Commit Graph

890 Commits

Author SHA1 Message Date
Jonas Maebe
6f5905684f + support for specifying the name of fields added to anonymous record
types that are constructed on the fly by the high level typed const
    builder

git-svn-id: trunk@31249 -
2015-07-30 16:57:58 +00:00
Jonas Maebe
e34a16d1f1 * change rtti_mangledname resultdef into TSymStr
git-svn-id: trunk@31245 -
2015-07-30 16:57:45 +00:00
Jonas Maebe
b21610050f * don't store reusable defs based on other defs from localsymtables into the
static/globalsymtable. The static/global symtable is stored to the PPU
    while the localsymtables are not, which means we get dangling references
    when reloading such defs from the ppu afterwards

git-svn-id: trunk@31149 -
2015-06-23 21:23:25 +00:00
Jonas Maebe
508828958c * changed getprocaddressprocvar() into a tprocvardef.getreusableprocaddr()
class method

git-svn-id: trunk@31148 -
2015-06-23 21:23:22 +00:00
Jonas Maebe
f40ea04540 * changed getarraydef() into a tarraydef.getreusable() class method
git-svn-id: trunk@31147 -
2015-06-23 21:23:18 +00:00
Jonas Maebe
f402b0d7df * changed getpointerdef() into a tpointerdef.getreusable() class method
o allows removing the ugly x86 hacks

git-svn-id: trunk@31144 -
2015-06-22 08:17:49 +00:00
Jonas Maebe
3f736f6114 * handle the loading of VMT entries at the node level, so it's done in a
type-safe way (for LLVM, and also internal consistency checking between
    the VMT as generated in nobj.pas and ncgvmt.pas)
   o also converted the VMT validity checking to the node level

git-svn-id: trunk@30950 -
2015-05-31 16:50:47 +00:00
Jonas Maebe
48ffe7b033 * don't create RTTI for internally generated recorddefs (doing so can cause
undefined references to RTTI for some of the fields, and we don't need
    it anyway because these types are not user-visible and we don't generate
    variables of these types)

git-svn-id: trunk@30945 -
2015-05-31 16:50:33 +00:00
Jonas Maebe
bd8079f896 * cache and reuse procvardefs internally created to get the address of a
procdef (halves the amount of memory needed to compile the googlapi
    package using a build unit, as done by fpmake)

git-svn-id: trunk@30856 -
2015-05-14 15:36:20 +00:00
Jonas Maebe
fb8c61889b * when getting a copy of a special procdef (such as a con/destructor) as a
procvardef, change the proctype into a regular procedure/function,
    because there are no "con/destructor procvar" types and there is code
    in the paramanager that expects e.g. a constructor to be always owned by an
    abstractrecorddef (which is not necessarily the case for the created
    procvardefs)

git-svn-id: trunk@30786 -
2015-05-03 16:51:20 +00:00
nickysn
8cf469e9ef * base the default alignment on voidpointertype.alignment instead of
voidpointertype.size for all types, that are internally represented as a
  pointer (ansi/widestrings,dynarrays,classes)

git-svn-id: trunk@30683 -
2015-04-20 12:44:11 +00:00
Jonas Maebe
3be51e1455 * fixed endless recursion in tabstractrecorddef.contains_float_field() in
case a record contains an internal typed constant of its own type
    (which happened becuase such a typed constant is also fieldvarsym, solved
     by checking for sp_static) (mantis #27880)
  * fixed several other similar cases in the compiler where we are only
    interested in instance fields, but processed all fieldvarsyms

git-svn-id: trunk@30614 -
2015-04-16 21:25:22 +00:00
florian
e0b27e1688 * unified typing and ppu reading/writing of setdef fields regarding aint/asizeint
git-svn-id: trunk@30542 -
2015-04-11 11:36:01 +00:00
Jonas Maebe
c32b6b01ad * make sure the calcparas() is called when calling getcopyas() on a
procvardef

git-svn-id: trunk@30371 -
2015-03-29 18:16:54 +00:00
Jonas Maebe
25e6eaf07f * store the to be used recordalignmin and maxcrecordalign settings inside
(abstract)recordsymtables, so that these settings don't depend on the
    current user settings when internally creating record definitions

