October 2017 - Wine-Devel - list.winehq.org

Re: [PATCH] shlwapi: Support UNC paths in PathGetDriveNumberW.
by Marvin Oct. 9, 2017

Oct. 9, 2017

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=33459 Your paranoid android. === wxppro (32 bit path) === path.c:1458: Test failed: got = -1 === w2003std (32 bit path) === path.c:1458: Test failed: got = -1

1 0

Re: [PATCH] d3dx9/tests: Avoid bitwise operation.
by Fabian Maurer Oct. 9, 2017

Oct. 9, 2017

> if (!end1 && !end2) break; > > - status = !end1 ^ !end2; > + status = !end1 && !end2; > ok(!status, "Line %u, test %u: Mismatch in size, test declaration > is %s than expected.\n", line, test_id, end1 ? "shorter" : "longer"); > if (status) Doesn't this change the meaning of the code? You could also just make it into "status = FALSE", since when it's true the conditional break is used. Regards, Fabian Maurer

2 1

Re: [PATCH] quartz: Add missing parameter.
by Matteo Bruni Oct. 9, 2017

Oct. 9, 2017

2017-10-07 15:55 GMT+02:00 Andrey Gusev <andrey.goosev(a)gmail.com>: > Signed-off-by: Andrey Gusev <andrey.goosev(a)gmail.com> > --- > dlls/quartz/pin.c | 5 ++++- > dlls/quartz/pin.h | 5 ++++- > 2 files changed, 8 insertions(+), 2 deletions(-) > > diff --git a/dlls/quartz/pin.c b/dlls/quartz/pin.c > index a2651fd49b..355d87d9d5 100644 > --- a/dlls/quartz/pin.c > +++ b/dlls/quartz/pin.c > @@ -218,7 +218,10 @@ static HRESULT PullPin_Init(const IPinVtbl *PullPin_Vtbl, const PIN_INFO * pPinI > return S_OK; > } > > -HRESULT PullPin_Construct(const IPinVtbl *PullPin_Vtbl, const PIN_INFO * pPinInfo, SAMPLEPROC_PULL pSampleProc, LPVOID pUserData, QUERYACCEPTPROC pQueryAccept, CLEANUPPROC pCleanUp, REQUESTPROC pCustomRequest, STOPPROCESSPROC pDone, LPCRITICAL_SECTION pCritSec, IPin ** ppPin) > +HRESULT PullPin_Construct(const IPinVtbl *PullPin_Vtbl, const PIN_INFO * pPinInfo, > + SAMPLEPROC_PULL pSampleProc, LPVOID pUserData, QUERYACCEPTPROC pQueryAccept, > + CLEANUPPROC pCleanUp, REQUESTPROC pCustomRequest, STOPPROCESSPROC pDone, > + LPCRITICAL_SECTION pCritSec, IPin ** ppPin) > { > PullPin * pPinImpl; > > diff --git a/dlls/quartz/pin.h b/dlls/quartz/pin.h > index 428fa6fe8e..2f29c929cf 100644 > --- a/dlls/quartz/pin.h > +++ b/dlls/quartz/pin.h > @@ -102,7 +102,10 @@ typedef struct PullPin > #define Req_Pause 3 > > /*** Constructors ***/ > -HRESULT PullPin_Construct(const IPinVtbl *PullPin_Vtbl, const PIN_INFO * pPinInfo, SAMPLEPROC_PULL pSampleProc, LPVOID pUserData, QUERYACCEPTPROC pQueryAccept, CLEANUPPROC pCleanUp, STOPPROCESSPROC, REQUESTPROC pCustomRequest, LPCRITICAL_SECTION pCritSec, IPin ** ppPin); > +HRESULT PullPin_Construct(const IPinVtbl *PullPin_Vtbl, const PIN_INFO * pPinInfo, > + SAMPLEPROC_PULL pSampleProc, LPVOID pUserData, QUERYACCEPTPROC pQueryAccept, > + CLEANUPPROC pCleanUp, REQUESTPROC pCustomRequest, STOPPROCESSPROC pDone, > + LPCRITICAL_SECTION pCritSec, IPin ** ppPin); > > /**************************/ > /*** Pin Implementation ***/ > -- > 2.13.6 Maybe I'm blind (I read this from the cell phone, so definitely possible) but I don't see that you're adding any parameter, rather swapping two arguments in the header only. Just making sure that's what you meant to do.

