fpc/compiler
Jonas Maebe 096e1c45d6 * don't keep records in registers if they contain a field that spans the
boundary between two awords, as our subscript code does not handle
    them correctly (second part of mantis #29669)

git-svn-id: trunk@33180 -
2016-03-06 13:20:25 +00:00
..
aarch64 * support marking defs created via the getreusable*() class methods as 2015-11-04 20:46:18 +00:00
arm + VOpVMov2VOp optimization 2016-02-28 20:13:16 +00:00
avr * increase ram size of the avrsim controller target 2016-02-07 11:17:55 +00:00
generic Moved tcontrollerdatatype out into cpuinfo. 2015-09-07 20:36:54 +00:00
html
i386 * the SEGFS and SEGGS prefixes are 386+ 2016-01-11 15:51:40 +00:00
i8086 * the SEGFS and SEGGS prefixes are 386+ 2016-01-11 15:51:40 +00:00
jvm * fixed initialising an array of ansichar typed constant using a string 2016-03-05 15:32:18 +00:00
llvm + handle the asd_cpu directive (commented out, but without producing an internal 2016-02-26 20:50:41 +00:00
m68k - Removed remains of tai_labeled_instruction, already commented out probably in 1.0 times. 2016-02-09 02:25:27 +00:00
mips * Handle possible relocation types in assembler reader using a single AS_RELTYPE token, rather than with individual tokens for each case. Since possible relocations are target-dependent, this will allow to support any amount of them without modifying the base tattreader class. 2016-02-23 21:28:46 +00:00
msg + implemented a new {$ASMCPU XXX} directive, allowing to specify a different 2016-02-29 22:25:25 +00:00
powerpc + handle the asd_cpu directive (commented out, but without producing an internal 2016-02-26 20:50:41 +00:00
powerpc64 * support marking defs created via the getreusable*() class methods as 2015-11-04 20:46:18 +00:00
ppcgen + be able to specify endianess in the assembler command line 2016-01-09 14:28:25 +00:00
sparc * Handle possible relocation types in assembler reader using a single AS_RELTYPE token, rather than with individual tokens for each case. Since possible relocations are target-dependent, this will allow to support any amount of them without modifying the base tattreader class. 2016-02-23 21:28:46 +00:00
systems Added x86_64-embedded target. Patch from Benjamin Rosseaux 2016-02-14 10:57:00 +00:00
utils Merged revision(s) 28690 from branches/svenbarth/packages: 2016-01-22 15:47:29 +00:00
x86 * made the i8086 internal assembler cpu instruction set checking follow the 2016-03-02 01:03:52 +00:00
x86_64 Added x86_64-embedded target. Patch from Benjamin Rosseaux 2016-02-14 10:57:00 +00:00
aasmbase.pas * factored out the generation of the Objective-C section names, so we can 2016-01-07 22:06:21 +00:00
aasmcnst.pas + emit_pchar_const() helper for high level typed const builder 2016-01-10 14:01:24 +00:00
aasmdata.pas * put the code for pure assembler routines in a separate asmlist, so 2015-09-12 23:32:21 +00:00
aasmsym.pas
aasmtai.pas + added new asm directive 'asd_cpu' (specifies the CPU's instruction set) in the 2016-02-25 21:47:13 +00:00
aggas.pas + added new asm directive 'asd_cpu' (specifies the CPU's instruction set) in the 2016-02-25 21:47:13 +00:00
aopt.pas * Fixed notes "var is assigned but not used". 2015-09-17 13:03:04 +00:00
aoptbase.pas * compilation fixed 2016-01-24 16:52:49 +00:00
aoptcs.pas
aoptda.pas
aoptobj.pas o refactored some peephole optimizer code: 2016-01-24 15:25:16 +00:00
assemble.pas * handle the asd_cpu asd_cpu directive and keep track of the currently selected 2016-02-27 00:50:01 +00:00
blockutl.pas Extend tstaticvarsym (and by extension tabstractnormalvarsym) with the capability to create it as unregistered if needed. 2015-11-20 11:15:30 +00:00
browcol.pas
catch.pas
ccharset.pas
cclasses.pas * since T(Tag)HashSet always forces the size to a power of 2, use "and" 2015-10-14 16:44:02 +00:00
cfidwarf.pas
cfileutl.pas * removed hack to disable dircache for DragonFly BSD now that its dirent 2015-12-24 10:54:53 +00:00
cg64f32.pas * add "doregister" parameter to torddef.create, and don't register temporary 2015-10-13 15:59:06 +00:00
