Commit Graph

280 Commits

Author SHA1 Message Date
Jonas Maebe
28c20cfc5e * the default string type for the JVM target is no longer automatically
unicodestring = java.lang.String. The reason this was the default in
    the past is that this was the first string type that was implemented,
    and without it being the default most code involving string operations
    would fail. Now the default strings types are the same as for other
    targets
  + new {$modeswitch unicodestrings} directive, that when activated
    *together* with {$h+},
   1) changes char into an alias for widechar
   2) changes string into an alias for unicodestring
   3) changes the preferred string evaluation type (in case of uncertainty)
      to unicodestring
    {$modeswitch unicodestrings} with {$h-} does not change anything at all
    regarding the string type (it still changes the char type)
  + new uuchar unit that redefines char as widechar, and which is automatically
    included by the compiler if {$modeswitch unicodestrings} is enabled

git-svn-id: branches/jvmbackend@18781 -
2011-08-20 08:35:47 +00:00
Jonas Maebe
37aa2d8443 + full support for sets on the JVM target
o sets of enums are handled as JUEnumSet instances, others as JUBitSet
     derivatives (both smallsets and varsets, to make interoperability with
     Java easier)
   o special handling of set constants: these have to be constructed at run
     time. In case of constants in the code, create an internal constsym to
     represent them. These and regular constsyms are then aliased by an
     another internal staticvarsym that is used to initialise them in the
     unit initialisation code.
   o until they are constructed at run time, set constants are encoded as
     constant Java strings (with the characters containing the set bits)
   o hlcgobj conversion of tcginnode.pass_generate_code() for the genjumps
     part (that's the only part of the generic code that's used by the JVM
     target)
   o as far as explicit typecasting support is concerned, currently the
     following ones are supported (both from/to setdefs): ordinal types,
     enums, any other set types (whose size is the same on native targets)
   o enum setdefs also emit signatures
   o overloading routines for different ordinal set types, or for different
     enum set types, is not supported on the JVM target

git-svn-id: branches/jvmbackend@18662 -
2011-08-20 08:22:22 +00:00
Jonas Maebe
466f6751c8 + getpointerdef() function that returns a pointerdef for the passed def.
Multiple calls to getpointerdef() for the same def in a single module
    return the same pointerdef

git-svn-id: branches/jvmbackend@18644 -
2011-08-20 08:20:51 +00:00
Jonas Maebe
b60e50b2b0 - reverted 679f3968901f0d565644d59e046eb25d5402f027, because it requires
support for concat_multi for ansistrings to prevent compilation
    problems there

git-svn-id: branches/jvmbackend@18545 -
2011-08-20 08:09:10 +00:00
Jonas Maebe
e403ddca0c * enabled generic multi-string concatenation optimization now that var
parameters for strings work (although they're not as efficient as
    function results for the JVM target)

git-svn-id: branches/jvmbackend@18535 -
2011-08-20 08:08:05 +00:00
Jonas Maebe
3e5b70d80d * cpu64bitaddr -> cpu64bitalu check to determine whether we need special
handling for 64 add-nodes
  * call the generic ordinal handler for 64 add nodes on the JVM

git-svn-id: branches/jvmbackend@18518 -
2011-08-20 08:06:35 +00:00
Jonas Maebe
d1a1d30e04 * unless compiled with -dnounsupported, the compiler will now accept not
only JVM constructs that are already implemented, but also ones that
    will be supported in the future but that aren't implemented yet (to
    make it easier to already adapt code to the future changes)

git-svn-id: branches/jvmbackend@18498 -
2011-08-20 08:04:51 +00:00
Jonas Maebe
990a4ccb64 - disabled genmultistringadd() for the JVM target for now
git-svn-id: branches/jvmbackend@18479 -
2011-08-20 08:03:19 +00:00
Jonas Maebe
7200245337 + cs_unicodestrings global switch to indicate that the default string type
is unicodestring (does *not* change the type of "char" to "unicodechar"
    (yet)). Not yet (un)selectable via a directive.
  + systems_default_unicodestring set containing systems on which the
    default string type is unicodestring

