# 7.1. Using GHC¶

## 7.1.1. Getting started: compiling programs¶

In this chapter you’ll find a complete reference to the GHC command-line syntax, including all 400+ flags. It’s a large and complex system, and there are lots of details, so it can be quite hard to figure out how to get started. With that in mind, this introductory section provides a quick introduction to the basic usage of GHC for compiling a Haskell program, before the following sections dive into the full syntax.

Let’s create a Hello World program, and compile and run it. First, create a file hello.hs containing the Haskell code:

main = putStrLn "Hello, World!"


To compile the program, use GHC like this:

$ghc hello.hs  (where $ represents the prompt: don’t type it). GHC will compile the source file hello.hs, producing an object file hello.o and an interface file hello.hi, and then it will link the object file to the libraries that come with GHC to produce an executable called hello on Unix/Linux/Mac, or hello.exe on Windows.

By default GHC will be very quiet about what it is doing, only printing error messages. If you want to see in more detail what’s going on behind the scenes, add -v to the command line.

Then we can run the program like this:

$./hello Hello World!  If your program contains multiple modules, then you only need to tell GHC the name of the source file containing the Main module, and GHC will examine the import declarations to find the other modules that make up the program and find their source files. This means that, with the exception of the Main module, every source file should be named after the module name that it contains (with dots replaced by directory separators). For example, the module Data.Person would be in the file Data/Person.hs on Unix/Linux/Mac, or Data\Person.hs on Windows. ## 7.1.2. Options overview¶ GHC’s behaviour is controlled by options, which for historical reasons are also sometimes referred to as command-line flags or arguments. Options can be specified in three ways: ### 7.1.2.1. Command-line arguments¶ An invocation of GHC takes the following form: ghc [argument...]  Command-line arguments are either options or file names. Command-line options begin with -. They may not be grouped: -vO is different from -v -O. Options need not precede filenames: e.g., ghc *.o -o foo. All options are processed and then applied to all files; you cannot, for example, invoke ghc -c -O1 Foo.hs -O2 Bar.hs to apply different optimisation levels to the files Foo.hs and Bar.hs. Note Note that command-line options are order-dependent, with arguments being evaluated from left-to-right. This can have seemingly strange effects in the presence of flag implication. For instance, consider -fno-specialise and -O1 (which implies -fspecialise). These two command lines mean very different things: -fno-specialise -O1 -fspecialise will be enabled as the -fno-specialise is overriden by the -O1. -O1 -fno-specialise -fspecialise will not be enabled, since the -fno-specialise overrides the -fspecialise implied by -O1. ### 7.1.2.2. Command line options in source files¶ Sometimes it is useful to make the connection between a source file and the command-line options it requires quite tight. For instance, if a Haskell source file deliberately uses name shadowing, it should be compiled with the -Wno-name-shadowing option. Rather than maintaining the list of per-file options in a Makefile, it is possible to do this directly in the source file using the OPTIONS_GHC pragma {-# OPTIONS_GHC -Wno-name-shadowing #-} module X where ...  OPTIONS_GHC is a file-header pragma (see OPTIONS_GHC pragma). Only dynamic flags can be used in an OPTIONS_GHC pragma (see Dynamic and Mode options). Note that your command shell does not get to the source file options, they are just included literally in the array of command-line arguments the compiler maintains internally, so you’ll be desperately disappointed if you try to glob etc. inside OPTIONS_GHC. Note The contents of OPTIONS_GHC are appended to the command-line options, so options given in the source file override those given on the command-line. It is not recommended to move all the contents of your Makefiles into your source files, but in some circumstances, the OPTIONS_GHC pragma is the Right Thing. (If you use -keep-hc-file and have OPTION flags in your module, the OPTIONS_GHC will get put into the generated .hc file). ### 7.1.2.3. Setting options in GHCi¶ Options may also be modified from within GHCi, using the :set command. ## 7.1.3. Dynamic and Mode options¶ Each of GHC’s command line options is classified as dynamic or mode: Mode: A mode may be used on the command line only. You can pass only one mode flag. For example, --make or -E. The available modes are listed in Modes of operation. Dynamic: A dynamic flag may be used on the command line, in a OPTIONS_GHC pragma in a source file, or set using :set in GHCi. The flag reference tables (Flag reference) lists the status of each flag. ## 7.1.4. Meaningful file suffixes¶ File names with “meaningful” suffixes (e.g., .lhs or .o) cause the “right thing” to happen to those files. .hs A Haskell module. .lhs A “literate Haskell” module. .hspp A file created by the preprocessor. .hi A Haskell interface file, probably compiler-generated. .hie An extended Haskell interface file, produced by the Haskell compiler. .hc Intermediate C file produced by the Haskell compiler. .c A C file not produced by the Haskell compiler. .ll An llvm-intermediate-language source file, usually produced by the compiler. .bc An llvm-intermediate-language bitcode file, usually produced by the compiler. .s An assembly-language source file, usually produced by the compiler. .o An object file, produced by an assembler. Files with other suffixes (or without suffixes) are passed straight to the linker. ## 7.1.5. Modes of operation¶ GHC’s behaviour is firstly controlled by a mode flag. Only one of these flags may be given, but it does not necessarily need to be the first option on the command-line. For instance, $ ghc Main.hs --make -o my-application


If no mode flag is present, then GHC will enter --make mode (Using ghc --make) if there are any Haskell source files given on the command line, or else it will link the objects named on the command line to produce an executable.