git-svn-id: branches/hlcgllvm@30343 -
2015-03-27 21:25:56 +00:00
Jonas Maebe
765044aede * fixed a number of (mostly false) warnings regarding uninitialised
variables with -Oodfa

git-svn-id: branches/hlcgllvm@30244 -
2015-03-15 21:10:50 +00:00
Jonas Maebe
bd203a5b57 * synchronised with trunk till r30240
git-svn-id: branches/hlcgllvm@30241 -
2015-03-15 19:44:58 +00:00
florian
f0984fb628 * abstractrecorddef.contains_float_field should search recursively, resolves #27580
git-svn-id: trunk@30168 -
2015-03-11 22:05:14 +00:00
Jonas Maebe
67b8aceaee * synchronized with privatetrunk till r30095
git-svn-id: branches/hlcgllvm@30101 -
2015-03-05 20:32:15 +00:00
florian
0a17a08b28 - remove accidently committed (commented though) debug statement
git-svn-id: trunk@30000 -
2015-02-24 20:58:49 +00:00
Jonas Maebe
79a06b1514 + iphonesim/x86_64 target (64 bit iOS simulator)
git-svn-id: trunk@29970 -
2015-02-23 22:56:09 +00:00
Jonas Maebe
0203f5e32f * search iphoneall rather than cocoaall for Cocoa types on Darwin/AArch64
git-svn-id: trunk@29922 -
2015-02-23 22:53:04 +00:00
Jonas Maebe
851b3499b4 * aarch64 only/always uses its vector fpu
git-svn-id: trunk@29917 -
2015-02-23 22:52:48 +00:00
Jonas Maebe
73a9dc8077 * fixed formatting
git-svn-id: trunk@29826 -
2015-02-23 22:48:05 +00:00
florian
bc672e8fb2 * do not mess with the global symbol when compiling the implementation part of a unit
git-svn-id: trunk@29810 -
2015-02-23 22:09:46 +00:00
florian
cab98a58d0 + parameters of nested procedure variable parameters are taken into account during name mangling, resolves #27515
* de-duplicated parameter name mangling code

git-svn-id: trunk@29791 -
2015-02-22 09:16:20 +00:00
florian
726a78be08 * fix possible range check errors, resolves #27461
git-svn-id: trunk@29743 -
2015-02-17 21:14:47 +00:00
Jonas Maebe
071282b286 * add methods of objcprotocols to the list of visible "helper" methods when
loading them from ppu (like was already done when they were initially
    compiler), so they can be called via the "id" type even if no classes are
    in scope that implement them

git-svn-id: trunk@29686 -
2015-02-13 15:43:34 +00:00
svenbarth
ba75fd62f0 Fix for Mantis #27348. Applied patch by Do-wan Kim plus comments for the reason.
symdef.pas:
  * tprocdef.defaultmangledname: add some kind of differentiator to the result def, so that x(integer;integer) and x(integer):integer generate different mangled names
  * make_mangledname: also use the returndef instead of a potentially uninitialized variable like was done for tprocdef.defaultmangledname and also add a prefix (so that the two are compatible)

+ added test

git-svn-id: trunk@29579 -
2015-01-30 13:49:25 +00:00
svenbarth
943a094683 Fix for Mantis #27320.
symdef.pas, tprocdef:
  * defaultmangledname: don't use the potentially unassigned hp, but use - as intended - returndef if it is set (analog to the non-CRC part)

+ added test

git-svn-id: trunk@29537 -
2015-01-23 16:47:23 +00:00
Jonas Maebe
146e3bcb9e * synchronised with trunk till r29513
git-svn-id: branches/blocks@29516 -
2015-01-21 23:28:34 +00:00
Jonas Maebe
e97f31541e - reverted partial commit of merge
git-svn-id: branches/blocks@29515 -
2015-01-21 23:24:37 +00:00
Jonas Maebe
a76bbf363d * synchronised with trunk r29513
git-svn-id: branches/blocks@29514 -
2015-01-21 23:18:29 +00:00
sergei
d4969ca2ee * Fixed subscripting records of size OS_PAIR on big endian targets and enabled placing such records into registers. Tested and shows no regressions on mips-linux.
git-svn-id: trunk@29493 -
2015-01-17 11:30:19 +00:00
svenbarth
5848637db3 Fix generation of class names for specializations.
symdef.pas, tstoreddef:
  * is_specialization: fix exit condition to correctly recognize specializations

