wine-devel February 2013

wine-devel@winehq.org

62 participants
126 discussions

Re: advapi32: Report administrator ownership by default for registry objects (try 3).
by Alexandre Julliard 19 Feb '13

19 Feb '13

"Erich E. Hoover" <ehoover(a)mymail.mines.edu> writes: > The attached patch fixes an issue where the Opera installer expects > registry keys to return ownership information (Bug #32904), concluding > the cleanup of [Get|Set]NamedSecurityInfo. This patch addresses the > issue by ensuring that NULL owner and group information is never > returned for registry keys and instead returns the Administrators SID > (S-1-5-32-544) when this information is unavailable, which is > sufficient to keep the Opera installer from encountering an error. > The included test shows that this behavior is the case for system keys > on Windows, which returns the Administrators SID for system registry > keys (such as HKLM\Software). Surely this is not specific to GetNamedSecurityInfo. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [14/14] gdiplus: Reimplement GdipDrawPath.
by Vincent Povirk 19 Feb '13

19 Feb '13

> Please don't apply this [14/14] patch, it causes a test failure. I'm sending > it to show how this is going to look like eventually. Yeah, GdipWidenPath isn't good enough for this yet. But I agree this is the way to go, and the rest of the series looks good.

2 2

Re: [PATCH 5/8] vbscript: Added RegExp.Execute tests
by Marvin 19 Feb '13

19 Feb '13

2 1

Re: [PATCH 7/8] vbscript: Added tests for invalid parameters in regular expression functions
by Marvin 19 Feb '13

19 Feb '13

2 1

Re: [PATCH 5/8] vbscript: Added RegExp.Execute tests
by Marvin 19 Feb '13

19 Feb '13

1 0

IE 6 related urlmon:protocol failures
by Francois Gouget 18 Feb '13

18 Feb '13

I have investigated the following test failure on the new WineTestBot: urlmon: protocol.c:1051: Test failed: hrResult = 800401e4, expected: 00000000 protocol.c:1056: Test failed: dwError == ERROR_SUCCESS protocol.c:613: Test failed: expected ReportData The results from test.winehq.org show that they only happen on systems that have urlmon.dll version 6.0, which I guess corresponds to Internet Explorer 6.0. More specifically: * The test detects systems that have Internet Explorer 5.5 or older, based on missing functions, declares too old and skips the tests: 5.0.3700.6705 wtb-w2kassp4 5.0.3888.1400 wtb-w2kprosp4 5.50.4807.2300 fg-win2000-ie55 * The test runs and fails on these systems: 6.0.2800.1649 fg-win2000-up1 6.0.2800.1649 fg-win2000-uplast 6.0.2900.3562 fg-winxp-2sp 6.0.2900.3698 wtb-wxpmce 6.0.2900.6049 wtb-wxphmnlsp3 * The test runs and succeeds on systems that have urlmon 7.0 or greater: 7.0.6000.16850 fg-winxp-2spie7 7.0.6000.16850 fg-winxp-3sp 8.0.6001.18806 fg-winxp-3spie8 8.0.6001.19100 fg-winxp-last So I will be applying the latest updates to the new TestBot's Windows XP VM and that should solve the issue in that case. However I think the test should still not fail. So I'm looking for a volunteer to either get the test to succeed with urlmon 6.0, or find a way to detect it and skip these three tests. -- Francois Gouget <fgouget(a)codeweavers.com>

2 4

Re: [PATCH 1/2] wininet: Address string should never be converted to UNICODE in status notifications
by Marvin 18 Feb '13

18 Feb '13

2 1

Re: msi: Don't ignore failure to remove an existing product.
by Alexandre Julliard 18 Feb '13

18 Feb '13

Hans Leidekker <hans(a)codeweavers.com> writes: > --- > dlls/msi/action.c | 5 ++++- > 1 file changed, 4 insertions(+), 1 deletion(-) It fails here: ../../../tools/runtest -q -P wine -M msi.dll -T ../../.. -p msi_test.exe.so action.c && touch action.ok action.c:6397: Test failed: Expected ERROR_SUCCESS, got 2 action.c:6400: Test failed: Expected ERROR_SUCCESS, got 2 make[1]: *** [action.ok] Error 2 -- Alexandre Julliard julliard(a)winehq.org

1 0

winmm: More compatible waveIn/Out[Un]Prepare WHDR_* flag handling.
by Joerg-Cyril.Hoehle＠t-systems.com 18 Feb '13

18 Feb '13

Hi, One thing to note is a slight change when Prepare is called twice in a row. When the ACM is used, the former Wine code would prepare the ACM again, leaking the memory of the previous preparation. Now that code is by-passed as the header is prepared exactly once, the first time. What if the app changes dwBufferLength in between? We don't know what exactly native does. My tests only show that *some* code is by-passed, because when WHDR_PREPARE is found already set, the other flags are not repeatedly cleared. All calls return NOERROR. The 16 bit code has a comment http://source.winehq.org/source/dlls/mmsystem.dll16/message16.c#L430 > /* dwBufferLength can be reduced between prepare & write */ > if (... && wh32->dwBufferLength < wh16->dwBufferLength) { I've checked neither - what native does in that case when the ACM is involved, nor - how Wine drives the ACM in that case. It's reasonable to assume that only dwBufferLength at the time of waveOutWrite should be relevant. Let's hope that dwBufferLength from the time of Prepare doesn't matter in Wine's 32 bit ACM. Regards, Jörg Höhle

1 0

Re: [PATCH 1/2] wined3d: Make (wined3d_)surface_depth_blt_fbo handle locations other than SFLAG_INTEXTURE.
by Henri Verbeet 18 Feb '13