The available mode flags are:

--interactive

Interactive mode, which is also available as ghci. Interactive mode is described in more detail in Using GHCi.

--make

In this mode, GHC will build a multi-module Haskell program automatically, figuring out dependencies for itself. If you have a straightforward Haskell program, this is likely to be much easier, and faster, than using make. Make mode is described in Using ghc --make.

This mode is the default if there are any Haskell source files mentioned on the command line, and in this case the --make option can be omitted.

-e ⟨expr⟩

Expression-evaluation mode. This is very similar to interactive mode, except that there is a single expression to evaluate (⟨expr⟩) which is given on the command line. See Expression evaluation mode for more details.

-E

Stop after preprocessing (.hspp file)

-C

Stop after generating C (.hc file)

-S

Stop after generating assembly (.s file)

-c

Stop after generating object (.o) file

This is the traditional batch-compiler mode, in which GHC can compile source files one at a time, or link objects together into an executable. See Batch compiler mode.

-M

Dependency-generation mode. In this mode, GHC can be used to generate dependency information suitable for use in a Makefile. See Dependency generation.

--frontend ⟨module⟩

Run GHC using the given frontend plugin. See Frontend plugins for details.

--mk-dll

DLL-creation mode (Windows only). See Creating a DLL.

--help
-?

Cause GHC to spew a long usage message to standard output and then exit.

--show-iface ⟨file⟩

Read the interface in ⟨file⟩ and dump it as text to stdout. For example ghc --show-iface M.hi.

--supported-extensions
--supported-languages

Print the supported language extensions.

--show-options

Print the supported command line options. This flag can be used for autocompletion in a shell.

--info

--version
-V

Print a one-line string including GHC’s version number.

--numeric-version

Print GHC’s numeric version number only.

--print-libdir

hello
olleh


### 7.1.5.3. Batch compiler mode¶

In batch mode, GHC will compile one or more source files given on the command line.

The first phase to run is determined by each input-file suffix, and the last phase is determined by a flag. If no relevant flag is present, then go all the way through to linking. This table summarises:

Phase of the compilation system Suffix saying “start here” Flag saying “stop after” (suffix of) output file
literate pre-processor .lhs   .hs
C pre-processor (opt.) .hs (with -cpp) -E .hspp
Haskell compiler .hs -C, -S .hc, .s
C compiler (opt.) .hc or .c -S .s
assembler .s -c .o
linker ⟨other⟩   a.out

Thus, a common invocation would be:

ghc -c Foo.hs


to compile the Haskell source file Foo.hs to an object file Foo.o.

Note

What the Haskell compiler proper produces depends on what backend code generator is used. See GHC Backends for more details.

Note

Pre-processing is optional, the -cpp flag turns it on. See Options affecting the C pre-processor for more details.

Note

The option -E runs just the pre-processing passes of the compiler, dumping the result in a file.

Note

The option -C is only available when GHC is built in unregisterised mode. See Unregisterised compilation for more details.

#### 7.1.5.3.1. Overriding the default behaviour for a file¶

As described above, the way in which a file is processed by GHC depends on its suffix. This behaviour can be overridden using the -x ⟨suffix⟩ option:

-x ⟨suffix⟩

Causes all files following this option on the command line to be processed as if they had the suffix ⟨suffix⟩. For example, to compile a Haskell module in the file M.my-hs, use ghc -c -x hs M.my-hs.

## 7.1.6. Verbosity options¶

See also the --help, --version, --numeric-version, and --print-libdir modes in Modes of operation.

-v

The -v option makes GHC verbose: it reports its version number and shows (on stderr) exactly how it invokes each phase of the compilation system. Moreover, it passes the -v flag to most phases; each reports its version number (and possibly some other information).

Please, oh please, use the -v option when reporting bugs! Knowing that you ran the right bits in the right order is always the first thing we want to verify.

