Re: [fpc-devel] data alignment

2005-08-13 Thread Florian Klaempfl
Jan Ruzicka wrote: > Hi Rostya and Florian > > Is there a way to align code? (specially in loops.) > A Profiler is complaining about an unaligned loop. > It gives following suggestion: > "... With gcc, use the -falign-loops=16 compiler flag (broken in gcc > 3.1; gcc 3.3 or later should be used)."

Re: [fpc-devel] WinCE port docs at wiki

2005-08-13 Thread Yury Sidorov
1. I got FPC to work with WinCE console. Expect a rtl sources update soon. 2. Since H2PAS dont eat MS headers the easiest way to port is to use win32 Windows unit and remove all functions which are missing in WinCE. The following list of WinCE functions can be used as reference: http://www.rai

Re: [fpc-devel] WinCE port docs at wiki

2005-08-13 Thread Florian Klaempfl
Yury Sidorov wrote: > 2. Since H2PAS dont eat MS headers the easiest way to port is to use > win32 Windows unit and remove all functions which are missing in WinCE. > The following list of WinCE functions can be used as reference: > http://www.rainer-keuchel.de/wince/dllexports/ipaq-coredll.txt >

Re: [fpc-devel] WinCE port docs at wiki

2005-08-13 Thread Yury Sidorov
From: "Florian Klaempfl" <[EMAIL PROTECTED]> Yury Sidorov wrote: 2. Since H2PAS dont eat MS headers the easiest way to port is to use win32 Windows unit and remove all functions which are missing in WinCE. The following list of WinCE functions can be used as reference: http://www.rainer-keuche

Re: [fpc-devel] WinCE port docs at wiki

2005-08-13 Thread Florian Klaempfl
Yury Sidorov wrote: > From: "Florian Klaempfl" <[EMAIL PROTECTED]> > >> Yury Sidorov wrote: >> >>> 2. Since H2PAS dont eat MS headers the easiest way to port is to use >>> win32 Windows unit and remove all functions which are missing in WinCE. >>> The following list of WinCE functions can be used

Re: [fpc-devel] WinCE port docs at wiki

2005-08-13 Thread Oro06
Yury Sidorov a écrit : 2. Since H2PAS dont eat MS headers the easiest way to port is to use win32 Windows unit and remove all functions which are missing in WinCE. The following list of WinCE functions can be used as reference: define like this are droped #define CALLBACKPASCAL // it seem

Re: [fpc-devel] WinCE port docs at wiki

2005-08-13 Thread Luiz Américo
Yury Sidorov wrote: 3. I think sqllite is dependent from Windows,SysUtils,Classes,etc. We need to port them first. No. Vanilla sqlite has no dependencies, is just a translation of the sqlite header, so if sqlite already works in wince with c compilers it should work with fpc too. The uni

Re: [fpc-devel] WinCE port docs at wiki

2005-08-13 Thread Yury Sidorov
From: "Oro06" <[EMAIL PROTECTED]> i'll do that. ( rainer site is a little bit old (wince2.11,3) since there has been wince4.0, 4.1 and now most of systems are 4.2 ) i started with 4.2 headers, i used h2pas for types and consts and do the rest by hand, you're righ it's long task compared arm an

Re: [fpc-devel] According to Valgrind, something is wrong in IndexByte

2005-08-13 Thread Micha Nelissen
On Sat, 13 Aug 2005 00:03:01 -0400 Jan Ruzicka <[EMAIL PROTECTED]> wrote: > On Aug 12, 2005, at 05:04, Micha Nelissen wrote: > > > But if I use my own version (which I use for delphi and is i386 > > specific), it reports no errors in this location, for > > Dbf_Common/Dbf_Memo. > > > > i386 Inde

Re: [fpc-devel] WinCE port windows.h

2005-08-13 Thread Oro06
1°)actual (wince 4.2) windows.h need at : windef.h, types.h, winbase.h, wingdi.h, winuser.h, winreg.h, shellapi.h, ole2.h, imm.h, tchar.h, excpt.h, strsafe.h actual window.pp (from fpc 2.1.X) look at : base.inc, errors.inc, defines.inc, struct.inc, messages.inc, ascfun.inc, unifun.inc, unidef.

Re: [fpc-devel] WinCE port windows.h

2005-08-13 Thread Yury Sidorov
1) The main problem is that Windows unit for win32 is generated automatically from GNU headers. Manual modifcations to it will make it frozen and make automatic generaton impossible. I know that WinCE API is very close to Win32 and it is possible to create shared Windows unit for Win32, Win64, W

[fpc-devel] Constant arrays sometimes produce incorrect data tables

2005-08-13 Thread David Butler
Hi all I picked up this weird code generation problem (it was also present in 2.0.0). I reported it as bug 4277 today. Unfortunately I can only reproduce it in a specific context, so please e-mail me for the unit. The problem is when you define an array of constant strings. For example: const

Re: [fpc-devel] WinCE port windows.h

2005-08-13 Thread Florian Klaempfl
Oro06 wrote: > 1°)actual (wince 4.2) windows.h need at : > windef.h, types.h, winbase.h, wingdi.h, winuser.h, winreg.h, shellapi.h, > ole2.h, imm.h, tchar.h, excpt.h, strsafe.h > > actual window.pp (from fpc 2.1.X) look at : > base.inc, errors.inc, defines.inc, struct.inc, messages.inc, ascfun.in

Re: [fpc-devel] Constant arrays sometimes produce incorrect data tables

2005-08-13 Thread Anton Tichawa
David Butler wrote: Hi all I picked up this weird code generation problem (it was also present in 2.0.0). I reported it as bug 4277 today. Unfortunately I can only reproduce it in a specific context, so please e-mail me for the unit. The problem is when you define an array of constant strings.

Re: [fpc-devel] Constant arrays sometimes produce incorrect data tables

2005-08-13 Thread David Butler
Hi Anton Thanks for the suggestion. I checked this by changing String to AnsiString, both in the declaration and where the array is accessed, but it still compiles differently in the two units. It seems that where the usual pattern is to store the table as "Ptr, Ptr, Ptr, ..., Value, Value, Valu

[fpc-devel] StrToNetAddr6

2005-08-13 Thread Johann Glaser
Hi! It seems that the StrToNetAddr6 function is not implemented yet (http://svn.freepascal.org/svn/fpc/trunk/rtl/inc/sockets.inc). I wrote an implementation (attached as file including some testcases). I hope you can use it for the Sockets unit. Bye Hansi Program IP6; Function StrToNetAddr6(I

Re: [fpc-devel] According to Valgrind, something is wrong in IndexByte

2005-08-13 Thread Jan Ruzicka
On Aug 13, 2005, at 10:41, Micha Nelissen wrote: On Sat, 13 Aug 2005 00:03:01 -0400 Jan Ruzicka <[EMAIL PROTECTED]> wrote: On Aug 12, 2005, at 05:04, Micha Nelissen wrote: But if I use my own version (which I use for delphi and is i386 specific), it reports no errors in this location, for D