+// the TCP connection alive. See RFC 7828.
+type EDNS0_TCP_KEEPALIVE struct {
+	Code    uint16 // Always EDNSTCPKEEPALIVE
+	Length  uint16 // the value 0 if the TIMEOUT is omitted, the value 2 if it is present;
+	Timeout uint16 // an idle timeout value for the TCP connection, specified in units of 100 milliseconds, encoded in network byte order.
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_TCP_KEEPALIVE) Option() uint16 { return EDNS0TCPKEEPALIVE }
+
+func (e *EDNS0_TCP_KEEPALIVE) pack() ([]byte, error) {
+	if e.Timeout != 0 && e.Length != 2 {
+		return nil, errors.New("dns: timeout specified but length is not 2")
+	}
+	if e.Timeout == 0 && e.Length != 0 {
+		return nil, errors.New("dns: timeout not specified but length is not 0")
+	}
+	b := make([]byte, 4+e.Length)
+	binary.BigEndian.PutUint16(b[0:], e.Code)
+	binary.BigEndian.PutUint16(b[2:], e.Length)
+	if e.Length == 2 {
+		binary.BigEndian.PutUint16(b[4:], e.Timeout)
+	}
+	return b, nil
+}
+
+func (e *EDNS0_TCP_KEEPALIVE) unpack(b []byte) error {
+	if len(b) < 4 {
+		return ErrBuf
+	}
+	e.Length = binary.BigEndian.Uint16(b[2:4])
+	if e.Length != 0 && e.Length != 2 {
+		return errors.New("dns: length mismatch, want 0/2 but got " + strconv.FormatUint(uint64(e.Length), 10))
+	}
+	if e.Length == 2 {
+		if len(b) < 6 {
+			return ErrBuf
+		}
+		e.Timeout = binary.BigEndian.Uint16(b[4:6])
+	}
+	return nil
+}
+
+func (e *EDNS0_TCP_KEEPALIVE) String() (s string) {
+	s = "use tcp keep-alive"
+	if e.Length == 0 {
+		s += ", timeout omitted"
+	} else {
+		s += fmt.Sprintf(", timeout %dms", e.Timeout*100)
+	}
+	return
+}
+
+// EDNS0_PADDING option is used to add padding to a request/response. The default
+// value of padding SHOULD be 0x0 but other values MAY be used, for instance if
+// compression is applied before encryption which may break signatures.
+type EDNS0_PADDING struct {
+	Padding []byte
+}
+
+// Option implements the EDNS0 interface.
+func (e *EDNS0_PADDING) Option() uint16        { return EDNS0PADDING }
+func (e *EDNS0_PADDING) pack() ([]byte, error) { return e.Padding, nil }
+func (e *EDNS0_PADDING) unpack(b []byte) error { e.Padding = b; return nil }
+func (e *EDNS0_PADDING) String() string        { return fmt.Sprintf("%0X", e.Padding) }
--- a/vendor/github.com/miekg/dns/format.go
+++ b/vendor/github.com/miekg/dns/format.go
@ -0,0 +1,87 @@
+package dns
+
+import (
+	"net"
+	"reflect"
+	"strconv"
+)
+
+// NumField returns the number of rdata fields r has.
+func NumField(r RR) int {
+	return reflect.ValueOf(r).Elem().NumField() - 1 // Remove RR_Header
+}
+
+// Field returns the rdata field i as a string. Fields are indexed starting from 1.
+// RR types that holds slice data, for instance the NSEC type bitmap will return a single
+// string where the types are concatenated using a space.
+// Accessing non existing fields will cause a panic.
+func Field(r RR, i int) string {
+	if i == 0 {
+		return ""
+	}
+	d := reflect.ValueOf(r).Elem().Field(i)
+	switch k := d.Kind(); k {
+	case reflect.String:
+		return d.String()
+	case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
+		return strconv.FormatInt(d.Int(), 10)
+	case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
+		return strconv.FormatUint(d.Uint(), 10)
+	case reflect.Slice:
+		switch reflect.ValueOf(r).Elem().Type().Field(i).Tag {
+		case `dns:"a"`:
+			// TODO(miek): Hmm store this as 16 bytes
+			if d.Len() < net.IPv6len {
+				return net.IPv4(byte(d.Index(0).Uint()),
+					byte(d.Index(1).Uint()),
+					byte(d.Index(2).Uint()),
+					byte(d.Index(3).Uint())).String()
+			}
+			return net.IPv4(byte(d.Index(12).Uint()),
+				byte(d.Index(13).Uint()),
+				byte(d.Index(14).Uint()),
+				byte(d.Index(15).Uint())).String()
+		case `dns:"aaaa"`:
+			return net.IP{
+				byte(d.Index(0).Uint()),
+				byte(d.Index(1).Uint()),
+				byte(d.Index(2).Uint()),
+				byte(d.Index(3).Uint()),
+				byte(d.Index(4).Uint()),
+				byte(d.Index(5).Uint()),
+				byte(d.Index(6).Uint()),
+				byte(d.Index(7).Uint()),
+				byte(d.Index(8).Uint()),
+				byte(d.Index(9).Uint()),
+				byte(d.Index(10).Uint()),
+				byte(d.Index(11).Uint()),
+				byte(d.Index(12).Uint()),
+				byte(d.Index(13).Uint()),
+				byte(d.Index(14).Uint()),
+				byte(d.Index(15).Uint()),
+			}.String()
+		case `dns:"nsec"`:
+			if d.Len() == 0 {
+				return ""
+			}
+			s := Type(d.Index(0).Uint()).String()
+			for i := 1; i < d.Len(); i++ {
+				s += " " + Type(d.Index(i).Uint()).String()
+			}
+			return s
+		default:
+			// if it does not have a tag its a string slice
+			fallthrough
+		case `dns:"txt"`:
+			if d.Len() == 0 {
+				return ""
+			}
+			s := d.Index(0).String()
+			for i := 1; i < d.Len(); i++ {
+				s += " " + d.Index(i).String()
+			}
+			return s
+		}
+	}
+	return ""
+}
--- a/vendor/github.com/miekg/dns/fuzz.go
+++ b/vendor/github.com/miekg/dns/fuzz.go
@ -0,0 +1,23 @@
+// +build fuzz
+
+package dns
+
+func Fuzz(data []byte) int {
+	msg := new(Msg)
+
+	if err := msg.Unpack(data); err != nil {
+		return 0
+	}
+	if _, err := msg.Pack(); err != nil {
+		return 0
+	}
+
+	return 1
+}
+
+func FuzzNewRR(data []byte) int {
+	if _, err := NewRR(string(data)); err != nil {
+		return 0
+	}
+	return 1
+}
--- a/vendor/github.com/miekg/dns/generate.go
+++ b/vendor/github.com/miekg/dns/generate.go
@ -0,0 +1,159 @@
+package dns
+
+import (
+	"bytes"
+	"errors"
+	"fmt"
+	"strconv"
+	"strings"
+)
+
+// Parse the $GENERATE statement as used in BIND9 zones.
+// See http://www.zytrax.com/books/dns/ch8/generate.html for instance.
+// We are called after '$GENERATE '. After which we expect:
+// * the range (12-24/2)
+// * lhs (ownername)
+// * [[ttl][class]]
+// * type
+// * rhs (rdata)
+// But we are lazy here, only the range is parsed *all* occurrences
+// of $ after that are interpreted.
+// Any error are returned as a string value, the empty string signals
+// "no error".
+func generate(l lex, c chan lex, t chan *Token, o string) string {
+	step := 1
+	if i := strings.IndexAny(l.token, "/"); i != -1 {
+		if i+1 == len(l.token) {
+			return "bad step in $GENERATE range"
+		}
+		if s, err := strconv.Atoi(l.token[i+1:]); err == nil {
+			if s < 0 {
+				return "bad step in $GENERATE range"
+			}
+			step = s
+		} else {
+			return "bad step in $GENERATE range"
+		}
+		l.token = l.token[:i]
+	}
+	sx := strings.SplitN(l.token, "-", 2)
+	if len(sx) != 2 {
+		return "bad start-stop in $GENERATE range"
+	}
+	start, err := strconv.Atoi(sx[0])
+	if err != nil {
+		return "bad start in $GENERATE range"
+	}
+	end, err := strconv.Atoi(sx[1])
+	if err != nil {
+		return "bad stop in $GENERATE range"
+	}
+	if end < 0 || start < 0 || end < start {
+		return "bad range in $GENERATE range"
+	}
+
+	<-c // _BLANK
+	// Create a complete new string, which we then parse again.
+	s := ""
+BuildRR:
+	l = <-c
+	if l.value != zNewline && l.value != zEOF {
+		s += l.token
+		goto BuildRR
+	}
+	for i := start; i <= end; i += step {
+		var (
+			escape bool
+			dom    bytes.Buffer
+			mod    string
+			err    error
+			offset int
+		)
+
+		for j := 0; j < len(s); j++ { // No 'range' because we need to jump around
+			switch s[j] {
+			case '\\':
+				if escape {
+					dom.WriteByte('\\')
+					escape = false
+					continue
+				}
+				escape = true
+			case '$':
+				mod = "%d"
+				offset = 0
+				if escape {
+					dom.WriteByte('$')
+					escape = false
+					continue
+				}
+				escape = false
+				if j+1 >= len(s) { // End of the string
+					dom.WriteString(fmt.Sprintf(mod, i+offset))
+					continue
+				} else {
+					if s[j+1] == '$' {
+						dom.WriteByte('$')
+						j++
+						continue
+					}
+				}
+				// Search for { and }
+				if s[j+1] == '{' { // Modifier block
+					sep := strings.Index(s[j+2:], "}")
+					if sep == -1 {
+						return "bad modifier in $GENERATE"
+					}
+					mod, offset, err = modToPrintf(s[j+2 : j+2+sep])
+					if err != nil {
+						return err.Error()
+					}
+					j += 2 + sep // Jump to it
+				}
+				dom.WriteString(fmt.Sprintf(mod, i+offset))
+			default:
+				if escape { // Pretty useless here
+					escape = false
+					continue
+				}
+				dom.WriteByte(s[j])
+			}
+		}
+		// Re-parse the RR and send it on the current channel t
+		rx, err := NewRR("$ORIGIN " + o + "\n" + dom.String())
+		if err != nil {
+			return err.Error()
+		}
+		t <- &Token{RR: rx}
+		// Its more efficient to first built the rrlist and then parse it in
+		// one go! But is this a problem?
+	}
+	return ""
+}
+
+// Convert a $GENERATE modifier 0,0,d to something Printf can deal with.
+func modToPrintf(s string) (string, int, error) {
+	xs := strings.SplitN(s, ",", 3)
+	if len(xs) != 3 {
+		return "", 0, errors.New("bad modifier in $GENERATE")
+	}
+	// xs[0] is offset, xs[1] is width, xs[2] is base
+	if xs[2] != "o" && xs[2] != "d" && xs[2] != "x" && xs[2] != "X" {
+		return "", 0, errors.New("bad base in $GENERATE")
+	}
+	offset, err := strconv.Atoi(xs[0])
+	if err != nil || offset > 255 {
+		return "", 0, errors.New("bad offset in $GENERATE")
+	}
+	width, err := strconv.Atoi(xs[1])
+	if err != nil || width > 255 {
+		return "", offset, errors.New("bad width in $GENERATE")
+	}
+	switch {
+	case width < 0:
+		return "", offset, errors.New("bad width in $GENERATE")
+	case width == 0:
+		return "%" + xs[1] + xs[2], offset, nil
+	}
+	return "%0" + xs[1] + xs[2], offset, nil
+}
--- a/vendor/github.com/miekg/dns/labels.go
+++ b/vendor/github.com/miekg/dns/labels.go
@ -0,0 +1,191 @@
+package dns
+
+// Holds a bunch of helper functions for dealing with labels.
+
+// SplitDomainName splits a name string into it's labels.
+// www.miek.nl. returns []string{"www", "miek", "nl"}
+// .www.miek.nl. returns []string{"", "www", "miek", "nl"},
+// The root label (.) returns nil. Note that using
+// strings.Split(s) will work in most cases, but does not handle
+// escaped dots (\.) for instance.
+// s must be a syntactically valid domain name, see IsDomainName.
+func SplitDomainName(s string) (labels []string) {
+	if len(s) == 0 {
+		return nil
+	}
+	fqdnEnd := 0 // offset of the final '.' or the length of the name
+	idx := Split(s)
+	begin := 0
+	if s[len(s)-1] == '.' {
+		fqdnEnd = len(s) - 1
+	} else {
+		fqdnEnd = len(s)
+	}
+
+	switch len(idx) {
+	case 0:
+		return nil
+	case 1:
+		// no-op
+	default:
+		end := 0
+		for i := 1; i < len(idx); i++ {
+			end = idx[i]
+			labels = append(labels, s[begin:end-1])
+			begin = end
+		}
+	}
+
+	labels = append(labels, s[begin:fqdnEnd])
+	return labels
+}
+
+// CompareDomainName compares the names s1 and s2 and
+// returns how many labels they have in common starting from the *right*.
+// The comparison stops at the first inequality. The names are downcased
+// before the comparison.
+//
+// www.miek.nl. and miek.nl. have two labels in common: miek and nl
+// www.miek.nl. and www.bla.nl. have one label in common: nl
+//
+// s1 and s2 must be syntactically valid domain names.
+func CompareDomainName(s1, s2 string) (n int) {
+	// the first check: root label
+	if s1 == "." || s2 == "." {
+		return 0
+	}
+
+	l1 := Split(s1)
+	l2 := Split(s2)
+
+	j1 := len(l1) - 1 // end
+	i1 := len(l1) - 2 // start
+	j2 := len(l2) - 1
+	i2 := len(l2) - 2
+	// the second check can be done here: last/only label
+	// before we fall through into the for-loop below
+	if equal(s1[l1[j1]:], s2[l2[j2]:]) {
+		n++
+	} else {
+		return
+	}
+	for {
+		if i1 < 0 || i2 < 0 {
+			break
+		}
+		if equal(s1[l1[i1]:l1[j1]], s2[l2[i2]:l2[j2]]) {
+			n++
+		} else {
+			break
+		}
+		j1--
+		i1--
+		j2--
+		i2--
+	}
+	return
+}
+
+// CountLabel counts the the number of labels in the string s.
+// s must be a syntactically valid domain name.
+func CountLabel(s string) (labels int) {
+	if s == "." {
+		return
+	}
+	off := 0
+	end := false
+	for {
+		off, end = NextLabel(s, off)
+		labels++
+		if end {
+			return
+		}
+	}
+}
+
+// Split splits a name s into its label indexes.
+// www.miek.nl. returns []int{0, 4, 9}, www.miek.nl also returns []int{0, 4, 9}.
+// The root name (.) returns nil. Also see SplitDomainName.
+// s must be a syntactically valid domain name.
+func Split(s string) []int {
+	if s == "." {
+		return nil
+	}
+	idx := make([]int, 1, 3)
+	off := 0
+	end := false
+
+	for {
+		off, end = NextLabel(s, off)
+		if end {
+			return idx
+		}
+		idx = append(idx, off)
+	}
+}
+
+// NextLabel returns the index of the start of the next label in the
+// string s starting at offset.
+// The bool end is true when the end of the string has been reached.
+// Also see PrevLabel.
+func NextLabel(s string, offset int) (i int, end bool) {
+	quote := false
+	for i = offset; i < len(s)-1; i++ {
+		switch s[i] {
+		case '\\':
+			quote = !quote
+		default:
+			quote = false
+		case '.':
+			if quote {
+				quote = !quote
+				continue
+			}
+			return i + 1, false
+		}
+	}
+	return i + 1, true
+}
+
+// PrevLabel returns the index of the label when starting from the right and
+// jumping n labels to the left.
+// The bool start is true when the start of the string has been overshot.
+// Also see NextLabel.
+func PrevLabel(s string, n int) (i int, start bool) {
+	if n == 0 {
+		return len(s), false
+	}
+	lab := Split(s)
+	if lab == nil {
+		return 0, true
+	}
+	if n > len(lab) {
+		return 0, true
+	}
+	return lab[len(lab)-n], false
+}
+
+// equal compares a and b while ignoring case. It returns true when equal otherwise false.
+func equal(a, b string) bool {
+	// might be lifted into API function.
+	la := len(a)
+	lb := len(b)
+	if la != lb {
+		return false
+	}
+
+	for i := la - 1; i >= 0; i-- {
+		ai := a[i]
+		bi := b[i]
+		if ai >= 'A' && ai <= 'Z' {
+			ai |= ('a' - 'A')
+		}
+		if bi >= 'A' && bi <= 'Z' {
+			bi |= ('a' - 'A')
+		}
+		if ai != bi {
+			return false
+		}
+	}
+	return true
+}
--- a/vendor/github.com/miekg/dns/msg.go
+++ b/vendor/github.com/miekg/dns/msg.go
--- a/vendor/github.com/miekg/dns/msg_generate.go
+++ b/vendor/github.com/miekg/dns/msg_generate.go
@ -0,0 +1,348 @@
+//+build ignore
+
+// msg_generate.go is meant to run with go generate. It will use
+// go/{importer,types} to track down all the RR struct types. Then for each type
+// it will generate pack/unpack methods based on the struct tags. The generated source is
+// written to zmsg.go, and is meant to be checked into git.
+package main
+
+import (
+	"bytes"
+	"fmt"
+	"go/format"
+	"go/importer"
+	"go/types"
+	"log"
+	"os"
+	"strings"
+)
+
+var packageHdr = `
+// Code generated by "go run msg_generate.go"; DO NOT EDIT.
+
+package dns
+
+`
+
+// getTypeStruct will take a type and the package scope, and return the
+// (innermost) struct if the type is considered a RR type (currently defined as
+// those structs beginning with a RR_Header, could be redefined as implementing
+// the RR interface). The bool return value indicates if embedded structs were
+// resolved.
+func getTypeStruct(t types.Type, scope *types.Scope) (*types.Struct, bool) {
+	st, ok := t.Underlying().(*types.Struct)
+	if !ok {
+		return nil, false
+	}
+	if st.Field(0).Type() == scope.Lookup("RR_Header").Type() {
+		return st, false
+	}
+	if st.Field(0).Anonymous() {
+		st, _ := getTypeStruct(st.Field(0).Type(), scope)
+		return st, true
+	}
+	return nil, false
+}
+
+func main() {
+	// Import and type-check the package
+	pkg, err := importer.Default().Import("github.com/miekg/dns")
+	fatalIfErr(err)
+	scope := pkg.Scope()
+
+	// Collect actual types (*X)
+	var namedTypes []string
+	for _, name := range scope.Names() {
+		o := scope.Lookup(name)
+		if o == nil || !o.Exported() {
+			continue
+		}
+		if st, _ := getTypeStruct(o.Type(), scope); st == nil {
+			continue
+		}
+		if name == "PrivateRR" {
+			continue
+		}
+
+		// Check if corresponding TypeX exists
+		if scope.Lookup("Type"+o.Name()) == nil && o.Name() != "RFC3597" {
+			log.Fatalf("Constant Type%s does not exist.", o.Name())
+		}
+
+		namedTypes = append(namedTypes, o.Name())
+	}
+
+	b := &bytes.Buffer{}
+	b.WriteString(packageHdr)
+
+	fmt.Fprint(b, "// pack*() functions\n\n")
+	for _, name := range namedTypes {
+		o := scope.Lookup(name)
+		st, _ := getTypeStruct(o.Type(), scope)
+
+		fmt.Fprintf(b, "func (rr *%s) pack(msg []byte, off int, compression map[string]int, compress bool) (int, error) {\n", name)
+		fmt.Fprint(b, `off, err := rr.Hdr.pack(msg, off, compression, compress)
+if err != nil {
+	return off, err
+}
+headerEnd := off
+`)
+		for i := 1; i < st.NumFields(); i++ {
+			o := func(s string) {
+				fmt.Fprintf(b, s, st.Field(i).Name())
+				fmt.Fprint(b, `if err != nil {
+return off, err
+}
+`)
+			}
+
+			if _, ok := st.Field(i).Type().(*types.Slice); ok {
+				switch st.Tag(i) {
+				case `dns:"-"`: // ignored
+				case `dns:"txt"`:
+					o("off, err = packStringTxt(rr.%s, msg, off)\n")
+				case `dns:"opt"`:
+					o("off, err = packDataOpt(rr.%s, msg, off)\n")
+				case `dns:"nsec"`:
+					o("off, err = packDataNsec(rr.%s, msg, off)\n")
+				case `dns:"domain-name"`:
+					o("off, err = packDataDomainNames(rr.%s, msg, off, compression, compress)\n")
+				default:
+					log.Fatalln(name, st.Field(i).Name(), st.Tag(i))
+				}
+				continue
+			}
+
+			switch {
+			case st.Tag(i) == `dns:"-"`: // ignored
+			case st.Tag(i) == `dns:"cdomain-name"`:
+				o("off, err = PackDomainName(rr.%s, msg, off, compression, compress)\n")
+			case st.Tag(i) == `dns:"domain-name"`:
+				o("off, err = PackDomainName(rr.%s, msg, off, compression, false)\n")
+			case st.Tag(i) == `dns:"a"`:
+				o("off, err = packDataA(rr.%s, msg, off)\n")
+			case st.Tag(i) == `dns:"aaaa"`:
+				o("off, err = packDataAAAA(rr.%s, msg, off)\n")
+			case st.Tag(i) == `dns:"uint48"`:
+				o("off, err = packUint48(rr.%s, msg, off)\n")
+			case st.Tag(i) == `dns:"txt"`:
+				o("off, err = packString(rr.%s, msg, off)\n")
+
+			case strings.HasPrefix(st.Tag(i), `dns:"size-base32`): // size-base32 can be packed just like base32
+				fallthrough
+			case st.Tag(i) == `dns:"base32"`:
+				o("off, err = packStringBase32(rr.%s, msg, off)\n")
+
+			case strings.HasPrefix(st.Tag(i), `dns:"size-base64`): // size-base64 can be packed just like base64
+				fallthrough
+			case st.Tag(i) == `dns:"base64"`:
+				o("off, err = packStringBase64(rr.%s, msg, off)\n")
+
+			case strings.HasPrefix(st.Tag(i), `dns:"size-hex:SaltLength`):
+				// directly write instead of using o() so we get the error check in the correct place
+				field := st.Field(i).Name()
+				fmt.Fprintf(b, `// Only pack salt if value is not "-", i.e. empty
+if rr.%s != "-" {
+  off, err = packStringHex(rr.%s, msg, off)
+  if err != nil {
+    return off, err
+  }
+}
+`, field, field)
+				continue
+			case strings.HasPrefix(st.Tag(i), `dns:"size-hex`): // size-hex can be packed just like hex
+				fallthrough
+			case st.Tag(i) == `dns:"hex"`:
+				o("off, err = packStringHex(rr.%s, msg, off)\n")
+
+			case st.Tag(i) == `dns:"octet"`:
+				o("off, err = packStringOctet(rr.%s, msg, off)\n")
+			case st.Tag(i) == "":
+				switch st.Field(i).Type().(*types.Basic).Kind() {
+				case types.Uint8:
+					o("off, err = packUint8(rr.%s, msg, off)\n")
+				case types.Uint16:
+					o("off, err = packUint16(rr.%s, msg, off)\n")
+				case types.Uint32:
+					o("off, err = packUint32(rr.%s, msg, off)\n")
+				case types.Uint64:
+					o("off, err = packUint64(rr.%s, msg, off)\n")
+				case types.String:
+					o("off, err = packString(rr.%s, msg, off)\n")
+				default:
+					log.Fatalln(name, st.Field(i).Name())
+				}
+			default:
+				log.Fatalln(name, st.Field(i).Name(), st.Tag(i))
+			}
+		}
+		// We have packed everything, only now we know the rdlength of this RR
+		fmt.Fprintln(b, "rr.Header().Rdlength = uint16(off-headerEnd)")
+		fmt.Fprintln(b, "return off, nil }\n")
+	}
+
+	fmt.Fprint(b, "// unpack*() functions\n\n")
+	for _, name := range namedTypes {
+		o := scope.Lookup(name)
+		st, _ := getTypeStruct(o.Type(), scope)
+
+		fmt.Fprintf(b, "func unpack%s(h RR_Header, msg []byte, off int) (RR, int, error) {\n", name)
+		fmt.Fprintf(b, "rr := new(%s)\n", name)
+		fmt.Fprint(b, "rr.Hdr = h\n")
+		fmt.Fprint(b, `if noRdata(h) {
+return rr, off, nil
+	}
+var err error
+rdStart := off
+_ = rdStart
+
+`)
+		for i := 1; i < st.NumFields(); i++ {
+			o := func(s string) {
+				fmt.Fprintf(b, s, st.Field(i).Name())
+				fmt.Fprint(b, `if err != nil {
+return rr, off, err
+}
+`)
+			}
+
+			// size-* are special, because they reference a struct member we should use for the length.
+			if strings.HasPrefix(st.Tag(i), `dns:"size-`) {
+				structMember := structMember(st.Tag(i))
+				structTag := structTag(st.Tag(i))
+				switch structTag {
+				case "hex":
+					fmt.Fprintf(b, "rr.%s, off, err = unpackStringHex(msg, off, off + int(rr.%s))\n", st.Field(i).Name(), structMember)
+				case "base32":
+					fmt.Fprintf(b, "rr.%s, off, err = unpackStringBase32(msg, off, off + int(rr.%s))\n", st.Field(i).Name(), structMember)
+				case "base64":
+					fmt.Fprintf(b, "rr.%s, off, err = unpackStringBase64(msg, off, off + int(rr.%s))\n", st.Field(i).Name(), structMember)
+				default:
+					log.Fatalln(name, st.Field(i).Name(), st.Tag(i))
+				}