git-svn-id: branches/jvmbackend@18415 -
2011-08-20 07:57:45 +00:00
Jonas Maebe
8c7f1e1013 * only transform multiplication nodes to sqr, not any kind of addnode
(fix for 18252)

git-svn-id: trunk@18263 -
2011-08-18 12:47:41 +00:00
Jonas Maebe
b5fc32750b * check whether both left and right are reals rather than twice whether left
is real before transforming a multiply into an sqr operation (fixes make
    cycle after r18252, don't know where or even how it caused a wrong
    optimization)

git-svn-id: trunk@18257 -
2011-08-18 09:12:13 +00:00
florian
e796a878ca * convert <real x>*<real x> into sqr(<real x>), it might reduces register
pressure and/or memory accesses without a drawback

git-svn-id: trunk@18252 -
2011-08-17 16:00:16 +00:00
pierre
9ec1213be1 * Reorganize code to avoid warnings about unreachable code
git-svn-id: trunk@17868 -
2011-06-29 09:59:06 +00:00
florian
d35d1ed357 + initial support for pascal booleans with sizes 2, 4 and 8
git-svn-id: branches/pasboolxx@17836 -
2011-06-26 15:02:37 +00:00
florian
6e1d5e0aa0 * handle pointer+type param. correctly, resolves #19548
* convert exit(x) into result:=x; exit;, enables further optimizations in the future

git-svn-id: trunk@17829 -
2011-06-25 18:05:49 +00:00
paul
d21bbc4548 compiler: reduce amount of hints and warnings
git-svn-id: trunk@17350 -
2011-04-20 02:58:52 +00:00
florian
cb003b631b * use helper for dword*dword if needed
git-svn-id: branches/avr@17026 -
2011-02-26 20:59:41 +00:00
florian
d5455cf6ac + software multiplication support for OS_16,OS_S16,OS_32,OS_S32
git-svn-id: branches/avr@17020 -
2011-02-26 20:33:01 +00:00
paul
906ca89f4e compiler:
- allow pointermath for pchar, pwidechar, array[] of char and array[] of widechar without pointermath directive, 
  - don't check tpointerdef().has_pointer_math if this is not a tpointerdef

This fixes testsuite regression (tbs/tb0439.pp)

git-svn-id: trunk@16679 -
2011-01-02 04:35:05 +00:00
Jonas Maebe
2222f2c44b * renamed is_class_or_interface_or_dispinterface_or_objc() into
is_implicit_pointer_object_type() to better indicate the purpose of
    that routine, and to avoid having to change its name every time
    a new object type with this property is added

git-svn-id: trunk@16664 -
2010-12-30 15:19:54 +00:00
paul
86d3e41442 compiler: replace MODESWTICH POINTERARITHMETICS with $POINTERMATH directive, disable pointer arithmetic in delphi mode by default (compatible with delphi), enable pointer arithmetic in fpc/objfpc modes as they had it enabled by default before, add has_pointer_math field to tpointerdef to allow pointer arithmetic with such pointer even if pointer math mode is off (delphi compatible) + tests
git-svn-id: trunk@16651 -
2010-12-29 03:24:28 +00:00
Jonas Maebe
769eb3f604 * moved the initialisation of the nodetree and tai class reference constants
from the unit initialisation sections to the variable declaration sections
    to prevent the base units from overriding derived classes (based on patch
    by Hans-Peter Dietrich, mantis #17516)

git-svn-id: trunk@16118 -
2010-10-09 18:52:22 +00:00
Jonas Maebe
94d976bc87 * when simplifying ordinal expressions during inlining, keep the resultdef
that was set during the typecheck pass because typeconversion nodes
    may have been optimised away previously and sometimes the resultdef is
    important (e.g. for the value of callparanodes) (mantis #17458)