1 0

Re: [PATCH 1/2] ntdll/tests: Don't abort if NtCancelFileIoEx() is missing.
by Zebediah Figura Oct. 9, 2017

Oct. 9, 2017

On 10/09/2017 07:29 AM, Jacek Caban wrote: > Hi Zebediah, > > On 07.10.2017 00:18, Zebediah Figura wrote: >> - ok(U(iosb).Status == STATUS_CANCELLED, "Wrong iostatus %x\n", U(iosb).Status); >> - ok(WaitForSingleObject(hEvent, 0) == 0, "hEvent not signaled\n"); >> + ok(U(iosb).Status == STATUS_CANCELLED, "Wrong iostatus %x\n", U(iosb).Status); >> + ok(WaitForSingleObject(hEvent, 0) == 0, "hEvent not signaled\n"); >> + >> + CloseHandle(hPipe); >> + } >> + else >> + skip("NtCancelIoFileEx not available\n"); > > > Please use win_skip() instead. > > > Thanks, > > Jacek > > > Whoops, thanks. Resent.

1 0

Re: [PATCH] gdiplus: Added GdipSetImageAttributesNoOp()
by Vincent Povirk Oct. 9, 2017

Oct. 9, 2017

This could use some tests. My guess is that there's an additional set of flags for "no op" that overrides the others.

1 0

Re: [PATCH V3] msvcp100: add time_get::do_get and time_get::get functions
by Piotr Caban Oct. 6, 2017

Oct. 6, 2017

On 10/06/17 20:21, Vijay Kiran Kamuju wrote: > +#define call_time_get_char_do_get(this, ret, s, e, base, err, t, fmt, mod) CALL_VTBL_FUNC(this, 12, istreambuf_iterator_char*, \ > + (const time_get_char*, istreambuf_iterator_char*, istreambuf_iterator_char, istreambuf_iterator_char, ios_base*, int*, struct tm*, char, char), \ > + (this, ret, s, e, base, err, t, fmt, mod)) > +#else > +#define call_time_get_char_do_get(this, ret, s, e, base, err, t, fmt, mod) CALL_VTBL_FUNC(this, 20, istreambuf_iterator_char*, \ > + (const time_get_char*, istreambuf_iterator_char*, istreambuf_iterator_char, istreambuf_iterator_char, ios_base*, int*, struct tm*, char, char), \ > + (this, ret, s, e, base, err, t, fmt, mod)) You're calling wrong virtual functions in call_time_get_char_do_get macro. Second argument of CALL_VTBL_FUNC macro is function offset in virtual function table. Thanks, Piotr

1 0

Re: [v2] msvcrt: Introduce vf[w]printf_helper function.
by Piotr Caban Oct. 6, 2017

Oct. 6, 2017

Hi, On 10/02/17 21:46, Gijs Vermeulen wrote: > -int CDECL MSVCRT_vfprintf(MSVCRT_FILE* file, const char *format, __ms_va_list valist) > +static int vfprintf_vfwprintf_helper(MSVCRT_FILE* file, const char *format, const MSVCRT_wchar_t *format_wide, > + __ms_va_list valist, MSVCRT__locale_t locale, DWORD options, BOOL wide) I would prefer to have a separate helper for vfprintf and vfwprintf functions. Also please reorder the parameters so it's more similar to _stdio_common_vfprintf. > + if(!MSVCRT_CHECK_PMT( file != NULL )) return -1; > + if(!(wide && MSVCRT_CHECK_PMT( format != NULL ))) return -1; This doesn't work as expected. It should look like this: if(!wide && !MSVCRT_CHECK_PMT(... Thanks, Piotr