+				fmt.Fprint(b, `if err != nil {
+return rr, off, err
+}
+`)
+				continue
+			}
+
+			if _, ok := st.Field(i).Type().(*types.Slice); ok {
+				switch st.Tag(i) {
+				case `dns:"-"`: // ignored
+				case `dns:"txt"`:
+					o("rr.%s, off, err = unpackStringTxt(msg, off)\n")
+				case `dns:"opt"`:
+					o("rr.%s, off, err = unpackDataOpt(msg, off)\n")
+				case `dns:"nsec"`:
+					o("rr.%s, off, err = unpackDataNsec(msg, off)\n")
+				case `dns:"domain-name"`:
+					o("rr.%s, off, err = unpackDataDomainNames(msg, off, rdStart + int(rr.Hdr.Rdlength))\n")
+				default:
+					log.Fatalln(name, st.Field(i).Name(), st.Tag(i))
+				}
+				continue
+			}
+
+			switch st.Tag(i) {
+			case `dns:"-"`: // ignored
+			case `dns:"cdomain-name"`:
+				fallthrough
+			case `dns:"domain-name"`:
+				o("rr.%s, off, err = UnpackDomainName(msg, off)\n")
+			case `dns:"a"`:
+				o("rr.%s, off, err = unpackDataA(msg, off)\n")
+			case `dns:"aaaa"`:
+				o("rr.%s, off, err = unpackDataAAAA(msg, off)\n")
+			case `dns:"uint48"`:
+				o("rr.%s, off, err = unpackUint48(msg, off)\n")
+			case `dns:"txt"`:
+				o("rr.%s, off, err = unpackString(msg, off)\n")
+			case `dns:"base32"`:
+				o("rr.%s, off, err = unpackStringBase32(msg, off, rdStart + int(rr.Hdr.Rdlength))\n")
+			case `dns:"base64"`:
+				o("rr.%s, off, err = unpackStringBase64(msg, off, rdStart + int(rr.Hdr.Rdlength))\n")
+			case `dns:"hex"`:
+				o("rr.%s, off, err = unpackStringHex(msg, off, rdStart + int(rr.Hdr.Rdlength))\n")
+			case `dns:"octet"`:
+				o("rr.%s, off, err = unpackStringOctet(msg, off)\n")
+			case "":
+				switch st.Field(i).Type().(*types.Basic).Kind() {
+				case types.Uint8:
+					o("rr.%s, off, err = unpackUint8(msg, off)\n")
+				case types.Uint16:
+					o("rr.%s, off, err = unpackUint16(msg, off)\n")
+				case types.Uint32:
+					o("rr.%s, off, err = unpackUint32(msg, off)\n")
+				case types.Uint64:
+					o("rr.%s, off, err = unpackUint64(msg, off)\n")
+				case types.String:
+					o("rr.%s, off, err = unpackString(msg, off)\n")
+				default:
+					log.Fatalln(name, st.Field(i).Name())
+				}
+			default:
+				log.Fatalln(name, st.Field(i).Name(), st.Tag(i))
+			}
+			// If we've hit len(msg) we return without error.
+			if i < st.NumFields()-1 {
+				fmt.Fprintf(b, `if off == len(msg) {
+return rr, off, nil
+	}
+`)
+			}
+		}
+		fmt.Fprintf(b, "return rr, off, err }\n\n")
+	}
+	// Generate typeToUnpack map
+	fmt.Fprintln(b, "var typeToUnpack = map[uint16]func(RR_Header, []byte, int) (RR, int, error){")
+	for _, name := range namedTypes {
+		if name == "RFC3597" {
+			continue
+		}
+		fmt.Fprintf(b, "Type%s: unpack%s,\n", name, name)
+	}
+	fmt.Fprintln(b, "}\n")
+
+	// gofmt
+	res, err := format.Source(b.Bytes())
+	if err != nil {
+		b.WriteTo(os.Stderr)
+		log.Fatal(err)
+	}
+
+	// write result
+	f, err := os.Create("zmsg.go")
+	fatalIfErr(err)
+	defer f.Close()
+	f.Write(res)
+}
+
+// structMember will take a tag like dns:"size-base32:SaltLength" and return the last part of this string.
+func structMember(s string) string {
+	fields := strings.Split(s, ":")
+	if len(fields) == 0 {
+		return ""
+	}
+	f := fields[len(fields)-1]
+	// f should have a closing "
+	if len(f) > 1 {
+		return f[:len(f)-1]
+	}
+	return f
+}
+
+// structTag will take a tag like dns:"size-base32:SaltLength" and return base32.
+func structTag(s string) string {
+	fields := strings.Split(s, ":")
+	if len(fields) < 2 {
+		return ""
+	}
+	return fields[1][len("\"size-"):]
+}
+
+func fatalIfErr(err error) {
+	if err != nil {
+		log.Fatal(err)
+	}
+}
--- a/vendor/github.com/miekg/dns/msg_helpers.go
+++ b/vendor/github.com/miekg/dns/msg_helpers.go
@ -0,0 +1,637 @@
+package dns
+
+import (
+	"encoding/base32"
+	"encoding/base64"
+	"encoding/binary"
+	"encoding/hex"
+	"net"
+	"strconv"
+)
+
+// helper functions called from the generated zmsg.go
+
+// These function are named after the tag to help pack/unpack, if there is no tag it is the name
+// of the type they pack/unpack (string, int, etc). We prefix all with unpackData or packData, so packDataA or
+// packDataDomainName.
+
+func unpackDataA(msg []byte, off int) (net.IP, int, error) {
+	if off+net.IPv4len > len(msg) {
+		return nil, len(msg), &Error{err: "overflow unpacking a"}
+	}
+	a := append(make(net.IP, 0, net.IPv4len), msg[off:off+net.IPv4len]...)
+	off += net.IPv4len
+	return a, off, nil
+}
+
+func packDataA(a net.IP, msg []byte, off int) (int, error) {
+	// It must be a slice of 4, even if it is 16, we encode only the first 4
+	if off+net.IPv4len > len(msg) {
+		return len(msg), &Error{err: "overflow packing a"}
+	}
+	switch len(a) {
+	case net.IPv4len, net.IPv6len:
+		copy(msg[off:], a.To4())
+		off += net.IPv4len
+	case 0:
+		// Allowed, for dynamic updates.
+	default:
+		return len(msg), &Error{err: "overflow packing a"}
+	}
+	return off, nil
+}
+
+func unpackDataAAAA(msg []byte, off int) (net.IP, int, error) {
+	if off+net.IPv6len > len(msg) {
+		return nil, len(msg), &Error{err: "overflow unpacking aaaa"}
+	}
+	aaaa := append(make(net.IP, 0, net.IPv6len), msg[off:off+net.IPv6len]...)
+	off += net.IPv6len
+	return aaaa, off, nil
+}
+
+func packDataAAAA(aaaa net.IP, msg []byte, off int) (int, error) {
+	if off+net.IPv6len > len(msg) {
+		return len(msg), &Error{err: "overflow packing aaaa"}
+	}
+
+	switch len(aaaa) {
+	case net.IPv6len:
+		copy(msg[off:], aaaa)
+		off += net.IPv6len
+	case 0:
+		// Allowed, dynamic updates.
+	default:
+		return len(msg), &Error{err: "overflow packing aaaa"}
+	}
+	return off, nil
+}
+
+// unpackHeader unpacks an RR header, returning the offset to the end of the header and a
+// re-sliced msg according to the expected length of the RR.
+func unpackHeader(msg []byte, off int) (rr RR_Header, off1 int, truncmsg []byte, err error) {
+	hdr := RR_Header{}
+	if off == len(msg) {
+		return hdr, off, msg, nil
+	}
+
+	hdr.Name, off, err = UnpackDomainName(msg, off)
+	if err != nil {
+		return hdr, len(msg), msg, err
+	}
+	hdr.Rrtype, off, err = unpackUint16(msg, off)
+	if err != nil {
+		return hdr, len(msg), msg, err
+	}
+	hdr.Class, off, err = unpackUint16(msg, off)
+	if err != nil {
+		return hdr, len(msg), msg, err
+	}
+	hdr.Ttl, off, err = unpackUint32(msg, off)
+	if err != nil {
+		return hdr, len(msg), msg, err
+	}
+	hdr.Rdlength, off, err = unpackUint16(msg, off)
+	if err != nil {
+		return hdr, len(msg), msg, err
+	}
+	msg, err = truncateMsgFromRdlength(msg, off, hdr.Rdlength)
+	return hdr, off, msg, err
+}
+
+// pack packs an RR header, returning the offset to the end of the header.
+// See PackDomainName for documentation about the compression.
+func (hdr RR_Header) pack(msg []byte, off int, compression map[string]int, compress bool) (off1 int, err error) {
+	if off == len(msg) {
+		return off, nil
+	}
+
+	off, err = PackDomainName(hdr.Name, msg, off, compression, compress)
+	if err != nil {
+		return len(msg), err
+	}
+	off, err = packUint16(hdr.Rrtype, msg, off)
+	if err != nil {
+		return len(msg), err
+	}
+	off, err = packUint16(hdr.Class, msg, off)
+	if err != nil {
+		return len(msg), err
+	}
+	off, err = packUint32(hdr.Ttl, msg, off)
+	if err != nil {
+		return len(msg), err
+	}
+	off, err = packUint16(hdr.Rdlength, msg, off)
+	if err != nil {
+		return len(msg), err
+	}
+	return off, nil
+}
+
+// helper helper functions.
+
+// truncateMsgFromRdLength truncates msg to match the expected length of the RR.
+// Returns an error if msg is smaller than the expected size.
+func truncateMsgFromRdlength(msg []byte, off int, rdlength uint16) (truncmsg []byte, err error) {
+	lenrd := off + int(rdlength)
+	if lenrd > len(msg) {
+		return msg, &Error{err: "overflowing header size"}
+	}
+	return msg[:lenrd], nil
+}
+
+func fromBase32(s []byte) (buf []byte, err error) {
+	for i, b := range s {
+		if b >= 'a' && b <= 'z' {
+			s[i] = b - 32
+		}
+	}
+	buflen := base32.HexEncoding.DecodedLen(len(s))
+	buf = make([]byte, buflen)
+	n, err := base32.HexEncoding.Decode(buf, s)
+	buf = buf[:n]
+	return
+}
+
+func toBase32(b []byte) string { return base32.HexEncoding.EncodeToString(b) }
+
+func fromBase64(s []byte) (buf []byte, err error) {
+	buflen := base64.StdEncoding.DecodedLen(len(s))
+	buf = make([]byte, buflen)
+	n, err := base64.StdEncoding.Decode(buf, s)
+	buf = buf[:n]
+	return
+}
+
+func toBase64(b []byte) string { return base64.StdEncoding.EncodeToString(b) }
+
+// dynamicUpdate returns true if the Rdlength is zero.
+func noRdata(h RR_Header) bool { return h.Rdlength == 0 }
+
+func unpackUint8(msg []byte, off int) (i uint8, off1 int, err error) {
+	if off+1 > len(msg) {
+		return 0, len(msg), &Error{err: "overflow unpacking uint8"}
+	}
+	return uint8(msg[off]), off + 1, nil
+}
+
+func packUint8(i uint8, msg []byte, off int) (off1 int, err error) {
+	if off+1 > len(msg) {
+		return len(msg), &Error{err: "overflow packing uint8"}
+	}
+	msg[off] = byte(i)
+	return off + 1, nil
+}
+
+func unpackUint16(msg []byte, off int) (i uint16, off1 int, err error) {
+	if off+2 > len(msg) {
+		return 0, len(msg), &Error{err: "overflow unpacking uint16"}
+	}
+	return binary.BigEndian.Uint16(msg[off:]), off + 2, nil
+}
+
+func packUint16(i uint16, msg []byte, off int) (off1 int, err error) {
+	if off+2 > len(msg) {
+		return len(msg), &Error{err: "overflow packing uint16"}
+	}
+	binary.BigEndian.PutUint16(msg[off:], i)
+	return off + 2, nil
+}
+
+func unpackUint32(msg []byte, off int) (i uint32, off1 int, err error) {
+	if off+4 > len(msg) {
+		return 0, len(msg), &Error{err: "overflow unpacking uint32"}
+	}
+	return binary.BigEndian.Uint32(msg[off:]), off + 4, nil
+}
+
+func packUint32(i uint32, msg []byte, off int) (off1 int, err error) {
+	if off+4 > len(msg) {
+		return len(msg), &Error{err: "overflow packing uint32"}
+	}
+	binary.BigEndian.PutUint32(msg[off:], i)
+	return off + 4, nil
+}
+
+func unpackUint48(msg []byte, off int) (i uint64, off1 int, err error) {
+	if off+6 > len(msg) {
+		return 0, len(msg), &Error{err: "overflow unpacking uint64 as uint48"}
+	}
+	// Used in TSIG where the last 48 bits are occupied, so for now, assume a uint48 (6 bytes)
+	i = (uint64(uint64(msg[off])<<40 | uint64(msg[off+1])<<32 | uint64(msg[off+2])<<24 | uint64(msg[off+3])<<16 |
+		uint64(msg[off+4])<<8 | uint64(msg[off+5])))
+	off += 6
+	return i, off, nil
+}
+
+func packUint48(i uint64, msg []byte, off int) (off1 int, err error) {
+	if off+6 > len(msg) {
+		return len(msg), &Error{err: "overflow packing uint64 as uint48"}
+	}
+	msg[off] = byte(i >> 40)
+	msg[off+1] = byte(i >> 32)
+	msg[off+2] = byte(i >> 24)
+	msg[off+3] = byte(i >> 16)
+	msg[off+4] = byte(i >> 8)
+	msg[off+5] = byte(i)
+	off += 6
+	return off, nil
+}
+
+func unpackUint64(msg []byte, off int) (i uint64, off1 int, err error) {
+	if off+8 > len(msg) {
+		return 0, len(msg), &Error{err: "overflow unpacking uint64"}
+	}
+	return binary.BigEndian.Uint64(msg[off:]), off + 8, nil
+}
+
+func packUint64(i uint64, msg []byte, off int) (off1 int, err error) {
+	if off+8 > len(msg) {
+		return len(msg), &Error{err: "overflow packing uint64"}
+	}
+	binary.BigEndian.PutUint64(msg[off:], i)
+	off += 8
+	return off, nil
+}
+
+func unpackString(msg []byte, off int) (string, int, error) {
+	if off+1 > len(msg) {
+		return "", off, &Error{err: "overflow unpacking txt"}
+	}
+	l := int(msg[off])
+	if off+l+1 > len(msg) {
+		return "", off, &Error{err: "overflow unpacking txt"}
+	}
+	s := make([]byte, 0, l)
+	for _, b := range msg[off+1 : off+1+l] {
+		switch b {
+		case '"', '\\':
+			s = append(s, '\\', b)
+		default:
+			if b < 32 || b > 127 { // unprintable
+				var buf [3]byte
+				bufs := strconv.AppendInt(buf[:0], int64(b), 10)
+				s = append(s, '\\')
+				for i := 0; i < 3-len(bufs); i++ {
+					s = append(s, '0')
+				}
+				for _, r := range bufs {
+					s = append(s, r)
+				}
+			} else {
+				s = append(s, b)
+			}
+		}
+	}
+	off += 1 + l
+	return string(s), off, nil
+}
+
+func packString(s string, msg []byte, off int) (int, error) {
+	txtTmp := make([]byte, 256*4+1)
+	off, err := packTxtString(s, msg, off, txtTmp)
+	if err != nil {
+		return len(msg), err
+	}
+	return off, nil
+}
+
+func unpackStringBase32(msg []byte, off, end int) (string, int, error) {
+	if end > len(msg) {
+		return "", len(msg), &Error{err: "overflow unpacking base32"}
+	}
+	s := toBase32(msg[off:end])
+	return s, end, nil
+}
+
+func packStringBase32(s string, msg []byte, off int) (int, error) {
+	b32, err := fromBase32([]byte(s))
+	if err != nil {
+		return len(msg), err
+	}
+	if off+len(b32) > len(msg) {
+		return len(msg), &Error{err: "overflow packing base32"}
+	}
+	copy(msg[off:off+len(b32)], b32)
+	off += len(b32)
+	return off, nil
+}
+
+func unpackStringBase64(msg []byte, off, end int) (string, int, error) {
+	// Rest of the RR is base64 encoded value, so we don't need an explicit length
+	// to be set. Thus far all RR's that have base64 encoded fields have those as their
+	// last one. What we do need is the end of the RR!
+	if end > len(msg) {
+		return "", len(msg), &Error{err: "overflow unpacking base64"}
+	}
+	s := toBase64(msg[off:end])
+	return s, end, nil
+}
+
+func packStringBase64(s string, msg []byte, off int) (int, error) {
+	b64, err := fromBase64([]byte(s))
+	if err != nil {
+		return len(msg), err
+	}
+	if off+len(b64) > len(msg) {
+		return len(msg), &Error{err: "overflow packing base64"}
+	}
+	copy(msg[off:off+len(b64)], b64)
+	off += len(b64)
+	return off, nil
+}
+
+func unpackStringHex(msg []byte, off, end int) (string, int, error) {
+	// Rest of the RR is hex encoded value, so we don't need an explicit length
+	// to be set. NSEC and TSIG have hex fields with a length field.
+	// What we do need is the end of the RR!
+	if end > len(msg) {
+		return "", len(msg), &Error{err: "overflow unpacking hex"}
+	}
+
+	s := hex.EncodeToString(msg[off:end])
+	return s, end, nil
+}
+
+func packStringHex(s string, msg []byte, off int) (int, error) {
+	h, err := hex.DecodeString(s)
+	if err != nil {
+		return len(msg), err
+	}
+	if off+(len(h)) > len(msg) {
+		return len(msg), &Error{err: "overflow packing hex"}
+	}
+	copy(msg[off:off+len(h)], h)
+	off += len(h)
+	return off, nil
+}
+
+func unpackStringTxt(msg []byte, off int) ([]string, int, error) {
+	txt, off, err := unpackTxt(msg, off)
+	if err != nil {
+		return nil, len(msg), err
+	}
+	return txt, off, nil
+}
+
+func packStringTxt(s []string, msg []byte, off int) (int, error) {
+	txtTmp := make([]byte, 256*4+1) // If the whole string consists out of \DDD we need this many.
+	off, err := packTxt(s, msg, off, txtTmp)
+	if err != nil {
+		return len(msg), err
+	}
+	return off, nil
+}
+
+func unpackDataOpt(msg []byte, off int) ([]EDNS0, int, error) {
+	var edns []EDNS0
+Option:
+	code := uint16(0)
+	if off+4 > len(msg) {
+		return nil, len(msg), &Error{err: "overflow unpacking opt"}
+	}
+	code = binary.BigEndian.Uint16(msg[off:])
+	off += 2
+	optlen := binary.BigEndian.Uint16(msg[off:])
+	off += 2
+	if off+int(optlen) > len(msg) {
+		return nil, len(msg), &Error{err: "overflow unpacking opt"}
+	}
+	switch code {
+	case EDNS0NSID:
+		e := new(EDNS0_NSID)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0SUBNET:
+		e := new(EDNS0_SUBNET)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0COOKIE:
+		e := new(EDNS0_COOKIE)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0UL:
+		e := new(EDNS0_UL)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0LLQ:
+		e := new(EDNS0_LLQ)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0DAU:
+		e := new(EDNS0_DAU)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0DHU:
+		e := new(EDNS0_DHU)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0N3U:
+		e := new(EDNS0_N3U)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	case EDNS0PADDING:
+		e := new(EDNS0_PADDING)
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	default:
+		e := new(EDNS0_LOCAL)
+		e.Code = code
+		if err := e.unpack(msg[off : off+int(optlen)]); err != nil {
+			return nil, len(msg), err
+		}
+		edns = append(edns, e)
+		off += int(optlen)
+	}
+
+	if off < len(msg) {
+		goto Option
+	}
+
+	return edns, off, nil
+}
+
+func packDataOpt(options []EDNS0, msg []byte, off int) (int, error) {
+	for _, el := range options {
+		b, err := el.pack()
+		if err != nil || off+3 > len(msg) {
+			return len(msg), &Error{err: "overflow packing opt"}
+		}
+		binary.BigEndian.PutUint16(msg[off:], el.Option())      // Option code
+		binary.BigEndian.PutUint16(msg[off+2:], uint16(len(b))) // Length
+		off += 4
+		if off+len(b) > len(msg) {
+			copy(msg[off:], b)
+			off = len(msg)
+			continue
+		}
+		// Actual data
+		copy(msg[off:off+len(b)], b)
+		off += len(b)
+	}
+	return off, nil
+}
+
+func unpackStringOctet(msg []byte, off int) (string, int, error) {
+	s := string(msg[off:])
+	return s, len(msg), nil
+}
+
+func packStringOctet(s string, msg []byte, off int) (int, error) {
+	txtTmp := make([]byte, 256*4+1)
+	off, err := packOctetString(s, msg, off, txtTmp)
+	if err != nil {
+		return len(msg), err
+	}
+	return off, nil
+}
+
+func unpackDataNsec(msg []byte, off int) ([]uint16, int, error) {
+	var nsec []uint16
+	length, window, lastwindow := 0, 0, -1
+	for off < len(msg) {
+		if off+2 > len(msg) {
+			return nsec, len(msg), &Error{err: "overflow unpacking nsecx"}
+		}
+		window = int(msg[off])
+		length = int(msg[off+1])
+		off += 2
+		if window <= lastwindow {
+			// RFC 4034: Blocks are present in the NSEC RR RDATA in
+			// increasing numerical order.
+			return nsec, len(msg), &Error{err: "out of order NSEC block"}
+		}
+		if length == 0 {
+			// RFC 4034: Blocks with no types present MUST NOT be included.
+			return nsec, len(msg), &Error{err: "empty NSEC block"}
+		}
+		if length > 32 {
+			return nsec, len(msg), &Error{err: "NSEC block too long"}
+		}
+		if off+length > len(msg) {
+			return nsec, len(msg), &Error{err: "overflowing NSEC block"}
+		}
+
+		// Walk the bytes in the window and extract the type bits
+		for j := 0; j < length; j++ {
+			b := msg[off+j]
+			// Check the bits one by one, and set the type
+			if b&0x80 == 0x80 {
+				nsec = append(nsec, uint16(window*256+j*8+0))
+			}
+			if b&0x40 == 0x40 {
+				nsec = append(nsec, uint16(window*256+j*8+1))
+			}
+			if b&0x20 == 0x20 {
+				nsec = append(nsec, uint16(window*256+j*8+2))
+			}
+			if b&0x10 == 0x10 {
+				nsec = append(nsec, uint16(window*256+j*8+3))
+			}
+			if b&0x8 == 0x8 {
+				nsec = append(nsec, uint16(window*256+j*8+4))
+			}
+			if b&0x4 == 0x4 {
+				nsec = append(nsec, uint16(window*256+j*8+5))
+			}
+			if b&0x2 == 0x2 {
+				nsec = append(nsec, uint16(window*256+j*8+6))
+			}
+			if b&0x1 == 0x1 {
+				nsec = append(nsec, uint16(window*256+j*8+7))
+			}
+		}
+		off += length
+		lastwindow = window
+	}
+	return nsec, off, nil
+}
+
+func packDataNsec(bitmap []uint16, msg []byte, off int) (int, error) {
+	if len(bitmap) == 0 {
+		return off, nil
+	}
+	var lastwindow, lastlength uint16
+	for j := 0; j < len(bitmap); j++ {
+		t := bitmap[j]
+		window := t / 256
+		length := (t-window*256)/8 + 1
+		if window > lastwindow && lastlength != 0 { // New window, jump to the new offset
+			off += int(lastlength) + 2
+			lastlength = 0
+		}
+		if window < lastwindow || length < lastlength {
+			return len(msg), &Error{err: "nsec bits out of order"}
+		}
+		if off+2+int(length) > len(msg) {
+			return len(msg), &Error{err: "overflow packing nsec"}
+		}
+		// Setting the window #
+		msg[off] = byte(window)
+		// Setting the octets length
+		msg[off+1] = byte(length)
+		// Setting the bit value for the type in the right octet
+		msg[off+1+int(length)] |= byte(1 << (7 - (t % 8)))
+		lastwindow, lastlength = window, length
+	}
+	off += int(lastlength) + 2
+	return off, nil
+}
+
+func unpackDataDomainNames(msg []byte, off, end int) ([]string, int, error) {
+	var (
+		servers []string
+		s       string
+		err     error
+	)
+	if end > len(msg) {
+		return nil, len(msg), &Error{err: "overflow unpacking domain names"}
+	}
+	for off < end {
+		s, off, err = UnpackDomainName(msg, off)
+		if err != nil {
+			return servers, len(msg), err
+		}
+		servers = append(servers, s)
+	}
+	return servers, off, nil
+}
+
+func packDataDomainNames(names []string, msg []byte, off int, compression map[string]int, compress bool) (int, error) {
+	var err error
+	for j := 0; j < len(names); j++ {
+		off, err = PackDomainName(names[j], msg, off, compression, false && compress)
+		if err != nil {
+			return len(msg), err
+		}
+	}
+	return off, nil
+}
--- a/vendor/github.com/miekg/dns/nsecx.go
+++ b/vendor/github.com/miekg/dns/nsecx.go
@ -0,0 +1,106 @@
+package dns
+
+import (
+	"crypto/sha1"
+	"hash"
+	"strings"
+)
+
+type saltWireFmt struct {
+	Salt string `dns:"size-hex"`
+}
+
+// HashName hashes a string (label) according to RFC 5155. It returns the hashed string in uppercase.
+func HashName(label string, ha uint8, iter uint16, salt string) string {
+	saltwire := new(saltWireFmt)
+	saltwire.Salt = salt
+	wire := make([]byte, DefaultMsgSize)
+	n, err := packSaltWire(saltwire, wire)
+	if err != nil {
+		return ""
+	}
+	wire = wire[:n]
