July 2006 - Wine-Devel - list.winehq.org

Re: Asynchronous Direct3D rendering design sugeestion
by H. Verbeet July 22, 2006

July 22, 2006

On 20/07/06, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > for us. The reason why it doesn't are apparently drawStridedSlow and the > brute force setting of gl states. The debugging facilies seem to have really > bad effects on performance too. Well, that should be fixed anyway, of course. > Yet I think some things should be done, like having vertex and pixel format > fixups in a seperate thread, or shader perhaps compilation. The D3D objects > have to be protected with critical sections too. I think that should be dependant on whether the device was created with D3DCREATE_MULTITHREADED or not. I'm not sure how much can be gained by moving shader compilation into it's own thread, but I suppose it's worth some tests.

2 1

Re: wined3d: Fix regression from dynamic constants patch
by Ivan Gyurdiev July 22, 2006

July 22, 2006

> > There were plenty of ways to fix this. One would have been to unroll > all of the memset/memcpy calls into functions that explicitly set > every value on the stateblock. While this might have been more > "correct", it would be harder to maintain. I argue that an explicit initializer is always better than an implicit one. An implicit initializer is basically programmer laziness imho (I consider myself very lazy :) When you add a new feature, you should check its initialization one way or the other - can't rely on calloc assumptions, the code has to be verified. So, while you're doing that, it's easy to add an initializer. > Plus, we might lose some performance from numerous sets vs. > a single memset routine. Code correctness, clarity and reliability is much more important than any performance considerations imho. Saving 1 cpu cycle here or there does not matter in the long run. > So, I chose to make the number of constants more fixed again (less > dynamic). The number of available constants is now capped at 256, > which is the minimum required number for shader model 3.0. Software shaders require 8192 constants. They're not supported right now, but that gives you an idea of how many we need to support. Anyway, I'm not sure which way is better - maybe you're right that making things dynamic is not worth the trouble. However, I think that shows lack of flexibility in the stateblock, and if we ever need to make other things dynamic in the future we might run into the same issue. > The app > will use the min() of 256 and the number reported by the card So it uses the lower value? How come?

2 1

Re: wine-devel Digest, Vol 12, Issue 71
by Yuriy July 22, 2006

July 22, 2006

Uh-oh. I'm already well into making this. I guess we have a little mis-/lack-of communication. https://wiki.kubuntu.org/KDEGuidanceWineSpec As for ideas/testing -- yes please! It's part of guidance in kde svn. http://websvn.kde.org/trunk/playground/base/guidance/?rev=563975#dirlist ~Yuriy > Date: Thu, 20 Jul 2006 10:48:55 -0600 > From: "Brian Vincent" <brian.vincent(a)gmail.com> > Subject: KDE Wine Kcontrol module > To: "Wine Develop" <wine-devel(a)winehq.org> > Cc: davidsansome(a)gmail.com, kevin.krammer(a)gmx.at > Message-ID: > <3010968c0607200948v6f4f634avdcdbbbf4515ff069(a)mail.gmail.com> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > I just noticed KDE accepted a project for their "Season of KDE" > program for a Wine KControl module: > http://developer.kde.org/seasonofkde/project.php?kde_wine.xml > > Not sure if anyone from Wine wants to get in contact with the author, > but we might be able to help out with ideas / testing / etc. For > example, the two things that come to mind are that it should implement > everything in winecfg and also have a button to pop up regedit. Maybe > it should also be able to pop up taskmgr too. > > -Brian Vincent

3 3

Re: wined3d: Implement more GLSL instructions
by Chris July 21, 2006

July 21, 2006

On Friday 21 July 2006 11:19, Jason Green wrote: > + /* Calculate reflection vector (Assume normal is normalized): RF = 2*(N.E)*N -E */ > + shader_addline(buffer, "tmp0.w = dot(vec3(tmp0), %s);\n", src1_str); > + shader_addline(buffer, "tmp0 = tmp0.w * tmp0;\n"); > + shader_addline(buffer, "tmp0 = (2.0 * tmp0) - %s;\n", src1_str); I'm hardly an expert, but couldn't this be simplified to: shader_addline(buffer, "tmp0 = reflect(-%s, vec3(tmp0));\n", src1_str);

