[fpc-devel] Minor error in ppudump from revision 2594
In browsing through the latest changes to the ppudump program, I noticed something that looks to be an error. It causes any object/class with virtual method(s) to also be flagged as being "IsForward". The following patch corrects it Index: ppudump.pp===--- ppudump.pp (revision 3919)+++ ppudump.pp (working copy)@@ -1165,7 +1165,7 @@ const symopts=13; symopt : array[1..symopts] of tsymopt=(- (mask:oo_has_virtual; str:'IsForward'),+ (mask:oo_is_forward; str:'IsForward'), (mask:oo_has_virtual; str:'HasVirtual'), (mask:oo_has_private; str:'HasPrivate'), (mask:oo_has_protected; str:'HasProtected'), ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
[fpc-devel] Problems with WINDOWS define (Revision 2291)
Revision 2291 adds a WINDOWS define for all Win platforms. I'm sure Florian had a very good reason for doing this but it seems to cause some problems for TP-compatible programs which already have some sort of mysterious meaning attached to that define (specifically, unzip, plex and pyacc). The following patch corrects this problem. I see that Peter fixed unzip with revision 2299 but that is a pretty extensive patch. Mine is much smaller and takes a simpler approach. Index: unzip.pp===--- unzip.pp (revision 2293)+++ unzip.pp (working copy)@@ -48,6 +48,9 @@ {$IFDEF FPC} {$DEFINE BIT32}+ {$IFDEF WINDOWS}+ {$UNDEF WINDOWS}+ {$ENDIF} {$ENDIF} {$IFDEF OS2}Index: plex.pas===--- plex.pas (revision 2293)+++ plex.pas (working copy)@@ -64,6 +64,11 @@ {$IFDEF DPMI} {$M 32768} {$ENDIF}+{$IFDEF FPC}+ {$IFDEF WINDOWS}+ {$UNDEF WINDOWS}+ {$ENDIF}+{$ENDIF} {$IFDEF Windows} {$M 32768,0} {$ENDIF}Index: pyacc.pas===--- pyacc.pas (revision 2293)+++ pyacc.pas (working copy)@@ -105,6 +105,11 @@ {$IFDEF DPMI} {$M 32768} {$ENDIF}+{$IFDEF FPC}+ {$IFDEF WINDOWS}+ {$UNDEF WINDOWS}+ {$ENDIF}+{$ENDIF} {$IFDEF Windows} {$M 32768,0} {$ENDIF} Dave ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
Re: [fpc-devel] make cycle fails on win32
Revision 1957 fixes this problem nicely. Thank you, Peter. As always, you're great!! Dave ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
Re: [fpc-devel] trunk/fcl/Makefile.fpc doesn't include db dir
I'm having the same problem and submitted bug report #4605 about it Vincent Snijders [EMAIL PROTECTED] wrote: >Any news about the win32 make cycle problems?>C:/lazarus/source/fpc/2.1/compiler/ppc1.exe -Ur -Xs -OG2p3 -n -Fui386 -Fusystems > -FuC:/lazarus/source/fpc/2.1/rtl/units/i386-win32 -Fii386 -FE. -FUi386/units/i3>86-win32 -dRELEASE -di386 -dGDB -dBROWSERLOG -Fux86 pp.pas>i386\units\i386-win32\pp.o(.text+0x210):pp.pas: undefined reference to `$DBGDWAR>F$_Ld1' ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
[fpc-devel] Re: Errror making fpc: menus.pas(1079, 1) Fatal: Internal error 4321
In a message dated 10/1/2005 5:01:25 A.M. Central Daylight Time, Bram Kuijvenhoven <[EMAIL PROTECTED]> writes: >Hi,>>When I try to make fpc from SVN sources, I get the following error.>>make fv_smart FPC=C:/svn/export/fpc/compiler/ppc386.exe RELEASE=1>make[1]: Entering directory `C:/svn/export/fpc'>make -C fv smart>make[2]: Entering directory `C:/svn/export/fpc/fv'>make all LINKSMART=1 CREATESMART=1>make[3]: Entering directory `C:/svn/export/fpc/fv'>c:/fpc/2.0.0/bin/i386-win32/gmkdir.exe -p units/i386-win32>C:/svn/export/fpc/compiler/ppc386.exe -XX -CX -Ur -Xs -OG2p3 -n ->FuC:/svn/export/fpc/rtl/units/i386-win32 -FE. -FUunits/i386-win32 -gl -di386 ->dRELEASE buildfv.pas>menus.pas(1079,1) Fatal: Internal error 4321>menus.pas(30,8) Fatal: Compilation aborted>make[3]: *** [buildfv.ppu] Error 1>make[3]: Leaving directory `C:/svn/export/fpc/fv'>make[2]: *** [fpc_smart] Error 2>make[2]: Leaving directory `C:/svn/export/fpc/fv'>make[1]: *** [fv_smart] Error 2>make[1]: Leaving directory `C:/svn/export/fpc'>make: *** [build-stamp.i386-win32] Error 2>>Details:> I have fpc 2.0.0 in my path environmental variable.> SVN revision is 1238> Platform: i386-win32 (Win XP SP2)> I export the fpc/trunk SVN repository to an empty directory and then run> make all OPT=-gl>>Apparently, the just built compiler (C:/svn/export/fpc/compiler/ppc386.exe) >doesn't like menus.pas from the FV units, but I have no idea why an 'internal error >4321' occurs.>>Regards,>>Bram I have also had this problem and reported it as bug #4391. Seems to be a problem with Smartlinking and FVision module. Temporary work-around is to edit fpc/makefile on the target fv_smart: MAKE -C fv smart change that to fv_smart: MAKE -C fv all This disables smartlinking for FVision, but then make all works Hope that helps Dave fpc/makefile and ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
