* harbour/include/hbatomic.h
! fixed inline assembler code for atomic inc/dec operations.
It's late and I haven't wrote anything in assembler using
Intel syntax for years.
* reenabled inline assembler code for atomic inc/dec operations
for POCC/XCC builds - such version should work. Please check.
* include/hbatomic.h
! Typo (__inline__ -> __inline).
! Temply disabled new _MSC_VER inline asm code, because it causes this with POCC5:
include\hbatomic.h(268): error #3114: [asm] Invalid combination of opcode and operands.
include\hbatomic.h(273): error #3114: [asm] Invalid combination of opcode and operands.
* harbour/include/hbatomic.h
+ added atomic inc/dec inline assembler code for 32bit MSVC builds
and probably also to other compilers which defines _MSC_VER
macro (XCC/POCC). Thanks to Viktor for help. Please make some tests.
* harbour/source/compiler/hbopt.c
! disabled one assert() in PCODE optimization code.
Mindaugas, I left note which explains why.
* include/hbdefs.h
+ Enabled new Harbour types. Still incomplete, and relying
on old unsafe types in places.
* contrib/gtwvg/gtwvg.c
! Warning.
* contrib/gtwvg/wvgcuig.c
! Cast errors.
* config/win/msvc.cf
+ Enabled C++ mode. This should result in stricter compiler
warnings/errors, and also enables to use standard init
method (non-C++ method is broken when using LTCG
and potentially in other scenarios like IA64). Note, this
doesn't mean we'd allow non-ANSI C usage inside Harbour,
so no C++ features are allowd. This won't cause any
drawbacks for the user AFAIK, if you know otherwise, pls
speak up. [Notice that we're using C++ mode for owatcom
on all platforms.]
* doc/whatsnew.txt
+ Documented win/owatcom and OpenSSL incompatibility.
(maybe it can be resolved, I didn't spend time on it)
* INSTALL
* C compiler -> ANSI C compiler.
* contrib/hbgt/strasint.c
% Using HB_ISDIGIT()
! Fixed to not include hbtrace.h directly.
* make_gnu.bat
% make clean and install phases merged in HB_BUILD_DLL mode.
* bin/postinst.bat
! Fix to recent change.
* include/hbdefs.h
! Moved #include hbtrace.h after HB_EXPORT is defined.
(hack)
* make_gnu_xmingwce.sh
- Disabled modifying PATH when host OS is Windows.
; NOTE: Hack.
* include/hbtrace.h
+ Added HB_EXPORT.
* doc/whatsnew.txt
* contrib/hbfimage/fi_winfu.c
* contrib/hbfimage/fi_wrp.c
* contrib/hbfimage/Makefile
- Restored previous state, because latest changed addressing
Linux build failure were causing problems for the Windows
platform.
; TOFIX: We must clean our types, to avoid these problems
and all bad hacks to work it around.
* mpkg_win.bat
* mpkg_win.nsi
+ Added files ChangeLog, TODO, ERRATA, INSTALL to root dir.
* Some changed to batch targeting robustness for situations
which cannot normally happen in normal cases, but it's
useful for debugging.
* source/compiler/hbusage.c
+ Added -u+ to help screen.
* One remaining non-ASCII char cleaned from credit list.
* include/hbpp.h
* source/pp/ppcore.c
! Renamed HB_PP_MAX_REPATS to HB_PP_MAX_REPEATS.
* ChangeLog
* Added one line to my prev entry.
* harbour/include/hbcompdf.h
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/hbcomp.c
* harbour/source/compiler/ppcomp.c
+ added support for extended definition files in Harbour compiler
switches: -u+<file>
This option can be repeated if user wants to use more then one
file with extended definitions. The definitions are added after
standard ones and after defines specified by -D<name>[=<value>]
Harbour compiler switch.
The behavior of -u[<file>] remains unchanged and -u or -u<file>
can can be used together with -u+<file>.
Note for xHarbour users: it's not exactly the same behavior as
in xHarbour compiler.
! fixed memory leak when repeated -u[<file>] options are used.
* harbour/include/hbpp.h
! changed the NOT operator precedence - it should be bigger then
logical .and./.or. operators in #if expressions
* harbour/source/vm/thread.c
! fixed typo HB_OS_WIM -> HB_OS_WIN - thanks to Francesco.
