February 2007 - Wine-Devel - list.winehq.org

RE: unicoding xcopy - question on wcschr
by Ann & Jason Edmeades Feb. 23, 2007

Feb. 23, 2007

For reference: 'obviously'... adding this line to the makefile solves it MODCFLAGS = -fno-builtin (Copied from msvcrt\tests) Jason

1 0

unicoding xcopy - question on wcschr
by Ann & Jason Edmeades Feb. 23, 2007

Feb. 23, 2007

Hi, Trying to convert my xcopy to Unicode proved simple on windows, but when putting into wine has proved more interesting. Most of the wide string calls are like lstrcmpW and they all work fine, but I really need the Unicode equivalent to strchr and strrchr. As far as I can tell my only option is wcschr. If I build as-is, I get errors saying wcschr is implicitly defined but the link etc is all ok (from a url http://fgouget.free.fr/wine/PrgApps/Chap7.shtml#filerev it implies this is because I am using the GNU C headers, and they don't define it in string.h). Grepping the source, I can only find a hit inside the msvcrt/string.h or wchar.h. If I add to the makefile EXTRAINCL = -I$(TOPSRCDIR)/include/msvcrt, to ensure I pick up the wine headers, I now get warnings like: ./../include/msvcrt/ctype.h:72: warning: conflicting types for built-in function 'iswalnum' How do I #include / link a program using this function? It all just worked under windows :-) Jason

1 0

Progress in implementation of Direct3D
by Mirek Feb. 23, 2007

Feb. 23, 2007

Hi, i would like to say "Thanks for your work!" to all wine developers. In past months there was realy big progress in implementation of Direct3D in wine (that part is most important for me) and other parts such Dinput, so I would like to introduce you some more sreenshots from latest Direct3D applications. I will try to post them on official wine website. To run third benchmark in 3DMark 2005 I used wined3d_3dmark05.diff.txt patch (from H. Verbeet) To run Call of Dutty 2 I used cod2.diff patch (my latest modification of patch against current CVS) All screenshots are from today CVS 20.02.2007 without patch "[9/10] WineD3D: Use VBOs for index buffers" which caused some regressions. http://headline.czela.net/Mirek/wine/ or via IP: http://62.240.181.87/Mirek/wine/ --- dlls/wined3d/directx.c | 7 ++++--- 1 files changed, 4 insertions(+), 3 deletions(-) diff --git a/dlls/wined3d/directx.c b/dlls/wined3d/directx.c index 841e16f..ca3eaaf 100644 --- a/dlls/wined3d/directx.c +++ b/dlls/wined3d/directx.c @@ -585,7 +585,7 @@ BOOL IWineD3DImpl_FillGLCaps(IWineD3D *iface, Display* display) { glGetIntegerv(GL_MAX_TEXTURE_UNITS_ARB, &gl_max); TRACE_(d3d_caps)(" FOUND: ARB Multitexture support - GL_MAX_TEXTURE_UNITS_ARB=%u\n", gl_max); gl_info->supported[ARB_MULTITEXTURE] = TRUE; - gl_info->max_textures = min(MAX_TEXTURES, gl_max); + gl_info->max_textures = 8; gl_info->max_texture_stages = min(MAX_TEXTURES, gl_max); gl_info->max_samplers = max(gl_info->max_samplers, gl_max); } else if (strcmp(ThisExtn, "GL_ARB_texture_cube_map") == 0) { @@ -2196,7 +2196,7 @@ static HRESULT WINAPI IWineD3DImpl_GetDeviceCaps(IWineD3D *iface, UINT Adapter, *pCaps->MaxVertexBlendMatrixIndex = 1; *pCaps->MaxAnisotropy = GL_LIMITS(anisotropy); - *pCaps->MaxPointSize = GL_LIMITS(pointsize); + *pCaps->MaxPointSize = 64.0f; *pCaps->VertexProcessingCaps = WINED3DVTXPCAPS_DIRECTIONALLIGHTS | @@ -2294,7 +2294,8 @@ static HRESULT WINAPI IWineD3DImpl_GetDeviceCaps(IWineD3D *iface, UINT Adapter, *pCaps->NumSimultaneousRTs = GL_LIMITS(buffers); - *pCaps->StretchRectFilterCaps = 0; + *pCaps->StretchRectFilterCaps = WINED3DPTFILTERCAPS_MINFLINEAR | + WINED3DPTFILTERCAPS_MAGFLINEAR; *pCaps->VertexTextureFilterCaps = 0; if(*pCaps->VertexShaderVersion == WINED3DVS_VERSION(3,0)) {

3 2

0.9.31 not suitable for gaming
by Pavel Troller Feb. 23, 2007

Feb. 23, 2007

Hi! I've tried 0.9.31 and it seems that there is a regression in at least two games: - Fallout Tactics (BOS.exe): Shows a dialog stating "C:\dev\phoenix\display\win32\win32_window.cpp(873): **fatal error**: Could not create display buffers" and then another one about abnormal termination of the program and quits. - Steam based games, mainly HL2: During game load, the screen goes black. Moving the cursor causes sound effects as it hovers across the menus and buttons, so the program runs, but nothing can be seen. Both those games are known to work with wine at about a half between .29 and .30. I tried it with basic .31 as well as with today's git update. Are those regressions known and is it a work in progress, or should I make a bisecting session to find the problem ? With regards, Pavel Troller

