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

Re: configure: Detect MinGW-w64 Clang cross toolchain

by Alexandre Julliard

André Hentschel <nerv(a)dawncrow.de> writes: > @@ -171,17 +171,19 @@ LIBS="$ac_wine_check_funcs_save_LIBS"]) > > dnl **** Check for a mingw program, trying the various mingw prefixes **** > dnl > -dnl Usage: WINE_CHECK_MINGW_PROG(variable,prog,[value-if-not-found],[path]) > +dnl Usage: WINE_CHECK_MINGW_PROG(variable,progs,[value-if-not-found],[path]) > dnl > AC_DEFUN([WINE_CHECK_MINGW_PROG], > [case "$host_cpu" in > i[[3456789]]86*) > ac_prefix_list="m4_foreach([ac_wine_prefix],[w64-mingw32, pc-mingw32, mingw32msvc, mingw32], > - m4_foreach([ac_wine_cpu],[i686,i586,i486,i386],[ac_wine_cpu-ac_wine_prefix-$2 ])) > + m4_foreach([ac_wine_cpu],[i686,i586,i486,i386], > + m4_foreach([ac_wine_prog],[$2],[ac_wine_cpu-ac_wine_prefix-ac_wine_prog ]))) > mingw32-$2" ;; Do we need to duplicate all the names? Or would it be sufficient to check only the w64-mingw one? -- Alexandre Julliard julliard(a)winehq.org

Dec. 30, 2015

Re: [PATCH] ole32: Properly implement OleRegGetUserType()

by Nikolay Sivov

On 30.12.2015 11:57, Dmitry Timoshkov wrote: >> dlls/ole32/ole2.c | 85 +++++++++++++++++++++++++++++-------------------------- >> 1 file changed, 45 insertions(+), 40 deletions(-) > > It would be helpful to know why this implementaion is considered more > "proper" than an existing one, especially taking into account lack of > any tests or reference to a bug report. > String 'form' argument was ignored before, and it's wrong. When there's no additional forms registered it falls back to form 1 as it always did.

Dec. 30, 2015

Re: [PATCH] ole32: Properly implement OleRegGetUserType()

by Dmitry Timoshkov

> dlls/ole32/ole2.c | 85 +++++++++++++++++++++++++++++-------------------------- > 1 file changed, 45 insertions(+), 40 deletions(-) It would be helpful to know why this implementaion is considered more "proper" than an existing one, especially taking into account lack of any tests or reference to a bug report. -- Dmitry.

Dec. 30, 2015

Re: [PATCH 1/4] ntdll: Use correct parameter types instead of casting.

by Sebastian Lackner

On 30.12.2015 06:37, Alex Henrie wrote: > Casting has not been necessary here ever since this code was converted > to use wineserver requests circa Wine 1.0. > > Signed-off-by: Alex Henrie <alexhenrie24(a)gmail.com> > --- > dlls/ntdll/file.c | 13 ++++--------- > 1 file changed, 4 insertions(+), 9 deletions(-) > > diff --git a/dlls/ntdll/file.c b/dlls/ntdll/file.c > index 1027b54..6524f32 100644 > --- a/dlls/ntdll/file.c > +++ b/dlls/ntdll/file.c > @@ -411,10 +411,8 @@ static struct async_fileio *alloc_fileio( DWORD size, HANDLE handle, PIO_APC_ROU > } > > /* callback for irp async I/O completion */ > -static NTSTATUS irp_completion( void *user, IO_STATUS_BLOCK *io, NTSTATUS status, void **apc, void **arg ) > +static NTSTATUS irp_completion( struct async_irp *async, IO_STATUS_BLOCK *io, NTSTATUS status, void **apc, void **arg ) > { It would no longer match the prototype in dlls/ntdll/server.c then: NTSTATUS (*func)(void *, IO_STATUS_BLOCK *, NTSTATUS, void **, void **) = wine_server_get_ptr( call->async_io.func );

