Build Seed7


The way to compile the interpreter depends on the operating system and the development tools used. You need a stand-alone C compiler and a make utility to compile the interpreter. A C compiler, which is only usable from an IDE, is not so useful, since some Seed7 programs (e.g. The Seed7 compiler s7c) need to call the C compiler as well. In case a make utility is missing under Windows the program make7.exe can be downloaded from

https://sourceforge.net/projects/seed7/files/bin

In this directory is also a Seed7 installer for Windows (seed7_05_20230709_win.exe).


THE MAKEFILES

Several makefiles are prepared in 'seed7/src' for various combinations of operating system, make utility, C compiler and shell:

makefile nameoperating system make prog C compiler shell
mk_linux.mak Linux/Unix/BSD (g)make gcc sh
mk_clang.mak Linux/Unix/BSD (g)make clang sh
mk_icc.mak Linux/Unix/BSD (g)make icc sh
mk_tcc_l.mak Linux/Unix/BSD (g)make tcc sh
mk_cygw.mak Windows (Cygwin) (g)make gcc sh
mk_msys.mak Windows (MSYS) mingw32-make gcc sh
mk_mingw.mak Windows (MinGW) mingw32-make gcc cmd.exe
mk_nmake.mak Windows (MinGW) nmake gcc cmd.exe
mk_msvc.mak Windows (MSVC) nmake cl cmd.exe
mk_bcc32.mak Windows (bcc32) make bcc32 cmd.exe
mk_bccv5.mak Windows (bcc32) make bcc32 V5.5 cmd.exe
mk_clangw.mak Windows (clang) (g)make clang cmd.exe
mk_tcc_w.mak Windows (tcc) (g)make tcc cmd.exe
mk_djgpp.mak DOS (g)make gcc cmd.exe
mk_osx.mak macOS make gcc sh
mk_osxcl.mak macOS make clang sh
mk_freebsd.mk FreeBSD make clang/gcc sh
mk_emccl.mak Linux/Unix/BSD make emcc + gcc sh
mk_emccw.mak Windows (emcc) mingw32-make emcc + gcc cmd.exe

Detailed explanations how to compile the interpreter (s7) can be found in the chapters below. In the optimal case you just copy one makefile from above to 'makefile' and do (with the corresponding make program):

make depend
make

When the interpreter (s7) is compiled successfully the executable and the libraries are placed in the 'bin' directory. Additionally a symbolic link to the executable is placed in the 'prg' directory (Under Windows symbolic links are not supported, so a copy of the executable is placed in the 'prg' directory).


REMAINING BUILD STEPS

After a successful compilation of the Seed7 interpreter (s7) the Seed7 compiler (s7c) can be compiled with:

make s7c

The compiler executable is copied to the 'bin' directory.

You can check interpreter and compiler with a test suite (chk_all.sd7). This step is optional and it will take some time. It is recommended to run the test suite but you can also run it later. To run the test suite do:

make test

There are some tools written in Seed7. If you compile these tools now they are installed with the final step. You can skip this step and compile (and install) the tools later. Compile the tools with:

make utils

Finally Seed7 can be installed with:

sudo make install

To do several compilation attempts in succession you need to execute the command

make clean

before you do 'make depend' again.


COMPILING UNDER LINUX

For Linux the compilation process is the simplest. The file 'makefile' is (almost) identical to 'mk_linux.mak' and that file is already prepared to compile under Linux. Go to the 'seed7/src' directory and type:

make depend
make

After the compilation the interpreter executable can be found in the 'bin' directory and the 'prg' directory will contain a link to the executable.

In some distributions the X11 library directory is not in $PATH. In this case the value of the SYSTEM_DRAW_LIBS variable in the makefile must be changed. Replace the -lX11 by the path and the name of your libXll (e.g. /usr/X11/lib/libX11.so or /usr/X11R6/lib/libX11.a).

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS.


COMPILING UNDER BSD AND UNIX

Probably you can do it just the same way as under Linux. If there are problems and the makefile needs to be adjusted please let me know. If you send the changed makefile to me I can include it in the release.


COMPILING UNDER WINDOWS WITH GCC FROM MINGW

If gcc and mingw32-make from MinGW are installed use a console, go to the 'seed7\src' directory and type:

copy /Y mk_mingw.mak makefile
mingw32-make depend
mingw32-make

To compile under MSYS (a Unix shell window from the MinGW tools) go to the 'seed7/src' directory and type:

cp mk_msys.mak makefile
mingw32-make depend
mingw32-make

If gmake uses Unix shell commands even if started from a Windows console, you can use gmake together with mk_msys.mak from a Windows console also.

To compile with gcc from MinGW and nmake from Windows use a console, go to the 'seed7\src' directory and type:

copy /Y mk_nmake.mak makefile
nmake depend
nmake

After the compilation the interpreter executable can be found in the 'bin' directory and it is also copied to prg/s7.exe.

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS. Keep in mind to use the same 'make' utility as before.


COMPILING UNDER WINDOWS WITH CL FROM MSVC

To compile Seed7 with cl and nmake (C compiler and make utility from a big big software company) it is necessary, that cl and nmake can be executed from a console window. The script vcvarsall.bat sets up the environment for cl and nmake. This script is in the directory

...Visual Studio\2019\Community\VC\Auxiliary\Build

or for a professional installation in the directory

...Visual Studio\2019\Professional\VC\Auxiliary\Build

Note that 2019 is the year of the MSVC release. In a 32-bit operating system use a console and execute:

...Studio\2019\Community\VC\Auxiliary\Build\vcvarsall x86

In older versions of MSVS the script was named vcvars32.bat. In a 64-bit operating system you need to execute:

...Studio\2019\Community\VC\Auxiliary\Build\vcvarsall x64

Afterwards the environment is set up for MSVC. In newer versions of MSVC the include files and libraries are in the "Windows Kits" directory. In this case it is necessary to define the environment variables INCLUDE and LIB. To do that look for: Control Panel -> System and Security -> System -> Advanced system settings -> Environment Variables. With "Windows Kits" the include files are in the directory:

...Windows Kits\<main_version>\Include\<detailled_version>

The actual include files are in the sub directories 'um', 'ucrt' and 'shared'. Add the absolute paths of these three directories to the INCLUDE environment variable. E.g.:

C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\um
C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\ucrt
C:\Program Files (x86)\Windows Kits\10\Include\10.0.19041.0\shared

On such installations the libraries are in the directory:

...Windows Kits\<main_version>\Lib\<detailled_version>

For 32-bit compilation the actual libraries are in the sub directories 'um\x86' and 'ucrt\x86'. For 64-bit compilation the actual libraries are in the sub directories 'um\x64' and 'ucrt\x64'. Add the absolute paths of the 32-bit or 64-bit directories to the LIB environment variable. E.g.:

C:\Program Files (x86)\Windows Kits\10\Lib\10.0.19041.0\um\x86
C:\Program Files (x86)\Windows Kits\10\Lib\10.0.19041.0\ucrt\x86

or for 64-bit compilation:

C:\Program Files (x86)\Windows Kits\10\Lib\10.0.19041.0\um\x64
C:\Program Files (x86)\Windows Kits\10\Lib\10.0.19041.0\ucrt\x64

Don't forget to restart a console window after environment variables have changed.

After everything has been set up for MSVC you can go to the 'seed7\src' directory and type:

copy /Y mk_msvc.mak makefile
nmake depend
nmake

If it is necessary to compile again you need to do a

nmake clean

to clean the files from a previous compilation attempt.

After the compilation the interpreter executable can be found in the 'bin' directory and it is also copied to prg/s7.exe. Note that the build process saves all environment variables in the file seed7/bin/cl_env.ini. Every time the Seed7 compiler calls cl it restores the environment variables from the file seed7/bin/cl_env.ini beforehand. This change of the environment happens in the Seed7 compiler and the original environment of the console remains unchanged. This means that the Seed7 compiler (s7c) can be used without the need to call vcvarsall.bat.

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS. Keep in mind to use the 'nmake' command as before.


COMPILING UNDER WINDOWS WITH BCC32

To compile Seed7 with bcc32 and make (C compiler and make utility from a smaller software company) use a console, go to the 'seed7\src' directory and type:

copy /Y mk_bcc32.mak makefile
make depend
make

After the compilation the interpreter executable can be found in the 'bin' directory and it is also copied to prg/s7.exe. For the older version of bcc32 (Version 5.5, which is available in the Internet) use:

copy /Y mk_bccv5.mak makefile
make depend
make

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS.


COMPILING UNDER WINDOWS WITH TCC

To compile Seed7 with tcc (Tiny C Compiler) and mingw32-make you need version 0.9.27 of tcc. The support for tcc under Windows is experimental.

You need to decide, which archiver utility should be used to create libraries. In the makefile mk_tcc_w.mak the value ARCHIVER defines the archiver utility to be used. For the ar utility from MinGW (ar should be in the search path) use the following lines (this is the default):

ARCHIVER = ar -r
# ARCHIVER = tiny_libmaker

As an alternative the tiny_libmaker utility can be used. Note that tiny_libmaker supports only 32-bit executables. You might need to copy tiny_libmaker.exe to the directory tcc. You need to change mk_tcc_w.mak also. Use the lines:

# ARCHIVER = ar -r
ARCHIVER = tiny_libmaker

Additionally you need some things:

  • The include file winsock2.h (copy it to tcc\include)

Then use a console, go to the 'seed7\src' directory and type:

copy /Y mk_tcc_w.mak makefile
make depend
make

After the compilation the interpreter executable can be found in the 'bin' directory and it is also copied to prg/s7.exe. Note that tcc for Windows has some bugs. The program chk_all.sd7 shows this bugs. Therefore Seed7 does not support the compilation with tcc under windows officially.

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS. Keep in mind to use the same 'make' utility as before.


COMPILING UNDER WINDOWS WITH CYGWIN

To compile Seed7 with Cygwin several Cygwin packages need to be installed. From the category Devel the following packages are needed: binutils, gcc-core, libncurses-devel and make. From category X11 the package libX11-devel is needed. To use graphic programs it is necessary to install Cygwin/X. Instructions to install Cygwin/X can be found in the Internet. To test the installation start a Cygwin window and do:

make --version
gcc --version

If the Cygwin gcc does not have the name gcc it is necessary to edit the file 'mk_cygw.mak'. The line containing

CC = gcc

must be changed to define the name of your gcc. If gcc and make work well change to the 'seed7/src' directory and do:

cp mk_cygw.mak makefile
make depend
make

After the compilation the interpreter executable can be found in the 'bin' directory and the 'prg' directory will contain a link to the executable.

Cygwin sometimes has problems with symlinks. This is indicated by an error message which ends with:

bin/as.exe: cannot execute binary file

The presence of a Cygwin symlink problem can be checked in the directory /usr/i686-pc-cygwin/bin (or the corresponding directory mentioned in the error message). If the file /usr/i686-pc-cygwin/bin/as.exe is a small file with less than 100 bytes and starts with

!<symlink>

it is some Cygwin symlink. To fix this error just copy the symlink destination over the file as.exe (after making a backup of the original symlink file as.exe). Additionally it might also be necessary to do the same for ld.exe (and maybe for some other symlinks as well).

