Author: | Andreas Rumpf |
---|---|
Version: | 2.2.0 |
"Look at you, hacker. A pathetic creature of meat and bone, panting and sweating as you run through my corridors. How can you challenge a perfect, immortal machine?"
Introduction
This document describes the usage of the Nim compiler on the different supported platforms. It is not a definition of the Nim programming language (which is covered in the manual).
Nim is free software; it is licensed under the MIT License.
Compiler Usage
Command-line switches
All options that take a PATH or DIR argument are subject to path substitution:
- $nim: The global nim prefix path
- $lib: The stdlib path
- $home and ~: The user's home path
- $config: The directory of the module currently being compiled
- $projectname: The project file's name without file extension
- $projectpath and $projectdir: The project file's path
- $nimcache: The nimcache path
Basic command-line switches are:
Usage:
nim command [options] [projectfile] [arguments]
Command:
Arguments: arguments are passed to the program being run (if --run option is selected)
Options:
Note, single letter options that take an argument require a colon. E.g. -p:PATH.
Advanced command-line switches are:
Advanced commands:Runtime checks (see -x):
Advanced options:
List of warnings
Each warning can be activated individually with --warning:NAME:on|off or in a push pragma with {.warning[NAME]:on|off.}.
Name | Description |
---|---|
CannotOpenFile | Some file not essential for the compiler's working could not be opened. |
OctalEscape | The code contains an unsupported octal sequence. |
Deprecated | The code uses a deprecated symbol. |
ConfigDeprecated | The project makes use of a deprecated config file. |
SmallLshouldNotBeUsed | The letter 'l' should not be used as an identifier. |
EachIdentIsTuple | The code contains a confusing var declaration. |
CStringConv | Warn about dangerous implicit conversions to cstring. |
EnumConv | Warn about conversions from enum to enum. |
AnyEnumConv | Warn about any conversions to an enum type. |
HoleEnumConv | Warn about conversion to an enum with holes. These conversions are unsafe. |
ResultUsed | Warn about the usage of the built-in result variable. |
User | Some user-defined warning. |
List of hints
Each hint can be activated individually with --hint:NAME:on|off or in a push pragma with {.hint[NAME]:on|off.}.
Name | Description |
---|---|
CC | Shows when the C compiler is called. |
CodeBegin | |
CodeEnd | |
CondTrue | |
Conf | A config file was loaded. |
ConvToBaseNotNeeded | |
ConvFromXtoItselfNotNeeded | |
Dependency | |
Exec | Program is executed. |
ExprAlwaysX | |
ExtendedContext | |
GCStats | Dumps statistics about the Garbage Collector. |
GlobalVar | Shows global variables declarations. |
Link | Linking phase. |
Name | |
Path | Search paths modifications. |
Pattern | |
Performance | |
Processing | Artifact being compiled. |
QuitCalled | |
Source | The source line that triggered a diagnostic message. |
StackTrace | |
Success, SuccessX | Successful compilation of a library or a binary. |
User | |
UserRaw | |
XDeclaredButNotUsed | Unused symbols in the code. |
Verbosity levels
Level | Description |
---|---|
0 | Minimal output level for the compiler. |
1 | Displays compilation of all the compiled files, including those imported by other modules or through the compile pragma. This is the default level. |
2 | Displays compilation statistics, enumerates the dynamic libraries that will be loaded by the final binary, and dumps to standard output the result of applying a filter to the source code if any filter was used during compilation. |
3 | In addition to the previous levels dumps a debug stack trace for compiler developers. |
Compile-time symbols
Through the -d:x or --define:x switch you can define compile-time symbols for conditional compilation. The defined switches can be checked in source code with the when statement and defined proc. The typical use of this switch is to enable builds in release mode (-d:release) where optimizations are enabled for better performance. Another common use is the -d:ssl switch to activate SSL sockets.
