August 2007 - Wine-Devel - list.winehq.org

Sent a patch twice

by Rene Kok

I never submitted a patch before. It wasn't a patch I created just resubmitting it. However I accidentally sent it to the wine-patches mailing list twice. The first time I wasn't subscribed, which I prefer not to because the huge amount of mail I'll get when I do :-) I subscribed and sent it again and now it's in there twice, oops! Anyway hope it's ok.

Aug. 4, 2007

Re: mshtml: Remove const in front of REFIID as that is a macro that already starts with const.

by Andrew Talbot

Andrew Talbot wrote: > So, in case anyone is still awake and interested: > > static const REFIID tid_ids[]; > > is equivalent to > > static const GUID *const const tid_ids[]; (Note the erroneous double > const.) > > or > static GUID const *const const tid_ids[]; > > as I would prefer to put it. So, to conclude, you don't want the extra > const, and it may not have been the const you thought it was, either. :) > Whoops! Hopefully, you had all fallen asleep before reading that bit, but because we are talking about a #define, not a typedef, the declaration would, in fact, be equivalent to: static const const GUID *const tid_ids[]; (I checked it with gcc -E.) But it's still a const too far. -- Andy.

Aug. 4, 2007

Re: mshtml: Remove const in front of REFIID as that is a macro that already starts with const.

by Andrew Talbot

Michael Stefaniuc wrote: > Jacek, > > I'm not sure about this patch. gcc 3.1.1 produced a warning about a > double const. The rules of what is const depending on where in the > declaration it is placed are mind bongling (especially at 1am) and I > don't know what the intention was here so please review. > > -static const REFIID tid_ids[] = { > +static REFIID tid_ids[] = { > &IID_IHTMLWindow2 > }; > It would be easier if one were obliged to always put const (and volatile) qualifiers to the right of what they qualify, then the whole declaration would read from right to left. Thus, here, char const * const *p; p is a pointer to a constant pointer to a constant char. The almost-universally adopted convention of putting the left-most qualifer on the left of the type it qualifies, as in const char *ch; rather than the equivalent char const *ch; rather spoils this flow. Here, static const REFIID tid_ids[]; could be rewritten static REFIID const tid_ids[]; And REFIID in a C setting is defined as: const GUID * __MIDL_CONST (where __MIDL_CONST is "const", if rpcproxy.h has not been included). So, in case anyone is still awake and interested: static const REFIID tid_ids[]; is equivalent to static const GUID *const const tid_ids[]; (Note the erroneous double const.) or static GUID const *const const tid_ids[]; as I would prefer to put it. So, to conclude, you don't want the extra const, and it may not have been the const you thought it was, either. :) -- Andy.

Aug. 4, 2007

Re: Bug introduced in 0.9.42 in printer dialog

by Vitaly Lipatov

В сообщении от 3 августа 2007 Christian Pradelli написал(a): > A bug was introduced in version 0.9.42, that makes printer dialogs to not > work. > When you press Accept button in the printer dialog the I receive the > followin error: > > fixme:commdlg:PRINTDLG_UpdatePrintDlgA No lpdm ptr? > fixme:commdlg:PRINTDLG_WMCommandA Update printdlg was not successful! As I know this bug introduced in winspool, not printer dialog. > > Should I post to bugzilla? Sure. -- Vitaly Lipatov, ALT Linux Team Russia, Saint-Petersburg, www.etersoft.ru

Aug. 4, 2007

Re: [PATCH] wined3d: Pass correct coord parameter to texture*DProj functions

by H. Verbeet

On 04/08/07, Vitaly Budovski <vbudovski(a)gmail.com> wrote: > --- > dlls/wined3d/glsl_shader.c | 9 +++++++++ > 1 files changed, 9 insertions(+), 0 deletions(-) > This patch is plain wrong. The caller of shader_glsl_get_sample_function() is responsible for correctly adding the correct component for projected lookups, since that isn't always the "next" component, but in fact depends on the instruction that does the sampling. Of course if the author of this patch had actually read the source for the callers of shader_glsl_get_sample_function() (pshader_glsl_tex() in particular...) or alternatively had been familiar with D3D shader assembly, he would have known that. It does make me wonder though, particularly given the lack of any explanation for the patch, if this is actually supposed to fix anything?

Aug. 4, 2007

Re: netapi32: Add lmuseflg.h and include it from lmuse.h.

by Kai Blin

On Saturday 04 August 2007 03:20:00 Francois Gouget wrote: > +/* > + * Copyright (C) the Wine project I think you need a year in the copyright statement. Cheers, Kai -- Kai Blin WorldForge developer http://www.worldforge.org/ Wine developer http://wiki.winehq.org/KaiBlin Samba team member http://www.samba.org/samba/team/ -- Will code for cotton.

