November 2011 - Wine-devel - list.winehq.org

Re: msvcrt: Don't hold FILE critical section in some functions
by Alexandre Julliard 15 Nov '11

15 Nov '11

Piotr Caban <piotr(a)codeweavers.com> writes: > @@ -3117,25 +3098,20 @@ int CDECL MSVCRT_fputc(int c, MSVCRT_FILE* file) > */ > int CDECL MSVCRT__flsbuf(int c, MSVCRT_FILE* file) > { > - MSVCRT__lock_file(file); > - > /* Flush output buffer */ > if(file->_bufsiz == 0 && !(file->_flag & MSVCRT__IONBF)) { > msvcrt_alloc_buffer(file); > } > if(!(file->_flag & MSVCRT__IOWRT)) { > - if(file->_flag & MSVCRT__IORW) { > + if(file->_flag & MSVCRT__IORW) > file->_flag |= MSVCRT__IOWRT; > - } else { > - MSVCRT__unlock_file(file); > + else That doesn't look safe. -- Alexandre Julliard julliard(a)winehq.org

2 2

Re: wininet: Give a name to the critical sections, make them static and delete them when unloading the dll.
by Marvin 15 Nov '11

15 Nov '11

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=15430 Your paranoid android. === WINEBUILD (build) === Make failed

1 0

Re: [4/4] ddraw: Size handling fixes for IDirect3DDevice::GetDesc
by Henri Verbeet 15 Nov '11

15 Nov '11

On 15 November 2011 00:13, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > + if (!HWDesc) > + { > + WARN("HWDesc is NULL, returning DDERR_INVALIDPARAMS.\n"); > + return DDERR_INVALIDPARAMS; > + } > + else if(!check_d3ddevicedesc_size(HWDesc->dwSize)) > + { > + WARN("HWDesc->dwSize is %u, returning DDERR_INVALIDPARAMS.\n", HWDesc->dwSize); > + return DDERR_INVALIDPARAMS; > + } > + else if (!HelDesc) > + { > + WARN("HelDesc is NULL, returning DDERR_INVALIDPARAMS.\n"); > + return DDERR_INVALIDPARAMS; > + } > + else if(!check_d3ddevicedesc_size(HelDesc->dwSize)) > + { > + WARN("HelDesc->dwSize is %u, returning DDERR_INVALIDPARAMS.\n", HelDesc->dwSize); > + return DDERR_INVALIDPARAMS; > + } The "else" is redundant here. Please use consistent formatting. > + hw_caps.dwSize = sizeof(hel_caps) + 1; Strictly speaking that should be "sizeof(hw_caps) + 1". > + case 172: /* DirectX 3, IDirect3DDevice1 */ > + ok(hw_caps.dwMaxTextureRepeat == 0xfefefefe, "dwMaxTextureRepeat was modified: %#x.\n", > + hw_caps.dwMaxTextureRepeat); > + ok(hel_caps.dwMaxTextureRepeat == 0xfefefefe, "dwMaxTextureRepeat was modified: %#x.\n", > + hel_caps.dwMaxTextureRepeat); > + /* drop through */ > + case 204: /* DirectX 5, IDirect3DDevice2 */ > + ok(hw_caps.dwMaxTextureRepeat == 0xfefefefe, "dwMaxTextureRepeat was modified: %#x.\n", > + hw_caps.dwMaxTextureRepeat); > + ok(hel_caps.dwMaxTextureRepeat == 0xfefefefe, "dwMaxTextureRepeat was modified: %#x.\n", > + hel_caps.dwMaxTextureRepeat); This doesn't make a lot of sense, both cases are the same.

1 0

Re: wined3d: Disable WINED3DFMT_B8G8R8_UNORM.
by Stefan Dösinger 15 Nov '11

15 Nov '11

On Tuesday 15 November 2011 14:21:09 Matteo Bruni wrote: > + case WINED3DFMT_B8G8R8_UNORM: /* Usable in <=dx7 but not > enumerated in any case */ > + TRACE_(d3d_caps)("[FAILED] - Not enumerated on Windows\n"); > + return FALSE; The comment "Usable in <=dx7" is misleading. If I remember correctly your attempts at writing a test showed that systemmem surfaces can be created, but not video memory surfaces. I wouldn't be surprised if d3d8/9 also allow systemmem or scratch B8G8R8 surfaces. You can also remove the special handling for D3DFMT_R8G8B8 from d3d8/9's CheckDeviceFormat handling implementation now that wined3d doesn't report it as supported.

