September 2014 - Wine-Devel

Re: (resend)[3/6]dwrite: Implement GetInformationalStrings
by Alexandre Julliard Sept. 5, 2014

Sept. 5, 2014

Aric Stewart <aric(a)codeweavers.com> writes: > @@ -53,6 +54,8 @@ struct dwrite_font_data { > DWRITE_FONT_SIMULATIONS simulations; > DWRITE_FONT_METRICS metrics; > > + IDWriteLocalizedStrings *info_strings[23]; Where does this 23 come from? -- Alexandre Julliard julliard(a)winehq.org

2 3

Re: [3/6]dwrite: Implement GetInformationalStrings
by Alexandre Julliard Sept. 5, 2014

Sept. 5, 2014

Aric Stewart <aric(a)codeweavers.com> writes: > read strings from OpenType 'name' table opentype.c: In function ‘get_name_font_table_data’: opentype.c:470:12: error: variable ‘format’ set but not used [-Werror=unused-but-set-variable] -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: dpnet: Stub interface IDirectPlay8LobbyClient (try 4 - resend)
by Alexandre Julliard Sept. 5, 2014

Sept. 5, 2014

Alistair Leslie-Hughes <leslie_alistair(a)hotmail.com> writes: > Hi, > Moved to be client.c. > Flags are printed as 0x%08x. > > Putting the tests in with client.c doesn't make much sense, since > they wont share code. Creating a new file for a 10-line test makes even less sense. Maybe at some point it will be big enough to have its own file, but for now please avoid creating new files when not necessary. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: configure: Fix detection of ncurses library on Archlinux.
by Alexandre Julliard Sept. 4, 2014

Sept. 4, 2014

Sebastian Lackner <sebastian(a)fds-team.de> writes: > On Archlinux the libncurses library is just an ASCII text file > containing "INPUT(-lncursesw)" to force the linker to use libncursesw > instead. AC_LINK_IFELSE succeeds, but afterwards, when > WINE_CHECK_SONAME tries to figure out the name it fails because it > doesn't recognize libncursesw.so as the same library. Using a regexp > as pattern is the easiest solution to fix this. That doesn't do the right thing on Mac OS: #define SONAME_LIBNCURSES " /sw/lib/ncurses/libncurses.5.dylib (compatibility version 5.0.0, current version 5.0.0)" -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH 1/5] widl: Introduce an "aggregate_return" function attribute.
by Alexandre Julliard Sept. 4, 2014

Sept. 4, 2014

Henri Verbeet <hverbeet(a)codeweavers.com> writes: > For MSVC style handling of aggregate return values. This probably doesn't make > anyone particularly happy; I'm open to suggestions. It may be possible to do > this without an attribute, based on the function declaration, but then we'd > have to figure out how MSVC decides when to use registers and when to use the > stack for the return value. The attribute is certainly ugly. Also it seems to me you are doing this the wrong way around; you don't want to add a return parameter, you want to avoid one. Check how div() is handled. -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: (resend/renumber)[6/14]dwrite: code to add fonts to a fontfamily
by Alexandre Julliard Sept. 4, 2014

Sept. 4, 2014

Aric Stewart <aric(a)codeweavers.com> writes: > @@ -968,6 +968,28 @@ static const IDWriteFontFamilyVtbl fontfamilyvtbl = { > dwritefontfamily_GetMatchingFonts > }; > > +static HRESULT fontfamily_AddFont(struct dwrite_fontfamily *family, IDWriteFont *font) > +{ > + struct dwrite_font *font_data = impl_from_IDWriteFont(font); > + > + if (family->font_count + 1 >= family->alloc) > + { > + struct dwrite_font_data **new_list; > + UINT32 new_alloc; > + new_alloc = family->alloc * 2; > + new_list = heap_realloc(family->fonts, sizeof(*family->fonts) * new_alloc); > + if (!new_list) > + return E_OUTOFMEMORY; > + family->fonts = new_list; > + family->alloc = new_alloc; > + } > + > + family->fonts[family->font_count] = font_data->data; > + InterlockedIncrement(&font_data->data->ref); > + family->font_count++; > + return S_OK; > +} Please don't add dead code. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [PATCH] ddrawex: Don't leak surfaces.
by Henri Verbeet Sept. 4, 2014