Aug. 4, 2007

Re: [PATCH] winecfg: Fix crash caused by calling set_reg_key with NULL value

by Nigel Liang

On 8/3/07, Evan Stade <estade(a)gmail.com> wrote: > On 8/3/07, Nigel Liang <ncliang(a)gmail.com> wrote: > > diff --git a/programs/winecfg/winecfg.c b/programs/winecfg/winecfg.c > > index 84dd4c8..05f322c 100644 > > --- a/programs/winecfg/winecfg.c > > +++ b/programs/winecfg/winecfg.c > > @@ -468,11 +468,15 @@ void set_reg_key(HKEY root, const char * > > > > wpath = HeapAlloc(GetProcessHeap(), 0, (strlen(path)+1)*sizeof(WCHAR)); > > wname = HeapAlloc(GetProcessHeap(), 0, (strlen(name)+1)*sizeof(WCHAR)); > > - wvalue = HeapAlloc(GetProcessHeap(), 0, (strlen(value)+1)*sizeof(WCHAR)); > > > > MultiByteToWideChar(CP_ACP, 0, path, -1, wpath, strlen(path)+1); > > MultiByteToWideChar(CP_ACP, 0, name, -1, wname, strlen(name)+1); > > - MultiByteToWideChar(CP_ACP, 0, value, -1, wvalue, strlen(value)+1); > > + > > + if (value) > > + { > > + wvalue = HeapAlloc(GetProcessHeap(), 0, (strlen(value)+1)*sizeof(WCHAR)); > > + MultiByteToWideChar(CP_ACP, 0, value, -1, wvalue, strlen(value)+1); > > + } > > > > set_reg_key_ex(root, wpath, wname, wvalue, REG_SZ); > > > > -- > > 1.4.1 > > > > > > > > > > You need to initialize wvalue to NULL > > -- > Evan Stade > Why is that? The code is not dependent on wvalue being NULL, but it checks "value" to catch NULL pointers passed in from the caller. I think you may have missed that... ;-) -Nigel

Aug. 4, 2007

Re: [PATCH] winecfg: Fix crash caused by calling set_reg_key with NULL value

by Evan Stade

On 8/3/07, Nigel Liang <ncliang(a)gmail.com> wrote: > diff --git a/programs/winecfg/winecfg.c b/programs/winecfg/winecfg.c > index 84dd4c8..05f322c 100644 > --- a/programs/winecfg/winecfg.c > +++ b/programs/winecfg/winecfg.c > @@ -468,11 +468,15 @@ void set_reg_key(HKEY root, const char * > > wpath = HeapAlloc(GetProcessHeap(), 0, (strlen(path)+1)*sizeof(WCHAR)); > wname = HeapAlloc(GetProcessHeap(), 0, (strlen(name)+1)*sizeof(WCHAR)); > - wvalue = HeapAlloc(GetProcessHeap(), 0, (strlen(value)+1)*sizeof(WCHAR)); > > MultiByteToWideChar(CP_ACP, 0, path, -1, wpath, strlen(path)+1); > MultiByteToWideChar(CP_ACP, 0, name, -1, wname, strlen(name)+1); > - MultiByteToWideChar(CP_ACP, 0, value, -1, wvalue, strlen(value)+1); > + > + if (value) > + { > + wvalue = HeapAlloc(GetProcessHeap(), 0, (strlen(value)+1)*sizeof(WCHAR)); > + MultiByteToWideChar(CP_ACP, 0, value, -1, wvalue, strlen(value)+1); > + } > > set_reg_key_ex(root, wpath, wname, wvalue, REG_SZ); > > -- > 1.4.1 > > > > You need to initialize wvalue to NULL -- Evan Stade

Aug. 4, 2007

Re: user32 - set_active_window uses SendMessage instead ofPostMessage for WM_ACTIVATEAPP messages.

by Dmitry Timoshkov

"Peter Dons Tychsen" <donpedro(a)tdcadsl.dk> wrote: > 1) Yes i did testing on Windows-XP. I did it by putting together various > examples and by checking with InSendMessage() for all cases. This > clearly showed that WM_ACTIVATEAPP was always posted and never sent. Looks like you misunderstood the InSendMessage behaviour and that led to some confusion. Wine test harness for message sequences works just fine, there is no need to blame it. -- Dmitry.

Aug. 3, 2007

Re: Nigel Liang : winecfg: Read/write registry in unicode.

by Nigel Liang

On 8/3/07, James Hawkins <truiken(a)gmail.com> wrote: > On 8/3/07, Alexandre Julliard <julliard(a)wine.codeweavers.com> wrote: > > Module: wine > > Branch: master > > Commit: 2d5a89bab087d09f3e3876e29ae32a2e863273fe > > URL: http://source.winehq.org/git/wine.git/?a=commit;h=2d5a89bab087d09f3e3876e29… > > > > Author: Nigel Liang <ncliang(a)gmail.com> > > Date: Thu Aug 2 16:33:24 2007 -0700 > > > > winecfg: Read/write registry in unicode. > > > > This patch breaks setting the windows version in winecfg (and I assume > everything else that uses the registry): > Ah~ This is because set_reg_key is called with NULL value. Caught this for get_reg_key, but missed set_reg_key. Patch sent to fix this: http://www.winehq.org/pipermail/wine-patches/2007-August/042325.html Thanks for the catch James. -Nigel

Aug. 3, 2007