July 2008 - Wine-devel - list.winehq.org

failing gdiplus:matrix tests
by James Hawkins 12 Jul '08

12 Jul '08

Hi Nikolay, Tests you added in the following commit fail on all platforms. Did you run the tests in Windows? commit d7999a008be5d7e95fcd86019adc79b0359ae8d4 Author: Nikolay Sivov <bunglehead(a)gmail.com> Date: Tue Jul 8 01:32:36 2008 +0400 gdiplus: GdipInvertMatrix implementation with tests. -- James Hawkins

1 0

failing gdiplus:font tests
by James Hawkins 12 Jul '08

12 Jul '08

Hi Adam, Tests you added in the following commit rely on the Arial font being installed, which is not necessarily the case. You can check out [1] for the test output of a system that does not have the Arial font. You should either test a font that is guaranteed to be installed (Tahoma?) or use some API to enumerate the available fonts and test one of those. commit fe3d38d1a0e19e503048f8338f8f26cd0668d364 Author: Adam Petaccia <adam(a)tpetaccia.com> Date: Sat Jun 21 13:02:56 2008 -0400 gdiplus: Tests for new font functions. [1] http://test.winehq.org/data/8fd6f0e26ae28f2ba4938e2fbcc4851f47baa53c/wine_x… Thanks, James Hawkins

1 0

Re: [2/3] appwiz.cpl: Read uninstallable applications from registry, add to list.
by Nikolay Sivov 11 Jul '08

11 Jul '08

Owen Rudge wrote: + while (iter) + { + if (iter->title) + HeapFree(hHeap, 0, iter->title); + + if (iter->path) + HeapFree(hHeap, 0, iter->path); + + if (iter->icon) + HeapFree(hHeap, 0, iter->icon); + + if (iter->publisher) + HeapFree(hHeap, 0, iter->icon); + + if (iter->version) + HeapFree(hHeap, 0, iter->icon); + + next = iter->next; + HeapFree(hHeap, 0, iter); + iter = next; + } Seems to be a copy-paste

2 1

Re: [dplayx 02/29] Tests for DirectPlayCreate
by Michael Karcher 11 Jul '08

11 Jul '08

Hello Ismael, reviewing this patch shows some issues. > +#define _okHR(expected, result) \ > + ok( (expected) == (result), \ > + "expected=%d(%s) got=%d(%s)\n", \ > + HRESULT_CODE(expected), dpResult2str(expected), \ > + HRESULT_CODE(result), dpResult2str(result) ); I dislike the order of the parameters and the name. You would write ok( hr == DP_OK, "Something failed with %x\n", hr); In this line, "hr" is before "DP_OK". So as an analogy, I also want to write _okHR( hr, DD_OK ); Why do you start the macro name with an underscore? And last but not least: As I see it, a static inline would do the job too, and is always preferrable to macros. > + heap = HeapCreate( 0, 0, 0 ); Do you have any reason to create your own heap? Each windows process already comes equipped with a standard heap you can obtain using the API function GetProcessHeap, so this line: > + IUnknown* pUnk = HeapAlloc( heap, 0, sizeof(LPDWORD) ); would then look like: + IUnknown* pUnk = HeapAlloc( GetProcessHeap(), 0, sizeof(LPDWORD) ); Another point is that this line is totally wrong anyway. An IUnknown* is a pointer to an COM object, which starts with a vtable pointer. You let pUnk point to uninitialized memory. While it doesn't matter in this case (as DirectPlayCreate will never access what the pointer points to), it is really bad style. I understand the intention to test that aggregation is not supported for DirectPlay, but you still should keep the contract of the creation function that pUnk either is NULL or points to a valid COM object. A basic COM object is just a pointer to a vtable that contains a QueryInterface method that returns "this" for IUnknown and fails in every other case, an AddRef method that returns 2 and does nothing, and a Release method that returns 1 and does nothing else. > + /* pUnk==NULL, pDP!=NULL */ > + hr = DirectPlayCreate( NULL, &pDP, NULL ); > + _okHR( DPERR_INVALIDPARAMS, hr ); > + hr = DirectPlayCreate( (LPGUID) &_GUID_NULL, &pDP, NULL ); > + _okHR( DP_OK, hr ); > + hr = DirectPlayCreate( (LPGUID) &DPSPGUID_TCPIP, &pDP, NULL ); > + todo_wine _okHR( DP_OK, hr ); Here you are leaking DirectPlay objects. If DirectPlayCreate worked, you have to release the object before you overwrite pDP (by passing it to DirectPlayCreate again) using IDirectPlay_Release(pDP). Regards, Michael Karcher

