May 2010 - Wine-Devel - list.winehq.org

Re: [2/5] rundll32: Fail if a rundll32 window cannot be created.

by Alexandre Julliard

Andrew Nguyen <anguyen(a)codeweavers.com> writes: > --- > programs/rundll32/rundll32.c | 16 +++++++++++++--- > 1 files changed, 13 insertions(+), 3 deletions(-) This will prevent initializing the prefix without an X display. Why do you need this? -- Alexandre Julliard julliard(a)winehq.org

May 3, 2010

Re: Valgrind's wine_cp_wcstombs warnings

by Dan Kegel

Hi Joerg, did you try the --track-origins option? That can sometimes give more detail (see e.g. the traces in this old message): On Thu, Jul 31, 2008 at 10:14 AM, Dan Kegel <dank(a)kegel.com> wrote: > Hi Colin, > your change > http://source.winehq.org/git/wine.git/?a=commitdiff;h=c2c1a55f1c08f35ae02b8… > seems to have triggered some valgrind warnings, > http://kegel.com/wine/valgrind/logs/2008-07-31-07.46/vg-kernel32_codepage-d… > e.g. > > + Use of uninitialised value of size 4 > + at wcstombs_sbcs (wctomb.c:186) > + by wine_cp_wcstombs (wctomb.c:457) > + by WideCharToMultiByte (locale.c:1936) > + by test_string_conversion (codepage.c:288) > + by func_codepage (codepage.c:351) > + by run_test (test.h:488) > + by main (test.h:537) > + Uninitialised value was created by a stack allocation > + at test_string_conversion (codepage.c:216) > > + Use of uninitialised value of size 4 > + at wcstombs_sbcs_slow (wctomb.c:246) > + by wine_cp_wcstombs (wctomb.c:453) > + by WideCharToMultiByte (locale.c:1936) > + by test_string_conversion (codepage.c:288) > + by func_codepage (codepage.c:352) > + by run_test (test.h:488) > + by main (test.h:537) > + Uninitialised value was created by a stack allocation > + at test_string_conversion (codepage.c:216) > > Could you have a look? >

May 3, 2010

Valgrind's wine_cp_wcstombs warnings

by Joerg-Cyril.Hoehle＠t-systems.com

Hi, running the MCI tests with Valgrind generates a lot of output like follows ==13170== Use of uninitialised value of size 4 ==13170== at 0x4035369: wine_cp_wcstombs (wctomb.c:147) ...[every line from 148 to 161] ==13170== Use of uninitialised value of size 4 ==13170== at 0x403550A: wine_cp_wcstombs (wctomb.c:162) I suspect the A<->W mapping is still not fully correct (I know at least one more case, but I'm not sure it gets triggered by the tests) but perhaps it's also the W->UNIX.UTF-8 conversions internal to Wine which cause these hits. How do I tell? Is there any way to get more precise reporting for the above messages (e.g. a backtrace)? The present output is not helpful at identifying the origin of the uninitialised read. Perhaps Valgrind is broken? Why does it mention "size 4" when all access in the offending function is either "char *" or "short *" according to the source http://source.winehq.org/source/libs/wine/wctomb.c#L145 ? Thanks for your help, Jörg Höhle

May 3, 2010

Re: ddraw: Fix missing D3D misc caps

by Stefan Dösinger

Am 02.05.2010 um 22:07 schrieb Ričardas Barkauskas: > Fixes missing D3D misc caps needed by Independence War 2: Edge Of Chaos (bug 7012) The patch seems ok

May 3, 2010

Re: [ddraw] Fix discrepancy in alpha mask for WINED3DFMT_B2G3R3A8_UNORM.

by Stefan Dösinger

Am 02.05.2010 um 08:26 schrieb Vincent Pelletier: > > PixelFormat_DD2WineD3D returns WINED3DFMT_B2G3R3A8_UNORM when alpha mask > is 0xFF00, which is consistent with constant name (A8). The patch looks ok to me.

May 3, 2010

Re: mshtml/tests: Fix return value of ActiveScript_SetScriptState.

by Jacek Caban

Hi Gerald, On 5/2/10 9:16 PM, Gerald Pfeifer wrote: > IActiveScriptSite_OnStateChange is described to return S_OK upon > success, so instead of ignoring its return value and unconditionally > returning S_OK it strikes me that we should return its result instead. > In this case returning S_OK is fine, but there is missing test for OnStateChange return value. Jacek

May 3, 2010

Re: gdi32/tests: Remove two variables which are not really used from test_clipping.

by Gerald Pfeifer

On Mon, 3 May 2010, Nikolay Sivov wrote: > Bitmap are useless after this change. Darn, I may have sent the wrong version of the patch. Let me look into this tonight or tomorrow. You have sharp eyes, Nikolay! :-) Gerald

May 3, 2010

Re: gdi32/tests: Remove two variables which are not really used from test_clipping.

