October 2012 - Wine-devel - list.winehq.org

Re: [PATCH] services: Fix non common sizeof usages. (resend)
by Alexandre Julliard 31 Oct '12

31 Oct '12

Christian Costa <titan.costa(a)gmail.com> writes: > Use parentesis is more common in Wine code and this makes all sizeof conform to file style anyway. You can clean this up if you are changing the code for other reasons, but otherwise there's no reason to do this. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: hh: Avoid handle leak (coverity)
by Alexandre Julliard 31 Oct '12

31 Oct '12

Frédéric Delanoy <frederic.delanoy(a)gmail.com> writes: > @@ -27,9 +27,13 @@ int WINAPI WinMain(HINSTANCE hInst, HINSTANCE hPrevInst, LPSTR cmdline, int cmds > > HMODULE hModule; > DOWINMAIN *doWinMain; > + int res; > > hModule = LoadLibraryA("hhctrl.ocx"); > doWinMain = (DOWINMAIN*) GetProcAddress(hModule, "doWinMain"); > > - return doWinMain(hInst, cmdline); > + res = doWinMain(hInst, cmdline); > + FreeLibrary(hModule); > + > + return res; Freeing things on process exit is not useful. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH] user32: Fix error handling in MapWindowPoints, ClientToScreen and ScreenToClient and add tests for them. (try 6) (resend)
by Dmitry Timoshkov 31 Oct '12

31 Oct '12

Christian Costa <titan.costa(a)gmail.com> wrote: > Based ont a patch of Rico Schuller. I'd suggest to make sure that the test with appropriate todo_wine statements is accepted first, and only then start thinking about a proper fix, or even leave creating the fix to someone else. -- Dmitry.

2 1

What parts of gettext does Wine need at runtime?
by Scott Ritchie 31 Oct '12

31 Oct '12

Is libgettextpo0 (in 32 and 64 bit) sufficient? Or do we need the sort of utilities in the gettext-base package? I'm trying to fix https://bugs.launchpad.net/ubuntu/precise/+source/gettext/+bug/954029 which was largely caused by my overcautiousness here.

2 1

Re: [PATCH 3/4] msvcp90: in istream<>::tellg don't use sentry
by Piotr Caban 30 Oct '12

30 Oct '12

On 10/30/12 10:16 PM, Daniel Lehman wrote: + tpos.pos = 0xdeadbeefdeadbeef; Long long constants are not supported in C89. You can use 0xdeadbeef here.

1 0

xapofx1_3.dll?
by Joerg-Cyril.Hoehle＠t-systems.com 30 Oct '12

30 Oct '12

Hi, why does Wine include a stub xapofx1_1.dll but no xapofx1_3.dll? Installing Wallace & Gromit 102 "The Last Resort" / "Urlaub unter Tage" adds xapofx1_3.dll XAudio2_4.dll and X3DAudio1_6.dll to system32 (though IIRC I asked the installer not to care about DirectX). This app loads none of these libraries (well, I solely tested winecfg's default xp mode, not w7. Presumably those dlls are only relevant since Vista with mmdevapi and the app uses DSound on older systems). http://bugs.winehq.org/show_bug.cgi?id=27279 is about apps linked against this dll such that they won't even start. Regards, Jörg Höhle

2 1

Re: scrrun: Add basic IDictionary getHasVal support
by Piotr Caban 30 Oct '12

30 Oct '12

Hi, On 10/30/12 09:51, Alistair Leslie-Hughes wrote: +static LONG create_hash_val(BSTR name) +{ + LONG hash = 0; + int len = SysStringLen(name); + int i; + + for(i =0; i < len; i++) + { + hash += name[i]; + } + + return hash; +} This hashing function will have lots of conflicts when used in hash-table. It's quite easy to guess how it's done in native dll: for(i=0; i<len; i++) hash = (hash<<4) + hash + name[i]; hash %= 1201; It would be nice to add some tests for strings containing '\0' characters and for other VARIANT types. I don't know how scrrun stores data internally, if it's using a hash-table, the function may change after adding some elements.

1 0

Re: ntdll: Do not execute callbacks past DeleteTimer(INVALID_HANDLE_VALUE)
by Alexandre Julliard 30 Oct '12

30 Oct '12

<Joerg-Cyril.Hoehle(a)t-systems.com> writes: > You know this patch from March this year. Initially, this patch was > the first of 2-3. I decided to release it on its own after finding > out yesterday that in July, Fran,cois Gouget added a todo_wine for a > bug that I fixed monthes earlier! Now I had to augment my patch > with a one-liner removing that todo_wine from kernel32/tests. > > Let me thank him for writing a test case that demonstrates the > effectiveness of my patch. Or lack thereof ;-) ../../../tools/runtest -q -P wine -M kernel32.dll -T ../../.. -p kernel32_test.exe.so sync.c && touch sync.ok sync.c:808: Test failed: Timer callback 0 expected 1 got 0 sync.c:1166: Tests skipped: one-time initialization API not supported make[1]: *** [sync.ok] Error 1 -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH 4/5] include: Don't add d3d8 LP* interface typedefs for __WINESRC__.
by Alexandre Julliard 30 Oct '12

