* harbour/include/hbapi.h
* harbour/include/hbvmpub.h
* harbour/source/vm/classes.c
+ added HB_ARRAY_OBJ() and changed HB_IS_OBJECT() to use it
instead of HB_OBJ_CLASS() - it allows to use HB_IS_OBJECT()
without including hbapicls.h
* harbour/contrib/rdd_ads/ads1.c
* synced with Toninho modifications in xHarbour:
(! Inverted checking for memo data types, verifying binary first,
to prevent freeze in some empty ADT tabes.)
* harbour/include/Makefile
+ added usrrdd.ch
* harbour/include/hbapiitm.h
* harbour/source/vm/itemapi.c
+ added hb_itemPutSymbol()
* harbour/source/vm/hvm.c
! fixed HB_P_POPVARIABLE - it should work like HB_P_POPMEMVAR
This Ron's code which illustrates it:
PROCEDURE Main()
USE Test
First := First
CLOSE
? First
RETURN
In practice it means that we do not need this PCODE at all and
we should replace it in compiler by HB_P_POPMEMVAR and reuse
it in the future for differ things.
* harbour/source/vm/memvars.c
* code cleanup
* harbour/source/vm/runner.c
! fixed possible memory leaks when corrupted .hrb file is loaded
- harbour/Common.mak
+ harbour/common.mak
* harbour/make_b32.bat
* harbour/make_vc.bat
* harbour/makefile.bc
* harbour/makefile.vc
! Filename casing correction. (fixed in two previous commits)
! Missing CVS ID added. (fixed in previous commit)
! Few minor formatting, typos fixed.
+ Few build envvars prefixed with HB_ (some renamed)
* Few minor cleanups not related the actual make process.
; Thanks Marek for the great work.
- harbour/make_b16.bat
- harbour/make_vc64.bat
- harbour/mkbcdll.bat
- harbour/hrbdll.bc
- harbour/makefile.nt
- harbour/makefile64.nt
- harbour/hrbdll.vc
- removed no more needed files
* harbour/makefile.bc
* harbour/makefile.vc
* harbour/make_b32.bat
* harbour/make_vc.bat
* totaly rewritten to support sharing common file lists,
automatic DLL creation and using implicit rules as a
default approach
+ harbour/Common.mak
+ common file lists and definitions shared between Bcc
and Msc makefiles
TODO : add automatic building of contrib libs/dll(s)
* harbour/include/hbstack.h
* harbour/source/vm/estack.c
- removed hb_stackTopItem()
+ added hb_stackItemBasePtr(), hb_stackAllocItem()
* changed the item initialization/clearing in push/pop operation
Now only complex item are cleared and it's not guaranteed that
the top item will be set to NIL. In fact it was not guarantied
even before my modifications. You coule be sure only that the
allocated item will not be complex one and you can safely pass
pointer to it to any other functions. It allow to clean a little
bit some code and remove redundant and repeated setting HB_IT_NIL
for allocated items.
Now using hb_stackPush() can be necessary only in some very seldom
cases. hb_stackAllocItem() allocates new item and returns pointer
to this item. After some code cleaning hb_stackPush() is not used
by core Harbour code at all so if you have it in your sources then
please check if you should not update them.
* harbour/source/rdd/usrrdd/usrrdd.c
* use hb_vmPushNil() instead of hb_stackPush()
* harbour/source/vm/arrays.c
* changed hb_arrayNew() to be safe for automatic GC activation
* some minor speed improvement
* harbour/source/vm/codebloc.c
* changed hb_codeblockNew() and hb_codeblockMacroNew() to be safe
for automatic GC activation
* harbour/source/vm/debug.c
* use only stack function/macros instead of direct accessing hb_stack
I hope that I haven't break anything.
* harbour/source/vm/hvm.c
! use hb_stackAllocItem() instead of hb_stackTopItem() to make our
HVM reentrant safe. This modifications also fixed some possible
bugs which could be exploited by other modules which have to execute
.prg code - f.e. in RDD when relation codeblock has to be executed
to position record in fieldget/fieldput operations.