5 19

Re: How complete will Wine ever be for older apps?
by Juan Lang Feb. 23, 2007

Feb. 23, 2007

> I'm just trying to build up a picture in my head of whether there > will come a day when Wine users can pick an app designed for a > previous generation of Windows and expect it to work flawlessly > (ignoring things that try to manipulate the hardware directly, I > suppose)? Or will there always be incompatibilities? There will (likely) always be incompatibilities. Wine development has always been driven by app compatibility - someone picks an app and tries to run it, and tries to fix problems they come across. We don't aim for feature completeness - Windows is far, far too big to replicate with the resources we have. So, some apps (the ones developers have worked on, and likely others) will work quite well, while others (those that haven't been worked on) may not work as well. --Juan ____________________________________________________________________________________ TV dinner still cooling? Check out "Tonight's Picks" on Yahoo! TV. http://tv.yahoo.com/

2 1

RE: a big fan (now with bad news)
by Jeffrey Sabarese Feb. 23, 2007

Feb. 23, 2007

hello there! thanks to all of you who replied to my inquiry about "ies4linux" vs compatibility w/ such "Windows-native" apps as Bradbury "TopStyle", HomeSite, etc... As Joris stated, IE6 sp1 is available at the MS site. I downloaded my own copy, but i haven't tried it yet because i think i "ruined" my WinE this evening. HOW? as my UserID (NOT as root), i "re installed" the [.reg] files in the ies4linux "faux directories" (as in a command such as :: [ wine '../win.reg' ]) . I tried [ yum remove wine ], and then [yum install wine] which went without a hitch, but there's no change in my Primary User (the broken one) HOWEVER, WinE continues to run perfectly under other user accounts, and ROOT. My QUESTION: How do i "restore" my WinE to a working state for my UserID? (because only UserID is affected) I wish to not destroy my Program Files while "repairing" WinE around it--- if possible. >From the primary UserID only, i can NOT launch [ winecfg.exe ] for example in terminal, or anything WinE that i can tell. ([while root, and other users were apparently not affected]) please help me to restore my "existing" Program files setup, if possible. it is no "tragedy" if i can not, but i'd like to try. in any event-- i do need to know how to "restore" it, maintaining program files or not. thank you!! best regards, -JS. On 2/21/07, SorinN <nemes.sorin(a)gmail.com> wrote: > Indeed - ies4linux - don't use Wine folders - on other words - it use > winelib but is on other place, where TopStyle and other windows > programs don't know to look. I will try to talk with it's developer to > suggest him to use an existing Wine installation ( is Wine exist > already ). > > > > 2007/2/19, Jeffrey Sabarese <core5fedora(a)gmail.com>: > > hi there! I have to wonder if i don't use WinE more than the "average" > > Fedora Core 5 user. I am forever trying to figure out how to run my > > old Windows apps here in the linux environment, using your incredible > > software. > > -- /* ************************************************ */ Quanta+ Web Development IDE: http://quanta.kdewebdev.org/ Quanta+ Wiki: http://tinyurl.com/3dzqxk Other Quanta Tips: http://invisiblepixels.org/progs/quanta.html /* ************************************************ */

2 1

Re: winex11.drv: Set the refresh rate for the virtual desktop to 60
by H. Verbeet Feb. 23, 2007

Feb. 23, 2007

On 23/02/07, Jan Zerebecki <jan.wine(a)zerebecki.de> wrote: > From: Jan Zerebecki <jan.wine(a)zerebecki.de> > Changelog: > winex11.drv: Set the refresh rate for the virtual desktop to 60 > instead of 0. If HKCU\Software\Wine\X11 Driver\FakeRefreshRate is > set, it is interpreted as a comma delimited list of refresh rates > to use instead. > --- Perhaps there should be a winecfg option for this, together with the one that already exists for the resolution.

4 5

Re: xcopy support try 1...
by Detlef Riekenberg Feb. 23, 2007

Feb. 23, 2007

On Do, 2007-02-22 at 21:52 +0000, Ann & Jason Edmeades wrote: > Firstly, I am not sure I have included everything correctly to define a new > directory in the build tree, With git, you only need to add all the new files to the Index ("git-update-index --add programs/xcopy/Makefile.in" as Example) and only that changes should be send. tools/make_makefiles does the rest, so you compile and test your new app. > I have only implemented the basics, Converting to unicode later is much harder. I suggest to use UNICODE from the begining. An example UNICODE console app is "programs/msiexec/*" When you really want to have also an ansi-version (for DOS or win9x as example), you can use TCHAR and test your app with and without "#define UNICODE". UNICODE is prefered and you avoid converting a lot of data, when the app is running. > which is file and subdirectory > copying plus the simple reporting information. The Patch would be even smaller, when you do no copy at all or just do a Filecopy without recursion / Parameter-Parsing (wildcards ) > + * According to http://www.microsoft.com/technet/archive/msdos/comm8.mspx?mfr=true Please remove that ref. Many URLs are a dead link really fast. -- By by ... Detlef

2 1

Re: ntdll: implement NtAllocateLocallyUniqueId with server call
by Robert Shearman Feb. 22, 2007

Feb. 22, 2007

Juan Lang wrote: > This introduces a warning in trace.c about dump_luid being undeclared. > trace.c seems to be at least partially generated, so I wasn't sure about > sending a patch for it. Let me know how I should handle that and I'll fix > it up. > Hi Juan, dump_luid needs to be in trace.c. As you say, it is partially generated: > /* Everything below this line is generated automatically by > tools/make_requests > */ > /* ### make_requests begin ### */ So just put dump_luid above that line and it'll be fine. -- Rob Shearman