cgbase.pas + initialize DS with the current unit's data segment in the function entry code 2015-09-03 21:44:16 +00:00
cghlcpu.pas * synchronized with privatetrunk till r30095 2015-03-05 20:32:15 +00:00
cgobj.pas * correctly handle negative paraloc shift values for non power-of-2 sized 2015-12-10 22:45:56 +00:00
cgutils.pas * define avr => cpu8bitalu 2015-09-13 21:07:23 +00:00
cmsgs.pas
comphook.pas + don't abuse status.currentsourcepath to pass on PPU names if the source is unavailable, causes side effects with other frontends (like the IDE). Improved solution for Mantis #27588 2015-03-29 19:05:19 +00:00
compiler.pas Use general_e_exception_raised if Compiler generates an uncaught exception 2015-09-23 16:49:24 +00:00
compinnr.inc Reverted revision 33036. This feature is too controverse to be left in. 2016-02-04 19:26:51 +00:00
comprsrc.pas + ELF/ppc64le resource support 2015-03-14 18:36:32 +00:00
constexp.pas Fix cycle for x86_64-win64 target with -Cr option 2016-01-28 15:07:40 +00:00
COPYING.txt
cpid.pas
crefs.pas
cresstr.pas * emit all resource strings and internal dynamic string data using the 2015-12-25 21:05:26 +00:00
cstreams.pas Merged revision(s) 32302-32305, 32310 from branches/svenbarth/packages: 2016-02-19 17:13:58 +00:00
cutils.pas * Removed lot of unused vars. 2015-09-17 12:48:58 +00:00
cwindirs.pp
dbgbase.pas
dbgdwarf.pas * add support in the debug info for the dummy s128bit/u128bit types added 2015-12-27 11:45:24 +00:00
dbgstabs.pas * add support in the debug info for the dummy s128bit/u128bit types added 2015-12-27 11:45:24 +00:00
dbgstabx.pas
defcmp.pas * better estimation of compatibility of array constructors with open arrays, resolves issue #29244 2016-01-04 23:10:45 +00:00
defutil.pas - removed symcpu dependencies 2015-06-13 22:47:59 +00:00
dirparse.pas
elfbase.pas
entfile.pas Merged revision(s) 32302-32305, 32310 from branches/svenbarth/packages: 2016-02-19 17:13:58 +00:00
export.pas Merged revision(s) 28792 from branches/svenbarth/packages: 2016-01-22 13:57:41 +00:00
expunix.pas Merged revision(s) 29037 from branches/svenbarth/packages: 2016-01-22 17:02:40 +00:00
finput.pas + implemented exporting of functions for the win16 target 2015-09-27 00:48:35 +00:00
fmodule.pas * free a bunch of extra data after a unit is compiled, as it is no longer 2015-10-30 15:12:37 +00:00
fpccrc.pas
fpcdefs.inc Allow use of stabs for 64-bit systems with -dUSE_STABS_64 2015-10-29 09:23:30 +00:00
fppu.pas * compilation fixed 2016-02-28 18:44:41 +00:00
gendef.pas
globals.pas + implemented a new {$ASMCPU XXX} directive, allowing to specify a different 2016-02-29 22:25:25 +00:00
globstat.pas
globtype.pas Add pi_has_open_array_parameter to proc_info.flags as this requires special handling for i8086 huge memory model to restore DS register correctly 2016-01-11 15:02:10 +00:00
hlcg2ll.pas removed a redundant zero extension from location_force_reg, which occured if the source location was a smaller than 32bit subreg, and the destination was 64bit. hregister is always overwritten in the codepath below anyway. 2015-12-13 02:08:24 +00:00
hlcgobj.pas Add pi_has_open_array_parameter to proc_info.flags as this requires special handling for i8086 huge memory model to restore DS register correctly 2016-01-11 15:02:10 +00:00
htypechk.pas Fix for Mantis #29546. 2016-01-31 17:14:44 +00:00
impdef.pas
import.pas
ldscript.pas
link.pas * removed DLLsource global variable and replaced its uses with 2015-12-17 15:23:05 +00:00
macho.pas
machoutils.pas
Makefile m68k: disabled the force-enabled stack checking for m68k-amiga when compiling the compiler, which was enabled since forever. the current Amiga RTL startup code supports setting stack size programmatically, therefore very low default stack on Amiga is no longer a problem 2016-01-05 02:14:40 +00:00
Makefile.fpc m68k: disabled the force-enabled stack checking for m68k-amiga when compiling the compiler, which was enabled since forever. the current Amiga RTL startup code supports setting stack size programmatically, therefore very low default stack on Amiga is no longer a problem 2016-01-05 02:14:40 +00:00