Sept. 4, 2014

On 4 September 2014 13:18, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > dds_get_outer AddRefs the inner surface when it creates the wrapper > surface. The wrapper releases only one reference when it is destroyed, > leaking one of the two. AddRefing in dds_get_outer makes sense because > this is also used by e.g. GetAttachedSurface, where it creates a new > wrapper for surfaces ddrawex didn't know about yet. > > Whether native ddrawex.dll properly handles GetAttachedSurfaces and > friends is an open question. We already know it doesn't properly handle > IDirectDraw4::GetSurfaceFromDC, even though the equivalent ddraw3 method > works. On a related note, I think it's unfortunate that dds_get_outer() creates a wrapper surface with refcount 1 if it doesn't exist yet, but doesn't AddRef() existing interfaces when returning them. That means the caller effectively needs to know if the wrapper already exists or not in order to do proper refcounting.

1 0

[PATCH 1/4] reg.exe: Add path_get_key() to remove boilerplate
by Hugh McMaster Sept. 4, 2014

Sept. 4, 2014

On Wed, 3 Sep 2014 14:54:44 +0200, Jonathan Vollebregt wrote: > + if (path[0] == '\\') > + { > + path++; > + } > + if (path[0] == '\\') > + { > + reg_message(STRING_NO_REMOTE); > + return NULL; > + } This would be valid if the path was something like \\remote-machine\HKLM. But if the first if block was true and the second block was false, the registry path is invalid. Windows does not accept registry paths such as \HKLM. In summary: True + True = valid path, but not supported in Wine. True + False = invalid path False + True = impossible False + False = valid path The original check was better: if (key_name[0] == '\\' && key_name[1] == '\\') { reg_message(STRING_NO_REMOTE); return 1; } It was also unclear why you perform the path[0] == '\\' check in path_get_key() and path_get_rootkey_name(). One common check should be enough. The original version of this patch did this (now superseded). But I see you made some changes around the reg_delete() function.

2 1

Re: [5/17]dwrite: Shift to using localized string for family name
by Nikolay Sivov Sept. 3, 2014

Sept. 3, 2014

> +HRESULT clone_localizedstring(IDWriteLocalizedStrings *iface, IDWriteLocalizedStrings **strings) This looks unnecessary, as you can do create_localizedstrings() and a series of add_localizedstring() after.

2 2

Re: (resend/renumber)[4/14]dwrite: implement a list of font data for IDWriteFontFamily
by Alexandre Julliard Sept. 3, 2014

Sept. 3, 2014

Aric Stewart <aric(a)codeweavers.com> writes: > @@ -1121,6 +1188,24 @@ static void get_font_properties(struct dwrite_font *font, HDC hdc) > } > } > > +static HRESULT create_font_from_data(struct dwrite_font_data *data, IDWriteFont **font) > +{ > + HRESULT hr; > + struct dwrite_font *This; > + > + hr = create_font_base(font); > + if (FAILED(hr)) > + return hr; > + > + This = impl_from_IDWriteFont(*font); > + > + heap_free(This->data); > + This->data = data; > + InterlockedIncrement(&This->data->ref); > + > + return S_OK; > +} That looks unnecessarily complicated. Why allocate and initialize the data only to immediately replace it? -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: wine.inf: Add HKLM, Software\Microsoft\Cryptography, "MachineGuid".
by Alexandre Julliard Sept. 3, 2014

Sept. 3, 2014

Hans Leidekker <hans(a)codeweavers.com> writes: > The UK version of the Wizard 101 downloader needs this. > --- > loader/wine.inf.in | 1 + > 1 file changed, 1 insertion(+) It should be created using the existing code in advapi32. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: (resend/renumber)[14/14]dwrite: Add Custom Font Collections
by Marvin Sept. 3, 2014

