Uwe Bonnes wrote:
> Hallo,
>
> some application ( AvrStudio.exe version 3 from www.atmel.com) uses
> many Commctrl features. Present CVS delivers somehow btnPtr->iString
> == -1 and that makes wine crash either in
> TRACE ("lpText: \"%s\"\n", debugstr_w(lpText));
> or later in some of the
> DrawTextW (hdc, lpText, -1, ...);
> calls.
>
>
> Question 1: Should debugstr_xx cope with invalid arguments
>
> Question 2: What should be done with btnPtr->iString == -1?
>
> This argument seems to have some special meaning in Comctl.
>
> Is something like appended patch enough?
>
> Bye
>
> Uwe Bonnes [EMAIL PROTECTED]
>
> Free Software: If you contribute nothing, expect nothing
> --
>
>
> Index: wine/dlls/comctl32/toolbar.c
> ===================================================================
> RCS file: /home/wine/wine/dlls/comctl32/toolbar.c,v
> retrieving revision 1.47
> diff -u -r1.47 toolbar.c
> --- wine/dlls/comctl32/toolbar.c 2000/06/04 01:46:27 1.47
> +++ wine/dlls/comctl32/toolbar.c 2000/06/04 18:01:05
> @@ -82,7 +82,7 @@
> TRACE ("iString: %lx\n", btnPtr->iString);
>
> /* get a pointer to the text */
> - if (HIWORD(btnPtr->iString) != 0)
> + if ((HIWORD(btnPtr->iString) != 0) && (btnPtr->iString!=-1))
> lpText = (LPWSTR)btnPtr->iString;
> else if ((btnPtr->iString >= 0) && (btnPtr->iString <
infoPtr->nNumStrings))
> lpText = infoPtr->strings[btnPtr->iString];
>
Hi,
I have not seen this yet. Perhaps this is another undocumented toolbar
feature? A string callback feature seems reasonable. Are strings missing in
the toolbar of your test app?
If your test app doesn't crash with your patch it is ok. A little comment in
the source code about this strange (undocumented) behaviour would be nice.
Regards,
Eric Kohl
