November 2008 - Wine-Devel - list.winehq.org

dead stores in msxml3
by ricardo filipe Nov. 18, 2008

Nov. 18, 2008

hi leslie. i'm tracking some dead stores with Clang and there are some in dlls/msxml3/saxreader.c i'm hoping you can tell me if the stores are to be removed for performance or if they should be checked in each iteration for correctness: file dlls/msxml3/saxreader.c lines: 1163 1166 1213 1216 regards, ricardo _________________________________________________________________ Confira vídeos com notícias do NY Times, gols direto do Lance, videocassetadas e muito mais no MSN Video! http://video.msn.com/?mkt=pt-br

2 1

Re: dlls/ntdll: fixed dead store (llvm/clang)
by Austin English Nov. 18, 2008

Nov. 18, 2008

On Tue, Nov 18, 2008 at 4:22 PM, ricardo filipe <ricardo_barbano(a)hotmail.com> wrote: > > > ________________________________ > Conheça já o Windows Live Spaces, o site de relacionamentos do Messenger! > Crie já o seu! > > > You forgot the patch. -- -Austin

1 0

Re: dlls/oleaut32: add tests for OleLoadPictureFile(Ex)
by Jeremy Drake Nov. 18, 2008

Nov. 18, 2008

On Tue, 18 Nov 2008, Michael Karcher wrote: > Looks OK to me, but what happened to your finding about > INET_E_RESOURCE_NOT_FOUND? Was it a red herring? I saw that on XP, but I couldn't reproduce it on Windows 2000. I need to investigate this further. Is there a mechansim to do windows-version-specific tests? > > I ended up having to add OleLoadPicutreFile and OleLoadPictureFileEx to > > the .spec and .c files as actual functions (for now just returning > > E_NOTIMPL) in order to get the test to even compile. > Might be that they don't appear in the ELF export table otherwise, and > as we don't crossbuild the tests, the imports are done by the ELF > linker/loader. BTW: In general, if you add stubs like this, please add a > FIXME call showing the parameters and saying stub. How do I show the VARIANT in the FIXME call? > This fragment repeats over and over, sometimes with PICTYPE_BITMAP > instead. Might be worth refactoring that into a function. todo_wine > works for function boundaries. If you refactor, you should pass that > function __LINE__ information and print it in your OK calls to know > which of the tests is failing if something fails. > > You do a lot of tests like this. Setting disp to "(void*)0xdeadbeef" > instead before allows you to differentiate between disp unmodified and > disp explicitly set to zero. Good ideas. I'll do that.

2 1

Re: cryptdlg(2/2): Implement FormatVerisignExtension
by Alexandre Julliard Nov. 18, 2008

Nov. 18, 2008

"Juan Lang" <juan.lang(a)gmail.com> writes: > + if (dwFormatStrType & CRYPT_FORMAT_STR_NO_HEX) > + { > + static BOOL stringLoaded = FALSE; > + DWORD bytesNeeded; > + > + if (!stringLoaded) > + { > + LoadStringW(hInstance, IDS_VERISIGN_STATEMENT, verisign_statement, > + sizeof(verisign_statement) / sizeof(verisign_statement[0])); > + stringLoaded = TRUE; > + } There's no need to store a copy of the string. You should load it directly into the destination buffer. -- Alexandre Julliard julliard(a)winehq.org

2 2

Another source of hardware statistics!
by Dan Kegel Nov. 18, 2008

Nov. 18, 2008

Unity3d is a game engine for casual gaming (not hardcore gaming). They have some good data on their user base's hardware at http://unity3d.com/webplayer/hwstats e.g. their breakdown of video cards for this quarter is at http://unity3d.com/webplayer/hwstats/pages/web-2008Q4-gfxcard.html Their top 3 graphics cards are Intel 945 (GMA 950) 13.3% Nvidia GeForce 6100 7.5% ATI Radeon X300/X550/Xpress200/1100 6.9% And believe it or not, 8% of their users still only have 256MB of RAM... - Dan

1 0

Re: fix a crash in widl on Vista x86
by Rob Shearman Nov. 18, 2008

Nov. 18, 2008

2008/11/17 <netzimme(a)aim.com>: > Changelog: > Add a FIXME for the crash in widl on Vista x86. > > diff --git a/tools/widl/typegen.c b/tools/widl/typegen.c > index fcb141e..45467e5 100644 > --- a/tools/widl/typegen.c > +++ b/tools/widl/typegen.c > @@ -3137,7 +3137,7 @@ size_t get_size_procformatstring(const > statement_list_t *stmts, type_pred_t pred > size += get_size_procformatstring(stmt->u.lib->stmts, pred) - 1; > continue; > } > - else if (stmt->type != STMT_TYPE && stmt->u.type->type != > RPC_FC_IP) > + else if (stmt->type != STMT_TYPE || stmt->u.type->type != > RPC_FC_IP) > continue; > > iface = stmt->u.type; Looks good. I had the same fix queued up in my tree. However, your changelog message needs some work. The bug isn't specific to Vista x86, but is probably specific to the IDL file you are compiling. -- Rob Shearman