Dec. 30, 2015

Re: 0001-winedbg-Add-a-debug-button-on-crash-dialog-get-rid-of-

by Jactry Zeng

Hi Changhui, I like this patch, too. '&Debug' and 'Show &Details' use a same shortcut key here. @@ -58,6 +50,7 @@ BEGIN IDC_STATIC_TXT2,"SysLink",0,27,60,224,100 DEFPUSHBUTTON "Close", IDOK, 205, 151, 60, 16, WS_TABSTOP PUSHBUTTON "Show &Details", ID_DETAILS, 140, 151, 60, 16, WS_TABSTOP + PUSHBUTTON "&Debug", ID_DEBUG, 75, 151, 60, 16, WS_TABSTOP END I think push this debug button in the left corner seems better? Maybe we also need some text in IDC_STATIC_TXT2 to let users know what it means. Thanks. 2015-12-28 17:21 GMT+08:00 Changhui LIU <liuchanghui(a)linuxdeepin.com>: > > > > > > > ------------------ > Regards, > Changhui Liu > > > > > -- Regards, Jactry Zeng

Dec. 30, 2015

Re: 0001-winedbg-Add-a-debug-button-on-crash-dialog-get-rid-of-

by Qian Hong

On Wed, Dec 30, 2015 at 2:06 AM, Alex Henrie <alexhenrie24(a)gmail.com> wrote: > I'm pretty sure that the Debug button on the crash dialog is hidden so > that people who don't know what they're doing don't try to use it. I have the same understand with you, but I think it is hidden too deep, I guess I'm not the only developer who don't know this feature until Changhui's patch. I don't think it is too harmful if someone "who don't know what they're doing" try to use it. Maybe some 12 years old hacker will play with the feature and become curious and then grow to a future Wine developer, why not? :p -- Regards, Qian Hong - http://www.winehq.org

Dec. 30, 2015

Re: 0001-winedbg-Add-a-debug-button-on-crash-dialog-get-rid-of-

by Alex Henrie

I'm pretty sure that the Debug button on the crash dialog is hidden so that people who don't know what they're doing don't try to use it. -Alex

Dec. 29, 2015

Re: [PATCH 1/1] msvcr120: Enable MSVCRT__nextafterf when _MSVCR_VER>=120.

by Piotr Caban

On 12/28/15 13:52, YongHao Hu wrote: > Hi, thank you for your comment and a new patch had sent. > I used tools/make_specfiles to update other spec, so I missed ucrtbase > and api-ms-win-crt-math-l1-1-0/api-ms-win-crt-math-l1-1-0. > Is that ucrtbase and msvcr120 a group of dll in tools/make_specfiles? I'm not sure what you mean. Ucrtbase is not updated automatically by make_specfiles script. The script is not very useful for dlls that use PARENTSRC anyway. Cheers, Piotr

Dec. 29, 2015

Re: [PATCH] ole32: Zero out pointer in CreateFileMoniker() on error

by Marvin

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=19236 Your paranoid android. === wvistau64 (32 bit moniker) === moniker: unhandled exception c0000005 at 0042A889 === w2008s64 (32 bit moniker) === moniker: unhandled exception c0000005 at 0042A889 === w7u (32 bit moniker) === moniker: unhandled exception c0000005 at 0042A889 === w7pro64 (32 bit moniker) === moniker: unhandled exception c0000005 at 0042A889 === w8 (32 bit moniker) === moniker: unhandled exception c0000005 at 0042A889 === w864 (32 bit moniker) === moniker: unhandled exception c0000005 at 0042A889 === w1064 (32 bit moniker) === moniker: unhandled exception c0000005 at 0042A889 === wvistau64 (64 bit moniker) === moniker: unhandled exception c0000005 at 00000000004241FA === w2008s64 (64 bit moniker) === moniker: unhandled exception c0000005 at 00000000004241FA === w7pro64 (64 bit moniker) === moniker: unhandled exception c0000005 at 00000000004241FA === w864 (64 bit moniker) === moniker: unhandled exception c0000005 at 00000000004241FA === w1064 (64 bit moniker) === moniker: unhandled exception c0000005 at 00000000004241FA

Dec. 29, 2015

Re: [2/2] shell32: ShellExecute() should not exec non-executable files.

by Francois Gouget

On Tue, 29 Dec 2015, Francois Gouget wrote: [...] > + > + if ((sei_tmp.lpVerb && *sei_tmp.lpVerb) || PathIsExeW(wszApplicationName)) Please don't apply. I think this test is wrong. It should probably be checking that no was specified, or maybe that no class was specified. -- Francois Gouget <fgouget(a)codeweavers.com>

Dec. 29, 2015