Re: [fpc-devel] make install/make zipinstall .svn problem (windows)
From: [EMAIL PROTECTED] (Marco van de Voort)>> When I try to run make install or make zipinstall from my fpc svn repository >>directory, I get the errors copied&pasted below. It seems somehow the .svn dirs >>were copied, and this causes porblems later on with the readonly files that are >>in the .svn directory. (The permission denied error seems to be caused by >>the mentioned files being readonly.)>> >> I'm running Win XP and the fpc that is in the PATH is 2.0.0.>> >> Is it possible to fix the makefiles somehow to circumvent this problem?>This is a known problem, and considered a non-fix because Peter is working>on a replacement system for the makefile system. I workaround it by always>SVN exporting before building. >I haven't tried, but a global "attrib -r /s" in the source dir before>installing might help also. The attrib approach does work, although this sort of defeats the purpose/intent of the svn design. I do have a patch which fixes this in the current makefile system. I have not pursued offering it since I knew the entire process was being revised. I would be happy to provide it. In and of itself, it is not particularly large although it does require that all makefiles be re-generated. If you would like it, please contact me off-list to discuss how you would like me to package it. Dave ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
[fpc-devel] Re: removed MaxAvail,MemAvail,HeapSize
In a message dated 5/23/2005 5:00:24 A.M. Central Daylight Time, Peter Vreman writes: {$ifdef fpc}const memavail=high(ptrint);maxavail=high(ptrint);{$endif} Peter, you are, indeed, a brilliant man!!! There is true beauty in simplicity. ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
[fpc-devel] Problem with install/binw32
It would appear that the set of win32 binaries that Peter installed in the install/binw32 directory of cvs on 2/7 contains a back-level version of a couple programs, as compared to the versions available from the binary download (1.9.6 released 1/1/2005). Specifically, the download version of ld, strip and windres are all "2.15.91 20040904" while the cvs install/binw32 versions of these three are "2.9.5". Using the 2.9.5 version of ld causes the following error when doing a make (linking the ide) libcygwin.a(pseudo-reloc.o)(.text+0x52)undefined reference to __RUNTIME_PSEUDO_RELOC_LIST_END__ libcygwin.a(pseudo-reloc.o)(.text+0x59)undefined reference to __RUNTIME_PSEUDO_RELOC_LIST_END__ ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
Re:[fpc-devel] FPC 1.9.6 (a.k.a. 2.0.0-RC1) is out
In a message dated 1/2/2005 12:31:43 P.M. Central Standard Time, [EMAIL PROTECTED] writes: The Free Pascal Team is pleased to announce that FPC version 1.9.6(or 2.0.0-RC1) is available and ready for download. Thanks, Michael. Downloaded it this morning and it seems to be working nicely. Do have a couple issues a couple questions and a monster KUDO. Will do the issues first: 1) The Win32 installer does not install the utils folder. It's in the download, but doesn't get installed. Easily corrected by just unzipping it manually. 2) There is no demo folder, and hence, can't do the simple check with hello.pp that is referenced in the README 3) The source directories on the download sites are still empty. Are they still coming or am I looking in the wrong place, which leads to my first question: 1) I have been keeping up-to-date at the source level with CVS updates every couple days. Can I just keep building on my current source tree with CVS updates or do I need to do a complete new CVS checkout (please say yes, my network connection stinks!!) 2) It's really nice to have debugger (gdb) support in the IDE again. I get it every time I load a new binary and lose it again every time I install a snapshot compiling the sources. What the heck do I have to do to keep getting the IDE and GDB connected under Windows. I've downloaded every concievable flavor of GDB I can find (the GNU .gz version, Cygwin, MingW) I'm sure I've got every thing I need to keep it working but I'll be darned if I can figure out what I need to do/configure to make it all work together. And now the praise and adulation: Michael et al, your work on the documentation is nothing short of herculean. While not directly related to this release, the Doc-HTML library is great and provides a great context-sensitive help under Lazarus. The new RTL.PDF file that came with this release is OUTSTANDING. For a non-Delphi (and basically non-OOP) oldtimer like me, I think this will become my "Newest Testament". Thanks for all the hard and great work Dave ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