1 0

Re: wine.inf: Update the timezone information.
by Uwe Bonnes Feb. 22, 2007

Feb. 22, 2007

>>>>> "Francois" == Francois Gouget <fgouget(a)codeweavers.com> writes: Francois> --- I have used two sources to update Wine's timezone data: ... Francois> The display name was set to the Olson database ids and I have Francois> kept it that way as it seems to match what is displayed on Francois> Linux. Wherever we used old names I switched to the new names Francois> as specified in the Olson database. Francois> We still have a few differences with the Windows data and in a Francois> few instances I diverged from the CLDR mapping to tzids. I Francois> will now summarize all these differences in bug 7295. How do internaltional version of Windows get the localized names of the pletora of timezones? -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

2 1

WineConf 2007 Fundraiser proposal
by Tom Wickline Feb. 22, 2007

Feb. 22, 2007

Hello Everyone, On February 14th I made a suggestion on the wine-conf mailing list that we should have a annual Fund Raiser to help raise funds for our annual Conference. http://www.winehq.org/pipermail/wineconf/2007-February/000397.html Jeremy replied with his thoughts/suggestions and some back ground info on how the WPF funds have helped in the past. http://www.winehq.org/pipermail/wineconf/2007-February/000403.html >From this, I cobbled together a extremely rough Donations page.. Keep in mind this is still very much a work in progress and the language within it needs major revisions. http://www.winehq.org/pipermail/wineconf/2007-February/000425.html The current discussion is now on whether or not a $1,000 donation is sufficient for a front page listing for one year since our search rank in so high. Dan suggested that we join adsence, and I replied that I had already suggested this in the past. http://www.winehq.org/pipermail/wineconf/2007-February/000437.html At this point there are many undecided questions about the original suggestion of us having a fund raiser, as well as if we should join adsence. If you have a opinion on this subject we would like (even suggest) that you to air them at this time before decisions are made. http://www.winehq.org/pipermail/wineconf/2007-February/000440.html Questions: Should we have a annual fundraiser to help with conference expenses? Thus far the consciences is Yes. What is a realistic goal? I believe $10,000 is very doable. What amount should we receive for a front page listing 'gold sponsor'? Is $10,000 to much or to small of a amount? If we decide on $10,000 the goal should be revised to about $25,000 Should the Wine project join the adsence network? With our high google page rank we are practically guaranteed to rake in a mint :D Comments and suggestions are at this time highly suggested! Tom

2 2

How complete will Wine ever be for older apps?
by Jules Richardson Feb. 22, 2007

Feb. 22, 2007

Sort of an open-ended question to developers :) It's obvious that Wine functionality will always lag behind Windows releases somewhat. But how accurate / complete do people think it will eventually be against older Windows releases? Is much development work even still being done on things that don't quite work as expected within apps that were current at the time of say NT4? I'm just trying to build up a picture in my head of whether there will come a day when Wine users can pick an app designed for a previous generation of Windows and expect it to work flawlessly (ignoring things that try to manipulate the hardware directly, I suppose)? Or will there always be incompatibilities? cheers Jules

1 0

Re: [DSOUND] Implement a locking mechanism
by M.B. Lankhorst Feb. 22, 2007

Feb. 22, 2007

Robert Reif wrote: > You shouldn't need to make any changes to direct sound. You just need > to remove the flag DSDDESC_DONTNEEDPRIMARYLOCK from alsa and make the > necessary changes in alsa. The direct sound hardware primary buffer > locking support is already in primary.c. You are right, partially, primary.c already has support for it, but that is only used when the application chooses to write to the primary buffer directly, in most cases the primary buffer is only used directly by the directsound mixer, and that's the case this patch is trying to handle. However the problem I have right now is that I miss some magic values so sound is a bit cumbersome, maybe someone with knowledge of dsound knows what I should put in there and be able to help me out a little. Maarten PS: CC to people you address in your emails, not everyone receives messages from wine-devel mailing lists.

1 0

message sequence testing - message flags
by Lei Zhang Feb. 22, 2007

Feb. 22, 2007

Hi, In dlls/user32/tests/msg.c, there's the following enum: typedef enum { sent=0x1, posted=0x2, parent=0x4, wparam=0x8, lparam=0x10, defwinproc=0x20, beginpaint=0x40, optional=0x80, hook=0x100, winevent_hook=0x200 } msg_flags_t; Can someone explain what these flags mean? parent defwinproc beginpaint hook winevent_hook - Lei

2 1

Re: [DSOUND] Implement a locking mechanism
by Robert Reif Feb. 22, 2007

Feb. 22, 2007

Maarten Lankhorst wrote: >[DSOUND] Implement a locking mechanism > >Attempt to implement locking in dsounds mixer. > >Sadly horribly broken, I'm posting this in the hope someone knows what >magic values to fill in at the first "if (lock)" .. I really have no clue. > >As stated before it is required to be used with previous winealsa patch. > >Right now it also calls Unlock with lockamount as argument, this should >probably be changed to how much is actually written to (frag), but that >part doesn't really matter much I think, it's more important to get the >first Lock() right. > > You shouldn't need to make any changes to direct sound. You just need to remove the flag DSDDESC_DONTNEEDPRIMARYLOCK from alsa and make the necessary changes in alsa. The direct sound hardware primary buffer locking support is already in primary.c.