git-svn-id: trunk@16101 -
2010-10-07 15:08:52 +00:00
florian
8b1889d033 * now compilation on i386 was broken, fixed
git-svn-id: trunk@15757 -
2010-08-09 20:00:19 +00:00
florian
33c67c7972 * compilation fixed on x86-64
git-svn-id: trunk@15756 -
2010-08-09 19:51:28 +00:00
florian
66571ee4c8 * use extended as default real type in calculations only if fputype=x87
git-svn-id: trunk@15750 -
2010-08-08 16:17:43 +00:00
Jonas Maebe
025ec34e4d + "CExtended" type that is the same as "extended", but conforming to the
properties/behaviour of the equivalent of Extended in C (i.e., to
    "long double" on i386 and x86_64 platforms that support a 10 byte
    long double, and to "double" elsewhere)

git-svn-id: trunk@14912 -
2010-02-14 13:45:58 +00:00
Jonas Maebe
38b1271a21 * fixed wrong range check errors for "qword and constant" after
r14892

git-svn-id: trunk@14904 -
2010-02-13 20:44:03 +00:00
Jonas Maebe
5e1a09f817 * fixed getting the "original" def in taddnode.simplify by stopping at
type conversions for absolute variables (fixes test/tprec13 on, a.o.,
    ppc64 after r14892)

git-svn-id: trunk@14896 -
2010-02-13 11:30:43 +00:00
Jonas Maebe
85984c2d8f * insert proper type conversions when optimising logical operations and
compares by avoiding unnecessary sign extensions (fixes bug reported in
    http://lists.freepascal.org/lists/fpc-pascal/2010-January/023907.html )
  * never throw away int2int type conversions on bitpacked loads, because
    in these cases the proper bits still need to be selected

git-svn-id: trunk@14892 -
2010-02-12 18:13:08 +00:00
paul
009ef9184d compiler: add dispinterfaces to most of places where something is comparing to usual interfaces since they are not very different things
git-svn-id: trunk@14751 -
2010-01-19 15:24:39 +00:00
florian
a58e07df38 * optimize <expr> and/or <expr> if they are equal and boolean
git-svn-id: trunk@14714 -
2010-01-17 14:39:58 +00:00
paul
74cc1e0bab compiler: allow 'as' for dispinterface, allow comparisons of dispinterface and pointers (issue #0015530, issue #0015529)
git-svn-id: trunk@14663 -
2010-01-16 11:37:25 +00:00
florian
c1f3d8dcaa * unified names of system_*/systems_* sets
git-svn-id: trunk@14566 -
2010-01-07 18:16:20 +00:00
Jonas Maebe
335e159c11 Merged revisions 13351-13373,13376-13457 via svnmerge from
svn+ssh://jonas@svn.freepascal.org/FPC/svn/fpc/trunk

git-svn-id: branches/objc@13458 -
2009-07-26 14:31:50 +00:00
Jonas Maebe
63ce04fd00 * set addn resultdef when evaluating constant string concatenation
instead of letting the resulting stringconstn decide for itself
    (mantis #14174)

git-svn-id: trunk@13419 -
2009-07-22 16:16:11 +00:00
Jonas Maebe
5a2ccfff52 --WARNING: start build process with FPC 2.2.4; won't work when
starting with a previous 2.3.1 or compiler built from the objc branch
  + added basic objcprotocol support (only for external protocols
    currently)
     o use in type declaration: "type xp = objcprotocol ... end;"
     o when defining a root class that implements it:
       "type yc = objcclass(xp) ... end" (note: no support yet
       for something like "objcclass(id,xp)" or so)
     o when defining a non-root class that implements a protocol:
       "type zc = objcclass(nsobject,xp) ... end"
     o includes support for "required" and "optional" sections
     o no support yet for the objcprotocol(<protocol>) expression
       that enables getting a class instance representing the
       protocol (e.g., for use with "conformsToProtocol:")
     o message names have to specified in protocol declarations,
       but if an objcclass implements a protocol, the message names do
       not have to be repeated (but if they are, they have to match;
       the same goes when overriding inherited methods)
  + allow specifying the external name of Objective-C classes and
    protocols, since classes and protocols can have the same name
    (and you cannot use the same Pascal identifier in such caseq)
  + added NSObject protocol, and make the NSObject class use it
  + added missing NSObject class methods that have the same name
    as instance methods (added "class" name prefix to avoid clashes)
  * fixed several cases where the compiler did not treat Objective-C
    classes/protocols the same as Object Pascal classes/interfaces
    (a.o., forward declarations, alignment, regvars, several type
     conversions, ...)
  * allow "override" directive in objcclass declarations, and print
    a hint if it's forgotten in an external declaration (because it
    doesn't really matter there, and may make automated header
    conversion harder than necessary) and an error if will be used in
    a non-external declaration (because it is not possible to start
    a new vmt entry-tree in Objective-C, you can only override parent
    methods)
  * reject objcclasses/protocols as parameters to typeof()
  * don't try to test VMT validity of objcclasses/protocols