18 Feb '13

On 17 February 2013 18:10, Matteo Bruni <mbruni(a)codeweavers.com> wrote: > @@ -6168,18 +6172,10 @@ HRESULT surface_load_location(struct wined3d_surface *surface, DWORD location, c > > if (surface->resource.usage & WINED3DUSAGE_DEPTHSTENCIL) > { > - if (location == SFLAG_INTEXTURE) > - { > - struct wined3d_context *context = context_acquire(device, NULL); > - surface_load_ds_location(surface, context, location); > - context_release(context); > - return WINED3D_OK; > - } > - else > - { > - FIXME("Unimplemented location %s for depth/stencil buffers.\n", debug_surflocation(location)); > - return WINED3DERR_INVALIDCALL; > - } > + struct wined3d_context *context = context_acquire(device, NULL); > + surface_load_ds_location(surface, context, location); > + context_release(context); > + return WINED3D_OK; > } This is probably wrong (and the original code is probably questionable too), surface_load_ds_location() is only really meant for onscreen <-> offscreen blits. The original code is meant for the case where AlwaysOffscreen is disabled and a depth stencil is used as texture. The implication there is that the only possible locations are SFLAG_INTEXTURE and SFLAG_INDRAWABLE, it doesn't really work with multisampling. Fortunately you can't directly texture from multisampled depth stencil surfaces either. In principle we could make surface_load_ds_location() the generic function for depth / stencil location transfers, but currently it isn't. If you're hitting it with the RESZ code it probably only works by accident.