2 1

Wine f0.9.17 fails to build
by Segin July 21, 2006

July 21, 2006

Well, it seems that we have a bit of a problem building Wine 0.9.17 on Gentoo... make[2]: Entering directory `/var/tmp/portage/wine-0.9.17/work/wine-0.9.17/dlls/gphoto2.ds' ../../tools/makedep -I. -I. -I../../include -I../../include -I/usr/include/gphoto2 -I/usr/include -I/usr/include/gphoto2 -I/usr/include -C. capability.c ds_ctrl.c ds_image.c gphoto2_main.c ui.c rsrc.rc /usr/include/io.h:29: config.h must not be included by a header file make[2]: *** [depend] Error 1 make[2]: Leaving directory `/var/tmp/portage/wine-0.9.17/work/wine-0.9.17/dlls/gphoto2.ds' make[1]: *** [gphoto2.ds/__depend__] Error 2 make[1]: Leaving directory `/var/tmp/portage/wine-0.9.17/work/wine-0.9.17/dlls' make: *** [dlls/__depend__] Error 2 I don't know if this is my fault, or what, because I have recently reinstalled gphoto2 from scratch. -- The real problem with C++ for kernel modules is: the language just sucks. -- Linus Torvalds

1 0

ntdll: Heap implementation has a performance bottleneck
by Michael Kaufmann July 21, 2006

July 21, 2006

Hi, I have discovered that Wine's heap implementation (HeapAlloc(), HeapFree(), etc. in dlls/ntdll/heap.c) is very slow if many small memory blocks are allocated and freed. Somebody has reported this before: http://www.winehq.org/pipermail/wine-devel/2005-February/034095.html HomeSite's syntax checker needs 120 seconds to load on Wine because of Wine's poor heap implementation. Most users will think that the program has crashed. I have created a trace of HomeSite's memory requests and a program to replay them. For details and performance measurements, please look at bug 5709 ( http://bugs.winehq.org/show_bug.cgi?id=5709 ). According to the source code, Wine's heap implementation is described in the book "Windows 95 System Programming Secrets". This implementation is inefficient: KB225099 ( http://support.microsoft.com/kb/225099/en-us ) says that the heap implementation of Windows NT 4.0 SP4 and 2000 is much better than the heap implementation of Windows 95. Does anybody know some details about the heap implementation of Windows NT/2000/XP ? Regards, Michael

3 3

Re: [kernel] add dummy HeapSetInformation()
by Robert Shearman July 21, 2006

July 21, 2006

dank(a)vroo.pair.com wrote: > /*********************************************************************** >+ * HeapSetInformation (KERNEL32.@) >+ * Change options for a given heap. >+ * >+ * NOTES >+ * All options are ignored. >+ * >+ * RETURNS >+ * TRUE: Success >+ * FALSE: Failure >+ */ >+BOOL WINAPI HeapSetInformation( >+ HANDLE heap, /* [in] Handle to the heap */ >+ DWORD class, /* [in] enum describing which option to set */ >+ LPCVOID poption, /* [in] Pointer to option value */ >+ DWORD optionlen /* [in] Length of option value */ >+) { >+ /* Only option available is Low Fragmentation Heap, which >+ * is just a performance thing. Allow but ignore; this should let >+ * Lionhead Studios' Black & White 2 run. >+ */ >+ return TRUE; >+} >+ > > You should make this function print a fixme and a minor nitpick that NOTES should probably be BUGS instead. -- Rob Shearman

4 3

Re: wined3d: Fix for for vertex shaders
by Ivan Gyurdiev July 21, 2006

July 21, 2006

> + /* Make sure the fog value is positive - values above 1.0 are ignored */ > + if (reg_maps->fog) > + shader_addline(&buffer, "MAX result.fogcoord, TMP_FOG, 0.0;\n"); > + > What do you mean by "values above 1.0 are ignored" ? Why does the GLSL implementation of this clamp between 0 and 1, and the ARB implementation only does a lower bound ? (sorry, I don't remember if you responded when I asked this last time)

1 0

KDE Wine Kcontrol module
by Brian Vincent July 20, 2006

July 20, 2006

I just noticed KDE accepted a project for their "Season of KDE" program for a Wine KControl module: http://developer.kde.org/seasonofkde/project.php?kde_wine.xml Not sure if anyone from Wine wants to get in contact with the author, but we might be able to help out with ideas / testing / etc. For example, the two things that come to mind are that it should implement everything in winecfg and also have a button to pop up regedit. Maybe it should also be able to pop up taskmgr too. -Brian Vincent

2 1

Wine, cygwin, X, and ssh working pretty well these days...
by Dan Kegel July 20, 2006

July 20, 2006

To my surprise, I was able to verify a Wine bugfix today even though I'm in a hotel room with a Windows-only laptop. cygwin's X server worked like a champ, and I was able to ssh into my development box, git pull, build, and install and run a windows app fine remotely. It was just a bit slow, but I'm not complaining!

1 0