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS.


COMPILING WITH EMCC FROM EMSCRIPTEN

Emscripten provides a C compiler (emcc) that targets JavaScript and WebAssembly. How to download and install Emscripten is described here.

The makefiles mk_emccl.mak and mk_emccw.mak are provided for compiling with emcc under Linux and Windows. Besides emcc you need also gcc (under Windows use gcc from MinGW) and node.js. When you download emsdk you get also a version of node.js.

To compile Seed7 under Linux use a terminal window and activate the PATH and other environment variables for emcc (with 'source ./emsdk_env.sh' in the emsdk directory). Afterwards go to the directory 'seed7/src' (with cd) and type:

cp mk_emccl.mak makefile
make depend
make

To compile Seed7 under Windows use a console window and activate the PATH and other environment variables for emcc (with 'emsdk_env.bat' in the emsdk directory). Afterwards go to the directory 'seed7\src' (with cd) and type:

copy /Y mk_emccw.mak makefile
mingw32-make depend
mingw32-make

After compilation of Seed7 the Seed7 interpreter (s7.js) can be started (in the directory 'seed7/prg') with:

node s7.js hello

Note that the Emscripten version of Seed7 is experimental. Due to limitations of Emscripten and missing Seed7 driver libraries several things do not work as they should:

  • In Emscripten stdout is always line buffered.
  • Reading from stdin is not possible (EOF is reached immediately).
  • Processes cannot be started.
  • Sockets cannot be used.
  • Graphics from node.js would need libraries that access win32 or X11 via native calls.

When you execute

node s7.js aS7Program

you might get an error like:

exception thrown: RangeError: Maximum call stack size exceeded

In this case you can raise the allowed stack size. Execute node.js as follows (chose a sufficient stack-size):

node --stack-size=8192 s7.js chkstr

You can also raise the allowed stack size of node.js permanently. Under Windows this is done with the command editbin (from Visual-C). The stack is increased (in the directory of node.exe) with:

editbin /stack:33554432 node.exe

If the stack problems are triggered by emscripten the file .emscripten in your home directory must be also adjusted. Change the NODE_JS entry to:

NODE_JS = ['node', '--stack-size=8192']

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS. Keep in mind to use the same 'make' utility as before.


COMPILING UNDER DOS WITH DJGPP

You need gcc and make from DJGPP. Make sure that the search PATH leads to gcc and make from DJGPP. Additionally the environment variable DJGPP must be set to a path leading to DJGPP.ENV. Note that make from DJGPP has some limitations:

  • It does not understand the option -f. So copying mk_djgpp.mak to makefile is really mandatory.
  • It does not support redirecting command output. So you will see all errors and warnings created by test programs.

Use the command line, go to the 'seed7\src' directory and type:

copy /Y mk_djgpp.mak makefile
make depend
make

After the compilation the interpreter executable can be found in the 'bin' directory and it is also copied to prg/s7.exe. If your get errors you can try mk_djgp2.mak instead. DOS usually supports only files with 8.3 file names (8 Ascii character name + dot + 3 Ascii character extension). The C source and header files of Seed7 all use this convention. Seed7 library files use longer file names. The DOS version of Seed7 maps long file names like reference.s7i to REFERE~1.S7I. Dosbox uses the same mapping so it should just work. Dosemu uses a different mapping, so the libraries are not found. In that case I suggest to copy files with long names in the 'lib' directory:

copy reference.s7i REFERE~1.S7I
copy hashsetof.s7i HASHSE~1.S7I
copy environment.s7i ENVIRO~1.S7I
copy null_file.s7i NULL_F~1.S7I
copy external_file.s7i EXTERN~1.S7I
copy clib_file.s7i CLIB_F~1.S7I
copy enable_io.s7i ENABLE~1.S7i
copy graph_file.s7i GRAPH_~1.S7I

Note that the DOS version of Seed7 currently does not support graphics, sockets, processes and databases.

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS.


COMPILING UNDER MAC OS

To compile under macOS you need You need to install Xcode. Xcode can be installed from the App Store or from Apple's developer site if a specific version is needed.

For the next steps a terminal window is needed. A terminal window is opened by clicking the launchpad icon in the dock, typing terminal in the search field, then clicking on the terminal icon.

The Homebrew package manager needs to be installed. Use a terminal window and the following command to install it:

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

Homebrew uses the brew command to install software. If the brew command is unknown it might be necessary to add it to the PATH. If the zsh shell is used this can be done with:

(echo; echo 'eval "$(/opt/homebrew/bin/brew shellenv)"') >> ~/.zprofile
eval "$(/opt/homebrew/bin/brew shellenv)"

Depending on the version of macOS it might be necessary to also install XQuartz (the X11 support of macOS). XQuartz can be installed using brew:

brew install --cask xquartz

After installing XQuartz the computer should be rebooted. If the x11 lib is not found it can be installed using brew:

brew install libx11
brew install libxrender

If /usr/local/include/X11 does not exist the brew location must be linked to to /usr/local/include, so the C compiler can find it:

sudo mkdir /usr/local/include
sudo ln -s /opt/homebrew/include/X11 /usr/local/include/X11

In newer versions of macOS gcc is based on clang. You can use this gcc, but it is not the original gcc. To obtain the original gcc you have to download it from a repository like Homebrew. To compile Seed7 with gcc (the clang based gcc or the original gcc) start a terminal, go to the 'seed7/src' directory and type:

cp mk_osx.mak makefile
make depend
make

To compile Seed7 with clang use the following commands:

cp mk_osxcl.mak makefile
make depend
make

After the compilation the interpreter executable can be found in the 'bin' directory and the 'prg' directory will contain a link to the executable.

Under macOS the X11 library is usually found in '/usr/X11R6/lib'. For unknown reasons the Xcode (macOS) linker normally does not search libraries in '/usr/X11R6/lib'. Therefore 'mk_osx.mak' defines the following linker flag:

LDFLAGS = -L/usr/X11R6/lib

Besides this 'mk_osx.mak' is almost identical to 'mk_linux.mak' and 'makefile'. If the X11 library is in a different directory you need to change the LDFLAGS value to that directory.

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS.


COMPILING UNDER ANDROID

To compile under Android you need to install the termux app on the Android device. Afterwards use the following commands:

pkg install clang
pkg install make
pkg install git
git clone https://github.com/ThomasMertes/seed7.git
cd seed7/src

Under termux the command gcc refers also to clang. There are two ways to compile Seed7 with termux. The first way uses the makefile mk_clang.mak. To compile with mk_clang.mak use the following commands:

make -f mk_clang.mak depend
make -f mk_clang.mak
make -f mk_clang.mak s7c

The second way to compile Seed7 uses makefile. To compile with makefile use the following commands:

make depend
make
make s7c

The remaining steps to build and install Seed7 are explained above in the chapter REMAINING BUILD STEPS. If you decided to use the 'make' parameter

-f mk_cygw.mak

you must continue to use this parameter with 'make'.


PACKAGES FOR DATABASES UNDER LINUX

Seed7 supports database access. Therefore the client library packages of the databases must be installed. On my computer the names of the client library packages are:

MariaDb:  mariadb-client
Sqlite:  sqlite3
PostgreSql:  postgresql
ODBC:  unixODBC
Firebird:  firebird
SQL Server:  libtdsodbc0

The client libraries are sufficient for Seed7 to access the databases. Seed7 provides replacements for the C header files used by the database client libraries. To use the original header files under Linux database client development packages must be installed. On my computer the names of the client development packages are:

MariaDb:  libmariadb-devel
Sqlite:  sqlite3-devel
PostgreSql:  postgresql-devel
ODBC:  unixODBC-devel
Firebird:  libfbclient-devel
SQL Server:  freetds-devel

WHAT TO DO IF ERRORS HAPPEN DURING THE COMPILATION?

In most cases errors indicate that some development package of your distribution is missing. If your operating system is Linux, BSD or Unix not all development packages with header files might be installed. In this case you get some errors after typing 'make depend'.

  • Errors such as
    chkccomp.c:56:20: fatal error: stdlib.h: No such file or directory
    s7.c:30:20: fatal error: stdlib.h: No such file or directory
    

indicate that the development package of the C library is missing. I don't know the name of this package in your distribution (under Ubuntu it has the name libc6-dev), but you can search for C development libraries and header files.

  • Errors such as
    con_inf.c:54:18: error: term.h: No such file or directory
    kbd_inf.c:53:18: error: term.h: No such file or directory
    trm_inf.c:47:18: error: term.h: No such file or directory
    

indicate that the curses or ncurses development package is missing. I don't know the name of this package in your distribution (under Ubuntu it has the name libncurses5-dev and under openSUSE Tumbleweed the name is ncurses-devel), but you can search in your package manager for a curses/ncurses package which mentions that it contains the header files. To execute programs you need also to install the non-developer package of curses/ncurses (in most cases it will already be installed because it is needed by other packages).

  • Errors such as
    drw_x11.c:38:19: error: X11/X.h: No such file or directory
    drw_x11.c:39:22: error: X11/Xlib.h: No such file or directory
    drw_x11.c:40:23: error: X11/Xutil.h: No such file or directory
    drw_x11.c:45:24: error: X11/keysym.h: No such file or directory
    

indicate that the X11 development package is missing. Under Ubuntu this package has the name libx11-dev and is described as:

X11 client-side library (development headers)

Under openSUSE Tumbleweed this package is named libX11-devel and is described as:

Development files for the Core X11 protocol library

Note that under X11 'client' means: The program which wants to draw. A X11 'server' is the place where the drawings are displayed. So you have to search for a X11 client developer package with headers. If you use X11 in some way (you don't do everything from the text console) the non-developer package of X11 will already be installed.

  • Errors such as
    common.h:35:17: fatal error: gmp.h: No such file or directory
    

indicate that the GMP development package is missing. I don't know the name of this package in your distribution. Under Linux Mint this package has the name libgmp-dev and is described as: Multiprecision arithmetic library developers tools.

  • Errors such as
    gcc chkccomp.c -o chkccomp
    chkccomp.c:28:10: fatal error: base.h: No such file or directory
    compilation terminated.
    

    or

    del version.h
    process_begin: CreateProcess(NULL, del version.h, ...) failed.
    make (e=2): The system cannot find the file specified.
    mingw32-make: *** [clean] Error 2
    

indicate that your makefile contains commands for the cmd.exe (or command.com) Windows console, but your 'make' program uses a Unix shell (/usr/bin/sh) to execute them. Either use a makefile which uses Unix shell commands (e.g. mk_msys.mak or mk_cygw.mak) or take care that the 'make' program uses cmd.exe (or command.com) to execute the commands.