Additionally, you may pass a value along with the symbol: -d:x=y which may be used in conjunction with the compile-time define pragmas to override symbols during build time.
Compile-time symbols are completely case insensitive and underscores are ignored too. --define:FOO and --define:foo are identical.
Compile-time symbols starting with the nim prefix are reserved for the implementation and should not be used elsewhere.
Name | Description |
---|---|
nimStdSetjmp | Use the standard setjmp()/longjmp() library functions for setjmp-based exceptions. This is the default on most platforms. |
nimSigSetjmp | Use sigsetjmp()/siglongjmp() for setjmp-based exceptions. |
nimRawSetjmp | Use _setjmp()/_longjmp() on POSIX and _setjmp()/longjmp() on Windows, for setjmp-based exceptions. It's the default on BSDs and BSD-like platforms, where it's significantly faster than the standard functions. |
nimBuiltinSetjmp | Use __builtin_setjmp()/__builtin_longjmp() for setjmp-based exceptions. This will not work if an exception is being thrown and caught inside the same procedure. Useful for benchmarking. |
Configuration files
Note: The project file name is the name of the .nim file that is passed as a command-line argument to the compiler.
The nim executable processes configuration files in the following directories (in this order; later files overwrite previous settings):
- $nim/config/nim.cfg, /etc/nim/nim.cfg (UNIX) or <Nim's installation directory>\config\nim.cfg (Windows). This file can be skipped with the --skipCfg command line option.
- If environment variable XDG_CONFIG_HOME is defined, $XDG_CONFIG_HOME/nim/nim.cfg or ~/.config/nim/nim.cfg (POSIX) or %APPDATA%/nim/nim.cfg (Windows). This file can be skipped with the --skipUserCfg command line option.
- $parentDir/nim.cfg where $parentDir stands for any parent directory of the project file's path. These files can be skipped with the --skipParentCfg command-line option.
- $projectDir/nim.cfg where $projectDir stands for the project file's path. This file can be skipped with the --skipProjCfg command-line option.
- A project can also have a project-specific configuration file named $project.nim.cfg that resides in the same directory as $project.nim. This file can be skipped with the --skipProjCfg command-line option.
Command-line settings have priority over configuration file settings.
The default build of a project is a debug build. To compile a release build define the release symbol:
nim c -d:release myproject.nim
To compile a dangerous release build define the danger symbol:
nim c -d:danger myproject.nim
Search path handling
Nim has the concept of a global search path (PATH) that is queried to determine where to find imported modules or include files. If multiple files are found an ambiguity error is produced.
nim dump shows the contents of the PATH.
However before the PATH is used the current directory is checked for the file's existence. So if PATH contains $lib and $lib/bar and the directory structure looks like this:
$lib/x.nim $lib/bar/x.nim foo/x.nim foo/main.nim other.nim
And main imports x, foo/x is imported. If other imports x then both $lib/x.nim and $lib/bar/x.nim match but $lib/x.nim is used as it is the first match.
Generated C code directory
The generated files that Nim produces all go into a subdirectory called nimcache. Its full path is
- $XDG_CACHE_HOME/nim/$projectname(_r|_d) or ~/.cache/nim/$projectname(_r|_d) on Posix
- $HOME\nimcache\$projectname(_r|_d) on Windows.
The _r suffix is used for release builds, _d is for debug builds.
This makes it easy to delete all generated files.
The --nimcache compiler switch can be used to to change the nimcache directory.
However, the generated C code is not platform-independent. C code generated for Linux does not compile on Windows, for instance. The comment on top of the C file lists the OS, CPU, and CC the file has been compiled for.
Compiler Selection
To change the compiler from the default compiler (at the command line):
nim c --cc:llvm_gcc --compile_only myfile.nim
This uses the configuration defined in config\nim.cfg for llvm_gcc.
If nimcache already contains compiled code from a different compiler for the same project, add the -f flag to force all files to be recompiled.
