You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

210 lines
9.2 KiB

# n2n on macOS
In order to use n2n on macOS, you first need to install support for TUN/TAP interfaces:
```bash
brew tap homebrew/cask
brew cask install tuntap
```
If you are on a modern version of macOS (i.e. Catalina), the commands above will ask you to enable the TUN/TAP kernel extension in System Preferences → Security & Privacy → General.
For more information refer to vendor documentation or the [Apple Technical Note](https://developer.apple.com/library/content/technotes/tn2459/_index.html).
added mingw test platform (#829) * Provide a minimal reimplementation of our autoconf, to try windows builds * Try building with windows * Fix thinko in spelling * Ensure shell script runs inside a shell * Add a hack to aid include discovery * Just keep adding tech debt... * Assume that we will have slashes in some of the replacement strings and avoid that char with sed * Restore one slash * Hack around the tools makefile interdependancy bug * A correct cflags include hack for each compile dir * Ensure we link against winsock (note, even though this says 32bit, it should link the 64bit library ... I think) * Bad link ordering if we dont use LDLIBS * Remove unused make variable * Remove makefile duplication using inheritance (this does mean you can no longer cd tools; make, but must do make tools) * Add missing library for win32 * Show OS variable * Make hack autoconf more robust for tests on non gitlab runners * Remove no longer used substitutions from hack autoconf * Add missing include path to tools under win32 * Build the win32 subdir when the compiler is Msys * The different subdirs have different dependancies * Ensure we can find the include files * Fix library link ordering * Ensure the tools dir can find the special win32 lib * Deal with the differing basic type sizes on both linux/64bit and windows/64bit * Document the steps to mimic the github windows/mingw build locally - to allow for simpler debugging * Ensure branch name in instructions matches my test branch name * Clarify the shell needed to build with mingw * Since the makefile depends on knowing the OS, raise a fatal error if we cannot determine this * Handling different compile environments is hard. - Linux: sane and reasonable results for both uname -s (=Linux) and uname -o (=GNU/Linux) - Windows/Mingw: insane results for uname -s (=MSYS_NT-$MAJOR.$MINOR-$BUILDNR) but sane results for uname -o (Msys) - Macos: sane results for uname -s (=Darwin) but does not support uname -o at all * Revamp the way that Mingw is detected * Avoid attempting to generate gcovr report when running under windows * Whoops, isolate the right step * Fix spelling mistake * win32/Makefile: Remove unused setting and add comment * ensure that all win32 includes use the same expected path * Allow simpler cross compilation by letting configure pass the CC and AR environment through * Avoid multiple '_CRT_SECURE_NO_WARNINGS redefined' warnings * Convert to a consolidated CONFIG_TARGET variable to select any different compile options * Use the more generic printf defines to avoid warnings on mingw * Update mingw build docs * English better for reader happy make * Address a number of mingw compiler warnings * Fix Visual C compile * Be sure to document some of the hacky nature of the mingw build
3 years ago
# Build on Windows (Visual Studio)
## Requirements
In order to build on Windows the following tools should be installed:
- Visual Studio. For a minimal install, the command line only build tools can be
downloaded and installed from https://visualstudio.microsoft.com/downloads/#build-tools-for-visual-studio-2017.
- CMake
- (optional) The OpenSSL library. Prebuild binaries can be downloaded from https://slproweb.com/products/Win32OpenSSL.html.
The full version is required, i.e. not the "Light" version. The Win32 version of it is usually required for a standard build.
> NOTE: In order to skip OpenSSL compilation, edit `CMakeLists.txt` and replace **– is this still valid?**
>
> ```plaintext
> OPTION(N2N_OPTION_AES "USE AES" ON)
> with
> OPTION(N2N_OPTION_AES "USE AES" OFF)
> ```
NOTE: To statically link OpenSSL, add the `-DOPENSSL_USE_STATIC_LIBS=true` option to the `cmake` command below.
- If compilation throws a "config.h: No such file or directory" error, an `include/config.h` file needs to be obtained from an already configured Linux compilation and put into the `include/` directory as discussed [here](https://github.com/ntop/n2n/issues/366).