1 0

NT driver address space
by Alexander Morozov Nov. 18, 2008

Nov. 18, 2008

Hi, Under Wine NT drivers are running in user address space. Some drivers do not work correctly because of this. For example Guardant USB token driver crashes when loading but this problem can be fixed with such patch (really I tested with our wine fork): diff --git a/dlls/ntdll/virtual.c b/dlls/ntdll/virtual.c index 78e38dc..f607f22 100644 --- a/dlls/ntdll/virtual.c +++ b/dlls/ntdll/virtual.c @@ -904,6 +904,7 @@ static NTSTATUS map_image( HANDLE hmapping, int fd, char *base, SIZE_T total_siz server_enter_uninterrupted_section( &csVirtual, &sigset ); + if (0x10000 == base) base = 0x81000000; if (base >= (char *)0x110000) /* make sure the DOS area remains free */ status = map_view( &view, base, total_size, mask, FALSE, VPROT_COMMITTED | VPROT_READ | VPROT_EXEC | VPROT_WRITECOPY | VPROT_IMAGE ); diff --git a/libs/wine/mmap.c b/libs/wine/mmap.c index ff117cc..4c18683 100644 --- a/libs/wine/mmap.c +++ b/libs/wine/mmap.c @@ -346,7 +346,7 @@ void mmap_init(void) #if defined(__i386__) && !defined(__FreeBSD__) && !defined(__FreeBSD_kernel__) /* commented out until FreeBSD gets fixed */ char stack; char * const stack_ptr = &stack; - char *user_space_limit = (char *)0x7ffe0000; + char *user_space_limit = (char *)0xaffe0000; reserve_malloc_space( 8 * 1024 * 1024 ); I see two ways of fixing this. 1. Modify wine to run winedevice.exe in kernel address space (> 0x80000000) by checking name of executable. 2. Modify ExAllocatePool and some other functions to use kernel address space, modify functions which is used by drivers (ntoskrnl.exe exports etc.) to use only this memory allocation functions, modify winedevice.exe to load driver and create stack in kernel address space. What is better decision? Thanks, Alexander Morozov

1 0

Re: Patch for [Bug 16087] wine spews innumerable "Unable to check compatibility for Format..." errors in some games
by Henri Verbeet Nov. 18, 2008

Nov. 18, 2008

2008/11/18 Stefan Dösinger <stefan(a)codeweavers.com>: > I think it would be better f getColorBits returned a proper error to the > caller, and the caller handles these errors without spamming the FIXME or > ERR channel. That would seem better to me than prefiltering the parameters > It does, it returns FALSE for "unsupported" formats. The only thing the patch fixes is the FIXME & ERR output. I think there are places where we'd like getColorBits() to work properly for float formats though.

1 0

Re: dlls/cabinet: fix dead stores (llvm/clang)
by Dmitry Timoshkov Nov. 18, 2008

Nov. 18, 2008

"ricardo filipe" <ricardo_barbano(a)hotmail.com> wrote: > --- a/dlls/cabinet/fci.c > +++ b/dlls/cabinet/fci.c > @@ -462,7 +462,7 @@ static cab_ULONG fci_get_checksum(const void *pv, UINT cb, CHECKSUM seed) > case 2: > ul |= (((ULONG)(*pb++)) << 8); > case 1: > - ul |= *pb++; > + ul |= *pb; > default: > break; > } Is there any reason that you ignored 2 similar cases above? -- Dmitry.

3 2

Re: Patch for [Bug 16087] wine spews innumerable "Unable to check compatibility for Format..." errors in some games
by Henri Verbeet Nov. 18, 2008

Nov. 18, 2008

> + switch(Format) > + { > + case WINED3DFMT_X8R8G8B8: > + case WINED3DFMT_R8G8B8: > + case WINED3DFMT_A8R8G8B8: > + case WINED3DFMT_A2R10G10B10: > + case WINED3DFMT_X1R5G5B5: > + case WINED3DFMT_A1R5G5B5: > + case WINED3DFMT_R5G6B5: > + case WINED3DFMT_X4R4G4B4: > + case WINED3DFMT_A4R4G4B4: > + case WINED3DFMT_R3G3B2: > + case WINED3DFMT_A8P8: > + case WINED3DFMT_P8: > + /* taken from utils.c/getColorBits() */ > + break; > + default: > + /* getColorBits doesn't work here, so avoid the silly error messages. */ > + return FALSE; > + } As you mention in the bug report, this is not a proper fix. We should probably fix the format table to use a shift and a bitcount.