Sept. 3, 2014

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 https://testbot.winehq.org/JobDetails.pl?Key=8774 Your paranoid android. === wvistau64 (32 bit font) === font.c:1163: Test failed: expected refcount 1, got 2 font.c:1174: Test failed: expected refcount 2, got 3 font.c:1175: Test failed: expected refcount 2, got 3 === w2008s64 (32 bit font) === font.c:1163: Test failed: expected refcount 1, got 2 font.c:1174: Test failed: expected refcount 2, got 3 font.c:1175: Test failed: expected refcount 2, got 3 === wvistau64 (64 bit font) === font.c:1163: Test failed: expected refcount 1, got 2 font.c:1174: Test failed: expected refcount 2, got 3 font.c:1175: Test failed: expected refcount 2, got 3 === w2008s64 (64 bit font) === font.c:1163: Test failed: expected refcount 1, got 2 font.c:1174: Test failed: expected refcount 2, got 3 font.c:1175: Test failed: expected refcount 2, got 3

1 0

[PATCH 2/5] reg.exe: Add path_get_key() to remove boilerplate
by Hugh McMaster Sept. 3, 2014

Sept. 3, 2014

Hi Jonathan, On Monday, 1 September 2014, 15:05:31 +0200, Jonathan Vollebregt wrote: + CompareStringW(CP_ACP,NORM_IGNORECASE, + path,strlenW(short_HKEY_name[i]), + short_HKEY_name[i],strlenW(short_HKEY_name[i]))==2 || + CompareStringW(CP_ACP,NORM_IGNORECASE, + path,strlenW(long_HKEY_name[i]), + long_HKEY_name[i],strlenW(long_HKEY_name[i]))==2) 1. CompareString does not accept code page identifiers. It accepts a locale.[1] You could use LOCAL_USER_DEFAULT or GetThreadLocale() instead. 2. Please use CSTR_EQUAL instead of 2. 3. You should also avoid calling strlenW four times. int len_short_HKEY_name = strlenW(short_HKEY_name[i]); or something similar. Alternatively, since wine/unicode.h is already being included, consider using strncmpiW. 4. In patches 2 to 5, please mimic the style of the surrounding/original code. That being said, if there are variations in style, just use the most common style. + if(path_get_rootkey_name(path)) + { + return HKEYs[(path_get_rootkey_name(path) - &long_HKEY_name[0][0]) / 20]; + } + else + { + return NULL; + } Single line code in if/else blocks can be compacted to use less space. E.g.: if (path_get_rootkey_name(path)) return HKEYs[(path_get_rootkey_name(path) - &long_HKEY_name[0][0]) / 20]; else return NULL; > +static const WCHAR * path_get_rootkey_name(LPWSTR path) Asterisks should be placed next to the object name or type, e.g. static const WCHAR *path_get_rootkey_name(LPWSTR path). > + if(!key) > + if (RegQueryValueW(key,value_name,NULL,NULL)==ERROR_SUCCESS) > + RegSetValueExW(key,value_name,0,reg_type,reg_data,reg_count); Also, code is easier to read if there is space after keywords, arguments and around == (equivalent to). You have been slightly inconsistent throughout. [1] http://msdn.microsoft.com/en-us/library/windows/desktop/dd317759%28v=vs.85%… -- Hugh McMaster

2 2

Re: winewrapper: In 64 bit trees the loader is called wine64.
by Alexandre Julliard Sept. 3, 2014

Sept. 3, 2014

Francois Gouget <fgouget(a)free.fr> writes: > This fixes running 'make test' in 64 bit trees. I don't mind adding this, but I would recommend running tests with a proper WoW64 setup instead. -- Alexandre Julliard julliard(a)winehq.org

2 2

[PATCH 3/5] reg.exe: Add wchar/type conversion functions
by Hugh McMaster Sept. 3, 2014