2 2

Re: Wine release 1.5.24
by Gennady Telegin 17 Feb '13

17 Feb '13

Hi, I want to unsubscribe from this list, how to do it? Best, Gennady Telegin On Sat, Feb 16, 2013 at 12:30 AM, Alexandre Julliard <julliard(a)winehq.org>wrote: > The Wine development release 1.5.24 is now available. > > What's new in this release (see below for details): > - Keyboard and mouse wheel support in the Mac driver. > - Regular expression support in VB Script. > - Many RichEdit code cleanups. > - Various bug fixes. > > The source is available from the following locations: > > http://prdownloads.sourceforge.net/wine/wine-1.5.24.tar.bz2 > http://mirrors.ibiblio.org/wine/source/1.5/wine-1.5.24.tar.bz2 > > Binary packages for various distributions will be available from: > > http://www.winehq.org/download > > You will find documentation on http://www.winehq.org/documentation > > You can also get the current source directly from the git > repository. Check http://www.winehq.org/git for details. > > Wine is available thanks to the work of many people. See the file > AUTHORS in the distribution for the complete list. > > ---------------------------------------------------------------- > > Bugs fixed in 1.5.24 (total 38): > > 6512 power-tab-editor freezes at end of a song > 8598 msvcrt file IO functions do not handle unicode properly in text > mode > 12908 Battle Zone I does not start. > 16685 redraw problem in AIM_675 > 17380 CListCtrl: wrong icon spacing > 17762 Citavi: Application is very slow > 17763 Citavi: mouse doesn't catch links > 18079 AutoCorect: does not properly display custom checkboxes made by > Alcinoe > 20294 sigma photo pro crashes in X11DRV_DIB_DeleteDIBSection > 24089 EverQuest: Escape to Norrath: login screen is blank without > native ie6 > 24315 Verizon Media Manager crashes on exit (VZMediaagent.exe > 24361 Crashday: game is unusably slow during gameplay > 24474 Simpsons Hit & Run sound bug > 24554 Black screen in Everquest 2 (lighting issue?) > 25576 Microsoft Flight Simulator X: Demo/Deluxe Edition, menu blank > without native ie7 > 25584 Racedriver:GRID sound hardware acceleration not working > 25958 DreamStation 1 free music tracker visually freezes under some > conditions > 27011 Lucent Heart: patcher window is blank > 27905 HUNTED The Demon Forge: Sound does not work in the game (in > movies this works) > 29718 IE4 setup wants wininet.dll.LoadUrlCacheContent > 29873 Guild Wars: Water graphic is missing > 30008 Resource Hacker logo on about window has black background > 30183 Fchart installation crashes > 30246 EA Origin:Crashes when using openssl 1.0.1 > 30839 BSTR cache corrupts most of cached BSTR entries > 31308 Remote Tools for Visual Studio 2012 RC installer for Windows on > ARM (WoA) crashes because TPIDRURW (user TLS register) needs to be set to > TEB address > 31706 Sacred Underworld: Some models glow like a Christmas tree > 31977 YoudaoDict crash at start > 32520 EVE Online and other games want d3d11.dll.D3D11CreateDevice > 32669 Ghost Master: invisible mouse pointer unless 'Enhanced Cursor' > option selected > 32808 installer of PPTV: needs unimplemented > atl100.dll.AtlComModuleRegisterClassObjects > 32818 Adrenalin Extreme Show: launcher.exe has repainting issues > 32842 TurboTax 2012 needs shlwapi.dll IsInternetESCEnabled stub > 32858 Crash dialog Details button Russian text doesn't fit > 32862 Microsoft Expression Design 4 (Free Version) needs > WindowsCodecsExt.dll (Microsoft Windows Codecs Extended Library) > 32882 Grand Theft Auto IV doesn't start, aborts with a GLX error. > 32909 QQDownload 3.9 needs unimplemented msvcr80.dll._wstat32i64 > 32929 Microsoft Expression Design 4 (Free Version) needs > windowscodecsext.dll.WICCreateColorTransform_Proxy > > ---------------------------------------------------------------- > > Changes since 1.5.23: > > Akihiro Sagawa (1): > gdi32: Ensure a fixed-pitch full-width character has double advance > of a half-width character. > > Alexander Morozov (1): > ole32: Avoid a deadlock when a being loaded DLL calls > CoRegisterClassObject from its DLL_PROCESS_ATTACH handler. > > Alexandre Julliard (28): > winemac: Don't move off-screen windows to a random position. > gdi32: Return the correct module handle for the initial display > driver load. > explorer: Retrieve the graphics driver module from gdi32. > explorer: Return a simple boolean instead of a window in the > wine_create_desktop entry point. > wintab32: Retrieve the graphics driver module from gdi32. > imm32: Retrieve the graphics driver module from gdi32. > winex11: Ignore color key and exposures when using the null surface. > user32: Avoid releasing a potentially null pointer. > server: Also exclude the top-level client rectangle for windows that > have a pixel format. > server: Return the window paint flags in the get_visible_region > request. > user32: Don't paint to the surface if the window has a pixel format, > regardless of the state of the top parent. > wininet: Don't reset the available size after the async has been > queued. > user32: Prefer loading color cursors in LoadImage. > ntdll: Add a helper function to check that a fault address lies in a > known virtual memory view. > ntdll: Check for invalid %gs value in 32-bit code. > winex11: Preserve the client window when changing the window visual. > server: Don't support debug registers in ptrace if sys/user.h is > missing. > loader: Define our own auxv structure. > loader: Rename the elf_hash function to avoid conflicts. > odbccu32: Remove an unnecessary source file. > twain_32: Remove an unnecessary DllMain entry point. > cryptnet: Remove an unnecessary DllMain entry point. > dpnhpast: Remove an unnecessary DllMain entry point. > iphlpapi: Remove an unnecessary DllMain entry point. > ktmw32: Remove an unnecessary DllMain entry point. > sensapi: Remove an unnecessary DllMain entry point. > usp10: Remove an unnecessary DllMain entry point. > wtsapi32: Remove an unnecessary DllMain entry point. > > Alistair Leslie-Hughes (1): > fusion: Return correct error for null public token. > > Andrew Talbot (8): > jscript: Avoid signed-unsigned integer comparisons. > mapi32: Avoid signed-unsigned integer comparisons. > kernel32: Avoid signed-unsigned integer comparisons. > mscoree: Avoid signed-unsigned integer comparisons. > mmdevapi: Avoid signed-unsigned integer comparisons. > mlang: Avoid signed-unsigned integer comparisons. > msdaps: Avoid signed-unsigned integer comparisons. > msctf: Avoid signed-unsigned integer comparisons. > > André Hentschel (4): > ntdll: Set TPIDRURW to TEB on thread init for ARM. > make_xftmpl: Avoid comparing a char with EOF. > winedbg: Add PState to cpu_register_map for ARM64. > ntdll: Implement RtlUnwind for ARM. > > Aric Stewart (6): > usp10: Fix typo with Script_Khmer_Numeric. > usp10: Split the core Itemize function out to be able to support > differences between the normal and OpenType versions. > usp10: Update diacritical script handling for non OpenType > Itemization. > usp10: Free default_language items when freeing script cache. > usp10: Fix memory leak when a font has 0 languages, scripts or > features. > usp10: Correct error with ScriptLayout log2vis. > > Austin English (4): > windowscodecsext: Add a stub dll. > shlwapi: Add a stub for IsInternetESCEnabled. > d3d11: Add a stub for D3D11CreateDevice. > wininet: Add a stub for LoadUrlCacheContent. > > Daniel Jelinski (6): > comctl32/listview: Correct icon spacing calculation when set to 0 or > -1. > comctl32/listview: Always use large icon size when calculating icon > spacing. > comctl32/listview: Do not touch icon spacing if set explicitly. > user32: Fix CopyImage implementation when changing cursor to icon. > comctl32/listview: Better handling of custom colors in report view. > comctl32/listview: Fix LVM_SETICONSPACING on 64bit machines. > > Detlef Riekenberg (2): > advapi32/tests: Skip tests on non-English systems. > kernel32/tests: Skip tests on non-English systems. > > Dmitry Timoshkov (14): > wbemprox: Free name in create_signature (Coverity). > comctl32: Test appropriate tool window for active state. > gdi32: Add support for 16-bit QUERYESCSUPPORT queries. > gdi32: Add a bunch of tests for CreateDC/ResetDC with a custom > DEVMODE. > winspool.drv: Turn macros into helper functions. > wineps.drv: Add DEVMODE dumping to make debugging device state > easier. > winspool.drv: There is no need to speculate about wineps.drv default > DEVMODE. > gdi32: Make sure that actual clipping region is updated for a EMF DC. > gdi32: Add more EMF clipping tests. > gdi32: OffsetClipRgn should update actual clipping region for a EMF > DC. > gdi32: IntersectClipRect should update actual clipping region for a > EMF DC. > gdi32: ExcludeClipRect should update actual clipping region for a > EMF DC. > gdi32: SelectClipPath should update actual clipping region for a EMF > DC. > kernel32: There is no need for lock/unlock heap in GlobalAlloc. > > Erich Hoover (2): > advapi32: Support registry objects in GetNamedSecurityInfo. > advapi32: Support registry objects in