In order to run n2n, you will need the following:
- The TAP drivers should be installed into the system. They can be installed from
http://build.openvpn.net/downloads/releases, search for "tap-windows".
- If OpenSSL has been linked dynamically, the corresponding `.dll` file should be available
onto the target computer.
added mingw test platform (#829) * Provide a minimal reimplementation of our autoconf, to try windows builds * Try building with windows * Fix thinko in spelling * Ensure shell script runs inside a shell * Add a hack to aid include discovery * Just keep adding tech debt... * Assume that we will have slashes in some of the replacement strings and avoid that char with sed * Restore one slash * Hack around the tools makefile interdependancy bug * A correct cflags include hack for each compile dir * Ensure we link against winsock (note, even though this says 32bit, it should link the 64bit library ... I think) * Bad link ordering if we dont use LDLIBS * Remove unused make variable * Remove makefile duplication using inheritance (this does mean you can no longer cd tools; make, but must do make tools) * Add missing library for win32 * Show OS variable * Make hack autoconf more robust for tests on non gitlab runners * Remove no longer used substitutions from hack autoconf * Add missing include path to tools under win32 * Build the win32 subdir when the compiler is Msys * The different subdirs have different dependancies * Ensure we can find the include files * Fix library link ordering * Ensure the tools dir can find the special win32 lib * Deal with the differing basic type sizes on both linux/64bit and windows/64bit * Document the steps to mimic the github windows/mingw build locally - to allow for simpler debugging * Ensure branch name in instructions matches my test branch name * Clarify the shell needed to build with mingw * Since the makefile depends on knowing the OS, raise a fatal error if we cannot determine this * Handling different compile environments is hard. - Linux: sane and reasonable results for both uname -s (=Linux) and uname -o (=GNU/Linux) - Windows/Mingw: insane results for uname -s (=MSYS_NT-$MAJOR.$MINOR-$BUILDNR) but sane results for uname -o (Msys) - Macos: sane results for uname -s (=Darwin) but does not support uname -o at all * Revamp the way that Mingw is detected * Avoid attempting to generate gcovr report when running under windows * Whoops, isolate the right step * Fix spelling mistake * win32/Makefile: Remove unused setting and add comment * ensure that all win32 includes use the same expected path * Allow simpler cross compilation by letting configure pass the CC and AR environment through * Avoid multiple '_CRT_SECURE_NO_WARNINGS redefined' warnings * Convert to a consolidated CONFIG_TARGET variable to select any different compile options * Use the more generic printf defines to avoid warnings on mingw * Update mingw build docs * English better for reader happy make * Address a number of mingw compiler warnings * Fix Visual C compile * Be sure to document some of the hacky nature of the mingw build
3 years ago
NOTE: Sticking to this tool chain has historically meant that resulting
executables are more likely to be able to communicate with Linux or other
OS builds, however efforts are being made to address this concern.
## Build (CLI)
In order to build from the command line, open a terminal window and run the following commands:
```batch
md build
cd build
cmake ..
MSBuild.exe edge.vcxproj /t:Build /p:Configuration=Release
MSBuild.exe supernode.vcxproj /t:Build /p:Configuration=Release
MSBuild.exe n2n-benchmark.vcxproj /t:Build /p:Configuration=Release
```
NOTE: If CMake has problems finding the installed OpenSSL library, try to download the official cmake and invoke it with
`C:\Program Files\CMake\bin\cmake`.
NOTE: Visual Studio might not add `MSBuild.exe`'s path to the environment variable %PATH% so you might have difficulties finding and executing it without giving the full path. Regular installations seem to have it reside at `"C:\Program Files (x86)\Microsoft Visual Studio\2019\BuildTools\MSBuild\Current\Bin\MSBuild.exe"`
The compiled `.exe` files should now be available in the `build\Release` directory.
## Run
The `edge.exe` program reads the `edge.conf` file located into the current directory if no option is provided.
