fpc/rtl
marco f54d92326e * remove leftover matrix unit dep
git-svn-id: trunk@27208 -
2014-03-20 18:40:44 +00:00
..
aix
amiga
android
arm - MIPS,SPARC and ARM-wince: removed remaining references to softfloat stuff. 2014-03-20 17:46:01 +00:00
atari - Don't include softfpu code on SPARC, unless bootstrapping with 2.6. 2014-03-20 17:20:58 +00:00
avr
beos * Removed SysInitFPU/SysResetFPU from (remaining) system units initialization code. It is redundant there because it is called from fpc_cpuinit, or, in case of i386, from fpc_cpucodeinit. 2014-03-17 10:53:52 +00:00
bsd * Removed SysInitFPU/SysResetFPU from (remaining) system units initialization code. It is redundant there because it is called from fpc_cpuinit, or, in case of i386, from fpc_cpucodeinit. 2014-03-17 10:53:52 +00:00
charmaps
darwin * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
embedded * partially merged the mips-embedded branch of Michael Ring: 2014-03-19 21:25:38 +00:00
emx * Clarified comments regarding TIB layout and stack bounds on OS/2. 2014-03-14 15:18:23 +00:00
freebsd * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
gba * Since StackTop=StackBottom+StackLength, any two variables are sufficient to describe stack, the third one is redundant. 2014-03-12 19:48:03 +00:00
go32v2 * Since StackTop=StackBottom+StackLength, any two variables are sufficient to describe stack, the third one is redundant. 2014-03-12 19:48:03 +00:00
haiku * Removed SysInitFPU/SysResetFPU from (remaining) system units initialization code. It is redundant there because it is called from fpc_cpuinit, or, in case of i386, from fpc_cpucodeinit. 2014-03-17 10:53:52 +00:00
i386 + functions for FMA and AVX2 detection 2014-03-20 15:47:51 +00:00
i8086 * i8086: added assembler implementation of fpc_round_real and removed remaining softfloat dependencies. 2014-03-20 02:28:47 +00:00
inc * Get rid of last softfloat dependencies in genmath.inc. Replaced remaining calls to float_raise with expressions which generate appropriate floating-point exceptions at runtime. This is probably somewhat slower on actual softfloat targets, but the change only affects exceptional control flow, so not a significant issue. 2014-03-20 11:04:00 +00:00
java
jvm
linux * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
m68k - Removed mathuh.inc files which are identical for all targets (except m68k, see below), their contents moved into math.pp. 2014-03-18 20:01:15 +00:00
macos
mips - MIPS,SPARC and ARM-wince: removed remaining references to softfloat stuff. 2014-03-20 17:46:01 +00:00
mipsel - Removed mathuh.inc files which are identical for all targets (except m68k, see below), their contents moved into math.pp. 2014-03-18 20:01:15 +00:00
morphos * Removed SysInitFPU/SysResetFPU from (remaining) system units initialization code. It is redundant there because it is called from fpc_cpuinit, or, in case of i386, from fpc_cpucodeinit. 2014-03-17 10:53:52 +00:00
msdos * fixed do_filepos and do_seekend not returning the high 16 bits of the file 2014-03-17 22:01:09 +00:00
nativent
nds * Since StackTop=StackBottom+StackLength, any two variables are sufficient to describe stack, the third one is redundant. 2014-03-12 19:48:03 +00:00
netbsd * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
netware * remove leftover matrix unit dep 2014-03-20 18:40:44 +00:00
netwlibc * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
objpas * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
openbsd * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
os2 * Removed SysInitFPU/SysResetFPU from (remaining) system units initialization code. It is redundant there because it is called from fpc_cpuinit, or, in case of i386, from fpc_cpucodeinit. 2014-03-17 10:53:52 +00:00
palmos
powerpc - Removed mathuh.inc files which are identical for all targets (except m68k, see below), their contents moved into math.pp. 2014-03-18 20:01:15 +00:00
powerpc64 - Removed mathuh.inc files which are identical for all targets (except m68k, see below), their contents moved into math.pp. 2014-03-18 20:01:15 +00:00
qnx
solaris * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
sparc - MIPS,SPARC and ARM-wince: removed remaining references to softfloat stuff. 2014-03-20 17:46:01 +00:00
symbian
ucmaps
unix
watcom
wii * Since StackTop=StackBottom+StackLength, any two variables are sufficient to describe stack, the third one is redundant. 2014-03-12 19:48:03 +00:00
win + Win32/Win64 SEH: STATUS_FLOAT_MULTIPLE_(FAULTS|TRAPS) need additional handling. Makes floating-point exceptions consistent between x87 and SSE. 2014-03-17 22:49:29 +00:00
win32 * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
win64 * rtl-unicode package, containing unicodeducet, collation units and *bidi* units 2014-03-20 18:28:38 +00:00
wince * Since StackTop=StackBottom+StackLength, any two variables are sufficient to describe stack, the third one is redundant. 2014-03-12 19:48:03 +00:00
x86_64 + functions for FMA and AVX2 detection 2014-03-20 15:47:51 +00:00
COPYING.FPC
COPYING.txt
fpmake.inc - Removed mathuh.inc files which are identical for all targets (except m68k, see below), their contents moved into math.pp. 2014-03-18 20:01:15 +00:00
fpmake.pp - Removed mathuh.inc files which are identical for all targets (except m68k, see below), their contents moved into math.pp. 2014-03-18 20:01:15 +00:00
Makefile
Makefile.fpc
README.txt

This is the Run-Time Library (RTL) tree for Free Pascal.

To recompile the RTL, edit the main makefile. The makefiles NEED a GNU make
compatible make, they need unix-like 'rm' and 'mv' commands, as well as some
others. You can find these in the gnuutils package on the ftp site.

The main makefile is located ABOVE the RTL tree. It uses the FPC
makefile.fpc to guess reasonable defaults for everything it needs.
(these files can be found in base.zip on the FTP site)

The only variable that you may want to set are
FPC             - What compiler to use. Use an absolute path. 
                  (default is ppc386)
INSTALL_UNITDIR - Where to install the RTL units 
OPT             - any special options you want to set for the compiler.

In principle, you can also descend into the subdirectory of your OS, and
type 'make' there, that should also compile everything.

The tree contains subdirectories for all the supported operating systems,
as well as all processor architectures. The processor directories contain
low-level routines which are required for the system unit (if they are not 
available in high-level language form), as well as optimized versions of
the pascal generic routines (the generic routine source code is localed in
the inc subdirectory).

The following directories are not supported officially and may not work
correctly with FreePascal v1.0:

go32v1 - The DJGPP go32v1 DOS extender (no longer supported)
palmos - The PalmOS directory for the Dragonball (incomplete)


Enjoy.