SetNamedSecurityInfo. > > Francois Gouget (5): > kernel32/tests: Reduce the size of the console buffer to speed up > the tests. > rpcrt4/tests: Trace the GUID when it is wrong. > d3d11: Make debug_d3d_driver_type() static. > ws2_32/tests: Fix the GetAcceptExSockaddrs() test. > winetest: Trace the time it takes to run each test. > > Frédéric Delanoy (1): > msi: Fix memory leak (coverity). > > Hans Leidekker (12): > windowscodecs: Implement IWICColorContext::InitializeFromFilename. > windowscodecs: Add a stub IWICColorTransform implementation. > windowscodecs: Convert the source bitmap to the requested pixel > format in IWICColorTransform::Initialize. > windowscodecs: Implement > IWICComponentFactory::CreateDecoderFromFileHandle. > wbemprox: Trace the qualifier value in IWbemClassObject::GetNames. > windowscodecsext: Implement WICCreateColorTransform_Proxy and > IWICColorTransform_Initialize_Proxy. > windowscodecs: Implement IWICBitmapFrameDecode::GetColorContexts for > TIFF images. > msiexec: Add /unregserver support. > msiexec: Convert DoRegServer to Unicode. > wininet/tests: Fix a test failure. > msi/tests: Move a couple of tests from install.c to msi.c. > msi: Don't remove products if msidbUpgradeAttributesOnlyDetect is > set. > > Henri Verbeet (2): > wined3d: Add WINED3DFMT_INST to the format table. > configure: Include Xlib.h before Xrandr.h. > > Huw Davies (23): > riched20: Actually pass the ANSI string to the break proc. > riched20: Pass the character ptr and string length directly to the > break proc helper. > riched20: Pass a character ptr and length to SplitParagraph. > riched20: Pass character ptrs to the whitespace finding functions. > riched20: Move the text to the paragraph level. > riched20: Move SplitRun to wrap.c and make it static. > riched20: Move the split point calculation to wrap.c. > riched20: Move run extent calculation to wrap.c. > riched20: The run width is always passed so we don't need to > calculate it again. > riched20: Pass the run to the text drawing routine. > riched20: Move underline pen creation to a helper function. > riched20: Move y offset calculation to a helper function. > riched20: Move text color determination to a helper function. > riched20: Move the white space painting operations to a common > function. > riched20: Remove the special case tab handling now that we handle > that in draw_space. > riched20: Make it possible to perform point -> char conversion while > holding a context. > riched20: The string len parameter is no longer necessary. > riched20: Use the point from char helpers to calculate the selection > rectangle. > riched20: Simplify selection painting by drawing a selected run > twice with appropriate clipping. > riched20: Move underline drawing to a common function. > riched20: Move the range clamping to the helper now that the range > isn't used later on. > riched20: Move the password string hack to just before it's used. > riched20: Use ME_PointFromChar to calculate the caret position. > > Jacek Caban (13): > atl100: Added AtlComModuleRegisterClassObjects implementation (based > on AtlModuleRegisterClassObjects). > oleaut32: COM clean up of ITypeLib2 interface implementation. > mshtml: Use VT_ERROR as return type in DispCallFunc call. > oleaut32: Return error if VT_HRESULT is used as return type in > DispCallFunc. > urlmon: Clean up handling tested URLs. > urlmon: Fixed QueryInfo tests during BINDSTATUS_PROXYDETECTING > notification. > urlmon: Skip reported_url tests when binding to object. > urlmon: Fixed tests on IEs that report BINDSTATUS_DECODING. > oleaut32: Don't free a string in SysFreeString if it's already in > cache. > oleaut32: Added some SysFreeString tests. > urlmon: Allow available_bytes to be set before protocol_continue > call. > wininet: Set available bytes in InternetQueryDataAvailable even if > it ends up in async call. > vbscript: Added support for a few more constants. > > Jactry Zeng (4): > msvcr100: Added _wstat32i64 implementation. > msvcr100: Added _stat32i64 implementation. > msvcr80: Sync _wstat32i64 and _stat32i64 implementation. > msvcr90: Sync _wstat32i64 and _stat32i64 implementation. > > Jeff Latimer (2): > wmvcore: Add stub implementation for WMCreateEditor. > wmvcore: Fix a typo. > > Juan Lang (2): > iphlpapi: Fix HRESULT/win32 error code confusion. > cryptui: Use add_usage_to_oid correctly. > > Julian Rüger (1): > po: Update German translation. > > Jörg Höhle (4): > ntdll: Use the monotonic time counter also for timer queues. > winmm: More compatible midiIn/Out[Un]Prepare MHDR_* flag handling. > winmm: More compatible waveIn/Out[Un]Prepare WHDR_* flag handling. > winmm: Handle GetTickCount overflow in MIDI player. > > Ken Thomases (25): > winemac: Implement ScrollDC(). > winemac: Implement support for minimizing windows. > winemac: Build a map from Mac virtual key codes to Win32 vkeys and > scan codes based on Mac keyboard layout. > winemac: Rebuild key map when Mac keyboard layout changes. > winemac: Generate KEY_PRESS/RELEASE events from Cocoa key events. > user32: In TranslateMessage, handle ToUnicode returning 2 or more > chars. > winemac: Implement ToUnicodeEx(). > winemac: Ignore Cocoa's window frame adjustments for off-screen > windows. > winemac: Implement Beep(). > winemac: Implement MapVirtualKeyEx(). > winemac: Implement VkKeyScanEx(). > winemac: Implement GetKeyNameText(). > winemac: Implement fake support for the active Windows keyboard > layout. > winemac: Implement SysCommand() to prevent Alt keypresses from > activating system menu. > winemac: Add WineApplication methods to convert from Cocoa to Win32 > coordinate space. > winemac: Implement MOUSE_MOVED(_ABSOLUTE) events. > winemac: Process clicks in an inactive window in addition to > activating it. > secur32: On Mac, add support for cipher suites defined in 10.8 SDK. > user32: Improve mouse wheel scrolling in edit control. > user32: Improve mouse wheel scrolling in listbox control. > comctl32: Eliminate redundant local variable (gcWheelDelta == > -wheelDelta). > comctl32: Improve mouse wheel scrolling in listview control. > comctl32: Improve mouse wheel scrolling in treeview control. > winemac: Implement MOUSE_SCROLL events. > winemac: Prevent Cocoa from consuming certain key-down events. > > Ludger Sprenker (1): > windowscodecs: Add enum type WICTiffCompressionOption. > > Marcus Meissner (1): > comdlg32: Free files on errorpath (Coverity). > > Michael Stefaniuc (20): > ddraw/tests: Use assignment instead of memcpy to copy structs. > crypt32: Use assignment instead of memcpy to copy structs. > wined3d: Remove ERR() on HeapAlloc failure for small sizes known at > compile time. > ieframe: Let C look like C. > krnl386.exe16: Let C look like C. > gphoto2.ds: Let C look like C. > iphlpapi: Let C look like C. > qedit: Let C look like C. > d3drm: Let C look like C. > riched20: Let C look like C. > usp10: Let C look like C. > d3d10core: Remove ERR() on HeapAlloc failure for small sizes known > at compile time. > d3d8: Remove ERR() on HeapAlloc failure for small sizes known at > compile time. > d3d9: Remove ERR() on HeapAlloc failure for small sizes known at > compile time. > d3dx9_36: Remove ERR() on HeapAlloc failure for small sizes known at > compile time. > d3drm: Remove ERR() on HeapAlloc failure for small sizes known at > compile time. > amstream: Remove ERR() on HeapAlloc failure for small sizes known at > compile time. > d3dcompiler_43: Remove ERR() on HeapAlloc failure for small sizes > known at compile time. > dmime: Remove ERR() on HeapAlloc failure for small sizes known at > compile time. > wineoss.drv: Use offsetof to calculate the size of a struct with > variable length array. > > Nikolay Sivov (3): > oledb32: Add a table of init properties, support Persist Security > Info switch in init string. > xmllite: Store start data pointer for comments. > xmllite: Basic test for chunk value read vs full value read. > > Nozomi Kodama (1): > d3dx9: Use const instead of CONST. > > Piotr Caban (11): > jscript: Rename jsheap_t to heap_pool_t. > jscript: Don't use jsstr_t in JSRegExp structure. > jscript: Don't use jsstr_t in js_NewRegExp. > jscript: Move code operating on JSRegExp out of do_regexp_match_next > function. > jscript: Make context generic in internal regular expression > functions. > jscript: Moved generic regular expressions code to separate file. > vbscript: Added IRegExp2 stub. > vbscript: Added IRegExp2_QueryInterface tests. > vbscript: Added partial implementation of IDispatch methods in > IRegExp2. > vbscript: Added IRegExp2::Pattern implementation. > vbscript: Rename vbsheap to heap_pool. > > Roman Dadkov (3): > ole32/tests: Add test for SetClientSite. > mpr/tests: Test to check the return values of WNetGetUniversalNameA. > ole32/tests: Add test for OleRun. > > Stefan Dösinger (2): > wined3d: Use the ARBfp ffp pipeline only if ARBfp is supported. > wined3d: Reserve one vs constant on r200. > > Vijay Kiran Kamuju (1): > msvcr90: Update the manifest version. > > -- > Alexandre Julliard > julliard(a)winehq.org > >