* harbour/source/vm/itemapi.c
* harbour/source/vm/memvars.c
* use hb_stackAllocItem()
The above modifications finish stack usage and making our HVM reentrant
safe. There is only one thing which have to be fixed yet. It's HVM
support for xbase++ syntax in things like:
func(&<paramList>), aVal[&<indexList>], {&<itemList>}
To fix it we will have to change PCODE generated by compiler and
replace some PCODEs used for current code by the new one. I will
wait with this modifications for Ryszard.
With the above exception now we are ready to implement destructors,
add automatic GC and begin to work on MT though before that I'd like
to clean the RT error support.
Please carefully check if all is correct. It was modification in core
code and even small typo can break whole application.
* harbour/include/hbstack.h
* harbour/source/vm/estack.c
- removed hb_stackTopItem()
+ added hb_stackItemBasePtr(), hb_stackAllocItem()
* changed the item initialization/clearing in push/pop operation
Now only complex item are cleared and it's not guaranteed that
the top item will be set to NIL. In fact it was not guarantied
even before my modifications. You coule be sure only that the
allocated item will not be complex one and you can safely pass
pointer to it to any other functions. It allow to clean a little
bit some code and remove redundant and repeated setting HB_IT_NIL
for allocated items.
Now using hb_stackPush() can be necessary only in some very seldom
cases. hb_stackAllocItem() allocates new item and returns pointer
to this item. After some code cleaning hb_stackPush() is not used
by core Harbour code at all so if you have it in your sources then
please check if you should not update them.
* harbour/source/rdd/usrrdd/usrrdd.c
* use hb_vmPushNil() instead of hb_stackPush()
* harbour/source/vm/arrays.c
* changed hb_arrayNew() to be safe for automatic GC activation
* some minor speed improvement
* harbour/source/vm/codebloc.c
* changed hb_codeblockNew() and hb_codeblockMacroNew() to be safe
for automatic GC activation
* harbour/source/vm/debug.c
* use only stack function/macros instead of direct accessing hb_stack
I hope that I haven't break anything.
* harbour/source/vm/hvm.c
! use hb_stackAllocItem() instead of hb_stackTopItem() to make our
HVM reentrant safe. This modifications also fixed some possible
bugs which could be exploited by other modules which have to execute
.prg code - f.e. in RDD when relation codeblock has to be executed
to position record in fieldget/fieldput operations.
* source/compiler/gencli.c
! .NET IL implementation for HB_P_POPLOCALNEAR and HB_P_PUSHLOCALNEAR
Notice: These opcodes are responsable for loading the values on the locals and
also for pushing the locals contents onto the virtual machine stack.
You can already run tests like this:
function Main()
local c := "Hello world from Harbour.NET !"
? c
return nil
Remember the process to build a .NET app:
Harbour.exe -gi -n hello.prg --> hello.il
ilasm.exe hello.il --> hello.eor Ryszard.
With the above exception now we are ready to implement destructors,
add automatic GC and begin to work on MT though before that I'd like
to clean the RT error support.
Please carefully check if all is correct. It was modification in core
code and even small typo can break whole application.
* harbour/bin/hb-func.sh
* harbour/include/hbgtcore.h
* use HB_FUNC*() macros
* harbour/source/rdd/dbcmd.c
! added fix for GPF in APPEND FROM when source file has repeated
fields with the same names.
* harbour/source/rtl/hbgtcore.c
- removed forgotten printf() used as debug message
* harbour/source/rtl/gtgui/gtdef.c
* register "GUI" GT as default one by executing
hb_gtSetDefault( "GUI" ) at startup
* harbour/source/vm/memvars.c
* indenting
* harbour/include/hbrddcdx.h
* harbour/include/hbrdddbf.h
* harbour/include/hbrdddbt.h
* harbour/include/hbrddfpt.h
* harbour/include/hbrddntx.h
* harbour/source/rdd/dbf1.c
* harbour/source/rdd/dbfcdx/dbfcdx1.c
* harbour/source/rdd/dbfdbt/dbfdbt1.c
* harbour/source/rdd/dbffpt/dbffpt1.c
* harbour/source/rdd/dbfntx/dbfntx1.c
* remove SYSNAME() method - use the default WA one.