5 9

Re: extending setupapi
by Juan Lang 11 Jul '08

11 Jul '08

Hi Pieter, overall your patch looks good, but it has some minor errors that get in the way: - struct DeviceInfo *devInfo = SETUPDI_AllocateDeviceInfo(set, set->cDevices, - instanceId, phantom); + struct DeviceInfo *devInfo; TRACE("%p, %s, %d, %s, %d\n", set, debugstr_guid(guid), devInst, debugstr_w(instanceId), phantom); + devInfo = SETUPDI_AllocateDeviceInfo(set, set->cDevices, + instanceId, phantom); + This change is unrelated to the function you're implementing, please send it in a separate patch if you want this change to go in. + TRACE("%s, %s, %p, %08x, %p, %s, %p\n", debugstr_guid(class), + debugstr_a(enumstr), parent, flags, deviceset, + debugstr_a(machine), reserved); + Same here.. - HKEY interfacesKey = SetupDiOpenClassRegKeyExW(interface, KEY_READ, - DIOCR_INTERFACE, NULL, NULL); + HKEY interfacesKey; TRACE("%p, %s, %s, %08x\n", DeviceInfoSet, debugstr_guid(interface), debugstr_w(enumstr), flags); + interfacesKey = SetupDiOpenClassRegKeyExW(interface, KEY_READ, + DIOCR_INTERFACE, NULL, NULL); + and here.. @@ -3549,6 +3558,10 @@ HKEY WINAPI SetupDiOpenClassRegKeyExW( LPCWSTR lpKeyName; LONG l; + TRACE("%s, %08x, %08x, %s, %p\n", debugstr_guid(ClassGuid), + samDesired, Flags, + debugstr_w(MachineName), Reserved); + and here. +@ stdcall SetupDiOpenDeviceInterfaceRegKey(ptr ptr long long) @ stdcall SetupDiOpenDeviceInterfaceA(ptr str long ptr) -@ stub SetupDiOpenDeviceInterfaceRegKey In general we alphabetize function names in spec files, so please don't change the order. Thanks for your help with Wine. --Juan

1 0

Re: quartz: Fix incorrect use of mtCurrent in transform filter
by Alexandre Julliard 11 Jul '08

11 Jul '08

"Maarten Lankhorst" <m.b.lankhorst(a)gmail.com> writes: > --- > mtCurrent is cleared when connecting, so this may cause weird bugs > during connection. This doesn't work: ../../../tools/runtest -q -P wine -M qedit.dll -T ../../.. -p qedit_test.exe.so mediadet.c && touch mediadet.ok err:quartz:OutputPin_Connect cbFormat: 0 pin.c:943: OutputPin_Connect: Assertion `pmtCandidate->pbFormat' failed. wine: Assertion failed at address 0x601b0b87 (thread 0009), starting debugger... Unhandled exception: assertion failed in 32-bit code (0x601b0b87). Register dump: CS:0023 SS:002b DS:002b ES:002b FS:0063 GS:006b EIP:601b0b87 ESP:0032fa50 EBP:0032fa58 EFLAGS:00000206( - 00 - -IP1) EAX:00000000 EBX:00001acf ECX:00001acf EDX:00000006 ESI:0032faf8 EDI:602c8ff4 Stack dump: 0x0032fa50: 602c8ff4 602d19f0 0032fb84 601b2281 0x0032fa60: 00000006 0032faf8 00000000 7d57b460 0x0032fa70: 00000068 7d57b468 0032fa9c 601ede11 0x0032fa80: 0032fabc 7d57b468 7d57b4cc 0032fb94 0x0032fa90: 602c8ff4 0000004a 7d57b468 0032fb68 0x0032faa0: 601e7c7b 7d57b468 0000004a 0032fbac Backtrace: =>1 0x601b0b87 gsignal+0x3f() in libc.so.6 (0x0032fa58) 2 0x601b2281 abort+0x101() in libc.so.6 (0x0032fb84) 3 0x601a9ed4 __assert_fail+0xf0() in libc.so.6 (0x0032fbc8) 4 0x60d7af84 OutputPin_Connect+0x184(iface=0x12f510, pReceivePin=0x12fe50, pmt=0x0) [/home/julliard/wine/wine/dlls/quartz/pin.c:965] in quartz (0x0032fc28) 5 0x60d2a52a MediaDet_put_Filename+0x5aa(iface=<register ESI not in topmost frame>, newVal=0x122834) [/home/julliard/wine/wine/dlls/qedit/mediadet.c:416] in qedit (0x0032fd58) 6 0x6053f375 func_mediadet+0x4d5() [/home/julliard/wine/wine/dlls/qedit/tests/mediadet.c:150] in qedit_test (0x0032fe68) 7 0x60540280 run_test+0x130(name=0x1103ee) [/home/julliard/wine/wine/dlls/qedit/tests/../../../include/wine/test.h:454] in qedit_test (0x0032fea8) 8 0x60540483 main+0x133(argc=<register ECX not in topmost frame>, argv=<register ECX not in topmost frame>) [/home/julliard/wine/wine/dlls/qedit/tests/../../../include/wine/test.h:502] in qedit_test (0x0032fed8) 9 0x60540c38 __wine_spec_exe_entry+0x88(peb=0x7ffdf000) [/home/julliard/wine/wine/dlls/winecrt0/exe_entry.c:36] in qedit_test (0x0032ff08) 10 0x7b87ade8 start_process+0x98(arg=0x0) [/home/julliard/wine/wine/dlls/kernel32/process.c:904] in kernel32 (0x0032ffe8) 11 0x60028b47 wine_switch_to_stack+0x17() in libwine.so.1 (0x00000000) -- Alexandre Julliard julliard(a)winehq.org