The default compiler is defined at the top of config\nim.cfg. Changing this setting affects the compiler used by koch to (re)build Nim.
To use the CC environment variable, use nim c --cc:env myfile.nim. To use the CXX environment variable, use nim cpp --cc:env myfile.nim. --cc:env is available since Nim version 1.4.
Cross-compilation
To cross compile, use for example:
nim c --cpu:i386 --os:linux --compileOnly --genScript myproject.nim
Then move the C code and the compile script compile_myproject.sh to your Linux i386 machine and run the script.
Another way is to make Nim invoke a cross compiler toolchain:
nim c --cpu:arm --os:linux myproject.nim
For cross compilation, the compiler invokes a C compiler named like $cpu.$os.$cc (for example arm.linux.gcc) with options defined in $cpu.$os.$cc.options.always. The configuration system is used to provide meaningful defaults. For example, for Linux on a 32-bit ARM CPU, your configuration file should contain something like:
arm.linux.gcc.path = "/usr/bin" arm.linux.gcc.exe = "arm-linux-gcc" arm.linux.gcc.linkerexe = "arm-linux-gcc" arm.linux.gcc.options.always = "-w -fmax-errors=3"
Cross-compilation for Windows
To cross-compile for Windows from Linux or macOS using the MinGW-w64 toolchain:
nim c -d:mingw myproject.nim # `nim r` also works, running the binary via `wine` or `wine64`: nim r -d:mingw --eval:'import os; echo "a" / "b"'
Use --cpu:i386 or --cpu:amd64 to switch the CPU architecture.
The MinGW-w64 toolchain can be installed as follows:
apt install mingw-w64 # Ubuntu yum install mingw32-gcc yum install mingw64-gcc # CentOS - requires EPEL brew install mingw-w64 # OSX
Cross-compilation for Android
There are two ways to compile for Android: terminal programs (Termux) and with the NDK (Android Native Development Kit).
The first one is to treat Android as a simple Linux and use Termux to connect and run the Nim compiler directly on android as if it was Linux. These programs are console-only programs that can't be distributed in the Play Store.
Use regular nim c inside termux to make Android terminal programs.
Normal Android apps are written in Java, to use Nim inside an Android app you need a small Java stub that calls out to a native library written in Nim using the NDK. You can also use native-activity to have the Java stub be auto-generated for you.
Use nim c -c --cpu:arm --os:android -d:androidNDK --noMain:on to generate the C source files you need to include in your Android Studio project. Add the generated C files to CMake build script in your Android project. Then do the final compile with Android Studio which uses Gradle to call CMake to compile the project.
Because Nim is part of a library it can't have its own C-style main() so you would need to define your own android_main and init the Java environment, or use a library like SDL2 or GLFM to do it. After the Android stuff is done, it's very important to call NimMain() in order to initialize Nim's garbage collector and to run the top level statements of your program.
proc NimMain() {.importc.} proc glfmMain*(display: ptr GLFMDisplay) {.exportc.} = NimMain() # initialize garbage collector memory, types and stack
The name NimMain can be influenced via the --nimMainPrefix:prefix switch. Use --nimMainPrefix:MyLib and the function to call is named MyLibNimMain.
Cross-compilation for iOS
To cross-compile for iOS you need to be on a macOS computer and use XCode. Normal languages for iOS development are Swift and Objective C. Both of these use LLVM and can be compiled into object files linked together with C, C++ or Objective C code produced by Nim.
Use nim c -c --os:ios --noMain:on to generate C files and include them in your XCode project. Then you can use XCode to compile, link, package and sign everything.
Because Nim is part of a library it can't have its own C-style main() so you would need to define main that calls autoreleasepool and UIApplicationMain to do it, or use a library like SDL2 or GLFM. After the iOS setup is done, it's very important to call NimMain() to initialize Nim's garbage collector and to run the top-level statements of your program.
proc NimMain() {.importc.} proc glfmMain*(display: ptr GLFMDisplay) {.exportc.} = NimMain() # initialize garbage collector memory, types and stack
Note: XCode's "make clean" gets confused about the generated nim.c files, so you need to clean those files manually to do a clean build.