* harbour/source/vm/hvm.c
! fixed last C stack usage of HB_ITEM - now HVM should be cleaned
and we begin to implement automatic GC activation
* harbour/config/w32/mingw32.cf
* updated for current xHarbour state - not tested but I guess it
works for Lorenzo
* harbour/config/w32/mingw32.cf
! Fix to make it work in environments where "gcc @__lib__.tmp"
construct is not possible. The "@" expansion is done by gcc
when using Cygwin, but it's not done (and left to the shell,
which doesn't do it either when using CMD/COMMAND) when using
plain MinGW.
* harbour/include/hbexpra.c
* harbour/include/hbexprb.c
* indenting
* harbour/source/vm/hvm.c
* harbour/source/vm/macro.c
! fixed HB_P_MACROPUSHINDEX
- removed hb_vm_iExtraIndex
- fixed GPF in direct uninitialized HB_ITEM usage
- make it safe for HVM reentrant and pending RT errors
There is a set of global hb_vm*Extra* variables which
tries to extend HVM by adding some additional registers
to it. This code does not work as it should even know
and it's absolutely not reentrant safe. It will have to
be fixed before we will have .prg destructor and MT support.
As I can see the same problems has xHarbour.
* harbour/include/hbapi.h
* harbour/source/vm/proc.c
+ added BOOL hb_procinfo( int iLevel, char * szName,
USHORT * puiLine, char * szFile )
* harbour/include/hbstack.h
* harbour/source/vm/estack.c
+ added hb_stackTotalItems(), hb_stackDateBuffer(),
hb_stackGetStaticsBase(), hb_stackSetStaticsBase(),
hb_stackBaseProcInfo()
* harbour/source/vm/garbage.c
* indenting
* harbour/source/vm/classes.c
* harbour/source/vm/codebloc.c
* harbour/source/vm/extend.c
* harbour/source/vm/fm.c
* harbour/source/vm/hvm.c
* replaced direct stack access by stack functions/macros
* harbour/source/vm/debug.c
* removed hb_stackLenGlobal() and replace it by hb_stackTopOffset()
Warning!!! hb_stackTopOffset() returnes results smaller by 1.
The TOP stack item is undefined and should not be accessed by
debuger.
* changed HB_DBG_VMSTKGLIST() to operate on functions/macros.
It's possible that the above will badly interact with core debugger
code so I stop farther modifications leaving them for someone who
knows debugger core code.
* include/hbapi.h
* source/vm/garbage.c
* changed hb_gcUnregistedSweep to static function
* fixed handling of sweeper functions
* there is no need to call hb_UnregisterSweep - this
is done automatically inside hb_gcFree and during
garbage collecting
* harbour/source/rtl/gtgui/Makefile
* harbour/source/rtl/gtgui/gtgui.c
+ harbour/source/rtl/gtgui/gtdef.c
+ added GT version info
* modified the hack which overwrite the default GT name
Now original GTGUI is always present and only separate
file add linker bindings. It will be cleaner solution which
will allow to use REQUEST HB_GT_GUI in valid programs and
libraries.
* harbour/source/rtl/gtwin/gtwin.c
* indenting
* harbour/source/vm/garbage.c
* minor modification: use HB_GC_PTR() to take HB_GRABAGE pointer
+ added alternate code for sweep function execution - disabled by
default, maybe it will be used when we agree the sweep function
behavior
* harbour/contrib/rdd_ads/ads1.c
* BCC warnings clean-up - synced with xHarbour
* harbour/source/rtl/empty.c
* added HB_IT_SYMBOL
* TODO
* assigned SQLRDD to me
* source/compiler/harbour.c
* source/compiler/harbour.y
* fixed one more warning (function with no RETURN) in cases:
IF tmp>0
RETURN 1
ELSE
RETURN 0
ENDIF
* include/hbapi.h
* source/vm/garbage.c
* fixed to call hb_gcUnregisterSweep before the memory
block clean-up
* harbour/source/rtl/gtgui/Makefile
* harbour/source/rtl/gtgui/gtgui.c
* hacked GTGUI for W32 to be compiled using the same name as RTL default
GT driver (GTWIN).
* harbour/include/hbgtcore.h
* harbour/source/rtl/hbgtcore.c
* changed 'char *' to 'const char *' in hb_gtSetDefault(), hb_gtLoad()
* harbour/source/rtl/errorsys.prg
! Fixed to use OutErr() instead of QOut() for CA-Cl*pper
compatibility. (Thanks Chen)
! Fixed to set ErrorLevel to 1 in case of a RT error.