+	name := make([]byte, 255)
+	off, err := PackDomainName(strings.ToLower(label), name, 0, nil, false)
+	if err != nil {
+		return ""
+	}
+	name = name[:off]
+	var s hash.Hash
+	switch ha {
+	case SHA1:
+		s = sha1.New()
+	default:
+		return ""
+	}
+
+	// k = 0
+	s.Write(name)
+	s.Write(wire)
+	nsec3 := s.Sum(nil)
+	// k > 0
+	for k := uint16(0); k < iter; k++ {
+		s.Reset()
+		s.Write(nsec3)
+		s.Write(wire)
+		nsec3 = s.Sum(nsec3[:0])
+	}
+	return toBase32(nsec3)
+}
+
+// Cover returns true if a name is covered by the NSEC3 record
+func (rr *NSEC3) Cover(name string) bool {
+	nameHash := HashName(name, rr.Hash, rr.Iterations, rr.Salt)
+	owner := strings.ToUpper(rr.Hdr.Name)
+	labelIndices := Split(owner)
+	if len(labelIndices) < 2 {
+		return false
+	}
+	ownerHash := owner[:labelIndices[1]-1]
+	ownerZone := owner[labelIndices[1]:]
+	if !IsSubDomain(ownerZone, strings.ToUpper(name)) { // name is outside owner zone
+		return false
+	}
+
+	nextHash := rr.NextDomain
+	if ownerHash == nextHash { // empty interval
+		return false
+	}
+	if ownerHash > nextHash { // end of zone
+		if nameHash > ownerHash { // covered since there is nothing after ownerHash
+			return true
+		}
+		return nameHash < nextHash // if nameHash is before beginning of zone it is covered
+	}
+	if nameHash < ownerHash { // nameHash is before ownerHash, not covered
+		return false
+	}
+	return nameHash < nextHash // if nameHash is before nextHash is it covered (between ownerHash and nextHash)
+}
+
+// Match returns true if a name matches the NSEC3 record
+func (rr *NSEC3) Match(name string) bool {
+	nameHash := HashName(name, rr.Hash, rr.Iterations, rr.Salt)
+	owner := strings.ToUpper(rr.Hdr.Name)
+	labelIndices := Split(owner)
+	if len(labelIndices) < 2 {
+		return false
+	}
+	ownerHash := owner[:labelIndices[1]-1]
+	ownerZone := owner[labelIndices[1]:]
+	if !IsSubDomain(ownerZone, strings.ToUpper(name)) { // name is outside owner zone
+		return false
+	}
+	if ownerHash == nameHash {
+		return true
+	}
+	return false
+}
+
+func packSaltWire(sw *saltWireFmt, msg []byte) (int, error) {
+	off, err := packStringHex(sw.Salt, msg, 0)
+	if err != nil {
+		return off, err
+	}
+	return off, nil
+}
--- a/vendor/github.com/miekg/dns/privaterr.go
+++ b/vendor/github.com/miekg/dns/privaterr.go
@ -0,0 +1,149 @@
+package dns
+
+import (
+	"fmt"
+	"strings"
+)
+
+// PrivateRdata is an interface used for implementing "Private Use" RR types, see
+// RFC 6895. This allows one to experiment with new RR types, without requesting an
+// official type code. Also see dns.PrivateHandle and dns.PrivateHandleRemove.
+type PrivateRdata interface {
+	// String returns the text presentaton of the Rdata of the Private RR.
+	String() string
+	// Parse parses the Rdata of the private RR.
+	Parse([]string) error
+	// Pack is used when packing a private RR into a buffer.
+	Pack([]byte) (int, error)
+	// Unpack is used when unpacking a private RR from a buffer.
+	// TODO(miek): diff. signature than Pack, see edns0.go for instance.
+	Unpack([]byte) (int, error)
+	// Copy copies the Rdata.
+	Copy(PrivateRdata) error
+	// Len returns the length in octets of the Rdata.
+	Len() int
+}
+
+// PrivateRR represents an RR that uses a PrivateRdata user-defined type.
+// It mocks normal RRs and implements dns.RR interface.
+type PrivateRR struct {
+	Hdr  RR_Header
+	Data PrivateRdata
+}
+
+func mkPrivateRR(rrtype uint16) *PrivateRR {
+	// Panics if RR is not an instance of PrivateRR.
+	rrfunc, ok := TypeToRR[rrtype]
+	if !ok {
+		panic(fmt.Sprintf("dns: invalid operation with Private RR type %d", rrtype))
+	}
+
+	anyrr := rrfunc()
+	switch rr := anyrr.(type) {
+	case *PrivateRR:
+		return rr
+	}
+	panic(fmt.Sprintf("dns: RR is not a PrivateRR, TypeToRR[%d] generator returned %T", rrtype, anyrr))
+}
+
+// Header return the RR header of r.
+func (r *PrivateRR) Header() *RR_Header { return &r.Hdr }
+
+func (r *PrivateRR) String() string { return r.Hdr.String() + r.Data.String() }
+
+// Private len and copy parts to satisfy RR interface.
+func (r *PrivateRR) len() int { return r.Hdr.len() + r.Data.Len() }
+func (r *PrivateRR) copy() RR {
+	// make new RR like this:
+	rr := mkPrivateRR(r.Hdr.Rrtype)
+	newh := r.Hdr.copyHeader()
+	rr.Hdr = *newh
+
+	err := r.Data.Copy(rr.Data)
+	if err != nil {
+		panic("dns: got value that could not be used to copy Private rdata")
+	}
+	return rr
+}
+func (r *PrivateRR) pack(msg []byte, off int, compression map[string]int, compress bool) (int, error) {
+	off, err := r.Hdr.pack(msg, off, compression, compress)
+	if err != nil {
+		return off, err
+	}
+	headerEnd := off
+	n, err := r.Data.Pack(msg[off:])
+	if err != nil {
+		return len(msg), err
+	}
+	off += n
+	r.Header().Rdlength = uint16(off - headerEnd)
+	return off, nil
+}
+
+// PrivateHandle registers a private resource record type. It requires
+// string and numeric representation of private RR type and generator function as argument.
+func PrivateHandle(rtypestr string, rtype uint16, generator func() PrivateRdata) {
+	rtypestr = strings.ToUpper(rtypestr)
+
+	TypeToRR[rtype] = func() RR { return &PrivateRR{RR_Header{}, generator()} }
+	TypeToString[rtype] = rtypestr
+	StringToType[rtypestr] = rtype
+
+	typeToUnpack[rtype] = func(h RR_Header, msg []byte, off int) (RR, int, error) {
+		if noRdata(h) {
+			return &h, off, nil
+		}
+		var err error
+
+		rr := mkPrivateRR(h.Rrtype)
+		rr.Hdr = h
+
+		off1, err := rr.Data.Unpack(msg[off:])
+		off += off1
+		if err != nil {
+			return rr, off, err
+		}
+		return rr, off, err
+	}
+
+	setPrivateRR := func(h RR_Header, c chan lex, o, f string) (RR, *ParseError, string) {
+		rr := mkPrivateRR(h.Rrtype)
+		rr.Hdr = h
+
+		var l lex
+		text := make([]string, 0, 2) // could be 0..N elements, median is probably 1
+	Fetch:
+		for {
+			// TODO(miek): we could also be returning _QUOTE, this might or might not
+			// be an issue (basically parsing TXT becomes hard)
+			switch l = <-c; l.value {
+			case zNewline, zEOF:
+				break Fetch
+			case zString:
+				text = append(text, l.token)
+			}
+		}
+
+		err := rr.Data.Parse(text)
+		if err != nil {
+			return nil, &ParseError{f, err.Error(), l}, ""
+		}
+
+		return rr, nil, ""
+	}
+
+	typeToparserFunc[rtype] = parserFunc{setPrivateRR, true}
+}
+
+// PrivateHandleRemove removes defenitions required to support private RR type.
+func PrivateHandleRemove(rtype uint16) {
+	rtypestr, ok := TypeToString[rtype]
+	if ok {
+		delete(TypeToRR, rtype)
+		delete(TypeToString, rtype)
+		delete(typeToparserFunc, rtype)
+		delete(StringToType, rtypestr)
+		delete(typeToUnpack, rtype)
+	}
+	return
+}
--- a/vendor/github.com/miekg/dns/rawmsg.go
+++ b/vendor/github.com/miekg/dns/rawmsg.go
@ -0,0 +1,49 @@
+package dns
+
+import "encoding/binary"
+
+// rawSetRdlength sets the rdlength in the header of
+// the RR. The offset 'off' must be positioned at the
+// start of the header of the RR, 'end' must be the
+// end of the RR.
+func rawSetRdlength(msg []byte, off, end int) bool {
+	l := len(msg)
+Loop:
+	for {
+		if off+1 > l {
+			return false
+		}
+		c := int(msg[off])
+		off++
+		switch c & 0xC0 {
+		case 0x00:
+			if c == 0x00 {
+				// End of the domainname
+				break Loop
+			}
+			if off+c > l {
+				return false
+			}
+			off += c
+
+		case 0xC0:
+			// pointer, next byte included, ends domainname
+			off++
+			break Loop
+		}
+	}
+	// The domainname has been seen, we at the start of the fixed part in the header.
+	// Type is 2 bytes, class is 2 bytes, ttl 4 and then 2 bytes for the length.
+	off += 2 + 2 + 4
+	if off+2 > l {
+		return false
+	}
+	//off+1 is the end of the header, 'end' is the end of the rr
+	//so 'end' - 'off+2' is the length of the rdata
+	rdatalen := end - (off + 2)
+	if rdatalen > 0xFFFF {
+		return false
+	}
+	binary.BigEndian.PutUint16(msg[off:], uint16(rdatalen))
+	return true
+}
--- a/vendor/github.com/miekg/dns/reverse.go
+++ b/vendor/github.com/miekg/dns/reverse.go
@ -0,0 +1,38 @@
+package dns
+
+// StringToType is the reverse of TypeToString, needed for string parsing.
+var StringToType = reverseInt16(TypeToString)
+
+// StringToClass is the reverse of ClassToString, needed for string parsing.
+var StringToClass = reverseInt16(ClassToString)
+
+// StringToOpcode is a map of opcodes to strings.
+var StringToOpcode = reverseInt(OpcodeToString)
+
+// StringToRcode is a map of rcodes to strings.
+var StringToRcode = reverseInt(RcodeToString)
+
+// Reverse a map
+func reverseInt8(m map[uint8]string) map[string]uint8 {
+	n := make(map[string]uint8, len(m))
+	for u, s := range m {
+		n[s] = u
+	}
+	return n
+}
+
+func reverseInt16(m map[uint16]string) map[string]uint16 {
+	n := make(map[string]uint16, len(m))
+	for u, s := range m {
+		n[s] = u
+	}
+	return n
+}
+
+func reverseInt(m map[int]string) map[string]int {
+	n := make(map[string]int, len(m))
+	for u, s := range m {
+		n[s] = u
+	}
+	return n
+}
--- a/vendor/github.com/miekg/dns/sanitize.go
+++ b/vendor/github.com/miekg/dns/sanitize.go
@ -0,0 +1,84 @@
+package dns
+
+// Dedup removes identical RRs from rrs. It preserves the original ordering.
+// The lowest TTL of any duplicates is used in the remaining one. Dedup modifies
+// rrs.
+// m is used to store the RRs temporary. If it is nil a new map will be allocated.
+func Dedup(rrs []RR, m map[string]RR) []RR {
+	if m == nil {
+		m = make(map[string]RR)
+	}
+	// Save the keys, so we don't have to call normalizedString twice.
+	keys := make([]*string, 0, len(rrs))
+
+	for _, r := range rrs {
+		key := normalizedString(r)
+		keys = append(keys, &key)
+		if _, ok := m[key]; ok {
+			// Shortest TTL wins.
+			if m[key].Header().Ttl > r.Header().Ttl {
+				m[key].Header().Ttl = r.Header().Ttl
+			}
+			continue
+		}
+
+		m[key] = r
+	}
+	// If the length of the result map equals the amount of RRs we got,
+	// it means they were all different. We can then just return the original rrset.
+	if len(m) == len(rrs) {
+		return rrs
+	}
+
+	j := 0
+	for i, r := range rrs {
+		// If keys[i] lives in the map, we should copy and remove it.
+		if _, ok := m[*keys[i]]; ok {
+			delete(m, *keys[i])
+			rrs[j] = r
+			j++
+		}
+
+		if len(m) == 0 {
+			break
+		}
+	}
+
+	return rrs[:j]
+}
+
+// normalizedString returns a normalized string from r. The TTL
+// is removed and the domain name is lowercased. We go from this:
+// DomainName<TAB>TTL<TAB>CLASS<TAB>TYPE<TAB>RDATA to:
+// lowercasename<TAB>CLASS<TAB>TYPE...
+func normalizedString(r RR) string {
+	// A string Go DNS makes has: domainname<TAB>TTL<TAB>...
+	b := []byte(r.String())
+
+	// find the first non-escaped tab, then another, so we capture where the TTL lives.
+	esc := false
+	ttlStart, ttlEnd := 0, 0
+	for i := 0; i < len(b) && ttlEnd == 0; i++ {
+		switch {
+		case b[i] == '\\':
+			esc = !esc
+		case b[i] == '\t' && !esc:
+			if ttlStart == 0 {
+				ttlStart = i
+				continue
+			}
+			if ttlEnd == 0 {
+				ttlEnd = i
+			}
+		case b[i] >= 'A' && b[i] <= 'Z' && !esc:
+			b[i] += 32
+		default:
+			esc = false
+		}
+	}
+
+	// remove TTL.
+	copy(b[ttlStart:], b[ttlEnd:])
+	cut := ttlEnd - ttlStart
+	return string(b[:len(b)-cut])
+}
--- a/vendor/github.com/miekg/dns/scan.go
+++ b/vendor/github.com/miekg/dns/scan.go
--- a/vendor/github.com/miekg/dns/scan_rr.go
+++ b/vendor/github.com/miekg/dns/scan_rr.go
--- a/vendor/github.com/miekg/dns/scanner.go
+++ b/vendor/github.com/miekg/dns/scanner.go
@ -0,0 +1,56 @@
+package dns
+
+// Implement a simple scanner, return a byte stream from an io reader.
+
+import (
+	"bufio"
+	"context"
+	"io"
+	"text/scanner"
+)
+
+type scan struct {
+	src      *bufio.Reader
+	position scanner.Position
+	eof      bool // Have we just seen a eof
+	ctx      context.Context
+}
+
+func scanInit(r io.Reader) (*scan, context.CancelFunc) {
+	s := new(scan)
+	s.src = bufio.NewReader(r)
+	s.position.Line = 1
+
+	ctx, cancel := context.WithCancel(context.Background())
+	s.ctx = ctx
+
+	return s, cancel
+}
+
+// tokenText returns the next byte from the input
+func (s *scan) tokenText() (byte, error) {
+	c, err := s.src.ReadByte()
+	if err != nil {
+		return c, err
+	}
+	select {
+	case <-s.ctx.Done():
+		return c, context.Canceled
+	default:
+		break
+	}
+
+	// delay the newline handling until the next token is delivered,
+	// fixes off-by-one errors when reporting a parse error.
+	if s.eof == true {
+		s.position.Line++
+		s.position.Column = 0
+		s.eof = false
+	}
+	if c == '\n' {
+		s.eof = true
+		return c, nil
+	}
+	s.position.Column++
+	return c, nil
+}
--- a/vendor/github.com/miekg/dns/server.go
+++ b/vendor/github.com/miekg/dns/server.go
@ -0,0 +1,719 @@
+// DNS server implementation.
+
+package dns
+
+import (
+	"bytes"
+	"crypto/tls"
+	"encoding/binary"
+	"io"
+	"net"
+	"sync"
+	"time"
+)
+
+// Maximum number of TCP queries before we close the socket.
+const maxTCPQueries = 128
+
+// Handler is implemented by any value that implements ServeDNS.
+type Handler interface {
+	ServeDNS(w ResponseWriter, r *Msg)
+}
+
+// A ResponseWriter interface is used by an DNS handler to
+// construct an DNS response.
+type ResponseWriter interface {
+	// LocalAddr returns the net.Addr of the server
+	LocalAddr() net.Addr
+	// RemoteAddr returns the net.Addr of the client that sent the current request.
+	RemoteAddr() net.Addr
+	// WriteMsg writes a reply back to the client.
+	WriteMsg(*Msg) error
+	// Write writes a raw buffer back to the client.
+	Write([]byte) (int, error)
+	// Close closes the connection.
+	Close() error
+	// TsigStatus returns the status of the Tsig.
+	TsigStatus() error
+	// TsigTimersOnly sets the tsig timers only boolean.
+	TsigTimersOnly(bool)
+	// Hijack lets the caller take over the connection.
+	// After a call to Hijack(), the DNS package will not do anything with the connection.
+	Hijack()
+}
+
+type response struct {
+	hijacked       bool // connection has been hijacked by handler
+	tsigStatus     error
+	tsigTimersOnly bool
+	tsigRequestMAC string
+	tsigSecret     map[string]string // the tsig secrets
+	udp            *net.UDPConn      // i/o connection if UDP was used
+	tcp            net.Conn          // i/o connection if TCP was used
+	udpSession     *SessionUDP       // oob data to get egress interface right
+	remoteAddr     net.Addr          // address of the client
+	writer         Writer            // writer to output the raw DNS bits
+}
+
+// ServeMux is an DNS request multiplexer. It matches the
+// zone name of each incoming request against a list of
+// registered patterns add calls the handler for the pattern
+// that most closely matches the zone name. ServeMux is DNSSEC aware, meaning
+// that queries for the DS record are redirected to the parent zone (if that
+// is also registered), otherwise the child gets the query.
+// ServeMux is also safe for concurrent access from multiple goroutines.
+type ServeMux struct {
+	z map[string]Handler
+	m *sync.RWMutex
+}
+
+// NewServeMux allocates and returns a new ServeMux.
+func NewServeMux() *ServeMux { return &ServeMux{z: make(map[string]Handler), m: new(sync.RWMutex)} }
+
+// DefaultServeMux is the default ServeMux used by Serve.
+var DefaultServeMux = NewServeMux()
+
+// The HandlerFunc type is an adapter to allow the use of
+// ordinary functions as DNS handlers.  If f is a function
+// with the appropriate signature, HandlerFunc(f) is a
+// Handler object that calls f.
+type HandlerFunc func(ResponseWriter, *Msg)
+
+// ServeDNS calls f(w, r).
+func (f HandlerFunc) ServeDNS(w ResponseWriter, r *Msg) {
+	f(w, r)
+}
+
+// HandleFailed returns a HandlerFunc that returns SERVFAIL for every request it gets.
+func HandleFailed(w ResponseWriter, r *Msg) {
+	m := new(Msg)
+	m.SetRcode(r, RcodeServerFailure)
+	// does not matter if this write fails
+	w.WriteMsg(m)
+}
+
+func failedHandler() Handler { return HandlerFunc(HandleFailed) }
+
+// ListenAndServe Starts a server on address and network specified Invoke handler
+// for incoming queries.
+func ListenAndServe(addr string, network string, handler Handler) error {
+	server := &Server{Addr: addr, Net: network, Handler: handler}
+	return server.ListenAndServe()
+}
+
+// ListenAndServeTLS acts like http.ListenAndServeTLS, more information in
+// http://golang.org/pkg/net/http/#ListenAndServeTLS
+func ListenAndServeTLS(addr, certFile, keyFile string, handler Handler) error {
+	cert, err := tls.LoadX509KeyPair(certFile, keyFile)
+	if err != nil {
+		return err
+	}
+
+	config := tls.Config{
+		Certificates: []tls.Certificate{cert},
+	}
+
+	server := &Server{
+		Addr:      addr,
+		Net:       "tcp-tls",
+		TLSConfig: &config,
+		Handler:   handler,
+	}
+
+	return server.ListenAndServe()
+}
+
+// ActivateAndServe activates a server with a listener from systemd,
+// l and p should not both be non-nil.
+// If both l and p are not nil only p will be used.
+// Invoke handler for incoming queries.
+func ActivateAndServe(l net.Listener, p net.PacketConn, handler Handler) error {
+	server := &Server{Listener: l, PacketConn: p, Handler: handler}
+	return server.ActivateAndServe()
+}
+
+func (mux *ServeMux) match(q string, t uint16) Handler {
+	mux.m.RLock()
+	defer mux.m.RUnlock()
+	var handler Handler
+	b := make([]byte, len(q)) // worst case, one label of length q
+	off := 0
+	end := false
+	for {
+		l := len(q[off:])
+		for i := 0; i < l; i++ {
+			b[i] = q[off+i]
+			if b[i] >= 'A' && b[i] <= 'Z' {
+				b[i] |= ('a' - 'A')
+			}
+		}
+		if h, ok := mux.z[string(b[:l])]; ok { // causes garbage, might want to change the map key
+			if t != TypeDS {
+				return h
+			}
+			// Continue for DS to see if we have a parent too, if so delegeate to the parent
+			handler = h
+		}
+		off, end = NextLabel(q, off)
+		if end {
+			break
+		}
+	}
+	// Wildcard match, if we have found nothing try the root zone as a last resort.
+	if h, ok := mux.z["."]; ok {
+		return h
+	}
+	return handler
+}
+
+// Handle adds a handler to the ServeMux for pattern.
+func (mux *ServeMux) Handle(pattern string, handler Handler) {
+	if pattern == "" {
+		panic("dns: invalid pattern " + pattern)
+	}
+	mux.m.Lock()
+	mux.z[Fqdn(pattern)] = handler
+	mux.m.Unlock()
+}
+
+// HandleFunc adds a handler function to the ServeMux for pattern.
+func (mux *ServeMux) HandleFunc(pattern string, handler func(ResponseWriter, *Msg)) {
+	mux.Handle(pattern, HandlerFunc(handler))
+}
+
+// HandleRemove deregistrars the handler specific for pattern from the ServeMux.
+func (mux *ServeMux) HandleRemove(pattern string) {
+	if pattern == "" {
+		panic("dns: invalid pattern " + pattern)
+	}
+	mux.m.Lock()
+	delete(mux.z, Fqdn(pattern))
+	mux.m.Unlock()
+}
+
+// ServeDNS dispatches the request to the handler whose
+// pattern most closely matches the request message. If DefaultServeMux
+// is used the correct thing for DS queries is done: a possible parent
+// is sought.
+// If no handler is found a standard SERVFAIL message is returned
+// If the request message does not have exactly one question in the
+// question section a SERVFAIL is returned, unlesss Unsafe is true.
+func (mux *ServeMux) ServeDNS(w ResponseWriter, request *Msg) {
+	var h Handler
+	if len(request.Question) < 1 { // allow more than one question
+		h = failedHandler()
+	} else {
+		if h = mux.match(request.Question[0].Name, request.Question[0].Qtype); h == nil {
+			h = failedHandler()
+		}
+	}
+	h.ServeDNS(w, request)
+}
+
+// Handle registers the handler with the given pattern
+// in the DefaultServeMux. The documentation for
+// ServeMux explains how patterns are matched.
+func Handle(pattern string, handler Handler) { DefaultServeMux.Handle(pattern, handler) }
+
+// HandleRemove deregisters the handle with the given pattern
+// in the DefaultServeMux.
+func HandleRemove(pattern string) { DefaultServeMux.HandleRemove(pattern) }
+
+// HandleFunc registers the handler function with the given pattern
+// in the DefaultServeMux.
+func HandleFunc(pattern string, handler func(ResponseWriter, *Msg)) {
+	DefaultServeMux.HandleFunc(pattern, handler)
+}
+
+// Writer writes raw DNS messages; each call to Write should send an entire message.
+type Writer interface {
+	io.Writer
+}
+
+// Reader reads raw DNS messages; each call to ReadTCP or ReadUDP should return an entire message.
+type Reader interface {
+	// ReadTCP reads a raw message from a TCP connection. Implementations may alter
+	// connection properties, for example the read-deadline.
+	ReadTCP(conn net.Conn, timeout time.Duration) ([]byte, error)
+	// ReadUDP reads a raw message from a UDP connection. Implementations may alter
+	// connection properties, for example the read-deadline.
+	ReadUDP(conn *net.UDPConn, timeout time.Duration) ([]byte, *SessionUDP, error)
+}
+
+// defaultReader is an adapter for the Server struct that implements the Reader interface
+// using the readTCP and readUDP func of the embedded Server.
+type defaultReader struct {
+	*Server
+}
+
+func (dr *defaultReader) ReadTCP(conn net.Conn, timeout time.Duration) ([]byte, error) {
+	return dr.readTCP(conn, timeout)
+}
+
+func (dr *defaultReader) ReadUDP(conn *net.UDPConn, timeout time.Duration) ([]byte, *SessionUDP, error) {
