November 2012 - Wine-Devel - list.winehq.org

Re: [PATCH 2/4] [cmd] Prevent external env vars causing tests to fail
by Marvin Nov. 19, 2012

Nov. 19, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=23015 Your paranoid android. === WNT4WSSP6 (32 bit) === batch.c:312: Test failed: unexpected char 0x68 position 0 in line 860 (got 'h=%h i=a j=b k=c l=e m= o=%o', wanted 'h=%h i=a j=b k=c l=e m=%m o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 862 (got 'h=%h i=a j=b k=c l=d e f g m= n=%n o=%o', wanted 'h=%h i=a j=b k=c l=d e f g m=%m n=%n o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 863 (got 'h=%h i=a j=c k= l= m= n=%n o=%o', wanted 'h=%h i=a j=c k= l= m=%m n=%n o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 864 (got 'h=%h i=b j=c k= l= m= n=%n o=%o', wanted 'h=%h i=b j=c k= l= m=%m n=%n o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 865 (got 'h=%h i=b j=c k= l= m= n=%n o=%o', wanted 'h=%h i=b j=c k= l= m=%m n=%n o=%o')

1 0

Re: [PATCH 1/4] [cmd] Add support for tokens= (for /f)
by Marvin Nov. 19, 2012

Nov. 19, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=23014 Your paranoid android. === WNT4WSSP6 (32 bit) === batch.c:312: Test failed: unexpected char 0x68 position 0 in line 860 (got 'h=%h i=a j=b k=c l=e m= o=%o', wanted 'h=%h i=a j=b k=c l=e m=%m o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 862 (got 'h=%h i=a j=b k=c l=d e f g m= n=%n o=%o', wanted 'h=%h i=a j=b k=c l=d e f g m=%m n=%n o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 863 (got 'h=%h i=a j=c k= l= m= n=%n o=%o', wanted 'h=%h i=a j=c k= l= m=%m n=%n o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 864 (got 'h=%h i=b j=c k= l= m= n=%n o=%o', wanted 'h=%h i=b j=c k= l= m=%m n=%n o=%o') batch.c:312: Test failed: unexpected char 0x68 position 0 in line 865 (got 'h=%h i=b j=c k= l= m= n=%n o=%o', wanted 'h=%h i=b j=c k= l= m=%m n=%n o=%o')

1 0

Re: [PATCH 4/4] [kernel] Support opening of console for read
by Christian Costa Nov. 19, 2012

Nov. 19, 2012

+ * write_console_input + * + * Helper function for bare_console_fetch_input calling WriteConsoleInput + * The handle passed to ReadConsole etc may not have been opened for write, + * and hence the calls to WriteConsoleInput may fail. This routine handles such + * a failure, and retries the write by opening the console for write, using + * it and then closing it again. + * + * Returns as per WriteConsoleInputW (TRUE on success) + */ +BOOL WINAPI write_console_input( HANDLE handle, const INPUT_RECORD *buffer, + DWORD count, LPDWORD written ) +{ + BOOL result = WriteConsoleInputW(handle, buffer, count, written); + + /* If this fails with access denied */ + if (!result && GetLastError() == ERROR_ACCESS_DENIED) { + HANDLE hConRW = CreateFileW(coninW, GENERIC_READ | GENERIC_WRITE, + FILE_SHARE_READ, NULL, OPEN_EXISTING, + FILE_ATTRIBUTE_NORMAL, 0); + if (hConRW != INVALID_HANDLE_VALUE) { + result = WriteConsoleInputW(hConRW, buffer, count, written); + CloseHandle(hConRW); + } + } + + return result; +} Hi Jason, That would be better to use 4 spaces identation as the rest of the file. Christian

1 0

Re: [PATCH 1/2] usp10: Fixed wgBlank, wgDefault, wgInvalid and wgKashida (try 4).
by Alexandre Julliard Nov. 19, 2012