(this was once removed by me as a fix about 7 years ago,
but I can't remember what was the reason. It looks to be
a wrong fix back then.)
* harbour/utils/hbtest/rt_array.prg
* Removed one codeblock unused var warning.
* harbour/source/rtl/Makefile
+ harbour/source/rtl/gtclip.c
+ harbour/source/rtl/gttone.c
* harbour/include/hbgtcore.h
* separated low level TONE and ClipBoard code for W32
new functions:
hb_gt_w32_Tone()
hb_gt_w32_SetClipboard()
hb_gt_w32_GetClipboard()
Please update non GNU make files
* harbour/source/rtl/gtwin/gtwin.c
* harbour/source/rtl/gtwvt/gtwvt.c
* use hb_gt_w32_*() for common GT code
* harbour/config/w32/global.cf
+ harbour/source/rtl/gtgui/Makefile
+ harbour/source/rtl/gtgui/gtgui.c
+ added GTGUI GT driver - it inherits from GTNUL and add support
for TONE and CLIPBOARD
* harbour.spec
* include/hbver.h
* updated to set version 0.46.2
* include/hbapi.h
* source/vm/garbage.c
+ added possibility to register a sweep function that will scan
memory block allocated with hb_gcAlloc
* source/compiler/harbour.c
* source/compiler/harbour.y
* fixed 'unreachable code' warning in IF/ELSEIF
* fixed 'variable unused' warnings when -w3 switvh was used
* harbour.spec
* include/hbver.h
* updated to set version 0.46.2
* include/hbapi.h
* source/vm/garbage.c
+ added possibility to register a sweep function that will scan
memory block allocated with hb_gcAlloc
* source/compiler/harbour.c
* source/compiler/harbour.y
* fixed 'unreachable code' warning in IF/ELSEIF
* fixed 'variable unused' warnings whe of a simple "nothing to do" message.
2003-05-09 20:30 UTC-0400 David G. Holm <dholm@jsd-llc.com>
* Makefile
+ config/doc.cf
+ doc/Makefile
* doc/gmake.txt
+ doc/en/Makefile
+ doc/es/Makefile
+ Added the ability to install files from doc directory using
the new environment variable HB_DOC_INSTALL, which is left
undefined by default. All files from doc, doc/en, and doc/es
are installed if the HB_DOC_INSTALL variable is defined when
you run 'make install'.
2003-05-09source/rtl/achoice.prg
* harbour/source/rtl/checkbox.prg
* harbour/source/rtl/dbedit.prg
* harbour/source/rtl/getsys.prg
* harbour/source/rtl/listbox.prg
* harbour/source/rtl/persist.prg
* harbour/source/rtl/pushbtn.prg
* harbour/source/rtl/radiobtn.prg
* harbour/source/rtl/tbcolumn.prg
* harbour/source/rtl/tbrowse.prg
* harbour/source/rtl/tclass.prg
* harbour/source/rtl/teditor.prg
* harbour/source/rtl/tget.prg
* harbour/source/rtl/tgetlist.prg
* harbour/source/rtl/tlabel.prg
* harbour/source/rtl/treport.prg
* harbour/utils/hbdoc/genchm.prg
* harbour/utils/hbdoc/genhtm.prg
* harbour/utils/hbdoc/genng.prg
* harbour/utils/hbdoc/genos2.prg
* harbour/utils/hbdoc/genrtf.prg
* harbour/utils/hbdoc/gentrf.prg
* harbour/utils/hbdoc/html.prg
* harbour/utils/hbdoc/ng.prg
* harbour/utils/hbdoc/os2.prg
* harbour/utils/hbdoc/rtf.prg
* harbour/utils/hbmake/checks.prg
* harbour/utils/hbmake/hbmake.prg
* harbour/utils/hbmake/hbmutils.prg
* harbour/utils/hbmake/radios.prg
* harbour/utils/hbtest/rt_hvm.prg
+ HB_SYMBOL_UNUSED() added for .prg code. It can suppress unused
var warnings where applicable. The code won't generate any final
pcodes, so it won't cause overhead. The only current downside
is that it cannot be used inside codeblocks.