* harbour/include/hbthread.h
* harbour/source/vm/thread.c
* emulate POSIX thread conditional variables in Windows builds.
It should fix some possible bad behavior in MS-Windows Harbour
MT programs.
* harbour/source/rtl/hbinet.c
! fixed typo in HB_INETDATAREADY() - the second parameter (timeout)
was ignored.
* harbour/contrib/examples/uhttpd/uhttpdc.c
* fixed typo in conditional iMsec usage.
* harbour/contrib/examples/uhttpd/uhttpd.prg
* set default timeout for socket operations to 3 sec.
* use timeout parameter in HB_INETDATAREADY() and
increased default timeout for accept to 250 millisecs.
Please correct this modification if you prefer different values.
* INSTALL
+ Added short information for Darwin/MacPorts and contribs.
* contrib/hbmysql/Makefile
* contrib/gtalleg/Makefile
* contrib/rddsql/sddmy/Makefile
* contrib/rddsql/sddpg/Makefile
* contrib/hbpgsql/Makefile
+ Enabled on Darwin (for MacPorts).
* include/hbdefs.h
* contrib/hbfimage/fi_wrp.c
* contrib/hbfimage/Makefile
+ Enabled on Darwin (for MacPorts).
+ Added ugly hack to make it compile on Darwin.
* contrib/hbtpathy/Makefile
+ Enabled on Darwin.
* include/hbextern.ch
+ Added HB_SETENV()
* utils/hbmk2/hbmk2.prg
* Minor cleanups.
* contrib/examples/uhttpd/uhttpd.hbm
- Removed -n. It's now automatic.
; TODO: Remove the rest.
* INSTALL
* doc/man/hbmk.1
+ config/win/icc.cf
* utils/hbmk2/hbmk2.prg
+ Added Intel(R) C++ compiler support for Windows.
+ Added initial support also in hbmk2.
; This compile is actively developed and its famous of
generating the fastest running code of all C compilers.
It also outputs an extremely detailed warning list.
And the best thing: It compiled Harbour without a single
hitch. hbtest is also running fine. hbmk2 is able to
create executables.
* harbour/include/hbapi.h
* harbour/include/hbwince.h
* harbour/source/common/hbgete.c
* harbour/source/common/hbwince.c
+ added BOOL hb_setenv( const char * szName, const char * szValue )
It sets or deletes (szValue==NULL) environment variable.
Please add support for other compilers/OS-es or define
unsupported platforms to eliminate TODO warning.
For sure OS2 version have to be updated.
* harbour/source/rtl/gete.c
+ added .prg function:
HB_SETENV( <cEnvName>, [<cNewVal>] [, <lOsCP>] ) -> <lOK>
which sets or deletes (when <cNewVal> is NIL) process environment
variable <cEnvName>. <lOsCP> controls character translation to
OS encoding. By default is enabled.
TOFIX: <cEnvName> OS character encoding is missing in [HB_]_GETENV()
* harbour/include/hbdefs.h
* harbour/source/vm/hvm.c
* respect HB_START_PROCEDURE in all builds if such public
function/procedure exists.
Now whole Harbour code can be compiled with -n1 in all builds.
Warning for Windows/DOS users:
if you have main() function in your .prg code then it will be
automatically detected as application startup entry just like
in other systems.
Otherwise the 1-st linked function will be used like so far.
+ added HB_START_PROC_STRICT macro which causes that HVM ignores
HB_FS_FIRST attribute. Before this modification HB_START_PROCEDURE
has such meaning.
* harbour/include/hbcomp.h
* harbour/include/hbcompdf.h
* harbour/include/hbexpra.c
* harbour/source/compiler/hbmain.c
* harbour/source/compiler/cmdcheck.c
* harbour/source/compiler/hbcomp.c
* harbour/source/compiler/genc.c
* harbour/source/compiler/harbour.yyc
* harbour/source/compiler/harbour.y
* harbour/source/compiler/harbour.yyh
* harbour/source/compiler/ppcomp.c
* harbour/source/compiler/hbdbginf.c
* harbour/source/compiler/genhrb.c
* harbour/source/compiler/hbusage.c
! fixed static variables indexes in the generated PCODE when static
variables are declared inside extended codeblocks - in some cases
wrong PCODE was generated
! fixed initialization PCODE for thread static variables declared
inside extended codeblocks
! fixed modulename generated for debugger to always contain function name
part for extended codeblock declared as file wide static variables.