1 0

Re: msvcp100: add time_get::do_get and time_get::get functions
by Piotr Caban Oct. 6, 2017

Oct. 6, 2017

Hi, On 10/06/17 18:37, Vijay Kiran Kamuju wrote: > +#if _MSVCP_VER <= 100 > +#define call_time_get_char_do_get(this, ret, s, e, base, err, t, fmt, mod) CALL_VTBL_FUNC(this, 12, istreambuf_iterator_char*, \ > + (const time_get_char*, istreambuf_iterator_char*, istreambuf_iterator_char, istreambuf_iterator_char, ios_base*, int*, struct tm*, char, char), \ > + (this, ret, s, e, base, err, t)) > +#else > +#define call_time_get_char_do_get(this, ret, s, e, base, err, t, fmt, mod) CALL_VTBL_FUNC(this, 20, istreambuf_iterator_char*, \ > + (const time_get_char*, istreambuf_iterator_char*, istreambuf_iterator_char, istreambuf_iterator_char, ios_base*, int*, struct tm*, char, char), \ > + (this, ret, s, e, base, err, t)) > +#endif This will not compile (too few arguments are passed). > @@ -11613,6 +11644,9 @@ void __asm_dummy_vtables(void) { > VTABLE_ADD_FUNC(locale_facet__Decref) > #endif > VTABLE_ADD_FUNC(time_get_char_do_date_order) > +#if _MSVCP_VER >= 100 > + VTABLE_ADD_FUNC(time_get_char_do_get) > +#endif > VTABLE_ADD_FUNC(time_get_char_do_get_time) > VTABLE_ADD_FUNC(time_get_char_do_get_date) > VTABLE_ADD_FUNC(time_get_char_do_get_weekday) The order of virtual functions is incorrect. Thanks, Piotr

1 0

Re: ImmReleaseContext only return true
by Nikolay Sivov Oct. 6, 2017

Oct. 6, 2017

On 06.10.2017 18:55, r00t0 wrote: > BOOL WINAPI ImmReleaseContext(HWND hWnd, HIMC hIMC) > { > - static BOOL shown = FALSE; > - > - if (!shown) { > - FIXME("(%p, %p): stub\n", hWnd, hIMC); > - shown = TRUE; > - } > - return TRUE; > + TRACE("(%p, %p): stub\n", hWnd, hIMC); > + return TRUE; > } Hi, Tomasz. What's a point of this change? (the rest of the diff seems to be whitespace changes only)

1 0

Re: ImmReleaseContext only return true
by Fabian Maurer Oct. 6, 2017

Oct. 6, 2017

You seem to have a lot of unrelated white-space changes. > BOOL WINAPI ImmReleaseContext(HWND hWnd, HIMC hIMC) > { >- static BOOL shown = FALSE; >- >- if (!shown) { >- FIXME("(%p, %p): stub\n", hWnd, hIMC); >- shown = TRUE; >- } >- return TRUE; >+ TRACE("(%p, %p): stub\n", hWnd, hIMC); >+ return TRUE; > } What's the point behind this change? Regards, Fabian Maurer

1 0

Re: msxml3: Use correct string length in saxreader_get_cdata_chunk
by Nikolay Sivov Oct. 6, 2017

Oct. 6, 2017

On 05.10.2017 15:59, Piotr Caban wrote: > @@ -1972,6 +1972,7 @@ static BSTR saxreader_get_cdata_chunk(const xmlChar *str, int len) > BSTR bstr = bstr_from_xmlCharN(str, len), ret; > WCHAR *ptr; > > + len = SysStringLen(bstr); Hi. Just making sure I understand this correctly, the issue is that we're using length counted in xmlChars as WCHAR length?