2 1

Re: [PATCH] Support a protection scheme which reset the GS selector - v2
by Sergey Guralnik 17 Feb '13

17 Feb '13

On 2013-02-07 17:00, Alessandro Pignotti wrote: >+ opcode=(unsigned char*)context->Eip; >+ if (*opcode==0x65 && /* GS segment instruction prefix */ >+ context->SegGs!=ntdll_get_thread_data()->gs) Segment-override prefix may be preceded by repeat or operand-size override prefixes. So instruction that refers to GS-segment doesn't always starts with 0x65. -- Sergey

3 2

Re: [PATCH] ntdll: Support a protection scheme which reset the GS selector - v3
by Ken Thomases 17 Feb '13

17 Feb '13

On Feb 17, 2013, at 4:18 AM, Alessandro Pignotti wrote: > Third attempt > > ---------------- > > Hi everyone, > > I've found two different games: > > -) Of orcs and men > -) The testament of sherlock holmes > > which are using a protection scheme which reset the GS segment selector, > possibly to confuse virtual machines. Since on linux the GS selector is > used internally by glibc the game would crash as soon as a native glibc > method is called by wine. The proposed patch tries to workaround this > problem by restoring the GS to a known working value in wine's SEGFAULS > handler and try the instruction again. > > Regards, > Alessandro Pignotti > > <0001-ntdll-Support-a-protection-scheme-which-reset-the-GS-v3.patch> Didn't Alexandre already fix this with <http://source.winehq.org/git/wine.git/?a=commit;h=64ddb263d25dfceeeae8457a0…>? -Ken

1 0

Fwd: Re: Fix catalyst brain damage to speed up Falcon BMS 2x
by Stanisław Halik 16 Feb '13

