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

Re: [7/7] cmd.exe: Support rd dir1 dir2 dir3 etc

by Alexandre Julliard

"Ann & Jason Edmeades" <us(a)edmeades.me.uk> writes: > diff --git a/programs/cmd/wcmd.h b/programs/cmd/wcmd.h > index d6357f0..a7a73ae 100644 > --- a/programs/cmd/wcmd.h > +++ b/programs/cmd/wcmd.h > @@ -27,6 +27,9 @@ > #include <stdio.h> > #include <ctype.h> > > +#include "wine/debug.h" > +WINE_DEFAULT_DEBUG_CHANNEL(cmd); Adding tracing is OK, but this should go in the C files that need it, not in a header. -- Alexandre Julliard julliard(a)winehq.org

Feb. 28, 2007

Re: [6/6] d3d9: Add a mova rounding test

by Alexandre Julliard

"H. Verbeet" <hverbeet(a)gmail.com> writes: > It should be noted that native d3d rounds to nearest even while Wine > doesn't. Applications are not supposed to depend on that though. > > Changelog: > - Add a mova rounding test The test fails here: ../../../tools/runtest -q -P wine -M d3d9.dll -T ../../.. -p d3d9_test.exe.so visual.c && touch visual.ok visual.c:405: Test failed: Expected color 00ffff00, got 00000000 (for input -2.400000) visual.c:405: Test failed: Expected color 00ffff00, got 00000000 (for input -1.600000) visual.c:405: Test failed: Expected color 0000ffff, got 00000000 (for input -0.400000) visual.c:405: Test failed: Expected color 0000ffff, got 00000000 (for input 0.400000) visual.c:405: Test failed: Expected color 00ff00ff, got 00000000 (for input 1.600000) visual.c:405: Test failed: Expected color 00ff00ff, got 00000000 (for input 2.400000) make: *** [visual.ok] Error 6 -- Alexandre Julliard julliard(a)winehq.org

Feb. 28, 2007

Re: wine and msys rxvt.exe

by Jakob Eriksson

Phil Krylov wrote: > > AFAIR MSys's rxvt.exe uses libW11.dll, which implements Xlib API on > top of Win32 USER/GDI (no extra X server). However, it could be still > difficult (probably because of API name clashes?) > Not sure why the message is about libX11.dll - probably it sees > $DISPLAY and tries to use real X? AFAIR rxvt can be used both with and without X server, so that's probable. regards, Jakob

Feb. 28, 2007

Re: msi: Add proper parameter handling to OLE automation interfaces.

by Robert Shearman

Misha Koshelev wrote: > +/* Macros for variant conversions */ > +#define V_DECLARE \ > + VARIANTARG vararg[pDispParams->cArgs]; \ > + int iVarCount; \ > + HRESULT hr = S_OK; \ > + memset(vararg, 0, sizeof(vararg)); > +#define V_PREPARE(n, type) \ > + VariantInit(&vararg[n]); \ > + hr = VariantChangeTypeEx(&vararg[n], &pDispParams->rgvarg[n], lcid, 0, VT_##type); \ > + if (FAILED(hr)) \ > + { \ > + *puArgErr = n; \ > + goto v_finish; \ > + } > +#define V_PREPAREOPT(n, type) \ > + VariantInit(&vararg[n]); \ > + if (FAILED(VariantChangeTypeEx(&vararg[n], &pDispParams->rgvarg[n], lcid, 0, VT_##type))) \ > + VariantClear(&vararg[n]); > +#define V_PREPARED(n) (V_VT(&vararg[n]) != VT_EMPTY) > +#define V(n, type) V_##type(&vararg[n]) > +#define V_RETURN(value, type) \ > + V_VT(pVarResult) = VT_##type; \ > + V_##type(pVarResult) = value; > +#define V_FAIL hr = E_FAIL; > +#define V_DEFAULT \ > + default: \ > + hr = DISP_E_MEMBERNOTFOUND; > +#define V_FINISH \ > + goto v_finish; /* suppress warnings about unused label v_free */ \ > +v_finish: \ > + for (iVarCount = 0; iVarCount < pDispParams->cArgs; iVarCount++) \ > + if (V_PREPARED(iVarCount)) \ > + VariantClear(&vararg[iVarCount]); \ > + return hr; > This is quite ugly. You can at replace at least some of this with calls to DispGetParam. A lot of the methods only have one parameter so will end up simpler without using these macros. -- Rob Shearman