30 Oct '12

Henri Verbeet <hverbeet(a)codeweavers.com> writes: > @@ -1128,6 +1105,21 @@ DECLARE_INTERFACE_(IDirect3DDevice8,IUnknown) > #define IDirect3DDevice8_DeletePatch(p,a) (p)->DeletePatch(a) > #endif > > +#ifndef __WINESRC__ > +typedef struct IDirect3D8 *LPDIRECT3D8; > +typedef struct IDirect3DDevice8 *LPDIRECT3DDEVICE8; > +typedef struct IDirect3DResource8 *LPDIRECT3DRESOURCE8, *PDIRECT3DRESOURCE8; > +typedef struct IDirect3DVertexBuffer8 *LPDIRECT3DVERTEXBUFFER8, *PDIRECT3DVERTEXBUFFER8; > +typedef struct IDirect3DVolume8 *LPDIRECT3DVOLUME8, *PDIRECT3DVOLUME8; > +typedef struct IDirect3DSwapChain8 *LPDIRECT3DSWAPCHAIN8, *PDIRECT3DSWAPCHAIN8; > +typedef struct IDirect3DSurface8 *LPDIRECT3DSURFACE8, *PDIRECT3DSURFACE8; > +typedef struct IDirect3DIndexBuffer8 *LPDIRECT3DINDEXBUFFER8, *PDIRECT3DINDEXBUFFER8; > +typedef struct IDirect3DBaseTexture8 *LPDIRECT3DBASETEXTURE8, *PDIRECT3DBASETEXTURE8; > +typedef struct IDirect3DTexture8 *LPDIRECT3DTEXTURE8, *PDIRECT3DTEXTURE8; > +typedef struct IDirect3DCubeTexture8 *LPDIRECT3DCUBETEXTURE8, *PDIRECT3DCUBETEXTURE8; > +typedef struct IDirect3DVolumeTexture8 *LPDIRECT3DVOLUMETEXTURE8, *PDIRECT3DVOLUMETEXTURE8; > +#endif I don't think we want to do that sort of thing, otherwise we'll end up with ifdef __WINESRC__ all over the standard headers. If we really want to disable interface typedefs this should be done in a generic way, preferably when generating headers from idl. -- Alexandre Julliard julliard(a)winehq.org

3 3

Re: [PATCH] kernel32: Implement CompareStringOrdinal. (try 2)
by Nikolay Sivov 30 Oct '12

30 Oct '12

On 10/30/2012 01:43, Christian Costa wrote: > Try 2: > - Don't use CompareStringEx. > - Add tests > --- > dlls/kernel32/kernel32.spec | 1 + > dlls/kernel32/locale.c | 28 ++++++++++++++++++++ > dlls/kernel32/tests/locale.c | 60 +++++++++++++++++++++++++++++++++++++++++- > 3 files changed, 88 insertions(+), 1 deletion(-) > > diff --git a/dlls/kernel32/kernel32.spec b/dlls/kernel32/kernel32.spec > index b7efa0f..0bd1adc 100644 > --- a/dlls/kernel32/kernel32.spec > +++ b/dlls/kernel32/kernel32.spec > @@ -201,6 +201,7 @@ > @ stdcall CompareStringA(long long str long str long) > @ stdcall CompareStringW(long long wstr long wstr long) > @ stdcall CompareStringEx(wstr long wstr long wstr long ptr ptr long) > +@ stdcall CompareStringOrdinal(wstr long wstr long long) > @ stdcall ConnectNamedPipe(long ptr) > @ stub ConsoleMenuControl > @ stub ConsoleSubst > diff --git a/dlls/kernel32/locale.c b/dlls/kernel32/locale.c > index c41442c..07eb688 100644 > --- a/dlls/kernel32/locale.c > +++ b/dlls/kernel32/locale.c > @@ -3047,6 +3047,34 @@ INT WINAPI CompareStringA(LCID lcid, DWORD flags, > return ret; > } > > +/****************************************************************************** > + * CompareStringOrdinal (KERNEL32.@) > + */ > +INT WINAPI CompareStringOrdinal(const WCHAR *str1, INT len1, const WCHAR *str2, INT len2, BOOL ignore_case) > +{ > + int ret, len; > + > + if (!str1 || !str2) > + { > + SetLastError(ERROR_INVALID_PARAMETER); > + return 0; > + } > + if (len1 < 0) len1 = strlenW(str1); > + if (len2 < 0) len2 = strlenW(str2); > + > + len = min(len1, len2); > + if (ignore_case) > + ret = strncmpiW(str1, str2, len); > + else > + ret = strncmpW(str1, str2, len); > + > + if ((ret < 0) || (!ret && (len1 < len2))) > + return CSTR_LESS_THAN; > + else if ((ret > 0) || (!ret && (len1 > len2))) > + return CSTR_GREATER_THAN; > + return CSTR_EQUAL; > +} > + This is almost the same as your first try. You're testing a trivial case only - ASCII range. I feel like it should behave more like memcmp for case insensitive comparison, that's what documentation mildly suggests.

3 6