+	return dr.readUDP(conn, timeout)
+}
+
+// DecorateReader is a decorator hook for extending or supplanting the functionality of a Reader.
+// Implementations should never return a nil Reader.
+type DecorateReader func(Reader) Reader
+
+// DecorateWriter is a decorator hook for extending or supplanting the functionality of a Writer.
+// Implementations should never return a nil Writer.
+type DecorateWriter func(Writer) Writer
+
+// A Server defines parameters for running an DNS server.
+type Server struct {
+	// Address to listen on, ":dns" if empty.
+	Addr string
+	// if "tcp" or "tcp-tls" (DNS over TLS) it will invoke a TCP listener, otherwise an UDP one
+	Net string
+	// TCP Listener to use, this is to aid in systemd's socket activation.
+	Listener net.Listener
+	// TLS connection configuration
+	TLSConfig *tls.Config
+	// UDP "Listener" to use, this is to aid in systemd's socket activation.
+	PacketConn net.PacketConn
+	// Handler to invoke, dns.DefaultServeMux if nil.
+	Handler Handler
+	// Default buffer size to use to read incoming UDP messages. If not set
+	// it defaults to MinMsgSize (512 B).
+	UDPSize int
+	// The net.Conn.SetReadTimeout value for new connections, defaults to 2 * time.Second.
+	ReadTimeout time.Duration
+	// The net.Conn.SetWriteTimeout value for new connections, defaults to 2 * time.Second.
+	WriteTimeout time.Duration
+	// TCP idle timeout for multiple queries, if nil, defaults to 8 * time.Second (RFC 5966).
+	IdleTimeout func() time.Duration
+	// Secret(s) for Tsig map[<zonename>]<base64 secret>. The zonename must be in canonical form (lowercase, fqdn, see RFC 4034 Section 6.2).
+	TsigSecret map[string]string
+	// Unsafe instructs the server to disregard any sanity checks and directly hand the message to
+	// the handler. It will specifically not check if the query has the QR bit not set.
+	Unsafe bool
+	// If NotifyStartedFunc is set it is called once the server has started listening.
+	NotifyStartedFunc func()
+	// DecorateReader is optional, allows customization of the process that reads raw DNS messages.
+	DecorateReader DecorateReader
+	// DecorateWriter is optional, allows customization of the process that writes raw DNS messages.
+	DecorateWriter DecorateWriter
+
+	// Shutdown handling
+	lock    sync.RWMutex
+	started bool
+}
+
+// ListenAndServe starts a nameserver on the configured address in *Server.
+func (srv *Server) ListenAndServe() error {
+	srv.lock.Lock()
+	defer srv.lock.Unlock()
+	if srv.started {
+		return &Error{err: "server already started"}
+	}
+	addr := srv.Addr
+	if addr == "" {
+		addr = ":domain"
+	}
+	if srv.UDPSize == 0 {
+		srv.UDPSize = MinMsgSize
+	}
+	switch srv.Net {
+	case "tcp", "tcp4", "tcp6":
+		a, err := net.ResolveTCPAddr(srv.Net, addr)
+		if err != nil {
+			return err
+		}
+		l, err := net.ListenTCP(srv.Net, a)
+		if err != nil {
+			return err
+		}
+		srv.Listener = l
+		srv.started = true
+		srv.lock.Unlock()
+		err = srv.serveTCP(l)
+		srv.lock.Lock() // to satisfy the defer at the top
+		return err
+	case "tcp-tls", "tcp4-tls", "tcp6-tls":
+		network := "tcp"
+		if srv.Net == "tcp4-tls" {
+			network = "tcp4"
+		} else if srv.Net == "tcp6-tls" {
+			network = "tcp6"
+		}
+
+		l, err := tls.Listen(network, addr, srv.TLSConfig)
+		if err != nil {
+			return err
+		}
+		srv.Listener = l
+		srv.started = true
+		srv.lock.Unlock()
+		err = srv.serveTCP(l)
+		srv.lock.Lock() // to satisfy the defer at the top
+		return err
+	case "udp", "udp4", "udp6":
+		a, err := net.ResolveUDPAddr(srv.Net, addr)
+		if err != nil {
+			return err
+		}
+		l, err := net.ListenUDP(srv.Net, a)
+		if err != nil {
+			return err
+		}
+		if e := setUDPSocketOptions(l); e != nil {
+			return e
+		}
+		srv.PacketConn = l
+		srv.started = true
+		srv.lock.Unlock()
+		err = srv.serveUDP(l)
+		srv.lock.Lock() // to satisfy the defer at the top
+		return err
+	}
+	return &Error{err: "bad network"}
+}
+
+// ActivateAndServe starts a nameserver with the PacketConn or Listener
+// configured in *Server. Its main use is to start a server from systemd.
+func (srv *Server) ActivateAndServe() error {
+	srv.lock.Lock()
+	defer srv.lock.Unlock()
+	if srv.started {
+		return &Error{err: "server already started"}
+	}
+	pConn := srv.PacketConn
+	l := srv.Listener
+	if pConn != nil {
+		if srv.UDPSize == 0 {
+			srv.UDPSize = MinMsgSize
+		}
+		// Check PacketConn interface's type is valid and value
+		// is not nil
+		if t, ok := pConn.(*net.UDPConn); ok && t != nil {
+			if e := setUDPSocketOptions(t); e != nil {
+				return e
+			}
+			srv.started = true
+			srv.lock.Unlock()
+			e := srv.serveUDP(t)
+			srv.lock.Lock() // to satisfy the defer at the top
+			return e
+		}
+	}
+	if l != nil {
+		srv.started = true
+		srv.lock.Unlock()
+		e := srv.serveTCP(l)
+		srv.lock.Lock() // to satisfy the defer at the top
+		return e
+	}
+	return &Error{err: "bad listeners"}
+}
+
+// Shutdown shuts down a server. After a call to Shutdown, ListenAndServe and
+// ActivateAndServe will return.
+func (srv *Server) Shutdown() error {
+	srv.lock.Lock()
+	if !srv.started {
+		srv.lock.Unlock()
+		return &Error{err: "server not started"}
+	}
+	srv.started = false
+	srv.lock.Unlock()
+
+	if srv.PacketConn != nil {
+		srv.PacketConn.Close()
+	}
+	if srv.Listener != nil {
+		srv.Listener.Close()
+	}
+	return nil
+}
+
+// getReadTimeout is a helper func to use system timeout if server did not intend to change it.
+func (srv *Server) getReadTimeout() time.Duration {
+	rtimeout := dnsTimeout
+	if srv.ReadTimeout != 0 {
+		rtimeout = srv.ReadTimeout
+	}
+	return rtimeout
+}
+
+// serveTCP starts a TCP listener for the server.
+// Each request is handled in a separate goroutine.
+func (srv *Server) serveTCP(l net.Listener) error {
+	defer l.Close()
+
+	if srv.NotifyStartedFunc != nil {
+		srv.NotifyStartedFunc()
+	}
+
+	reader := Reader(&defaultReader{srv})
+	if srv.DecorateReader != nil {
+		reader = srv.DecorateReader(reader)
+	}
+
+	handler := srv.Handler
+	if handler == nil {
+		handler = DefaultServeMux
+	}
+	rtimeout := srv.getReadTimeout()
+	// deadline is not used here
+	for {
+		rw, err := l.Accept()
+		srv.lock.RLock()
+		if !srv.started {
+			srv.lock.RUnlock()
+			return nil
+		}
+		srv.lock.RUnlock()
+		if err != nil {
+			if neterr, ok := err.(net.Error); ok && neterr.Temporary() {
+				continue
+			}
+			return err
+		}
+		go func() {
+			m, err := reader.ReadTCP(rw, rtimeout)
+			if err != nil {
+				rw.Close()
+				return
+			}
+			srv.serve(rw.RemoteAddr(), handler, m, nil, nil, rw)
+		}()
+	}
+}
+
+// serveUDP starts a UDP listener for the server.
+// Each request is handled in a separate goroutine.
+func (srv *Server) serveUDP(l *net.UDPConn) error {
+	defer l.Close()
+
+	if srv.NotifyStartedFunc != nil {
+		srv.NotifyStartedFunc()
+	}
+
+	reader := Reader(&defaultReader{srv})
+	if srv.DecorateReader != nil {
+		reader = srv.DecorateReader(reader)
+	}
+
+	handler := srv.Handler
+	if handler == nil {
+		handler = DefaultServeMux
+	}
+	rtimeout := srv.getReadTimeout()
+	// deadline is not used here
+	for {
+		m, s, err := reader.ReadUDP(l, rtimeout)
+		srv.lock.RLock()
+		if !srv.started {
+			srv.lock.RUnlock()
+			return nil
+		}
+		srv.lock.RUnlock()
+		if err != nil {
+			if netErr, ok := err.(net.Error); ok && netErr.Temporary() {
+				continue
+			}
+			return err
+		}
+		if len(m) < headerSize {
+			continue
+		}
+		go srv.serve(s.RemoteAddr(), handler, m, l, s, nil)
+	}
+}
+
+// Serve a new connection.
+func (srv *Server) serve(a net.Addr, h Handler, m []byte, u *net.UDPConn, s *SessionUDP, t net.Conn) {
+	w := &response{tsigSecret: srv.TsigSecret, udp: u, tcp: t, remoteAddr: a, udpSession: s}
+	if srv.DecorateWriter != nil {
+		w.writer = srv.DecorateWriter(w)
+	} else {
+		w.writer = w
+	}
+
+	q := 0 // counter for the amount of TCP queries we get
+
+	reader := Reader(&defaultReader{srv})
+	if srv.DecorateReader != nil {
+		reader = srv.DecorateReader(reader)
+	}
+Redo:
+	req := new(Msg)
+	err := req.Unpack(m)
+	if err != nil { // Send a FormatError back
+		x := new(Msg)
+		x.SetRcodeFormatError(req)
+		w.WriteMsg(x)
+		goto Exit
+	}
+	if !srv.Unsafe && req.Response {
+		goto Exit
+	}
+
+	w.tsigStatus = nil
+	if w.tsigSecret != nil {
+		if t := req.IsTsig(); t != nil {
+			secret := t.Hdr.Name
+			if _, ok := w.tsigSecret[secret]; !ok {
+				w.tsigStatus = ErrKeyAlg
+			}
+			w.tsigStatus = TsigVerify(m, w.tsigSecret[secret], "", false)
+			w.tsigTimersOnly = false
+			w.tsigRequestMAC = req.Extra[len(req.Extra)-1].(*TSIG).MAC
+		}
+	}
+	h.ServeDNS(w, req) // Writes back to the client
+
+Exit:
+	if w.tcp == nil {
+		return
+	}
+	// TODO(miek): make this number configurable?
+	if q > maxTCPQueries { // close socket after this many queries
+		w.Close()
+		return
+	}
+
+	if w.hijacked {
+		return // client calls Close()
+	}
+	if u != nil { // UDP, "close" and return
+		w.Close()
+		return
+	}
+	idleTimeout := tcpIdleTimeout
+	if srv.IdleTimeout != nil {
+		idleTimeout = srv.IdleTimeout()
+	}
+	m, err = reader.ReadTCP(w.tcp, idleTimeout)
+	if err == nil {
+		q++
+		goto Redo
+	}
+	w.Close()
+	return
+}
+
+func (srv *Server) readTCP(conn net.Conn, timeout time.Duration) ([]byte, error) {
+	conn.SetReadDeadline(time.Now().Add(timeout))
+	l := make([]byte, 2)
+	n, err := conn.Read(l)
+	if err != nil || n != 2 {
+		if err != nil {
+			return nil, err
+		}
+		return nil, ErrShortRead
+	}
+	length := binary.BigEndian.Uint16(l)
+	if length == 0 {
+		return nil, ErrShortRead
+	}
+	m := make([]byte, int(length))
+	n, err = conn.Read(m[:int(length)])
+	if err != nil || n == 0 {
+		if err != nil {
+			return nil, err
+		}
+		return nil, ErrShortRead
+	}
+	i := n
+	for i < int(length) {
+		j, err := conn.Read(m[i:int(length)])
+		if err != nil {
+			return nil, err
+		}
+		i += j
+	}
+	n = i
+	m = m[:n]
+	return m, nil
+}
+
+func (srv *Server) readUDP(conn *net.UDPConn, timeout time.Duration) ([]byte, *SessionUDP, error) {
+	conn.SetReadDeadline(time.Now().Add(timeout))
+	m := make([]byte, srv.UDPSize)
+	n, s, err := ReadFromSessionUDP(conn, m)
+	if err != nil {
+		return nil, nil, err
+	}
+	m = m[:n]
+	return m, s, nil
+}
+
+// WriteMsg implements the ResponseWriter.WriteMsg method.
+func (w *response) WriteMsg(m *Msg) (err error) {
+	var data []byte
+	if w.tsigSecret != nil { // if no secrets, dont check for the tsig (which is a longer check)
+		if t := m.IsTsig(); t != nil {
+			data, w.tsigRequestMAC, err = TsigGenerate(m, w.tsigSecret[t.Hdr.Name], w.tsigRequestMAC, w.tsigTimersOnly)
+			if err != nil {
+				return err
+			}
+			_, err = w.writer.Write(data)
+			return err
+		}
+	}
+	data, err = m.Pack()
+	if err != nil {
+		return err
+	}
+	_, err = w.writer.Write(data)
+	return err
+}
+
+// Write implements the ResponseWriter.Write method.
+func (w *response) Write(m []byte) (int, error) {
+	switch {
+	case w.udp != nil:
+		n, err := WriteToSessionUDP(w.udp, m, w.udpSession)
+		return n, err
+	case w.tcp != nil:
+		lm := len(m)
+		if lm < 2 {
+			return 0, io.ErrShortBuffer
+		}
+		if lm > MaxMsgSize {
+			return 0, &Error{err: "message too large"}
+		}
+		l := make([]byte, 2, 2+lm)
+		binary.BigEndian.PutUint16(l, uint16(lm))
+		m = append(l, m...)
+
+		n, err := io.Copy(w.tcp, bytes.NewReader(m))
+		return int(n), err
+	}
+	panic("not reached")
+}
+
+// LocalAddr implements the ResponseWriter.LocalAddr method.
+func (w *response) LocalAddr() net.Addr {
+	if w.tcp != nil {
+		return w.tcp.LocalAddr()
+	}
+	return w.udp.LocalAddr()
+}
+
+// RemoteAddr implements the ResponseWriter.RemoteAddr method.
+func (w *response) RemoteAddr() net.Addr { return w.remoteAddr }
+
+// TsigStatus implements the ResponseWriter.TsigStatus method.
+func (w *response) TsigStatus() error { return w.tsigStatus }
+
+// TsigTimersOnly implements the ResponseWriter.TsigTimersOnly method.
+func (w *response) TsigTimersOnly(b bool) { w.tsigTimersOnly = b }
+
+// Hijack implements the ResponseWriter.Hijack method.
+func (w *response) Hijack() { w.hijacked = true }
+
+// Close implements the ResponseWriter.Close method
+func (w *response) Close() error {
+	// Can't close the udp conn, as that is actually the listener.
+	if w.tcp != nil {
+		e := w.tcp.Close()
+		w.tcp = nil
+		return e
+	}
+	return nil
+}
--- a/vendor/github.com/miekg/dns/sig0.go
+++ b/vendor/github.com/miekg/dns/sig0.go
@ -0,0 +1,218 @@
+package dns
+
+import (
+	"crypto"
+	"crypto/dsa"
+	"crypto/ecdsa"
+	"crypto/rsa"
+	"encoding/binary"
+	"math/big"
+	"strings"
+	"time"
+)
+
+// Sign signs a dns.Msg. It fills the signature with the appropriate data.
+// The SIG record should have the SignerName, KeyTag, Algorithm, Inception
+// and Expiration set.
+func (rr *SIG) Sign(k crypto.Signer, m *Msg) ([]byte, error) {
+	if k == nil {
+		return nil, ErrPrivKey
+	}
+	if rr.KeyTag == 0 || len(rr.SignerName) == 0 || rr.Algorithm == 0 {
+		return nil, ErrKey
+	}
+	rr.Header().Rrtype = TypeSIG
+	rr.Header().Class = ClassANY
+	rr.Header().Ttl = 0
+	rr.Header().Name = "."
+	rr.OrigTtl = 0
+	rr.TypeCovered = 0
+	rr.Labels = 0
+
+	buf := make([]byte, m.Len()+rr.len())
+	mbuf, err := m.PackBuffer(buf)
+	if err != nil {
+		return nil, err
+	}
+	if &buf[0] != &mbuf[0] {
+		return nil, ErrBuf
+	}
+	off, err := PackRR(rr, buf, len(mbuf), nil, false)
+	if err != nil {
+		return nil, err
+	}
+	buf = buf[:off:cap(buf)]
+
+	hash, ok := AlgorithmToHash[rr.Algorithm]
+	if !ok {
+		return nil, ErrAlg
+	}
+
+	hasher := hash.New()
+	// Write SIG rdata
+	hasher.Write(buf[len(mbuf)+1+2+2+4+2:])
+	// Write message
+	hasher.Write(buf[:len(mbuf)])
+
+	signature, err := sign(k, hasher.Sum(nil), hash, rr.Algorithm)
+	if err != nil {
+		return nil, err
+	}
+
+	rr.Signature = toBase64(signature)
+
+	buf = append(buf, signature...)
+	if len(buf) > int(^uint16(0)) {
+		return nil, ErrBuf
+	}
+	// Adjust sig data length
+	rdoff := len(mbuf) + 1 + 2 + 2 + 4
+	rdlen := binary.BigEndian.Uint16(buf[rdoff:])
+	rdlen += uint16(len(signature))
+	binary.BigEndian.PutUint16(buf[rdoff:], rdlen)
+	// Adjust additional count
+	adc := binary.BigEndian.Uint16(buf[10:])
+	adc++
+	binary.BigEndian.PutUint16(buf[10:], adc)
+	return buf, nil
+}
+
+// Verify validates the message buf using the key k.
+// It's assumed that buf is a valid message from which rr was unpacked.
+func (rr *SIG) Verify(k *KEY, buf []byte) error {
+	if k == nil {
+		return ErrKey
+	}
+	if rr.KeyTag == 0 || len(rr.SignerName) == 0 || rr.Algorithm == 0 {
+		return ErrKey
+	}
+
+	var hash crypto.Hash
+	switch rr.Algorithm {
+	case DSA, RSASHA1:
+		hash = crypto.SHA1
+	case RSASHA256, ECDSAP256SHA256:
+		hash = crypto.SHA256
+	case ECDSAP384SHA384:
+		hash = crypto.SHA384
+	case RSASHA512:
+		hash = crypto.SHA512
+	default:
+		return ErrAlg
+	}
+	hasher := hash.New()
+
+	buflen := len(buf)
+	qdc := binary.BigEndian.Uint16(buf[4:])
+	anc := binary.BigEndian.Uint16(buf[6:])
+	auc := binary.BigEndian.Uint16(buf[8:])
+	adc := binary.BigEndian.Uint16(buf[10:])
+	offset := 12
+	var err error
+	for i := uint16(0); i < qdc && offset < buflen; i++ {
+		_, offset, err = UnpackDomainName(buf, offset)
+		if err != nil {
+			return err
+		}
+		// Skip past Type and Class
+		offset += 2 + 2
+	}
+	for i := uint16(1); i < anc+auc+adc && offset < buflen; i++ {
+		_, offset, err = UnpackDomainName(buf, offset)
+		if err != nil {
+			return err
+		}
+		// Skip past Type, Class and TTL
+		offset += 2 + 2 + 4
+		if offset+1 >= buflen {
+			continue
+		}
+		var rdlen uint16
+		rdlen = binary.BigEndian.Uint16(buf[offset:])
+		offset += 2
+		offset += int(rdlen)
+	}
+	if offset >= buflen {
+		return &Error{err: "overflowing unpacking signed message"}
+	}
+
+	// offset should be just prior to SIG
+	bodyend := offset
+	// owner name SHOULD be root
+	_, offset, err = UnpackDomainName(buf, offset)
+	if err != nil {
+		return err
+	}
+	// Skip Type, Class, TTL, RDLen
+	offset += 2 + 2 + 4 + 2
+	sigstart := offset
+	// Skip Type Covered, Algorithm, Labels, Original TTL
+	offset += 2 + 1 + 1 + 4
+	if offset+4+4 >= buflen {
+		return &Error{err: "overflow unpacking signed message"}
+	}
+	expire := binary.BigEndian.Uint32(buf[offset:])
+	offset += 4
+	incept := binary.BigEndian.Uint32(buf[offset:])
+	offset += 4
+	now := uint32(time.Now().Unix())
+	if now < incept || now > expire {
+		return ErrTime
+	}
+	// Skip key tag
+	offset += 2
+	var signername string
+	signername, offset, err = UnpackDomainName(buf, offset)
+	if err != nil {
+		return err
+	}
+	// If key has come from the DNS name compression might
+	// have mangled the case of the name
+	if strings.ToLower(signername) != strings.ToLower(k.Header().Name) {
+		return &Error{err: "signer name doesn't match key name"}
+	}
+	sigend := offset
+	hasher.Write(buf[sigstart:sigend])
+	hasher.Write(buf[:10])
+	hasher.Write([]byte{
+		byte((adc - 1) << 8),
+		byte(adc - 1),
+	})
+	hasher.Write(buf[12:bodyend])
+
+	hashed := hasher.Sum(nil)
+	sig := buf[sigend:]
+	switch k.Algorithm {
+	case DSA:
+		pk := k.publicKeyDSA()
+		sig = sig[1:]
+		r := big.NewInt(0)
+		r.SetBytes(sig[:len(sig)/2])
+		s := big.NewInt(0)
+		s.SetBytes(sig[len(sig)/2:])
+		if pk != nil {
+			if dsa.Verify(pk, hashed, r, s) {
+				return nil
+			}
+			return ErrSig
+		}
+	case RSASHA1, RSASHA256, RSASHA512:
+		pk := k.publicKeyRSA()
+		if pk != nil {
+			return rsa.VerifyPKCS1v15(pk, hash, hashed, sig)
+		}
+	case ECDSAP256SHA256, ECDSAP384SHA384:
+		pk := k.publicKeyECDSA()
+		r := big.NewInt(0)
+		r.SetBytes(sig[:len(sig)/2])
+		s := big.NewInt(0)
+		s.SetBytes(sig[len(sig)/2:])
+		if pk != nil {
+			if ecdsa.Verify(pk, hashed, r, s) {
+				return nil
+			}
+			return ErrSig
+		}
+	}
+	return ErrKeyAlg
+}
--- a/vendor/github.com/miekg/dns/singleinflight.go
+++ b/vendor/github.com/miekg/dns/singleinflight.go
@ -0,0 +1,57 @@
+// Copyright 2013 The Go Authors.  All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Adapted for dns package usage by Miek Gieben.
+
+package dns
+
+import "sync"
+import "time"
+
+// call is an in-flight or completed singleflight.Do call
+type call struct {
+	wg   sync.WaitGroup
+	val  *Msg
+	rtt  time.Duration
+	err  error
+	dups int
+}
+
+// singleflight represents a class of work and forms a namespace in
+// which units of work can be executed with duplicate suppression.
+type singleflight struct {
+	sync.Mutex                  // protects m
+	m          map[string]*call // lazily initialized
+}
+
+// Do executes and returns the results of the given function, making
+// sure that only one execution is in-flight for a given key at a
+// time. If a duplicate comes in, the duplicate caller waits for the
+// original to complete and receives the same results.
+// The return value shared indicates whether v was given to multiple callers.
+func (g *singleflight) Do(key string, fn func() (*Msg, time.Duration, error)) (v *Msg, rtt time.Duration, err error, shared bool) {
+	g.Lock()
+	if g.m == nil {
+		g.m = make(map[string]*call)
+	}
+	if c, ok := g.m[key]; ok {
+		c.dups++
+		g.Unlock()
+		c.wg.Wait()
+		return c.val, c.rtt, c.err, true
+	}
+	c := new(call)
+	c.wg.Add(1)
+	g.m[key] = c
+	g.Unlock()
+
+	c.val, c.rtt, c.err = fn()
+	c.wg.Done()
+
+	g.Lock()
+	delete(g.m, key)
+	g.Unlock()
+
+	return c.val, c.rtt, c.err, c.dups > 0
+}
--- a/vendor/github.com/miekg/dns/smimea.go
+++ b/vendor/github.com/miekg/dns/smimea.go
@ -0,0 +1,47 @@
+package dns
+
+import (
+	"crypto/sha256"
+	"crypto/x509"
+	"encoding/hex"
+)
+
+// Sign creates a SMIMEA record from an SSL certificate.
+func (r *SMIMEA) Sign(usage, selector, matchingType int, cert *x509.Certificate) (err error) {
+	r.Hdr.Rrtype = TypeSMIMEA
+	r.Usage = uint8(usage)
+	r.Selector = uint8(selector)
+	r.MatchingType = uint8(matchingType)
+
+	r.Certificate, err = CertificateToDANE(r.Selector, r.MatchingType, cert)
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+// Verify verifies a SMIMEA record against an SSL certificate. If it is OK
+// a nil error is returned.
+func (r *SMIMEA) Verify(cert *x509.Certificate) error {
+	c, err := CertificateToDANE(r.Selector, r.MatchingType, cert)
+	if err != nil {
+		return err // Not also ErrSig?
+	}
+	if r.Certificate == c {
+		return nil
+	}
+	return ErrSig // ErrSig, really?
+}
+
+// SMIMEAName returns the ownername of a SMIMEA resource record as per the
+// format specified in RFC 'draft-ietf-dane-smime-12' Section 2 and 3
+func SMIMEAName(email, domain string) (string, error) {
+	hasher := sha256.New()
+	hasher.Write([]byte(email))
+
+	// RFC Section 3: "The local-part is hashed using the SHA2-256
+	// algorithm with the hash truncated to 28 octets and
+	// represented in its hexadecimal representation to become the
+	// left-most label in the prepared domain name"
+	return hex.EncodeToString(hasher.Sum(nil)[:28]) + "." + "_smimecert." + domain, nil
+}
--- a/vendor/github.com/miekg/dns/tlsa.go
+++ b/vendor/github.com/miekg/dns/tlsa.go
@ -0,0 +1,47 @@
+package dns
+
+import (
+	"crypto/x509"
+	"net"
+	"strconv"
+)
+
+// Sign creates a TLSA record from an SSL certificate.