  • Errors such as
    s7.c:28:21: error: version.h: No such file or directory
    

indicate that you forgot to run 'make depend' before running 'make'. Since such an attempt produces several unneeded files it is necessary now to run 'make clean', 'make depend' and 'make'.

  • The 'make' utility sometimes writes an error like
    depend:1: *** multiple target patterns.  Stop.
    

This indicates that an old 'depend' file (probably a relict of an earlier 'make' command with a different makefile) does not fit to the current makefile. To get rid of this error remove the file 'depend'. After that the compilation process must be started from scratch with 'make clean', 'make depend' and 'make'.

  • If you use bcc32 an error like
    Error E2194: Could not find file 'Studio\7.0\bin\bcc32.exe'
    

can happen. This indicates that the search path where bcc32.exe is found contains a space and the 'make' program is not capable to handle spaces in a search path. Creating the batch file 'seed7\src\bcc32.bat' with the content

bcc32.exe %*

helps. As alternate solution 'bcc32.exe' and 'bcc32.cfg' can be copied to the 'seed7\src' directory (this solution should be avoided, since it does not consider updates of bcc32).

  • If you use cl (from msvc) an error like
    NMAKE : fatal error U1077: 'cl' : return code '0xc0000135'
    

indicates that you forgot to execute vcvars32 before executing 'make depend' or 'make'. A message box, which complains that mspdb100.dll was not found, indicates also that vcvars32 was not executed.

  • If you use clang under Windows an error like
    chkccomp.c:60:10: fatal error: 'stdlib.h' file not found
    

indicates that the include file search path is not set up. If clang is based on the VC toolchain it is necessary to call vcvars32 from the commandline to set up the environment.

  • A linker error like
    ld: library not found for -lX11
    

indicates that the linker was not able to find the X11 library. The X11 library can have the name libX11.so (dynamic library) or libX11.a (static library). You need to search for this library (the dynamic library should normally be preferred). After you found it the LDFLAGS definition in your makefile must be changed. E.g.: If you found libX11.so in '/usr/X11R6/lib' the LDFLAGS assignment should be changed from

LDFLAGS =

to

LDFLAGS = -L/usr/X11R6/lib
  • If you use icc an error like
    .../compiler/include/math.h(1216):
    error: identifier "_LIB_VERSION_TYPE" is undefined
      _LIBIMF_EXT _LIB_VERSIONIMF_TYPE _LIBIMF_PUBVAR _LIB_VERSIONIMF;
    

indicates, that the math.h include file of icc uses _LIB_VERSION_TYPE. Recently the support for_LIB_VERSION_TYPE has been removed from glibc. Fortunately there is a definition of _LIB_VERSION_TYPE in math.h just a few lines above in a part deactivated by an #if. The #if line starts with:

#if (!defined(__linux__) || !defined(__USE_MISC)) && ...

I added a condition and the line now starts with:

#if (1 || !defined(__linux__) || !defined(__USE_MISC)) && ...

This fixed the error.

  • Other errors

    If you got other errors I would like to know about. Please send an mail with detailed information to seed7-users@lists.sourceforge.net or to my mail address, which can be found at the Seed7 Homepage (look for Links). The detailed information should include:

    • Operating system
    • Distribution used
    • C compiler
    • The version of Seed7 you wanted to compile.
    • The complete log of error messages
    • The file src/version.h

WHAT ABOUT THE WARNINGS THAT HAPPEN DURING THE COMPILATION?

The warnings can usually be ignored. Seed7 is compiled with the highest warning level (-Wall). Additionally there are also some warnings requested (such as -Wstrict-prototypes) which are not part of -Wall. The warnings can be classified to the following cases:

  • Warnings about float used instead of double because of the prototype: There are functions which use float parameters or return float values. Gcc has the opinion that only double parameters and double results should be used and warns about that.
  • Warnings about conversion from double to float for isnan.
    fltlib.c:1144:27: warning: conversion from 'floatType' {aka 'double'} to 'float' may change value [-Wfloat-conversion]
     1144 |     if (unlikely(os_isnan(number) ||
    
    The conversion from double to float keeps NaN values and it does not convert a normal number into a NaN value. This way it is always safe to check for NaN after a double has been converted to float.
  • Warnings about unused parameter 'arguments'.
    actlib.c
    actlib.c(204): warning C4100: 'arguments': unreferenced formal parameter
    
    All primitive action functions use one parameter named 'arguments'. This is necessary to access primitive actions with function pointers. Some primitive actions do not use 'arguments' which causes this warning.
  • Warnings about signed/unsigned instead of unsigned/signed because of the prototype.
  • Warnings about 'variablename' may be used uninitialized.
    dcllib.c
    dcllib.c(512) : warning C4701: potentially uninitialized local variable 'created_object' used
    
    These are false complaints. Interestingly C compilers are not able to recognize if the states of two variables are connected. Such as a global fail_flag variable and a local condition variable (cond). The connection is: As long as fail_flag is FALSE the cond variable is initialized. If the fail_flag is TRUE the cond variable is not used and therefore it could be in an uninitialized state. At several places connected variable states are used which are not recognized by the optimizer and are therefore flagged with a warning. I accept such warnings in performance critical paths. I am not willing to do "unnecessary" initializations in performance critical paths of the program. At places that are not performance critical I do some of this "unnecessary" initializations just to avoid such warnings.
  • Warnings about unary minus operator applied to unsigned.
    int_rtl.c
    int_rtl.c(662): warning C4146: unary minus operator applied to unsigned type, result still unsigned
    
    Applying the unary minus operator to an unsigned value is done on purpose. In the two's complement representation the most negative integer has no corresponding positive integer. C considers negating the most negative number as undefined behavior. For unsigned integers there is no undefined behavior. So an unsigned is negated to have well defined behavior.
  • Warnings about unsafe functions that should be replaced.
    scanner.c
    scanner.c(203): warning C4996: 'strcpy': This function or variable may be unsafe. Consider using strcpy_s instead.
    
    There are doubts about the safety these "_s" functions. The buffer size must be provided as additional parameter. For code that does not consider buffer sizes there is a high probability that the size parameter is also wrong. For this reason these "safe" functions are not supported by many C libraries. So in portable C code they cannot be used.
  • Warnings about "deprecated" POSIX functions.
    traceutl.c
    traceutl.c(526): warning C4996: 'fileno': The POSIX name for this item is deprecated. Instead, use the ISO C and C++ conformant name: _fileno.
    
    The reasoning about ISO C and C++ conformant names is absurd. This is just an attemt to force a vendor lock-in. If all POSIX function names are changed the program will not compile elsewhere. To avoid the vendor lock-in the file os_decls.h defines macros like os_fileno which are used in the source files of Seed7. Depending on the operating system os_fileno is defined as fileno or _fileno. This avoids these strange warnings and a vendor lock-in.
  • Warnings about array subscript partly outside array bounds.
    s7.c:226:33: warning: array subscript 'struct rtlArrayStruct[0]' is partly outside array bounds of 'unsigned char[16]' [-Warray-bounds=]
    
    The structs of arrays and strings are defined with one element. When an array or string is created the actual number of elements is allocated with malloc(). For empty arrays and empty strings the allocated size is smaller than the corresponding struct (that contains one element). This is not a problem, because the non-existing element in of an empty array or string will never be accessed.
  • Warnings about unknown conversion type characters 'I' in format strings. The MinGW gcc checks format strings and warns about the type character 'I' format strings. These warnings can be ignored, because the actual C run-time library (msvcrt) accepts format strings with 'I'.
  • Warnings about passing argument with different width due to prototype: Some compilers write such warnings for formal boolean (boolType) parameters and actual boolean arguments. Since the types of parameter and argument are identical and a prototype is specified this warning can be considered as false alarm.

WHAT TO DO WITH ERRORS TRIGGERED BY SEED7 PROGRAMS?

Sometimes errors are triggered, when a Seed7 program runs. This can happen because of reasons unrelated to Seed7.

A run-time error like

error while loading shared libraries: libtinfo.so.5:
cannot open shared object file: No such file or directory

indicates that the library libtinfo.so.5 is missing. This happened, when starting a program that was dynamically linked to ncurses. Obviously the dependency of the ncurses package was set wrong. I fixed this by downloading libtinfo.so.5 and copying it to the directory /lib64.


HOW TO VERIFY THAT THE INTERPRETER WORKS CORRECT?

A comprehensive test of the s7 interpreter and the s7c compiler can be done in the directory 'prg' with the command:

./s7 chk_all

Under Windows using ./ might not work. Just omit the ./ and type:

s7 chk_all

The program 'chk_all' uses several check programs to do its work. First a check program is interpreted and the output is compared to a reference. Then the program is compiled and executed and this output is also checked. Finally the C code generated by the compiled compiler is checked against the C code generated by the interpreted compiler. The checks of the compiler are repeated with several compiler options. If everything works correct the output is (after the usual information from the interpreter):

compiling the compiler - okay
chkint ........... okay
chkovf ........... okay
chkflt ........... okay
chkbin ........... okay
chkchr ........... okay
chkstr ........... okay
chkidx ........... okay
chkbst ........... okay
chkarr ........... okay
chkprc ........... okay
chkbig ........... okay
chkbool ........... okay
chkenum ........... okay
chkbitdata ........... okay
chkset ........... okay
chkhsh ........... okay
chkfil ........... okay
chkexc ........... okay

This verifies that interpreter and compiler work correct.


HOW TO USE THE GMP LIBRARY?

The functions to operate with bigInteger values are defined in the file 'big_rtl.c'. This functions provide reasonable performance for the usual bigInteger computations. If for some reason the bigInteger performance of 'big_rtl.c' is not enough it is possible to replace 'big_rtl.c' with 'big_gmp.c'. The file 'big_gmp.c' uses the GMP library to do the bigInteger computations. To use the GMP library use the following steps:

You need the GMP library (one of gmp.lib/gmp.dll/gmp.a/gmp.so) and the gmp.h include file.

The use of GMP must be enabled explicit in the makefile. In the makefile (that you use) is an echo statement with:

#define USE_GMP 0

To enable GMP just change the 0 in this echo statement to 1. Remember to do a

make clean

after you changed the makefile. When Seed7 is compiled with

make depend

the program chkccomp.c checks if it is possible to compile and link a small test program with the GMP library. If the small test program works correct the definitions to use 'big_gmp.c' are written to version.h:

#define BIG_GMP_LIBRARY 2
#define BIGINT_LIB BIG_GMP_LIBRARY

Additionally the option to link the GMP library is added to SYSTEM_BIGINT_LIBS. SYSTEM_BIGINT_LIBS can be found in version.h:

#define SYSTEM_BIGINT_LIBS "-lgmp"

The file macros will also contain a definition of SYSTEM_BIGINT_LIBS:

SYSTEM_BIGINT_LIBS = -lgmp

If everything works as expected the GMP library will be used automatically.