In such case "(_INITSTATICS)" is used as function name.
Without it debugger does not register entering such codeblocks and
GPFs.
+ added support for automatic adding 1-st function with the same name
as compiled .prg file if some code statements are used before 1-st
function/procedure declared explicitly in .prg file. Such mode can
be enabled by -n2 new harbour compiler switch.
* source/rtl/Makefile
* include/hbextern.ch
+ source/rtl/hbstrfmt.c
+ Added HB_STRFORMAT() sprintf()-like formatting.
+ Added hb_StrFormat() C level function.
; Work of Mindaugas sent to the list, with some minor
cleanups applied. Please test on your platform/compiler.
* contrib/examples/uhttpd/uhttpd.prg
* contrib/examples/uhttpd/cgifunc.prg
* contrib/examples/uhttpd/session.prg
* contrib/examples/uhttpd/uhttpd.hbm
* xhb compatibility cleanups.
- xhb lib and header dependency removed.
* utils/hbmk2/hbmk2.prg
* Changed so that libs won't inherit their .hbm file paths.
+ Embedded .hbm files will inherit their parents dir.
* harbour/include/hbapi.h
* harbour/include/hbstack.h
* harbour/source/vm/hvm.c
* harbour/source/vm/estack.c
* harbour/source/vm/memvars.c
! fixed __M[V]CLEAR()/ __M[V]RESTORE() to be exactly Clipper compatible
and do not release PUBLIC GetList value. Here it's small example
which illustrates current behavior and exploits bug in previous
version:
memvar getlist
proc main()
getlist:="public:getlist"
? getlist
private getlist:="private:getlist"
? getlist
CLEAR MEMORY
? getlist
return
* harbour/source/compiler/gencc.c
+ added Harbour exception to the license - it's mine code with
few lines added/modified by Ryszard and Viktor. I hope they do
not have anything against.
* harbour/source/compiler/hbmain.c
* harbour/source/compiler/hbident.c
* minor formatting
* harbour/include/hbsetup.h
* minor modifications in macros
* disabled some GCC function attributes when ICC is used
* harbour/source/pp/ppcore.c
* casting
* harbour/source/vm/fm.c
* destroy dlmalloc mspace in hb_xexit() when USE_DL_PREFIX is used
otherwise call malloc_trim(0) to at least reduce allocated memory
* include/hbdefs.h
+ Added similar to HB_DONT_DEFINE_BOOL hack: HB_DONT_DEFINE_LONG.
(needed until we clear up our types.)
* contrib/hbfimage/fi_winfu.c
* contrib/hbfimage/fi_wrp.c
* contrib/hbfimage/Makefile
* doc/whatsnew.txt
% Deleted #includes/macro checks not necessary.
! Disabled body of fi_winfu.c when not compiled under
Windows. This fixes one Linux compilation error.
% Replaced HB_OS_WIN_USED with HB_DONT_DEFINE_BOOL and
HB_DONT_DEFINE_LONG, moved foreign headers before
Harbour headers, added non-Harbour Windows detection
in fi_wrp.c (I'm not sure that is necessary), so that
it now compiles on both Windows and non-Windows (Linux
tested) systems.
+ Renabled on Linux in Makefile.
; Please review/test these changes.
* harbour/include/hbgtinfo.ch
+ added HB_GTI_FONTSEL
* harbour/source/rtl/gtxwc/gtxwc.c
+ added support for HB_GTI_FONTSEL
+ added support for HB_GTI_PALETTE
+ harbour/tests/gtxfnt.prg
+ added small demonstration/test code for HB_GTI_FONTSEL
and XWindow GTs
* harbour/source/rtl/hbproces.c
* replaced execv() with execvp() in *nix builds
* harbour/include/hbthread.h
* harbour/source/vm/thread.c
! added support for conditional variables working like in PTHREADS
in OS2 builds - it fixes the problem with possible dead lock or
starvation effect which can also cause dead lock in some cases.
OS2 users please test current code.
* config/win/bcc32.cf
* Synced flags with non-GNU make. This adds MT support.
* config/win/install.cf
! Install command modified to not give errors and block
the whole build process if the target file already exists.
The workaround consisted of changing 'copy' to 'xcopy /d /y'.
* include/inkey.ch
+ Added comments on K_MM*DOWN codes.