MPWMake
msgidx.inc + implemented a new {$ASMCPU XXX} directive, allowing to specify a different 2016-02-29 22:25:25 +00:00
msgtxt.inc + implemented a new {$ASMCPU XXX} directive, allowing to specify a different 2016-02-29 22:25:25 +00:00
nadd.pas * make the resultdef of left and right equal when comparing a dynarray 2016-01-10 14:01:39 +00:00
nbas.pas + new parser-only node class to handle Delphi-mode inline specializations 2015-09-11 12:58:25 +00:00
ncal.pas * hidden high parameter must be of type ptrsinttype, resolves issue #29204 2015-12-26 20:48:50 +00:00
ncgadd.pas * converted register_maybe_adjust_setbase() to the high level code generator 2015-12-05 18:03:37 +00:00
ncgbas.pas + implemented a new {$ASMCPU XXX} directive, allowing to specify a different 2016-02-29 22:25:25 +00:00
ncgcal.pas Fix for Mantis #29372. 2016-01-23 22:19:00 +00:00
ncgcnv.pas * fixed the parameter order for thlcgobj.g_load_const_field_by_name() 2015-12-05 18:03:48 +00:00
ncgcon.pas * fixed shortstring constant emission after r28210: the terminating #0 2015-12-26 20:01:06 +00:00
ncgflw.pas * fixed type of stored exception class instance in on-node 2015-11-21 12:37:33 +00:00
ncghlmat.pas * replaced current_procinfo.currtrue/falselabel with storing the true/false 2015-08-27 18:28:57 +00:00
ncginl.pas * implement sizeof/typeof completely at the node level, based on 2015-12-27 14:22:06 +00:00
ncgld.pas - disable x86-specific workarounds when using LLVM (even when targeting x86) 2015-12-28 15:06:50 +00:00
ncgmat.pas + support mmregisters in the generic unaryminus code for floats 2015-09-12 23:33:50 +00:00
ncgmem.pas * fixed check regarding whether a field is in the first or second word of 2016-03-06 13:20:16 +00:00
ncgnstld.pas * ensure that we don't firstpass a tcgnestloadnode twice, as it 2015-12-28 15:06:36 +00:00
ncgnstmm.pas
ncgobjc.pas
ncgopt.pas
ncgrtti.pas Only write the IIDStr for Corba interfaces (it's completely redundant for COM ones) 2016-02-19 15:22:46 +00:00
ncgset.pas * converted register_maybe_adjust_setbase() to the high level code generator 2015-12-05 18:03:37 +00:00
ncgutil.pas (Note to self: always include the root of the working copy in the commit when dealing with merges -.-) 2016-01-22 16:58:33 +00:00
ncgvmt.pas (Note to self: always include the root of the working copy in the commit when dealing with merges -.-) 2016-01-22 16:58:33 +00:00
ncnv.pas * handle tc_equal typeconversions from conststring to a stringdef in 2016-01-10 14:01:58 +00:00
ncon.pas * support unicodestring constants in "case string of" (mantis #29353) 2016-01-20 21:11:26 +00:00
nflw.pas Check for nil before doing loopunrolling. 2016-02-04 15:53:36 +00:00
ngenutil.pas * if a function result has been migrated to the parentfpstruct because it is 2015-12-20 20:56:08 +00:00
ngtcon.pas * removed special-purpose code for constant chars while parsing typed string 2015-12-11 15:33:28 +00:00
ninl.pas Reverted revision 33036. This feature is too controverse to be left in. 2016-02-04 19:26:51 +00:00
nld.pas * always set the resultdef for variant array constructors correctly, so we 2015-11-04 20:46:10 +00:00
nmat.pas Avoid overflow error for single MinValue in rtl syshelph.inc for x86_64 CPU 2016-01-28 14:20:51 +00:00
nmem.pas * apparently, Darwin/x86-64 also uses an opaque ISA pointer on OS X 10.11 2016-02-22 08:44:00 +00:00
nobj.pas * give fields added for VMT definition of TP-style objects the same 2015-12-27 14:21:59 +00:00
nobjc.pas
node.pas + write also node complexity when dumping a node tree 2016-02-28 18:44:43 +00:00
nopt.pas
nset.pas * fixed "case <string> of" with multiple labels for a single statement after 2015-07-03 20:04:25 +00:00
nstate.pas
nutils.pas * support static fields in nested types in records, by always including 2015-11-24 16:04:19 +00:00
objcasm.pas * factored out the generation of the Objective-C section names, so we can 2016-01-07 22:06:21 +00:00
objcdef.pas