Nov. 19, 2012

Qian Hong <fracting(a)gmail.com> writes: > @@ -813,6 +926,12 @@ static HRESULT init_script_cache(const HDC hdc, SCRIPT_CACHE *psc) > return E_INVALIDARG; > } > sc->sfnt = (GetFontData(hdc, MS_MAKE_TAG('h','e','a','d'), 0, NULL, 0)!=GDI_ERROR); > + > + set_cache_blank_char(hdc, sc); > + set_cache_default_char(hdc, sc); > + set_cache_invalid_char(hdc, sc); > + set_cache_kashida_char(hdc, sc); > + > *psc = sc; > TRACE("<- %p\n", sc); > return S_OK; > @@ -1063,12 +1182,11 @@ HRESULT WINAPI ScriptGetFontProperties(HDC hdc, SCRIPT_CACHE *psc, SCRIPT_FONTPR > if (sfp->cBytes != sizeof(SCRIPT_FONTPROPERTIES)) > return E_INVALIDARG; > > - /* return something sensible? */ > - sfp->wgBlank = 0; > + sfp->wgBlank = get_cache_blank_char(psc); > sfp->wgDefault = get_cache_default_char(psc); > - sfp->wgInvalid = 0; > - sfp->wgKashida = 0xffff; > - sfp->iKashidaWidth = 0; > + sfp->wgInvalid = get_cache_invalid_char(psc); > + sfp->wgKashida = get_cache_kashida_char(psc); > + sfp->iKashidaWidth = 0; /* TODO */ It doesn't seem useful to have all these separate set and get functions, particularly since their implementation is very similar. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [1/5] windowscodecs: Implement QueryCapability in the PNG decoder.
by Vincent Povirk Nov. 19, 2012

Nov. 19, 2012

You should at least look at the stream and check that it looks like an image in a format the decoder can read. For PNG, you could maybe argue that given a correct magic number and a correctly-working decoder, the decoder must be able to read all the images or either the decoder or image is broken. Then again, you're not even checking the magic number.. For TGA, though, we don't have an exact magic number. It's entirely possible that we our TGA magic number matches some files that aren't really TGA images but are images in some other format that we can decode. We need QueryCapabilities to disambiguate the decoders in these cases. For JPEG and TIFF, which are extensible formats, there's also the possibility that an image has something in it that we don't support (and that the corresponding decoder on Windows doesn't support) but some other decoder on the system can. The method should also invalidate the decoder object, so we don't accidentally introduce code that reuses it.

2 3

Ubuntu 12.10 - anyone?
by Daniel Jelinski Nov. 19, 2012

Nov. 19, 2012

Hello, Is it possible to compile 32bit wine in 64bit ubuntu 12.10? I heard there were problems in 12.04 - were they resolved? Regards, Daniel

7 11

Re: [PATCH] regedit: fixed a fclose(stdin) issue (Coverity)
by Frédéric Delanoy Nov. 19, 2012

Nov. 19, 2012

On Sat, Nov 17, 2012 at 9:28 PM, Marcus Meissner <meissner(a)suse.de> wrote: > Hi, > > If we have multiple filenames on the commandline, and the second one is -, > we would fclose(stdin) as realname is not NULL. > > (And also double free realname.) > > So set it back to NULL after free and in the stdin case. > > Ciao, Marcus > --- > programs/regedit/regedit.c | 1 + > 1 Datei geändert, 1 Zeile hinzugefügt(+) > > diff --git a/programs/regedit/regedit.c b/programs/regedit/regedit.c > index acff701..b6f74ee 100644 > --- a/programs/regedit/regedit.c > +++ b/programs/regedit/regedit.c > @@ -143,6 +143,7 @@ static BOOL PerformRegAction(REGEDIT_ACTION action, LPSTR s) > if (strcmp(filename, "-") == 0) > { > reg_file = stdin; > + realname = NULL; > } > else > { realname is set to NULL at the start of every iteration, so this is not useful. Frédéric

1 0

Re: cmd: Fix possible NULL pointer access in heap_strdupW
by Frédéric Delanoy Nov. 19, 2012

Nov. 19, 2012

On Sat, Nov 17, 2012 at 10:52 PM, André Hentschel <nerv(a)dawncrow.de> wrote: > we already do that (ieframe,vbscript,wininet,.. are good examples), but not consistent all over the place > --- > programs/cmd/wcmd.h | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/programs/cmd/wcmd.h b/programs/cmd/wcmd.h > index 3cc4c0b..303d260 100644 > --- a/programs/cmd/wcmd.h > +++ b/programs/cmd/wcmd.h > @@ -142,7 +142,8 @@ static inline WCHAR *heap_strdupW(const WCHAR *str) > > size = (strlenW(str)+1)*sizeof(WCHAR); > ret = heap_alloc(size); > - memcpy(ret, str, size); > + if(ret) > + memcpy(ret, str, size); > } > > return ret; > -- > 1.8.0 heap_alloc is cmd was made infallible, so this isn't necessary (c6d24089dbc284f0fef5e1ce37b75ac04c6ebd94) Frédéric

1 0

Re: [PATCH 1/2] usp10/tests: Added tests for wgBlank, wgDefault, wgInvalid and wgKashida (try 3).
by Alexandre Julliard Nov. 18, 2012