1 0

Re: [PATCH 1/2] Implement GdipGetLogFontA (try3)
by Adam Petaccia Nov. 17, 2008

Nov. 17, 2008

Was there something wrong with this patch series? On Tue, 2008-11-11 at 11:24 -0500, Adam Petaccia wrote: > Changelog: > try3: > Only copy the bytes we'll use > Toss out bytesWritten > try2: > Fix memcpy() use > Add documentation > Add testcase > --- > dlls/gdiplus/font.c | 27 +++++++++++++++++++++++++++ > dlls/gdiplus/gdiplus.spec | 2 +- > include/gdiplusflat.h | 1 + > 3 files changed, 29 insertions(+), 1 deletions(-) > > diff --git a/dlls/gdiplus/font.c b/dlls/gdiplus/font.c > index 32713c7..7f815b6 100644 > --- a/dlls/gdiplus/font.c > +++ b/dlls/gdiplus/font.c > @@ -368,6 +368,33 @@ GpStatus WINGDIPAPI GdipGetFontUnit(GpFont *font, Unit *unit) > } > > /******************************************************************************* > + * GdipGetLogFontA [GDIPLUS.@] > + * > + * Converts a font into a LOGFONTA > + * > + * PARAMS > + * font [I] GpFont to create LOGFONTA object > + * graphics [I] Current graphics context > + * lfa [O] Resulting LOGFONTA object > + * RETURNS > + * InvalidParameter: font, graphics, or lfa was NULL. > + * Ok: otherwise > + */ > +GpStatus WINGDIPAPI GdipGetLogFontA(GpFont *font, GpGraphics *graphics, > + LOGFONTA *lfa) > +{ > + /* FIXME: use graphics */ > + if(!(font && graphics && lfa)) > + return InvalidParameter; > + > + memcpy(lfa, &font->lfw, FIELD_OFFSET(LOGFONTA, lfFaceName)); > + WideCharToMultiByte(CP_ACP, 0, font->lfw.lfFaceName, -1, > + lfa->lfFaceName, LF_FACESIZE, NULL, NULL); > + > + return Ok; > +} > + > +/******************************************************************************* > * GdipGetLogFontW [GDIPLUS.@] > */ > GpStatus WINGDIPAPI GdipGetLogFontW(GpFont *font, GpGraphics *graphics, > diff --git a/dlls/gdiplus/gdiplus.spec b/dlls/gdiplus/gdiplus.spec > index c25d4b7..dce558c 100644 > --- a/dlls/gdiplus/gdiplus.spec > +++ b/dlls/gdiplus/gdiplus.spec > @@ -311,7 +311,7 @@ > @ stdcall GdipGetLineSpacing(ptr long ptr) > @ stub GdipGetLineTransform > @ stdcall GdipGetLineWrapMode(ptr ptr) > -@ stub GdipGetLogFontA > +@ stdcall GdipGetLogFontA(ptr ptr ptr) > @ stdcall GdipGetLogFontW(ptr ptr ptr) > @ stdcall GdipGetMatrixElements(ptr ptr) > @ stub GdipGetMetafileDownLevelRasterizationLimit > diff --git a/include/gdiplusflat.h b/include/gdiplusflat.h > index ea4bec5..a2c6fc9 100644 > --- a/include/gdiplusflat.h > +++ b/include/gdiplusflat.h > @@ -81,6 +81,7 @@ GpStatus WINGDIPAPI GdipCreateFontFromDC(HDC,GpFont**); > GpStatus WINGDIPAPI GdipCreateFontFromLogfontA(HDC,GDIPCONST LOGFONTA*,GpFont**); > GpStatus WINGDIPAPI GdipCreateFontFromLogfontW(HDC,GDIPCONST LOGFONTW*,GpFont**); > GpStatus WINGDIPAPI GdipDeleteFont(GpFont*); > +GpStatus WINGDIPAPI GdipGetLogFontA(GpFont*,GpGraphics*,LOGFONTA*); > GpStatus WINGDIPAPI GdipGetLogFontW(GpFont*,GpGraphics*,LOGFONTW*); > GpStatus WINGDIPAPI GdipGetFamily(GpFont*, GpFontFamily**); > GpStatus WINGDIPAPI GdipGetFontUnit(GpFont*, Unit*);

2 1

Re: implement OleLoadPictureFile
by Austin English Nov. 17, 2008

Nov. 17, 2008

On Wed, Nov 12, 2008 at 2:16 AM, Jeremy Drake <wine(a)jdrake.com> wrote: > I ran into a function that was not implemented when attempting to run a > program I work on under wine. > > This patch implements OleLoadPictureFile based on the MSDN docs > (http://msdn.microsoft.com/en-us/library/ms221680.aspx) and what I saw > when stepping through in windbg on XP. > > A few notes: > I'm not sure I got the spec file right. The first param is a VARIANT, not > a VARIANT *, so I'm not sure ptr is the right thing. > > I'm not sure the right way to do the TRACE call with the VARIANT param. > > I did not implement OleLoadPictureFileEx. Technically, OleLoadPictureFile > should call OleLoadPictureFileEx(varFilename, LP_DEFAULT, LP_DEFAULT, > LP_DEFAULT, ppdispPicture) and OleLoadPictureEx should pass the sizeX, > sizeY, and flags to OleLoadPictureEx. > > When stepping through the function in windbg, I saw it call a function > OLEAUT32!CreateFileStream which does not exist in wine anywhere. Instead > of using that, I used the same mechanism as was used in OleLoadPicturePath > (GlobalAlloc the size of the file, read the whole file in, and > CreateStreamOnHGlobal). > > Since this is my first wine patch, I hope you will forgive me these > shortcomings. I'm not subscribed to any wine list, so please copy me on > any replies. > > Thanks, > Jeremy > > > > -- > Personifiers Unite! You have nothing to lose but Mr. Dignity! > > > Can you add a testcase to show that this behavior is correct? -- -Austin

5 12

re: Patch for [Bug 16087] wine spews innumerable "Unable to check compatibility for Format..." errors in some games
by Dan Kegel Nov. 17, 2008

Nov. 17, 2008

Andrew wrote: >When sending a patch to wine-patches put the patch number followed by >how many patches you are sending so the automated scripts can read it. yes, please! > For example "[1/1] sending only one patch". Poor example, please don't do it on single patches. You only need to do it for series of more than one patch. Also please don't send patches that don't apply cleanly against current git. (You'd be surprised how many people do this. It's ok if Alexandre can figure it out, but poor old Patchwatcher will reject it.) - Dan

1 0

Re: [1/5] WineD3D: Get some vertex pipeline infrastructure in place
by Henri Verbeet Nov. 17, 2008

Nov. 17, 2008

2008/11/17 Stefan Dösinger <stefan(a)codeweavers.com>: > - } else if (fixup || > + } else if (device->vertex_pipe->can_convert_d3dcolor || fixup || Shouldn't this be redundant if you already handle color conversion in IWineD3DVertexBufferImpl_FindDecl()? > - } else { > + } else if(!device->vertex_pipe->can_convert_d3dcolor) { I don't think this is how we want to do this. I think it makes more sense to ask the pipeline if it can handle a particular type for a particular usage. You can handle FLOAT16 support in the same way then. Also note that EXT_vertex_array_bgra adds support for D3DCOLOR, but only for diffuse, specular and generic attributes. That should be enough for anything we come across, but with a "can_convert_d3dcolor" flag like that you can't handle that extension properly.