+func (r *TLSA) Sign(usage, selector, matchingType int, cert *x509.Certificate) (err error) {
+	r.Hdr.Rrtype = TypeTLSA
+	r.Usage = uint8(usage)
+	r.Selector = uint8(selector)
+	r.MatchingType = uint8(matchingType)
+
+	r.Certificate, err = CertificateToDANE(r.Selector, r.MatchingType, cert)
+	if err != nil {
+		return err
+	}
+	return nil
+}
+
+// Verify verifies a TLSA record against an SSL certificate. If it is OK
+// a nil error is returned.
+func (r *TLSA) Verify(cert *x509.Certificate) error {
+	c, err := CertificateToDANE(r.Selector, r.MatchingType, cert)
+	if err != nil {
+		return err // Not also ErrSig?
+	}
+	if r.Certificate == c {
+		return nil
+	}
+	return ErrSig // ErrSig, really?
+}
+
+// TLSAName returns the ownername of a TLSA resource record as per the
+// rules specified in RFC 6698, Section 3.
+func TLSAName(name, service, network string) (string, error) {
+	if !IsFqdn(name) {
+		return "", ErrFqdn
+	}
+	p, err := net.LookupPort(network, service)
+	if err != nil {
+		return "", err
+	}
+	return "_" + strconv.Itoa(p) + "._" + network + "." + name, nil
+}
--- a/vendor/github.com/miekg/dns/tsig.go
+++ b/vendor/github.com/miekg/dns/tsig.go
@ -0,0 +1,386 @@
+package dns
+
+import (
+	"crypto/hmac"
+	"crypto/md5"
+	"crypto/sha1"
+	"crypto/sha256"
+	"crypto/sha512"
+	"encoding/binary"
+	"encoding/hex"
+	"hash"
+	"strconv"
+	"strings"
+	"time"
+)
+
+// HMAC hashing codes. These are transmitted as domain names.
+const (
+	HmacMD5    = "hmac-md5.sig-alg.reg.int."
+	HmacSHA1   = "hmac-sha1."
+	HmacSHA256 = "hmac-sha256."
+	HmacSHA512 = "hmac-sha512."
+)
+
+// TSIG is the RR the holds the transaction signature of a message.
+// See RFC 2845 and RFC 4635.
+type TSIG struct {
+	Hdr        RR_Header
+	Algorithm  string `dns:"domain-name"`
+	TimeSigned uint64 `dns:"uint48"`
+	Fudge      uint16
+	MACSize    uint16
+	MAC        string `dns:"size-hex:MACSize"`
+	OrigId     uint16
+	Error      uint16
+	OtherLen   uint16
+	OtherData  string `dns:"size-hex:OtherLen"`
+}
+
+// TSIG has no official presentation format, but this will suffice.
+
+func (rr *TSIG) String() string {
+	s := "\n;; TSIG PSEUDOSECTION:\n"
+	s += rr.Hdr.String() +
+		" " + rr.Algorithm +
+		" " + tsigTimeToString(rr.TimeSigned) +
+		" " + strconv.Itoa(int(rr.Fudge)) +
+		" " + strconv.Itoa(int(rr.MACSize)) +
+		" " + strings.ToUpper(rr.MAC) +
+		" " + strconv.Itoa(int(rr.OrigId)) +
+		" " + strconv.Itoa(int(rr.Error)) + // BIND prints NOERROR
+		" " + strconv.Itoa(int(rr.OtherLen)) +
+		" " + rr.OtherData
+	return s
+}
+
+// The following values must be put in wireformat, so that the MAC can be calculated.
+// RFC 2845, section 3.4.2. TSIG Variables.
+type tsigWireFmt struct {
+	// From RR_Header
+	Name  string `dns:"domain-name"`
+	Class uint16
+	Ttl   uint32
+	// Rdata of the TSIG
+	Algorithm  string `dns:"domain-name"`
+	TimeSigned uint64 `dns:"uint48"`
+	Fudge      uint16
+	// MACSize, MAC and OrigId excluded
+	Error     uint16
+	OtherLen  uint16
+	OtherData string `dns:"size-hex:OtherLen"`
+}
+
+// If we have the MAC use this type to convert it to wiredata. Section 3.4.3. Request MAC
+type macWireFmt struct {
+	MACSize uint16
+	MAC     string `dns:"size-hex:MACSize"`
+}
+
+// 3.3. Time values used in TSIG calculations
+type timerWireFmt struct {
+	TimeSigned uint64 `dns:"uint48"`
+	Fudge      uint16
+}
+
+// TsigGenerate fills out the TSIG record attached to the message.
+// The message should contain
+// a "stub" TSIG RR with the algorithm, key name (owner name of the RR),
+// time fudge (defaults to 300 seconds) and the current time
+// The TSIG MAC is saved in that Tsig RR.
+// When TsigGenerate is called for the first time requestMAC is set to the empty string and
+// timersOnly is false.
+// If something goes wrong an error is returned, otherwise it is nil.
+func TsigGenerate(m *Msg, secret, requestMAC string, timersOnly bool) ([]byte, string, error) {
+	if m.IsTsig() == nil {
+		panic("dns: TSIG not last RR in additional")
+	}
+	// If we barf here, the caller is to blame
+	rawsecret, err := fromBase64([]byte(secret))
+	if err != nil {
+		return nil, "", err
+	}
+
+	rr := m.Extra[len(m.Extra)-1].(*TSIG)
+	m.Extra = m.Extra[0 : len(m.Extra)-1] // kill the TSIG from the msg
+	mbuf, err := m.Pack()
+	if err != nil {
+		return nil, "", err
+	}
+	buf := tsigBuffer(mbuf, rr, requestMAC, timersOnly)
+
+	t := new(TSIG)
+	var h hash.Hash
+	switch strings.ToLower(rr.Algorithm) {
+	case HmacMD5:
+		h = hmac.New(md5.New, []byte(rawsecret))
+	case HmacSHA1:
+		h = hmac.New(sha1.New, []byte(rawsecret))
+	case HmacSHA256:
+		h = hmac.New(sha256.New, []byte(rawsecret))
+	case HmacSHA512:
+		h = hmac.New(sha512.New, []byte(rawsecret))
+	default:
+		return nil, "", ErrKeyAlg
+	}
+	h.Write(buf)
+	t.MAC = hex.EncodeToString(h.Sum(nil))
+	t.MACSize = uint16(len(t.MAC) / 2) // Size is half!
+
+	t.Hdr = RR_Header{Name: rr.Hdr.Name, Rrtype: TypeTSIG, Class: ClassANY, Ttl: 0}
+	t.Fudge = rr.Fudge
+	t.TimeSigned = rr.TimeSigned
+	t.Algorithm = rr.Algorithm
+	t.OrigId = m.Id
+
+	tbuf := make([]byte, t.len())
+	if off, err := PackRR(t, tbuf, 0, nil, false); err == nil {
+		tbuf = tbuf[:off] // reset to actual size used
+	} else {
+		return nil, "", err
+	}
+	mbuf = append(mbuf, tbuf...)
+	// Update the ArCount directly in the buffer.
+	binary.BigEndian.PutUint16(mbuf[10:], uint16(len(m.Extra)+1))
+
+	return mbuf, t.MAC, nil
+}
+
+// TsigVerify verifies the TSIG on a message.
+// If the signature does not validate err contains the
+// error, otherwise it is nil.
+func TsigVerify(msg []byte, secret, requestMAC string, timersOnly bool) error {
+	rawsecret, err := fromBase64([]byte(secret))
+	if err != nil {
+		return err
+	}
+	// Strip the TSIG from the incoming msg
+	stripped, tsig, err := stripTsig(msg)
+	if err != nil {
+		return err
+	}
+
+	msgMAC, err := hex.DecodeString(tsig.MAC)
+	if err != nil {
+		return err
+	}
+
+	buf := tsigBuffer(stripped, tsig, requestMAC, timersOnly)
+
+	// Fudge factor works both ways. A message can arrive before it was signed because
+	// of clock skew.
+	now := uint64(time.Now().Unix())
+	ti := now - tsig.TimeSigned
+	if now < tsig.TimeSigned {
+		ti = tsig.TimeSigned - now
+	}
+	if uint64(tsig.Fudge) < ti {
+		return ErrTime
+	}
+
+	var h hash.Hash
+	switch strings.ToLower(tsig.Algorithm) {
+	case HmacMD5:
+		h = hmac.New(md5.New, rawsecret)
+	case HmacSHA1:
+		h = hmac.New(sha1.New, rawsecret)
+	case HmacSHA256:
+		h = hmac.New(sha256.New, rawsecret)
+	case HmacSHA512:
+		h = hmac.New(sha512.New, rawsecret)
+	default:
+		return ErrKeyAlg
+	}
+	h.Write(buf)
+	if !hmac.Equal(h.Sum(nil), msgMAC) {
+		return ErrSig
+	}
+	return nil
+}
+
+// Create a wiredata buffer for the MAC calculation.
+func tsigBuffer(msgbuf []byte, rr *TSIG, requestMAC string, timersOnly bool) []byte {
+	var buf []byte
+	if rr.TimeSigned == 0 {
+		rr.TimeSigned = uint64(time.Now().Unix())
+	}
+	if rr.Fudge == 0 {
+		rr.Fudge = 300 // Standard (RFC) default.
+	}
+
+	// Replace message ID in header with original ID from TSIG
+	binary.BigEndian.PutUint16(msgbuf[0:2], rr.OrigId)
+
+	if requestMAC != "" {
+		m := new(macWireFmt)
+		m.MACSize = uint16(len(requestMAC) / 2)
+		m.MAC = requestMAC
+		buf = make([]byte, len(requestMAC)) // long enough
+		n, _ := packMacWire(m, buf)
+		buf = buf[:n]
+	}
+
+	tsigvar := make([]byte, DefaultMsgSize)
+	if timersOnly {
+		tsig := new(timerWireFmt)
+		tsig.TimeSigned = rr.TimeSigned
+		tsig.Fudge = rr.Fudge
+		n, _ := packTimerWire(tsig, tsigvar)
+		tsigvar = tsigvar[:n]
+	} else {
+		tsig := new(tsigWireFmt)
+		tsig.Name = strings.ToLower(rr.Hdr.Name)
+		tsig.Class = ClassANY
+		tsig.Ttl = rr.Hdr.Ttl
+		tsig.Algorithm = strings.ToLower(rr.Algorithm)
+		tsig.TimeSigned = rr.TimeSigned
+		tsig.Fudge = rr.Fudge
+		tsig.Error = rr.Error
+		tsig.OtherLen = rr.OtherLen
+		tsig.OtherData = rr.OtherData
+		n, _ := packTsigWire(tsig, tsigvar)
+		tsigvar = tsigvar[:n]
+	}
+
+	if requestMAC != "" {
+		x := append(buf, msgbuf...)
+		buf = append(x, tsigvar...)
+	} else {
+		buf = append(msgbuf, tsigvar...)
+	}
+	return buf
+}
+
+// Strip the TSIG from the raw message.
+func stripTsig(msg []byte) ([]byte, *TSIG, error) {
+	// Copied from msg.go's Unpack() Header, but modified.
+	var (
+		dh  Header
+		err error
+	)
+	off, tsigoff := 0, 0
+
+	if dh, off, err = unpackMsgHdr(msg, off); err != nil {
+		return nil, nil, err
+	}
+	if dh.Arcount == 0 {
+		return nil, nil, ErrNoSig
+	}
+
+	// Rcode, see msg.go Unpack()
+	if int(dh.Bits&0xF) == RcodeNotAuth {
+		return nil, nil, ErrAuth
+	}
+
+	for i := 0; i < int(dh.Qdcount); i++ {
+		_, off, err = unpackQuestion(msg, off)
+		if err != nil {
+			return nil, nil, err
+		}
+	}
+
+	_, off, err = unpackRRslice(int(dh.Ancount), msg, off)
+	if err != nil {
+		return nil, nil, err
+	}
+	_, off, err = unpackRRslice(int(dh.Nscount), msg, off)
+	if err != nil {
+		return nil, nil, err
+	}
+
+	rr := new(TSIG)
+	var extra RR
+	for i := 0; i < int(dh.Arcount); i++ {
+		tsigoff = off
+		extra, off, err = UnpackRR(msg, off)
+		if err != nil {
+			return nil, nil, err
+		}
+		if extra.Header().Rrtype == TypeTSIG {
+			rr = extra.(*TSIG)
+			// Adjust Arcount.
+			arcount := binary.BigEndian.Uint16(msg[10:])
+			binary.BigEndian.PutUint16(msg[10:], arcount-1)
+			break
+		}
+	}
+	if rr == nil {
+		return nil, nil, ErrNoSig
+	}
+	return msg[:tsigoff], rr, nil
+}
+
+// Translate the TSIG time signed into a date. There is no
+// need for RFC1982 calculations as this date is 48 bits.
+func tsigTimeToString(t uint64) string {
+	ti := time.Unix(int64(t), 0).UTC()
+	return ti.Format("20060102150405")
+}
+
+func packTsigWire(tw *tsigWireFmt, msg []byte) (int, error) {
+	// copied from zmsg.go TSIG packing
+	// RR_Header
+	off, err := PackDomainName(tw.Name, msg, 0, nil, false)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint16(tw.Class, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint32(tw.Ttl, msg, off)
+	if err != nil {
+		return off, err
+	}
+
+	off, err = PackDomainName(tw.Algorithm, msg, off, nil, false)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint48(tw.TimeSigned, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint16(tw.Fudge, msg, off)
+	if err != nil {
+		return off, err
+	}
+
+	off, err = packUint16(tw.Error, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint16(tw.OtherLen, msg, off)
+	if err != nil {
+		return off, err
+	}
+	off, err = packStringHex(tw.OtherData, msg, off)
+	if err != nil {
+		return off, err
+	}
+	return off, nil
+}
+
+func packMacWire(mw *macWireFmt, msg []byte) (int, error) {
+	off, err := packUint16(mw.MACSize, msg, 0)
+	if err != nil {
+		return off, err
+	}
+	off, err = packStringHex(mw.MAC, msg, off)
+	if err != nil {
+		return off, err
+	}
+	return off, nil
+}
+
+func packTimerWire(tw *timerWireFmt, msg []byte) (int, error) {
+	off, err := packUint48(tw.TimeSigned, msg, 0)
+	if err != nil {
+		return off, err
+	}
+	off, err = packUint16(tw.Fudge, msg, off)
+	if err != nil {
+		return off, err
+	}
+	return off, nil
+}
--- a/vendor/github.com/miekg/dns/types.go
+++ b/vendor/github.com/miekg/dns/types.go
--- a/vendor/github.com/miekg/dns/types_generate.go
+++ b/vendor/github.com/miekg/dns/types_generate.go
@ -0,0 +1,272 @@
+//+build ignore
+
+// types_generate.go is meant to run with go generate. It will use
+// go/{importer,types} to track down all the RR struct types. Then for each type
+// it will generate conversion tables (TypeToRR and TypeToString) and banal
+// methods (len, Header, copy) based on the struct tags. The generated source is
+// written to ztypes.go, and is meant to be checked into git.
+package main
+
+import (
+	"bytes"
+	"fmt"
+	"go/format"
+	"go/importer"
+	"go/types"
+	"log"
+	"os"
+	"strings"
+	"text/template"
+)
+
+var skipLen = map[string]struct{}{
+	"NSEC":  {},
+	"NSEC3": {},
+	"OPT":   {},
+	"CSYNC": {},
+}
+
+var packageHdr = `
+// Code generated by "go run types_generate.go"; DO NOT EDIT.
+
+package dns
+
+import (
+	"encoding/base64"
+	"net"
+)
+
+`
+
+var TypeToRR = template.Must(template.New("TypeToRR").Parse(`
+// TypeToRR is a map of constructors for each RR type.
+var TypeToRR = map[uint16]func() RR{
+{{range .}}{{if ne . "RFC3597"}}  Type{{.}}:  func() RR { return new({{.}}) },
+{{end}}{{end}}                    }
+
+`))
+
+var typeToString = template.Must(template.New("typeToString").Parse(`
+// TypeToString is a map of strings for each RR type.
+var TypeToString = map[uint16]string{
+{{range .}}{{if ne . "NSAPPTR"}}  Type{{.}}: "{{.}}",
+{{end}}{{end}}                    TypeNSAPPTR:    "NSAP-PTR",
+}
+
+`))
+
+var headerFunc = template.Must(template.New("headerFunc").Parse(`
+{{range .}}  func (rr *{{.}}) Header() *RR_Header { return &rr.Hdr }
+{{end}}
+
+`))
+
+// getTypeStruct will take a type and the package scope, and return the
+// (innermost) struct if the type is considered a RR type (currently defined as
+// those structs beginning with a RR_Header, could be redefined as implementing
+// the RR interface). The bool return value indicates if embedded structs were
+// resolved.
+func getTypeStruct(t types.Type, scope *types.Scope) (*types.Struct, bool) {
+	st, ok := t.Underlying().(*types.Struct)
+	if !ok {
+		return nil, false
+	}
+	if st.Field(0).Type() == scope.Lookup("RR_Header").Type() {
+		return st, false
+	}
+	if st.Field(0).Anonymous() {
+		st, _ := getTypeStruct(st.Field(0).Type(), scope)
+		return st, true
+	}
+	return nil, false
+}
+
+func main() {
+	// Import and type-check the package
+	pkg, err := importer.Default().Import("github.com/miekg/dns")
+	fatalIfErr(err)
+	scope := pkg.Scope()
+
+	// Collect constants like TypeX
+	var numberedTypes []string
+	for _, name := range scope.Names() {
+		o := scope.Lookup(name)
+		if o == nil || !o.Exported() {
+			continue
+		}
+		b, ok := o.Type().(*types.Basic)
+		if !ok || b.Kind() != types.Uint16 {
+			continue
+		}
+		if !strings.HasPrefix(o.Name(), "Type") {
+			continue
+		}
+		name := strings.TrimPrefix(o.Name(), "Type")
+		if name == "PrivateRR" {
+			continue
+		}
+		numberedTypes = append(numberedTypes, name)
+	}
+
+	// Collect actual types (*X)
+	var namedTypes []string
+	for _, name := range scope.Names() {
+		o := scope.Lookup(name)
+		if o == nil || !o.Exported() {
+			continue
+		}
+		if st, _ := getTypeStruct(o.Type(), scope); st == nil {
+			continue
+		}
+		if name == "PrivateRR" {
+			continue
+		}
+
+		// Check if corresponding TypeX exists
+		if scope.Lookup("Type"+o.Name()) == nil && o.Name() != "RFC3597" {
+			log.Fatalf("Constant Type%s does not exist.", o.Name())
+		}
+
+		namedTypes = append(namedTypes, o.Name())
+	}
+
+	b := &bytes.Buffer{}
+	b.WriteString(packageHdr)
+
+	// Generate TypeToRR
+	fatalIfErr(TypeToRR.Execute(b, namedTypes))
+
+	// Generate typeToString
+	fatalIfErr(typeToString.Execute(b, numberedTypes))
+
+	// Generate headerFunc
+	fatalIfErr(headerFunc.Execute(b, namedTypes))
+
+	// Generate len()
+	fmt.Fprint(b, "// len() functions\n")
+	for _, name := range namedTypes {
+		if _, ok := skipLen[name]; ok {
+			continue
+		}
+		o := scope.Lookup(name)
+		st, isEmbedded := getTypeStruct(o.Type(), scope)
+		if isEmbedded {
+			continue
+		}
+		fmt.Fprintf(b, "func (rr *%s) len() int {\n", name)
+		fmt.Fprintf(b, "l := rr.Hdr.len()\n")
+		for i := 1; i < st.NumFields(); i++ {
+			o := func(s string) { fmt.Fprintf(b, s, st.Field(i).Name()) }
+
+			if _, ok := st.Field(i).Type().(*types.Slice); ok {
+				switch st.Tag(i) {
+				case `dns:"-"`:
+					// ignored
+				case `dns:"cdomain-name"`, `dns:"domain-name"`, `dns:"txt"`:
+					o("for _, x := range rr.%s { l += len(x) + 1 }\n")
+				default:
+					log.Fatalln(name, st.Field(i).Name(), st.Tag(i))
+				}
+				continue
+			}
+
+			switch {
+			case st.Tag(i) == `dns:"-"`:
+				// ignored
+			case st.Tag(i) == `dns:"cdomain-name"`, st.Tag(i) == `dns:"domain-name"`:
+				o("l += len(rr.%s) + 1\n")
+			case st.Tag(i) == `dns:"octet"`:
+				o("l += len(rr.%s)\n")
+			case strings.HasPrefix(st.Tag(i), `dns:"size-base64`):
+				fallthrough
+			case st.Tag(i) == `dns:"base64"`:
+				o("l += base64.StdEncoding.DecodedLen(len(rr.%s))\n")
+			case strings.HasPrefix(st.Tag(i), `dns:"size-hex:`): // this has an extra field where the length is stored
+				o("l += len(rr.%s)/2\n")
+			case strings.HasPrefix(st.Tag(i), `dns:"size-hex`):
+				fallthrough
+			case st.Tag(i) == `dns:"hex"`:
+				o("l += len(rr.%s)/2 + 1\n")
+			case st.Tag(i) == `dns:"a"`:
+				o("l += net.IPv4len // %s\n")
+			case st.Tag(i) == `dns:"aaaa"`:
+				o("l += net.IPv6len // %s\n")
+			case st.Tag(i) == `dns:"txt"`:
+				o("for _, t := range rr.%s { l += len(t) + 1 }\n")
+			case st.Tag(i) == `dns:"uint48"`:
+				o("l += 6 // %s\n")
+			case st.Tag(i) == "":
+				switch st.Field(i).Type().(*types.Basic).Kind() {
+				case types.Uint8:
+					o("l++ // %s\n")
+				case types.Uint16:
+					o("l += 2 // %s\n")
+				case types.Uint32:
+					o("l += 4 // %s\n")
+				case types.Uint64:
+					o("l += 8 // %s\n")
+				case types.String:
+					o("l += len(rr.%s) + 1\n")
+				default:
+					log.Fatalln(name, st.Field(i).Name())
+				}
+			default:
+				log.Fatalln(name, st.Field(i).Name(), st.Tag(i))
+			}
+		}
+		fmt.Fprintf(b, "return l }\n")
+	}
+
+	// Generate copy()
+	fmt.Fprint(b, "// copy() functions\n")
+	for _, name := range namedTypes {
+		o := scope.Lookup(name)
+		st, isEmbedded := getTypeStruct(o.Type(), scope)
+		if isEmbedded {
+			continue
+		}
+		fmt.Fprintf(b, "func (rr *%s) copy() RR {\n", name)
+		fields := []string{"*rr.Hdr.copyHeader()"}
+		for i := 1; i < st.NumFields(); i++ {
+			f := st.Field(i).Name()
+			if sl, ok := st.Field(i).Type().(*types.Slice); ok {
+				t := sl.Underlying().String()
+				t = strings.TrimPrefix(t, "[]")
+				if strings.Contains(t, ".") {
+					splits := strings.Split(t, ".")
+					t = splits[len(splits)-1]
+				}
+				fmt.Fprintf(b, "%s := make([]%s, len(rr.%s)); copy(%s, rr.%s)\n",
+					f, t, f, f, f)
+				fields = append(fields, f)
+				continue
+			}
+			if st.Field(i).Type().String() == "net.IP" {
+				fields = append(fields, "copyIP(rr."+f+")")
+				continue
+			}
+			fields = append(fields, "rr."+f)
+		}
+		fmt.Fprintf(b, "return &%s{%s}\n", name, strings.Join(fields, ","))
+		fmt.Fprintf(b, "}\n")
+	}
+
+	// gofmt
+	res, err := format.Source(b.Bytes())
+	if err != nil {
+		b.WriteTo(os.Stderr)
+		log.Fatal(err)
+	}
+
+	// write result
+	f, err := os.Create("ztypes.go")
+	fatalIfErr(err)
+	defer f.Close()
+	f.Write(res)
+}
+
+func fatalIfErr(err error) {
+	if err != nil {
+		log.Fatal(err)
+	}
+}
--- a/vendor/github.com/miekg/dns/udp.go
+++ b/vendor/github.com/miekg/dns/udp.go
@ -0,0 +1,89 @@
+// +build !windows
+
+package dns
+
+import (
+	"net"
+
+	"golang.org/x/net/ipv4"
+	"golang.org/x/net/ipv6"
+)
+
+// SessionUDP holds the remote address and the associated
+// out-of-band data.
+type SessionUDP struct {
+	raddr   *net.UDPAddr
+	context []byte
+}
+
+// RemoteAddr returns the remote network address.
+func (s *SessionUDP) RemoteAddr() net.Addr { return s.raddr }
+
+// ReadFromSessionUDP acts just like net.UDPConn.ReadFrom(), but returns a session object instead of a
+// net.UDPAddr.
+func ReadFromSessionUDP(conn *net.UDPConn, b []byte) (int, *SessionUDP, error) {
+	oob := make([]byte, 40)
+	n, oobn, _, raddr, err := conn.ReadMsgUDP(b, oob)
+	if err != nil {
+		return n, nil, err
+	}
+	return n, &SessionUDP{raddr, oob[:oobn]}, err
+}
+
+// WriteToSessionUDP acts just like net.UDPConn.WriteTo(), but uses a *SessionUDP instead of a net.Addr.
+func WriteToSessionUDP(conn *net.UDPConn, b []byte, session *SessionUDP) (int, error) {
+	oob := correctSource(session.context)
+	n, _, err := conn.WriteMsgUDP(b, oob, session.raddr)
+	return n, err
+}
+
+func setUDPSocketOptions(conn *net.UDPConn) error {
+	// Try setting the flags for both families and ignore the errors unless they
+	// both error.
+	err6 := ipv6.NewPacketConn(conn).SetControlMessage(ipv6.FlagDst|ipv6.FlagInterface, true)
+	err4 := ipv4.NewPacketConn(conn).SetControlMessage(ipv4.FlagDst|ipv4.FlagInterface, true)
+	if err6 != nil && err4 != nil {
+		return err4
+	}
+	return nil
+}
+
+// parseDstFromOOB takes oob data and returns the destination IP.
+func parseDstFromOOB(oob []byte) net.IP {
+	// Start with IPv6 and then fallback to IPv4
+	// TODO(fastest963): Figure out a way to prefer one or the other. Looking at
+	// the lvl of the header for a 0 or 41 isn't cross-platform.
+	var dst net.IP
+	cm6 := new(ipv6.ControlMessage)
+	if cm6.Parse(oob) == nil {
+		dst = cm6.Dst
+	}
+	if dst == nil {
+		cm4 := new(ipv4.ControlMessage)
+		if cm4.Parse(oob) == nil {
+			dst = cm4.Dst
+		}
+	}
+	return dst
+}
+
+// correctSource takes oob data and returns new oob data with the Src equal to the Dst
+func correctSource(oob []byte) []byte {
+	dst := parseDstFromOOB(oob)
+	if dst == nil {
+		return nil
+	}
+	// If the dst is definitely an IPv6, then use ipv6's ControlMessage to
+	// respond otherwise use ipv4's because ipv6's marshal ignores ipv4
+	// addresses.
+	if dst.To4() == nil {