Feb. 28, 2007

Re: fc4 rpms

by Marcus Meissner

On Wed, Feb 28, 2007 at 09:43:02AM +0100, Andreas Bierfert wrote: > On Tue, 27 Feb 2007 18:17:04 -0800 > Bill Medland <billmedland(a)shaw.ca> wrote: > > > Is there any chance that someone with the ability can create the fedora > > core 4 wine rpms for version 0.9.30? > > > > Bill Medland > > If you want to I can upgrade the spec in fedora cvs to the newest versions as > well. It is sad that I cannot push builds for fc 3 and fc 4 anymore I might put > builds on my webpage if the audience is big enough... Fedora Core 4 RPMs are still available in the openSUSE buildservice. YUM base URL: http://software.opensuse.org/download/Emulators:/Wine/Fedora_Core_4/ Ciao, Marcus

Feb. 28, 2007

Re: fc4 rpms

by Andreas Bierfert

On Tue, 27 Feb 2007 18:17:04 -0800 Bill Medland <billmedland(a)shaw.ca> wrote: > Is there any chance that someone with the ability can create the fedora > core 4 wine rpms for version 0.9.30? > > Bill Medland If you want to I can upgrade the spec in fedora cvs to the newest versions as well. It is sad that I cannot push builds for fc 3 and fc 4 anymore I might put builds on my webpage if the audience is big enough... - Andreas -- Andreas Bierfert | http://awbsworld.de | GPG: C58CF1CB andreas.bierfert(a)lowlatency.de | http://lowlatency.de | signed/encrypted phone: +49 2402 102373 | cell: +49 173 5803043 | mail preferred