1 0

Re: [PATCH 2/3] server: Fix get_handle for global handles. It doesn't need process for them.
by Alexandre Julliard Feb. 21, 2007

Feb. 21, 2007

Vitaliy Margolen <wine-patches(a)kievinfo.com> writes: > Resend as-is. I need this function to work on global handles when > process is NULL. If you need this you are doing something wrong. Callers are not supposed to know that a handle is global. -- Alexandre Julliard julliard(a)winehq.org

2 6

Re: resend : gdi32:WidenPath partial implementation
by Alexandre Julliard Feb. 21, 2007

Feb. 21, 2007

Laurent Vromman <laurent(a)vromman.org> writes: > + > + dc = DC_GetDCPtr( hdc ); You shouldn't need that, the caller has done it already. Simply pass a DC* to this function. > + if(pPath->state != PATH_Closed) { > + ERR("Path Closed\n"); > + return FALSE; > + } > + > + hPen = SelectObject(hdc, GetStockObject(DC_PEN)); You shouldn't select a new pen into the DC, just use the current one. > + case PT_BEZIERTO: > + // should never happen because of the FlattenPath call Please avoid C++ comments. > + default: > + ERR("Got path flag %d\n", (INT)pPath->pFlags[i]); Please avoid casts like this, use the correct printf format instead. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [10/13] D3D9: Fix the circular converted vertex declaration reference
by H. Verbeet Feb. 21, 2007

Feb. 21, 2007

On 20/02/07, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > If anyone knows a nicer way to release the converted declaration from > SetVertexDeclaration feel free to comment. > > Not applying this patch won't affect patches 11-13 > Wouldn't it be better to directly compare This against convertedDecl from the device in IDirect3DVertexDeclaration9Impl_Release, rather than adding a convertedDecl flag to IDirect3DVertexDeclaration9Impl and setting / unsetting it? I think it makes things harder than it has to be. Also, afaik Ivan has some tests for the exact behaviour.

2 1

Working on JScript/VBScript support for MSI
by Misha Koshelev Feb. 20, 2007

Feb. 20, 2007

Hi, since MSDN states that any installer that requires JScript/VBScript must make sure to install it, and since some applications require these for install (bug #7353 and another one is there for VBScript for a different installer), I figure there is no reason why builtin MSI cannot use Microsoft Script if it is there (and plus when we have our own JScript/VBScript we will be able to use it off the bat with MSI). I have posted my initial work on this on the bugzilla for bug #7353. The first file (a patch, will attach also) is my patch which updates custom.c to have the appropriate handlers for the custom action types, extracts the script from the appropriate place, writes it to a file, and uses a framework very similar to the current thread framework for DLLs to call my helper call_script function. This first patch, in theory, should not have to change much (if at all), although I have not thoroughly tested it yet so it probably will. The second file (and attached here too) is dlls/msi/script.c, which is where the meat and potatoes of the actual call_script function is/will be. Right now it is pretty barebones, but it does load the script from the file and calls it appropriately. Now I am working on implementing the Session object, although it looks like a wrapper around MSI functions we already have so I think it should be doable. Anyways, any comments on my work so far will be appreciated. Misha