If chkccomp.c does not succceed with the small GMP test program the file version.h will contain definitions to use 'big_rtl.c' as bigInteger library:

#define BIG_RTL_LIBRARY 1
#define BIGINT_LIB BIG_RTL_LIBRARY

In this case either the include file gmp.h is missing or the GMP library cannot be linked. By default chkccomp.c uses "-lgmp" as option to link the GMP library. If a different option should be used it can be added to the makefile used to compile Seed7. The section to create chkccomp.h: can be extended with a line like

echo "#define BIGINT_LIBS \" option to link GMP \"" >> chkccomp.h

to write a definition of BIGINT_LIBS to the file chkccomp.h.

The GMP library might be used by other libraries as well. In some cases the other users of the GMP library use a custom memory allocator. By design all users of GMP share the same custom memory allocator. Memory could be freed with the wrong allocator and this corrupts the heap. If you see buffer overflow errors or corrupt heap memory you should switch back to the default bigInteger library.


SOURCE FILES


The C code of Seed7 can be grouped into several categories: Interpreter main, Parser, Interpreter core, Primitive action functions, General helper functions, Runtime library, Drivers and Compiler data library.


INTERPRETER MAIN

The main() function of the interpreter calls the Parser to create an internal program representation and afterwards it calls the Interpreter core to execute this program.

s7.c  Main program of the s7 Interpreter.

The interpreter main is licensed under the GPL.


PARSER

The Seed7 parser (analyzer) reads Seed7 programs from a file or string and creates an internal representation for them.

analyze.c  Main procedure of the analyzing phase.
syntax.c  Generate new syntax descriptions out of expressions.
token.c  Procedures to maintain objects of type tokentype.
parser.c  Main procedures of the parser.
name.c  Enter an object in a specified declaration level.
type.c  Parse and assign a type of an object.
expr.c  Read the next expression from the source file.
atom.c  Read the next object from the source file.
object.c  Procedures to maintain objects and lists.
scanner.c  Read the next symbol from the source file.
literal.c  Read next char or string literal from the source file.
numlit.c  Read the next number from the source file.
findid.c  Procedures to maintain the identifier table.
error.c  Submit normal compile time error messages.
infile.c  Procedures to open, close and read the source file.
libpath.c  Procedures to manage the include library search path.
symbol.c  Maintains the current symbol of the scanner.
info.c  Procedures for compile time info.
stat.c  Procedures for maintaining an analyze phase statistic.
fatal.c  Submit fatal compile time error messages.
match.c  Procedures to do static match on expressions.
act_comp.c Primitive actions for the action type.
prg_comp.c Primitive actions for the program type.

The Seed7 parser is licensed under the GPL.


INTERPRETER CORE

The interpreter core executes the internal representation of a program. In doing so primitive action functions are called.

exec.c  Main interpreter procedures.
doany.c  Procedures to call several Seed7 functions from C.
objutl.c  isit_.. and bld_.. functions for primitive datatypes.

The interpreter core is licensed under the GPL.


PRIMITIVE ACTION FUNCTIONS

The *lib.c files contain interpreter functions which execute "PRIMITIVE ACTIONS" (in doing so they can call functions from the runtime library. The PRIMITIVE ACTION functions are called from the interpreter core.

actlib.c  ACTION (ACT_*) actions
arrlib.c  array (ARR_*) actions
biglib.c  bigInteger (BIG_*) actions
binlib.c  binary (BIN_*) actions
blnlib.c  boolean (BLN_*) actions
bstlib.c  byte string (BST_*) actions
chrlib.c  char (CHR_*) actions
cmdlib.c  Directory, file and system command (CMD_*) actions
conlib.c  Text console (CON_*) actions
dcllib.c  Declaration (DCL_*) actions
drwlib.c  Drawing (DRW_*) actions
enulib.c  Enumeration (ENU_*) actions
fillib.c  PRIMITIVE_FILE (FIL_*) actions
fltlib.c  float (FLT_*) actions
hshlib.c  hash (HSH_*) actions
intlib.c  integer (INT_*) actions
itflib.c  interface (ITF_*) actions
kbdlib.c  Keyboard (KBD_*) actions
lstlib.c  List (LST_*) actions
pcslib.c  Process (PCS_*) actions
pollib.c  Poll (POL_*) actions
prclib.c  proc/statement (PRC_*) actions
prglib.c  Program (PRG_*) actions
reflib.c  reference (REF_*) actions
rfllib.c  ref_list (RFL_*) actions
sctlib.c  struct (SCT_*) actions
setlib.c  set (SET_*) actions
soclib.c  PRIMITIVE_SOCKET (SOC_*) actions
sqllib.c  database and sqlStatement (SQL_*) actions
strlib.c  string (STR_*) actions
timlib.c  time and duration (TIM_*) actions
typlib.c  type (TYP_*) actions
ut8lib.c  utf8_file (UT8_*) actions

The primitive action functions are licensed under the GPL.


GENERAL HELPER FUNCTIONS

General helper functions are used by Parser, Interpreter core and Primitive action functions.

runerr.c  Runtime error and exception handling procedures.
option.c  Reads and interprets the command line options.
primitiv.c Table definitions for all primitive actions.
syvarutl.c Maintains the interpreter system variables.
traceutl.c Tracing and protocol procedures.
actutl.c  Conversion of strings to ACTIONs and back.
executl.c  Initialization operation procedures used at runtime.
blockutl.c Procedures to maintain objects of type blocktype.
entutl.c  Procedures to maintain objects of type entitytype.
identutl.c Procedures to maintain objects of type identtype.
chclsutl.c Compute the type of a character very quickly.
arrutl.c  Maintain objects of type arraytype.

The general helper functions are licensed under the GPL.


RUNTIME LIBRARY

The functions defined in the runtime library are called from interpreted and compiled code (The Seed7 runtime library is linked to every compiled Seed7 program).

arr_rtl.c  Primitive actions for the array type.
bln_rtl.c  Primitive actions for the boolean type.
bst_rtl.c  Primitive actions for the byte string type.
chr_rtl.c  Primitive actions for the integer type.
cmd_rtl.c  Directory, file and other system functions.
con_rtl.c  Primitive actions for console/terminal output.
dir_rtl.c  Primitive actions for the directory type.
drw_rtl.c  Platform idependent drawing functions.
fil_rtl.c  Primitive actions for the C library file type.
flt_rtl.c  Primitive actions for the float type.
hsh_rtl.c  Primitive actions for the hash map type.
int_rtl.c  Primitive actions for the integer type.
itf_rtl.c  Primitive actions for the interface type.
pcs_rtl.c  Platform idependent process handling functions.
set_rtl.c  Primitive actions for the set type.
soc_rtl.c  Primitive actions for the socket type.
sql_rtl.c  Database access functions.
str_rtl.c  Primitive actions for the string type.
tim_rtl.c  Time access using the C capabilities.
ut8_rtl.c  Primitive actions for the UTF-8 file type.
heaputl.c  Procedures for heap allocation and maintenance.
numutl.c  Numeric utility functions.
sigutl.c  Driver shutdown and signal handling.
striutl.c  Procedures to work with wide char strings.

The runtime library is licensed under the LGPL.


DRIVERS

The drivers are also part of the runtime library. Depending on operating system and C compiler the makefile decides, which driver is used.

big_gmp.c  Alternate bigInteger library based on GMP.
big_rtl.c  Functions for the built-in bigInteger support.
cmd_unx.c  Command functions which call the Unix API.
cmd_win.c  Command functions which call the Windows API.
con_cap.c  Driver for termcap console access.
con_con.c  Driver for conio console access.
con_cur.c  Driver for curses console access.
con_dos.c  Driver for dos console access.
con_emc.c  Driver for emcc (JavaScript) console access.
con_inf.c  Driver for terminfo console access.
con_tcp.c  Driver for termcap console access.
con_wat.c  Driver for watcom console access.
con_win.c  Driver for windows console access.
con_x11.c  Driver for X11 text console access.
dir_dos.c  Directory functions which call the Dos API.
dir_win.c  Directory functions which call the Windows API.
dll_dos.c  Dynamic link library support (dummy).
dll_unx.c  Dynamic link library (*.so) support.
dll_win.c  Dynamic link library (*.dll) support.
drw_dos.c  Graphic access using the dos capabilities.
drw_emc.c  Graphic access using the browser.
drw_win.c  Graphic access using the windows capabilities.
drw_x11.c  Graphic access using the X11 capabilities.
emc_utl.c  Emscripten utility functions.
fil_dos.c  File functions which call the Dos API.
fil_emc.c  File functions which use Emscripten.
fil_unx.c  File functions which call the Unix API.
fil_win.c  File functions which call the Windows API.
gkb_emc.c  Keyboard and mouse access using the browser.
gkb_rtl.c  Generic keyboard support for graphics keyboard.
gkb_win.c  Keyboard and mouse access for windows.
gkb_x11.c  Keyboard and mouse access with X11 capabilities.
kbd_inf.c  Driver for terminfo keyboard access.
kbd_poll.c Driver for terminfo keyboard access using poll().
kbd_rtl.c  Platform idependent console keyboard support.
pcs_dos.c  Process functions using the dos capabilities.
pcs_unx.c  Process functions which use the Unix API.
pcs_win.c  Process functions which use the Windows API.
pol_dos.c  Poll type and function using DOS capabilities.
pol_sel.c  Poll type and function based on select function.
pol_unx.c  Poll type and function using UNIX capabilities.
soc_none.c Dummy functions for the socket type.
sql_base.c Basic database functions.
sql_db2.c  Database access functions for Db2.
sql_fire.c Database access functions for Firebird/InterBase.
sql_ifx.c  Database access functions for Informix.
sql_lite.c Database access functions for SQLite.
sql_my.c  Database access functions for MariaDB and MySQL.
sql_oci.c  Database access functions for OCI.
sql_odbc.c Database access functions for the ODBC interface.
sql_post.c Database access functions for PostgreSQL.
sql_srv.c  Database access functions for MS SQL-Server.
sql_tds.c  Database access functions for Tabular Data Stream.
stat_win.c Define functions used by os_stat macros.
trm_cap.c  Driver for termcap screen access.
trm_inf.c  Driver for terminfo screen access.
tim_dos.c  Time functions which call the Dos API.
tim_emc.c  Time functions for Emscripten.
tim_unx.c  Time functions which call the Unix API.
tim_win.c  Time functions which call the Windows API.

The drivers are licensed under the LGPL.


COMPILER DATA LIBRARY

The functions in the compiler data library manipulate compiler and interpreter internal data such as "type", "ref_list" and "reference". The compiler data library is linked to the interpreter and to some compiled Seed7 programs such as the Seed7 compiler (s7c) itself.

typ_data.c Primitive actions for the type type.
rfl_data.c Primitive actions for the ref_list type.
ref_data.c Primitive actions for the reference type.
listutl.c  Procedures to maintain objects of type listtype.
flistutl.c Procedures for free memory list maintenance.
typeutl.c  Procedures to maintain objects of type typetype.
datautl.c  Procedures to maintain objects of type identtype.

