February 2003 - Wine-Devel - list.winehq.org

{G|S}etThreadContext question
by Rein Klazes Feb. 5, 2003

Feb. 5, 2003

hi, I am trying to add some support for the floating point unit to winedbg. I have one question to the experts. The debugger code gets/sets the CPU registers through a {G|S}etThreadContext() from the server, adding the flag CONTEXT_FLOATING_POINT should make sure to include the FP registers. This doesn't work because the server code clears the CONTEXT_FLOATING_POINT flag at some point (see attached patch). My patch is simply preventing this, but it may well be that the FP registers should be available at this point. So is this the correct fix? Rein. -- Rein Klazes rklazes(a)xs4all.nl

2 3

CVS Compile error.
by Patrik Båt Feb. 5, 2003

Feb. 5, 2003

Hello people out there! Im trying to compile the cvs source of WINE and gets error: gcc -c -I. -I. -I../../include -I../../include -g -O2 -Wall -mpreferred-stack-boundary=2 -fno-keep-static-consts -D__int8=char -D__int16=short -D__int32=int "-D__int64=long long" -fPIC -D__WINE__ -fshort-wchar -D_REENTRANT -I/usr/X11R6/include -o dpnet_main.o dpnet_main.c cc1: Invalid option `-fshort-wchar' make[2]: *** [dpnet_main.o] Error 1 make[2]: Leaving directory `/root/winecvs/wine/dlls/dpnet' make[1]: *** [dpnet/libdpnet.so] Error 2 make[1]: Leaving directory `/root/winecvs/wine/dlls' make: *** [dlls] Error 2 Is it anyone ho have a clue? I updated the CVS source today, 2003-02-05 (YYYY-MM-DD). ./configure dosen't complaine... My system is: Linux 2.4.20, Debian 3.0 unstable. -- Patrik Båt <patrik(a)researchelectronics.se> Research Electronics AB

4 7

Re: CVS Compile error.
by Patrik Båt Feb. 5, 2003

Feb. 5, 2003

hello, testing som stuff here.... root(a)lightning:~/winecvs/wine# gcc -fshort-wchar test.c cc1: Invalid option `-fshort-wchar' root(a)lightning:~/winecvs/wine# with 2.54 and root(a)lightning:~/winecvs/wine# gcc -fshort-wchar test.c root(a)lightning:~/winecvs/wine# cat test.c #include <stdio.h> int main() { printf("hello!"); } with 3.0 -- Patrik Båt <patrik.bat(a)home.se>

1 0

Opengl Glide
by Kristoffer Ericson Feb. 5, 2003

Feb. 5, 2003

Greetings, I've been trying to get StarSiege Tribes 1 to work on wine, basicly this is what's currently wrong: * As soon as any key is pressed, wine quits = Fix with native dinput.dll version dx4->dx6 * Opengl isn't detected and impossible to choose = Currently no fix, although it works to use glidewrappers. Don't know what Tribes uses to detect opengl. The basic errors i'm getting in the terminal is. fixme:dinput:IDirectInputDevice2AImpl_EnumObjects (!stub) fixme:dsound:IDirectSoundImpl_SetCooperativeLevel err:dinput:SysMouseAImpl_Unacquire Unacquiring a not-acquired device !!! The first fixme, makes me hope that if that is fixed it might get Tribes to work using our own builtin dinput.dll My other question is this, if i want to use glide (got a voodoo3) how do i enable and use this in wine? Do i do a ./configure --enable-glide or --with-glide ? And so on.. I've compiled Mesa with glide support and installed the glidev3 package (using gentoo). Haven't been able to find any info on the homepage or mailing lists, and if needed i could compile a nfo file covering the subject once i understand it better. I would like to point out also that Tribes is working quite well in software mode and since late fall last year i've noticed that many problems has dissapeared. Btw, using daily cvs so got the latest patches. Best wishes Kristoffer _________________________________________________________________ Tired of spam? Get advanced junk mail protection with MSN 8. http://join.msn.com/?page=features/junkmail