+		cm := new(ipv6.ControlMessage)
+		cm.Src = dst
+		oob = cm.Marshal()
+	} else {
+		cm := new(ipv4.ControlMessage)
+		cm.Src = dst
+		oob = cm.Marshal()
+	}
+	return oob
+}
--- a/vendor/github.com/miekg/dns/udp_windows.go
+++ b/vendor/github.com/miekg/dns/udp_windows.go
@ -0,0 +1,37 @@
+// +build windows
+
+package dns
+
+import "net"
+
+// SessionUDP holds the remote address
+type SessionUDP struct {
+	raddr *net.UDPAddr
+}
+
+// RemoteAddr returns the remote network address.
+func (s *SessionUDP) RemoteAddr() net.Addr { return s.raddr }
+
+// ReadFromSessionUDP acts just like net.UDPConn.ReadFrom(), but returns a session object instead of a
+// net.UDPAddr.
+// TODO(fastest963): Once go1.10 is released, use ReadMsgUDP.
+func ReadFromSessionUDP(conn *net.UDPConn, b []byte) (int, *SessionUDP, error) {
+	n, raddr, err := conn.ReadFrom(b)
+	if err != nil {
+		return n, nil, err
+	}
+	session := &SessionUDP{raddr.(*net.UDPAddr)}
+	return n, session, err
+}
+
+// WriteToSessionUDP acts just like net.UDPConn.WriteTo(), but uses a *SessionUDP instead of a net.Addr.
+// TODO(fastest963): Once go1.10 is released, use WriteMsgUDP.
+func WriteToSessionUDP(conn *net.UDPConn, b []byte, session *SessionUDP) (int, error) {
+	n, err := conn.WriteTo(b, session.raddr)
+	return n, err
+}
+
+// TODO(fastest963): Once go1.10 is released and we can use *MsgUDP methods
+// use the standard method in udp.go for these.
+func setUDPSocketOptions(*net.UDPConn) error { return nil }
+func parseDstFromOOB([]byte, net.IP) net.IP  { return nil }
--- a/vendor/github.com/miekg/dns/update.go
+++ b/vendor/github.com/miekg/dns/update.go
@ -0,0 +1,106 @@
+package dns
+
+// NameUsed sets the RRs in the prereq section to
+// "Name is in use" RRs. RFC 2136 section 2.4.4.
+func (u *Msg) NameUsed(rr []RR) {
+	if u.Answer == nil {
+		u.Answer = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		u.Answer = append(u.Answer, &ANY{Hdr: RR_Header{Name: r.Header().Name, Ttl: 0, Rrtype: TypeANY, Class: ClassANY}})
+	}
+}
+
+// NameNotUsed sets the RRs in the prereq section to
+// "Name is in not use" RRs. RFC 2136 section 2.4.5.
+func (u *Msg) NameNotUsed(rr []RR) {
+	if u.Answer == nil {
+		u.Answer = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		u.Answer = append(u.Answer, &ANY{Hdr: RR_Header{Name: r.Header().Name, Ttl: 0, Rrtype: TypeANY, Class: ClassNONE}})
+	}
+}
+
+// Used sets the RRs in the prereq section to
+// "RRset exists (value dependent -- with rdata)" RRs. RFC 2136 section 2.4.2.
+func (u *Msg) Used(rr []RR) {
+	if len(u.Question) == 0 {
+		panic("dns: empty question section")
+	}
+	if u.Answer == nil {
+		u.Answer = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		r.Header().Class = u.Question[0].Qclass
+		u.Answer = append(u.Answer, r)
+	}
+}
+
+// RRsetUsed sets the RRs in the prereq section to
+// "RRset exists (value independent -- no rdata)" RRs. RFC 2136 section 2.4.1.
+func (u *Msg) RRsetUsed(rr []RR) {
+	if u.Answer == nil {
+		u.Answer = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		u.Answer = append(u.Answer, &ANY{Hdr: RR_Header{Name: r.Header().Name, Ttl: 0, Rrtype: r.Header().Rrtype, Class: ClassANY}})
+	}
+}
+
+// RRsetNotUsed sets the RRs in the prereq section to
+// "RRset does not exist" RRs. RFC 2136 section 2.4.3.
+func (u *Msg) RRsetNotUsed(rr []RR) {
+	if u.Answer == nil {
+		u.Answer = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		u.Answer = append(u.Answer, &ANY{Hdr: RR_Header{Name: r.Header().Name, Ttl: 0, Rrtype: r.Header().Rrtype, Class: ClassNONE}})
+	}
+}
+
+// Insert creates a dynamic update packet that adds an complete RRset, see RFC 2136 section 2.5.1.
+func (u *Msg) Insert(rr []RR) {
+	if len(u.Question) == 0 {
+		panic("dns: empty question section")
+	}
+	if u.Ns == nil {
+		u.Ns = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		r.Header().Class = u.Question[0].Qclass
+		u.Ns = append(u.Ns, r)
+	}
+}
+
+// RemoveRRset creates a dynamic update packet that deletes an RRset, see RFC 2136 section 2.5.2.
+func (u *Msg) RemoveRRset(rr []RR) {
+	if u.Ns == nil {
+		u.Ns = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		u.Ns = append(u.Ns, &ANY{Hdr: RR_Header{Name: r.Header().Name, Ttl: 0, Rrtype: r.Header().Rrtype, Class: ClassANY}})
+	}
+}
+
+// RemoveName creates a dynamic update packet that deletes all RRsets of a name, see RFC 2136 section 2.5.3
+func (u *Msg) RemoveName(rr []RR) {
+	if u.Ns == nil {
+		u.Ns = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		u.Ns = append(u.Ns, &ANY{Hdr: RR_Header{Name: r.Header().Name, Ttl: 0, Rrtype: TypeANY, Class: ClassANY}})
+	}
+}
+
+// Remove creates a dynamic update packet deletes RR from a RRSset, see RFC 2136 section 2.5.4
+func (u *Msg) Remove(rr []RR) {
+	if u.Ns == nil {
+		u.Ns = make([]RR, 0, len(rr))
+	}
+	for _, r := range rr {
+		r.Header().Class = ClassNONE
+		r.Header().Ttl = 0
+		u.Ns = append(u.Ns, r)
+	}
+}
--- a/vendor/github.com/miekg/dns/version.go
+++ b/vendor/github.com/miekg/dns/version.go
@ -0,0 +1,15 @@
+package dns
+
+import "fmt"
+
+// Version is current version of this library.
+var Version = V{1, 0, 4}
+
+// V holds the version of this library.
+type V struct {
+	Major, Minor, Patch int
+}
+
+func (v V) String() string {
+	return fmt.Sprintf("%d.%d.%d", v.Major, v.Minor, v.Patch)
+}
--- a/vendor/github.com/miekg/dns/xfr.go
+++ b/vendor/github.com/miekg/dns/xfr.go
@ -0,0 +1,260 @@
+package dns
+
+import (
+	"fmt"
+	"time"
+)
+
+// Envelope is used when doing a zone transfer with a remote server.
+type Envelope struct {
+	RR    []RR  // The set of RRs in the answer section of the xfr reply message.
+	Error error // If something went wrong, this contains the error.
+}
+
+// A Transfer defines parameters that are used during a zone transfer.
+type Transfer struct {
+	*Conn
+	DialTimeout    time.Duration     // net.DialTimeout, defaults to 2 seconds
+	ReadTimeout    time.Duration     // net.Conn.SetReadTimeout value for connections, defaults to 2 seconds
+	WriteTimeout   time.Duration     // net.Conn.SetWriteTimeout value for connections, defaults to 2 seconds
+	TsigSecret     map[string]string // Secret(s) for Tsig map[<zonename>]<base64 secret>, zonename must be in canonical form (lowercase, fqdn, see RFC 4034 Section 6.2)
+	tsigTimersOnly bool
+}
+
+// Think we need to away to stop the transfer
+
+// In performs an incoming transfer with the server in a.
+// If you would like to set the source IP, or some other attribute
+// of a Dialer for a Transfer, you can do so by specifying the attributes
+// in the Transfer.Conn:
+//
+//	d := net.Dialer{LocalAddr: transfer_source}
+//	con, err := d.Dial("tcp", master)
+//	dnscon := &dns.Conn{Conn:con}
+//	transfer = &dns.Transfer{Conn: dnscon}
+//	channel, err := transfer.In(message, master)
+//
+func (t *Transfer) In(q *Msg, a string) (env chan *Envelope, err error) {
+	timeout := dnsTimeout
+	if t.DialTimeout != 0 {
+		timeout = t.DialTimeout
+	}
+	if t.Conn == nil {
+		t.Conn, err = DialTimeout("tcp", a, timeout)
+		if err != nil {
+			return nil, err
+		}
+	}
+	if err := t.WriteMsg(q); err != nil {
+		return nil, err
+	}
+	env = make(chan *Envelope)
+	go func() {
+		if q.Question[0].Qtype == TypeAXFR {
+			go t.inAxfr(q, env)
+			return
+		}
+		if q.Question[0].Qtype == TypeIXFR {
+			go t.inIxfr(q, env)
+			return
+		}
+	}()
+	return env, nil
+}
+
+func (t *Transfer) inAxfr(q *Msg, c chan *Envelope) {
+	first := true
+	defer t.Close()
+	defer close(c)
+	timeout := dnsTimeout
+	if t.ReadTimeout != 0 {
+		timeout = t.ReadTimeout
+	}
+	for {
+		t.Conn.SetReadDeadline(time.Now().Add(timeout))
+		in, err := t.ReadMsg()
+		if err != nil {
+			c <- &Envelope{nil, err}
+			return
+		}
+		if q.Id != in.Id {
+			c <- &Envelope{in.Answer, ErrId}
+			return
+		}
+		if first {
+			if in.Rcode != RcodeSuccess {
+				c <- &Envelope{in.Answer, &Error{err: fmt.Sprintf(errXFR, in.Rcode)}}
+				return
+			}
+			if !isSOAFirst(in) {
+				c <- &Envelope{in.Answer, ErrSoa}
+				return
+			}
+			first = !first
+			// only one answer that is SOA, receive more
+			if len(in.Answer) == 1 {
+				t.tsigTimersOnly = true
+				c <- &Envelope{in.Answer, nil}
+				continue
+			}
+		}
+
+		if !first {
+			t.tsigTimersOnly = true // Subsequent envelopes use this.
+			if isSOALast(in) {
+				c <- &Envelope{in.Answer, nil}
+				return
+			}
+			c <- &Envelope{in.Answer, nil}
+		}
+	}
+}
+
+func (t *Transfer) inIxfr(q *Msg, c chan *Envelope) {
+	serial := uint32(0) // The first serial seen is the current server serial
+	axfr := true
+	n := 0
+	qser := q.Ns[0].(*SOA).Serial
+	defer t.Close()
+	defer close(c)
+	timeout := dnsTimeout
+	if t.ReadTimeout != 0 {
+		timeout = t.ReadTimeout
+	}
+	for {
+		t.SetReadDeadline(time.Now().Add(timeout))
+		in, err := t.ReadMsg()
+		if err != nil {
+			c <- &Envelope{nil, err}
+			return
+		}
+		if q.Id != in.Id {
+			c <- &Envelope{in.Answer, ErrId}
+			return
+		}
+		if in.Rcode != RcodeSuccess {
+			c <- &Envelope{in.Answer, &Error{err: fmt.Sprintf(errXFR, in.Rcode)}}
+			return
+		}
+		if n == 0 {
+			// Check if the returned answer is ok
+			if !isSOAFirst(in) {
+				c <- &Envelope{in.Answer, ErrSoa}
+				return
+			}
+			// This serial is important
+			serial = in.Answer[0].(*SOA).Serial
+			// Check if there are no changes in zone
+			if qser >= serial {
+				c <- &Envelope{in.Answer, nil}
+				return
+			}
+		}
+		// Now we need to check each message for SOA records, to see what we need to do
+		t.tsigTimersOnly = true
+		for _, rr := range in.Answer {
+			if v, ok := rr.(*SOA); ok {
+				if v.Serial == serial {
+					n++
+					// quit if it's a full axfr or the the servers' SOA is repeated the third time
+					if axfr && n == 2 || n == 3 {
+						c <- &Envelope{in.Answer, nil}
+						return
+					}
+				} else if axfr {
+					// it's an ixfr
+					axfr = false
+				}
+			}
+		}
+		c <- &Envelope{in.Answer, nil}
+	}
+}
+
+// Out performs an outgoing transfer with the client connecting in w.
+// Basic use pattern:
+//
+//	ch := make(chan *dns.Envelope)
+//	tr := new(dns.Transfer)
+//	go tr.Out(w, r, ch)
+//	ch <- &dns.Envelope{RR: []dns.RR{soa, rr1, rr2, rr3, soa}}
+//	close(ch)
+//	w.Hijack()
+//	// w.Close() // Client closes connection
+//
+// The server is responsible for sending the correct sequence of RRs through the
+// channel ch.
+func (t *Transfer) Out(w ResponseWriter, q *Msg, ch chan *Envelope) error {
+	for x := range ch {
+		r := new(Msg)
+		// Compress?
+		r.SetReply(q)
+		r.Authoritative = true
+		// assume it fits TODO(miek): fix
+		r.Answer = append(r.Answer, x.RR...)
+		if err := w.WriteMsg(r); err != nil {
+			return err
+		}
+	}
+	w.TsigTimersOnly(true)
+	return nil
+}
+
+// ReadMsg reads a message from the transfer connection t.
+func (t *Transfer) ReadMsg() (*Msg, error) {
+	m := new(Msg)
+	p := make([]byte, MaxMsgSize)
+	n, err := t.Read(p)
+	if err != nil && n == 0 {
+		return nil, err
+	}
+	p = p[:n]
+	if err := m.Unpack(p); err != nil {
+		return nil, err
+	}
+	if ts := m.IsTsig(); ts != nil && t.TsigSecret != nil {
+		if _, ok := t.TsigSecret[ts.Hdr.Name]; !ok {
+			return m, ErrSecret
+		}
+		// Need to work on the original message p, as that was used to calculate the tsig.
+		err = TsigVerify(p, t.TsigSecret[ts.Hdr.Name], t.tsigRequestMAC, t.tsigTimersOnly)
+		t.tsigRequestMAC = ts.MAC
+	}
+	return m, err
+}
+
+// WriteMsg writes a message through the transfer connection t.
+func (t *Transfer) WriteMsg(m *Msg) (err error) {
+	var out []byte
+	if ts := m.IsTsig(); ts != nil && t.TsigSecret != nil {
+		if _, ok := t.TsigSecret[ts.Hdr.Name]; !ok {
+			return ErrSecret
+		}
+		out, t.tsigRequestMAC, err = TsigGenerate(m, t.TsigSecret[ts.Hdr.Name], t.tsigRequestMAC, t.tsigTimersOnly)
+	} else {
+		out, err = m.Pack()
+	}
+	if err != nil {
+		return err
+	}
+	if _, err = t.Write(out); err != nil {
+		return err
+	}
+	return nil
+}
+
+func isSOAFirst(in *Msg) bool {
+	if len(in.Answer) > 0 {
+		return in.Answer[0].Header().Rrtype == TypeSOA
+	}
+	return false
+}
+
+func isSOALast(in *Msg) bool {
+	if len(in.Answer) > 0 {
+		return in.Answer[len(in.Answer)-1].Header().Rrtype == TypeSOA
+	}
+	return false
+}
+
+const errXFR = "bad xfr rcode: %d"
--- a/vendor/github.com/miekg/dns/zcompress.go
+++ b/vendor/github.com/miekg/dns/zcompress.go
@ -0,0 +1,118 @@
+// Code generated by "go run compress_generate.go"; DO NOT EDIT.
+
+package dns
+
+func compressionLenHelperType(c map[string]int, r RR) {
+	switch x := r.(type) {
+	case *AFSDB:
+		compressionLenHelper(c, x.Hostname)
+	case *CNAME:
+		compressionLenHelper(c, x.Target)
+	case *DNAME:
+		compressionLenHelper(c, x.Target)
+	case *HIP:
+		for i := range x.RendezvousServers {
+			compressionLenHelper(c, x.RendezvousServers[i])
+		}
+	case *KX:
+		compressionLenHelper(c, x.Exchanger)
+	case *LP:
+		compressionLenHelper(c, x.Fqdn)
+	case *MB:
+		compressionLenHelper(c, x.Mb)
+	case *MD:
+		compressionLenHelper(c, x.Md)
+	case *MF:
+		compressionLenHelper(c, x.Mf)
+	case *MG:
+		compressionLenHelper(c, x.Mg)
+	case *MINFO:
+		compressionLenHelper(c, x.Rmail)
+		compressionLenHelper(c, x.Email)
+	case *MR:
+		compressionLenHelper(c, x.Mr)
+	case *MX:
+		compressionLenHelper(c, x.Mx)
+	case *NAPTR:
+		compressionLenHelper(c, x.Replacement)
+	case *NS:
+		compressionLenHelper(c, x.Ns)
+	case *NSAPPTR:
+		compressionLenHelper(c, x.Ptr)
+	case *NSEC:
+		compressionLenHelper(c, x.NextDomain)
+	case *PTR:
+		compressionLenHelper(c, x.Ptr)
+	case *PX:
+		compressionLenHelper(c, x.Map822)
+		compressionLenHelper(c, x.Mapx400)
+	case *RP:
+		compressionLenHelper(c, x.Mbox)
+		compressionLenHelper(c, x.Txt)
+	case *RRSIG:
+		compressionLenHelper(c, x.SignerName)
+	case *RT:
+		compressionLenHelper(c, x.Host)
+	case *SIG:
+		compressionLenHelper(c, x.SignerName)
+	case *SOA:
+		compressionLenHelper(c, x.Ns)
+		compressionLenHelper(c, x.Mbox)
+	case *SRV:
+		compressionLenHelper(c, x.Target)
+	case *TALINK:
+		compressionLenHelper(c, x.PreviousName)
+		compressionLenHelper(c, x.NextName)
+	case *TKEY:
+		compressionLenHelper(c, x.Algorithm)
+	case *TSIG:
+		compressionLenHelper(c, x.Algorithm)
+	}
+}
+
+func compressionLenSearchType(c map[string]int, r RR) (int, bool) {
+	switch x := r.(type) {
+	case *AFSDB:
+		k1, ok1 := compressionLenSearch(c, x.Hostname)
+		return k1, ok1
+	case *CNAME:
+		k1, ok1 := compressionLenSearch(c, x.Target)
+		return k1, ok1
+	case *MB:
+		k1, ok1 := compressionLenSearch(c, x.Mb)
+		return k1, ok1
+	case *MD:
+		k1, ok1 := compressionLenSearch(c, x.Md)
+		return k1, ok1
+	case *MF:
+		k1, ok1 := compressionLenSearch(c, x.Mf)
+		return k1, ok1
+	case *MG:
+		k1, ok1 := compressionLenSearch(c, x.Mg)
+		return k1, ok1
+	case *MINFO:
+		k1, ok1 := compressionLenSearch(c, x.Rmail)
+		k2, ok2 := compressionLenSearch(c, x.Email)
+		return k1 + k2, ok1 && ok2
+	case *MR:
+		k1, ok1 := compressionLenSearch(c, x.Mr)
+		return k1, ok1
+	case *MX:
+		k1, ok1 := compressionLenSearch(c, x.Mx)
+		return k1, ok1
+	case *NS:
+		k1, ok1 := compressionLenSearch(c, x.Ns)
+		return k1, ok1
+	case *PTR:
+		k1, ok1 := compressionLenSearch(c, x.Ptr)
+		return k1, ok1
+	case *RT:
+		k1, ok1 := compressionLenSearch(c, x.Host)
+		return k1, ok1
+	case *SOA:
+		k1, ok1 := compressionLenSearch(c, x.Ns)
+		k2, ok2 := compressionLenSearch(c, x.Mbox)
+		return k1 + k2, ok1 && ok2
+	}
+	return 0, false
+}
--- a/vendor/github.com/miekg/dns/zmsg.go
+++ b/vendor/github.com/miekg/dns/zmsg.go
--- a/vendor/github.com/miekg/dns/ztypes.go
+++ b/vendor/github.com/miekg/dns/ztypes.go
@ -0,0 +1,863 @@
+// Code generated by "go run types_generate.go"; DO NOT EDIT.
+
+package dns
+
+import (
+	"encoding/base64"
+	"net"
+)
+
+// TypeToRR is a map of constructors for each RR type.
+var TypeToRR = map[uint16]func() RR{
+	TypeA:          func() RR { return new(A) },
+	TypeAAAA:       func() RR { return new(AAAA) },
+	TypeAFSDB:      func() RR { return new(AFSDB) },
+	TypeANY:        func() RR { return new(ANY) },
+	TypeAVC:        func() RR { return new(AVC) },
+	TypeCAA:        func() RR { return new(CAA) },
+	TypeCDNSKEY:    func() RR { return new(CDNSKEY) },
+	TypeCDS:        func() RR { return new(CDS) },
+	TypeCERT:       func() RR { return new(CERT) },
+	TypeCNAME:      func() RR { return new(CNAME) },
+	TypeCSYNC:      func() RR { return new(CSYNC) },
+	TypeDHCID:      func() RR { return new(DHCID) },
+	TypeDLV:        func() RR { return new(DLV) },
+	TypeDNAME:      func() RR { return new(DNAME) },
+	TypeDNSKEY:     func() RR { return new(DNSKEY) },
+	TypeDS:         func() RR { return new(DS) },
+	TypeEID:        func() RR { return new(EID) },
+	TypeEUI48:      func() RR { return new(EUI48) },
+	TypeEUI64:      func() RR { return new(EUI64) },
+	TypeGID:        func() RR { return new(GID) },
+	TypeGPOS:       func() RR { return new(GPOS) },
+	TypeHINFO:      func() RR { return new(HINFO) },
+	TypeHIP:        func() RR { return new(HIP) },
+	TypeKEY:        func() RR { return new(KEY) },
+	TypeKX:         func() RR { return new(KX) },
+	TypeL32:        func() RR { return new(L32) },
+	TypeL64:        func() RR { return new(L64) },
+	TypeLOC:        func() RR { return new(LOC) },
+	TypeLP:         func() RR { return new(LP) },
+	TypeMB:         func() RR { return new(MB) },
+	TypeMD:         func() RR { return new(MD) },
+	TypeMF:         func() RR { return new(MF) },
+	TypeMG:         func() RR { return new(MG) },
+	TypeMINFO:      func() RR { return new(MINFO) },
+	TypeMR:         func() RR { return new(MR) },
+	TypeMX:         func() RR { return new(MX) },
+	TypeNAPTR:      func() RR { return new(NAPTR) },
+	TypeNID:        func() RR { return new(NID) },
+	TypeNIMLOC:     func() RR { return new(NIMLOC) },
+	TypeNINFO:      func() RR { return new(NINFO) },
+	TypeNS:         func() RR { return new(NS) },
+	TypeNSAPPTR:    func() RR { return new(NSAPPTR) },
+	TypeNSEC:       func() RR { return new(NSEC) },
+	TypeNSEC3:      func() RR { return new(NSEC3) },
+	TypeNSEC3PARAM: func() RR { return new(NSEC3PARAM) },
+	TypeOPENPGPKEY: func() RR { return new(OPENPGPKEY) },
+	TypeOPT:        func() RR { return new(OPT) },
+	TypePTR:        func() RR { return new(PTR) },
+	TypePX:         func() RR { return new(PX) },
+	TypeRKEY:       func() RR { return new(RKEY) },
+	TypeRP:         func() RR { return new(RP) },
+	TypeRRSIG:      func() RR { return new(RRSIG) },
+	TypeRT:         func() RR { return new(RT) },
+	TypeSIG:        func() RR { return new(SIG) },
+	TypeSMIMEA:     func() RR { return new(SMIMEA) },
+	TypeSOA:        func() RR { return new(SOA) },
+	TypeSPF:        func() RR { return new(SPF) },
+	TypeSRV:        func() RR { return new(SRV) },
+	TypeSSHFP:      func() RR { return new(SSHFP) },
+	TypeTA:         func() RR { return new(TA) },
+	TypeTALINK:     func() RR { return new(TALINK) },
+	TypeTKEY:       func() RR { return new(TKEY) },
+	TypeTLSA:       func() RR { return new(TLSA) },
+	TypeTSIG:       func() RR { return new(TSIG) },
+	TypeTXT:        func() RR { return new(TXT) },
+	TypeUID:        func() RR { return new(UID) },
+	TypeUINFO:      func() RR { return new(UINFO) },
+	TypeURI:        func() RR { return new(URI) },
+	TypeX25:        func() RR { return new(X25) },
+}
+
+// TypeToString is a map of strings for each RR type.
+var TypeToString = map[uint16]string{
+	TypeA:          "A",
+	TypeAAAA:       "AAAA",
+	TypeAFSDB:      "AFSDB",
+	TypeANY:        "ANY",
+	TypeATMA:       "ATMA",
+	TypeAVC:        "AVC",
+	TypeAXFR:       "AXFR",
+	TypeCAA:        "CAA",
+	TypeCDNSKEY:    "CDNSKEY",
+	TypeCDS:        "CDS",
+	TypeCERT:       "CERT",
+	TypeCNAME:      "CNAME",
+	TypeCSYNC:      "CSYNC",
+	TypeDHCID:      "DHCID",
+	TypeDLV:        "DLV",
+	TypeDNAME:      "DNAME",
+	TypeDNSKEY:     "DNSKEY",
+	TypeDS:         "DS",
+	TypeEID:        "EID",
+	TypeEUI48:      "EUI48",
+	TypeEUI64:      "EUI64",
+	TypeGID:        "GID",
+	TypeGPOS:       "GPOS",
+	TypeHINFO:      "HINFO",
+	TypeHIP:        "HIP",
+	TypeISDN:       "ISDN",
+	TypeIXFR:       "IXFR",
+	TypeKEY:        "KEY",
+	TypeKX:         "KX",
+	TypeL32:        "L32",
+	TypeL64:        "L64",
+	TypeLOC:        "LOC",
+	TypeLP:         "LP",
+	TypeMAILA:      "MAILA",
+	TypeMAILB:      "MAILB",
+	TypeMB:         "MB",
+	TypeMD:         "MD",
+	TypeMF:         "MF",
+	TypeMG:         "MG",
+	TypeMINFO:      "MINFO",
+	TypeMR:         "MR",
+	TypeMX:         "MX",
+	TypeNAPTR:      "NAPTR",
+	TypeNID:        "NID",
+	TypeNIMLOC:     "NIMLOC",
+	TypeNINFO:      "NINFO",
+	TypeNS:         "NS",
+	TypeNSEC:       "NSEC",
+	TypeNSEC3:      "NSEC3",
+	TypeNSEC3PARAM: "NSEC3PARAM",
+	TypeNULL:       "NULL",
+	TypeNXT:        "NXT",
+	TypeNone:       "None",
+	TypeOPENPGPKEY: "OPENPGPKEY",
+	TypeOPT:        "OPT",
+	TypePTR:        "PTR",
+	TypePX:         "PX",
+	TypeRKEY:       "RKEY",
+	TypeRP:         "RP",
+	TypeRRSIG:      "RRSIG",
+	TypeRT:         "RT",
+	TypeReserved:   "Reserved",
+	TypeSIG:        "SIG",
+	TypeSMIMEA:     "SMIMEA",
+	TypeSOA:        "SOA",
+	TypeSPF:        "SPF",
+	TypeSRV:        "SRV",