objcgutl.pas * converted objcgutl (Objective-C metadata generation) to the high level 2016-01-10 17:22:20 +00:00
objcutil.pas Merged revision(s) 28691, 31914 from branches/svenbarth/packages: 2016-01-15 16:06:06 +00:00
ogbase.pas * in the internal assembler writer, when starting a new object file, set the 2016-03-02 01:12:56 +00:00
ogcoff.pas Move some of the local reloc calculation out of aasmcpu, and into COFF and ELF writers. 2016-01-06 23:15:08 +00:00
ogelf.pas Move some of the local reloc calculation out of aasmcpu, and into COFF and ELF writers. 2016-01-06 23:15:08 +00:00
oglx.pas * refactored the internal linker ar object reader object creation to allow using 2015-08-14 13:52:45 +00:00
ogmacho.pas * store a pointer to the used tasminfo record in every assembler writer, so 2015-09-12 23:32:13 +00:00
ogmap.pas * the code for displaying a symbol address in the linker map file moved to a 2015-08-24 22:12:00 +00:00
ognlm.pas * store a pointer to the used tasminfo record in every assembler writer, so 2015-09-12 23:32:13 +00:00
ogomf.pas + introduced a new type TRelocDataInt for use as the 'data' parameter for 2016-01-13 17:57:36 +00:00
omfbase.pas * support writing omf libraries with size greater than 251 blocks 2015-11-07 16:54:23 +00:00
optbase.pas
optconstprop.pas
optcse.pas * improve cse on record/array accesses 2016-02-28 18:44:45 +00:00
optdead.pas * factored out the check for whether we have to generate dotted function 2015-03-14 18:35:25 +00:00
optdeadstore.pas
optdfa.pas * Fixed notes "var is assigned but not used". 2015-09-17 13:03:04 +00:00
options.pas * allow very small heaps for avr 2016-02-07 18:28:13 +00:00
optloop.pas
opttail.pas
optutils.pas
optvirt.pas
owar.pas * refactored the internal linker ar object reader object creation to allow using 2015-08-14 13:52:45 +00:00
owbase.pas * support writing more than 1024 zeros in tobjectwriter.writezeros 2015-08-24 01:09:53 +00:00
owomflib.pas * support writing omf libraries with size greater than 251 blocks 2015-11-07 16:54:23 +00:00
parabase.pas + support for constant call parameters on LLVM 2015-12-28 15:06:07 +00:00
paramgr.pas * only copy the shiftval field from the original para location when creating 2015-12-05 18:03:59 +00:00
parser.pas Use general_e_exception_raised if Compiler generates an uncaught exception 2015-09-23 16:49:24 +00:00
pass_1.pas
pass_2.pas * fixed compilation of the compiler with -dEXTDEBUG after r31588 2015-09-12 11:55:54 +00:00
pbase.pas
pdecl.pas * initialize variables to avoid warnings 2015-11-21 14:14:30 +00:00
pdecobj.pas Implement support for parsing "generic [class] procedure" and "generic [class] function" in non-Delphi modes. Since "generic" is a prefix it is quite ugly to implement, but from a Pascal language point of view it fits better than "procedure generic xyz". 2015-11-20 16:50:58 +00:00
pdecsub.pas Fix a few typos discovered by Blaise 2015-12-11 16:31:00 +00:00
pdecvar.pas * initialize variables to avoid warnings 2015-11-21 14:14:30 +00:00
pexports.pas Merged revision(s) 28691, 31914 from branches/svenbarth/packages: 2016-01-15 16:06:06 +00:00
pexpr.pas Reverted accidentally committed inline if-then-else expression (committed in r33048 when I reverted the IfThen()... -.- ) 2016-02-19 21:51:56 +00:00
pgentype.pas Fully enable generic functions, procedures and methods by allowing inline specializations in procedure bodies as well. 2015-11-21 16:52:09 +00:00
pgenutil.pas Convert the Boolean parameters accept_equal, typeonly and hadspecialize of comp_expr(), sub_expr() and factor() to a set parameter and adjust all calls. 2015-11-07 21:45:03 +00:00
pinline.pas o basic extended pascal support: 2015-11-15 15:13:36 +00:00
pmodules.pas (Note to self: always include the root of the working copy in the commit when dealing with merges -.-) 2016-01-22 16:58:33 +00:00
pp.lpi
pp.pas - removed leftover alpha, ia64 and vis code 2015-08-28 22:31:29 +00:00
pparautl.pas * hidden high parameter must be of type ptrsinttype, resolves issue #29204 2015-12-26 20:48:50 +00:00
ppc68k.lpi