2 1

Re: Initialize listview item size.
by Dimitrie O. Paun Feb. 5, 2003

Feb. 5, 2003

On February 3, 2003 11:58 am, Duane Clark wrote: > This is a separate bug from the other listview patches. If an app sends > a LISTVIEW_Paint to a new listview before adding any items (which an app > was :-) the item size was not getting set, causing subsequent > LISTVIEW_Paint calls to clip the painting. > > Changelog: > Make sure item size is initialized when the first item is added. This is an optimization I've put in to avoid a lot of complex calculations while the listview is initially populated (typically before it's being painted). You are right, we have to detect the change nItemCount from 0 to 1, to do all those calculations. But since we don't actually need them until we paint, I'd like to defer them as much as possible. Moreover, with your patch, we don't catch the SetItemCount case. What about this patch instead: Index: dlls/comctl32/listview.c =================================================================== RCS file: /var/cvs/wine/dlls/comctl32/listview.c,v retrieving revision 1.338 diff -u -r1.338 listview.c --- dlls/comctl32/listview.c 24 Jan 2003 00:54:59 -0000 1.338 +++ dlls/comctl32/listview.c 5 Feb 2003 06:48:41 -0000 @@ -7850,7 +7850,7 @@ { TRACE("(hdc=%p)\n", hdc); - if (infoPtr->bFirstPaint) + if (infoPtr->bFirstPaint && infoPtr->nItemCount) { UINT uView = infoPtr->dwStyle & LVS_TYPEMASK; -- Dimi.

2 1

Re: Code Duplication
by Rolf Kalbermatter Feb. 5, 2003

Feb. 5, 2003

Matthew Davison <m.davison(a)virgin.net> wrote: >During my work to clean up the Unicode -> ascii cross-calls I have >noticed some A:W pairs where both are implemented in essentially the >same way. Would it be sensible in these case to make the A function into >a call to the W one, to reduce code size, or would I be wasting my time. I have noticed the same, and looking through it I have for myself come to the conclusion that if the function is rather simple, such as only calling a few A/W specific WinAPIs there is no gain in trying to make the A function call the W function other than an additional possible overhead in execution. Just make sure both functions remain at the same position in the source so that a prospective developer who patches one of the functions can't miss the other one. I for myself handle the rule: If the 'A calls W function with Unicode translation' results in more lines of source code than the direct A implementation I won't spend time to try to convert them. If the function however is a little more complex there is always the issue of maintaining two different functions in parallel and then making the A function call the W function is the most sensible approach indeed. Rolf Kalbermatter

1 0

wine ICMP code
by Jason Algol Feb. 5, 2003

Feb. 5, 2003

hi guys, in dlls/icmp/icmp_main.c would it be possible to use system("/bin/ping..."); if not running as uid 0?, (or using pipes if the ouput needs to be parsed) as /bin/ping is obviously usually suid root and if the required options are available as arguments to /bin/ping then wine can do as requested without running as root, this would be useful to me with something im working on. if you guys do consider it a possibility, i would be willing to research it and write the patch, id just like the go-ahead from you :) if you can think of any other way round it, id be greatful to hear your suggestions. _________________________________________________________________ Worried what your kids see online? Protect them better with MSN 8 http://join.msn.com/?page=features/parental&pgmarket=en-gb&XAPID=186&DI=1059