The name NimMain can be influenced via the --nimMainPrefix:prefix switch. Use --nimMainPrefix:MyLib and the function to call is named MyLibNimMain.
Cross-compilation for Nintendo Switch
Simply add --os:nintendoswitch to your usual nim c or nim cpp command and set the passC and passL command line switches to something like:
nim c ... --d:nimAllocPagesViaMalloc --mm:orc --passC="-I$DEVKITPRO/libnx/include" ... --passL="-specs=$DEVKITPRO/libnx/switch.specs -L$DEVKITPRO/libnx/lib -lnx"
or setup a nim.cfg file like so:
#nim.cfg --mm:orc --d:nimAllocPagesViaMalloc --define:nimInheritHandles --passC="-I$DEVKITPRO/libnx/include" --passL="-specs=$DEVKITPRO/libnx/switch.specs -L$DEVKITPRO/libnx/lib -lnx"
The devkitPro setup must be the same as the default with their new installer here for Mac/Linux or here for Windows.
For example, with the above-mentioned config:
nim c --os:nintendoswitch switchhomebrew.nim
This will generate a file called switchhomebrew.elf which can then be turned into an nro file with the elf2nro tool in the devkitPro release. Examples can be found at the nim-libnx github repo.
There are a few things that don't work because the devkitPro libraries don't support them. They are:
- Waiting for a subprocess to finish. A subprocess can be started, but right now it can't be waited on, which sort of makes subprocesses a bit hard to use
- Dynamic calls. Switch OS (Horizon) doesn't support dynamic libraries, so dlopen/dlclose are not available.
- mqueue. Sadly there are no mqueue headers.
- ucontext. No headers for these either. No coroutines for now :(
- nl_types. No headers for this.
- As mmap is not supported, the nimAllocPagesViaMalloc option has to be used.
GPU Compilation
Compiling for GPU computation can be achieved with --cc:nvcc for CUDA with nvcc, or with --cc:hipcc for AMD GPUs with HIP. Both compilers require building for C++ with nim cpp.
Here's a very simple CUDA kernel example using emit, which can be compiled with nim cpp --cc:nvcc --define:"useMalloc" hello_kernel.nim assuming you have the CUDA toolkit installed.
{.emit: """ __global__ void add(int a, int b) { int c; c = a + b; } """.} proc main() = {.emit: """ add<<<1,1>>>(2,7); """.} main()
DLL generation
Note: The same rules apply to lib*.so shared object files on UNIX. For better readability only the DLL version is described here.
Nim supports the generation of DLLs. However, there must be only one instance of the GC per process/address space. This instance is contained in nimrtl.dll. This means that every generated Nim DLL depends on nimrtl.dll. To generate the "nimrtl.dll" file, use the command:
nim c -d:release lib/nimrtl.nim
To link against nimrtl.dll use the command:
nim c -d:useNimRtl myprog.nim
Additional compilation switches
The standard library supports a growing number of useX conditional defines affecting how some features are implemented. This section tries to give a complete list.