Sept. 3, 2014

On Mon, 1 Sep 2014 15:05:32 +0200, Jonathan Vollebregt wrote: +#define MAX_TYPE_LENGTH 32 +static const WCHAR type_names[][MAX_TYPE_LENGTH] = { + {'R','E','G','_','N','O','N','E',0}, + {'R','E','G','_','S','Z',0}, + {'R','E','G','_','E','X','P','A','N','D','_','S','Z',0}, + {'R','E','G','_','B','I','N','A','R','Y',0}, + {'R','E','G','_','D','W','O','R','D',0}, + {'R','E','G','_','D','W','O','R','D','_','B','I','G','_','E','N','D','I','A','N',0}, + {0}, + {'R','E','G','_','M','U','L','T','I','_','S','Z',0}, +}; + + + Why are you defining MAX_TYPE_LENGTH as 32? The longest is 21 (including the null terminator) or, if you include REG_DWORD_LITTLE_ENDIAN, the longest is 24. Also, what is the purpose of {0}? And you have included three newlines. -- Hugh McMaster

1 0

Re: [17/17]dwrite: Add Custom Font Collections
by Marvin Sept. 2, 2014

Sept. 2, 2014

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 https://testbot.winehq.org/JobDetails.pl?Key=8763 Your paranoid android. === build (build) === Patch failed to apply

1 0

Re: [4/17]dwrite: Implement GetGlyphIndices from the CMAP table
by Marvin Sept. 2, 2014

Sept. 2, 2014

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 https://testbot.winehq.org/JobDetails.pl?Key=8762 Your paranoid android. === build (build) === Patch failed to apply

1 0

Re: wininet: Fixed most tests on IE11.
by Marvin Sept. 2, 2014

Sept. 2, 2014

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 https://testbot.winehq.org/JobDetails.pl?Key=8759 Your paranoid android. === w7u (32 bit http) === http.c:2818: Test failed: HttpSendRequest failed http.c:2823: Test failed: HttpQueryInfo failed http.c:2824: Test failed: got status 0, expected 200

1 0

Re: wined3d: Fix trace output of emulated_textureram. (try 2)
by Henri Verbeet Sept. 2, 2014

Sept. 2, 2014

On 1 September 2014 18:12, Johannes Brandstätter <jbrandst(a)2ds.eu> wrote: > - wined3d_settings.emulated_textureram = TmpVideoMemorySize *1024*1024; > - TRACE("Use %iMB = %d byte for emulated_textureram\n", > + wined3d_settings.emulated_textureram = (UINT64)TmpVideoMemorySize *1024*1024; > + TRACE("Use %iMiB = %llu byte for emulated_textureram\n", Unfortunately we can't use %ll in Wine, you'll have to wine_dbgstr_longlong(). You can look at other places in wined3d for examples of how it's used.

1 0

Re: dwrite: Add the system font collection to the dwrite factory object
by Alexandre Julliard Sept. 1, 2014

