September 2015 - Wine-devel

Re: [PATCH 2/3] wineps.drv: Add support for PostScript Format 2 standard glyph names (try 3).
by Huw Davies 09 Sep '15

09 Sep '15

On 8 Sep 2015, at 18:30, Erich E. Hoover wrote: > +WORD get_post2_name_index(BYTE *post2header, DWORD size, WORD index) > +{ > + USHORT numberOfGlyphs = GET_BE_WORD(post2header); > + DWORD offset = (1 + index) * sizeof(USHORT); > + > + if((1 + numberOfGlyphs) * sizeof(USHORT) > size) > + { > + FIXME("Total number of glyphs (%d) exceeds PostScript Format 2 table size (%d)\n", > + numberOfGlyphs, size); > + return 0; /* .notdef */ > + } Personally, I'd remove this check, you only added it because you misunderstood something I suggested. What we actually care about is covered by the check below. > + if(offset + sizeof(USHORT) > size || index >= numberOfGlyphs) > + { > + FIXME("Index '%d' exceeds PostScript Format 2 table size (%d)\n", index, numberOfGlyphs); > + return 0; /* .notdef */ > + } In the failure case we need to not return 0 (.notdef), but should end up using the existing "gxxxx" glyph name. This is because glyphs are shown using the glyphshow PostScript operator, so we need unique names for every glyph. Note though that zero should be returned if the glyph's name really is .notdef. I guess returning 0xffff on failure would work, or something like that. Otherwise, the rest of the series looks good to me. Huw.

1 0

Re: d3dx9_36: Implement D3DXGetShaderOutputSemantics
by Marvin 09 Sep '15

09 Sep '15

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=16592 Your paranoid android. === build (build) === Patch failed to apply

1 0

Re: [2/2] oleaut32/tests: Add a test to show that support for coercion to arrays of variants is specific to vararg methods.
by Marvin 08 Sep '15

08 Sep '15

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=16571 Your paranoid android. === w2000pro (32 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === wxppro (32 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === w2003std (32 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === wvistau64 (32 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === w2008s64 (32 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === w7u (32 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === w8 (32 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === wvistau64 (64 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 === w2008s64 (64 bit tmarshal) === tmarshal.c:1522: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008

1 0

Re: [PATCH 2/3] wineps.drv: Add support for PostScript Format 2 standard glyph names.
by Huw Davies 08 Sep '15

08 Sep '15

On 8 Sep 2015, at 17:09, Erich E. Hoover wrote: > This version adds an additional check that the number of glyphs > reported to be in the table fits within the size of the table data. By checking size is big enough for numberOfGlyphs, I meant that you need to check whether size is big enough to read the first two bytes of the extended (version 2) table. Huw.

1 0

FW: [PATCH] msxml3: treat namespace as attribute in IXMLDOMElement::getAttribute (try 2)
by Daniel Lehman 08 Sep '15

08 Sep '15

Anything I need to add or change on this? https://source.winehq.org/patches/data/114075 thanks Daniel -----Original Message----- From: wine-patches [mailto:wine-patches-bounces(a)winehq.org] On Behalf Of Daniel Lehman Sent: Friday, August 28, 2015 2:49 PM To: 'wine-patches(a)winehq.org' Subject: [PATCH] msxml3: treat namespace as attribute in IXMLDOMElement::getAttribute (try 2) Try 2: add tests for failures

1 0

RFC - msacm32: Quit on unsupported destination format in PCM_FormatSuggest
by Bruno Jesus 08 Sep '15

08 Sep '15

Since this DLL is far away from my comprehension I'm first sending this to wine-devel. Current code [1] suggest that only WAVE_FORMAT_PCM is supported so do not allow the function to succeed if the destination format is not PCM. Fixes bug https://bugs.winehq.org/show_bug.cgi?id=14193 [1] 858 if (!(adfs->fdwSuggest & ACM_FORMATSUGGESTF_WFORMATTAG)) { 859 if (adfs->pwfxSrc->wFormatTag != WAVE_FORMAT_PCM) { 860 WARN("not possible\n"); 861 return ACMERR_NOTPOSSIBLE; 862 } 863 adfs->pwfxDst->wFormatTag = adfs->pwfxSrc->wFormatTag; 864 }

2 1

Re: [3/3] oleaut32/tests: Add a test to show that support for coercion to arrays of variants is specific to vararg methods.
by Marvin 08 Sep '15

08 Sep '15

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=16567 Your paranoid android. === w2000pro (32 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === wxppro (32 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2003std (32 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === wvistau64 (32 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2008s64 (32 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === w7u (32 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === w8 (32 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === wvistau64 (64 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2008s64 (64 bit tmarshal) === tmarshal.c:1558: Test failed: expected DISP_E_TYPEMISMATCH, got 0x80020008 tmarshal.c:1629: Test failed: IDispatch_Invoke failed with error 0x80020008

1 0

Re: [2/3] oleaut32: Add support for handling VT_VARIANT|VT_ARRAY|VT_BYREF argument type.
by Marvin 08 Sep '15

08 Sep '15

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=16566 Your paranoid android. === w2000pro (32 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === wxppro (32 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2003std (32 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === wvistau64 (32 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2008s64 (32 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === w7u (32 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === w8 (32 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === wvistau64 (64 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2008s64 (64 bit tmarshal) === tmarshal.c:1605: Test failed: IDispatch_Invoke failed with error 0x80020008

1 0

Re: [1/3] oleaut32/tests: Add another test for calling a varargs method.
by Marvin 08 Sep '15

08 Sep '15

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=16565 Your paranoid android. === w2000pro (32 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === wxppro (32 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2003std (32 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === wvistau64 (32 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2008s64 (32 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === w7u (32 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === w8 (32 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === wvistau64 (64 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008 === w2008s64 (64 bit tmarshal) === tmarshal.c:1606: Test failed: IDispatch_Invoke failed with error 0x80020008

1 0

Re: PATCH: widl: implement move-if-changed logic
by Jacek Caban 08 Sep '15

08 Sep '15

Hi Marcus, On 09/03/15 13:07, Marcus Meissner wrote: > From d8df64a7261bc0febcdd48d4ed818dc1f36f3471 Mon Sep 17 00:00:00 2001 > From: Marcus Meissner <meissner(a)suse.de> > Date: Thu, 3 Sep 2015 12:11:13 +0200 > Subject: [PATCH] widl: implement a move-if-changed logic > > This avoids unnecessary rebuilds if the generated code by widl > does not change. While I like the idea and it would often save a bug chunk of build time, I don't think it would work well with make. If you don't touch output file, then following make invocation will execute widl again, which is undesired. Thanks, Jacek

2 1