ppc8086.lpi + started implementing a win16 target :) 2015-09-05 02:47:42 +00:00
ppcaarch64.lpi
ppcarm.lpi
ppcavr.lpi
ppcjvm.lpi
ppcmips.lpi
ppcmipsel.lpi
ppcppc64.lpi
ppcppc.lpi
ppcsparc.lpi
ppheap.pas
ppu.pas Merged revision(s) 28690 from branches/svenbarth/packages: 2016-01-22 15:47:29 +00:00
ppx86_64.lpi
procinfo.pas * replaced current_procinfo.currtrue/falselabel with storing the true/false 2015-08-27 18:28:57 +00:00
pstatmnt.pas * In pure-assembler procedures, don't allocate/free any registers around the only assembler node. Such procedures don't use virtual registers or register allocator. 2015-12-07 18:32:35 +00:00
psub.pas (Note to self: always include the root of the working copy in the commit when dealing with merges -.-) 2016-01-22 16:58:33 +00:00
psystem.pas - Removed remains of tai_labeled_instruction, already commented out probably in 1.0 times. 2016-02-09 02:25:27 +00:00
ptconst.pas * synchronized with privatetrunk till r30095 2015-03-05 20:32:15 +00:00
ptype.pas Implement support for parsing "generic [class] procedure" and "generic [class] function" in non-Delphi modes. Since "generic" is a prefix it is quite ugly to implement, but from a Pascal language point of view it fits better than "procedure generic xyz". 2015-11-20 16:50:58 +00:00
raatt.pas * Handle possible relocation types in assembler reader using a single AS_RELTYPE token, rather than with individual tokens for each case. Since possible relocations are target-dependent, this will allow to support any amount of them without modifying the base tattreader class. 2016-02-23 21:28:46 +00:00
rabase.pas
rasm.pas
rautils.pas * set operand size information when accessing fields in assembly 2015-12-02 15:53:22 +00:00
README.txt
regvars.pas
rescmn.pas
rgbase.pas
rgobj.pas * Make trgobj.uses_registers method return True when registers of appropriate type are specified in list of used registers (after assembler blocks), but not used otherwise. This is a missing piece of r30011, enabling code generator to actually save/restore such registers. Resolves #28421. 2015-12-07 19:12:20 +00:00
scandir.pas + implemented a new {$ASMCPU XXX} directive, allowing to specify a different 2016-02-29 22:25:25 +00:00
scanner.pas * Replaced hacks with resetting 'c' to zero and decreasing inputpointer by boolean parameter to skipcomment and skipoldtpcomment. This parameter specifies whether first character of comment should be read. 2016-01-03 17:07:15 +00:00
script.pas
switches.pas * add llvminfo to override optimizer switches in case of {$ifdef llvm} 2016-01-10 14:01:35 +00:00
symbase.pas
symconst.pas * converted objcgutl (Objective-C metadata generation) to the high level 2016-01-10 17:22:20 +00:00
symcreat.pas * fixed copy/paste error in r31285 2016-01-10 14:02:02 +00:00
symdef.pas * don't keep records in registers if they contain a field that spans the 2016-03-06 13:20:25 +00:00
symsym.pas * don't recalculate the regable info when dereferencing a symbol's vardef: 2016-03-06 13:20:22 +00:00
symtable.pas Merged revision(s) 28689 from branches/svenbarth/packages: 2016-01-22 15:26:30 +00:00
symtype.pas * converted the result of fullownerhierarchyname to tsymstr so it (and 2015-11-21 12:37:04 +00:00
symutil.pas
systems.inc Add -Aas-sol to use native Solaris assembler 2015-10-29 15:10:27 +00:00
systems.pas * Fix default OS target for i386 and sparc processor on solaris OS 2016-01-27 15:32:55 +00:00
tgobj.pas * suppress (harmless) range error 2015-05-31 16:50:50 +00:00
tokens.pas o basic extended pascal support: 2015-11-15 15:13:36 +00:00
verbose.pas + don't abuse status.currentsourcepath to pass on PPU names if the source is unavailable, causes side effects with other frontends (like the IDE). Improved solution for Mantis #27588 2015-03-29 19:05:19 +00:00
version.pas - removed leftover alpha, ia64 and vis code 2015-08-28 22:31:29 +00:00
widestr.pas * fixed accidentally committed (harmless) typo in r32202 2015-11-04 13:03:45 +00:00
wpo.pas
wpobase.pas
wpoinfo.pas Merged revision(s) 28689 from branches/svenbarth/packages: 2016-01-22 15:26:30 +00:00