+	TypeSSHFP:      "SSHFP",
+	TypeTA:         "TA",
+	TypeTALINK:     "TALINK",
+	TypeTKEY:       "TKEY",
+	TypeTLSA:       "TLSA",
+	TypeTSIG:       "TSIG",
+	TypeTXT:        "TXT",
+	TypeUID:        "UID",
+	TypeUINFO:      "UINFO",
+	TypeUNSPEC:     "UNSPEC",
+	TypeURI:        "URI",
+	TypeX25:        "X25",
+	TypeNSAPPTR:    "NSAP-PTR",
+}
+
+func (rr *A) Header() *RR_Header          { return &rr.Hdr }
+func (rr *AAAA) Header() *RR_Header       { return &rr.Hdr }
+func (rr *AFSDB) Header() *RR_Header      { return &rr.Hdr }
+func (rr *ANY) Header() *RR_Header        { return &rr.Hdr }
+func (rr *AVC) Header() *RR_Header        { return &rr.Hdr }
+func (rr *CAA) Header() *RR_Header        { return &rr.Hdr }
+func (rr *CDNSKEY) Header() *RR_Header    { return &rr.Hdr }
+func (rr *CDS) Header() *RR_Header        { return &rr.Hdr }
+func (rr *CERT) Header() *RR_Header       { return &rr.Hdr }
+func (rr *CNAME) Header() *RR_Header      { return &rr.Hdr }
+func (rr *CSYNC) Header() *RR_Header      { return &rr.Hdr }
+func (rr *DHCID) Header() *RR_Header      { return &rr.Hdr }
+func (rr *DLV) Header() *RR_Header        { return &rr.Hdr }
+func (rr *DNAME) Header() *RR_Header      { return &rr.Hdr }
+func (rr *DNSKEY) Header() *RR_Header     { return &rr.Hdr }
+func (rr *DS) Header() *RR_Header         { return &rr.Hdr }
+func (rr *EID) Header() *RR_Header        { return &rr.Hdr }
+func (rr *EUI48) Header() *RR_Header      { return &rr.Hdr }
+func (rr *EUI64) Header() *RR_Header      { return &rr.Hdr }
+func (rr *GID) Header() *RR_Header        { return &rr.Hdr }
+func (rr *GPOS) Header() *RR_Header       { return &rr.Hdr }
+func (rr *HINFO) Header() *RR_Header      { return &rr.Hdr }
+func (rr *HIP) Header() *RR_Header        { return &rr.Hdr }
+func (rr *KEY) Header() *RR_Header        { return &rr.Hdr }
+func (rr *KX) Header() *RR_Header         { return &rr.Hdr }
+func (rr *L32) Header() *RR_Header        { return &rr.Hdr }
+func (rr *L64) Header() *RR_Header        { return &rr.Hdr }
+func (rr *LOC) Header() *RR_Header        { return &rr.Hdr }
+func (rr *LP) Header() *RR_Header         { return &rr.Hdr }
+func (rr *MB) Header() *RR_Header         { return &rr.Hdr }
+func (rr *MD) Header() *RR_Header         { return &rr.Hdr }
+func (rr *MF) Header() *RR_Header         { return &rr.Hdr }
+func (rr *MG) Header() *RR_Header         { return &rr.Hdr }
+func (rr *MINFO) Header() *RR_Header      { return &rr.Hdr }
+func (rr *MR) Header() *RR_Header         { return &rr.Hdr }
+func (rr *MX) Header() *RR_Header         { return &rr.Hdr }
+func (rr *NAPTR) Header() *RR_Header      { return &rr.Hdr }
+func (rr *NID) Header() *RR_Header        { return &rr.Hdr }
+func (rr *NIMLOC) Header() *RR_Header     { return &rr.Hdr }
+func (rr *NINFO) Header() *RR_Header      { return &rr.Hdr }
+func (rr *NS) Header() *RR_Header         { return &rr.Hdr }
+func (rr *NSAPPTR) Header() *RR_Header    { return &rr.Hdr }
+func (rr *NSEC) Header() *RR_Header       { return &rr.Hdr }
+func (rr *NSEC3) Header() *RR_Header      { return &rr.Hdr }
+func (rr *NSEC3PARAM) Header() *RR_Header { return &rr.Hdr }
+func (rr *OPENPGPKEY) Header() *RR_Header { return &rr.Hdr }
+func (rr *OPT) Header() *RR_Header        { return &rr.Hdr }
+func (rr *PTR) Header() *RR_Header        { return &rr.Hdr }
+func (rr *PX) Header() *RR_Header         { return &rr.Hdr }
+func (rr *RFC3597) Header() *RR_Header    { return &rr.Hdr }
+func (rr *RKEY) Header() *RR_Header       { return &rr.Hdr }
+func (rr *RP) Header() *RR_Header         { return &rr.Hdr }
+func (rr *RRSIG) Header() *RR_Header      { return &rr.Hdr }
+func (rr *RT) Header() *RR_Header         { return &rr.Hdr }
+func (rr *SIG) Header() *RR_Header        { return &rr.Hdr }
+func (rr *SMIMEA) Header() *RR_Header     { return &rr.Hdr }
+func (rr *SOA) Header() *RR_Header        { return &rr.Hdr }
+func (rr *SPF) Header() *RR_Header        { return &rr.Hdr }
+func (rr *SRV) Header() *RR_Header        { return &rr.Hdr }
+func (rr *SSHFP) Header() *RR_Header      { return &rr.Hdr }
+func (rr *TA) Header() *RR_Header         { return &rr.Hdr }
+func (rr *TALINK) Header() *RR_Header     { return &rr.Hdr }
+func (rr *TKEY) Header() *RR_Header       { return &rr.Hdr }
+func (rr *TLSA) Header() *RR_Header       { return &rr.Hdr }
+func (rr *TSIG) Header() *RR_Header       { return &rr.Hdr }
+func (rr *TXT) Header() *RR_Header        { return &rr.Hdr }
+func (rr *UID) Header() *RR_Header        { return &rr.Hdr }
+func (rr *UINFO) Header() *RR_Header      { return &rr.Hdr }
+func (rr *URI) Header() *RR_Header        { return &rr.Hdr }
+func (rr *X25) Header() *RR_Header        { return &rr.Hdr }
+
+// len() functions
+func (rr *A) len() int {
+	l := rr.Hdr.len()
+	l += net.IPv4len // A
+	return l
+}
+func (rr *AAAA) len() int {
+	l := rr.Hdr.len()
+	l += net.IPv6len // AAAA
+	return l
+}
+func (rr *AFSDB) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Subtype
+	l += len(rr.Hostname) + 1
+	return l
+}
+func (rr *ANY) len() int {
+	l := rr.Hdr.len()
+	return l
+}
+func (rr *AVC) len() int {
+	l := rr.Hdr.len()
+	for _, x := range rr.Txt {
+		l += len(x) + 1
+	}
+	return l
+}
+func (rr *CAA) len() int {
+	l := rr.Hdr.len()
+	l++ // Flag
+	l += len(rr.Tag) + 1
+	l += len(rr.Value)
+	return l
+}
+func (rr *CERT) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Type
+	l += 2 // KeyTag
+	l++    // Algorithm
+	l += base64.StdEncoding.DecodedLen(len(rr.Certificate))
+	return l
+}
+func (rr *CNAME) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Target) + 1
+	return l
+}
+func (rr *DHCID) len() int {
+	l := rr.Hdr.len()
+	l += base64.StdEncoding.DecodedLen(len(rr.Digest))
+	return l
+}
+func (rr *DNAME) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Target) + 1
+	return l
+}
+func (rr *DNSKEY) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Flags
+	l++    // Protocol
+	l++    // Algorithm
+	l += base64.StdEncoding.DecodedLen(len(rr.PublicKey))
+	return l
+}
+func (rr *DS) len() int {
+	l := rr.Hdr.len()
+	l += 2 // KeyTag
+	l++    // Algorithm
+	l++    // DigestType
+	l += len(rr.Digest)/2 + 1
+	return l
+}
+func (rr *EID) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Endpoint)/2 + 1
+	return l
+}
+func (rr *EUI48) len() int {
+	l := rr.Hdr.len()
+	l += 6 // Address
+	return l
+}
+func (rr *EUI64) len() int {
+	l := rr.Hdr.len()
+	l += 8 // Address
+	return l
+}
+func (rr *GID) len() int {
+	l := rr.Hdr.len()
+	l += 4 // Gid
+	return l
+}
+func (rr *GPOS) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Longitude) + 1
+	l += len(rr.Latitude) + 1
+	l += len(rr.Altitude) + 1
+	return l
+}
+func (rr *HINFO) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Cpu) + 1
+	l += len(rr.Os) + 1
+	return l
+}
+func (rr *HIP) len() int {
+	l := rr.Hdr.len()
+	l++    // HitLength
+	l++    // PublicKeyAlgorithm
+	l += 2 // PublicKeyLength
+	l += len(rr.Hit) / 2
+	l += base64.StdEncoding.DecodedLen(len(rr.PublicKey))
+	for _, x := range rr.RendezvousServers {
+		l += len(x) + 1
+	}
+	return l
+}
+func (rr *KX) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Preference
+	l += len(rr.Exchanger) + 1
+	return l
+}
+func (rr *L32) len() int {
+	l := rr.Hdr.len()
+	l += 2           // Preference
+	l += net.IPv4len // Locator32
+	return l
+}
+func (rr *L64) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Preference
+	l += 8 // Locator64
+	return l
+}
+func (rr *LOC) len() int {
+	l := rr.Hdr.len()
+	l++    // Version
+	l++    // Size
+	l++    // HorizPre
+	l++    // VertPre
+	l += 4 // Latitude
+	l += 4 // Longitude
+	l += 4 // Altitude
+	return l
+}
+func (rr *LP) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Preference
+	l += len(rr.Fqdn) + 1
+	return l
+}
+func (rr *MB) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Mb) + 1
+	return l
+}
+func (rr *MD) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Md) + 1
+	return l
+}
+func (rr *MF) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Mf) + 1
+	return l
+}
+func (rr *MG) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Mg) + 1
+	return l
+}
+func (rr *MINFO) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Rmail) + 1
+	l += len(rr.Email) + 1
+	return l
+}
+func (rr *MR) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Mr) + 1
+	return l
+}
+func (rr *MX) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Preference
+	l += len(rr.Mx) + 1
+	return l
+}
+func (rr *NAPTR) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Order
+	l += 2 // Preference
+	l += len(rr.Flags) + 1
+	l += len(rr.Service) + 1
+	l += len(rr.Regexp) + 1
+	l += len(rr.Replacement) + 1
+	return l
+}
+func (rr *NID) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Preference
+	l += 8 // NodeID
+	return l
+}
+func (rr *NIMLOC) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Locator)/2 + 1
+	return l
+}
+func (rr *NINFO) len() int {
+	l := rr.Hdr.len()
+	for _, x := range rr.ZSData {
+		l += len(x) + 1
+	}
+	return l
+}
+func (rr *NS) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Ns) + 1
+	return l
+}
+func (rr *NSAPPTR) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Ptr) + 1
+	return l
+}
+func (rr *NSEC3PARAM) len() int {
+	l := rr.Hdr.len()
+	l++    // Hash
+	l++    // Flags
+	l += 2 // Iterations
+	l++    // SaltLength
+	l += len(rr.Salt) / 2
+	return l
+}
+func (rr *OPENPGPKEY) len() int {
+	l := rr.Hdr.len()
+	l += base64.StdEncoding.DecodedLen(len(rr.PublicKey))
+	return l
+}
+func (rr *PTR) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Ptr) + 1
+	return l
+}
+func (rr *PX) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Preference
+	l += len(rr.Map822) + 1
+	l += len(rr.Mapx400) + 1
+	return l
+}
+func (rr *RFC3597) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Rdata)/2 + 1
+	return l
+}
+func (rr *RKEY) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Flags
+	l++    // Protocol
+	l++    // Algorithm
+	l += base64.StdEncoding.DecodedLen(len(rr.PublicKey))
+	return l
+}
+func (rr *RP) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Mbox) + 1
+	l += len(rr.Txt) + 1
+	return l
+}
+func (rr *RRSIG) len() int {
+	l := rr.Hdr.len()
+	l += 2 // TypeCovered
+	l++    // Algorithm
+	l++    // Labels
+	l += 4 // OrigTtl
+	l += 4 // Expiration
+	l += 4 // Inception
+	l += 2 // KeyTag
+	l += len(rr.SignerName) + 1
+	l += base64.StdEncoding.DecodedLen(len(rr.Signature))
+	return l
+}
+func (rr *RT) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Preference
+	l += len(rr.Host) + 1
+	return l
+}
+func (rr *SMIMEA) len() int {
+	l := rr.Hdr.len()
+	l++ // Usage
+	l++ // Selector
+	l++ // MatchingType
+	l += len(rr.Certificate)/2 + 1
+	return l
+}
+func (rr *SOA) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Ns) + 1
+	l += len(rr.Mbox) + 1
+	l += 4 // Serial
+	l += 4 // Refresh
+	l += 4 // Retry
+	l += 4 // Expire
+	l += 4 // Minttl
+	return l
+}
+func (rr *SPF) len() int {
+	l := rr.Hdr.len()
+	for _, x := range rr.Txt {
+		l += len(x) + 1
+	}
+	return l
+}
+func (rr *SRV) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Priority
+	l += 2 // Weight
+	l += 2 // Port
+	l += len(rr.Target) + 1
+	return l
+}
+func (rr *SSHFP) len() int {
+	l := rr.Hdr.len()
+	l++ // Algorithm
+	l++ // Type
+	l += len(rr.FingerPrint)/2 + 1
+	return l
+}
+func (rr *TA) len() int {
+	l := rr.Hdr.len()
+	l += 2 // KeyTag
+	l++    // Algorithm
+	l++    // DigestType
+	l += len(rr.Digest)/2 + 1
+	return l
+}
+func (rr *TALINK) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.PreviousName) + 1
+	l += len(rr.NextName) + 1
+	return l
+}
+func (rr *TKEY) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Algorithm) + 1
+	l += 4 // Inception
+	l += 4 // Expiration
+	l += 2 // Mode
+	l += 2 // Error
+	l += 2 // KeySize
+	l += len(rr.Key) / 2
+	l += 2 // OtherLen
+	l += len(rr.OtherData) / 2
+	return l
+}
+func (rr *TLSA) len() int {
+	l := rr.Hdr.len()
+	l++ // Usage
+	l++ // Selector
+	l++ // MatchingType
+	l += len(rr.Certificate)/2 + 1
+	return l
+}
+func (rr *TSIG) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Algorithm) + 1
+	l += 6 // TimeSigned
+	l += 2 // Fudge
+	l += 2 // MACSize
+	l += len(rr.MAC) / 2
+	l += 2 // OrigId
+	l += 2 // Error
+	l += 2 // OtherLen
+	l += len(rr.OtherData) / 2
+	return l
+}
+func (rr *TXT) len() int {
+	l := rr.Hdr.len()
+	for _, x := range rr.Txt {
+		l += len(x) + 1
+	}
+	return l
+}
+func (rr *UID) len() int {
+	l := rr.Hdr.len()
+	l += 4 // Uid
+	return l
+}
+func (rr *UINFO) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.Uinfo) + 1
+	return l
+}
+func (rr *URI) len() int {
+	l := rr.Hdr.len()
+	l += 2 // Priority
+	l += 2 // Weight
+	l += len(rr.Target)
+	return l
+}
+func (rr *X25) len() int {
+	l := rr.Hdr.len()
+	l += len(rr.PSDNAddress) + 1
+	return l
+}
+
+// copy() functions
+func (rr *A) copy() RR {
+	return &A{*rr.Hdr.copyHeader(), copyIP(rr.A)}
+}
+func (rr *AAAA) copy() RR {
+	return &AAAA{*rr.Hdr.copyHeader(), copyIP(rr.AAAA)}
+}
+func (rr *AFSDB) copy() RR {
+	return &AFSDB{*rr.Hdr.copyHeader(), rr.Subtype, rr.Hostname}
+}
+func (rr *ANY) copy() RR {
+	return &ANY{*rr.Hdr.copyHeader()}
+}
+func (rr *AVC) copy() RR {
+	Txt := make([]string, len(rr.Txt))
+	copy(Txt, rr.Txt)
+	return &AVC{*rr.Hdr.copyHeader(), Txt}
+}
+func (rr *CAA) copy() RR {
+	return &CAA{*rr.Hdr.copyHeader(), rr.Flag, rr.Tag, rr.Value}
+}
+func (rr *CERT) copy() RR {
+	return &CERT{*rr.Hdr.copyHeader(), rr.Type, rr.KeyTag, rr.Algorithm, rr.Certificate}
+}
+func (rr *CNAME) copy() RR {
+	return &CNAME{*rr.Hdr.copyHeader(), rr.Target}
+}
+func (rr *CSYNC) copy() RR {
+	TypeBitMap := make([]uint16, len(rr.TypeBitMap))
+	copy(TypeBitMap, rr.TypeBitMap)
+	return &CSYNC{*rr.Hdr.copyHeader(), rr.Serial, rr.Flags, TypeBitMap}
+}
+func (rr *DHCID) copy() RR {
+	return &DHCID{*rr.Hdr.copyHeader(), rr.Digest}
+}
+func (rr *DNAME) copy() RR {
+	return &DNAME{*rr.Hdr.copyHeader(), rr.Target}
+}
+func (rr *DNSKEY) copy() RR {
+	return &DNSKEY{*rr.Hdr.copyHeader(), rr.Flags, rr.Protocol, rr.Algorithm, rr.PublicKey}
+}
+func (rr *DS) copy() RR {
+	return &DS{*rr.Hdr.copyHeader(), rr.KeyTag, rr.Algorithm, rr.DigestType, rr.Digest}
+}
+func (rr *EID) copy() RR {
+	return &EID{*rr.Hdr.copyHeader(), rr.Endpoint}
+}
+func (rr *EUI48) copy() RR {
+	return &EUI48{*rr.Hdr.copyHeader(), rr.Address}
+}
+func (rr *EUI64) copy() RR {
+	return &EUI64{*rr.Hdr.copyHeader(), rr.Address}
+}
+func (rr *GID) copy() RR {
+	return &GID{*rr.Hdr.copyHeader(), rr.Gid}
+}
+func (rr *GPOS) copy() RR {
+	return &GPOS{*rr.Hdr.copyHeader(), rr.Longitude, rr.Latitude, rr.Altitude}
+}
+func (rr *HINFO) copy() RR {
+	return &HINFO{*rr.Hdr.copyHeader(), rr.Cpu, rr.Os}
+}
+func (rr *HIP) copy() RR {
+	RendezvousServers := make([]string, len(rr.RendezvousServers))
+	copy(RendezvousServers, rr.RendezvousServers)
+	return &HIP{*rr.Hdr.copyHeader(), rr.HitLength, rr.PublicKeyAlgorithm, rr.PublicKeyLength, rr.Hit, rr.PublicKey, RendezvousServers}
+}
+func (rr *KX) copy() RR {
+	return &KX{*rr.Hdr.copyHeader(), rr.Preference, rr.Exchanger}
+}
+func (rr *L32) copy() RR {
+	return &L32{*rr.Hdr.copyHeader(), rr.Preference, copyIP(rr.Locator32)}
+}
+func (rr *L64) copy() RR {
+	return &L64{*rr.Hdr.copyHeader(), rr.Preference, rr.Locator64}
+}
+func (rr *LOC) copy() RR {
+	return &LOC{*rr.Hdr.copyHeader(), rr.Version, rr.Size, rr.HorizPre, rr.VertPre, rr.Latitude, rr.Longitude, rr.Altitude}
+}
+func (rr *LP) copy() RR {
+	return &LP{*rr.Hdr.copyHeader(), rr.Preference, rr.Fqdn}
+}
+func (rr *MB) copy() RR {
+	return &MB{*rr.Hdr.copyHeader(), rr.Mb}
+}
+func (rr *MD) copy() RR {
+	return &MD{*rr.Hdr.copyHeader(), rr.Md}
+}
+func (rr *MF) copy() RR {
+	return &MF{*rr.Hdr.copyHeader(), rr.Mf}
+}
+func (rr *MG) copy() RR {
+	return &MG{*rr.Hdr.copyHeader(), rr.Mg}
+}
+func (rr *MINFO) copy() RR {
+	return &MINFO{*rr.Hdr.copyHeader(), rr.Rmail, rr.Email}
+}
+func (rr *MR) copy() RR {
+	return &MR{*rr.Hdr.copyHeader(), rr.Mr}
+}
+func (rr *MX) copy() RR {
+	return &MX{*rr.Hdr.copyHeader(), rr.Preference, rr.Mx}
+}
+func (rr *NAPTR) copy() RR {
+	return &NAPTR{*rr.Hdr.copyHeader(), rr.Order, rr.Preference, rr.Flags, rr.Service, rr.Regexp, rr.Replacement}
+}
+func (rr *NID) copy() RR {
+	return &NID{*rr.Hdr.copyHeader(), rr.Preference, rr.NodeID}
+}
+func (rr *NIMLOC) copy() RR {
+	return &NIMLOC{*rr.Hdr.copyHeader(), rr.Locator}
+}
+func (rr *NINFO) copy() RR {
+	ZSData := make([]string, len(rr.ZSData))
+	copy(ZSData, rr.ZSData)
+	return &NINFO{*rr.Hdr.copyHeader(), ZSData}
+}
+func (rr *NS) copy() RR {
+	return &NS{*rr.Hdr.copyHeader(), rr.Ns}
+}
+func (rr *NSAPPTR) copy() RR {
+	return &NSAPPTR{*rr.Hdr.copyHeader(), rr.Ptr}
+}
+func (rr *NSEC) copy() RR {
+	TypeBitMap := make([]uint16, len(rr.TypeBitMap))
+	copy(TypeBitMap, rr.TypeBitMap)
+	return &NSEC{*rr.Hdr.copyHeader(), rr.NextDomain, TypeBitMap}
+}
+func (rr *NSEC3) copy() RR {
+	TypeBitMap := make([]uint16, len(rr.TypeBitMap))
+	copy(TypeBitMap, rr.TypeBitMap)
+	return &NSEC3{*rr.Hdr.copyHeader(), rr.Hash, rr.Flags, rr.Iterations, rr.SaltLength, rr.Salt, rr.HashLength, rr.NextDomain, TypeBitMap}
+}
+func (rr *NSEC3PARAM) copy() RR {
+	return &NSEC3PARAM{*rr.Hdr.copyHeader(), rr.Hash, rr.Flags, rr.Iterations, rr.SaltLength, rr.Salt}
+}
+func (rr *OPENPGPKEY) copy() RR {
+	return &OPENPGPKEY{*rr.Hdr.copyHeader(), rr.PublicKey}
+}
+func (rr *OPT) copy() RR {
+	Option := make([]EDNS0, len(rr.Option))
+	copy(Option, rr.Option)
+	return &OPT{*rr.Hdr.copyHeader(), Option}
+}
+func (rr *PTR) copy() RR {
+	return &PTR{*rr.Hdr.copyHeader(), rr.Ptr}
+}
+func (rr *PX) copy() RR {
+	return &PX{*rr.Hdr.copyHeader(), rr.Preference, rr.Map822, rr.Mapx400}
+}
+func (rr *RFC3597) copy() RR {
+	return &RFC3597{*rr.Hdr.copyHeader(), rr.Rdata}
+}
+func (rr *RKEY) copy() RR {
+	return &RKEY{*rr.Hdr.copyHeader(), rr.Flags, rr.Protocol, rr.Algorithm, rr.PublicKey}
+}
+func (rr *RP) copy() RR {
+	return &RP{*rr.Hdr.copyHeader(), rr.Mbox, rr.Txt}
+}
+func (rr *RRSIG) copy() RR {
+	return &RRSIG{*rr.Hdr.copyHeader(), rr.TypeCovered, rr.Algorithm, rr.Labels, rr.OrigTtl, rr.Expiration, rr.Inception, rr.KeyTag, rr.SignerName, rr.Signature}
+}
+func (rr *RT) copy() RR {
+	return &RT{*rr.Hdr.copyHeader(), rr.Preference, rr.Host}
+}
+func (rr *SMIMEA) copy() RR {
+	return &SMIMEA{*rr.Hdr.copyHeader(), rr.Usage, rr.Selector, rr.MatchingType, rr.Certificate}
+}
+func (rr *SOA) copy() RR {
+	return &SOA{*rr.Hdr.copyHeader(), rr.Ns, rr.Mbox, rr.Serial, rr.Refresh, rr.Retry, rr.Expire, rr.Minttl}
+}
+func (rr *SPF) copy() RR {
+	Txt := make([]string, len(rr.Txt))
+	copy(Txt, rr.Txt)
+	return &SPF{*rr.Hdr.copyHeader(), Txt}
+}
+func (rr *SRV) copy() RR {
+	return &SRV{*rr.Hdr.copyHeader(), rr.Priority, rr.Weight, rr.Port, rr.Target}
+}
+func (rr *SSHFP) copy() RR {
+	return &SSHFP{*rr.Hdr.copyHeader(), rr.Algorithm, rr.Type, rr.FingerPrint}
+}
+func (rr *TA) copy() RR {
+	return &TA{*rr.Hdr.copyHeader(), rr.KeyTag, rr.Algorithm, rr.DigestType, rr.Digest}
+}
+func (rr *TALINK) copy() RR {
+	return &TALINK{*rr.Hdr.copyHeader(), rr.PreviousName, rr.NextName}
+}
+func (rr *TKEY) copy() RR {
+	return &TKEY{*rr.Hdr.copyHeader(), rr.Algorithm, rr.Inception, rr.Expiration, rr.Mode, rr.Error, rr.KeySize, rr.Key, rr.OtherLen, rr.OtherData}
+}
+func (rr *TLSA) copy() RR {
+	return &TLSA{*rr.Hdr.copyHeader(), rr.Usage, rr.Selector, rr.MatchingType, rr.Certificate}
+}
+func (rr *TSIG) copy() RR {
+	return &TSIG{*rr.Hdr.copyHeader(), rr.Algorithm, rr.TimeSigned, rr.Fudge, rr.MACSize, rr.MAC, rr.OrigId, rr.Error, rr.OtherLen, rr.OtherData}
+}
+func (rr *TXT) copy() RR {
+	Txt := make([]string, len(rr.Txt))
+	copy(Txt, rr.Txt)
+	return &TXT{*rr.Hdr.copyHeader(), Txt}
+}
+func (rr *UID) copy() RR {
+	return &UID{*rr.Hdr.copyHeader(), rr.Uid}
+}
+func (rr *UINFO) copy() RR {
+	return &UINFO{*rr.Hdr.copyHeader(), rr.Uinfo}
+}
+func (rr *URI) copy() RR {
+	return &URI{*rr.Hdr.copyHeader(), rr.Priority, rr.Weight, rr.Target}
+}
+func (rr *X25) copy() RR {
+	return &X25{*rr.Hdr.copyHeader(), rr.PSDNAddress}
+}
--- a/vendor/golang.org/x/crypto/AUTHORS
+++ b/vendor/golang.org/x/crypto/AUTHORS
@ -1,3 +1,3 @@
 # This source code refers to The Go Authors for copyright purposes.
 # The master list of authors is in the main Go distribution,
-# visible at http://tip.golang.org/AUTHORS.
+# visible at https://tip.golang.org/AUTHORS.
--- a/vendor/golang.org/x/crypto/CONTRIBUTORS
+++ b/vendor/golang.org/x/crypto/CONTRIBUTORS
@ -1,3 +1,3 @@
 # This source code was written by the Go contributors.
 # The master list of contributors is in the main Go distribution,
-# visible at http://tip.golang.org/CONTRIBUTORS.
+# visible at https://tip.golang.org/CONTRIBUTORS.
--- a/vendor/golang.org/x/crypto/LICENSE
+++ b/vendor/golang.org/x/crypto/LICENSE
--- a/vendor/golang.org/x/crypto/PATENTS
+++ b/vendor/golang.org/x/crypto/PATENTS
--- a/vendor/golang.org/x/crypto/blowfish/block.go
+++ b/vendor/golang.org/x/crypto/blowfish/block.go
--- a/vendor/golang.org/x/crypto/blowfish/cipher.go
+++ b/vendor/golang.org/x/crypto/blowfish/cipher.go
@ -6,7 +6,7 @@
 package blowfish