2 3

Re: [9/10] WineD3D: Use VBOs for index buffers
by Mirek Feb. 20, 2007

Feb. 20, 2007

Hi, this patch caused regression in many Direct3D games. TES IV: Oblivion can't start game (menu works), Flatout II is completly broken, Alpine Sky Racing 2007 can't run, HalfLife 2 EO can't run. I tried rewrite GLint vbo; to GLuint vbo; but it doesn't help. Mirek Stefan Dösinger napsal(a): > > ------------------------------------------------------------------------ > > From 4d52be700666b6b1e853bad296c502b5b0d76189 Mon Sep 17 00:00:00 2001 > From: Stefan Doesinger <stefan(a)codeweavers.com> > Date: Mon, 19 Feb 2007 15:12:39 +0100 > Subject: [PATCH] WineD3D: Use VBOs for index buffers > > This patch uses (if available) gl vertex buffer objects for static d3d index buffers > to store indices in video memory. Therefore a new index buffer state is added which > binds the vbo to GL_ELEMENT_ARRAY_BUFFER. DrawStrided* use the index buffer stride to > tell if indices are used for drawing because a NULL index data pointer is valid now. > DrawStridedSlow has to read the indices from the index buffer if it is used, because > it can't use the data from the vbo(or only very inefficiently). > --- > dlls/wined3d/device.c | 68 ++++++++++++++++++++++++++++++++++++--- > dlls/wined3d/drawprim.c | 14 ++++++-- > dlls/wined3d/indexbuffer.c | 49 +++++++++++++++++++++++++++- > dlls/wined3d/state.c | 12 +++++++ > dlls/wined3d/wined3d_private.h | 8 ++++- > 5 files changed, 139 insertions(+), 12 deletions(-) > > diff --git a/dlls/wined3d/device.c b/dlls/wined3d/device.c > index 18f3e75..af0e96a 100644 > --- a/dlls/wined3d/device.c > +++ b/dlls/wined3d/device.c > @@ -443,21 +443,67 @@ static HRESULT WINAPI IWineD3DDeviceImpl_CreateVertexBuffer(IWineD3DDevice *ifac > return WINED3D_OK; > } > > +static void CreateIndexBufferVBO(IWineD3DDeviceImpl *This, IWineD3DIndexBufferImpl *object) { > + GLenum error, glUsage; > + TRACE("Creating VBO for Index Buffer %p\n", object); > + > + ENTER_GL(); > + while(glGetError()); > + > + GL_EXTCALL(glGenBuffersARB(1, &object->vbo)); > + error = glGetError(); > + if(error != GL_NO_ERROR || object->vbo == 0) { > + ERR("Creating a vbo failed, continueing without vbo for this buffer\n"); > + object->vbo = 0; > + return; > + } > + > + GL_EXTCALL(glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, object->vbo)); > + error = glGetError(); > + if(error != GL_NO_ERROR) { > + ERR("Failed to bind index buffer, continueing without vbo for this buffer\n"); > + goto out; > + } > + > + /* Use static write only usage for now. Dynamic index buffers stay in sysmem, and due to the sysmem > + * copy no readback will be needed > + */ > + glUsage = GL_STATIC_DRAW; > + GL_EXTCALL(glBufferDataARB(GL_ELEMENT_ARRAY_BUFFER_ARB, object->resource.size, NULL, glUsage)); > + error = glGetError(); > + if(error != GL_NO_ERROR) { > + ERR("Failed to initialize the index buffer\n"); > + goto out; > + } > + LEAVE_GL(); > + TRACE("Successfully created vbo %d for index buffer %p\n", object->vbo, object); > + return; > + > +out: > + GL_EXTCALL(glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, 0)); > + GL_EXTCALL(glDeleteBuffersARB(1, &object->vbo)); > + LEAVE_GL(); > + object->vbo = 0; > +} > + > static HRESULT WINAPI IWineD3DDeviceImpl_CreateIndexBuffer(IWineD3DDevice *iface, UINT Length, DWORD Usage, > WINED3DFORMAT Format, WINED3DPOOL Pool, IWineD3DIndexBuffer** ppIndexBuffer, > HANDLE *sharedHandle, IUnknown *parent) { > IWineD3DDeviceImpl *This = (IWineD3DDeviceImpl *)iface; > IWineD3DIndexBufferImpl *object; > TRACE("(%p) Creating index buffer\n", This); > - > + > /* Allocate the storage for the device */ > D3DCREATERESOURCEOBJECTINSTANCE(object,IndexBuffer,WINED3DRTYPE_INDEXBUFFER, Length) > - > - /*TODO: use VBO's */ > - if (Pool == WINED3DPOOL_DEFAULT ) { /* Allocate some system memory for now */ > + > + if (Pool == WINED3DPOOL_DEFAULT ) { /* We need a local copy for drawStridedSlow */ > object->resource.allocatedMemory = HeapAlloc(GetProcessHeap(), HEAP_ZERO_MEMORY,object->resource.size); > } > > + if(Pool != WINED3DPOOL_SYSTEMMEM && !