Nov. 18, 2012

Qian Hong <fracting(a)gmail.com> writes: > - Merge tests functions. > > --- > dlls/usp10/tests/usp10.c | 156 ++++++++++++++++++++++++++++++++++++++++++++++ > 1 file changed, 156 insertions(+) > The test fails: ../../../tools/runtest -q -P wine -M usp10.dll -T ../../.. -p usp10_test.exe.so usp10.c && touch usp10.ok usp10.c:1958: Test failed: wgBlank 0000 tmBreakChar 0020 usp10.c:1958: Test failed: wgBlank 0000 tmBreakChar 0020 usp10.c:1958: Test failed: wgBlank 0000 tmBreakChar 0020 usp10.c:1958: Test failed: wgBlank 0000 tmBreakChar 0020 etc. -- Alexandre Julliard julliard(a)winehq.org

2 3

certificate problem accessing https://source.winehq.org ?
by Dan Kegel Nov. 17, 2012

Nov. 17, 2012

$ git clone https://source.winehq.org/git/wine.git newwine-git Cloning into 'newwine-git'... error: server certificate verification failed. CAfile: /etc/ssl/certs/ca-certificates.crt CRLfile: none while accessing https://source.winehq.org/git/wine.git/info/refs Maybe that's related to the big fat red warning Chrome gives when visiting https://source.winehq.org/ Happily, just leaving off the 's' is a workaround.

2 2

Re: [PATCH 2/7] advapi: Implement SetNamedSecurityInfoW on top of SetSecurityInfo (resend 3).
by Alexandre Julliard Nov. 16, 2012

Nov. 16, 2012

"Erich E. Hoover" <ehoover(a)mymail.mines.edu> writes: > + hfile = CreateFileW( pObjectName, access, FILE_SHARE_READ|FILE_SHARE_WRITE|FILE_SHARE_DELETE, > + NULL, OPEN_EXISTING, FILE_FLAG_BACKUP_SEMANTICS, 0 ); > + if (hfile == INVALID_HANDLE_VALUE) > + return FALSE; FALSE doesn't do what you want here. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH v3] dsound: create a primary_pwfx separately from pwfx
by Andrew Eikum Nov. 16, 2012

Nov. 16, 2012

This breaks the Giants: Citizen Kabuto demo on coreaudio: http://www.fileplanet.com/55371/download/Giants:-Citizen-Kabuto-Demo I've attached a +dsound,+coreaudio,+tid log (could certainly be a winecoreaudio bug, but I don't have time to analyze right now). Also, > diff --git a/dlls/dsound/primary.c b/dlls/dsound/primary.c > index 8da12e2..ff4598c 100644 > --- a/dlls/dsound/primary.c > +++ b/dlls/dsound/primary.c > @@ -85,11 +174,20 @@ HRESULT DSOUND_ReopenDevice(DirectSoundDevice *device, BOOL forcewave) > + if (FAILED(hres))) { Fails to compile. Andrew

1 0

Re: [2/2] msxml3: Stub support for external entity resolver
by Marvin Nov. 16, 2012

Nov. 16, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=22981 Your paranoid android. === WXPPROSP3 (32 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000 === WVISTAADM (32 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000 === W2K8SE (32 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000 === W7PRO (32 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000 === W7PROX64 (32 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000 === TEST64_W7SP1 (32 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000 === W7PROX64 (64 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000 === TEST64_W7SP1 (64 bit saxreader) === saxreader.c:2230: Test failed: got 0x00000000

1 0

Re: msi: Fix a couple of leaks (coverity)
by Hans Leidekker Nov. 16, 2012

Nov. 16, 2012

On Fri, 2012-11-16 at 11:50 +0100, Frédéric Delanoy wrote: > diff --git a/dlls/msi/files.c b/dlls/msi/files.c > index 11913ef..98d7513 100644 > --- a/dlls/msi/files.c > +++ b/dlls/msi/files.c > @@ -336,6 +336,7 @@ UINT ACTION_InstallFiles(MSIPACKAGE *package) > if (rc != ERROR_SUCCESS) > { > ERR("Unable to load media info for %s (%u)\n", debugstr_w(file->File), rc); > + msi_free(mi); > return ERROR_FUNCTION_FAILED; This is a loop, so you should instead call msi_free_media_info to also free any strings allocated in a previous call to msi_load_media_info.