Here is an example `edge.conf` file:
```plaintext
-c=mycommunity
-k=mysecretpass
# supernode IP address
-l=1.2.3.4:5678
# edge IP address
-a=192.168.100.1
```
The `supernode.exe` program reads the `supernode.conf` file located into the current directory if no option is provided.
Here is an example `supernode.conf` file:
```plaintext
-p=5678
```
See `edge.exe --help` and `supernode.exe --help` for a full list of supported options.
added mingw test platform (#829) * Provide a minimal reimplementation of our autoconf, to try windows builds * Try building with windows * Fix thinko in spelling * Ensure shell script runs inside a shell * Add a hack to aid include discovery * Just keep adding tech debt... * Assume that we will have slashes in some of the replacement strings and avoid that char with sed * Restore one slash * Hack around the tools makefile interdependancy bug * A correct cflags include hack for each compile dir * Ensure we link against winsock (note, even though this says 32bit, it should link the 64bit library ... I think) * Bad link ordering if we dont use LDLIBS * Remove unused make variable * Remove makefile duplication using inheritance (this does mean you can no longer cd tools; make, but must do make tools) * Add missing library for win32 * Show OS variable * Make hack autoconf more robust for tests on non gitlab runners * Remove no longer used substitutions from hack autoconf * Add missing include path to tools under win32 * Build the win32 subdir when the compiler is Msys * The different subdirs have different dependancies * Ensure we can find the include files * Fix library link ordering * Ensure the tools dir can find the special win32 lib * Deal with the differing basic type sizes on both linux/64bit and windows/64bit * Document the steps to mimic the github windows/mingw build locally - to allow for simpler debugging * Ensure branch name in instructions matches my test branch name * Clarify the shell needed to build with mingw * Since the makefile depends on knowing the OS, raise a fatal error if we cannot determine this * Handling different compile environments is hard. - Linux: sane and reasonable results for both uname -s (=Linux) and uname -o (=GNU/Linux) - Windows/Mingw: insane results for uname -s (=MSYS_NT-$MAJOR.$MINOR-$BUILDNR) but sane results for uname -o (Msys) - Macos: sane results for uname -s (=Darwin) but does not support uname -o at all * Revamp the way that Mingw is detected * Avoid attempting to generate gcovr report when running under windows * Whoops, isolate the right step * Fix spelling mistake * win32/Makefile: Remove unused setting and add comment * ensure that all win32 includes use the same expected path * Allow simpler cross compilation by letting configure pass the CC and AR environment through * Avoid multiple '_CRT_SECURE_NO_WARNINGS redefined' warnings * Convert to a consolidated CONFIG_TARGET variable to select any different compile options * Use the more generic printf defines to avoid warnings on mingw * Update mingw build docs * English better for reader happy make * Address a number of mingw compiler warnings * Fix Visual C compile * Be sure to document some of the hacky nature of the mingw build
3 years ago
# Build on Windows (MinGW)
These steps were tested on a fresh install of Windows 10 Pro with all patches
applied as of 2021-09-29.
- Install Chocolatey (Following instructions on https://chocolatey.org/install)
- from an admin cmd prompt
- choco install git mingw make
- All the remaining commands must be run from inside a bash shell ("C:\Program Files\Git\usr\bin\bash.exe")
- git clone $THIS_REPO
- cd n2n
- ./scripts/hack_fakeautoconf
- make
- make test
Due to the hack used to replace autotools on windows, any build created this
way will currently have inaccurate build version numbers.
Note: MinGW builds have a history of incompatibility reports with other OS
builds, please see [#617](https://github.com/ntop/n2n/issues/617) and [#642](https://github.com/ntop/n2n/issues/642).
However, if the tests pass, you should have a high confidence that your build
will be compatible.
# General Building Options
## Compiler Optimizations
The easiest way to boosting speed is by allowing the compiler to apply optimization to the code. To let the compiler know, the configuration process can take in the optionally specified compiler flag `-O3`:
`./configure CFLAGS="-O3"`
The `tools/n2n-benchmark` tool reports speed-ups of 200% or more! There is no known risk in terms of instable code or so.