(Usage & WINED3DUSAGE_DYNAMIC) && GL_SUPPORT(ARB_VERTEX_BUFFER_OBJECT)) { > + CreateIndexBufferVBO(This, object); > + } > + > TRACE("(%p) : Len=%d, Use=%x, Format=(%u,%s), Pool=%d - Memory@%p, Iface@%p\n", This, Length, Usage, Format, > debug_d3dformat(Format), Pool, object, object->resource.allocatedMemory); > *ppIndexBuffer = (IWineD3DIndexBuffer *) object; > @@ -2557,10 +2603,15 @@ static HRESULT WINAPI IWineD3DDeviceImpl_SetIndices(IWineD3DDevice *iface, IWine > return WINED3D_OK; > } > > - /* So far only the base vertex index is tracked */ > + /* The base vertex index affects the stream sources, while > + * The index buffer is a seperate index buffer state > + */ > if(BaseVertexIndex != oldBaseIndex) { > IWineD3DDeviceImpl_MarkStateDirty(This, STATE_STREAMSRC); > } > + if(oldIdxs != pIndexData) { > + IWineD3DDeviceImpl_MarkStateDirty(This, STATE_INDEXBUFFER); > + } > return WINED3D_OK; > } > > @@ -4224,9 +4275,11 @@ static HRESULT WINAPI IWineD3DDeviceImpl_DrawIndexedPrimitive(IWineD3DDevice * > UINT idxStride = 2; > IWineD3DIndexBuffer *pIB; > WINED3DINDEXBUFFER_DESC IdxBufDsc; > + GLint vbo; > > pIB = This->stateBlock->pIndexData; > This->stateBlock->streamIsUP = FALSE; > + vbo = ((IWineD3DIndexBufferImpl *) pIB)->vbo; > > TRACE("(%p) : Type=(%d,%s), min=%d, CountV=%d, startIdx=%d, countP=%d\n", This, > PrimitiveType, debug_d3dprimitivetype(PrimitiveType), > @@ -4245,7 +4298,7 @@ static HRESULT WINAPI IWineD3DDeviceImpl_DrawIndexedPrimitive(IWineD3DDevice * > } > > drawPrimitive(iface, PrimitiveType, primCount, 0, NumVertices, startIndex, > - idxStride, ((IWineD3DIndexBufferImpl *) pIB)->resource.allocatedMemory, minIndex); > + idxStride, vbo ? NULL : ((IWineD3DIndexBufferImpl *) pIB)->resource.allocatedMemory, minIndex); > > return WINED3D_OK; > } > @@ -4310,6 +4363,7 @@ static HRESULT WINAPI IWineD3DDeviceImpl_DrawIndexedPrimitiveUP(IWineD3DDevice * > This->stateBlock->loadBaseVertexIndex = 0; > /* Mark the state dirty until we have nicer tracking of the stream source pointers */ > IWineD3DDeviceImpl_MarkStateDirty(This, STATE_VDECL); > + IWineD3DDeviceImpl_MarkStateDirty(This, STATE_INDEXBUFFER); > > drawPrimitive(iface, PrimitiveType, PrimitiveCount, 0 /* vertexStart */, NumVertices, 0 /* indxStart */, idxStride, pIndexData, MinVertexIndex); > > @@ -4332,8 +4386,10 @@ static HRESULT WINAPI IWineD3DDeviceImpl_DrawPrimitiveStrided (IWineD3DDevice *i > * that value. > */ > IWineD3DDeviceImpl_MarkStateDirty(This, STATE_VDECL); > + IWineD3DDeviceImpl_MarkStateDirty(This, STATE_INDEXBUFFER); > This->stateBlock->baseVertexIndex = 0; > This->up_strided = DrawPrimStrideData; > + This->stateBlock->streamIsUP = TRUE; > drawPrimitive(iface, PrimitiveType, PrimitiveCount, 0, 0, 0, 0, NULL, 0); > This->up_strided = NULL; > return WINED3D_OK; > diff --git a/dlls/wined3d/drawprim.c b/dlls/wined3d/drawprim.c > index 2cc3a15..7cf6877 100644 > --- a/dlls/wined3d/drawprim.c > +++ b/dlls/wined3d/drawprim.c > @@ -571,7 +571,7 @@ static void drawStridedFast(IWineD3DDevice *iface,UINT numberOfVertices, GLenum > const void *idxData, short idxSize, ULONG minIndex, ULONG startIdx, ULONG startVertex) { > IWineD3DDeviceImpl *This = (IWineD3DDeviceImpl *)iface; > > - if (idxData != NULL /* This crashes sometimes!*/) { > + if (idxSize != 0 /* This crashes sometimes!*/) { > TRACE("(%p) : glElements(%x, %d, %d, ...)\n", This, glPrimitiveType, numberOfVertices, minIndex); > idxData = idxData == (void *)-1 ? NULL : idxData; > #if 1 > @@ -629,7 +629,15 @@ static void drawStridedSlow(IWineD3DDevice *iface, WineDirect3DVertexStridedData > TRACE("Using slow vertex array code\n"); > > /* Variable Initialization */ > - if (idxData != NULL) { > + if (idxSize != 0) { > + /* Immediate mode drawing can't make use of indices in a vbo - get the data from the index buffer. > + * If the index buffer has no vbo(not supported or other reason), or with user pointer drawing > + * idxData will be != NULL > + */ > + if(idxData == NULL) { > + idxData = ((IWineD3DIndexBufferImpl *) This->stateBlock->pIndexData)->resource.allocatedMemory; > + } > + > if (idxSize == 2) pIdxBufS = (const short *) idxData; > else pIdxBufL = (const long *) idxData; > } > @@ -1165,7 +1173,7 @@ inline void drawStridedInstanced(IWineD3DDevice *iface, WineDirect3DVertexStride > IWineD3DDeviceImpl *This = (IWineD3DDeviceImpl *) iface; > IWineD3DStateBlockImpl *stateblock = This->stateBlock; > > - if (idxData == NULL) { > + if (idxSize == 0) { > /* This is a nasty thing. MSDN says no hardware supports that and apps have to use software vertex processing. > * We don't support this for now > * > diff --git a/dlls/wined3d/indexbuffer.c b/dlls/wined3d/indexbuffer.c > index 39a270d..0f768ac 100644 > --- a/dlls/wined3d/indexbuffer.c > +++ b/dlls/wined3d/indexbuffer.c > @@ -58,6 +58,15 @@ static ULONG WINAPI IWineD3DIndexBufferImpl_Release(IWineD3DIndexBuffer *iface) > ULONG ref = InterlockedDecrement(&This->resource.ref); > TRACE("(%p) : Releasing from %d\n", This, ref + 1); > if (ref == 0) { > + if(This->vbo) { > + ENTER_GL(); > + GL_EXTCALL(glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, 0)); > + checkGLcall("glBindBufferARB"); > + GL_EXTCALL(glDeleteBuffersARB(1, &This->vbo)); > + checkGLcall("glDeleteBuffersARB"); > + LEAVE_GL(); > + } > + > IWineD3DResourceImpl_CleanUp((IWineD3DResource *)iface); > HeapFree(GetProcessHeap(), 