3 2

Re: [Wine-d3d] [dx8]: warcraft 3
by Christian Costa Feb. 5, 2003

Feb. 5, 2003

Raphaël Junqueira wrote: >-----BEGIN PGP SIGNED MESSAGE----- >Hash: SHA1 > >Le Samedi 25 Janvier 2003 15:07, Raphaël Junqueira a écrit : > >>Hi, >> >> i'm trying to get warcraft3 displaying something and looking at log i >>didn't understand why i can see anything (all rendering traces are ok). >>But when i used: >>wine --debugmsg +d3d,+opengl,+x11drv >>/mnt/win_c2/Jeux/Warcraft3/War3Demo.exe 2> /usr/src/wine/logs/war3.log >> >>i have found X11DRV interesting traces: >><> >>trace:x11drv:X11DRV_CreateWindow win 0x10021 window 0,0,0,0 client 0,0,0,0 >>whole 0,0,0,0 X client 0,0,0,0 xwin 2600001/2600002 >>trace:x11drv:X11DRV_ShowWindow hwnd=0x10021, cmd=1 >>trace:x11drv:X11DRV_SetWindowPos hwnd 0x10021, swp (0,0)-(0,0) flags >>00000063 trace:x11drv:X11DRV_SetWindowPos current (0,0)-(0,0), style >>86080000 warn:x11drv:SWP_DoOwnedPopups (0x10021) hInsertAfter = (nil) >>trace:x11drv:SWP_DoNCCalcSize (0,0)-(0,0) >>trace:x11drv:X11DRV_sync_whole_window_position setting win 2600001 pos >>0,0,0x0 after 400c7db4 changes=40 >>trace:x11drv:set_visible_style hwnd 0x10021 (2600001) set 1 visible 0 empty >>1 trace:x11drv:X11DRV_SetWindowPos hwnd 0x10021, swp (0,0)-(0,0) flags >>00000013 trace:x11drv:X11DRV_SetWindowPos current (0,0)-(0,0), style >>96080000 warn:x11drv:SWP_DoOwnedPopups (0x10021) hInsertAfter = (nil) >>trace:x11drv:X11DRV_sync_whole_window_position setting win 2600001 pos >>0,0,0x0 after 400c7db4 changes=40 >>trace:x11drv:X11DRV_SetWindowPos status flags = 1803 >>trace:x11drv:X11DRV_SetWindowPos status flags = 1863 >><> >> >>it seems that the created window have a (0,0) size !!!! >> > >i'm alone here ????? > Don't cry... You're not alone! ;-) > >i don't understand why this stupid game create an empty window > Turning x11drv and win debug channels gives the following log. trace:win:WIN_CreateWindowEx "Warcraft III" "Warcraft III" ex=00040008 style=86080000 -2147483648,-2147483648 -2147483648x-2147483648 parent=(nil) menu=(nil) inst=0x400000 params=0x44f500c0 trace:win:dump_window_styles style: WS_POPUP WS_CLIPSIBLINGS WS_CLIPCHILDREN WS_SYSMENU trace:win:dump_window_styles exstyle: WS_EX_TOPMOST WS_EX_APPWINDOW trace:win:WIN_CreateWindowEx winproc type is 2 (WIN_PROC_32A) trace:win:WIN_SetRectangles win 0x10021 window (0,0)-(0,0) client (0,0)-(0,0) trace:x11drv:X11DRV_CreateWindow hwnd 0x10021 cs 0,0 0x0 trace:x11drv:X11DRV_SetWindowText text_cp = 0 trace:win:WIN_SetRectangles win 0x10021 window (0,0)-(0,0) client (0,0)-(0,0) trace:x11drv:X11DRV_CreateWindow win 0x10021 window 0,0,0,0 client 0,0,0,0 whole 0,0,0,0 X client 0,0,0,0 xwin 2c00001/2c00002 trace:win:WIN_SetWindowLong 0x10021 -21 44f500c0 2 trace:win:WIN_CreateWindowEx created window 0x10021 trace:x11drv:X11DRV_ShowWindow hwnd=0x10021, cmd=1 trace:win:SetWindowPos hwnd 0x10021, after (nil), 0,0 (0x0), flags 00000063 trace:win:dump_winpos_flags flags: SWP_NOSIZE SWP_NOMOVE SWP_FRAMECHANGED SWP_SHOWWINDOW trace:x11drv:X11DRV_SetWindowPos hwnd 0x10021, swp (0,0)-(0,0) flags 00000063 trace:x11drv:X11DRV_SetWindowPos current (0,0)-(0,0), style 86080000 warn:x11drv:SWP_DoOwnedPopups (0x10021) hInsertAfter = (nil) The log shows that the game does not create a windows with a (0,0) but a pop-up where xpos = ypos = width = height = CW_USEDEFAULT (0x8000000 or in decimal -2147483648). For pop-up and childs windows if witdh = CW_USEDEFAULT, width and height are set to 0. For overlapped windows, the width and height are set to make the lower-right corner of the windows corresponds to the lower_right corner of the screen. So the question is why does the game create a pop-up instead of a standard windows. That would be cool if we can solve this issue. Forcing the screen size make the game demo displays something like menus and so on. It's encouraging. :-) > > >helppppppppp > >>see attached log >> >>Raphael >> Christian

