February 2011 - Wine-devel - list.winehq.org

re: ntoskrnl.exe: unimplemented function ntoskrnl.exe.IoGetDeviceInterfaces
by Dan Kegel 02 Feb '11

02 Feb '11

Qian wrote: >1. If a function is implemented in wine dll as a stup, then usually >this function won't cause crash, >and the wine debug will output "fix:me ...stup" rather then >"Unimplemented function", is it right? I think "stub" is somewhat ambiguous, it can mean either something without an implementation, or something with a very minimal implementation. But yes, if something is listed in dlls/*/*.spec as a stub, you get 'Unimplemented function', and if something is listed in grep FIXME.*stub dlls/*/*.c as a stub, you get "fixme: ... stub". >2. If a function is listed in the Wine API guide, >http://source.winehq.org/WineAPI/ , and marked as stup, >then this function must be implemented as a stup in wine source code, >is it right? The Wine API guide is not a very reliable guide, it might be out of date. Better to just look in the source code. - Dan

2 1

Re: crypt32: Fix quoting-behaviour
by Juan Lang 01 Feb '11

01 Feb '11

Hi Christian, I'm afraid this patch is incorrect. In particular, - chars = CertRDNValueToStrA( - rdn->rgRDNAttr[j].dwValueType, - &rdn->rgRDNAttr[j].Value, psz ? psz + ret : NULL, - psz ? csz - ret : 0); + if ((dwStrType & 0x000000ff) == CERT_X500_NAME_STR) + chars = quote_string_a( + rdn->rgRDNAttr[j].dwValueType, + &rdn->rgRDNAttr[j].Value, psz ? psz + ret : NULL, + psz ? csz - ret : 0); + else + chars = CertRDNValueToStrA( + rdn->rgRDNAttr[j].dwValueType, + &rdn->rgRDNAttr[j].Value, psz ? psz + ret : NULL, + psz ? csz - ret : 0); This change (and the corresponding one to CertNameToStrW) is incorrect. I tested a couple of strings with CERT_SIMPLE_NAME_STR, and they are also quoted. I can try to take these over if you like, or you can continue to work on them. Let me know what you prefer. --Juan

1 0

Re: [dlls/shdocvw] Initialize variable (Coverity)
by Nikolay Sivov 01 Feb '11

01 Feb '11

On 2/1/2011 17:14, Paul Vriens wrote: > Hi, > > Addresses Coverity #1090. > > If CoCreateInstance() fails we check the uninitialized urlfile variable. Same story with urlobj. In fact it looks better to try CoCreateInstance first, before anything else like lstrcpy and just return earlier on failure.

