December 2015 - Wine-Devel - list.winehq.org

Re: Out-of-date gcc and ld on the testbot build VM

by Dmitry Timoshkov

Hugh McMaster <hugh.mcmaster(a)outlook.com> wrote: > >> The testbot seems to have an out-of-date version of mingw32, or at least, its linker, ld.exe. > >> > >> I'm trying to run tests for GetConsoleScreenBufferInfoEx, a function available on Vista and higher. > > > > Then using GetProcAddress() is the only solution. > > Yes. Until the build software can be updated. Since the tests must be executed under versions of Windows earlier than Vista it's still needed to fallback to GetProcAddress(). -- Dmitry.

Dec. 28, 2015

Re: Out-of-date gcc and ld on the testbot build VM

by Nikolay Sivov

On 28.12.2015 13:04, Hugh McMaster wrote: > On Monday, 28 Dec 2015 15:20:23 +0800, Dmitry Timoshkov wrote: >> Hugh McMaster wrote: >>> The testbot seems to have an out-of-date version of mingw32, or at least, its linker, ld.exe. >>> >>> I'm trying to run tests for GetConsoleScreenBufferInfoEx, a function available on Vista and higher. >> >> Then using GetProcAddress() is the only solution. > > Yes. Until the build software can be updated. > If it's Vista+ you need GetProcAddress() regardless.

Dec. 28, 2015

RE: Out-of-date gcc and ld on the testbot build VM

by Hugh McMaster

On Monday, 28 Dec 2015 15:20:23 +0800, Dmitry Timoshkov wrote: > Hugh McMaster wrote: >> The testbot seems to have an out-of-date version of mingw32, or at least, its linker, ld.exe. >> >> I'm trying to run tests for GetConsoleScreenBufferInfoEx, a function available on Vista and higher. > > Then using GetProcAddress() is the only solution. Yes. Until the build software can be updated.

Dec. 28, 2015

Re: shell32: add SHCreateSessionKey stub (try 2)

by Nikolay Sivov

> +/************************************************************************* > + * SHCreateSessionKey [SHELL32.723] NT 4.0 > + * > + */ > +HRESULT WINAPI SHCreateSessionKey(REGSAM access, HKEY *hkey) > +{ > + FIXME("stub: %u %p\n", access, hkey); > + return ERROR_CALL_NOT_IMPLEMENTED; > +} 'access' is a 32bit bit mask, return value is HRESULT.

Dec. 28, 2015

Re: shell32: add SHCreateSessionKey stub

by Nikolay Sivov

On 28.12.2015 8:15, Austin English wrote: > +HRESULT WINAPI SHCreateSessionKey(REGSAM access, HKEY *hkey) > +{ > + FIXME("stub: %d %p\n", access, hkey); > + return ERROR_SUCCESS; > +} This is wrong regarding tracing format, return value and formatting. I'll try to find time this week to send an actual implementation. It's fine to have fixed stub in before that of course.

Dec. 28, 2015

Re: Out-of-date gcc and ld on the testbot build VM

by Dmitry Timoshkov

Hugh McMaster <hugh.mcmaster(a)outlook.com> wrote: > The testbot seems to have an out-of-date version of mingw32, or at least, its linker, ld.exe. > > I'm trying to run tests for GetConsoleScreenBufferInfoEx, a function available on Vista and higher. Then using GetProcAddress() is the only solution. -- Dmitry.

Dec. 28, 2015

Out-of-date gcc and ld on the testbot build VM

by Hugh McMaster

The testbot seems to have an out-of-date version of mingw32, or at least, its linker, ld.exe. I'm trying to run tests for GetConsoleScreenBufferInfoEx, a function available on Vista and higher. But the build VM's linker is failing with an undefined reference: reg.o: In function `test_GetConsoleScreenBufferInfoEx': /home/winetest/tools/testbot/var/build-mingw32/programs/reg/tests/../../../../wine/programs/reg/tests/reg.c:489: undefined reference to `__imp__GetConsoleScreenBufferInfoEx(a)8' collect2: ld returned 1 exit status winegcc: i686-w64-mingw32-gcc failed My machine is able to compile the tests. It also has no trouble building a separate .exe using i686-w64-mingw32-gcc. My machine has: - mingw32 gcc version 4.8.2 (GCC) - mingw32 ld version: GNU ld (GNU Binutils) 2.23.52.20130620 -- Hugh McMaster

Dec. 28, 2015

Re: [RFC] Removing 200XXXXX versions from Bugzilla

by Rosanne DiMesio

On Sat, 26 Dec 2015 21:34:06 -0600 Austin English <austinenglish(a)gmail.com> wrote: > Howdy all, > > I'd like to remove the old 200XXXXX from Bugzilla, and replace them > with a pre-0.9 version instead. My reasoning is that it's obnoxious to > have to find the latest release in the middle of the list (or type it > in to find it), but if we removed all the 200XXXXX versions, it would > be the second to last (next to -unknown). > > For that matter, I'd be okay with merging everything pre-1.0, but I'm > mostly concerned about the 200XXXXX versions. > > Comments would be appreciated. > Yes, please. People have been asking for this for years. -- Rosanne DiMesio dimesio(a)earthlink.net

Dec. 27, 2015

Re: [PATCH] d3drm/tests: Add tests for IDirect3DRM*::LoadTexture (try 4).

by Aaryaman Vasishta

On Sun, Dec 27, 2015 at 12:40 AM, Stefan Dösinger <stefandoesinger(a)gmail.com > wrote: > You could add a few more checks for bpp and bytes_per_line in > test_bitmap_data(). Most of those would be inside the format-specific paths. > I've sent try 5 with some more tests and also added odd/even width tests for palettized textures, and format-specific depth/bpp tests along with the rest of the changes. Cheers, Aaryaman

Dec. 27, 2015

Re: ntdll: add RtlCreateUserProcess stub

by Austin English

On Sun, Dec 27, 2015 at 2:12 AM, Dmitry Timoshkov <dmitry(a)baikal.ru> wrote: > Austin English <austinenglish(a)gmail.com> wrote: > >> +NTSTATUS WINAPI RtlCreateUserProcess(UNICODE_STRING *path, ULONG attributes, RTL_USER_PROCESS_PARAMETERS *parameters, >> + PSECURITY_DESCRIPTOR *process_descriptor, PSECURITY_DESCRIPTOR *thread_descriptor, >> + HANDLE parent, BOOLEAN inherit, HANDLE debug, HANDLE exception, >> + PRTL_USER_PROCESS_INFORMATION info) > > Some of the parameters use wrong pointer information. Probably it would be > better to completely avoid Pxxx prefixes and use normal pointers for all > parameters? > > -- > Dmitry. Thanks for reviewing Dmitry. I'll revise and resend in the next day or two. -- -Austin

Dec. 27, 2015