May 2007 - Wine-devel - list.winehq.org

Re: [8/8] DDraw: Make IDirectDrawImpl thread safe
by Alexandre Julliard 11 May '07

11 May '07

Stefan Dösinger <stefan(a)codeweavers.com> writes: > @@ -2731,9 +2792,13 @@ IDirectDrawImpl_EnumSurfaces(IDirectDraw7 *iface, > desc = surf->surface_desc; > IDirectDrawSurface7_AddRef(ICOM_INTERFACE(surf, IDirectDrawSurface7)); > if(Callback( ICOM_INTERFACE(surf, IDirectDrawSurface7), &desc, Context) != DDENUMRET_OK) > + { > + DDOBJ_UNLOCK(This); > return DD_OK; > + } It don't think you can call an arbitrary callback function with the lock held. -- Alexandre Julliard julliard(a)winehq.org

2 1

sprintfW formatting in lib\wine\string.c
by Ann & Jason Edmeades 11 May '07

11 May '07

Whilst trying to convert cmd to Unicode, I've stumbled across a problem which I either need to work around or fix, and I am not sure which. WCHAR h[] = {'h','e','l','l','o','\0'}; WCHAR fmt[] = {':', '%', '1', '0','s',':','\0'}; char test[255]; WCHAR result[255]; sprintf(test, "'%10s'", "hello"); sprintfW(result, fmt, h); WINE_FIXME("%s vs %s\n", test, wine_dbgstr_w(result)); Gives: fixme:cmd:WCMD_list_directory ' hello' vs L":hello:" ie sprintf of %10s means right justified field of 10 characters sprintfW of %10s gives left justified field of n characters (I haven't tried if the string is > 10, as I don't care...) Basically, the format strings in cmd rely on the padding behaviour. Is this a bug in sprintfW which needs fixing, or should I work around it? Unfortunately, looking in the code to lib\wine\string.c, it just copies the field and ignores justification or min/max width precisions for strings, and as it is used all over the place there's also a concern of what would be changed by fixing it. Any suggestions for a workable alternative? Jason

2 1

Re: Adds NET START to net.exe
by Alexandre Julliard 11 May '07

11 May '07

"Tim Schwartz" <tim(a)sanityinternet.com> writes: > + if(!strcasecmp(argv[1], "start")) > + { > + if(argc < 3) > + { > + printf("Specify service name to start.\n"); > + return TRUE; > + } > + > + if(!net_service(NET_START, argv[2])) > + { > + return TRUE; > + } > + return FALSE; > + } > + > + return FALSE; The previous version was better here, TRUE and FALSE are not appropriate return values for main(), it's supposed to return an error code not a boolean. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: crypt32: implement CryptSIPLoad (try 3)
by Paul Vriens 11 May '07

11 May '07

Juan Lang wrote: > - todo_wine > - { > - ok ( ret, "Expected CryptSIPLoad to succeed\n"); > - /* This error will always be there as native searches for the function DllCanUnloadNow > - * in WINTRUST.DLL (in this case). This function is not available in WINTRUST.DLL. > - * For now there's no need to implement this is Wine as I doubt any program will rely on > - * this last error when the call succeeded. > - */ > - ok ( GetLastError() == ERROR_PROC_NOT_FOUND, > - "Expected ERROR_PROC_NOT_FOUND, got 0x%08x\n", GetLastError()); > - ok( sdi.pfGet != (pCryptSIPGetSignedDataMsg)0xdeadbeef, "Expected a function pointer to be loaded.\n"); > - } > + ok ( ret, "Expected CryptSIPLoad to succeed\n"); > + ok( sdi.pfGet != (pCryptSIPGetSignedDataMsg)0xdeadbeef, "Expected a function pointer to be loaded.\n"); > > /* The function addresses returned by CryptSIPLoad are actually the addresses of > * crypt32's own functions. A function calling these addresses will end up first > @@ -348,18 +337,8 @@ static void test_SIPLoad(void) > sdi.cbSize = sizeof(SIP_DISPATCH_INFO); > sdi.pfGet = (pCryptSIPGetSignedDataMsg)0xdeadbeef; > ret = CryptSIPLoad(&unknown2, 0, &sdi); > - todo_wine > - { > - ok ( ret, "Expected CryptSIPLoad to succeed\n"); > - /* This call on it's own would have resulted in a ERROR_PROC_NOT_FOUND, but the previous > - * call to CryptSIPLoad already loaded wintrust.dll. As this information is cached, > - * CryptSIPLoad will not try to search for the already mentioned DllCanUnloadNow. > - */ > - } > - ok ( GetLastError() == 0xdeadbeef, > - "Expected 0xdeadbeef, got 0x%08x\n", GetLastError()); > - todo_wine > - ok( sdi.pfGet != (pCryptSIPGetSignedDataMsg)0xdeadbeef, "Expected a function pointer to be loaded.\n"); > + ok ( ret, "Expected CryptSIPLoad to succeed\n"); > + ok( sdi.pfGet != (pCryptSIPGetSignedDataMsg)0xdeadbeef, "Expected a function pointer to be loaded.\n"); > > /* All OK, but other SIP */ > SetLastError(0xdeadbeef); > > > ------------------------------------------------------------------------ > > Hi Juan, Why did you remove all the comments from the test? It does explain (a little bit) how native does get to this specific lasterror. I agree that the lasterror is not very valuable. I've spent quite some time figuring this stuff out and it seems that (at least to my perception) information gets lost. Cheers, Paul.