1 0

Re: [PATCH 1/2] ntdll: windows refuses to load an image, if sections overlap the header size as given in the SizeOfHeaders field (resend)
by Alexandre Julliard 15 Nov '11

15 Nov '11

Bernhard Loos <bernhardloos(a)googlemail.com> writes: > --- > dlls/kernel32/tests/loader.c | 79 ++++++++++++++++++++++++++++++++++++++++++ > dlls/ntdll/virtual.c | 5 +++ > 2 files changed, 84 insertions(+), 0 deletions(-) > > diff --git a/dlls/kernel32/tests/loader.c b/dlls/kernel32/tests/loader.c > index 4d420ea..bcef492 100644 > --- a/dlls/kernel32/tests/loader.c > +++ b/dlls/kernel32/tests/loader.c > @@ -692,9 +692,88 @@ static void test_section_access(void) > } > } > > +static void test_section_mapping(void) > +{ > + static const char filler[0x2000]; > + static const char section_data[] = "section data 1234567890"; > + DWORD total, written; > + HANDLE hfile; > + HMODULE hlib; > + SYSTEM_INFO si; > + char temp_path[MAX_PATH]; > + char dll_name[MAX_PATH]; > + BOOL ret; > + > + GetSystemInfo(&si); > + > + /* prevent displaying of the "Unable to load this DLL" message box */ > + SetErrorMode(SEM_FAILCRITICALERRORS); > + > + GetTempPath(MAX_PATH, temp_path); > + > + GetTempFileName(temp_path, "ldr", 0, dll_name); Please merge that test case into the existing loader tests instead of adding a new function to do the same thing. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [PATCH 11/12] mshtml: Added more navigation tests
by Marvin 15 Nov '11

15 Nov '11

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=15423 Your paranoid android. === TEST64_W7SP1 (32 bit htmldoc) === htmldoc.c:2883: Test failed: unexpected cmd 109 of CGID_Explorer htmldoc.c:2692: Test failed: unexpected command 6058 htmldoc.c:4725: Test failed: expected Invoke_AMBIENT_PALETTE htmldoc.c:4726: Test failed: expected GetOptionKeyPath htmldoc.c:2702: Test failed: unexpected call Exec_ShellDocView_37 htmldoc.c:2883: Test failed: unexpected cmd 109 of CGID_Explorer htmldoc.c:2789: Test failed: unexpected command 143 htmldoc.c:1704: Test failed: unexpected call SetStatusText === TEST64_W7SP1 (64 bit htmldoc) === htmldoc.c:2883: Test failed: unexpected cmd 109 of CGID_Explorer htmldoc.c:2692: Test failed: unexpected command 6058 htmldoc.c:4725: Test failed: expected Invoke_AMBIENT_PALETTE htmldoc.c:4726: Test failed: expected GetOptionKeyPath htmldoc.c:2702: Test failed: unexpected call Exec_ShellDocView_37 htmldoc.c:2883: Test failed: unexpected cmd 109 of CGID_Explorer htmldoc.c:2789: Test failed: unexpected command 143 htmldoc.c:1704: Test failed: unexpected call SetStatusText

2 1

Re: [PATCH 1/2] userenv: Improved CreateEnvironmentBlock implementation (try3)
by Marvin 15 Nov '11

15 Nov '11

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=15426 Your paranoid android. === WVISTAADM (32 bit userenv) === Failure running script in VM: The specified guest user must be logged in interactively to perform this operation

1 0

Re: [PATCH 1/2] userenv: Improved CreateEnvironmentBlock implementation (try2)
by Alexandre Julliard 15 Nov '11

15 Nov '11

Piotr Caban <piotr(a)codeweavers.com> writes: > +static void set_wow64_environment(WCHAR **env) > +{ > + static const WCHAR archW[] = {'P','R','O','C','E','S','S','O','R','_','A','R','C','H','I','T','E','C','T','U','R','E',0}; > + static const WCHAR arch6432W[] = {'P','R','O','C','E','S','S','O','R','_','A','R','C','H','I','T','E','W','6','4','3','2',0}; This stuff needs to be done depending on the 32/64 architecture of the child process, so it probably shouldn't be done here. The variables will be reset on every process startup anyway. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [PATCH] urlmon: Improve parsing of SCHEME_MK URIs
by Jacek Caban 15 Nov '11

15 Nov '11

Hi Andrew, On 11/14/11 21:12, Andrew Eikum wrote: > --- > dlls/urlmon/tests/uri.c | 81 +++++++++++++++++++++++++++++++++++++++++++++++ > dlls/urlmon/uri.c | 15 +++++++-- > 2 files changed, 93 insertions(+), 3 deletions(-) The patch looks good to me, but could you please add a simpler test using CreateUri (using uri_tests array) as follow up. Thanks, Jacek

1 0

Re: ws2_32: Fix handling of NULL and empty hostname in getaddrinfo/GetAddrInfoW.
by Marvin 15 Nov '11

15 Nov '11

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=15418 Your paranoid android. === WNT4WSSP6 (32 bit sock) === Timeout === W2KPROSP4 (32 bit sock) === Timeout

1 0