2 2

Re: [PATCH] rsaenh: Fix length of some algorithm names in aProvEnumAlgsEx array.
by Michael Stefaniuc Oct. 5, 2017

Oct. 5, 2017

Hello Alex, On 10/05/2017 09:15 AM, Alex Henrie wrote: > From: Louis Lenders <xerox.xerox2000x(a)gmail.com> > > Patch from Wine Staging. > > Signed-off-by: Louis Lenders <xerox.xerox2000x(a)gmail.com> > Signed-off-by: Alex Henrie <alexhenrie24(a)gmail.com> > --- > dlls/rsaenh/rsaenh.c | 6 +++--- > 1 file changed, 3 insertions(+), 3 deletions(-) > > diff --git a/dlls/rsaenh/rsaenh.c b/dlls/rsaenh/rsaenh.c > index 3db828638b..97eb474374 100644 > --- a/dlls/rsaenh/rsaenh.c > +++ b/dlls/rsaenh/rsaenh.c > @@ -246,9 +246,9 @@ static const PROV_ENUMALGS_EX aProvEnumAlgsEx[5][RSAENH_MAX_ENUMALGS+1] = > - {CALG_SHA_256, 256,256, 256,CRYPT_FLAG_SIGNING, 6,"SHA-256", 30,"Secure Hash Algorithm (SHA-256)"}, > - {CALG_SHA_384, 384,384, 384,CRYPT_FLAG_SIGNING, 6,"SHA-384", 30,"Secure Hash Algorithm (SHA-384)"}, > - {CALG_SHA_512, 512,512, 512,CRYPT_FLAG_SIGNING, 6,"SHA-512", 30,"Secure Hash Algorithm (SHA-512)"}, > + {CALG_SHA_256, 256,256, 256,CRYPT_FLAG_SIGNING, 8,"SHA-256", 32,"Secure Hash Algorithm (SHA-256)"}, > + {CALG_SHA_384, 384,384, 384,CRYPT_FLAG_SIGNING, 8,"SHA-384", 32,"Secure Hash Algorithm (SHA-384)"}, > + {CALG_SHA_512, 512,512, 512,CRYPT_FLAG_SIGNING, 8,"SHA-512", 32,"Secure Hash Algorithm (SHA-512)"}, why don't you switch the whole big structure to automatically calculate the string sizes? Something like: #define X(str) sizeof(str), str ... {CALG_SHA_256, 256,256, 256,CRYPT_FLAG_SIGNING, X("SHA-256"), X("Secure Hash Algorithm (SHA-256)")}, ... #undef X The less the programmer can fat finger the merrier, especially in security related code. bye michael

1 0

Re: [PATCH] ntdll: Failure callback is optional in LdrResolveDelayLoadedAPI()
by Marvin Oct. 5, 2017

Oct. 5, 2017

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=33398 Your paranoid android. === w1064 (32 bit loader) === loader.c:3019: Test failed: Test 4: unexpected return value DEADBEEF. === w1064 (64 bit loader) === loader.c:3019: Test failed: Test 4: unexpected return value 00000000DEADBEEF.

1 0

Re: [3/3] gdiplus/tests: Add tests for gdi32 transform and clip caching.
by Marvin Oct. 5, 2017

Oct. 5, 2017

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=33389 Your paranoid android. === build (build) === Patch failed to apply

1 0

Re: [3/3] gdiplus/tests: Add tests for gdi32 transform and clip caching.
by Marvin Oct. 5, 2017

Oct. 5, 2017

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=33389 Your paranoid android. === wvistau64 (32 bit graphics) === Can't locate object method "new" via package "URI" at /home/testbot/lib/WineTestBot/VMs.pm line 203, <DATA> line 751. Child process died unexpectedly === w864 (32 bit graphics) === Can't locate object method "new" via package "URI" at /home/testbot/lib/WineTestBot/VMs.pm line 203, <DATA> line 751. Child process died unexpectedly

1 0

Re: [v2 PATCH] qedit: Support IAMTimelineGroup interface in IAMTimelineObj
by Nikolay Sivov Oct. 5, 2017