Sept. 1, 2014

Aric Stewart <aric(a)codeweavers.com> writes: > This makes is a properly allocated object It doesn't work: ../../../tools/runtest -q -P wine -T ../../.. -M dwrite.dll -p dwrite_test.exe.so layout && touch layout.ok wine: Unhandled page fault on execute access to 0x00000001 at address 0x1 (thread 0037), starting debugger... Unhandled exception: page fault on execute access to 0x00000001 in 32-bit code (0x00000001). Register dump: CS:0023 SS:002b DS:002b ES:002b FS:0063 GS:006b EIP:00000001 ESP:0032fc2c EBP:0032fc48 EFLAGS:00010202( R- -- I - - - ) EAX:0012b780 EBX:686675ac ECX:0032fca0 EDX:001100f8 ESI:0012e988 EDI:0012e980 Stack dump: 0x0032fc2c: 6865b668 0012b780 0012b780 0032fc88 0x0032fc3c: 6865b857 6863464d 00000000 0032fc88 0x0032fc4c: 6865b715 00110000 00000000 0012b780 0x0032fc5c: 68636210 6863464d 68634632 0012b780 0x0032fc6c: 00000000 6863464d 68634632 0032fca0 0x0032fc7c: 68636210 6863464d 68634632 0032fd48 Backtrace: =>0 0x00000001 (0x0032fc48) 1 0x6865b715 dwritetextformat_Release+0x44() in dwrite (0x0032fc88) 2 0x686309d5 func_layout+0x864() [/home/julliard/wine/wine/dlls/dwrite/tests/../../../include/dwrite.h:3596] in dwrite_test (0x0032fd48) 3 0x6862a0af main+0x37e(argc=<is not available>, argv=<is not available>) [/home/julliard/wine/wine/dlls/dwrite/tests/../../../include/wine/test.h:584] in dwrite_test (0x0032fe18) 4 0x686335a0 __wine_spec_exe_entry+0x7f(peb=<couldn't compute location>) [/home/julliard/wine/wine/dlls/winecrt0/exe_entry.c:36] in dwrite_test (0x0032fe58) 5 0x7b85f54c call_process_entry+0xb() in kernel32 (0x0032fe78) 6 0x7b860593 start_process+0x62(peb=<couldn't compute location>) [/home/julliard/wine/wine/dlls/kernel32/process.c:1097] in kernel32 (0x0032feb8) 7 0x7bc7fba0 call_thread_func_wrapper+0xb() in ntdll (0x0032fed8) 8 0x7bc82b4d call_thread_func+0x7c(entry=0x7b860530, arg=0x7ffdf000, frame=0x32ffc8) [/home/julliard/wine/wine/dlls/ntdll/signal_i386.c:2637] in ntdll (0x0032ffa8) 9 0x7bc7fb7e call_thread_entry_point+0x11() in ntdll (0x0032ffc8) 10 0x7bc53fee start_process+0x1d(kernel_start=0x7b860530) [/home/julliard/wine/wine/dlls/ntdll/loader.c:2856] in ntdll (0x0032ffe8) 11 0x680316dd wine_call_on_stack+0x1c() in libwine.so.1 (0x00000000) 12 0x6803179b wine_switch_to_stack+0x2a(func=0x7bc53fd0, arg=0x7b860530, stack=0x330000) [/home/julliard/wine/wine/libs/wine/port.c:59] in libwine.so.1 (0xffc74fc8) 13 0x7bc59ae9 LdrInitializeThunk+0x238(kernel_start=<couldn't compute location>, unknown2=<couldn't compute location>, unknown3=<couldn't compute location>, unknown4=<couldn't compute location>) [/home/julliard/wine/wine/dlls/ntdll/loader.c:2910] in ntdll (0xffc75008) 14 0x7b866db3 __wine_kernel_init+0xa02() [/home/julliard/wine/wine/dlls/kernel32/process.c:1269] in kernel32 (0xffc75f08) 15 0x7bc5a9e3 __wine_process_init+0x192() [/home/julliard/wine/wine/dlls/ntdll/loader.c:3119] in ntdll (0xffc75f98) 16 0x6802ede8 wine_init+0x2c7(argc=0x3, argv=0xffc764d4, error="", error_size=0x400) [/home/julliard/wine/wine/libs/wine/loader.c:958] in libwine.so.1 (0xffc75ff8) 17 0x7bf00d6c main+0x7b(argc=<is not available>, argv=<is not available>) [/home/julliard/wine/wine/loader/main.c:237] in <wine-loader> (0xffc76438) 18 0x682518c5 __libc_start_main+0xf4() in libc.so.6 (0x00000000) -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH] mshtml: Add IHTMLStyle5::minWidth property implementation.
by Marvin Sept. 1, 2014

Sept. 1, 2014

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 https://testbot.winehq.org/JobDetails.pl?Key=8743 Your paranoid android. === w2000pro (32 bit style) === style.c:2305: Test failed: Could not get IHTMLStyle5 iface: 80004002