-v⟨n⟩

To provide more control over the compiler’s verbosity, the -v flag takes an optional numeric argument. Specifying -v on its own is equivalent to -v3, and the other levels have the following meanings:

-v0
Disable all non-essential messages (this is the default).
-v1
Minimal verbosity: print one line per compilation (this is the default when --make or --interactive is on).
-v2
Print the name of each compilation phase as it is executed. (equivalent to -dshow-passes).
-v3
The same as -v2, except that in addition the full command line (if appropriate) for each compilation phase is also printed.
-v4
The same as -v3 except that the intermediate program representation after each compilation phase is also printed (excluding preprocessed and C/assembly files).
-fprint-potential-instances

When GHC can’t find an instance for a class, it displays a short list of some in the instances it knows about. With this flag it prints all the instances it knows about.

-fhide-source-paths

Starting with minimal verbosity (-v1, see -v), GHC displays the name, the source path and the target path of each compiled module. This flag can be used to reduce GHC’s output by hiding source paths and target paths.

The following flags control the way in which GHC displays types in error messages and in GHCi:

-fprint-unicode-syntax

When enabled GHC prints type signatures using the unicode symbols from the -XUnicodeSyntax extension. For instance,

ghci> :set -fprint-unicode-syntax
ghci> :t +v (>>)
(>>) ∷ Monad m ⇒ ∀ a b. m a → m b → m b

-fprint-explicit-foralls

Using -fprint-explicit-foralls makes GHC print explicit forall quantification at the top level of a type; normally this is suppressed. For example, in GHCi:

ghci> let f x = x
ghci> :t f
f :: a -> a
ghci> :set -fprint-explicit-foralls
ghci> :t f
f :: forall a. a -> a


However, regardless of the flag setting, the quantifiers are printed under these circumstances:

• For nested foralls, e.g.

ghci> :t GHC.ST.runST
GHC.ST.runST :: (forall s. GHC.ST.ST s a) -> a

• If any of the quantified type variables has a kind that mentions a kind variable, e.g.

ghci> :i Data.Type.Equality.sym
Data.Type.Equality.sym ::
forall k (a :: k) (b :: k).
(a Data.Type.Equality.:~: b) -> b Data.Type.Equality.:~: a
-- Defined in Data.Type.Equality


This flag also enables the printing of inferred type variables inside braces. See Inferred vs. specified type variables.

-fprint-explicit-kinds

Using -fprint-explicit-kinds makes GHC print kind arguments in types, which are normally suppressed. This can be important when you are using kind polymorphism. For example:

ghci> :set -XPolyKinds
ghci> data T a (b :: l) = MkT
ghci> :t MkT
MkT :: forall k l (a :: k) (b :: l). T a b
ghci> :set -fprint-explicit-kinds
ghci> :t MkT
MkT :: forall k l (a :: k) (b :: l). T @{k} @l a b
ghci> :set -XNoPolyKinds
ghci> :t MkT
MkT :: T @{*} @* a b


In the output above, observe that T has two kind variables (k and l) and two type variables (a and b). Note that k is an inferred variable and l is a specified variable (see Inferred vs. specified type variables), so as a result, they are displayed using slightly different syntax in the type T @{k} @l a b. The application of l (with @l) is the standard syntax for visible type application (see Visible type application). The application of k (with @{k}), however, uses a hypothetical syntax for visible type application of inferred type variables. This syntax is not currently exposed to the programmer, but it is nevertheless displayed when -fprint-explicit-kinds is enabled.

-fprint-explicit-runtime-reps

When -fprint-explicit-runtime-reps is enabled, GHC prints RuntimeRep type variables for levity-polymorphic types. Otherwise GHC will default these to LiftedRep. For example,

ghci> :t ($) ($) :: (a -> b) -> a -> b
ghci> :set -fprint-explicit-runtime-reps
ghci> :t ($) ($)
:: forall (r :: GHC.Types.RuntimeRep) a (b :: TYPE r).
(a -> b) -> a -> b

-fprint-explicit-coercions

Using -fprint-explicit-coercions makes GHC print coercions in types. When trying to prove the equality between types of different kinds, GHC uses type-level coercions. Users will rarely need to see these, as they are meant to be internal.

-fprint-equality-relations

Using -fprint-equality-relations tells GHC to distinguish between its equality relations when printing. For example, ~ is homogeneous lifted equality (the kinds of its arguments are the same) while ~~ is heterogeneous lifted equality (the kinds of its arguments might be different) and ~# is heterogeneous unlifted equality, the internal equality relation used in GHC’s solver. Generally, users should not need to worry about the subtleties here; ~ is probably what you want. Without -fprint-equality-relations, GHC prints all of these as ~. See also Equality constraints.