1 0

Re: [PATCH 2/4] kernel32: Partial implement CreateHardLinkTransactedA/W
by Alexandre Julliard Nov. 16, 2012

Nov. 16, 2012

Detlef Riekenberg <wine.dev(a)web.de> writes: > transaction is ignored for now I don't think that's a good idea, particularly not for DeleteFile. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH] amstream: Initialize correctly AM_MEDIA_TYPE struct.
by Michael Stefaniuc Nov. 16, 2012

Nov. 16, 2012

Hello Christian, On 11/15/2012 09:43 AM, Christian Costa wrote: > Fixes bug 32185. > --- > dlls/amstream/mediastreamfilter.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/dlls/amstream/mediastreamfilter.c b/dlls/amstream/mediastreamfilter.c > index 9b2bd8e..d5388e5 100644 > --- a/dlls/amstream/mediastreamfilter.c > +++ b/dlls/amstream/mediastreamfilter.c > @@ -132,6 +132,8 @@ static HRESULT WINAPI BasePinImp_GetMediaType(BasePin *This, int index, AM_MEDIA > MSPID purpose_id; > int i; > > + ZeroMemory(amt, sizeof(*amt)); > + this looks "odd". There is no check if amt is NULL, at least that's what caught my attention. So I've looked around and stumbled upon the AM_MEDIA_TYPE structure documentation http://msdn.microsoft.com/en-us/library/windows/desktop/dd373477%28v=vs.85%… "pUnk Not used. Set to NULL." Setting that to NULL in amstream's BasePinImp_GetMediaType() fixes the crash for me. But who's responsibility is to set / check that? That function or the caller aka test_media_streams() ? bye michael

2 1

Re: [PATCH 1/2] usp10/tests: Added tests for wgBlank, wgDefault, wgInvalid and wgKashida (try 2).
by Alexandre Julliard Nov. 15, 2012

Nov. 15, 2012

Qian Hong <fracting(a)gmail.com> writes: > @@ -3038,6 +3203,7 @@ START_TEST(usp10) > test_ScriptPlace(hdc); > > test_ScriptGetFontProperties(hdc); > + test_ScriptGetFontProperties2(hdc); There's no need to have two separate tests functions. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: advapi32: fix a too small buffer in CredUnmarshalCredentialW [try2]
by Hans Leidekker Nov. 15, 2012

Nov. 15, 2012

On Wed, 2012-11-14 at 16:28 +0100, Stefan Leichter wrote: > @@ -2053,6 +2053,8 @@ static BOOL cred_decode( const WCHAR *cred, unsigned int len, char *buf ) > char c0, c1, c2, c3; > const WCHAR *p = cred; > > + TRACE("%s\n", debugstr_wn(cred,len)); This string is already traced in CredUnmarshalCredentialW. > @@ -2134,6 +2136,7 @@ BOOL WINAPI CredUnmarshalCredentialW( LPCWSTR cred, PCRED_MARSHAL_TYPE type, PVO > case UsernameTargetCredential: > { > USERNAME_TARGET_CREDENTIAL_INFO *target; > + ULONGLONG size = 0; > > if (len < 9 || !cred_decode( cred + 3, 6, (char *)&size ) || !size || size % sizeof(WCHAR)) > { You should also perform a sanity check on 'size' to avoid overflow in calculations that follow.

2 3

Re: wined3d: Use backup swapchain DC for devices created with desktop window. (try 2)
by Stefan Dösinger Nov. 14, 2012

Nov. 14, 2012

Am Montag, 3. September 2012, 04:06:01 schrieb Adam Jakubek: > Previous version failed on test bots. > Added test case skipping when no device can be created with normal > (non-desktop) window. I skimmed bug #18490, and I'm wondering about two things: Does the app create a D3DDEVTYPE_REF or D3DDEVTYPE_NULLREF device? Once the device is created, what does it to with it? A +d3d log would provide some answers. This is some guesswork: I don't think the game creates a _REF device because REF devices require the user to install the reference rasterizer dll, as Matteo already mentioned. If it's a NULLREF device I wonder what the game is able to do with it. The bug report has a link which promises some clues, but the link is dead. Stefan

3 2

Re: fix a too small buffer in CredUnmarshalCredentialW
by Hans Leidekker Nov. 14, 2012

Nov. 14, 2012

On Wed, 2012-11-14 at 09:48 +0100, Stefan Leichter wrote: > + char buffer[6]; > + unsigned int buflen, *size = (unsigned int*) buffer; > > - if (len < 9 || !cred_decode( cred + 3, 6, (char *)&size ) || !size || size % sizeof(WCHAR)) > + if (len < 9 || !cred_decode( cred + 3, 6, buffer ) || !*size || *size % sizeof(WCHAR)) You're still truncating the decoded size. You should instead make 'size' a ULONGLONG and initialize it to 0. Is this little-endian ARM?