Oct. 5, 2017

On 05.10.2017 3:03, Alistair Leslie-Hughes wrote: > +static HRESULT WINAPI timelinegrp_QueryInterface(IAMTimelineGroup *iface, REFIID riid, void **object) > +{ > + TimelineObjImpl *This = impl_from_IAMTimelineGroup(iface); > + > + TRACE("(%p)->(%s %p)\n", This, debugstr_guid(riid), object); > + > + if (!object) > + return E_POINTER; There's no test for this condition. > + > + *object = NULL; > + if (IsEqualIID(riid, &IID_IUnknown) || > + IsEqualIID(riid, &IID_IAMTimelineGroup) ) > + *object = &This->IAMTimelineGroup_iface; > + else if (IsEqualIID(riid, &IID_IAMTimelineObj) ) > + *object = &This->IAMTimelineObj_iface; > + else > + WARN("(%p, %s,%p): not found\n", This, debugstr_guid(riid), object); > + > + if (!*object) > + return E_NOINTERFACE; > + > + IUnknown_AddRef((IUnknown*)*object); > + return S_OK; > +} That's not what I meant. Tests don't show that you need this. Please take a look how objects supporting multiple interfaces are implemented.

1 0

Re: [PATCH 4/4] user32: Don't invalidate ComboBox on CB_SETCURSEL message
by Marvin Oct. 4, 2017

Oct. 4, 2017

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=33383 Your paranoid android. === w7u (32 bit msg) === msg.c:12113: Test failed: WmMouseHoverSeq: 0: the msg sequence is not complete: expected 0118 - actual 0000 msg.c:12115: Test failed: wrong tme.dwFlags 00000001, expected 00000000 msg.c:12115: Test failed: wrong tme.hwndTrack 001C01BA, expected 00000000 msg.c:12115: Test failed: wrong tme.dwHoverTime 400, expected 0

1 0

Re: [PATCH] qedit: Support IAMTimelineGroup interface in IAMTimelineObj
by Nikolay Sivov Oct. 4, 2017

Oct. 4, 2017

On 10/04/2017 07:40 AM, Alistair Leslie-Hughes wrote: > --- a/dlls/qedit/tests/timeline.c > +++ b/dlls/qedit/tests/timeline.c > @@ -126,7 +126,7 @@ static void test_timelineobj_interfaces(void) > IAMTimelineGroup *group; > > hr = IAMTimelineObj_QueryInterface(obj, &IID_IAMTimelineGroup, (void **)&group); > - todo_wine ok(hr == S_OK, "got %08x\n", hr); > + ok(hr == S_OK, "got %08x\n", hr); > if(hr == S_OK) > IAMTimelineGroup_Release(group); You don't need if() after this patch. > +static HRESULT WINAPI timelinegrp_QueryInterface(IAMTimelineGroup *iface, REFIID riid, void **object) > +{ > + TimelineObjImpl *This = impl_from_IAMTimelineGroup(iface); > + > + TRACE("(%p)->(%s %p)\n", This, debugstr_guid(riid), object); > + > + if (!object) > + return E_POINTER; > + > + *object = NULL; > + if (IsEqualIID(riid, &IID_IUnknown) || > + IsEqualIID(riid, &IID_IAMTimelineGroup) ) > + *object = &This->IAMTimelineGroup_iface; > + else > + WARN("(%p, %s,%p): not found\n", This, debugstr_guid(riid), object); > + > + if (!*object) > + return E_NOINTERFACE; > + > + IUnknown_AddRef((IUnknown*)*object); > + return S_OK; > +} This looks wrong, but not impossible. Can you query IAMTimelineObj back from IAMTimelineGroup? Looks like we don't have tests showing that you can't do that.