git-svn-id: branches/objc@13375 -
2009-07-09 20:48:28 +00:00
Jonas Maebe
7e908cb1e3 * forbid "nil +/- ordinal" expressions (Delphi-compatible, and the
expression is meaningless since nil is not a valid address)

git-svn-id: trunk@13205 -
2009-05-27 17:50:55 +00:00
Jonas Maebe
17a7903e60 * fixed evaluating "nil +/- int_const" expressions (mantis #13815)
git-svn-id: trunk@13194 -
2009-05-25 18:34:04 +00:00
Jonas Maebe
79c70c52a7 * store/restore taddnode.resultrealdef to/from the ppufiles, and also
copy/compare it when copying/comparing taddnodes (necessary for inlined
    procedures, mantis #13596)

git-svn-id: trunk@13054 -
2009-04-27 21:03:11 +00:00
Jonas Maebe
626a723f8c * in {$h+} mode, always treat string constants as if they were ansistrings
when determining the result of string concatenations (mantis #13456)

git-svn-id: trunk@13004 -
2009-04-05 17:26:38 +00:00
Jonas Maebe
a00df1033f * changed "Mixing signed expressions and longwords gives a 64bit result"
from warning into a hint (it's not a potential error, but potential
    optimisation advice)

git-svn-id: trunk@12075 -
2008-11-13 19:49:46 +00:00
peter
af437d5beb * refactor the forward type declaration handling, remove
global typecanbeforward and move 'class of ..' parsing
    to ptype

git-svn-id: trunk@12045 -
2008-11-09 22:07:36 +00:00
florian
2e89b30e86 * fixed wrongly written system.$unicodetype type definition
* improved code generation for unicodestrings

git-svn-id: trunk@11776 -
2008-09-14 15:22:32 +00:00
florian
b178b08ba7 Merged revisions 11665-11738 via svnmerge from
http://svn.freepascal.org/svn/fpc/branches/unicodestring

........
  r11665 | florian | 2008-08-30 13:30:17 +0200 (Sat, 30 Aug 2008) | 1 line
  
  * continued to work on unicodestring type support
........
  r11666 | florian | 2008-08-30 19:02:26 +0200 (Sat, 30 Aug 2008) | 2 lines
  
  * expectloc for wide/ansi/unicode strings is LOC_CONSTANT or LOC_REGISTER now
........
  r11667 | florian | 2008-08-30 20:42:37 +0200 (Sat, 30 Aug 2008) | 1 line
  
  * more unicodestring stuff fixed, test results on win32 are already good
........
  r11670 | florian | 2008-08-30 23:21:48 +0200 (Sat, 30 Aug 2008) | 2 lines
  
  * first fixes for unix bootstrapping
........
  r11683 | ivost | 2008-09-01 12:46:39 +0200 (Mon, 01 Sep 2008) | 2 lines
  
      * fixed 64bit bug in iconvenc.pas
........
  r11689 | florian | 2008-09-01 23:12:34 +0200 (Mon, 01 Sep 2008) | 1 line
  
  * fixed several errors when building on unix
........
  r11694 | florian | 2008-09-03 20:32:43 +0200 (Wed, 03 Sep 2008) | 1 line
  