This directory contains the sources of the Free Pascal Compiler

If you want to compile/modify the compiler, please read first the
programmers manual.

To recompile the compiler, you can use the batch files :
 + mppc386.bat    if you want to build a cross compiler from i386 to m68k
 + mppcsparc      if you want to build a cross compiler from i386 to SPARC
 
 or
Use the make utility as following
  
      make OS_TARGET="compiler OS target" \
      CPU_TARGET="compiler CPU target" \
      FPCCPUOPT="Optimization level" \
      PP="compiler used to compile FPC" \
      COMPILER_OPTIONS="Options passed to compiler" \
      
      
If an option is omitted, then target CPU/OS will be same as current CPU/OS
 
Possibles targets are : linux go32v2 win32 os2 freebsd beos netbsd amiga haiku
atari sunos qnx netware openbsd wdosx palmos macos macosx emx
   
Possible compiler switches (* marks a currently required switch):
  -----------------------------------------------------------------
  GDB*                support of the GNU Debugger
  I386                generate a compiler for the Intel i386+
  x86_64              generate a compiler for the AMD x86-64 architecture
  M68K                generate a compiler for the M68000
  SPARC               generate a compiler for SPARC
  POWERPC             generate a compiler for the PowerPC
  VIS                 generate a compile for the VIS
  DEBUG               version with debug code is generated
  EXTDEBUG            some extra debug code is executed
  SUPPORT_MMX         only i386: releases the compiler switch
                      MMX which allows the compiler to generate
                      MMX instructions
  EXTERN_MSG          Don't compile the msgfiles in the compiler, always
                      use external messagefiles, default for TP
  NOAG386INT          no Intel Assembler output
  NOAG386NSM          no NASM output
  NOAG386BIN          leaves out the binary writer, default for TP
  NORA386DIR          No direct i386 assembler reader
  TEST_GENERIC        Test Generic version of code generator
                      (uses generic RTL calls)
  -----------------------------------------------------------------
  cpuflags            The target processor has status flags (on by default)
  cpufpemu            The target compiler will also support emitting software
                       floating point operations
  cpu64bitaddr        The targets use a 64-bit address space (pointers and
                       the default integer type are 64 bit)
  cpu64bitalu         The target cpu has 64-bit registers available (unless
                       cpu64bitaddr is also defined, pointers and default
                       integer type remain 32 bit, but the cpu can perform
                       64 bit calculations directly without needing helpers)
  -----------------------------------------------------------------

  Required switches for a i386 compiler be compiled by Free Pascal Compiler:
  GDB;I386

to build a compiler to SPARC target using a Win32/i386 you just use :
      make CPU_TARGET=SPARC