2 3

Re: Patch for [Bug 16087] wine spews innumerable "Unable to check compatibility for Format..." errors in some games
by Andrew Fenn Nov. 17, 2008

Nov. 17, 2008

Your patch adds extra whitespace on an empty line. You are mixing coding styles with your if statements. When sending a patch to wine-patches put the patch number followed by how many patches you are sending so the automated scripts can read it. For example "[1/1] sending only one patch". 2008/11/16 Bogdan Butnaru <bogdanb(a)gmail.com>: > Patch for http://bugs.winehq.org/show_bug.cgi?id=16087 > > -- > Bogdan Butnaru > > > >

1 0

Re: (try 5) comdlg32/tests: Windows XP+ cannot do a CreateViewWindow2 twice in rapid seccesion. handle this error.
by Alexandre Julliard Nov. 17, 2008

Nov. 17, 2008

Aric Stewart <aric(a)codeweavers.com> writes: > @@ -182,19 +182,26 @@ static UINT CALLBACK create_view_window2_hook(HWND dlg, UINT msg, WPARAM wParam, > > hr = IShellView2_GetCurrentInfo(shell_view2, &folder_settings); > ok(SUCCEEDED(hr), "GetCurrentInfo returned %#x\n", hr); > - ok(folder_settings.ViewMode == FVM_LIST, "view mode is %d, expected %d\n", folder_settings.ViewMode, FVM_LIST); > + ok(folder_settings.ViewMode == FVM_DETAILS, "view mode is %d, expected %d\n", folder_settings.ViewMode, FVM_DETAILS); > > hr = IShellView2_DestroyViewWindow(shell_view2); > ok(SUCCEEDED(hr), "DestroyViewWindow returned %#x\n", hr); > > - view_params.pvid = &VID_Details; > + folder_settings.ViewMode = FVM_LIST; > + folder_settings.fFlags = 0; > + view_params.pvid = NULL; > hr = IShellView2_CreateViewWindow2(shell_view2, &view_params); > - ok(SUCCEEDED(hr), "CreateViewWindow2 returned %#x\n", hr); > + /* > + * Windows XP is unable to recreate the ViewWindow2 returning > + * A Catastrophic failure error > + */ So why can't it recreate it? Do you need to clean up something else beforehand? -- Alexandre Julliard julliard(a)winehq.org

2 1

[wdm.h] AddDevice member
by Alessandro Nov. 17, 2008

Nov. 17, 2008

Hi, I've noticed that the AddDevice member of DRIVER_EXTENSION is not declared as a functio pointer. I've put up a patch to fix this. Alessandro Pignotti -- Vi Veri Veniversum Vivus Vici -Dr. Faustus - Marlowe Public GPG Key ID 0x650B3ED9 on subkeys.gpg.net Key Fingerprint 6243 AAD3 E3EC 52D8 DFAA 2A2F 9FCD 0457 650B 3ED9 Encrypted mails are welcome

1 0

Re: msxml3: Determine whether to call xsltInit() or not at run time.
by Michael Karcher Nov. 17, 2008

Nov. 17, 2008

Am Montag, den 17.11.2008, 15:09 +0100 schrieb Francois Gouget: > configure will need to be regenerated. I did not include it here since > the consensus seems to be that it should not be included. Just a nitpick: Same applies for config.h.in, then. > As a reminder, the issue is that it's not because the libxslt.so we > compile against has xlstInit() that the one we load at runtime will have > it too. And vice versa. Wait a moment. I don't know about Wine's dl* functions, but can't we use dlsym with the RTLD_DEFAULT handle to look in the libxslt msxml3.dll.so is linked with anyway? That would do away with the need to know the soname during compilation. As it is hardcoded into the executable during link, hard-coding the soname doesn't hurt that much too, but still it gets simpler Regards, Michael Karcher

2 1

Race in thread shutdown in imm32?
by Dan Kegel Nov. 17, 2008

Nov. 17, 2008

I'm seeing the following valgrind warning in three out of eight runs under heavy load: InvalidRead IMM_FreeThreadData:233 DllMain:382 __wine_spec_dll_entry:40 MODULE_InitDLL:911 LdrShutdownThread:2174 call_thread_func:403 start_thread:444 It kind of feels like a race between thread shutdown and process shutdown. Does that ring a bell with anyone? - Dan

3 3

New error: cp: X and X are the same file
by Dan Kegel Nov. 17, 2008

Nov. 17, 2008

I'm seeing a lot of errors like this lately: cp: `/home/dank/.local/share/icons/3550_shell32.0.xpm' and `/home/dank/.local/share/icons/3550_shell32.0.xpm' are the same file Did the recent rewrite of menu item creation break something?

1 0

Re: ntoskrnl.exe: Implement MmGetSystemRoutineAddress
by Dmitry Timoshkov Nov. 17, 2008

Nov. 17, 2008

"Alistair Leslie-Hughes" <leslie_alistair(a)hotmail.com> wrote: > + if(RtlUnicodeStringToAnsiString(&routineNameA, SystemRoutineName, TRUE) == STATUS_SUCCESS) > + { > + /* We only support functions exported from ntoskrnl.exe or hal.dll */ > + hMod = GetModuleHandleW(ntoskrnlW); > + if(hMod) > + { > + pFunc = GetProcAddress(hMod, routineNameA.Buffer); > + } You are calling GetModuleHandleW from inside of ntoskrnl.exe, how that could fail? > + if(!pFunc) > + { > + hMod = GetModuleHandleW(halW); > + if(hMod) > + { > + pFunc = GetProcAddress(hMod, routineNameA.Buffer); > + } > + } > + > + RtlFreeAnsiString(&routineNameA); > + } > + > + TRACE("Function addresss %p\n", pFunc); > + > + return pFunc; > +} Please follow the existing coding style, place a space after 'if' and avoid not necessary braces for single line statements after 'if'. -- Dmitry.