2 1

Implementation of PNG 'Get Count' and indexed reads...
by Max TenEyck Woodbury Nov. 14, 2012

Nov. 14, 2012

These are currently returning 'not implemented'. Would it make sense to have 'GetCount' always return '1' and have the indexed read simply read in the whole thing?

2 6

Re: [PATCH] netapi32/tests: Fix workstation test failures on XP
by Saulius Krasuckas Nov. 14, 2012

Nov. 14, 2012

* On Tue, 13 Nov 2012, Detlef Riekenberg wrote: > The workstation service might not running > > http://test.winehq.org/data/b5d96da32f955357b7d9341ee9fca68108078dce/xp_s2-… Indeed it isn't. I've disabled serveral services on this box. > --- a/dlls/netapi32/tests/wksta.c > +++ b/dlls/netapi32/tests/wksta.c > @@ -79,12 +79,17 @@ static void run_wkstausergetinfo_tests(void) > LPWKSTA_USER_INFO_1 ui1 = NULL; > LPWKSTA_USER_INFO_1101 ui1101 = NULL; > DWORD dwSize; > + NET_API_STATUS rc; > > /* Level 0 */ > - ok(pNetWkstaUserGetInfo(NULL, 0, (LPBYTE *)&ui0) == NERR_Success, > - "NetWkstaUserGetInfo is unsuccessful\n"); > + rc = pNetWkstaUserGetInfo(NULL, 0, (LPBYTE *)&ui0); > + if (rc == NERR_WkstaNotStarted) > + { > + skip("Workstation service not running\n"); > + return; > + } > + ok(!rc && ui0, "got %d and %p (expected NERR_Success and != NULL\n", rc, ui0); No to blame you, this is just generic inquiry for developers. Won't it be better to check service status directly in such case? S.

1 0

Re: [PATCH 1/2] user32: Add tests for MapWindowPoints, ClientToScreen and ScreenToClient. (try 3)
by Marvin Nov. 13, 2012

Nov. 13, 2012

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 http://testbot.winehq.org/JobDetails.pl?Key=22927 Your paranoid android. === WXPPROSP3 (32 bit win) === win.c:881: Test failed: wrong dwStyle: 84cb0000 != 94cb0000 for 000E0138 in hook HCBT_ACTIVATE

2 1

Re: [PATCH 2/2] usp10: Fixed wgBlank, wgDefault, wgInvalid and wgKashida.
by Nikolay Sivov Nov. 13, 2012

Nov. 13, 2012

On 11/13/2012 18:20, Qian Hong wrote: > Thanks Aric! > > --- > dlls/usp10/usp10.c | 133 +++++++++++++++++++++++++++++++++++++++++-- > dlls/usp10/usp10_internal.h | 1 + > 2 files changed, 128 insertions(+), 6 deletions(-) I think it's better to avoid macros if possible. > + if (!sc->sfnt) > + { > + if (GetTextMetricsW(hdc, &tmW)) > + sc->sfp.wgBlank = tmW.tmBreakChar; > + else > + { > + ERR("Get wgBlank failed, fallback to Numeric_space.\n"); > + sc->sfp.wgBlank = Numeric_space; > + } If it's a normal fallback case it's not an error I guess. > +{ > +#define Zero_width_space 0x200b /* Zero Width Space */ > +#define Invalid_char3 0xf71b /* Unknow, found by black box testing */ > +#define NON_EXIST_INDEX 0xffff /* Default non exist char index */ > + WCHAR invalid_chars[3] = {Numeric_space, Zero_width_space, Invalid_char3}; I don't see why this can't be hardcoded without macros and one line comment about meaning of this. > +#define Kashida_char 0x0640 /* Kashida */ > + WCHAR kashida_chars[] = {Kashida_char}; Same here, also naming is a bit strange. I think it's enough to name it 'kashida' and make it 'static const WCHAR'.

2 1

Re: [PATCH 2/2] usp10: Fixed wgBlank, wgDefault, wgInvalid and wgKashida.
by Alexandre Julliard Nov. 13, 2012

Nov. 13, 2012

Qian Hong <fracting(a)gmail.com> writes: > +static inline void set_cache_invalid_char(const HDC hdc, ScriptCache *sc) > +{ > +#define Zero_width_space 0x200b /* Zero Width Space */ > +#define Invalid_char3 0xf71b /* Unknow, found by black box testing */ > +#define NON_EXIST_INDEX 0xffff /* Default non exist char index */ > + WCHAR invalid_chars[3] = {Numeric_space, Zero_width_space, Invalid_char3}; There's no reason to add #defines for that kind of thing, just put the values in the string directly. -- Alexandre Julliard julliard(a)winehq.org

2 1