! Fixed a number (>200) of declared but unused variable /w2 error.
All parts were scanned and fixed (except contrib).
Note that there are several false "unreachable code" warnings,
and there are still quite a few unused var warnings inside
codeblocks. After fixing these /w2 switch could be added to
the standard build process to maintain code quality.
* harbour/include/common.ch
* harbour/source/debug/dbghelp.prg
* harbour/source/debug/dbgmenu.prg
* harbour/source/debug/dbgtarr.prg
* harbour/source/debug/dbgtobj.prg
* harbour/source/debug/dbgwa.prg
* harbour/source/debug/debugger.prg
* harbour/source/rdd/dbupdat.prg
* harbour/source/rtl/achoice.prg
* harbour/source/rtl/checkbox.prg
* harbour/source/rtl/dbedit.prg
* harbour/source/rtl/getsys.prg
* harbour/source/rtl/listbox.prg
* harbour/source/rtl/persist.prg
* harbour/source/rtl/pushbtn.prg
* harbour/source/rtl/radiobtn.prg
* harbour/source/rtl/tbcolumn.prg
* harbour/source/rtl/tbrowse.prg
* harbour/source/rtl/tclass.prg
* harbour/source/rtl/teditor.prg
* harbour/source/rtl/tget.prg
* harbour/source/rtl/tgetlist.prg
* harbour/source/rtl/tlabel.prg
* harbour/source/rtl/treport.prg
* harbour/utils/hbdoc/genchm.prg
* harbour/utils/hbdoc/genhtm.prg
* harbour/utils/hbdoc/genng.prg
* harbour/utils/hbdoc/genos2.prg
* harbour/utils/hbdoc/genrtf.prg
* harbour/utils/hbdoc/gentrf.prg
* harbour/utils/hbdoc/html.prg
* harbour/utils/hbdoc/ng.prg
* harbour/utils/hbdoc/os2.prg
* harbour/utils/hbdoc/rtf.prg
* harbour/utils/hbmake/checks.prg
* harbour/utils/hbmake/hbmake.prg
* harbour/utils/hbmake/hbmutils.prg
* harbour/utils/hbmake/radios.prg
* harbour/utils/hbtest/rt_hvm.prg
+ HB_SYMBOL_UNUSED() added for .prg code. It can suppress unused
var warnings where applicable. The code won't generate any final
pcodes, so it won't cause overhead. The only current downside
is that it cannot be used inside codeblocks.
! Fixed a number (>200) of declared but unused variable /w2 error.
All parts were scanned and fixed (except contrib).
Note that there are several false "unreachable code" warnings,
and there are still quite a few unused var warnings inside
codeblocks. After fixing these /w2 switch could be added to
the standard build process to maintain code quality.
* harbour/config/w32/dir.cf
+ added EOL at EOF
* harbour/config/w32/global.cf
* changed clean and dirbase commands to work cleanly with CMD.EXE
* harbour/source/rtl/gtwin/gtwin.c
* always try to allocate console
+ added workaround for MSYS console which does not support
WriteConsoleOutput()/ReadConsoleInput()
* harbour/source/rtl/gtwvt/gtwvt.c
* changed default codepage to OEM
* harbour/config/w32/dir.cf
+ added EOL at EOF
* harbour/config/w32/global.cf
* changed clean and dirbase commands to work cleanly with CMD.EXE
* harbour/source/rtl/gtwin/gtwin.c
* always try to allocate console
+ added workaround for MSYS console which does not support
WriteConsoleOutput()/ReadConsoleInput()
* harbour/source/rtl/gtwvt/gtwvt.c
* changed default codepage to OEM
* harbour/include/hbapicls.h
* harbour/source/vm/classes.c
+ added hb_objHasMesage( PHB_ITEM pObject, PHB_DYNS pMessage )
+ added hb_objGetpMethod() for backward binary compatibility with FWH
* hidden for 3-rd party users hb_objGetMethod() and hb_objPopSuperCast()
If for some reason 3-rd party code will need these methods please
tell me in which case and I'll add necessary functions. Please
try to not use any non-public function. I strongly prefer to
add new public functions for which we will know that they are used
by 3-rd party code then create problems with binary compatibility
after updating local HVM functions.