2 1

Trying to Port Over Adobe Photoshop
by aaronvaldes＠rogers.com Feb. 5, 2003

Feb. 5, 2003

Hello I am trying to port over adobe Photoshop. When I execute: $ ./configure --with-wine=/usr/local/wine I get the error: ng whether we are using the GNU C compiler... yes checking whether gcc accepts -g... yes checking for g++... g++ checking whether we are using the GNU C++ compiler... yes checking whether g++ accepts -g... yes checking how to run the C preprocessor... gcc -E checking whether ln -s works... yes checking for sqrt in -lm... yes checking for iswalnum in -lw... no checking for gethostbyname... yes checking for connect... yes checking "for gcc strength-reduce bug"... no checking "whether external symbols need an underscore prefix"... no checking "whether we can build a Linux dll"... yes checking "whether we need to define __i386__"... no checking "for g++ -fpermissive option"... yes checking "for g++ -fno-for-scope option"... yes checking for windef.h... configure: error: Could not find the Wine headers (windef.h) I would appreciate any help. I apologize if I dont have this formatted correctly. Aaron Valdes

3 2

Building crosstest ...
by Paul Millar Feb. 4, 2003

Feb. 4, 2003

Hi everyone, Apologies for the rambling email, but here's a bit of a brain-dump. I don't know how much of this people already know about ... After banging my head against a wall for the past week or so, I've put together a script that builds a cross-compiling mingw. The cross-compiler should be suitable for building .EXE conformance test executables and is available from: http://www.astro.gla.ac.uk/users/paulm/Scripts/build-mingw-cross.sh Have a look at the top of the script to see how to get things working. (yes, I know /. has just done this, but their HOWTO is old, circa '99) That should be enough to get mingw for cross-compiling under wine, but there's still problems with the crosstest target for me. Using the current CVS, crosstest depends on winebuild. But the cross-compilation doesn't actually use winebuild. Worse, it fails whilst trying to building winebuild (missing libraries), if you haven't already built wine. I guess winebuild is missing some dependencies. A simple patch: --- Index: dlls/Makefile.in =================================================================== RCS file: /home/wine/wine/dlls/Makefile.in,v retrieving revision 1.168 diff -u -r1.168 Makefile.in --- dlls/Makefile.in 24 Jan 2003 01:08:15 -0000 1.168 +++ dlls/Makefile.in 4 Feb 2003 00:15:22 -0000 @@ -1181,7 +1181,6 @@ x11drv/x11drv.spec.def: $(WINEBUILD) $(SUBDIRS): $(IMPORT_LIBS:%=%.$(IMPLIBEXT)) -$(SUBDIRS:%=%/__crosstest__): $(IMPORT_LIBS:%=%.a) # Map library name to the corresponding directory --- fixes this, although I'm not 100% its correct. It stops the .a libraries from being built, but are they needed for crosstest? After this, it builds the advapi32 tests ok, but there's problems parsing the files include/winsock{,2}.h when it tries to build the dsound tests... Just thought people would be interested. ---- Paul Millar

2 1