Define | Effect |
---|---|
release | Turns on the optimizer. More aggressive optimizations are possible, e.g.: --passC:-ffast-math (but see issue #10305) |
danger | Turns off all runtime checks and turns on the optimizer. |
useFork | Makes osproc use fork instead of posix_spawn. |
useNimRtl | Compile and link against nimrtl.dll. |
useMalloc | Makes Nim use C's malloc instead of Nim's own memory manager, albeit prefixing each allocation with its size to support clearing memory on reallocation. This only works with --mm:none, --mm:arc and --mm:orc. |
useRealtimeGC | Enables support of Nim's GC for soft realtime systems. See the documentation of the refc for further information. |
logGC | Enable GC logging to stdout. |
nodejs | The JS target is actually node.js. |
ssl | Enables OpenSSL support for the sockets module. |
memProfiler | Enables memory profiling for the native GC. |
uClibc | Use uClibc instead of libc. (Relevant for Unix-like OSes) |
checkAbi | When using types from C headers, add checks that compare what's in the Nim file with what's in the C header. This may become enabled by default in the future. |
tempDir | This symbol takes a string as its value, like --define:tempDir:/some/temp/path to override the temporary directory returned by os.getTempDir(). The value should end with a directory separator character. (Relevant for the Android platform) |
useShPath | This symbol takes a string as its value, like --define:useShPath:/opt/sh/bin/sh to override the path for the sh binary, in cases where it is not located in the default location /bin/sh. |
noSignalHandler | Disable the crash handler from system.nim. |
globalSymbols | Load all {.dynlib.} libraries with the RTLD_GLOBAL flag on Posix systems to resolve symbols in subsequently loaded libraries. |
lto | Enable link-time optimization in the backend compiler and linker. |
lto_incremental | Enable link-time optimization and additionally enable incremental linking for compilers that support it. Currently only clang and vcc. |
strip | Strip debug symbols added by the backend compiler from the executable. |
Additional Features
This section describes Nim's additional features that are not listed in the Nim manual. Some of the features here only make sense for the C code generator and are subject to change.
LineDir option
The --lineDir option can be turned on or off. If turned on the generated C code contains #line directives. This may be helpful for debugging with GDB.
StackTrace option
If the --stackTrace option is turned on, the generated C contains code to ensure that proper stack traces are given if the program crashes or some uncaught exception is raised.
LineTrace option
The --lineTrace option implies the stackTrace option. If turned on, the generated C contains code to ensure that proper stack traces with line number information are given if the program crashes or an uncaught exception is raised.
DynlibOverride
By default Nim's dynlib pragma causes the compiler to generate GetProcAddress (or their Unix counterparts) calls to bind to a DLL. With the dynlibOverride command line switch this can be prevented and then via --passL the static library can be linked against. For instance, to link statically against Lua this command might work on Linux:
nim c --dynlibOverride:lua --passL:liblua.lib program.nim
Backend language options
The typical compiler usage involves using the compile or c command to transform a .nim file into one or more .c files which are then compiled with the platform's C compiler into a static binary. However, there are other commands to compile to C++, Objective-C, or JavaScript. More details can be read in the Nim Backend Integration document.
Nim documentation tools
Nim provides the doc command to generate HTML documentation from .nim source files. Only exported symbols will appear in the output. For more details see the docgen documentation.
Nim idetools integration
Nim provides language integration with external IDEs through the idetools command. See the documentation of idetools for further information.
Nim for embedded systems
While the default Nim configuration is targeted for optimal performance on modern PC hardware and operating systems with ample memory, it is very well possible to run Nim code and a good part of the Nim standard libraries on small embedded microprocessors with only a few kilobytes of memory.
A good start is to use the any operating target together with the malloc memory allocator and the arc garbage collector. For example:
nim c --os:any --mm:arc -d:useMalloc [...] x.nim
- --mm:arc will enable the reference counting memory management instead of the default garbage collector. This enables Nim to use heap memory which is required for strings and seqs, for example.
- The --os:any target makes sure Nim does not depend on any specific operating system primitives. Your platform should support only some basic ANSI C library stdlib and stdio functions which should be available on almost any platform.
- The -d:useMalloc option configures Nim to use only the standard C memory manage primitives malloc(), free(), realloc().
If your platform does not provide these functions it should be trivial to provide an implementation for them and link these to your program.