The compiler data library is licensed under the GPL.


PROGRAMS USED BY THE MAKEFILES

The makefiles use programs to write definitions to version.h . These are stand-alone programs that are not linked to the interpreter or to the runtime library.

chkccomp.c  Check properties of C compiler and runtime.
setwpath.c  Set the search path (PATH variable) under Windows.
sudo.c  Execute command as administrator under Windows.

The programs used by the makefiles are licensed under the GPL.


MACROS WRITTEN TO VERSION.H BY THE MAKEFILE

One of the main jobs of the makefile is creating a version.h file. This is done with 'make depend' (or 'gmake depend' or 'nmake depend'). The version.h file contains several #defines which contain information over available features and the way they are available. Other #defines can be found in config.h.

PATH_DELIMITER: This is the path delimiter character used by the command shell of the operating system. It is defined as '/' for most operating systems, except for Windows, where it is defined as '\\'. The PATH_DELIMITER macro is used when the functions popen() and system() are called. Seed7 programs are portable and do not need to distinguish between different path delimiter characters. Instead Seed7 programs must always use '/' as path delimiter.
USE_WMAIN: Defined if the main function is named wmain. This is a way to support Unicode command line arguments under Windows. An alternate way to support Unicode command line arguments under Windows uses the functions getUtf16Argv and freeUtf16Argv (both defined in "cmd_win.c").
USE_WINMAIN: Defined if the main function is named WinMain.
USE_DO_EXIT: TRUE if the main function must be terminated with doExit().
SEARCH_PATH_DELIMITER: The paths in the PATH environment variable are separated with this character. Linux/Unix/BSD use ':' and Windows uses ';'. If it is unknown at compile-time it is set to 0 and at run-time a function tries to guess it.
AWAIT_WITH_POLL: The function timAwait() uses the function poll() to implement waiting for a time. Only one #define of AWAIT_WITH_xxx is allowed.
AWAIT_WITH_PPOLL: The function timAwait() uses the function ppoll() to implement waiting for a time. Only one #define of AWAIT_WITH_xxx is allowed.
AWAIT_WITH_SELECT: The function timAwait() uses the function select() to implement waiting for a time. Only one #define of AWAIT_WITH_xxx is allowed.
AWAIT_WITH_SIGACTION: The function timAwait() uses the functions sigaction(), setitimer(), pause() and a signal handler function to catch a SIGALRM signal. Only one #define of AWAIT_WITH_xxx is allowed.
AWAIT_WITH_SIGNAL: The function timAwait() uses the functions signal(), setitimer(), pause() and a signal handler function to catch a SIGALRM signal. Only one #define of AWAIT_WITH_xxx is allowed.
USE_TERMINFO: Defined if the console should be used with terminfo. Only one #define of USE_TERMxxx is allowed.
USE_TERMCAP: Defined if the console should be used with termcap. Only one #define of USE_TERMxxx is allowed.
SIGNAL_HANDLER_CAN_DO_IO: Defined if I/O in signal handlers works.
CONSOLE_WCHAR: Defined if the console uses wide characters.
CONSOLE_UTF8: Defined if the console uses UTF-8 characters.
OS_STRI_WCHAR: Defined if the system calls (os_...) use wide characters (type wchar_t) for string and path parameters. In this case functions like _wgetcwd(), wreaddir() and _wstati64() together with types like 'WDIR', 'wdirent' and 'struct _stati64' must be used. It is therefore necessary to define the os_... macros accordingly.
OS_STRI_UTF8: Defined if the system calls (os_...) use UTF-8 characters (type char) for string and path parameters. In this case functions like getcwd(), readdir() and stat() together with types like 'DIR', 'dirent' and 'struct stat' must be used. If this functions and types do not use the POSIX/SUS names it is necessary to define the os_... macros accordingly.
OS_STRI_USES_CODE_PAGE: Defined if the system calls (os_...) use characters (type char) encoded with a code page (only the code pages 437 and 850 are supported). In this case functions like getcwd(), readdir() and stat() together with types like 'DIR', 'dirent' and 'struct stat' must be used.
ESCAPE_SHELL_COMMANDS: Depending on the shell/os the C functions system() and popen() need to get processed shell commands. If the macro is defined, a backslash (\) is added before the chars tab, space, !, ", $, &, ', (, ), *, ;, <, >, ?, \, `, and |. If the macro is not defined (undef), the whole shell command is surrounded by double quotes (") if it contains a space, &, comma, ;, =, ^, ~ or the non-breaking space.
QUOTE_WHOLE_SHELL_COMMAND: Defined if shell commands, starting with double quotes ("), need to be quoted a again. In this case the shell command, together with its parameters, needs to be quoted with double quotes (The final command string starts with two double quotes).
OBJECT_FILE_EXTENSION: The extension used by the C compiler for object files (Several object files and libraries are linked together to an executable). Under Linux/Unix/BSD this is usually ".o" Under Windows this is ".o" for MinGW and Cygwin, but ".obj" for MSVC and bcc32.
LIBRARY_FILE_EXTENSION: The extension used by the linker for static libraries. Several object files can be grouped to a library. Under Linux/Unix/BSD this is usually ".a". Under Windows this is ".a" for MinGW and Cygwin, but ".lib" for other linkers.
EXECUTABLE_FILE_EXTENSION: The extension which is used by the operating system for executables. Since executable extensions are not used under Linux/Unix/BSD it is "" for them. Under Windows the value ".exe" is used.
LINKED_PROGRAM_EXTENSION: The extension of the file produced by compiling and linking a program. Normally this is identical to the EXECUTABLE_FILE_EXTENSION, but in case of Emscripten this is independent from the EXECUTABLE_FILE_EXTENSION.
INTERPRETER_FOR_LINKED_PROGRAM: Defines an interpreter that is used if compiler and linker create a file that must be interpreted.
C_COMPILER: Contains the command to call the stand-alone C compiler and linker (Most IDEs provide also a stand-alone compiler/linker). If the C compiler is called via a script C_COMPILER_SCRIPT is defined and C_COMPILER is not defined by a makefile. In that case C_COMPILER is defined by chkccomp.c together with the flag CALL_C_COMPILER_FROM_SHELL.
CPLUSPLUS_COMPILER: Contains the command to call the stand-alone C++ compiler and linker.
C_COMPILER_SCRIPT: Relative path of a script that calls the stand-alone C compiler and linker.
CALL_C_COMPILER_FROM_SHELL: Flag that is defined if the stand-alone C compiler and linker is called via a script (defined with C_COMPILER_SCRIPT).
CC_ERROR_FILDES: File descriptor to which the C compiler writes errors. The MSVC stand-alone C compiler (CL) writes the error messages to standard output (file descriptor 1). The C compilers of Linux/Unix/BSD and the compilers from MinGW and Cygwin write the error messages to the error output (file descriptor 2).
CC_FLAGS: Contains C compiler flags, which should be used when C programs are compiled.
CC_OPT_DEBUG_INFO: Contains the C compiler option to add source level debugging information to the object file.
CC_OPT_LINK_TIME_OPTIMIZATION: Contains the compiler option for link time optimization (e.g.: "-flto").
CC_OPT_NO_WARNINGS: Contains the C compiler option to suppress all warnings.
CC_OPT_OPTIMIZE_1: C compiler option to optimize with a level of -O1.
CC_OPT_OPTIMIZE_2: C compiler option to optimize with a level of -O2.
CC_OPT_OPTIMIZE_3: C compiler option to optimize with a level of -O3.
CC_OPT_TRAP_OVERFLOW: C compiler option to generate traps for signed integer overflow. An integer overflow in the compiled program will trigger the signal defined with OVERFLOW_SIGNAL. CC_OPT_TRAP_OVERFLOW is not defined, if the C compiler does not support such an option or if traps would not lead to a performance advantage.
CC_OPT_VERSION_INFO: C compiler option to write the C compiler version information.
CC_SOURCE_UTF8: Defined if the C compiler accepts UTF-8 encoded file names in #line directives. The file names from #line directives are used by the debugger to allow source code debugging.
CC_VERSION_INFO_FILEDES: File descriptor to which the C compiler writes its version info.
LINKER: Defined if C_COMPILER does just invoke the stand-alone C compiler. In that case LINKER contains the command to call the stand-alone linker.
LINKER_FLAGS: Standard linker options to link a compiled program. This contains options, that the linker always uses, and which are not covered by other LINKER_OPT_... settings.
LINKER_OPT_DEBUG_INFO: Linker option to add source level debugging information. With this option source level debugging information is added to the executable file. (e.g.: "-Z7" or "-v"). Many compiler/linker combinations don't need this option to do source level debugging.
LINKER_OPT_LTO_MANDATORY: TRUE if linking always requires the option CC_OPT_LINK_TIME_OPTIMIZATION.
LINKER_OPT_NO_DEBUG_INFO: Linker option to be used without source level debugging. This option can strip debug information (e.g.: "-Wl,--strip-debug").
LINKER_OPT_OUTPUT_FILE: Linker option to provide the output filename (e.g.: "-o "). If no such option exists the definition of LINKER_OPT_OUTPUT_FILE is omitted. In this case it is assumed that the linker replaces the OBJECT_FILE_EXTENSION of the file with the LINKED_PROGRAM_EXTENSION.
LINKER_OPT_SPECIAL_LIB: Linker option that needs to precede the special library. If no special library exists the definition of LINKER_OPT_SPECIAL_LIB is omitted.
LINKER_OPT_STACK_SIZE: Linker option to specify the stack size of the executable. The argument for LINKER_OPT_STACK_SIZE is the stack size in bytes as decimal number. It must follow LINKER_OPT_STACK_SIZE immediately. The Seed7 compiler uses an argument of either DEFAULT_STACK_SIZE or the value specified with the -S Seed7 compiler option. If no linker option for the stack size exists the definition of LINKER_OPT_STACK_SIZE is omitted.
LINKER_OPT_STATIC_LINKING: Contains the linker option to force static linking (e.g.: "-static").
DEFAULT_STACK_SIZE: Default stack size for a compiled executable. This value is used, if the Seed7 compiler is invoked without -S. If LINKER_OPT_STACK_SIZE exists DEFAULT_STACK_SIZE is used as argument for the stack size. Additionally DEFAULT_STACK_SIZE is used as argument for setupStack().
POTENTIAL_PARTIAL_LINKING_OPTIONS: A comma separated list of potential partial/incremental linking options. Defined if partial/incremental linking is possibly supported. This list is tested to find an option that supports partial linking. A working option is written to "version.h" and "macros" under the name LINKER_OPT_PARTIAL_LINKING. Usually LINKER_OPT_PARTIAL_LINKING is "-r" or "-r -nostdlib". In case partial works the source code can be compiled with the options -c and LINKER_OPT_PARTIAL_LINKING. This option produces a relocatable object as output. This is also known as partial linking. The tool objcopy is used also. Objcopy is used with the option -L symbolname which converts a global or weak symbol called symbolname into a local symbol. This way the symbol is not visible externally.
SYSTEM_LIBS: Options to link system libraries to a compiled program. This is intended for options to link libraries required by the Seed7 runtime library. E.g. libraries for math or socket.
SYSTEM_BIGINT_LIBS: Options to link system bigint libraries to a compiled program. This is intended for options to link libraries required by the Seed7 bigint.s7i runtime library (e.g.: "-lgmp").
SYSTEM_CONSOLE_LIBS: Options to link system console libraries to a compiled program. This is intended for options to link libraries required by the Seed7 console runtime library (e.g.: "-lncurses").
SYSTEM_DRAW_LIBS: Options to link system graphic libraries to a compiled program. This is intended for options to link libraries required by the Seed7 graphic runtime library (e.g.: "-lX11").
SYSTEM_DATABASE_LIBS: Options to link system database libraries to a compiled program. This is intended for options to link libraries required by the Seed7 database runtime libraries (e.g.: "-lmysqlclient").
SYSTEM_MATH_LIBS: Options to link system mathematic libraries to a compiled program. This is intended for options to link libraries required by the Seed7 math.s7i runtime library. (e.g.: "-lm").
DEFINE_COMMAND_LINE_TO_ARGV_W: Defined if the function CommandLineToArgvW() is missing or buggy. In this case the definition of CommandLineToArgvW() in cmd_win.c is be used instead.
ALLOW_REPLACEMENT_OF_SYSTEM_HEADERS: Defined in chkccomp.h, if system header files for X11 and ncurses can be replaced with header files provided by Seed7.
INT_DIV_BY_ZERO_POPUP: Defined if an integer division by zero may trigger a popup window. Consequently chkccomp.c defines CHECK_INT_DIV_BY_ZERO, to avoid the popup.
INT_DIV_OVERFLOW_INFINITE_LOOP: Defined if an integer overflow may trigger an infinite loop. In that case chkccomp.c defines INT_DIV_OVERFLOW and INT_REM_OVERFLOW as 0.