 // The code is a port of Bruce Schneier's C implementation.
-// See http://www.schneier.com/blowfish.html.
+// See https://www.schneier.com/blowfish.html.

 import "strconv"

--- a/vendor/golang.org/x/crypto/blowfish/const.go
+++ b/vendor/golang.org/x/crypto/blowfish/const.go
@ -4,7 +4,7 @@

 // The startup permutation array and substitution boxes.
 // They are the hexadecimal digits of PI; see:
-// http://www.schneier.com/code/constants.txt.
+// https://www.schneier.com/code/constants.txt.

 package blowfish

--- a/vendor/golang.org/x/crypto/cast5/cast5.go
+++ b/vendor/golang.org/x/crypto/cast5/cast5.go
--- a/vendor/golang.org/x/crypto/curve25519/const_amd64.h
+++ b/vendor/golang.org/x/crypto/curve25519/const_amd64.h
@ -3,6 +3,6 @@
 // license that can be found in the LICENSE file.

 // This code was translated into a form compatible with 6a from the public
-// domain sources in SUPERCOP: http://bench.cr.yp.to/supercop.html
+// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html

 #define REDMASK51     0x0007FFFFFFFFFFFF
--- a/vendor/golang.org/x/crypto/curve25519/const_amd64.s
+++ b/vendor/golang.org/x/crypto/curve25519/const_amd64.s
@ -3,7 +3,7 @@
 // license that can be found in the LICENSE file.

 // This code was translated into a form compatible with 6a from the public
-// domain sources in SUPERCOP: http://bench.cr.yp.to/supercop.html
+// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html

 // +build amd64,!gccgo,!appengine

--- a/vendor/golang.org/x/crypto/curve25519/cswap_amd64.s
+++ b/vendor/golang.org/x/crypto/curve25519/cswap_amd64.s
--- a/vendor/golang.org/x/crypto/curve25519/curve25519.go
+++ b/vendor/golang.org/x/crypto/curve25519/curve25519.go
@ -2,7 +2,7 @@
 // Use of this source code is governed by a BSD-style
 // license that can be found in the LICENSE file.

-// We have a implementation in amd64 assembly so this code is only run on
+// We have an implementation in amd64 assembly so this code is only run on
 // non-amd64 platforms. The amd64 assembly does not support gccgo.
 // +build !amd64 gccgo appengine

--- a/vendor/golang.org/x/crypto/curve25519/doc.go
+++ b/vendor/golang.org/x/crypto/curve25519/doc.go
@ -3,7 +3,7 @@
 // license that can be found in the LICENSE file.

 // Package curve25519 provides an implementation of scalar multiplication on
-// the elliptic curve known as curve25519. See http://cr.yp.to/ecdh.html
+// the elliptic curve known as curve25519. See https://cr.yp.to/ecdh.html
 package curve25519

 // basePoint is the x coordinate of the generator of the curve.
--- a/vendor/golang.org/x/crypto/curve25519/freeze_amd64.s
+++ b/vendor/golang.org/x/crypto/curve25519/freeze_amd64.s
@ -3,7 +3,7 @@
 // license that can be found in the LICENSE file.

 // This code was translated into a form compatible with 6a from the public
-// domain sources in SUPERCOP: http://bench.cr.yp.to/supercop.html
+// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html

 // +build amd64,!gccgo,!appengine

--- a/vendor/golang.org/x/crypto/curve25519/ladderstep_amd64.s
+++ b/vendor/golang.org/x/crypto/curve25519/ladderstep_amd64.s
@ -3,7 +3,7 @@
 // license that can be found in the LICENSE file.

 // This code was translated into a form compatible with 6a from the public
-// domain sources in SUPERCOP: http://bench.cr.yp.to/supercop.html
+// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html

 // +build amd64,!gccgo,!appengine

--- a/vendor/golang.org/x/crypto/curve25519/mont25519_amd64.go
+++ b/vendor/golang.org/x/crypto/curve25519/mont25519_amd64.go
--- a/vendor/golang.org/x/crypto/curve25519/mul_amd64.s
+++ b/vendor/golang.org/x/crypto/curve25519/mul_amd64.s
@ -3,7 +3,7 @@
 // license that can be found in the LICENSE file.

 // This code was translated into a form compatible with 6a from the public
-// domain sources in SUPERCOP: http://bench.cr.yp.to/supercop.html
+// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html

 // +build amd64,!gccgo,!appengine

--- a/vendor/golang.org/x/crypto/curve25519/square_amd64.s
+++ b/vendor/golang.org/x/crypto/curve25519/square_amd64.s
@ -3,7 +3,7 @@
 // license that can be found in the LICENSE file.

 // This code was translated into a form compatible with 6a from the public
-// domain sources in SUPERCOP: http://bench.cr.yp.to/supercop.html
+// domain sources in SUPERCOP: https://bench.cr.yp.to/supercop.html

 // +build amd64,!gccgo,!appengine

--- a/vendor/golang.org/x/crypto/ed25519/ed25519.go
+++ b/vendor/golang.org/x/crypto/ed25519/ed25519.go
@ -3,20 +3,20 @@
 // license that can be found in the LICENSE file.

 // Package ed25519 implements the Ed25519 signature algorithm. See
-// http://ed25519.cr.yp.to/.
+// https://ed25519.cr.yp.to/.
 //
 // These functions are also compatible with the “Ed25519” function defined in
-// https://tools.ietf.org/html/draft-irtf-cfrg-eddsa-05.
+// RFC 8032.
 package ed25519

 // This code is a port of the public domain, “ref10” implementation of ed25519
 // from SUPERCOP.

 import (
+	"bytes"
 	"crypto"
 	cryptorand "crypto/rand"
 	"crypto/sha512"
-	"crypto/subtle"
 	"errors"
 	"io"
 	"strconv"
@ -177,5 +177,5 @@ func Verify(publicKey PublicKey, message, sig []byte) bool {

 	var checkR [32]byte
 	R.ToBytes(&checkR)
-	return subtle.ConstantTimeCompare(sig[:32], checkR[:]) == 1
+	return bytes.Equal(sig[:32], checkR[:])
 }
--- a/vendor/golang.org/x/crypto/ed25519/internal/edwards25519/const.go
+++ b/vendor/golang.org/x/crypto/ed25519/internal/edwards25519/const.go
--- a/vendor/golang.org/x/crypto/ed25519/internal/edwards25519/edwards25519.go
+++ b/vendor/golang.org/x/crypto/ed25519/internal/edwards25519/edwards25519.go
--- a/Show More
+++ b/Show More
Author	SHA1	Message	Date
arraykeys@gmail.com	7df2d990e7	Merge branch 'dev'	2018-03-28 14:10:10 +08:00
snail007	900b75ddcd	Merge pull request #48 from yincongcyincong/dev Dev	2018-03-28 14:08:28 +08:00
yincongcyincong	a5d199fb1c	Update README.md	2018-03-28 14:06:19 +08:00
yincongcyincong	2d8190873f	Update README.md	2018-03-28 13:56:02 +08:00
yincongcyincong	3eff793ac2	Update README.md	2018-03-28 13:53:09 +08:00
arraykeys@gmail.com	27ce6e1bd2	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:34:08 +08:00
arraykeys@gmail.com	b11c7e632b	fix #47 Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:30:09 +08:00
arraykeys@gmail.com	d911be7d80	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:27:59 +08:00
arraykeys@gmail.com	bd056d74cc	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:25:47 +08:00
arraykeys@gmail.com	69b65a37ca	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:25:17 +08:00
arraykeys@gmail.com	dd52ad8a8a	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:22:37 +08:00
arraykeys@gmail.com	59b5ef2df4	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:21:02 +08:00
arraykeys@gmail.com	4a34566f08	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-28 13:14:06 +08:00
arraykeys@gmail.com	d81d5ffe06	socks5 client server done Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-26 18:42:44 +08:00
arraykeys@gmail.com	59c9148875	optimize timeout of http(s)\socks\nat Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-21 18:17:20 +08:00
arraykeys@gmail.com	40bce3e736	optimize http basic auth Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-21 16:57:46 +08:00
arraykeys@gmail.com	d4c0775b4a	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-16 16:16:11 +08:00
arraykeys@gmail.com	7f983152b7	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-14 09:37:57 +08:00
arraykeys@gmail.com	bab4325414	server add safe close user conn Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-13 18:36:34 +08:00
arraykeys@gmail.com	0d85c7dd7d	bridge add timeout Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-13 17:31:51 +08:00
arraykeys@gmail.com	f756d62b19	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-13 17:14:25 +08:00
arraykeys@gmail.com	2d1c1449aa	Merge branch 'master' of https://github.com/snail007/goproxy.git	2018-03-13 10:29:50 +08:00
arraykeys@gmail.com	f348298acd	v4.5	2018-03-13 10:29:26 +08:00
arraykeys@gmail.com	ad47104fc7	v4.5	2018-03-13 10:04:18 +08:00
snail007	4d4fb64b59	Update mux_bridge.go	2018-03-12 19:27:37 +08:00
arraykeys@gmail.com	9ce3a6e468	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-12 18:15:14 +08:00
arraykeys@gmail.com	34e9e362b9	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-12 17:31:35 +08:00
arraykeys@gmail.com	f87cbf73e8	Merge branch 'dev' of https://github.com/snail007/goproxy.git into dev	2018-03-12 13:59:47 +08:00
arraykeys@gmail.com	362ada2ebb	v4.5 Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-12 13:59:25 +08:00
snail007	b4b7221dab	Merge pull request #44 from snail007/dev Update README_ZH.md	2018-03-10 17:00:50 +08:00
snail007	2d8dc56f4e	Update README_ZH.md	2018-03-10 17:00:03 +08:00
arraykeys@gmail.com	1cf4313d12	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-09 17:16:21 +08:00
arraykeys@gmail.com	7bb8f19b90	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-09 16:02:05 +08:00
arraykeys@gmail.com	1263a4e751	Merge branch 'dev' of https://github.com/snail007/goproxy.git into dev	2018-03-09 09:50:38 +08:00
arraykeys@gmail.com	1a432a9b79	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-09 09:50:21 +08:00
snail007	8951fdbd59	Merge pull request #42 from yincongcyincong/dev Update README.md	2018-03-09 08:36:10 +08:00
yincongcyincong	77129367fe	Update README.md	2018-03-08 20:50:36 +08:00
arraykeys@gmail.com	ae2e1e0933	Merge branch 'dev' of https://github.com/snail007/goproxy.git into dev	2018-03-08 18:43:07 +08:00
arraykeys@gmail.com	4143f14fbd	optimise nat forwarding in different lan Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 18:42:50 +08:00
snail007	8e3e262c2f	Merge pull request #41 from yincongcyincong/dev Dev	2018-03-08 13:46:47 +08:00
arraykeys@gmail.com	6f11deab96	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 13:45:54 +08:00
arraykeys@gmail.com	a17acd7351	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 12:08:25 +08:00
arraykeys@gmail.com	1cbb4195e4	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 12:08:02 +08:00
arraykeys@gmail.com	c471dd8297	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 12:07:44 +08:00
arraykeys@gmail.com	25deffb7d6	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 12:07:19 +08:00
arraykeys@gmail.com	7eb0e0040e	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 12:06:42 +08:00
arraykeys@gmail.com	15994988be	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 12:04:26 +08:00
arraykeys@gmail.com	ae293a6102	Merge branch 'dev' of https://github.com/snail007/goproxy.git into dev	2018-03-08 11:52:06 +08:00
arraykeys@gmail.com	5ed4702b62	add kcp config args Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-08 11:51:34 +08:00
yc	287ddc3424	Merge branch 'snail007-dev' into dev	2018-03-08 09:56:52 +08:00
yc	57a4227007	fix conflick	2018-03-08 09:56:21 +08:00
yc	c9eacd1bf2	Merge branch 'dev' of https://github.com/yincongcyincong/goproxy into dev	2018-03-08 09:50:59 +08:00
yincongcyincong	5f38162fbb	Update README.md	2018-03-07 23:30:05 +08:00
snail007	c755f75a11	Update README_ZH.md	2018-03-07 23:09:08 +08:00
yincongcyincong	ac9eb64501	Update README.md	2018-03-07 13:34:16 +08:00
yincongcyincong	3dd013c13c	Update README.md	2018-03-07 08:01:32 +08:00
yincongcyincong	ab0205587a	Update README.md	2018-03-06 18:28:54 +08:00
arraykeys@gmail.com	70955878c9	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 18:11:13 +08:00
arraykeys@gmail.com	86f017d92f	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:38:37 +08:00
arraykeys@gmail.com	446cc3f9a7	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:37:23 +08:00
arraykeys@gmail.com	6529921d71	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:24:28 +08:00
arraykeys@gmail.com	52e441c111	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:21:44 +08:00
arraykeys@gmail.com	a6b169d336	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:20:33 +08:00
arraykeys@gmail.com	5514cfee6c	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:19:34 +08:00
arraykeys@gmail.com	b1de184bda	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:17:09 +08:00
arraykeys@gmail.com	af2405ba48	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:15:34 +08:00
arraykeys@gmail.com	bee80330b0	sps done Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-06 17:13:43 +08:00
arraykeys@gmail.com	f1de8659b7	prepare for sps Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-05 18:46:19 +08:00
arraykeys@gmail.com	bfc5835d82	Merge branch 'master' of https://github.com/snail007/goproxy.git into dev	2018-03-05 11:22:33 +08:00
arraykeys@gmail.com	82bc3e27d6	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-05 11:22:15 +08:00
arraykeys@gmail.com	5436a95430	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-05 11:15:30 +08:00
arraykeys@gmail.com	2c675f2cbe	Merge branch 'dev' of https://github.com/snail007/goproxy.git into dev	2018-03-05 09:46:07 +08:00
arraykeys@gmail.com	8e9427b0c0	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-05 09:46:00 +08:00
arraykeys@gmail.com	edcf78f77c	Merge branch 'master' into dev	2018-03-05 09:45:42 +08:00
snail007	8f88d14c07	Update README.md	2018-03-01 19:40:33 +08:00
snail007	1372801b6f	Update README_ZH.md	2018-03-01 19:39:08 +08:00
arraykeys@gmail.com	2af904f442	fix dns Resolve Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-01 15:01:08 +08:00
arraykeys@gmail.com	885b27e0d1	Merge branch 'dev' of https://github.com/snail007/goproxy.git into dev	2018-03-01 13:55:53 +08:00
arraykeys@gmail.com	0207e4731f	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-03-01 13:55:25 +08:00
arraykeys@gmail.com	038b6749a3	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-02-28 17:55:18 +08:00
snail007	ead577cbfb	Merge pull request #40 from yincongcyincong/dev Dev	2018-02-28 16:28:10 +08:00
yincongcyincong	26e9231e48	Update README.md	2018-02-28 16:23:30 +08:00
yincongcyincong	938ddd1141	Update README.md	2018-02-28 16:15:29 +08:00
yincongcyincong	68080539f7	Update README.md	2018-02-28 16:14:16 +08:00
yincongcyincong	5583b303be	Update README.md	2018-02-28 16:11:43 +08:00
yincongcyincong	9301c9b49b	Update README.md	2018-02-28 16:07:40 +08:00
yincongcyincong	8d2e210522	Update README.md	2018-02-28 16:05:47 +08:00
yincongcyincong	32661552ff	Update README.md	2018-02-28 14:09:23 +08:00
arraykeys@gmail.com	9a111a59bf	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-02-28 13:40:14 +08:00
arraykeys@gmail.com	f5e472ea9f	4.3 Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-02-28 13:27:13 +08:00
arraykeys@gmail.com	7599e2c793	add socks dns support	2018-02-27 18:33:22 +08:00
arraykeys@gmail.com	3726f5b9c3	update vendor Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-02-27 10:33:46 +08:00
arraykeys@gmail.com	dee517217e	add dns support Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-02-26 18:44:14 +08:00
arraykeys@gmail.com	983912e44e	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-02-26 10:33:12 +08:00
arraykeys@gmail.com	982390f4b2	Merge branch 'dev' of https://github.com/snail007/goproxy.git into dev	2018-02-26 10:28:01 +08:00
arraykeys@gmail.com	5e3f51a8b0	Signed-off-by: arraykeys@gmail.com <arraykeys@gmail.com>	2018-02-26 10:26:58 +08:00
snail007	675ae276f9	Merge pull request #38 from snail007/dev Update README.md	2018-02-20 17:41:29 +08:00
snail007	7899c45176	Update README.md	2018-02-20 17:40:53 +08:00
snail007	1dbb6feb57	Merge pull request #37 from snail007/dev update readme	2018-02-13 12:46:38 +08:00
snail007	661685d136	Merge pull request #36 from yincongcyincong/dev Update README.md	2018-02-13 12:44:40 +08:00
yincongcyincong	28947a0352	Update README.md	2018-02-13 08:20:18 +08:00
yincongcyincong	bd594684ce	Update README.md	2018-02-13 08:18:40 +08:00
snail007	fda3609873	Merge pull request #35 from snail007/dev Update README.md	2018-02-08 23:42:44 +08:00
snail007	d20db0c546	Update README.md	2018-02-08 23:42:04 +08:00
snail007	07fb22ae70	Merge pull request #34 from snail007/dev add tg	2018-02-08 23:41:03 +08:00
snail007	8d5c3944ad	Update README.md	2018-02-08 23:40:29 +08:00
snail007	7cf28aa9f4	Update README_ZH.md	2018-02-08 23:36:43 +08:00
snail007	731867b73c	Merge pull request #33 from snail007/dev Update README.md	2018-02-07 23:08:59 +08:00
snail007	1c6df2d9a2	Update README.md	2018-02-07 23:08:01 +08:00
snail007	ee5a248a39	Merge pull request #32 from snail007/dev update docs	2018-02-07 23:05:53 +08:00
snail007	c174f85656	Merge pull request #31 from yincongcyincong/dev Update README.md	2018-02-07 23:05:15 +08:00
yincongcyincong	2669aac7c9	Update README.md	2018-02-07 16:09:25 +08:00
snail007	f1aec74b11	Merge pull request #30 from snail007/dev update en docs	2018-02-06 14:57:06 +08:00
snail007	1d382b2bf6	Merge pull request #29 from yincongcyincong/dev Update README.md	2018-02-06 14:56:31 +08:00
yincongcyincong	55cac537b1	Update README.md	2018-02-06 14:50:31 +08:00
snail007	75258fa195	Merge pull request #28 from snail007/dev update en_docs	2018-02-06 14:43:31 +08:00
snail007	cef2ca6d8e	Merge pull request #27 from yincongcyincong/dev Update README.md	2018-02-06 14:40:17 +08:00
yincongcyincong	4b1651bb3e	Update README.md	2018-02-06 11:33:06 +08:00
snail007	78c116bca9	Merge pull request #26 from yincongcyincong/dev en readme	2018-02-06 11:14:41 +08:00
yincongcyincong	a7c46f5582	Update README.md	2018-02-06 08:09:16 +08:00
yincongcyincong	f947d35bc3	Update README.md	2018-02-05 18:00:50 +08:00
yincongcyincong	a49e0166d4	Update README.md	2018-02-05 17:50:39 +08:00
yincongcyincong	7272b592d5	Update README.md	2018-02-05 11:35:50 +08:00
snail007	cced739d0e	Update README_ZH.md	2018-02-05 10:57:07 +08:00
snail007	54ac46b3e4	Update README_ZH.md	2018-02-05 10:56:32 +08:00
snail007	8f9aa2fd64	Update README_ZH.md	2018-02-05 10:56:03 +08:00
snail007	c7b9cd5853	Update README_ZH.md	2018-02-05 10:55:28 +08:00
snail007	f9dfac55b0	Update README_ZH.md	2018-02-05 10:54:51 +08:00
snail007	b4ad1b5465	Update README_ZH.md	2018-02-05 10:54:11 +08:00
snail007	e2b2b7e255	Update README_ZH.md	2018-02-05 10:53:43 +08:00
snail007	80b691564c	Update README_ZH.md	2018-02-05 10:52:25 +08:00
yincongcyincong	dfc326b771	Update README.md	2018-02-01 18:01:10 +08:00
yincongcyincong	7cfde70a9f	Update README.md	2018-02-01 17:59:21 +08:00
arraykeys@gmail.com	20837ba983	Merge branch 'master' of https://github.com/snail007/goproxy.git	2018-02-01 15:26:08 +08:00
yincongcyincong	42ce2a4351	Update README.md	2018-01-31 12:57:11 +08:00
yincongcyincong	6574d5cd29	Update README.md	2018-01-31 12:41:52 +08:00
snail007	4e9ae9a8f5	Update README.md	2018-01-31 12:32:58 +08:00
snail007	65708a0f12	Update README.md	2018-01-31 12:31:29 +08:00