1 0

ComputeSphereVisibility: a patch
by paulo lesgaz Nov. 17, 2008

Nov. 17, 2008

Hi, here is a patch for a first try to implement ComputeSphereVisibility. Any feedback is welcome. David

3 5

Grumble. Kernel lockup in Intrepid.
by Dan Kegel Nov. 17, 2008

Nov. 17, 2008

Wonder why patchwatcher was down for 12 hours? It seems to have been a kernel problem. This is on an Ubuntu 8.10 32 bit Core 2 Duo system on a P5K Pro motherboard, with a PCI-E NVidia card. I have the proprietary Nvidia drivers installed, so I can't expect support from anybody, but I did want to at least kvetch in public somewhere. The build slave was unresponsive. Upon rebooting, I found the following in /var/log/messages: Nov 15 18:07:05 slave2 kernel: [1146762.712500] Modules linked in: nfs lockd nfs_acl sunrpc af_packet binfmt_misc rfcomm bridge stp bnep sco l2cap bluetooth ppdev ipv6 acpi_cpufreq cpufreq_stats cpufreq_userspace cpufreq_powersave cpufreq_conservative cpufreq_ondemand freq_table sbs wmi sbshc video output container pci_slot battery iptable_filter ip_tables x_tables ac sbp2 parport_pc lp parport snd_hda_intel psmouse pcspkr serio_raw evdev snd_pcm_oss snd_mixer_oss nvidia(P) snd_pcm i2c_core snd_seq_dummy iTCO_wdt iTCO_vendor_support snd_seq_oss snd_seq_midi snd_rawmidi snd_seq_midi_event snd_seq snd_timer snd_seq_device snd shpchp soundcore button pci_hotplug snd_page_alloc intel_agp agpgart ext3 jbd mbcache sr_mod cdrom sd_mod crc_t10dif sg pata_marvell pata_acpi usbhid hid natsemi ata_generic ohci1394 ieee1394 ata_piix libata scsi_mod dock sky2 ehci_hcd uhci_hcd usbcore thermal processor fan fbcon tileblit font bitblit softcursor fuse Nov 15 18:07:05 slave2 kernel: [1146762.712501] Nov 15 18:07:05 slave2 kernel: [1146762.712501] Pid: 21048, comm: wrc Tainted: P (2.6.27-7-generic #1) Nov 15 18:07:05 slave2 kernel: [1146762.712501] EIP: 0060:[<c0184320>] EFLAGS: 00000286 CPU: 1 Nov 15 18:07:05 slave2 kernel: [1146762.712501] EIP is at find_get_pages+0x70/0x110 Nov 15 18:07:05 slave2 kernel: [1146762.712501] EAX: ee96fa7c EBX: c1d5ac50 ECX: e2333d38 EDX: 00000000 Nov 15 18:07:05 slave2 kernel: [1146762.712501] ESI: e2333d30 EDI: e2333d38 EBP: e2333ce8 ESP: e2333cb8 Nov 15 18:07:05 slave2 kernel: [1146762.712501] DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068 Nov 15 18:07:05 slave2 kernel: [1146762.712501] CR0: 8005003b CR2: 09c28000 CR3: 35009000 CR4: 00000690 Nov 15 18:07:05 slave2 kernel: [1146762.712501] DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000 Nov 15 18:07:05 slave2 kernel: [1146762.712501] DR6: ffff0ff0 DR7: 00000403 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c018d4d7>] pagevec_lookup+0x27/0x40 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c018e2ac>] truncate_inode_pages_range+0x8c/0x360 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<f89e9c7f>] ? do_get_write_access+0x2df/0x4b0 [jbd] Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<f8a23174>] ? ext3_get_group_desc+0x14/0xd0 [ext3] Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c018e59f>] truncate_inode_pages+0x1f/0x30 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c0195d11>] vmtruncate+0x161/0x190 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01c85d2>] inode_setattr+0x62/0x190 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<f8a28ef9>] ext3_setattr+0xd9/0x210 [ext3] Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01c89d6>] fnotify_change+0x2d6/0x3b0 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01b9d81>] ? path_permission+0x31/0x40 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01b0f56>] do_truncate+0x76/0xa0 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c023528f>] ? apparmor_path_permission+0x5f/0x80 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01ba6f3>] may_open+0x193/0x210 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01bda75>] do_filp_open+0x115/0x790 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01c9619>] ? expand_files+0x9/0x60 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01c9750>] ? alloc_fd+0xe0/0x100 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01afff5>] do_sys_open+0x65/0x100 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c01b00fe>] sys_open+0x2e/0x40 Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c0103f7b>] sysenter_do_call+0x12/0x2f Nov 15 18:07:05 slave2 kernel: [1146762.712501] [<c0370000>] ? default_device_exit+0x60/0xb0 It crashed in the middle of a wine build. The last thing in the patchwatcher log was make[1]: Leaving directory `/home/patchslave/winezeug/patchwatcher/wine-continuous-workdir/active/dlls' make[2]: Entering BIOS reported the temperature was 56 degrees after rebooting, fwiw. This is the first problem on this machine since I set it up several weeks ago. I guess I'll just reboot it and hope...