MACROS WRITTEN TO VERSION.H BY CHKCCOMP.C

The program chkccomp.c is compiled and executed by the makefile. Chkccomp checks properties of C compiler, include files, linker, run-time library and operating systems. The result is written to version.h

UNISTD_H_PRESENT: TRUE if the include file <unistd.h> is present.
HAS_SIGNAL: TRUE if the function signal() is present.
HAS_SIGACTION: TRUE if the function sigaction() is present.
SIGNAL_RESETS_HANDLER: TRUE if a signal resets the signal handling to SIG_DFL. If it is FALSE the signal handler stays unchanged, if a signal is raised. If a signal handler is set with signal() an incoming signal might trigger a reset of the signal handling. This depends on the C run-time. If a signal handler is set with sigaction() an incoming signal always leaves the signal handling unchanged.
restrict: Defined if the C compiler does not support the restrict keyword.
likely: Macro that expands to a __builtin_expect expression, if __builtin_expect is supported by the C compiler. Without support of __builtin_expect the macro expands to its argument.
unlikely: Macro that expands to a __builtin_expect expression, if __builtin_expect is supported by the C compiler. Without support of __builtin_expect the macro expands to its argument.
NORETURN: Macro that expands to __attribute__ ((noreturn)), if the C compiler supports this attribute. Without support for this attribute the macro expands to analyze empty expression.
MACRO_DEFS: String with macro definitions for likely, unlikely and NORETURN.
NO_SOCKETS: Defined as -1. The meaning is: There is no socket library.
UNIX_SOCKETS: Defined as 1. The meaning is: The operating system uses Unix sockets.
WINSOCK_SOCKETS: Defined as 2. The meaning is: The operating system uses Windows sockets.
SOCKET_LIB: Defines the socket library used. The value is one of NO_SOCKETS, UNIX_SOCKETS and WINSOCK_SOCKETS.
NO_DIRECTORY: Defined as -1. The meaning is: No header file with definitions for opendir(), readdir() and closedir() is present.
DIRENT_DIRECTORY: Defined as 1. The meaning is: The prototypes of opendir(), readdir() and closedir() are in the header file <dirent.h>.
DIRECT_DIRECTORY: Defined as 2. The meaning is: The prototypes of opendir(), readdir() and closedir() are in the header file <direct.h>.
DIRDOS_DIRECTORY: Defined as 3. The meaning is: The opendir(), readdir() and closedir() functions from dir_dos.c are used. This functions are based on _dos_findfirst() and _dos_findnext().
DIRWIN_DIRECTORY: Defined as 4. The meaning is: The wopendir(), wreaddir() and wclosedir() functions from dir_win.c are used. This functions are based on FindFirstFileW() and FindNextFileW(). Additionally the file dir_win.c contains also definitions of opendir(), readdir() and closedir(). This functions are based on FindFirstFileA() and FindNextFileA().
DIR_LIB: Defines the directory library used. The value is one of NO_DIRECTORY, DIRENT_DIRECTORY, DIRECT_DIRECTORY, DIRDOS_DIRECTORY and DIRWIN_DIRECTORY.
os_DIR: Type to be used instead of 'DIR' under the target operating system. If not defined 'DIR' is used.
os_dirent_struct: Type to be used instead of 'struct dirent' under the target operating system. If not defined 'struct dirent' is used.
os_opendir: Function to be used instead of opendir() under the target operating system. If not defined opendir() is used.
os_readdir: Function to be used instead of readdir() under the target operating system. If not defined readdir() is used.
os_closedir: Function to be used instead of closedir() under the target operating system. If not defined closedir() is used.
os_chdir: Function to be used instead of chdir() under the target operating system. If not defined chdir() is used.
OS_GETCWD_MAX_BUFFER_SIZE: Defined if there is a maximum buffer size supported by os_getcwd().
OS_GETCWD_RETURNS_SLASH: Defined if os_getcwd() returns a path with slashes instead of PATH_DELIMITER.
os_getcwd: Function to be used instead of getcwd() under the target operating system. If not defined getcwd() is used.
os_mkdir(path,mode): Function to be used instead of mkdir() under the target operating system. If not defined mkdir() is used. Under Windows the mkdir() function usually has only one parameter while under Unix/Linux/BSD mkdir() has two parameters.
os_rmdir: Function to be used instead of rmdir() under the target operating system. If not defined rmdir() is used.
os_chmod: Function to be used instead of chmod() under the target operating system. If not defined chmod() is used.
os_chown(name,uid,gid): Function to be used instead of chown() under the target operating system. If not defined chown() is used. If chown() is not supported this macro is defined empty.
os_utimbuf_struct: Type to be used instead of 'struct utimbuf' under the target operating system. If not defined 'struct utimbuf' is used.
USE_ALTERNATE_UTIME: Defined if alternate_utime should be used instead of os_utime_orig.
os_utime_orig: Defined if the function alternate_utime is used instead of the function defined with os_utime. In this case os_utime_orig refers to the original os_utime function and the macro os_utime is redefined to refer to alternate_utime.
os_utime: Function to be used instead of utime() under the target operating system. If not defined utime() is used.
os_remove: Function to be used instead of remove() under the target operating system. If not defined remove() is used.
os_rename: Function to be used instead of rename() under the target operating system. If not defined rename() is used.
os_system: Function to be used instead of system() under the target operating system. If not defined system() is used.
os_fopen: Function to be used instead of fopen() under the target operating system. If not defined fopen() is used.
os_popen: Function to be used instead of popen() under the target operating system. If not defined popen() is used.
os_pclose: Function to be used instead of pclose() under the target operating system. If not defined pclose() is used.
DEFINE_WGETENV: Defined if an own implementation of wgetenv should be used.
os_getenv: Function to be used instead of getenv() under the target operating system. If not defined getenv() is used.
os_setenv: Function to be used instead of setenv() under the target operating system. If not defined setenv() is used.
FILENO_WORKS_FOR_NULL: TRUE if the fileno() function works for NULL and returns -1.
FOPEN_SUPPORTS_CLOEXEC_MODE: TRUE if fopen() supports a mode of "e".
HAS_FCNTL_SETFD_CLOEXEC: TRUE if fcntl() supports F_SETFD,FD_CLOEXEC.
HAS_GETADDRINFO: TRUE if the function getaddrinfo() is available.
HAS_PIPE2: TRUE if the function pipe2() is available.
HAS_SNPRINTF: TRUE if the function snprintf() is available.
HAS_VSNPRINTF: TRUE if the function vsnprintf() is available.
HAS_POPEN: TRUE if the function popen() is available.
STAT_MISSING: Defined if the function stat() is missing. This macro is only used in chkccomp.c.
os_fstat: Function to be used instead of fstat() under the target operating system. If not defined fstat() is used.
os_lstat: Function to be used instead of lstat() under the target operating system. If not defined lstat() is used.
os_stat: Function to be used instead of stat() under the target operating system. If not defined stat() is used.
os_stat_struct: Type to be used for the output parameter of os_stat() and os_lstat(). If not defined 'struct stat' is used.
os_fstat_struct: Type to be used for the output parameter of os_fstat(). If not defined 'os_stat_struct' is used.
os_fseek: Function to be used instead of fseek() under the target operating system. If it is not defined os_fsetpos() or os_lseek() are used.
os_ftell: Function to be used instead of ftell() under the target operating system. If it is not defined os_fgetpos() or os_lseek() are used.
os_off_t: Type used for os_fseek(), os_ftell(), offsetSeek(), offsetTell() and seekFileLength().
CHAR_SIZE: Size of a char in bits.
SHORT_SIZE: Size of a short int in bits.
INT_SIZE: Size of an int in bits.
LONG_SIZE: Size of a long int in bits.
INTTYPE_SIZE: Size of an inttype in bits (either 32 or 64).
FLOAT_SIZE: Size of a float in bits.
DOUBLE_SIZE: Size of a double in bits.
FLOATTYPE_SIZE: Size of an inttype in bits (either FLOAT_SIZE or DOUBLE_SIZE).
POINTER_SIZE: Size of a pointer in bits.
OS_OFF_T_SIZE: Size of os_off_t in bits.
TIME_T_SIZE: Size of time_t in bits.
INT_RANGE_IN_FLOATTYPE_MAX: Maximum from the continuous range of integers that map to floats. All integers from -INT_RANGE_IN_FLOATTYPE_MAX to INT_RANGE_IN_FLOATTYPE_MAX can be converted to 'floatType' and back to 'intType' without loss.
MINIMUM_TRUNC_ARGUMENT: Minimum value that trunc() or round() can convert. Values below MINIMUM_TRUNC_ARGUMENT raise RANGE_ERROR, if trunc() or round() is applied to them.
MAXIMUM_TRUNC_ARGUMENT: Maximum value that trunc() or round() can convert. Values above MAXIMUM_TRUNC_ARGUMENT raise RANGE_ERROR, if trunc() or round() is applied to them.
PIXEL_ALPHA_MASK Mask for the alpha channel in a pixel. 0 if there is no alpha channel.
PIXEL_RED_MASK Mask for the red color in a pixel. 0 if there is no mapping and drwRgbColor() should be used.
PIXEL_GREEN_MASK Mask for the green color in a pixel. 0 if there is no mapping and drwRgbColor() should be used.
PIXEL_BLUE_MASK Mask for the blue color in a pixel. 0 if there is no mapping and drwRgbColor() should be used.
RGB_TO_PIXEL_FLAG_NAME Name of the variable deciding between macro and drwRgbColor(). "" if no variable needs to be checked.
POINT_LIST_INT_SIZE Size of an integer in a ''pointList'' object. A ''pointList'' consists of (x, y) coordinate pairs where each x and y has the size POINT_LIST_INT_SIZE.