These also should be protected, and moved to
local the local headers belonging to GTs
supporting this feature (GTWVG notably).
* source/rtl/gtwvt/gtwvt.c
+ Protected non-Clipper compatible extension regarding
K_MM*DOWN with __HB_GTWVT_GEN_K_MMDOWN_EVENTS #define.
Also see notes in about app level code modification
needed if you want to continue using default Harbour build:
2009-02-18 12:20 UTC+0100 Viktor Szakats (harbour.01 syenar hu)
* include/hbextern.ch
* common.mak
* source/rtl/Makefile
* source/rtl/mouse53.c
+ Added HB_MMIDDLEDOWN() functions, plus MMIDDLEDOWN() alias.
These are always available regardless of the CLIP53
compatibility setting.
; TOFIX: K_MM*DOWN events _override_ C5.3 compatible
K_MOUSEMOVE events (in GTWVT only), which created
an unfortunate incompatibility with our target
C5.x compiler when using some specific GT. Ideas
are welcome how to fix this. We should probably
either send out both codes and let the inkey queue
do the filtering as requested by app (we need a
new filter then), or remove these event code and
generate proper K_MOUSEMOVE events instead. Apps
using this these incompatible keycodes are encouraged
to replace expressions:
'( key == K_KK*DOWN )' with the compatible:
'( key == K_MOUSEMOVE .AND. M*DOWN() )' to prepare
for the fix.
* harbour/include/hbapi.h
* harbour/source/vm/cmdarg.c
* harbour/source/vm/hvm.c
* added hb_cmdargUpdate() function which makes final argc/argv updating
(code moved from hb_cmdargInit())
! use hb_cmdargUpdate() - it fixes GPF in MT mode when some fs* functions
were called before allocating HVM stack
; TOFIX: change some hb_setGet*() functions to work without HVM stack.
* harbour/source/rtl/hbtoken.c
* formatting
* harbour/include/dbinfo.ch
+ added DBI_ISTEMPORARY
* harbour/include/hbapirdd.h
* harbour/source/rdd/wafunc.c
+ added hb_rddCreateTableTemp() C function to create temporary
table which are automatically deleted on close. All indexes and
memo files for such tables are also in temporary files and
automatically deleted on close. Please remember that Harbour uses
OS functions to create temporary files so it's installation dependent
how much disk space is available for them. Some RDDs may also do
not create temporary files but keep everything in memory if they
decide it's more efficient.
* harbour/source/rdd/dbcmd.c
+ added .prg function to create temporary files:
dbCreateTemp( <cAlias>, <aStruct>, <cRDD>, ;
<cCodePage>, <nConnection> ) -> <lSuccess>
Not all RDDs have to support it. It's verified by support for
DBI_ISTEMPORARY dbInfo() action.
* harbour/include/hbrdddbf.h
* harbour/include/hbrddnsx.h
* harbour/include/hbrddcdx.h
* harbour/include/hbrddntx.h
* harbour/source/rdd/dbf1.c
* harbour/source/rdd/dbffpt/dbffpt1.c
* harbour/source/rdd/dbfntx/dbfntx1.c
* harbour/source/rdd/dbfnsx/dbfnsx1.c
* harbour/source/rdd/dbfcdx/dbfcdx1.c
+ added support for temporary tables in native RDDs.
Most of above modifications were done by Viktor Szakats.
* bin/hbmk.bat
+ Added autodetection of Harbour dirs. So this batch will
now work when called from any directories, and even if
'/bin' dir is added to the PATH. IOW, it's not anymore
necessary to supply these envvars in normal cases:
HB_INSTALL_PREFIX, HB_BIN_INSTALL, HB_LIB_INSTALL,
HB_INC_INSTALL. If someone moves this batch file away
from its original place, these envvars can still be used
to point to the Harbour tree, but support for this may
be removed in the future to keep things simple.
+ Added support for multiple .prg files on the command line.
The first will be the main program.
+ Added support to supply .c files on the command line.
[these features need Windows NT or upper]
; TODO: Test OpenWatcom. I had to remove -fo and FILE
directives from script, if this was required we
need to find a solution to support multiple
input files.
* bin/hbmk_b32.bat
* bin/hbmk_vc.bat
+ Now these files can be called from any directories and even
when '/bin' dir is added to the PATH.
+ Added support for unlimited number of command line parameters.