0, This); > } > @@ -108,13 +117,49 @@ static HRESULT WINAPI IWineD3DIndexBufferImpl_GetParent(IWineD3DIndexBuffer *ifa > ****************************************************** */ > static HRESULT WINAPI IWineD3DIndexBufferImpl_Lock(IWineD3DIndexBuffer *iface, UINT OffsetToLock, UINT SizeToLock, BYTE** ppbData, DWORD Flags) { > IWineD3DIndexBufferImpl *This = (IWineD3DIndexBufferImpl *)iface; > - TRACE("(%p) : no real locking yet, offset %d, size %d, Flags=%x\n", This, OffsetToLock, SizeToLock, Flags); > + TRACE("(%p) : offset %d, size %d, Flags=%x\n", This, OffsetToLock, SizeToLock, Flags); > + > + InterlockedIncrement(&This->lockcount); > *ppbData = (BYTE *)This->resource.allocatedMemory + OffsetToLock; > + > + if(Flags & (WINED3DLOCK_READONLY | WINED3DLOCK_NO_DIRTY_UPDATE) || This->vbo == 0) { > + return WINED3D_OK; > + } > + > + if(This->dirtystart != This->dirtyend) { > + if(This->dirtystart > OffsetToLock) This->dirtystart = OffsetToLock; > + if(SizeToLock) { > + if(This->dirtyend < OffsetToLock + SizeToLock) This->dirtyend = OffsetToLock + SizeToLock; > + } else { > + This->dirtyend = This->resource.size; > + } > + } else { > + This->dirtystart = OffsetToLock; > + if(SizeToLock) > + This->dirtyend = OffsetToLock + SizeToLock; > + else > + This->dirtyend = This->resource.size; > + } > + > return WINED3D_OK; > } > static HRESULT WINAPI IWineD3DIndexBufferImpl_Unlock(IWineD3DIndexBuffer *iface) { > IWineD3DIndexBufferImpl *This = (IWineD3DIndexBufferImpl *)iface; > - TRACE("(%p) : stub\n", This); > + unsigned long locks = InterlockedDecrement(&This->lockcount);; > + TRACE("(%p)\n", This); > + > + /* For now load in unlock */ > + if(locks == 0 && This->vbo) { > + ENTER_GL(); > + GL_EXTCALL(glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, This->vbo)); > + checkGLcall("glBindBufferARB"); > + GL_EXTCALL(glBufferSubDataARB(GL_ELEMENT_ARRAY_BUFFER_ARB, > + This->dirtystart, This->dirtyend - This->dirtystart, This->resource.allocatedMemory + This->dirtystart)); > + checkGLcall("glBufferSubDataARB"); > + LEAVE_GL(); > + This->dirtystart = 0; > + This->dirtyend = 0; > + } > return WINED3D_OK; > } > static HRESULT WINAPI IWineD3DIndexBufferImpl_GetDesc(IWineD3DIndexBuffer *iface, WINED3DINDEXBUFFER_DESC *pDesc) { > diff --git a/dlls/wined3d/state.c b/dlls/wined3d/state.c > index 0b96e54..f78720a 100644 > --- a/dlls/wined3d/state.c > +++ b/dlls/wined3d/state.c > @@ -3062,6 +3062,17 @@ static void scissorrect(DWORD state, IWineD3DStateBlockImpl *stateblock, WineD3D > checkGLcall("glScissor"); > } > > +static void indexbuffer(DWORD state, IWineD3DStateBlockImpl *stateblock, WineD3DContext *context) { > + if(GL_SUPPORT(ARB_VERTEX_BUFFER_OBJECT)) { > + if(stateblock->streamIsUP || stateblock->pIndexData == NULL ) { > + GL_EXTCALL(glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, 0)); > + } else { > + IWineD3DIndexBufferImpl *ib = (IWineD3DIndexBufferImpl *) stateblock->pIndexData; > + GL_EXTCALL(glBindBufferARB(GL_ELEMENT_ARRAY_BUFFER_ARB, ib->vbo)); > + } > + } > +} > + > const struct StateEntry StateTable[] = > { > /* State name representative, apply function */ > @@ -4077,6 +4088,7 @@ const struct StateEntry StateTable[] = > { /*511, WINED3DTS_WORLDMATRIX(255) */ STATE_TRANSFORM(WINED3DTS_WORLDMATRIX(255)), transform_worldex }, > /* Various Vertex states follow */ > { /* , STATE_STREAMSRC */ STATE_VDECL, vertexdeclaration }, > + { /* , STATE_INDEXBUFFER */ STATE_INDEXBUFFER, indexbuffer }, > { /* , STATE_VDECL */ STATE_VDECL, vertexdeclaration }, > { /* , STATE_VSHADER */ STATE_VDECL, vertexdeclaration }, > { /* , STATE_VIEWPORT */ STATE_VIEWPORT, viewport }, > diff --git a/dlls/wined3d/wined3d_private.h b/dlls/wined3d/wined3d_private.h > index 3bf0c8e..be8cd6e 100644 > --- a/dlls/wined3d/wined3d_private.h > +++ b/dlls/wined3d/wined3d_private.h > @@ -426,8 +426,10 @@ typedef void (*APPLYSTATEFUNC)(DWORD state, IWineD3DStateBlockImpl *stateblock, > > #define STATE_STREAMSRC (STATE_TRANSFORM(WINED3DTS_WORLDMATRIX(255)) + 1) > #define STATE_IS_STREAMSRC(a) ((a) == STATE_STREAMSRC) > +#define STATE_INDEXBUFFER (STATE_STREAMSRC + 1) > +#define STATE_IS_INDEXBUFFER(a) ((a) == STATE_INDEXBUFFER) > > -#define STATE_VDECL (STATE_STREAMSRC + 1) > +#define STATE_VDECL (STATE_INDEXBUFFER + 1) > #define STATE_IS_VDECL(a) ((a) == STATE_VDECL) > > #define STATE_VSHADER (STATE_VDECL + 1) > @@ -793,6 +795,10 @@ typedef struct IWineD3DIndexBufferImpl > const IWineD3DIndexBufferVtbl *lpVtbl; > IWineD3DResourceClass resource; > > + GLuint vbo; > + UINT dirtystart, dirtyend; > + LONG lockcount; > + > /* WineD3DVertexBuffer specifics */ > } IWineD3DIndexBufferImpl; > > > > ------------------------------------------------------------------------ > >

