Re: [fpc-pascal] Sharp Zaurus ARM port stability?
Thank you Den. I'll do some homework and keep you posted of my findings and progress. Don't expect much progress in the next 3 weeks, I'm on holiday :) Regards, Alain Le 22 juil. 05, à 00:24, Den Jean a écrit : On Thursday 21 July 2005 06:53 pm, Alain Vitry wrote: Den, Could you point me to some reading/starting point on how to make those bindings ? Swig doesn't support pascal yet. I have a fresh QT4 bindings for pascal in mind. Thanks There are several ways: kalyptus: see cvs/fpc/utils/kalyptus google on kde bindings / qt bindings Florian started from the c-kalyptus one. http://www.lazarus.freepascal.org/list_archives/lazarus/08-17-02_01 -02-04/msg04138.html smoke: more for scripting languages smoke/refract: c++ headers to xml using the gcc parsing capability. C++ may look to be the perfect parser for c++, but qt is not c++ but an enhanced delphi like language (hence why the moc) so distance to qt language is big. http://wiki.dotgnu.info/Smoke/Refract freeclx: based on moc parsing. Moc converts qt language to c++. So Moc is close to the qt language and it gives you a very able parser. With some modifications to the lex/yacc you can let moc produce a pascal qt binding. But each time trolltech adds to the qt language, they renew the moc, then you need to port your adaptations to the moc lex/yacc again. We adapted the qt part of freeclx from qt 2.2.4 to qt 2.3.2 and qt/embedded The result is on http://cvs.sourceforge.net/viewcvs.py/qtforfpc/qt2/FreeClx/ qtc binding: as pascal can easily use c-libraries, the qtc binding seems a good base but difficult to maintain at each update of the qtc binding. Andreas Hausladen is very productive in this matter. (allready qt3 binding), I guess qt4 is in the make. Perhaps you can join forces with him. http://andy.jgknet.de/oss/qt/qt3forFPC/ http://andy.jgknet.de/oss/kylix/wiki/index.php/Qt3_for_Kylix kind regards, Den Jean ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
[fpc-pascal] MSG variable type, Windows GUI
Hi, I encountered an FPC Win32, tutorial that uses the a declaration of variable "msg", declared as a MSG type, shown below. Dev-Pas IDE, FPC 2.00 gives an "error in type definition" and I find no doc references to MSG, variable type. Was this an earlier convention in FPC and if so, how was it replaced? Thanks, Mike [clip] program Win32_Source03; uses windows; const AppName = 'Win32/ FPC - Source 03'; VAR active : BOOLEAN; msg : MSG; hWindow,msg : hwnd; [end clip] ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] MSG variable type, Windows GUI
Urbansound wrote: Hi, I encountered an FPC Win32, tutorial that uses the a declaration of variable "msg", declared as a MSG type, shown below. Dev-Pas IDE, FPC 2.00 gives an "error in type definition" and I find no doc references to MSG, variable type. Was this an earlier convention in FPC and if so, how was it replaced? Previous version of fpc allowed this construct while newer not use Message:msg instead of msg:msg and rename msg in the line below of that. [clip] program Win32_Source03; uses windows; const AppName = 'Win32/ FPC - Source 03'; VAR active : BOOLEAN; msg : MSG; hWindow,msg : hwnd; [end clip] active : BOOLEAN; message : MSG; hWindow,msg1 : hwnd; this should work Luiz ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] fast integer multiplication
Vincent Snijders wrote: > Hi, > > Suppose I have the following code: > > var > a,b: dword; > c: qword; > > begin > a := 1000; > b := 2000; > c := a * b; > writeln(c); > end. > > Now, although c is large enough to contain the result only the lower > dword is filled. I can force correct results by using c := qword(a) * b, > but the slow fpc_mul_qword is used. > > The code generated for the above sample is: > # [16] c:=a*b; > movlU_P$PROJECT1_A,%edx > movlU_P$PROJECT1_B,%eax > mull%edx > movl$0,%edx > movl%eax,U_P$PROJECT1_C > movl%edx,U_P$PROJECT1_C+4 > > What I want is the above code, but without the "movl $0,%edx" > instruction. Is there a way to do this (wihtout using fpc_mul_qword). Only assembler for now. Any suggestions how the compiler could be told to generate such code? ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
Re: [fpc-pascal] profiler
That is really a great tool. It shows even how is the assembler mapped to the source code. I was surprised when I learned that 90% of time spent in my function goes to a result assignment stalling the processor. the usual function f : TStructure var b: TStructure; begin ... some processing on b ... f := b; end; Jan On Jul 20, 2005, at 04:04, Jonas Maebe wrote: On 05 Jul 2005, at 06:36, Jan Ruzicka wrote: Was the profiling tested under OS X? No, but I would urge you to use Shark from CHUD (http://developer.apple.com/tools/performance/overview.html) It's much better and versatile than gprof, and only requires regular debugging information to work. Jonas ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal ___ fpc-pascal maillist - fpc-pascal@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-pascal