1 0

Re: [3/4] msi: Return an error from MSI_GetComponentPath if the buffer is too small.
by Marvin Sept. 1, 2014

Sept. 1, 2014

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 https://testbot.winehq.org/JobDetails.pl?Key=8742 Your paranoid android. === w7u (32 bit msi) === The test timed out

1 0

Re: wined3d: Fix trace output of emulated_textureram.
by Henri Verbeet Sept. 1, 2014

Sept. 1, 2014

On 31 August 2014 21:40, Johannes Brandstätter <jbrandst(a)2ds.eu> wrote: > wined3d_settings.emulated_textureram = > TmpVideoMemorySize *1024*1024; It looks like your MUA wrapped the patch, I'd recommend using git send-email for sending patches. > - TRACE("Use %iMB = %d byte for emulated_textureram\n", > + TRACE("Use %iMB = %u byte for emulated_textureram\n", > TmpVideoMemorySize, > wined3d_settings.emulated_textureram); This change is ok, but note that for large values you'd want "emulated_textureram" to be a UINT64. Currently it's going to wrap to 0 at 4 GiB.

2 2

Re: [PATCH 3/5] d3d10core/tests: Add a basic scissoring test.
by Marvin Sept. 1, 2014

Sept. 1, 2014

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 https://testbot.winehq.org/JobDetails.pl?Key=8734 Your paranoid android. === wvistau64 (32 bit device) === device.c:126: Test failed: Failed to create swapchain, hr 0x80070057. device: unhandled exception c0000005 at 00401DB0 === w2008s64 (32 bit device) === device.c:126: Test failed: Failed to create swapchain, hr 0x80070057. device: unhandled exception c0000005 at 00401DB0 === w8 (32 bit device) === device.c:126: Test failed: Failed to create swapchain, hr 0x80070057. device: unhandled exception c0000005 at 00401DB0 === wvistau64 (64 bit device) === device.c:126: Test failed: Failed to create swapchain, hr 0x80070057. device: unhandled exception c0000005 at 0000000000401C38 === w2008s64 (64 bit device) === device.c:126: Test failed: Failed to create swapchain, hr 0x80070057. device: unhandled exception c0000005 at 0000000000401C38

1 0

Re: wined3d: Remove an ugly NULL cast.
by Henri Verbeet Sept. 1, 2014

Sept. 1, 2014

On 1 September 2014 12:25, Francois Gouget <fgouget(a)free.fr> wrote: > --- > > This is mostly cosmetic I guess. > Well, no. This will introduce a warning on 64-bit because now you're casting to a pointer from an integer of a different size.

1 0

Wine test analysis August 28th
by Jeremy White Sept. 1, 2014

Sept. 1, 2014

Hi Folks, Good news - we've made progress! We're down to only 25 issues. I have now annotated every single failure: https://www.winehq.org/~jwhite/5aeb834dd4e7.html (although some of my annotations are out of date). It looks like the biggest win will come if Francois and Hans can figure out the msi timing stuff. I think there are a number of ones that might be 'easy': http://test.winehq.org/data/tests/gdi32:font.html Bug 35573, 35760, fails in certain locales http://test.winehq.org/data/tests/mmdevapi:capture.html I asked Jorg, but he may have a hard time getting to it. It's nice and consistent... http://test.winehq.org/data/tests/shell32:shlexec.html Nice and consistent; asked Misha to look into it, but he can't. http://test.winehq.org/data/tests/user32:sysparams.html Nice and consistent; I'd invited Frederic to work on this one, but looks like he could use help... Don't be shy, there are plenty of consistent failures to go around! From there, I think nudging Jacek, Nikolay, and Huw will get us our next big set of wins. Nudge, nudge, nudge. Cheers, Jeremy p.s. Reminder: my analysis always shows up at http://winehq.org/~jwhite/latest.html

5 5