2 3

Re: Rediffed patch from Rein Klazes:winex11.drv:Add pointer check in X11DRV_SetDIBits
by Felix Nawothnig Feb. 20, 2007

Feb. 20, 2007

Louis. Lenders wrote: > + if( IsBadReadPtr( bits, nrsrcbytes)) return 0; AFAIK you should never use that function since it introduces a race con. Try using an exception handler around the memory access itself. Felix

1 0

wine-0.9.31 version is absent from bugzilla
by Vitaliy Margolen Feb. 20, 2007

Feb. 20, 2007

Please whoever has powers to add new versions to bugzilla, can they please do that? Thanks. Vitaliy.

3 2

a big fan of your work
by Jeffrey Sabarese Feb. 20, 2007

Feb. 20, 2007

hi there! I have to wonder if i don't use WinE more than the "average" Fedora Core 5 user. I am forever trying to figure out how to run my old Windows apps here in the linux environment, using your incredible software. If ever you seek a "test subject", or feedback of any kind, please don't hesitate to include me as a candidate for some "volunteer work". If i may be so bold, i wish to address an issue related to WinE and some Web Development software: I have been able to successfully install, and use the Web Dev IDE, HomeSite+ (as in "the homesite which comes w/ dreamweaver nowadays"). previous to this, i was unable to get Bradbury's TopStyle Lite to operate under WinE due to the inability to "find Internet Exlorer", yet it seems as though HomeSite opens TopStyle w/out it's dependecy on IE keeping it from opening (bypassing the "warning" you might say). [can i just say it's sooo cool to finally have TopStyle here in Linux!!! thank you!). I guess you could say then that my question for you is: "How might i go about trying to get IE6 to be "recognized" by such WinE-apps as TopStyle?" I have installed, and use the "ies4linux" ( here for more info if you're unfamiliar with it: http://novicenotes.com/install-internet-explorer-on-linux-under-wine-run-ie… ), but though it works VERY well launching from the Fedora Terminal, i can't seem to get it to be recognized by "Windows" (WinE) as an available Browser. The apps i've tried which seem unable to integrate it are i. EditPlus 2.x (current) ii. TSW WebCoder 5 (as TSW Webcoder 2005 ( Delphi ) will NOT run on WinE, though i wish i did as i have a purchased licence) iii. TopStyle Lite iv. HomeSite Dreamweaver 8 will both run on WinE, and execute "normally", the ies4linux IE6-- yet, obviously there is a primary difference in that IE is NOT integrated into DreamWeaver as "part of the app", as it is in WebCoder and TopStyle. My best guess is that those apps have some security measure preventing a "non-standard" IE to become part of its processes. who knows. anyway-- i thought there was something else, but i suppose that's more than enough b.s. from me from one sitting. good luck in your work! if you have any ideas for my situation, i'd appriciate your input. again, please don't hesitate to inquire if you'd like for me to "test" anything for you and report my findings. i realize you probably have a whole lot of people already out there, but feel free to add me to the list. thanks, --Jeff Sabarese

3 2

Re: [7/10] WineD3D: No bounds checking is done on sampler / texture numbers
by H. Verbeet Feb. 20, 2007

Feb. 20, 2007

On 19/02/07, Stefan Dösinger <stefan(a)codeweavers.com> wrote: >- if(Sampler > GL_LIMITS(sampler_stages) || Sampler < 0 || Type > WINED3D_HIGHEST_SAMPLER_STATE || Type < 0) { This also checked the sampler state, not just the sampler.

4 5

Re: [PATCH 3/3] server: Check object's security when creating handles.
by Robert Shearman Feb. 20, 2007

Feb. 20, 2007

Vitaliy Margolen wrote: > After > checking object's SD against token we fail some tests. > What it seems to me is that some one tried to "optimize" this part in > windows and instead created a security problem. If you already have a valid handle to the object, then it isn't really a security problem to have another one with the same access rights. > @@ -450,25 +453,34 @@ obj_handle_t duplicate_handle( struct pr > unsigned int access, unsigned int attr, unsigned int options ) > { > obj_handle_t res; > + struct handle_entry *entry; > + unsigned int src_access, dst_access; > struct object *obj = get_handle_obj( src, src_handle, 0, NULL ); > > if (!obj) return 0; > - if (options & DUP_HANDLE_SAME_ACCESS) > + if ((entry = get_handle( src, src_handle ))) > + src_access = entry->access; > + else /* pseudo-handle, give it full access */ > { > - struct handle_entry *entry = get_handle( src, src_handle ); > - if (entry) > - access = entry->access; > - else /* pseudo-handle, give it full access */ > - { > - access = obj->ops->map_access( obj, GENERIC_ALL ); > - clear_error(); > - } > + src_access = obj->ops->map_access( obj, GENERIC_ALL ); > + clear_error(); > } > - access &= ~RESERVED_ALL; > + src_access &= ~RESERVED_ALL; > + > + if (options & DUP_HANDLE_SAME_ACCESS) > + dst_access = src_access; > + else > + dst_access = obj->ops->map_access( obj, access ) & ~RESERVED_ALL; > + > + /* asking for more or the same/less access rights */ > + access = ~src_access & dst_access ? dst_access : 0; > + > if (options & DUP_HANDLE_MAKE_GLOBAL) > res = alloc_global_handle( obj, access ); > else > res = alloc_handle( dst, obj, access, attr ); > + if (res) get_handle( dst, res )->access = dst_access; > + > It would seem better to add an extra parameter to alloc_handle and alloc_global_handle to tell them to not check access rather than adding this hack. It should use less processor time and it should make patch 2 not necessary. -- Rob Shearman

2 1