For targets with very restricted memory, it might be beneficial to pass some additional flags to both the Nim compiler and the C compiler and/or linker to optimize the build for size. For example, the following flags can be used when targeting a gcc compiler:
--opt:size -d:lto -d:strip
The --opt:size flag instructs Nim to optimize code generation for small size (with the help of the C compiler), the -d:lto flags enable link-time optimization in the compiler and linker, the -d:strip strips debug symbols.
Check the Cross-compilation section for instructions on how to compile the program for your target.
nimAllocPagesViaMalloc
Nim's default allocator is based on TLSF, this algorithm was designed for embedded devices. This allocator gets blocks/pages of memory via a currently undocumented osalloc API which usually uses POSIX's mmap call. On many environments mmap is not available but C's malloc is. You can use the nimAllocPagesViaMalloc define to use malloc instead of mmap. nimAllocPagesViaMalloc is currently only supported with --mm:arc or --mm:orc. (Since version 1.6)
nimPage256 / nimPage512 / nimPage1k
Adjust the page size for Nim's GC allocator. This enables using nimAllocPagesViaMalloc on devices with less RAM. The default page size requires too much RAM to work.
Recommended settings:
- < 32 kB of RAM use nimPage256
- < 512 kB of RAM use nimPage512
- < 2 MB of RAM use nimPage1k
Initial testing hasn't shown much difference between 512B or 1kB page sizes in terms of performance or latency. Using nimPages256 will limit the total amount of allocatable RAM.
nimMemAlignTiny
Sets MemAlign to 4 bytes which reduces the memory alignment to better match some embedded devices.
Thread stack size
Nim's thread API provides a simple wrapper around more advanced RTOS task features. Customizing the stack size and stack guard size can be done by setting -d:nimThreadStackSize=16384 or -d:nimThreadStackGuard=32.
Currently only Zephyr, NuttX and FreeRTOS support these configurations.
Nim for realtime systems
See the --mm:arc or --mm:orc memory management settings in MM for further information.
Signal handling in Nim
The Nim programming language has no concept of Posix's signal handling mechanisms. However, the standard library offers some rudimentary support for signal handling, in particular, segmentation faults are turned into fatal errors that produce a stack trace. This can be disabled with the -d:noSignalHandler switch.
Optimizing for Nim
Nim has no separate optimizer, but the C code that is produced is very efficient. Most C compilers have excellent optimizers, so usually it is not needed to optimize one's code. Nim has been designed to encourage efficient code: The most readable code in Nim is often the most efficient too.
However, sometimes one has to optimize. Do it in the following order:
- switch off the embedded debugger (it is slow!)
- turn on the optimizer and turn off runtime checks
- profile your code to find where the bottlenecks are
- try to find a better algorithm
- do low-level optimizations
This section can only help you with the last item.
Optimizing string handling
String assignments are sometimes expensive in Nim: They are required to copy the whole string. However, the compiler is often smart enough to not copy strings. Due to the argument passing semantics, strings are never copied when passed to subroutines. The compiler does not copy strings that are a result of a procedure call, because the callee returns a new string anyway. Thus it is efficient to do:
var s = procA() # assignment will not copy the string; procA allocates a new # string already
However, it is not efficient to do:
var s = varA # assignment has to copy the whole string into a new buffer!
For let symbols a copy is not always necessary:
let s = varA # may only copy a pointer if it safe to do so
The compiler optimizes string case statements: A hashing scheme is used for them if several different string constants are used. So code like this is reasonably efficient:
case normalize(k.key) of "name": c.name = v of "displayname": c.displayName = v of "version": c.version = v of "os": c.oses = split(v, {';'}) of "cpu": c.cpus = split(v, {';'}) of "authors": c.authors = split(v, {';'}) of "description": c.description = v of "app": case normalize(v) of "console": c.app = appConsole of "gui": c.app = appGUI else: quit(errorStr(p, "expected: console or gui")) of "license": c.license = UnixToNativePath(k.value) else: quit(errorStr(p, "unknown variable: " & k.key))