1 0

Re: [4/8] WineD3D: Destroying the resource frees all private data
by H. Verbeet 10 May '07

10 May '07

On 10/05/07, Stefan Dösinger <stefan(a)codeweavers.com> wrote: > - } > + }; Any particular reason for this change? > + if (Flags & WINED3DSPD_IUNKNOWN) { > + if(SizeOfData != 4) { > + WARN("IUnknown data with size %d, returning WINED3DERR_INVALIDCALL\n", SizeOfData); > + return WINED3DERR_INVALIDCALL; This looks rather questionable to me. Are you making an assumption about the size of a pointer here or is 4 some magic constant? > - list_add_tail(&This->resource.privateData, &data->entry); > - return WINED3D_OK; > - ... > + list_init(&data->entry); > + list_add_tail(&This->resource.privateData, &data->entry); > > return WINED3D_OK; Is the added list_init() intentional here? Looks redundant to me, if you're doing a list_add_tail() right afterwards anyway.

2 1

Re: [PATCH 2/2] msi: automation: Implement Installer::RegistryValue.
by Alexandre Julliard 10 May '07

10 May '07

Misha Koshelev <mk144210(a)bcm.edu> writes: > + case VT_BSTR: /* Return value of specified key if it exists */ > + if (ret == ERROR_SUCCESS) > + { > + ret = RegQueryValueExW(hkey, V_BSTR(&varg2), NULL, NULL, NULL, &dwSize); > + if (ret == ERROR_SUCCESS) > + { > + szString = msi_alloc(dwSize); > + if (szString) > + { > + LPWSTR szNewString = NULL; > + DWORD dwType, dwNewSize = 0; > + int idx; > + > + ret = RegQueryValueExW(hkey, V_BSTR(&varg2), NULL, &dwType, (LPBYTE)szString, &dwSize); > + switch (dwType) > + { > + /* Registry strings may not be null terminated */ > + case REG_MULTI_SZ: > + idx = (dwSize/sizeof(WCHAR))-1; > + while (idx >= 0 && !szString[idx]) idx--; > + for (; idx >= 0; idx--) > + if (!szString[idx]) szString[idx] = '\n'; > [...] This stuff clearly needs to be split out into a few helper functions. -- Alexandre Julliard julliard(a)winehq.org

2 1

[5/8] WineD3D: Implement DDBLT_DEPTHFILL on the active depth stencil target
by Stefan Dösinger 10 May '07

10 May '07

1 1

re: blizzard conference
by Dan Kegel 10 May '07

10 May '07

Hey, that's only an hour away, maybe I / Lei / Nigel can drop by. But what's the "itchy /etc/hosts bug"? Is that http://bugs.winehq.org/show_bug.cgi?id=7929 ? - Dan

8 12

Re: [5/5] widl: Remove var_t's ptr_level field and start write_pointers
by Dan Hipschman 10 May '07

10 May '07

On Thu, May 10, 2007 at 10:39:31AM -0700, Duane Clark wrote: > Dan Hipschman wrote: > >Well, I sent this quite a while ago and it hasn't shown up on the list yet. > >It's probably too big, so here's a link to the patch instead of inlining > >it. > > > > Patches over 80KB will go to a queue for moderation. I am the only one > moderating, so just be patient. I'll get around to it eventually. OK, I didn't know that and assumed it just got nuked. It appears to have just shown up on the list. :-) Thanks

1 0

Re: [3/5] widl: Simplify make_safearray
by Robert Shearman 10 May '07

10 May '07

Dan Hipschman wrote: > This patch just simplifies the make_safearray function. It didn't need > to be that complicated. I tested the generated headers and typelibs > with this patch applied (well, I do that with every patch in this series), > and this doesn't change anything. > Is dlls/oleaut32/tests/tmarshal.tlb identical compared to before this patch is applied? If not, you won't get test failures in Wine, but you will when running tests in Windows. -- Rob Shearman

2 1