  * fixed unix compilation
........
  r11695 | florian | 2008-09-03 21:01:04 +0200 (Wed, 03 Sep 2008) | 1 line
  
  * bootstrapping fix
........
  r11696 | florian | 2008-09-03 21:07:18 +0200 (Wed, 03 Sep 2008) | 1 line
  
  * more bootstrapping fixed
........
  r11698 | florian | 2008-09-03 22:47:54 +0200 (Wed, 03 Sep 2008) | 1 line
  
  + two missing compiler procs exported
........
  r11701 | florian | 2008-09-04 16:42:34 +0200 (Thu, 04 Sep 2008) | 2 lines
  
  + lazarus project for the linux rtl
........
  r11702 | florian | 2008-09-04 16:43:27 +0200 (Thu, 04 Sep 2008) | 2 lines
  
  + set unicode string procedures
........
  r11707 | florian | 2008-09-04 23:23:02 +0200 (Thu, 04 Sep 2008) | 2 lines
  
  * fixed several type casting stuff
........
  r11712 | florian | 2008-09-05 22:46:03 +0200 (Fri, 05 Sep 2008) | 1 line
  
  * fixed unicodestring compilation on windows after recent unix changes
........
  r11713 | florian | 2008-09-05 23:35:12 +0200 (Fri, 05 Sep 2008) | 1 line
  
  + UnicodeString support for Variants
........
  r11715 | florian | 2008-09-06 20:59:54 +0200 (Sat, 06 Sep 2008) | 1 line
  
  * patch by Martin Schreiber for UnicodeString streaming
........
  r11716 | florian | 2008-09-06 22:22:55 +0200 (Sat, 06 Sep 2008) | 2 lines
  
  * fixed test
........
  r11717 | florian | 2008-09-07 10:25:51 +0200 (Sun, 07 Sep 2008) | 1 line
  
  * fixed typo when converting tunicodestring to punicodechar
........
  r11718 | florian | 2008-09-07 11:29:52 +0200 (Sun, 07 Sep 2008) | 3 lines
  
  * fixed writing of UnicodeString properties
  * moved some helper routines to unicode headers
........
  r11734 | florian | 2008-09-09 22:38:55 +0200 (Tue, 09 Sep 2008) | 1 line
  
  * fixed bootstrapping
........
  r11735 | florian | 2008-09-10 11:25:28 +0200 (Wed, 10 Sep 2008) | 2 lines
  
  * first fixes for persisten unicodestrings
........
  r11736 | florian | 2008-09-10 14:31:00 +0200 (Wed, 10 Sep 2008) | 3 lines
  
  Initialized merge tracking via "svnmerge" with revisions "1-11663" from 
  http://svn.freepascal.org/svn/fpc/trunk
........
  r11737 | florian | 2008-09-10 21:06:57 +0200 (Wed, 10 Sep 2008) | 3 lines
  
  * fixed unicodestring <-> variant handling
  * fixed unicodestring property reading
........

git-svn-id: trunk@11739 -
2008-09-10 20:14:31 +00:00
yury
491f0fa1d8 * Replaced all user defined warnings by TODO comments to reduce compiler noise.
git-svn-id: trunk@11443 -
2008-07-23 11:00:03 +00:00
yury
fcceb9cfa1 * Removed/ifdefed/commented unused local variables.
git-svn-id: trunk@11430 -
2008-07-20 23:00:31 +00:00
florian
37372fb7ce * fixed copy&paste errors in cpunodefaultint
git-svn-id: trunk@11379 -
2008-07-13 18:15:17 +00:00
florian
0a4314206f * range check error checking for floats refactored
* never throw range/overflow check errors for floats in delphi mode, resolves #7584

git-svn-id: trunk@10940 -
2008-05-11 09:09:21 +00:00
peter
8f239d04b6 * cleanup and simplify the set type handling
git-svn-id: trunk@10432 -
2008-03-02 17:48:27 +00:00