16 Feb '13

Ben allowed me to forward this email. Kudos to him for all the knowledge his brainbox contains! -------- Original Message -------- Subject: Re: Fix catalyst brain damage to speed up Falcon BMS 2x Date: Sat, 16 Feb 2013 11:20:35 -0500 From: Ben Supnik <bsupnik(a)xsquawkbox.net> To: Stanislaw Halik <sthalik(a)misaki.pl> Hi Guys, I'm afraid I don't know enough about the _specific_ situation you guys are seeing. I can tell you guys a few things from my GL work: 1. The ATI OpenGL Linux team is pretty accessible; do you guys have anyone in the fglrx beta program? 2. What we found was that for stream-draw buffers that need to be orphaned, mapped, unmapped and drawn, there was a fixed overhead in the ATI drivers compared to NV; this 'performance gap' is cross-platform - both NV and ATI use the same GL stack (more or less) for Windows and Linux, and we saw the slow-down on both. 3. We originally were using map buffer (not map buffer range) with a NULL glBufferData to "orphan" the buffer (the equivalent of d3d map-discard). I think I tried MBR and it didn't fix it - both were expensive because the fundamental memory mapping operation was slow. 4. The slowness was in milliseconds, e.g. "this hits our fps by 20% or 30%" - but it wasn't "this is 3x slower because it stalled the GPU." So if you're seeing truly face-meltingly bad performance, like a total pipeline stall, you have a different bug. 5. As a general statement, the original glMapBuffer is subject to a lot of heuristic behavior in the drivers; app developers are very fast and loose with how they use it, so the driver vendors tend to try to make it do the fastest, most useful, least crash-y thing because the apps use it like monkeys on type-writers. By comparison, MBR came out later and has much more specific semantics for particular optimizations, as a result, the MBR implementation will often do exactly what you say, _even_ if it's slower. Getting even one flag wrong in MBR can cause it to hit a face-meltingly slow path. We worked around the perf cost of mapping a buffer on ATI hw by using pinned memory (but we do have a Linux-only bug where we get corrupt geometry with pinned memory - it works on Windows); I have some todo items to investigate the problem more thoroughly now. Cheers Ben On 2/16/13 5:47 AM, Stanislaw Halik wrote: > On 2013-02-16 09:04, Stefan Dösinger wrote: >> What you really want to do is figure out why GL_ARB_map_buffer_range >> is slow on fglrx, and make sure that the problem is really fglrx >> specific. I fixed a number of dynamic buffer performance problems in >> the past months, but there are still problems if we're falling back >> to draw_strided_slow for some reason, like fixed function material >> tracking. > > Thanks for reviewing this. > > Going to ask Ben Supnik from Laminar Research (X-Plane developer) and > BCC him, since he has apparently run into the same issue. There's much > info of fglrx woes (not really Linux specific, either) on > http://developer.x-plane.com/ > > He said publicly to be in contact with AMD themselves, and been friendly > to OSS by releasing an X-Plane Linux version, as well as overall cool > fellow. > > Ben, Please help! > >> Other than being wrong conceptually, you're disabling dynamic buffers >> the wrong way: The "proper" way would be to add a quirk to the >> quirk_table in directx.c that removes ARB_map_buffer_range from the >> list of supported extensions if the driver vendor is AMD. > > Like this? Patch attached. > > I've run into hard GPU hangs with fglrx 13.2, no VT switch either. This > helps: > > [Software\\Wine\\Direct3D] > "DirectDrawRenderer"="gdi" > "Multisampling"="disabled" > "OffscreenRenderingMode"="fbo" > "UseGLSL"="enabled" > > Lack of GLSL disables HDR apparently. > > Without GDI, there's some nasty display corruption on FBOs. > > Also Catalyst likes to hang display when switching from 3D to 2D and VT > switch helps. > > But with all this busywork, performance is near-native. Catalyst at > least supports indirect addressing (whatever that means) and doesn't > choke on > 128 temps... FYI Mesa bug submitted: > > https://bugs.freedesktop.org/show_bug.cgi?id=55420 > > -sh > -- Scenery Home Page: http://scenery.x-plane.com/ Scenery blog: http://www.x-plane.com/blog/ Plugin SDK: http://www.xsquawkbox.net/xpsdk/ X-Plane Wiki: http://wiki.x-plane.com/ Scenery mailing list: x-plane-scenery(a)yahoogroups.com Developer mailing list: x-plane-dev(a)yahoogroups.com

1 0

Re: Fix catalyst brain damage to speed up Falcon BMS 2x
by Stefan Dösinger 16 Feb '13

16 Feb '13

Am 16.02.2013 um 08:39 schrieb Stanisław Halik <sthalik(a)misaki.pl>: > + dynamic_buffer_ok = gl_info->supported[APPLE_FLUSH_BUFFER_RANGE] || > + (!gl_info->[GL_AMDX_debug_output] && gl_info->supported[ARB_MAP_BUFFER_RANGE]); What you really want to do is figure out why GL_ARB_map_buffer_range is slow on fglrx, and make sure that the problem is really fglrx specific. I fixed a number of dynamic buffer performance problems in the past months, but there are still problems if we're falling back to draw_strided_slow for some reason, like fixed function material tracking. If you absolutely have to do some fglrx hacks for performant dynamic buffers, look into GL_AMD_pinned_memory. I'd prefer not to use this extension though. Other than being wrong conceptually, you're disabling dynamic buffers the wrong way: The "proper" way would be to add a quirk to the quirk_table in directx.c that removes ARB_map_buffer_range from the list of supported extensions if the driver vendor is AMD.

4 8

Re: Fix catalyst brain damage to speed up Falcon BMS 2x
by Stanislaw Halik 16 Feb '13

16 Feb '13

Updated patch following review.