2 2

Re: gdiplus: C++ fixes for headers
by Paul Chitescu 01 Feb '11

01 Feb '11

On Tuesday 01 February 2011 09:49:46 am Iain Arnell wrote: > To fix inclusion of gdiplus.h under C++, this patch: > * adds missing class declaration for GpPen > * adds missing class declaration for CharacterRange > * fixes PenAlignment/GpPenAlignment declaration > * avoids redeclaration of PixelFormat > > Simple testcase: > > cat >test.cpp <<EOF > #include <windows.h> > #include <gdiplus.h> > EOF > wineg++ -c test.cpp > --- > include/gdiplusenums.h | 2 +- > include/gdiplusgpstubs.h | 2 ++ > include/gdiplusimaging.h | 4 ++-- > include/gdiplustypes.h | 25 +++++++++++++++++++++++++ > 4 files changed, 30 insertions(+), 3 deletions(-) > > diff --git a/include/gdiplusenums.h b/include/gdiplusenums.h > index 82f20eb..dc03b29 100644 > --- a/include/gdiplusenums.h > +++ b/include/gdiplusenums.h > @@ -734,7 +734,7 @@ typedef enum StringTrimming StringTrimming; > typedef enum FontStyle FontStyle; > typedef enum StringFormatFlags StringFormatFlags; > typedef enum HotkeyPrefix HotkeyPrefix; > -typedef enum PenAlignment GpPenAlignment; > +typedef enum PenAlignment PenAlignment; > typedef enum PaletteFlags PaletteFlags; > typedef enum ImageCodecFlags ImageCodecFlags; > typedef enum CombineMode CombineMode; > diff --git a/include/gdiplusgpstubs.h b/include/gdiplusgpstubs.h > index 3edc1f4..5a65cd8 100644 > --- a/include/gdiplusgpstubs.h > +++ b/include/gdiplusgpstubs.h > @@ -22,6 +22,7 @@ > #ifdef __cplusplus > > class GpGraphics {}; > +class GpPen {}; > class GpBrush {}; > class GpHatch : public GpBrush {}; > class GpSolidFill : public GpBrush {}; > @@ -92,6 +93,7 @@ typedef WrapMode GpWrapMode; > typedef Color GpColor; > typedef FlushIntention GpFlushIntention; > typedef CoordinateSpace GpCoordinateSpace; > +typedef PenAlignment GpPenAlignment; > typedef PenType GpPenType; > > #endif > diff --git a/include/gdiplusimaging.h b/include/gdiplusimaging.h > index 42a00cb..3e9f6ee 100644 > --- a/include/gdiplusimaging.h > +++ b/include/gdiplusimaging.h > @@ -110,7 +110,7 @@ public: > UINT Width; > UINT Height; > INT Stride; > - PixelFormat PixelFormat; > + INT PixelFormat; /* MSDN: "PixelFormat PixelFormat;" */ > VOID* Scan0; > UINT_PTR Reserved; > }; > @@ -177,7 +177,7 @@ typedef struct BitmapData > UINT Width; > UINT Height; > INT Stride; > - PixelFormat PixelFormat; > + INT PixelFormat; /* MSDN: "PixelFormat PixelFormat;" */ > VOID* Scan0; > UINT_PTR Reserved; /* undocumented: stores the lock mode */ > } BitmapData; > diff --git a/include/gdiplustypes.h b/include/gdiplustypes.h > index 4f97432..b84c74e 100644 > --- a/include/gdiplustypes.h > +++ b/include/gdiplustypes.h > @@ -199,6 +199,31 @@ public: > INT Height; > }; > > +class CharacterRange > +{ > +public: > + CharacterRange() > + { > + First = Length = 0; > + } > + > + CharacterRange(IN INT first, IN INT length) > + { > + First = first; > + Length = length; > + } > + > + CharacterRange& operator=(IN const CharacterRange& rhs) > + { > + First = rhs.First; > + Length = rhs.Length; > + return *this; > + } > +public: > + INT First; > + INT Length; > +}; > + > #else /* end of c++ typedefs */ > > typedef struct Point Wouldn't be better to wrap the offending C with: #ifdef __cplusplus extern "C" { #endif ... #ifdef __cplusplus } #endif

2 1

Re: programs: Add winemsibuilder.
by Joris Huizer 01 Feb '11

01 Feb '11

While skimming through the new patches, I found: > +WCHAR *encode_stream( const WCHAR *in ) > +{ > + DWORD c, next, count; > + WCHAR *out, *p; > + > + count = strlenW( in ); > + if (count > MAX_STREAM_NAME) > + return NULL; > + > + count += 2; > + out = HeapAlloc( GetProcessHeap(), 0, count > * sizeof(WCHAR) ); > + p = out; You aren't checking for out-of-memory condition here. As you seem to cover all the other possible errors, I thought to let you know HTH, Joris

1 0

Re: [PATCH 1/5] shell32: Added SHCNRF_InterruptLevel flag support to SHChangeNotifyRegister
by Alexandre Julliard 01 Feb '11

01 Feb '11

Piotr Caban <piotr(a)codeweavers.com> writes: > void InitChangeNotifications(void) > { > + watched_directories = 0; > + interrupts_server_terminate = CreateEventW(NULL, FALSE, FALSE, NULL); > + interrupts_server = CreateThread(NULL, 0, interrupts_server_thread, NULL, 0, NULL); > } This needs to be done on demand. You can't create a thread in every process on the off chance that it would use interrupt notifications. Probably you can use the thread pool for this. -- Alexandre Julliard julliard(a)winehq.org

1 0