## Hardware Features
Some parts of the code can be compiled to benefit from available hardware acceleration. It needs to be decided at compile-time. So, if compiling for a specific platform with known features (maybe the local one), it should be specified to the compiler, for example through the `-march=sandybridge` (you name it) or just `-march=native` for local use.
So far, the following portions of n2n's code benefit from hardware features:
```
AES: AES-NI
ChaCha20: SSE2, SSSE3
SPECK: SSE2, SSSE3, AVX2, AVX512, (NEON)
Pearson Hashing: AES-NI
Random Numbers: RDSEED, RDRND (not faster but more random seed)
```
The compilations flags could easily be combined:
`./configure CFLAGS="-O3 -march=native"`.
## OpenSSL Support
Some ciphers' speed can take advantage of OpenSSL support which is disabled by default as the built-in ciphers already prove reasonably fast in most cases. OpenSSL support can be configured using
`./configure --with-openssl`
which then will include OpenSSL 1.1 if found on the system. This can be combined with the hardware support and compiler optimizations such as
`./configure --with-openssl CFLAGS="-O3 -march=native"`
Please do no forget to `make clean` after (re-)configuration and before building (again) using `make`.
## ZSTD Compression Support
In addition to the built-in LZO1x for payload compression (`-z1` at the edge's commandline), n2n optionally supports [ZSTD](https://github.com/facebook/zstd). As of 2020, it is considered cutting edge and [praised](https://en.wikipedia.org/wiki/Zstandard) for reaching the currently technologically possible Pareto frontier in terms of CPU power versus compression ratio. ZSTD support can be configured using
`./configure --with-zstd`
which then will include ZSTD if found on the system. It will be available via `-z2` at the edges. Of course, it can be combined with the other features mentioned above:
`./configure --with-zstd --with-openssl CFLAGS="-O3 -march=native"`
Again, and this needs to be reiterated sufficiently often, please do no forget to `make clean` after (re-)configuration and before building (again) using `make`.
## Federation – Supernode Selection by Round Trip Time
If used with multiple supernodes, by default, an edge choses the least loaded supernode to connect to. This selection strategy is part of the [federation](Federation.md) feature and aims at a fair workload distribution among the supernodes. To serve special scenarios, an edge can be compiled to always connect to the supernode with the lowest round trip time, i.e. the "closest" with the lowest ping. However, this could result in not so fair workload distribution among supernodes. This option can be configured by defining the macro `SN_SELECTION_RTT` and affects edge's behaviour only:
`./configure CFLAGS="-DSN_SELECTION_RTT"`
which of course can be combined with the compiler optimizations mentioned above…
Note that the activation of this strategy requires a sufficiently accurate local day-of-time clock. It probably will fail on smaller systems using `uclibc` (instead of `glibc`) whose day-of-time clock is said to not provide sub-second accuracy.
## SPECK – ARM NEON Hardware Acceleration
By default, SPECK does not take advantage of ARM NEON hardware acceleration even if compiled with `-march=native`. The reason is that the NEON implementation proved to be slower than the 64-bit scalar code on Raspberry Pi 3B+, see [here](https://github.com/ntop/n2n/issues/563).
Your specific ARM mileage may vary, so it can be enabled by configuring the definition of the `SPECK_ARM_NEON` macro:
`./configure CFLAGS="-DSPECK_ARM_NEON"`
Just make sure that the correct architecture is set, too. `-march=native` usually works quite well.
3 years ago
## Disable Multicast Local Peer Detection
For better local peer detection, the edges try to detect local peers by sending REGISTER
packets to a certain multicast address. Also, edges listen to this address to eventually
fetch such packets.
If these packets disturb network's peace or even get forwarded by (other) edges through the
n2n network, this behavior can be disabled, just add
`-DSKIP_MULTICAST_PEERS_DISCOVERY`
to your `CFLAGS` when configuring, e.g.
`./configure --with-zstd CFLAGS="-O3 -march=native -DSKIP_MULTICAST_PEERS_DISCOVERY"`