[these features need Windows NT or upper]
* Changed to not interfere with 'echo off' setting of the
caller.
; NOTE: After this change it's possible to get rid of all
non-customized version of hbmk_b32.bat/hbmk_vc.bat
files in the source tree. I plan to remove these in
the near future. The rest can be simplified in most
places. This will only cause some inconvenience
for developers still using Win95/98/ME.
+ bin/hbmk_gcc.bat
+ bin/hbmk_ow.bat
+ Added make batch files to MingW/GCC and OpenWatcom.
* contrib/hbwin/win_tprn.prg
! Fixed recent typo.
* include/hbstdgen.ch
- Removed #undef __FILE__. I understand this isn't needed anymore.
* harbour/include/hbthread.h
* harbour/source/vm/thread.c
! fixed a bug on OS/2 MT implementation adding a new function,
_hb_cond_timed_wait() because when a semaphore gets posted it has
to be reset, otherwise it does not stop new threads entering it in a wait.
* harbour/include/hbpp.h
* harbour/source/pp/hbpp.c
* harbour/source/pp/ppcore.c
* harbour/source/compiler/hbmain.c
+ added support for dynamically calculated #define results
! removed old code which was adding __FILE__ #define constant
It was not working correctly like in other compilers ignoring
#include
+ added __FILE__ and __LINE__ #define directives using dynamically
calculated results - now it works as expected respecting #include
directives and the implementation is local to PP without any hacks
in core compiler code.
* harbour/contrib/hbct/files.c
* added support for optional milliseconds part in HB_FSETDATETIME()/
HB_FGETDATETIME().
* harbour/source/rtl/philes.c
! minor fix in return value
* include/hbdefs.h
* contrib/gtwvg/gtwvg.h
* contrib/gtwvg/wvggui.h
* contrib/examples/terminal/trm_client.prg
+ Added HB_EXTERN_START/END.
* include/hbdefs.h
+ Added some comments from the original mail from Przemek.
+ harbour/include/hbwapi.h
As this header will be required fot many libraries in contrib
I have placed it in core folder. It will pave me the way to
isolated WINAPI wrappers as discussed earlier. Before I begin to
shift functions in harbour/contrib/hbwin please feel free to
include in this header whatever you feel will be required
for such transition.
* harbour/include/hbextern.ch
* harbour/source/rtl/philes.c
+ added HB_FGETDATETIME(). A "brother" of function HB_FGETDATETIME().
; TODO: BTW, hb_fs*() functions can get a better precision
(milliseconds). It would be nice to support it also at .prg level.
* source/rtl/gtwvt/gtwvt.c
* BYTE -> CHAR (Windows) - UNDONE.
* include/hbdefs.h
+ Added (commented) new types. Work in progress, incomplete.
* contrib/hbssl/ssl.c
- Temporary disabled SSL_GET_READ_AHEAD() as a workaround for
BCC linking problem with SSL_get_read_ahead() OpenSSL
library function (symbol is unresolved due to missing
leading underscore in implib generated ssleay32.lib file).
* source/vm/set.c
% hb_fsFile() -> hb_fsFileExists()
* source/rtl/hbffind.c
* HB_WIN_32_MATCH -> HB_WIN_MATCH
* doc/whatsnew.txt
* include/hbsetup.h
* source/vm/fm.c
* #define HB_FM_WIN32_ALLOC renamed to HB_FM_WIN_ALLOC.
; INCOMPATIBLE: For those who are using this switch, please
update your make files.
* source/rtl/tpopup.prg
! Fixed to draw separator lines with BOX GT attributes,
so that they appear properly.
* harbour/include/hbapi.h
* harbour/source/vm/cmdarg.c
+ added const char * hb_cmdargARGVN( int argc )
* harbour/source/vm/cmdarg.c
* try to convert executable file name argument in argv[0] to
absolute path if it's relative one. Respect PATH envvar it necessary.
* harbour/source/common/hbdate.c
* formatting
* harbour/source/rtl/philes.c
* harbour/source/rtl/errorint.c
* harbour/source/rtl/filesys.c
! use hb_cmdargARGVN(n) instead of hb_cmdargARGV()[n] to avoid
possible GPF if some function will be called before argument
initialization
* removed additional logic which tries to convert relative paths
in file name from hb_fsBaseDirBuff() and HB_PROGNAME() functions