1 0

Re: [website] Simplified Chinese translation for release 1.5.24
by Jactry Zeng 16 Feb '13

16 Feb '13

Oops...I made a mistake, this is the patch. So sorry :-(

1 0

mmsystem 16bit questions
by Joerg-Cyril.Hoehle＠t-systems.com 15 Feb '13

15 Feb '13

Hi, I noticed some dubious differences among winmm16 functions. I believe there should be none, but perhaps I'm missing some 16bit idiosyncrasies? Please compare waveOutPrepareHeader16, Unprepare16 and waveOutWrite16: http://source.winehq.org/source/dlls/mmsystem.dll16/mmsystem.c#L1287 UINT16 WINAPI waveOutUnprepareHeader16(HWAVEOUT16 hWaveOut, /* [in] */ SEGPTR lpsegWaveOutHdr, /* [???] */ UINT16 uSize) /* [in] */ { [...] return MMSYSTDRV_Message(..., WODM_UNPREPARE, lpsegWaveOutHdr, uSize); UINT16 WINAPI waveOutWrite16(HWAVEOUT16 hWaveOut, /* [in] */ LPWAVEHDR lpsegWaveOutHdr, /* [???] NOTE: SEGPTR */ UINT16 uSize) /* [in] */ { [...] return MMSYSTDRV_Message(..., WODM_WRITE, (DWORD_PTR)lpsegWaveOutHdr, uSize); The mmsystem.spec knows no difference: pascal waveOutPrepareHeader(word segptr word) waveOutPrepareHeader16 pascal waveOutUnprepareHeader(word segptr word) waveOutUnprepareHeader16 pascal waveOutWrite(word segptr word) waveOutWrite16 I believe: - SEGPTR should be used everywhere the .spec says; - (DWORD_PTR) cast then becomes obsolete. Shall I submit a patch? Furthermore, how to perform a 16bit NULL pointer check? waveInUnPrepareHeader16 uses: LPWAVEHDR lpWaveOutHdr = MapSL(lpsegWaveOutHdr); if (lpWaveInHdr == NULL) return MMSYSERR_INVALPARAM; waveInAddBuffer16 directly uses: if (lpsegWaveInHdr == NULL) return MMSYSERR_INVALPARAM; Last but not least, some but not all of the functions check for a NULL pointer. Why only some? When the check is missing, it looks like MMSYSTDRV_Message will crash when calling: map = drvtype->mapmsg16to32W(msg, &param1, &param2); http://source.winehq.org/source/dlls/mmsystem.dll16/message16.c?v=wine-1.5.… I don't know whether the check should occur or if we should let the app crash (all I know is that native MCI_PLAY crashes on w2k when given MCI_FROM/TO or callback flags with a NULL MCI_PARMS pointer, whereas Wine returns INVALPARAM, whereas a NULL pointer is ok in native when not setting any flags that require access via the pointer). If the check shall be done, then I believe that it should occur centrally in MMSYSTDRV_Message, not within the wave/midi* functions, such that if an app calls e.g. waveInMessage16(WIDM_UNPREPARE) that dispatches to MMSYSTDRV_Message, the check is not bypassed. BTW, MMSYSTDRV_WaveOut_UnMap16To32W forgets to deallocate the wh32 when waveOutPrepare fails. Same with wave in and MIDI. Thank you for 16bit enlightening, Jörg Höhle

2 1

Re: [PATCH 2/5] vbscript: Added IRegExp2_QueryInterface tests
by Marvin 15 Feb '13

15 Feb '13

2 1

MSI signing
by Marc-André Lureau 15 Feb '13

15 Feb '13

Hello! PE signing is already publicly documented by Microsoft, and we have an open-source tool (osslsigncode) capable of signing PE and Cabinet. A signed MSI dump easily reveals that the signature is under the '\005DigitalSignature' stream, and it's a regular pkcs#7 signature, whose content info type is 1.3.6.1.4.1.311.2.1.30 (SPC_SIPINFO_OBJID whatever that might be). To ease debugging, I added asn1parse of signature to msitools with "msidump --signature". You can also compare two MSI signatures with msidiff (http://git.gnome.org/browse/msitools) I found this blog http://qistoph.blogspot.com.es/2012/01/manual-verify-pkcs7-signed-data-with… very helpful, so my testing setup currently is: - self signed pkcs12 certificate - minimal "starting point" MSI crafted with msibuild mini.msi -s '', modified with hex editor for 0 length streams - signtool from 7.0A sdk, used simply "sign /f cert.pfx /v /p passphrase mini.msi" >From there, I get a "canonical" message digest sha1 value of 98C090F33284468D64F30EE7B8C00894135D54C8. Only the stream content seems to be used to compute it. All modifications I can think of to the MSI do not seem to alter that sha1, none of: OLECF internal data, stream name, number of (empty) streams, total file length, position of the streams, or whatever I could tweak in the OLECF data seems to change the resulting sha1. So it very much looks like an initial "seed", somehow? Based on my limited knowledge of sha1 computation, I have tried initializing the sha1 by modifying its initial value manually (the h values) to match with the "canonical" sha1, and added 1 byte to a stream. I was hoping to get the same result, but that didn't work that way. It could be that my modified sha1 is wrong, or that they apply further transformation on the stream content before computing the checksum, although this would be surprising since they don't do that for PE or cabinet file. Any suggestion or help is welcome! -- Marc-André Lureau