+ added test

git-svn-id: trunk@29328 -
2014-12-26 18:14:52 +00:00
svenbarth
77df68b2d9 Fix basically a regression since the reworking of Delphi compatible generics. The typenames of specialization now (again) contains the type names of the parameters. Additionally they are nicely formatted (TypeName<UnitName1.TypeName1[,...]>. If the full string should be longer than 255 characters an ellipsis will be added accordingly instead of the remaining parameters.
symdef.pas, tabstractrecorddef:
  + new private field rttistring to hold the string generated by RttiName
  * instead of generating the RTTI name each time on the fly it is generated only once and additionally contains (symbollically) the amount of type parameters for generics and all type parameters for specializations including their units

git-svn-id: trunk@29275 -
2014-12-12 17:25:15 +00:00
marco
2578514853 * first dragonfly patch (existing most). Mantis #27091
git-svn-id: trunk@29213 -
2014-12-07 20:27:02 +00:00
sergei
07ab3fbec4 * Fixed RTTI name mangling for nested types, resolves #26069.
git-svn-id: trunk@28958 -
2014-11-01 14:54:16 +00:00
Jonas Maebe
95b95497d4 * mark the typesyms generated in trecorddef.create_global_internal() as used
so that they don't cause hints about unused types

git-svn-id: branches/hlcgllvm@28762 -
2014-10-06 20:53:43 +00:00
Jonas Maebe
854fe2230a * store the vmt field rather than only its offset in tobjectdef, so we can
easily find it again later (since it's hidden and doesn't have a fixed
    name, it's hard otherwise). Needed for the typed constant builder.

git-svn-id: branches/hlcgllvm@28760 -
2014-10-06 20:53:38 +00:00
Jonas Maebe
489594f02f * fixed copy/paste error in r28322
git-svn-id: branches/hlcgllvm@28759 -
2014-10-06 20:53:36 +00:00
Jonas Maebe
0def3629cc * factored out adding a single field to a record based on a def
git-svn-id: branches/hlcgllvm@28757 -
2014-10-06 20:53:30 +00:00
Jonas Maebe
864b36fbe5 * create the recorddef at the start when building an anonymous recorddef,
so that we can already refer to its def while we are sill parsing
    individual elements

git-svn-id: branches/hlcgllvm@28750 -
2014-10-06 20:53:12 +00:00
Jonas Maebe
b18ba8e85b * syncrhonised with trunk up till r28471
git-svn-id: branches/hlcgllvm@28495 -
2014-08-19 20:23:11 +00:00
Jonas Maebe
63a3ff3c7f * integrate the functionality to create an address-only procvardef from a
procdef into tprocdef.getcopyas()

git-svn-id: branches/hlcgllvm@28474 -
2014-08-19 20:22:06 +00:00
Jonas Maebe
6f01bb56d6 * automatically call calcparas adter creating a procvardef from a procvdef,
so it doesn't have to be done explicitly anymore afterwards

git-svn-id: branches/hlcgllvm@28473 -
2014-08-19 20:22:03 +00:00
Jonas Maebe
83c7dbb7a6 * fixed copy/paste error that caused all warnings to be disabled after
compiling the symdef unit while building a compiler for a non-64 bit
    code generator

git-svn-id: trunk@28462 -
2014-08-18 23:03:54 +00:00
Jonas Maebe
ee76e8dbf5 * fixed warning about always-true comparison when compiling a 16 bit target
git-svn-id: trunk@28461 -
2014-08-18 23:02:12 +00:00
Jonas Maebe
5c75b6dd6b * synchronised with trunk up till r28402
git-svn-id: branches/hlcgllvm@28403 -
2014-08-13 16:04:30 +00:00
Jonas Maebe
6e1d370417 * split off the texceptaddr declaration into rtl/inc/excepth.inc, so it can
be included at the start of the implementation of the system unit (before
    the rest of except.inc)
  * catch declarations in/loading from the system unit of the TExceptAddr type
  * use this type instead of hardcoded size constants in the compiler
  * in generic code that is active for all targets, puts its use in a virtual
    method since it's only valid for targets using setjmp/longjmp-style
    exception handling (and the record is not defined at all in the JVM RTL)

git-svn-id: branches/hlcgllvm@28376 -
2014-08-10 19:40:29 +00:00