Re: [fpc-devel] Optimization error causes pyacc and plex to fail
In a message dated 1/3/2005 9:02:50 A.M. Central Standard Time, [EMAIL PROTECTED] writes: > I downloaded the new binary release Michael posted this morning. The > versions in that exhibit the same behavior > I re-ran all my tests with the same results. plex quits > way too early and pyacc dies an ugly death if compiled with -OG2 and > work fine when compiled with either -Og2 or -OG1. It's fixed now, thanks!Jonas Just tested with the latest CVS and both are now working great. Thanks for all your help. Dave --- Begin Message --- I downloaded the new binary release Michael posted this morning. The versions in that exhibit the same behavior (think I was already using that as I had done a CVS update and rebuild on 12/31). I see that Marco is getting the same behavior using the H2PAS.Y file under LINUX, while I get it using both my .y and .l files and the pyacc.y file that is part of the TPLY distro under Windows XP, so I don't thinks its an environment thing. In any case, I re-ran all my tests with the same results. plex quits way too early and pyacc dies an ugly death if compiled with -OG2 and work fine when compiled with either -Og2 or -OG1. I have attached my .l and .y files, the logs of the compile/execution and also three folders of .s files (I added -al to the compiles for this round of tests). The OptG2 folder contains the .s files for the non-working version (-OG2). The OptG1 folder has the .s's using -OG1 and the Optlittleg2 folder (Windows case-insensitivity, you know) has the results using -Og2. Hope this helps It's fixed now, thanks! Jonas --- End Message --- ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel
[fpc-devel] Optimization error causes pyacc and plex to fail
Can't seem to get bug submission to work so I'm sending this here. The 1.9.x distributed binaries of plex and pyacc both fail, as well as any created from a 1.9.5 source snapshot. plex ends normally but only creates 2 states. pyacc fails with a RTE 216 on a move in procedure setunion of yaccbase. The makefile compiles these programs with the -OG2p3 option. Changing this to either -Og2p3 or -OG1p3 and they both work correctly. I am using unmodified versions of both programs so am only attaching the output of each scenario. Feel free to contact me if you need anything else. (* plex compiled with -OG2p3 *) >plex -v -o xrlex.l TP Lex Version 4.1a [April 2000], Copyright (c) 1990-2000 Albert Graef parse ... DFA construction ... DFA optimization ... code generation ... DONE 169 lines, 44 rules, 173/1200 p, 2/600 s, 0/1200 t. (see xrlex.lst for more information) (* plex compiled with -Og2p3 or -OG1p3 *) >plex -v -o xrlex.l TP Lex Version 4.1a [April 2000], Copyright (c) 1990-2000 Albert Graef parse ... DFA construction ... DFA optimization ... code generation ... DONE 169 lines, 44 rules, 173/1200 p, 61/600 s, 102/1200 t. (see xrlex.lst for more information) (* pyacc compiled with -OG2p3 *) >pyacc -v xrpars.y TP Yacc Version 4.1a [April 2000], Copyright (c) 1990-2000 Albert Graef parse ... sort ... closures ... Runtime error 216 at $0040CA1C $0040CA1C MOVE, line 76 of F:/FPCDevel/SrcCVS123104/fpc/rtl/i386/i386.inc $00407FA5 SETUNION, line 349 of YaccBase.pas $0040208B CLOSURES, line 94 of YaccClos.pas $00407452 GENERATE_PARSER, line 580 of YaccSem.pas $00409B20 YYACTION, line 250 of pyacc.pas $00409716 YYPARSE, line 2079 of pyacc.pas $0040B5F5 main, line 2479 of pyacc.pas (* pyacc compiled with -Og2p3 or -OG1p3 *) >pyacc -v xrpars.y TP Yacc Version 4.1a [April 2000], Copyright (c) 1990-2000 Albert Graef parse ... sort ... closures ... first sets ... LR0 set ... lookaheads ... code generation ... DONE 1396 lines, 433/900 rules, 691/1200 s, 3334/9600 i, 3138/9600 t, 487/1200 r. 1 shift/reduce conflicts. (see xrpars.lst for more information) ___ fpc-devel maillist - fpc-devel@lists.freepascal.org http://lists.freepascal.org/mailman/listinfo/fpc-devel