-fprint-expanded-synonyms

When enabled, GHC also prints type-synonym-expanded types in type errors. For example, with this type synonyms:

type Foo = Int
type Bar = Bool
type MyBarST s = ST s Bar


This error message:

Couldn't match type 'Int' with 'Bool'
Expected type: ST s Foo
Actual type: MyBarST s


Becomes this:

Couldn't match type 'Int' with 'Bool'
Expected type: ST s Foo
Actual type: MyBarST s
Type synonyms expanded:
Expected type: ST s Int
Actual type: ST s Bool

-fprint-typechecker-elaboration

When enabled, GHC also prints extra information from the typechecker in warnings. For example:

main :: IO ()
main = do
return $let a = "hello" in a return ()  This warning message: A do-notation statement discarded a result of type ‘[Char]’ Suppress this warning by saying ‘_ <- ($) return let a = "hello" in a’
or by using the flag -fno-warn-unused-do-bind


Becomes this:

A do-notation statement discarded a result of type ‘[Char]’
Suppress this warning by saying
‘_ <- (\$)
return
let
AbsBinds [] []
{Exports: [a <= a
<>]
Exported types: a :: [Char]
[LclId, Str=DmdType]
Binds: a = "hello"}
in a’
or by using the flag -fno-warn-unused-do-bind

-fdiagnostics-color=⟨always|auto|never⟩

Causes GHC to display error messages with colors. To do this, the terminal must have support for ANSI color codes, or else garbled text will appear. The default value is auto, which means GHC will make an attempt to detect whether terminal supports colors and choose accordingly.

The precise color scheme is controlled by the environment variable GHC_COLORS (or GHC_COLOURS). This can be set to colon-separated list of key=value pairs. These are the default settings:

header=:message=1:warning=1;35:error=1;31:fatal=1;31:margin=1;34


Each value is expected to be a Select Graphic Rendition (SGR) substring. The formatting of each element can inherit from parent elements. For example, if header is left empty, it will inherit the formatting of message. Alternatively if header is set to 1 (bold), it will be bolded but still inherits the color of message.

Currently, in the primary message, the following inheritance tree is in place:

• message
• header
• warning
• error
• fatal

In the caret diagnostics, there is currently no inheritance at all between margin, warning, error, and fatal.

The environment variable can also be set to the magical values never or always, which is equivalent to setting the corresponding -fdiagnostics-color flag but with lower precedence.

-fdiagnostics-show-caret

Controls whether GHC displays a line of the original source code where the error was detected. This also affects the associated caret symbol that points at the region of code at fault. The flag is on by default.

-ferror-spans

Causes GHC to emit the full source span of the syntactic entity relating to an error message. Normally, GHC emits the source location of the start of the syntactic entity only.

For example:

test.hs:3:6: parse error on input where'


becomes:

test296.hs:3:6-10: parse error on input where'


And multi-line spans are possible too:

test.hs:(5,4)-(6,7):
Conflicting definitions for a'
Bound at: test.hs:5:4
test.hs:6:7
In the binding group for: a, b, a


Note that line numbers start counting at one, but column numbers start at zero. This choice was made to follow existing convention (i.e. this is how Emacs does it).

-freverse-errors

Causes GHC to output errors in reverse line-number order, so that the errors and warnings that originate later in the file are displayed first.

-H ⟨size⟩

Set the minimum size of the heap to ⟨size⟩. This option is equivalent to +RTS -Hsize, see RTS options to control the garbage collector.

-Rghc-timing

Prints a one-line summary of timing statistics for the GHC run. This option is equivalent to +RTS -tstderr, see RTS options to control the garbage collector.

## 7.1.7. Platform-specific Flags¶

Some flags only make sense for particular target platforms.

-msse2

(x86 only, added in GHC 7.0.1) Use the SSE2 registers and instruction set to implement floating point operations when using the native code generator. This gives a substantial performance improvement for floating point, but the resulting compiled code will only run on processors that support SSE2 (Intel Pentium 4 and later, or AMD Athlon 64 and later). The LLVM backend will also use SSE2 if your processor supports it but detects this automatically so no flag is required.

SSE2 is unconditionally used on x86-64 platforms.

-msse4.2

(x86 only, added in GHC 7.4.1) Use the SSE4.2 instruction set to implement some floating point and bit operations when using the native code generator. The resulting compiled code will only run on processors that support SSE4.2 (Intel Core i7 and later). The LLVM backend will also use SSE4.2 if your processor supports it but detects this automatically so no flag is required.

## 7.1.8. Miscellaneous flags¶

Some flags only make sense for a particular use case.

-ghcversion-file ⟨path to ghcversion.h⟩

When GHC is used to compile C files, GHC adds package include paths and includes ghcversion.h directly. The compiler will lookup the path for the ghcversion.h file from the rts package in the package database. In some cases, the compiler’s package database does not contain the rts package, or one wants to specify a specific ghcversions.h to be included. This option can be used to specify the path to the ghcversions.h` file to be included. This is primarily intended to be used by GHC’s build system.