1 0

New winetricks 20080710: new verb ogg, fixes for firefox and divx. Schroedinger support!
by Dan Kegel 11 Jul '08

11 Jul '08

Another week, another winetricks. The big news is that Maarten now has a good directshow filter for the Schroedinger codec, and that it runs on Wine finally. To see it in action, do rm winetricks wget http://kegel.com/wine/winetricks wget http://kegel.com/wine/codecdemo bash codecdemo then choose the options to install ogg and xvid; then choose the option to play the example ogg dirac file Acadiane.ogg. It should work (with some artifacts). This is big news because Schroedinger is potentially a good enough, free enough codec to achieve world domination one of these days (once all the bugs are worked out). It's also big news because now I owe Maarten a beer. Online as always at http://kegel.com/wine/winetricks or http://winezeug.googlecode.com Changes since July 4th's wine-devel annoucement: r101 Added Schroedinger directshow filter (w00t!) r100 Add ogg codec to winetricks and codecdemo r99 add missing quotes on backtick; should fix problem reported by Philippe Roubach? r98 Updated divx r97 Fixed URL for firefox3 (thanks Shing-Shong Shei), Fixed silent install for firefox3 Added firefox synonym for firefox3

1 0

New winetricks 20080710: new verb ogg, fixes for firefox and divx.
by Dan Kegel 11 Jul '08

11 Jul '08

Another week, another winetricks. The big news is that Maarten now has a good directshow filter for the Schroedinger codec, and that it runs on Wine finally. To see it in action, do rm winetricks wget http://kegel.com/wine/winetricks wget http://kegel.com/wine/codecdemo bash codecdemo then choose the options to install ogg and xvid; then choose the option to play the example ogg dirac file Acadiane.ogg. It should work (with some artifacts). This is big news because Schroedinger is potentially a good enough, free enough codec to achieve world domination one of these days (once all the bugs are worked out). Online as always at http://kegel.com/wine/winetricks or http://winezeug.googlecode.com Changes since July 4th's wine-devel annoucement: r101 Added Schroedinger directshow filter (w00t!) r100 Add ogg codec to winetricks and codecdemo r99 add missing quotes on backtick; should fix problem reported by Philippe Roubach? r98 Updated divx r97 Fixed URL for firefox3 (thanks Shing-Shong Shei), Fixed silent install for firefox3 Added firefox synonym for firefox3

1 0

What's the holdup with RegisterHotKey and UnregisterHotKey??
by James Keane 11 Jul '08

11 Jul '08

Hey everyone! Just wanted to bring to your attention http://bugs.winehq.org/show_bug.cgi?id=1348 again, and ask what is the holdup? As far as I can see it appears a solution and working patch has been posted. I am interested in getting this into the main tree and wondering what kind of modifications have to be made to the patch supplied to make it happen. ------- Comment #8 From Robert Reif 2007-02-17 15:45:56 ------- The reason I was given by Alexandre is that the functionality should be moved into the server. Why is it not possible to accept this patch and then start working on moving it into the server? With a quick look the only changes from the patch and the git, is the change from dll/user to dll/user32. Cheers! James

3 5

Re: winhttp: main.c: Add stub implementation for WinHttpGetDefaultProxyConfiguration
by Zac Brown 11 Jul '08

11 Jul '08

Zac Brown wrote: > Add stub implementation for WinHttpGetDefaultProxyConfiguration > > Ignore this, formatting is wrong.

1 0