2 1

d3dx9: Add ID3DXSprite tests (review please)
by Tony Wasserka Nov. 16, 2008

Nov. 16, 2008

This is my first test, so I think it's better to ask for feedback on wine-devel, first. There are some points, where I wasn't sure what to do: line 26: Is using inline okay here? IIRC that's C++ only, but other tests do it the same way... line 82: Is it legal to demand A8R8G8B8 support here? Otherwise, how would I do this without making the code overcomplicated? For what other formats should I check then? line 120-123: I hope it's okay to use hardcoded matrix values here, as the point of that test is only to make sure that the same matrix gets returned... line 215-253: 1) Is it okay to initialize D3D that way? 2) Can I simplify the initialization even further? Also, I'd like to know wether the code in general is clean enough. Any other feedback is welcome, too. Best regards, Tony (Note1: This test only finishes successfully if my recent ID3DXSprite patch series is applied, but I already had all of them pass/todo'ed. Note2: visual tests will follow later, this is just for the basic functionality) >From 02587ae3cedb0b68fba1c360f63d2571c175bf07 Mon Sep 17 00:00:00 2001 From: Tony Wasserka <tony.wasserka(a)freenet.de> Date: Sun, 16 Nov 2008 12:26:06 +0100 Subject: [PATCH] d3dx9: Add ID3DXSprite tests --- dlls/d3dx9_36/tests/Makefile.in | 3 +- dlls/d3dx9_36/tests/core.c | 252 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 254 insertions(+), 1 deletions(-) create mode 100644 dlls/d3dx9_36/tests/core.c diff --git a/dlls/d3dx9_36/tests/Makefile.in b/dlls/d3dx9_36/tests/Makefile.in index 40500e2..1ac8b7e 100644 --- a/dlls/d3dx9_36/tests/Makefile.in +++ b/dlls/d3dx9_36/tests/Makefile.in @@ -3,9 +3,10 @@ TOPOBJDIR = ../../.. SRCDIR = @srcdir@ VPATH = @srcdir@ TESTDLL = d3dx9_36.dll -IMPORTS = d3dx9 kernel32 +IMPORTS = d3dx9 d3d9 user32 kernel32 CTESTS = \ + core.c \ math.c \ shader.c diff --git a/dlls/d3dx9_36/tests/core.c b/dlls/d3dx9_36/tests/core.c new file mode 100644 index 0000000..2d15859 --- /dev/null +++ b/dlls/d3dx9_36/tests/core.c @@ -0,0 +1,252 @@ +/* + * Tests for the D3DX9 core interfaces + * + * Copyright 2008 Tony Wasserka + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA + */ + +#define COBJMACROS +#include "wine/test.h" +#include <dxerr9.h> +#include "d3dx9core.h" + +static inline int get_ref(IUnknown *obj) +{ + if(!obj) return 0; + IUnknown_AddRef(obj); + return IUnknown_Release(obj); +} + +#define CHECK_REF(obj, exp) \ + ok(exp==get_ref((IUnknown*)(obj)), "Invalid refcount. Expected %d, got %d\n", exp, get_ref((IUnknown*)(obj))); + +#define CHECK_RELEASE(obj, exp) \ + { \ + if(obj) { \ + int ref=IUnknown_Release((IUnknown*)(obj)); \ + ok(ref==exp, "Invalid refcount. Expected %d, got %d\n", exp, ref); \ + } else skip("Object was NULL before Release\n"); \ + } + +#define CHECK_CALL(call, name, exp) \ + hr=(call); \ + ok(hr==(exp), "%s returned %#x, expected %#x\n", (name), hr, (exp)); + +#define admitted_error 0.0001f +static inline void check_mat(D3DXMATRIX got, D3DXMATRIX exp) +{ + int i, j, equal=1; + for (i=0; i<4; i++) + for (j=0; j<4; j++) + if (fabs(U(exp).m[i][j]-U(got).m[i][j]) > admitted_error) + equal=0; + + ok(equal, "Got matrix\n\t(%f,%f,%f,%f\n\t %f,%f,%f,%f\n\t %f,%f,%f,%f\n\t %f,%f,%f,%f)\n" + "Expected matrix=\n\t(%f,%f,%f,%f\n\t %f,%f,%f,%f\n\t %f,%f,%f,%f\n\t %f,%f,%f,%f)\n", + U(got).m[0][0],U(got).m[0][1],U(got).m[0][2],U(got).m[0][3], + U(got).m[1][0],U(got).m[1][1],U(got).m[1][2],U(got).m[1][3], + U(got).m[2][0],U(got).m[2][1],U(got).m[2][2],U(got).m[2][3], + U(got).m[3][0],U(got).m[3][1],U(got).m[3][2],U(got).m[3][3], + U(exp).m[0][0],U(exp).m[0][1],U(exp).m[0][2],U(exp).m[0][3], + U(exp).m[1][0],U(exp).m[1][1],U(exp).m[1][2],U(exp).m[1][3], + U(exp).m[2][0],U(exp).m[2][1],U(exp).m[2][2],U(exp).m[2][3], + U(exp).m[3][0],U(exp).m[3][1],U(exp).m[3][2],U(exp).m[3][3]); +} + +static void test_ID3DXSprite(IDirect3DDevice9 *device) +{ + ID3DXSprite *sprite=NULL; + IDirect3D9 *d3d; + IDirect3DDevice9 *cmpdev=NULL; + IDirect3DTexture9 *tex1=NULL, *tex2=NULL; + D3DXMATRIX mat, cmpmat; + D3DVIEWPORT9 vp; + RECT rect; + D3DXVECTOR3 pos, center; + HRESULT