POINT_LIST_ABSOLUTE TRUE if a ''pointList'' uses absolute coordinates. If it is FALSE the first (x, y) coordinate pair is absolute and the remaining (x, y) coordinates are relative to the previous coordinate pair.
POINT_LIST_ALIGNMENT Defined if ''pointList'' data (plist->mem) needs to be aligned. In this case it contains the number of zero bits (counting from the least significant bit) in the address of ''pointList'' data. The function pltAlign() checks if ''pointList'' data is aligned and returns an aligned copy if the original ''pointList'' data is not aligned.
NULL_DEVICE: Name of the NULL device. Under Linux/Unix/BSD this is "/dev/null". Under Windows this is "NUL:".
OS_PATH_HAS_DRIVE_LETTERS: TRUE if the absolute paths of the operating system use drive letters. Determined by chkccomp.c.
REDIRECT_FILDES_1: Shell parameter to redirect to the file descriptor 1. Under Linux/Unix/BSD and Windows this is ">". The file to which the standard output should be redirected must be appended. E.g.: >myFile.
REDIRECT_FILDES_2: Shell parameter to redirect to the file descriptor 2. Under Linux/Unix/BSD and Windows this is "2>". The file to which the error output should be redirected must be appended. E.g.: 2>myFile.
FTELL_SUCCEEDS_FOR_PIPE: The ftell() function should return -1 if it is called with a pipe (since a pipe is not seekable). This macro is TRUE if ftell() succeeds for pipes (it does not return -1). In this case the function improved_ftell is used which returns -1 if the check with fstat() does not verify that the parameter is a regular file.
FSEEK_SUCCEEDS_FOR_STDIN: TRUE if fseek(stdin, 0, SEEK_SET) succeeds (stdin is not redirected).
FOPEN_OPENS_DIRECTORIES: TRUE if fopen() is able to open directories (it does not return NULL).
FWRITE_WRONG_FOR_READ_ONLY_FILES: TRUE if fwrite() to a read only file succeeds (it does not return 0).
REMOVE_FAILS_FOR_EMPTY_DIRS: TRUE if remove() cannot remove an empty directory (it does not return 0).
HOME_DIR_ENV_VAR: Name of the environment variable that contains the path of the home directory.
INCLUDE_SYS_UTIME: Defined if utime() respectively _wutime() is defined in <sys/utime.h>.
USE_ALTERNATE_UTIME: Defined if utime() is not able to change the access and modification times of a directory (utime() does not return 0 and errno has the value EACCES).
RSHIFT_DOES_SIGN_EXTEND: The C standard specifies that the right shift of signed integers is implementation defined, if the shifted values are negative. The macro is TRUE if the sign of negative signed integers is preserved with a right shift ( -1 >> 1 == -1 ).
TWOS_COMPLEMENT_INTTYPE: TRUE if signed integers are represented as twos complement numbers. This allows some simplified range checks in compiled programs. This macro is defined if ~(-1) == 0 holds.
ONES_COMPLEMENT_INTTYPE: TRUE if signed integers are represented as ones complement numbers. This macro is currently not used.
LITTLE_ENDIAN_INTTYPE: TRUE if the byte ordering of integers is little endian.
BIG_ENDIAN_INTTYPE: TRUE if the byte ordering of integers is big endian.
MALLOC_ALIGNMENT: Contains the number of zero bits (counting from the least significant bit) in an address returned by malloc(). The memory management in the parser uses this alignment.
UNALIGNED_MEMORY_ACCESS_OKAY: Defined if integers can be read from unaligned memory positions. This macro is currently not used.
NO_EMPTY_STRUCTS: Defined if the C compiler considers an empty struct as syntax error.
HAS_GETRLIMIT: TRUE if the functions getrlimit() and setrlimit() are available.
MEMCMP_RETURNS_SIGNUM: TRUE if memcmp() returns only the values -1, 0 and 1.
HAS_WMEMCMP: TRUE if the function wmemcmp() is available.
WMEMCMP_RETURNS_SIGNUM: TRUE if wmemcmp() returns only the values -1, 0 and 1.
HAS_WMEMCHR: TRUE if the function wmemchr() is available.
HAS_WMEMSET: TRUE if the function wmemset() is available.
HAS_SETJMP: TRUE if the functions setjmp() and longjmp() are available.
HAS_SIGSETJMP: TRUE if the functions sigsetjmp() and siglongjmp() are available. If it is FALSE the functions setjmp() and longjmp() must be used instead.
HAS_SYMBOLIC_LINKS: TRUE if the operating system supports symlinks.
HAS_READLINK: TRUE if the function readlink() is available.
HAS_FIFO_FILES: TRUE if the operating system supports fifo (first in first out) files.
HAS_SELECT: TRUE if the function select() is available.
HAS_POLL: TRUE if the function poll() is available.
HAS_MMAP: TRUE if the function mmap() is available.
INT_DIV_OVERFLOW: Defines what happens in case of an integer division overflow. An integer division can overflow with: integer.first div -1 Possible values of INT_DIV_OVERFLOW are:
  1. the test program crashed.
  2. it triggers an endless loop.
  3. it triggers a popup window.
  4. it returns integer.first.
  5. it returns some other value.
  6. it raises SIGFPE.
  7. it raises SIGILL.
  8. it raises SIGABRT.
  9. it raises SIGTRAP.
INT_REM_OVERFLOW: Defines what happens in case of an integer remainder overflow. An integer remainder can overflow with: integer.first rem -1 Possible values of INT_REM_OVERFLOW are:
  1. the test program crashed.
  2. it triggers an endless loop.
  3. it triggers a popup window.
  4. it returns 0.
  5. it returns some other value.
  6. it raises SIGFPE.
  7. it raises SIGILL.
  8. it raises SIGABRT.
  9. it raises SIGTRAP.
DO_SIGFPE_WITH_DIV_BY_ZERO: TRUE if SIGFPE should be raised with an integer division by zero. If it is FALSE raise(SIGFPE) can be called instead. Under Windows it is necessary to trigger SIGFPE this way, to assure that the debugger can catch it. If the compiler (s7c) is called with the option -e the function triggerSigfpe() is used to raise SIGFPE, if an uncaught EXCEPTION occurs.
CHECK_INT_DIV_BY_ZERO: TRUE if integer divisions must be checked for a division by zero. This applies to the division operations div and mdiv. The generated C code should, if a division by zero occurs, raise the exception NUMERIC_ERROR instead of doing the illegal divide operation.
CHECK_INT_DIV_ZERO_BY_ZERO: TRUE if the C expression 0/0 might not trigger SIGFPE. C compilers assume that so called "undefined behavior" will not happen. According to the C standard a division by 0 triggers undefined behavior. This way a C compiler is allowed to optimize the expressions 0/0 and 0/variable to 0. Likewise the expression variable/variable can be optimized to 1. In Seed7 a division by zero is defined behavior, since it raises the exception NUMERIC_ERROR. This configuration setting applies to the division operations div and mdiv. The generated C code should, if a division by zero occurs, raise the exception NUMERIC_ERROR instead of allowing the C compiler to do its optimization.
CHECK_INT_REM_BY_ZERO: TRUE if integer remainder must be checked for a division by zero. This applies to the division operations rem and mod. The generated C code should, if a remainder by zero occurs, raise the exception NUMERIC_ERROR instead of doing the illegal divide operation.
CHECK_INT_REM_ZERO_BY_ZERO: TRUE if the C expression 0%0 might not trigger SIGFPE. C compilers assume that so called "undefined behavior" will not happen. According to the C standard a division by 0 triggers undefined behavior. This way a C compiler is allowed to optimize the expressions 0%0 and 0%variable to 0. Likewise the expression variable%variable can be optimized to 1. In Seed7 a division by zero is defined behavior, since it raises the exception NUMERIC_ERROR. This configuration setting applies to the division operations rem and mod. The generated C code should, if a division by zero occurs, raise the exception NUMERIC_ERROR instead of allowing the C compiler to do its optimization.
TURN_OFF_FP_EXCEPTIONS: In Seed7 floating point errors such as the division by zero should create values like Infinite and NaN which are defined in IEEE 754. Some C compilers/libraries raise exceptions for floating point errors. This macro is used to turn off such a behaviour.
DEFINE_MATHERR_FUNCTION: Some C compilers/libraries call the matherr() function for every floating point error and terminate the program, if the function is not present. To get the Seed7 behaviour of using the IEEE 754 values of Infinite and NaN this function must be defined and it must return 1.
DEFINE__MATHERR_FUNCTION: Some C compilers/libraries call the _matherr() function for every floating point error and terminate the program, if the function is not present. To get the Seed7 behaviour of using the IEEE 754 values of Infinite and NaN this function must be defined and it must return 1.
HAS_LOG2: TRUE if the function log2() is available.
HAS_CBRT: TRUE if the function cbrt() is available.
FLOAT_NAN_COMPARISON_OKAY: TRUE if comparisons between NaN and any other value return FALSE. Comparison refers to comparisons with == < > <= or >= .
FLOAT_ZERO_COMPARISON_OKAY: TRUE if comparisons consider negative zero as equal to positive zero. FALSE if this is not the case. Comparison refers to comparisons with == < > <= or >= .
SQRT_OF_NAN_OKAY: TRUE if sqrt(NaN) returns NaN.
SQRT_OF_NEGATIVE_OKAY: TRUE if sqrt(x) returns NaN for every x < 0.0.
EXP_OF_NAN_OKAY: TRUE if exp(NaN) returns NaN.
LDEXP_OF_NAN_OKAY: TRUE if ldexp(NaN, n) returns NaN.
LDEXP_OF_INFINITY_OKAY: TRUE if ldexp(Infinity, n) returns Infinity and ldexp(-Infinity, n) returns -Infinity.
LDEXP_SUBNORMAL_OKAY: TRUE if ldexp(x, n) works correctly for subnormal numbers.
FREXP_INFINITY_NAN_OKAY: TRUE if frexp(Infinity, &n) returns Infinity and frexp(-Infinity, &n) returns -Infinity and frexp(NaN, &n) returns NaN.