by Nikolay Sivov

On 5/3/2010 03:48, Gerald Pfeifer wrote: > --- > dlls/gdi32/tests/bitmap.c | 4 ---- > 1 files changed, 0 insertions(+), 4 deletions(-) > > diff --git a/dlls/gdi32/tests/bitmap.c b/dlls/gdi32/tests/bitmap.c > index 882d548..b570a8a 100644 > --- a/dlls/gdi32/tests/bitmap.c > +++ b/dlls/gdi32/tests/bitmap.c > @@ -2810,9 +2810,7 @@ static void test_GdiAlphaBlend(void) > static void test_clipping(void) > { > HBITMAP bmpDst; > - HBITMAP oldDst; > HBITMAP bmpSrc; > - HBITMAP oldSrc; > HRGN hRgn; > LPVOID bits; > BOOL result; > @@ -2830,11 +2828,9 @@ static void test_clipping(void) > > bmpDst = CreateDIBSection( hdcDst,&bmpinfo, DIB_RGB_COLORS,&bits, NULL, 0 ); > ok(bmpDst != NULL, "Couldn't create destination bitmap\n"); > - oldDst = SelectObject( hdcDst, bmpDst ); > > bmpSrc = CreateDIBSection( hdcSrc,&bmpinfo, DIB_RGB_COLORS,&bits, NULL, 0 ); > ok(bmpSrc != NULL, "Couldn't create source bitmap\n"); > - oldSrc = SelectObject( hdcSrc, bmpSrc ); > > result = BitBlt( hdcDst, 0, 0, 100, 100, hdcSrc, 100, 100, SRCCOPY ); > ok(result, "BitBlt failed\n"); > Bitmap are useless after this change.

May 3, 2010

Re: [PATCH] Fix Unicode escape string occurring when calling GetClipboardData(CF_TEXT).

by Dmitry Timoshkov

Sin-ta Hsiea <ibmibmibm.tw(a)gmail.com> wrote: > Here's a patch that fetch UTF8 string, and convert to local codepage. > + count = WideCharToMultiByte(CP_OEMCP, 0, lpwstr, -1, NULL, 0, NULL, NULL); > + hText = GlobalAlloc(GMEM_MOVEABLE | GMEM_DDESHARE, count); > + > + if (hText) > + { > + CHAR *text = GlobalLock(hText); > + WideCharToMultiByte(CP_OEMCP, 0, lpwstr, -1, text, -1, NULL, NULL); > + GlobalUnlock(hText); > + } CP_OEMCP is not a local codepage. -- Dmitry.

May 3, 2010

Re: [Wine] Potential bug in RtlRegisterWait()

by Joakim Tjernlund

Charles Davis <cdavis(a)mymail.mines.edu> wrote on 2010/05/03 01:18:25: > > On 5/2/10 3:02 PM, Joakim Tjernlund wrote: > > > > Noticed that RtlRegisterWait() will approve a zero timeout value(Milliseconds == 0) > > > > This makes ies4linux spin the CPU 100% when visiting a > > https page. The below crude patch fixes it: > > > > --- dlls/ntdll/threadpool.c.org 2010-05-02 22:37:08.000000000 +0200 > > +++ dlls/ntdll/threadpool.c 2010-05-02 22:46:45.000000000 +0200 > > @@ -427,6 +427,9 @@ > > > > TRACE( "(%p, %p, %p, %p, %d, 0x%x)\n", NewWaitObject, Object, Callback, > Context, Milliseconds, Flags ); > > > > + if (!Milliseconds) > > + return RPC_NT_INVALID_TIMEOUT; > > + > > wait_work_item = RtlAllocateHeap( GetProcessHeap(), 0, sizeof(*wait_work_item) ); > > if (!wait_work_item) > > return STATUS_NO_MEMORY; > > > > On a related note, whitout this patch wineserver also goes into an endless > > loop, doing brk() calls over and over until memory is exausted. Seems > > like a bug that too. > This looks like something you should post to wine-devel. Right, posting there but with a new patch. I think the orginal fix was wrong. Here is a new one: >From 2f5e169e286265bb82b17f76023c463bc2a96703 Mon Sep 17 00:00:00 2001 From: Joakim Tjernlund <Joakim.Tjernlund(a)transmode.se> Date: Mon, 3 May 2010 09:03:09 +0200 Subject: [PATCH] RegisterWaitForSingleObjectX(), fix tmo == 0 Timeout == 0 is special, from the manual: The time-out interval, in milliseconds. The function returns if the interval elapses, even if the object's state is nonsignaled. If dwMilliseconds is zero, the function tests the object's state and returns immediately. If dwMilliseconds is INFINITE, the function's time-out interval never elapses. 1) Only call Callback iff signaled 2) Implies WT_EXECUTEONLYONCE. 3) Callback should be called before returning. This is a crude fix not impl. 3) but is enough to get ies4linux around the 100% CPU problem. Not sure tmo == 0 implies returning a NULL ptr for phNewWaitObject too. --- dlls/ntdll/threadpool.c | 14 ++++++++++---- 1 files changed, 10 insertions(+), 4 deletions(-) diff --git a/dlls/ntdll/threadpool.c b/dlls/ntdll/threadpool.c index 44afb01..159baee 100644 --- a/dlls/ntdll/threadpool.c +++ b/dlls/ntdll/threadpool.c @@ -373,11 +373,17 @@ static DWORD CALLBACK wait_thread_proc(LPVOID Arg) wait_work_item->Context ); TimerOrWaitFired = TRUE; } - wait_work_item->CallbackInProgress = TRUE; - wait_work_item->Callback( wait_work_item->Context, TimerOrWaitFired ); - wait_work_item->CallbackInProgress = FALSE; + /* Milliseconds == 0 is special, only call Callback iff + signaled and treat as WT_EXECUTEONLYONCE */ + if (wait_work_item->Milliseconds || !TimerOrWaitFired) + { + wait_work_item->CallbackInProgress = TRUE; + wait_work_item->Callback( wait_work_item->Context, TimerOrWaitFired ); + wait_work_item->CallbackInProgress = FALSE; + } - if (wait_work_item->Flags & WT_EXECUTEONLYONCE) + if (!wait_work_item->Milliseconds || + wait_work_item->Flags & WT_EXECUTEONLYONCE) break; } else -- 1.6.4.4

May 3, 2010