hr; + + IDirect3DDevice9_GetDirect3D(device, &d3d); + if(FAILED(IDirect3D9_CheckDeviceFormat(d3d, 0, D3DDEVTYPE_HAL, D3DFMT_UNKNOWN, 0, D3DRTYPE_TEXTURE, D3DFMT_A8R8G8B8))) { + skip("D3DFMT_A8R8B8G8 not supported\n"); + return; + } + IDirect3D9_Release(d3d); + + if(FAILED(hr=IDirect3DDevice9_CreateTexture(device, 64, 64, 1, 0, D3DFMT_A8R8G8B8, D3DPOOL_DEFAULT, &tex1, NULL))) { + skip("Couldn't create first texture, CreateTexture returned %#x\n", hr); + return; + } + if(FAILED(hr=IDirect3DDevice9_CreateTexture(device, 32, 32, 1, 0, D3DFMT_A8R8G8B8, D3DPOOL_DEFAULT, &tex2, NULL))) { + skip("Couldn't create second texture, CreateTexture returned %#x\n", hr); + IDirect3DTexture9_Release(tex1); + return; + } + /* Test D3DXCreateSprite */ + CHECK_CALL(D3DXCreateSprite(device, NULL), "D3DXCreateSprite", D3DERR_INVALIDCALL); + CHECK_CALL(D3DXCreateSprite(NULL, &sprite), "D3DXCreateSprite", D3DERR_INVALIDCALL); + CHECK_CALL(D3DXCreateSprite(device, &sprite), "D3DXCreateSprite", D3D_OK); + + /* Test ID3DXSprite_GetDevice */ + CHECK_CALL(ID3DXSprite_GetDevice(sprite, NULL), "GetDevice", D3DERR_INVALIDCALL); + CHECK_CALL(ID3DXSprite_GetDevice(sprite, &cmpdev), "GetDevice", D3D_OK); /* cmpdev=NULL */ + CHECK_CALL(ID3DXSprite_GetDevice(sprite, &cmpdev), "GetDevice", D3D_OK); /* cmpdev!=NULL */ + IDirect3DDevice9_Release(device); + IDirect3DDevice9_Release(device); + + /* Test ID3DXSprite_GetTransform */ + CHECK_CALL(ID3DXSprite_GetTransform(sprite, NULL), "GetTransform", D3DERR_INVALIDCALL); + CHECK_CALL(ID3DXSprite_GetTransform(sprite, &mat), "GetTransform", D3D_OK); + if(SUCCEEDED(hr)) { + D3DXMATRIX identity; + D3DXMatrixIdentity(&identity); + check_mat(mat, identity); + } else skip("No matrix to test\n"); + + /* Test ID3DXSprite_SetTransform */ + /* Set a transform and test if it gets returned correctly */ + U(mat).m[0][0]=2.1f; U(mat).m[0][1]=6.5f; U(mat).m[0][2]=-9.6f; U(mat).m[0][3]=1.7f; + U(mat).m[1][0]=4.2f; U(mat).m[1][1]=-2.5f; U(mat).m[1][2]=2.1f; U(mat).m[1][3]=5.5f; + U(mat).m[2][0]=-2.6f; U(mat).m[2][1]=0.3f; U(mat).m[2][2]=8.6f; U(mat).m[2][3]=8.4f; + U(mat).m[3][0]=6.7f; U(mat).m[3][1]=-5.1f; U(mat).m[3][2]=6.1f; U(mat).m[3][3]=2.2f; + CHECK_CALL(ID3DXSprite_SetTransform(sprite, NULL), "SetTransform", D3DERR_INVALIDCALL); + CHECK_CALL(ID3DXSprite_SetTransform(sprite, &mat), "SetTransform", D3D_OK); + if(SUCCEEDED(hr)) { + hr=ID3DXSprite_GetTransform(sprite, &cmpmat); + if(SUCCEEDED(hr)) check_mat(cmpmat, mat); + else skip("GetTransform returned %#x\n", hr); + } else skip("No matrix to test\n"); + + /* Test ID3DXSprite_SetWorldViewLH/RH */ + todo_wine { + CHECK_CALL(ID3DXSprite_SetWorldViewLH(sprite, &mat, &mat), "SetWorldViewLH", D3D_OK); + CHECK_CALL(ID3DXSprite_SetWorldViewLH(sprite, NULL, &mat), "SetWorldViewLH", D3D_OK); + CHECK_CALL(ID3DXSprite_SetWorldViewLH(sprite, &mat, NULL), "SetWorldViewLH", D3D_OK); + CHECK_CALL(ID3DXSprite_SetWorldViewLH(sprite, NULL, NULL), "SetWorldViewLH", D3D_OK); + CHECK_CALL(ID3DXSprite_SetWorldViewRH(sprite, &mat, &mat), "SetWorldViewRH", D3D_OK); + CHECK_CALL(ID3DXSprite_SetWorldViewRH(sprite, NULL, &mat), "SetWorldViewRH", D3D_OK); + CHECK_CALL(ID3DXSprite_SetWorldViewRH(sprite, &mat, NULL), "SetWorldViewRH", D3D_OK); + CHECK_CALL(ID3DXSprite_SetWorldViewRH(sprite, NULL, NULL), "SetWorldViewRH", D3D_OK); + } + IDirect3DDevice9_BeginScene(device); + + /* Test ID3DXSprite_Begin*/ + CHECK_CALL(ID3DXSprite_Begin(sprite, 0), "Begin", D3D_OK); + IDirect3DDevice9_GetTransform(device, D3DTS_WORLD, &mat); + D3DXMatrixIdentity(&cmpmat); + check_mat(mat, cmpmat); + IDirect3DDevice9_GetTransform(device, D3DTS_VIEW, &mat); + check_mat(mat, cmpmat); + IDirect3DDevice9_GetTransform(device, D3DTS_PROJECTION, &mat); + IDirect3DDevice9_GetViewport(device, &vp); + D3DXMatrixOrthoOffCenterLH(&cmpmat, vp.X+0.5f, (float)vp.Width+vp.X+0.5f, (float)vp.Height+vp.Y+0.5f, vp.Y+0.5f, vp.MinZ, vp.MaxZ); + check_mat(mat, cmpmat); + + /* Test ID3DXSprite_Flush and ID3DXSprite_End */ + CHECK_CALL(ID3DXSprite_Flush(sprite), "Flush", D3D_OK); + CHECK_CALL(ID3DXSprite_End(sprite), "End", D3D_OK); + CHECK_CALL(ID3DXSprite_Flush(sprite), "Flush", D3DERR_INVALIDCALL); /* May not be called before next Begin */ + CHECK_CALL(ID3DXSprite_End(sprite), "End", D3DERR_INVALIDCALL); + + /* Test ID3DXSprite_Draw */ + CHECK_CALL(ID3DXSprite_Begin(sprite, 0), "Begin", D3D_OK); + if(FAILED(hr)) skip("Couldn't ID3DXSprite_Begin, can't test ID3DXSprite_Draw\n"); + else { /* Feed the sprite batch */ + int texref1, texref2; + + SetRect(&rect, 0, 0, 192, 128); + pos.x=0.f; pos.y=10.f; pos.z=0.f; + center.x=192.f; center.y=0.f; center.z=0.f; + + texref1=get_ref((IUnknown*)tex1); + texref2=get_ref((IUnknown*)tex2); + CHECK_CALL(ID3DXSprite_Draw(sprite, NULL, &rect, &center, &pos, D3DCOLOR_XRGB(255, 255, 255)), "Draw", D3DERR_INVALIDCALL); + CHECK_CALL(ID3DXSprite_Draw(sprite, tex1, &rect, &center, &pos, D3DCOLOR_XRGB(255, 255, 255)), "Draw", D3D_OK); + CHECK_CALL(ID3DXSprite_Draw(sprite, tex2, &rect, &center, &pos, D3DCOLOR_XRGB( 3, 45, 66)), "Draw", D3D_OK); + CHECK_REF(tex1, texref1+1); CHECK_REF(tex2, texref2+1); + CHECK_CALL(ID3DXSprite_Flush(sprite), "Flush", D3D_OK); + CHECK_REF(tex1, texref1); CHECK_REF(tex2, texref2); + + CHECK_CALL(ID3DXSprite_Draw(sprite, tex1, NULL, &center, &pos, D3DCOLOR_XRGB(255, 255, 255)), "Draw", D3D_OK); + CHECK_CALL(ID3DXSprite_Draw(sprite, tex1, &rect, NULL, &pos, D3DCOLOR_XRGB(255, 255, 255)), "Draw", D3D_OK); + CHECK_CALL(ID3DXSprite_Draw(sprite, tex1, &rect, &center, NULL, D3DCOLOR_XRGB(255, 255, 255)), "Draw", D3D_OK); + CHECK_CALL(ID3DXSprite_Draw(sprite, tex1, NULL, NULL, NULL, 0), "Draw", D3D_OK); + CHECK_CALL(ID3DXSprite_Flush(sprite), "Flush", D3D_OK); + CHECK_CALL(ID3DXSprite_Flush(sprite), "Flush", D3D_OK); /* Flushing twice should work */ + CHECK_CALL(ID3DXSprite_End(sprite), "End", D3D_OK); + } + + /* Test ID3DXSprite_OnLostDevice and ID3DXSprite_OnResetDevice */ + /* Both can be called twice */ + CHECK_CALL(ID3DXSprite_OnLostDevice(sprite), "OnLostDevice", D3D_OK); + CHECK_CALL(ID3DXSprite_OnLostDevice(sprite), "OnLostDevice", D3D_OK); + CHECK_CALL(ID3DXSprite_OnResetDevice(sprite), "OnResetDevice", D3D_OK); + CHECK_CALL(ID3DXSprite_OnResetDevice(sprite), "OnResetDevice", D3D_OK); + + /* Make sure everything works like before */ + CHECK_CALL(ID3DXSprite_Begin(sprite, 0), "Begin", D3D_OK); + CHECK_CALL(ID3DXSprite_Draw(sprite, tex2, &rect, &center, &pos, D3DCOLOR_XRGB(255, 255, 255)), "Draw", D3D_OK); + CHECK_CALL(ID3DXSprite_Flush(sprite), "Flush", D3D_OK); + CHECK_CALL(ID3DXSprite_End(sprite), "End", D3D_OK); + + /* OnResetDevice makes the interface "forget" the Begin call */ + CHECK_CALL(ID3DXSprite_Begin(sprite, 0), "Begin", D3D_OK); + CHECK_CALL(ID3DXSprite_OnResetDevice(sprite), "OnResetDevice", D3D_OK); + CHECK_CALL(ID3DXSprite_End(sprite), "End", D3DERR_INVALIDCALL); + + IDirect3DDevice9_EndScene(device); + CHECK_RELEASE(sprite, 0); + CHECK_RELEASE(tex2, 0); + CHECK_RELEASE(tex1, 0); +} + +START_TEST(core) +{ + HWND wnd=NULL; + IDirect3D9 *d3d = NULL; + IDirect3DDevice9 *device=NULL; + D3DPRESENT_PARAMETERS d3dpp; + HRESULT hr; + + wnd=CreateWindow("static", "d3dx9_test", 0, 0, 0, 0, 0, NULL, NULL, NULL, NULL); + d3d=Direct3DCreate9(D3D_SDK_VERSION); + if(!wnd) { + skip("Couldn't create application window\n"); + return; + } + if(!d3d) { + skip("Couldn't create Direct3D9 object\n"); + DestroyWindow(wnd); + return; + } + + ZeroMemory(&d3dpp, sizeof(d3dpp)); + d3dpp.Windowed = TRUE; + d3dpp.SwapEffect = D3DSWAPEFFECT_DISCARD; + hr=IDirect3D9_CreateDevice(d3d, D3DADAPTER_DEFAULT, D3DDEVTYPE_HAL, wnd, + D3DCREATE_MIXED_VERTEXPROCESSING, &d3dpp, &device); + if(FAILED(hr)) { + skip("Failed to create Direct3DDevice9 object %#x\n", hr); + IDirect3D9_Release(d3d); + DestroyWindow(wnd); + return; + } + + test_ID3DXSprite(device); + + CHECK_RELEASE(device, 0); + CHECK_RELEASE(d3d, 0); + if(wnd) DestroyWindow(wnd); +} -- 1.5.6.5

3 3

Re: [5/5] WineD3D: Move the fogdensity state to the misc state table
by Austin English Nov. 16, 2008

Nov. 16, 2008

On Sun, Nov 16, 2008 at 5:20 PM, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > fogstart and fogend will probably follow as well later, I still have to look > at those because they interfere with glEnable(GL_FOG) and the fog source > settings and are thus involved in the vertex pipeline as well > > > > > > You forgot the patch. -- -Austin

1 0