FREXP_SUBNORMAL_OKAY: TRUE if frexp(x, &n) works correctly for subnormal numbers.
FMOD_DIVIDEND_NAN_OKAY: TRUE if fmod(NaN, y) returns NaN.
FMOD_DIVISOR_NAN_OKAY: TRUE if fmod(x, NaN) returns NaN.
FMOD_DIVIDEND_INFINITY_OKAY: TRUE if both fmod(Infinity, y) and fmod(-Infinity, y) return NaN.
FMOD_DIVISOR_INFINITY_OKAY: TRUE if both fmod(x, Infinity) and fmod(x, -Infinity) return x and all of fmod(Infinity, Infinity), fmod(Infinity, -Infinity), fmod(-Infinity, Infinity) and fmod(-Infinity, -Infinity) return NaN.
FMOD_DIVISOR_ZERO_OKAY: TRUE if fmod(x, 0.0) returns NaN.
FLOAT_ZERO_TIMES_INFINITE_OKAY: TRUE if all of 0.0 * Infinity, 0.0 * -Infinity, Infinity * 0.0 and -Infinity * 0 return NaN.
NAN_MULTIPLICATION_OKAY: TRUE if both x * NaN and NaN * y return NaN.
NAN_DIVISION_OKAY: TRUE if both x / NaN and NaN / y return NaN.
MAX_ODD_FLOAT: Maximum float value that is odd. All larger float values are even.
MAX_ODD_DOUBLE: Maximum double value that is odd. All larger double values are even.
POW_OF_NAN_OKAY: TRUE if pow(NaN, 0.0) returns 1.0 and pow(NaN, anyOtherExponent) returns NaN.
POW_OF_ZERO_OKAY: TRUE if the pow() function works correct for a base of zero (0.0 or -0.0) and a negative exponent.
POW_OF_NEGATIVE_OKAY: TRUE if the pow() function works correct, if the base is negative.
POW_OF_ONE_OKAY: TRUE if the pow() function always returns 1.0 if the base is 1.0 (Even for an exponent of NaN).
POW_EXP_NAN_OKAY: TRUE if the pow() function always returns NaN if the exponent is NaN and the base is not 1.0 (pow(1.0, NaN) should return 1.0).
POW_EXP_MINUS_INFINITY_OKAY: TRUE if the pow() function works correct for an exponent of minus infinity.
STRTOD_ACCEPTS_INF: TRUE if strtod() accepts "INF" disregarding case and returns Infinity.
STRTOD_ACCEPTS_INFINITY: TRUE if strtod() accepts "INFINITY" disregarding case and returns Infinity.
STRTOD_ACCEPTS_NAN: TRUE if strtod() accepts "NAN" disregarding case and returns NaN.
STRTOD_ACCEPTS_SIGN_WITHOUT_DIGITS: TRUE if strtod() accepts "+" or "-" as numeric literal.
LOG_OF_NAN_OKAY: TRUE if log(NaN) returns NaN.
LOG_OF_ZERO_OKAY: TRUE if log(0.0) returns minus infinity.
LOG_OF_NEGATIVE_OKAY: TRUE if log() of a negative number returns NaN.
LOG10_OF_NAN_OKAY: TRUE if log10(NaN) returns NaN.
LOG10_OF_ZERO_OKAY: TRUE if log10(0.0) returns minus infinity.
LOG10_OF_NEGATIVE_OKAY: TRUE if log10() of a negative number returns NaN.
LOG2_OF_NAN_OKAY: TRUE if log2(NaN) returns NaN.
LOG2_OF_ZERO_OKAY: TRUE if log2(0.0) returns minus infinity.
LOG2_OF_NEGATIVE_OKAY: TRUE if log2() of a negative number returns NaN.
FLOAT_ZERO_DIV_ERROR: TRUE if floating point divisions by zero cause compilation errors. Some C compilers check if the dividend is 0.0 and classify a floating point division by zero as fatal compilation error. Some C compilers even trigger wrong behavior if the dividend is 0.0. In this case FLOAT_ZERO_DIV_ERROR is also TRUE. If FLOAT_ZERO_DIV_ERROR is TRUE the generated C code should avoid divisions by zero and generate code to return Infinity, -Infinity or NaN instead.
CHECK_FLOAT_DIV_BY_ZERO: TRUE if floating point divisions by zero don't conform to IEEE 754. According to IEEE 754 a floating point division by zero should return Infinity, -Infinity or NaN. In this case the compiler generates C code, which checks all float divisions ( / and /:= ) for division by zero. The generated C code should return Infinity, -Infinity or NaN instead of doing the divide operation.
PRINTF_MAXIMUM_FLOAT_PRECISION: Precision up to which writing a float with printf (using format %e or %f) will always work ok. This can be defined in a makefile and is used only in chkccomp.c.
PRINTF_FMT_E_MAXIMUM_FLOAT_PRECISION: Precision up to which writing a float with printf (using format %e) will always work ok.
LIMIT_FMT_E_MAXIMUM_FLOAT_PRECISION: Defined if a printf (using format %e) with a larger precision crashes. In that case it is defined with a low precision limit as string.
PRINTF_FMT_F_MAXIMUM_FLOAT_PRECISION: Precision up to which writing a float with printf (using format %f) will always work ok.
LIMIT_FMT_F_MAXIMUM_FLOAT_PRECISION: Defined if a printf (using format %f) with a larger precision crashes. In that case it is defined with a low precision limit as string.
ISNAN_WITH_UNDERLINE: Defined if the macro/function _isnan() is defined in <float.h> respectively <math.h> instead of isnan().
USE_ALTERNATE_LOCALTIME_R: Defined if the function alternate_localtime_r() should be used instead of localtime().
USE_LOCALTIME_R: Defined if the thread-safe function localtime_r() should be used instead of localtime().
USE_LOCALTIME_S: Defined if the thread-safe function localtime_s() should be used instead of localtime().
DEFINE_OS_ENVIRON: Defined if environ respectively _wenviron is not defined in <stdlib.h> or <unistd.h>. In this case environ respectively _wenviron must be defined.
INITIALIZE_OS_ENVIRON: Defined if environ respectively _wenviron is NULL unless getenv() is called.
LIMITED_CSTRI_LITERAL_LEN: 0 if the C compiler does not limit the length of string literals and at run-time the string literal has the correct value. -1 if the compiler prohibits long string literals. -2 if at run-time the whole string literal is wrong. >0 if at run-time the string literal is correct up to a position.
LIMITED_ARRAY_LITERAL_LEN: 0 if the C compiler does not limit the length of array literals and at run-time the array literal has the correct value. -1 if the compiler prohibits long array literals. -2 if at run-time the whole array literal is wrong. >0 if at run-time the array literal is correct up to a position.
NO_BIG_LIBRARY: Defined as -1. The meaning is: No library to to implement the bigInteger functions.
BIG_RTL_LIBRARY: Defined as 1. The meaning is: The big_rtl.c library is used to implement the bigInteger functions.
BIG_GMP_LIBRARY: Defined as 2. The meaning is: The big_gmp.c library is used to implement the bigInteger functions.
BIGINT_LIB: Defines the bigInteger library used. The value is one of NO_BIG_LIBRARY, BIG_RTL_LIBRARY and BIG_GMP_LIBRARY.
C_COMPILER_VERSION: Contains a string describing the version of the C compiler which compiled the Seed7 runtime libraries. This string can be used together with GET_CC_VERSION_INFO to compare the C compiler used to compile the Seed7 runtime libraries with the actual version of the C compiler.
CC_NO_OPT_OUTPUT_FILE: Defined if compiling and linking with one command does not work, with the option LINKER_OPT_OUTPUT_FILE.
SEED7_LIB: Contains the name of the Seed7 runtime library.
DRAW_LIB, CONSOLE_LIB, DATABASE_LIB, COMP_DATA_LIB and COMPILER_LIB: Contain names of other Seed7 runtime libraries.
MYSQL_DLL, SQLITE_DLL, POSTGRESQL_DLL, ODBC_DLL, OCI_DLL, FIRE_DLL, DB2_DLL, INFORMIX_DLL, SQL_SERVER_DLL and TDS_DLL: These macros contain lists of DLLs / shared libraries with absolute paths or just DLL / shared library names. If a library is needed at run-time the list of the corresponding macro is processed until loading a library succeeds. If all attempts to load a library fail you get the error: Searching dynamic libraries failed.
LIBINTL_DLL, LIBEAY32_DLL, LIBCRYPTO_DLL and LIBSSL_DLL: The PostgreSQL library (defined with POSTGRESQL_DLL) might depend on other libraries. If this is the case these macros are defined. The libraries defined by these macros are loaded before loading POSTGRESQL_DLL.
CC_ENVIRONMENT_INI: Path to an INI file with the environment for the C compiler. If it is "" the C compiler does not need environment settings. The INI file contains a snapshot of the environment variables at the time Seed7 is compiled. As snapshot it contains also environment variables that are not necessary for the C compiler.
S7_LIB_DIR: Directory containing the Seed7 runtime libraries. This path uses the standard path representation (a slash is used as path separator and instead of a drive letter like "C:" the path "/c" is used).
SEED7_LIBRARY: Directory containing the Seed7 include files. This path uses the standard path representation (a slash is used as path separator and instead of a drive letter like "C:" the path "/c" is used).
INT32TYPE: Name of a signed integer type that is 32 bits wide.
UINT32TYPE: Name of an unsigned integer type that is 32 bits wide.
INT64TYPE: Name of a signed integer type that is 64 bits wide.
UINT64TYPE: Name of an unsigned integer type that is 64 bits wide.
INT32TYPE_LITERAL_SUFFIX: The suffix used by the literals of the 32 bits wide integer type.
INT64TYPE_LITERAL_SUFFIX: The suffix used by the literals of the 64 bits wide integer type.

MACROS DEFINED IN CONFIG.H

WITH_STRI_CAPACITY: TRUE if the Seed7 runtime library uses strings with capacity. The capacity of a string can be larger than its size. Strings with capacity can be enlarged without calling realloc().
ALLOW_STRITYPE_SLICES: TRUE if the actual characters of a string can be stored elsewhere. This allows string slices without the need to copy characters.
ALLOW_BSTRITYPE_SLICES: TRUE if the actual characters of a bstring can be stored elsewhere. This allows bstring slices without the need to copy characters.
WITH_SQL: 1 if SQL should be supported, 0 otherwise.
MAP_ABSOLUTE_PATH_TO_DRIVE_LETTERS: Defined in config.h, if absolute paths (paths starting with '/') must be mapped to operating system paths with drive letter. E.g.: "/c" is mapped to the drive letter "C:".
FORBID_DRIVE_LETTERS: Defined in config.h, if a Seed7 path with drive letters must raise RANGE_ERROR.
EMULATE_ROOT_CWD: Defined in config.h, if the operating system uses drive letters and reading the directory "/" must return a list of available drives.