1 1

64bit integer overflow in 3 days
by Joerg-Cyril.Hoehle＠t-systems.com 15 Feb '13

15 Feb '13

Hi, You believe that 64bit integers give you plenty of room and time to avoid trouble like Y2k and Y2033, don't you? The trouble is, these large data types make using other large numbers more attractive than they used to be, e.g. counting time in 100 nanosecond units, representing one second as 10000000. Combining large numbers will cause an overflow sooner than you think. Specifically, the pattern X * numerator / denominator is troublesome. Let's analyse winealsa:AudioClock_GetPosition: QueryPerformanceCounter(&stamp); QueryPerformanceFrequency(&frequency); *qpctime = (stamp.QuadPart * (INT64)10000000) / frequency.QuadPart; QueryPerformanceCounter (QPC) returns time in an unspecified unit (frequency) and needs to convert that into 100 nanosecond units. Graphically, with g - garbage bits, s - shifted, X - valid bits: 6 5 4 3 2 1 4321098765432109876543210987654321098765432109876543210987654321 <<< shift left, multiply divide, shift right >>> Assuming QPC returns 64 valid bits: XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX QPC Multiply with (INT64)10000000 (log2 ~ 23): XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXsssssssssssssssssssssss *10000000 Divide by frequency (happens to be 10000000 too): gggggggggggggggggggggggXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX /10000000 (100ns) You start to see how the number of valid bits already shrunk a lot. Now you don't care about nanoseconds, you want seconds. Divide again: ggggggggggggggggggggggggggggggggggggggggggggggXXXXXXXXXXXXXXXXXX /10000000 (seconds) Ouch, only 18bits left to express seconds! How many days is that? (/ (ash 1 18) 3600.0 24) -> ~3.03 So after 3 days in Wine, mmdevapi's GetPosition clock has wrapped. 3 days is not much compared to over 50000 years that 64 bit integers allow to express with 100 nanosecond units, isn't it? I don't know if GetPosition is related to bug #26968 where a user reports that foobar stops playing audio after 4-7 days, but you get a clue. Cross-check: 0x40000 is 1^18 seconds ~ 3 days 0x2625A000000 same in 100 nanosecond units 0x16BCC41E900000000 the * (INT64)10000000 part in winealsa ^1 bit unsigned overflow (more with signed integers) 0x6BCC41E900000000 clipped 0xB4DAD65435 after /10000000, a value already reached after ~21 hours There are even less bits, because LARGE_INTEGER is signed. There are several instances of that X * p / q pattern in Wine. Examples: ntdll/threadpool.c:queue_current_time: return now.QuadPart * 1000 / freq.QuadPart; ntdll/kernel_main.c:GetTickCount64: return counter.QuadPart * 1000 / frequency.QuadPart; IOW, GetTickCount64 does not offer 64 valid bits. ntdll/time.c:monotonic_counter: return mach_absolute_time() * timebase.numer / timebase.denom / 100; This means that on MacOS, QPC may overflow even earlier because it does not offer 64 valid bits. I've prepared a few patches to fix some of these. Generally, I don't know what to recommend. Use double floating point intermediates and loose 11 bits for the exponent but at least get the magnitude right? Regards, Jörg Höhle

2 1

Re: d3dx9: Specify arguments in trace to ease debugging
by Alexandre Julliard 15 Feb '13

15 Feb '13

Nozomi Kodama <nozomi.kodama(a)yahoo.com> writes: > Hello. > > I took advantage of this minor patch to clean up and make uniform the > coding style of this file. Please don't do that. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: vbscript: Added support for a few more constants
by Marvin 15 Feb '13

15 Feb '13

Hi, While running your changed tests on Windows, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at http://testbot.winehq.org/JobDetails.pl?Key=24408 Your paranoid android. === WNT4WSSP6 (32 bit) === No test summary line found === W2KPROSP4 (32 bit) === No test summary line found === WXPPROSP3 (32 bit) === No test summary line found === W2K3R2SESP2 (32 bit) === No test summary line found === WVISTAADM (32 bit) === No test summary line found === W2K8SE (32 bit) === No test summary line found === W7PRO (32 bit) === No test summary line found === W7PROX64 (32 bit) === No test summary line found === TEST64_W7SP1 (32 bit) === No test summary line found === W7PROX64 (64 bit) === No test summary line found === TEST64_W7SP1 (64 bit) === No test summary line found

1 0

Re: d3dx9: Specify arguments in trace to ease debugging
by Rico Schüller 15 Feb '13

15 Feb '13

On 15.02.2013 08:20, Nozomi Kodama wrote: > @@ -721,9 +739,10 @@ D3DXMATRIX* WINAPI D3DXMatrixTransformation(D3DXMATRIX *pout, const D3DXVECTOR3 > > @@ -797,69 +821,70 @@ D3DXMATRIX* WINAPI D3DXMatrixTransformation2D(D3DXMATRIX *pout, const D3DXVECTOR Those two functions need a speed improvement. Normally we change only the code (and it's style) if we touch that area. I'm not a fan of changing that in one go in a huge patch, as it might introduce new mistakes, see below. I'm not sure if I caught all issues. > + rot.w = cos(rotation/2.0f); rot.w = cos(rotation / 2.0f); > > - TRACE("iface %p\n", iface); > + TRACE("iface %p, %pm\n", iface, pm); Typo. > + TRACE("iface %p, x %f, y %f, z %f\n", iface, x, y, z); > > + TRACE("(pout %p, pvpoint %p, pvnormal %p)\n", pout, pvpoint, pvnormal); Well, I would use the same style for both... I'd prefer the first one. Cheers Rico

1 0

Re: [PATCH 2/2] wininet: Set available bytes in InternetQueryDataAvailable even if it ends up in async call
by Marvin 14 Feb '13

14 Feb '13

2 1

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

wine-devel February 2013