Feb. 28, 2007

Re: [8/12] WineD3D: Create multithreading contexts for swapchains

by Mirek

Stefan Dösinger napsal(a): > This patch finishes the opengl side of multithreaded direct3d, which was > started with the state management rewrite(well, almost, no offscreen rendering > yet) It does not make Direct3D thread safe, because there is no protection > against race conditions. > > This patch improves multithreaded games a bit. They do not crash any more in a > gl call, but running them is kinda like a lottery, especially on smp systems. > > Vitaly and Andras have reported success with Prince of Persia 3D: Sands of > Time and some other games, and on my laptop Empire Earth is running. Empire > Earth deadlocks regularly on my dual core system at ddraw creation. I can even run Tom Clancy's Rainbow Six Vegas, but there are still problems with offscreen rendering. Mirek > > > ------------------------------------------------------------------------ > > From 49cc9a0aa6e06f559540758d5b0ed2252208a0ad Mon Sep 17 00:00:00 2001 > From: Stefan Doesinger <stefan(a)codeweavers.com> > Date: Sun, 25 Feb 2007 00:18:25 +0100 > Subject: [PATCH] WineD3D: Create multithreading contexts for swapchains > > This patch adds a routine to clone the main context of a swapchain to use it with a different thread on the > same drawable. This will enable multithreaded direct3d rendering, or better enable an attempt to do that. > From the opengl point of view it will work, but because concurrency control is not implemented yet(appart > of ENTER_GL() / LEAVE_GL() ) games may crash randomly, show strange behavior, ... > > The other problem is that this code was only tested with the open source radeon driver yet. While what > wined3d is doing here should be allowed by the spec, some drivers may have some bugs when using multiple > contexts on one drawable. > --- > dlls/wined3d/context.c | 4 ++-- > dlls/wined3d/swapchain.c | 30 ++++++++++++++++++++++++++++++ > dlls/wined3d/wined3d_private.h | 2 ++ > 3 files changed, 34 insertions(+), 2 deletions(-) > > diff --git a/dlls/wined3d/context.c b/dlls/wined3d/context.c > index 3a815b3..8545568 100644 > --- a/dlls/wined3d/context.c > +++ b/dlls/wined3d/context.c > @@ -626,8 +626,8 @@ void ActivateContext(IWineD3DDeviceImpl *This, IWineD3DSurface *target, ContextU > } > > if(!context) { > - /* TODO: Create a new context for the thread */ > - FIXME("Context creation for a new thread not implemented yet\n"); > + /* Create a new context for the thread */ > + context = IWineD3DSwapChainImpl_CreateContextForThread(swapchain); > } > } else { > context = ((IWineD3DSwapChainImpl *) swapchain)->context[0]; > diff --git a/dlls/wined3d/swapchain.c b/dlls/wined3d/swapchain.c > index 2977a05..3b8e3e8 100644 > --- a/dlls/wined3d/swapchain.c > +++ b/dlls/wined3d/swapchain.c > @@ -510,3 +510,33 @@ const IWineD3DSwapChainVtbl IWineD3DSwapChain_Vtbl = > IWineD3DSwapChainImpl_SetGammaRamp, > IWineD3DSwapChainImpl_GetGammaRamp > }; > + > +WineD3DContext *IWineD3DSwapChainImpl_CreateContextForThread(IWineD3DSwapChain *iface) { > + WineD3DContext *ctx; > + IWineD3DSwapChainImpl *This = (IWineD3DSwapChainImpl *) iface; > + WineD3DContext **newArray; > + > + TRACE("Creating a new context for swapchain %p, thread %d\n", This, GetCurrentThreadId()); > + > + ctx = CreateContext(This->wineD3DDevice, (IWineD3DSurfaceImpl *) This->frontBuffer, > + This->context[0]->display, This->win); > + if(!ctx) { > + ERR("Failed to create a new context for the swapchain\n"); > + return NULL; > + } > + > + newArray = HeapAlloc(GetProcessHeap(), 0, sizeof(*newArray) * This->num_contexts + 1); > + if(!newArray) { > + ERR("Out of memory when trying to allocate a new context array\n"); > + DestroyContext(This->wineD3DDevice, ctx); > + return NULL; > + } > + memcpy(newArray, This->context, sizeof(*newArray) * This->num_contexts); > + HeapFree(GetProcessHeap(), 0, This->context); > + newArray[This->num_contexts] = ctx; > + This->context = newArray; > + This->num_contexts++; > + > + TRACE("Returning context %p\n", ctx); > + return ctx; > +} > diff --git a/dlls/wined3d/wined3d_private.h b/dlls/wined3d/wined3d_private.h > index 45e69bf..fa4232c 100644 > --- a/dlls/wined3d/wined3d_private.h > +++ b/dlls/wined3d/wined3d_private.h > @@ -1341,6 +1341,8 @@ typedef struct IWineD3DSwapChainImpl > > extern const IWineD3DSwapChainVtbl IWineD3DSwapChain_Vtbl; > > +WineD3DContext *IWineD3DSwapChainImpl_CreateContextForThread(IWineD3DSwapChain *iface); > + > /***************************************************************************** > * Utility function prototypes > */ > > > ------------------------------------------------------------------------ > >

Feb. 28, 2007

Re: msi: automation: clean up comments.

by James Hawkins

On 2/27/07, Misha Koshelev <mk144210(a)bcm.tmc.edu> wrote: > Looking back at James comments, maybe I did have a little too much > commenting there :) > You're sending patches that change code that hasn't been committed yet. Instead, you need to go back and change the original patch and resend. -- James Hawkins

Feb. 28, 2007

Re: msi: Expand OLE automation interface parameter handling.

by James Hawkins

On 2/27/07, Misha Koshelev <mk144210(a)bcm.tmc.edu> wrote: > This applies on top of my other OLE automation patches. Added new > checks: > Why don't you wait until the other patches are committed before sending these new patches? It makes it hard to review, because I don't have your other patches applied. > - Parameter we are looking for is not passed? > - Only set puArgErr if it is passed > - Make sure pVarResult is valid before we write to it > - Don't return S_FAIL. MSDN doesn't mention it in the Invoke help, so we > won't use it. > You need to add tests for these items. -- James Hawkins

Feb. 28, 2007

Success story with Flash on Wine

by Dan Kegel

http://blog.brokenfunction.com/2007/02/27/debugging-flash-on-linux/ says "... How strange, the wine-powered standalone player is the most stable Flash player I've used yet. On Windows it would occasionally be unable to connect to the debugger until I restarted Eclipse, usually related to having to many players open at the same time. On Linux it will crash unexpectedly. Now I can have a dozen debuggers open at the same time with no crashes yet." Cool!

Feb. 28, 2007