1 0

Re: [Rediffed PATCH] ntdll: Add stub for RtlGetUnloadEventTrace
by Alex Henrie Oct. 4, 2017

Oct. 4, 2017

+#define RTL_UNLOAD_EVENT_TRACE_NUMBER 64 +RTL_UNLOAD_EVENT_TRACE RtlpUnloadEventTrace[RTL_UNLOAD_EVENT_TRACE_NUMBER]; Why did you add these? If you don't need them for the stub implementation, I recommend submitting a revised patch without them. -Alex

1 1

Wine 2.18 source code scanned with cppcheck 1.72
by Игорь Чебаненко Oct. 3, 2017

Oct. 3, 2017

2 1

Re: [v2 PATCH 1/2] advapi32: Handle WinLogonIdsSid in CreateWellKnownSid
by Alexandre Julliard Oct. 2, 2017

Oct. 2, 2017

Alistair Leslie-Hughes <leslie_alistair(a)hotmail.com> writes: > @@ -967,7 +967,7 @@ CreateWellKnownSid( WELL_KNOWN_SID_TYPE WellKnownSidType, > unsigned int i; > TRACE("(%d, %s, %p, %p)\n", WellKnownSidType, debugstr_sid(DomainSid), pSid, cbSid); > > - if (cbSid == NULL || (DomainSid && !IsValidSid(DomainSid))) > + if (cbSid == NULL || (DomainSid && !IsValidSid(DomainSid)) || WellKnownSidType == WinLogonIdsSid) It doesn't look right to special-case it like this. It probably needs more test cases to find out what's going on with such sids. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Update Japanese translation.
by Hugh McMaster Oct. 2, 2017

Oct. 2, 2017

Hi Akihiro, On Sunday, 1 October 2017 11:23 PM, Akihiro Sagawa wrote: > #: regedit.rc:213 > -#, fuzzy > -#| msgid "" > -#| "Are you sure you want to delete the registry key '%1' and all of its " > -#| "subkeys?" > msgid "" > "Are you sure you want to delete this registry key and all of its subkeys?" > -msgstr "本当にレジストリキー '%1' とそのサブキーを削除しますか?" > +msgstr "本当にこのレジストリキー '%1' とそのサブキーすべてを削除しますか?" You left the '%1' in the updated string. -- Hugh McMaster

2 1

Re: [PATCH 2/2] dwrite: Free the mappings array, but not every element.
by Huw Davies Oct. 2, 2017

Oct. 2, 2017

On 2 Oct 2017, at 10:26, Huw Davies <huw(a)codeweavers.com> wrote: > > Signed-off-by: Huw Davies <huw(a)codeweavers.com> > --- > dlls/dwrite/analyzer.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/dlls/dwrite/analyzer.c b/dlls/dwrite/analyzer.c > index 6808b468d6..183abed7f9 100644 > --- a/dlls/dwrite/analyzer.c > +++ b/dlls/dwrite/analyzer.c > @@ -2305,8 +2305,8 @@ static ULONG WINAPI fontfallbackbuilder_Release(IDWriteFontFallbackBuilder *ifac > IDWriteFontCollection_Release(mapping->collection); > heap_free(mapping->ranges); > heap_free(mapping->locale); > - heap_free(mapping); > } > + heap_free(fallbackbuilder->mappings); > > IDWriteFactory5_Release(fallbackbuilder->factory); > heap_free(fallbackbuilder); Ah, this is a dup of 137464. Huw.

1 0

Re: [PATCH 2/2] winhttp: Send basic authorization credentials with the first request if they've been set.
by Marvin Oct. 2, 2017

Oct. 2, 2017

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=33362 Your paranoid android. === w2008s64 (64 bit winhttp) === winhttp.c:1148: Test failed: request failed unexpectedly 400

2 1

Re: [PATCH] msvcrt: Don't crash if _vsnwprintf gets NULL as format string
by Marvin Oct. 1, 2017

Oct. 1, 2017

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=33350 Your paranoid android. === wxppro (32 bit printf) === 09e8:printf: unhandled exception c0000005 at 77C431F7 === w2003std (32 bit printf) === 05f0:printf: unhandled exception c0000005 at 77BD4190

2 1