March 2008 - Wine-devel - list.winehq.org

Re: [19/22] WineD3D: Enable V8U8 for legacy bump mapping with ati extensions
by Stefan Dösinger 18 Mar '08

18 Mar '08

Am Montag, 17. März 2008 19:46:45 schrieb Stefan Dösinger: > Here is an updated patch that takes Roderick's recent changes into account Yet another patch, it takes Rodericks patches from today into account. Feel free to apply whatever patch suits better. If none of them are applied I will resend all the patches when Roderick is done with his CheckDeviceFormat rewite

1 0

appdb request: "Browse Regressions"
by Dan Kegel 18 Mar '08

18 Mar '08

The nice reports in recent wwn's showing changes in appdb ratings make me want to use the appdb itself to see which apps have recently fallen in rating. Following past appdb practice, one would implement that by adding a new "Browse Regressions" menu item on the left, below Browse Apps, Browse Newest Apps, Downloadable Apps, and Browse Apps by Rating. I expect this will become increasingly useful as we approach 1.0. How hard would it be? (Also, it might be nice to have an "advanced browse" that let you filter and sort by date, rating, and rating change. Once that works well, we might not need all those other specific browse commands...) - Dan

5 6

Re: services.exe/advapi32[7/7]: move GetServiceDisplayName to services.exe and implement GetServiceKeyName
by Robert Shearman 17 Mar '08

17 Mar '08

Mikołaj Zalewski wrote: > + /* Windows function 0x14 must be using a different prototype - not compatible */ > + DWORD svcctl_GetServiceDisplayNameW( > + SvcCtlRpcHandle rpc_handle, > + [in] POLICY_HANDLE *hSCManager, > + [in] LPCWSTR lpServiceName, > + [in,out,size_is(cchBufSize)] WCHAR lpBuffer[], > + [in] DWORD cchBufSize, > + [out] DWORD *cchLength); > + > + /* Windows function 0x15 must be using a different prototype - not compatible */ > + DWORD svcctl_GetServiceKeyNameW( > + SvcCtlRpcHandle rpc_handle, > + [in] POLICY_HANDLE *hSCManager, > + [in] LPCWSTR lpServiceDisplayName, > + [in,out,size_is(cchBufSize)] WCHAR lpBuffer[], > + [in] DWORD cchBufSize, > + [out] DWORD *cchLength); > } > The lpBuffer parameter should be out-only instead of in and out. I believe this will make the prototype compatible. -- Rob Shearman

1 0

Re: services.exe/advapi32[5/7]: Move QueryServiceConfig to services.exe
by Robert Shearman 17 Mar '08

17 Mar '08

In the next patch you've found a problem: Mikołaj Zalewski wrote: > +#if 0 /* for some reason (rpcrt4 bug?) QueryServiceConfig for a non-NULL lpLoadOrder crashes Wine */ > The issue is to do with this code: > LPQUERY_SERVICE_CONFIGW lpServiceConfig, > DWORD cbBufSize, LPDWORD pcbBytesNeeded) > { > - WCHAR str_buffer[ MAX_PATH ]; > - LONG r; > - DWORD type, val, sz, total, n; > - LPBYTE p; > - HKEY hKey; > + QUERY_SERVICE_CONFIGW config; > struct sc_service *hsvc; > + DWORD total; > + DWORD err; > + BYTE *bufpos; > > TRACE("%p %p %d %p\n", hService, lpServiceConfig, > cbBufSize, pcbBytesNeeded); > @@ -1886,58 +1908,21 @@ QueryServiceConfigW( SC_HANDLE hService, > SetLastError( ERROR_INVALID_HANDLE ); > return FALSE; > } > - hKey = hsvc->hkey; > - > - /* TODO: Check which members are mandatory and what the registry types > - * should be. This should of course also be tested when a service is > - * created. > - */ > - > - /* calculate the size required first */ > - total = sizeof (QUERY_SERVICE_CONFIGW); > > - sz = sizeof(str_buffer); > - r = RegQueryValueExW( hKey, szImagePath, 0, &type, (LPBYTE) str_buffer, &sz ); > - if( ( r == ERROR_SUCCESS ) && ( type == REG_SZ || type == REG_EXPAND_SZ ) ) > + if ((err = svcctl_QueryServiceConfigW(hsvc->hdr.rpc_handle, &hsvc->hdr.server_handle, &config)) != 0) > The problem is that QUERY_SERVICE_CONFIGW contains pointers and the DCE/RPC programming model ensures that non-NULL pointers that are being unmarshalled into are used (presumably to reduce memory allocations). So the issue here is that you're not initialising config before passing it into svcctl_QueryServiceConfigW and it is blowing up just by chance on the pointer occupying the lpLoadOrderGroup; it could just as well have been lpBinaryName, lpServiceStartName or lpDisplayName. > + > + /* Windows function 0x11 must be using a different prototype - not compatible */ > + /* Robert Shearman thinks there should be a byte_count attribute but (as of Sep 2007) > + * this isn't supported by widl nor by rpcrt4 */ > + DWORD svcctl_QueryServiceConfigW( > + SvcCtlRpcHandle rpc_handle, > + [in] POLICY_HANDLE *handle, > + [out] QUERY_SERVICE_CONFIGW *config); > + > } > The byte_count attribute is officially deprecated by Microsoft and it's not part of the DCE/RPC standard, plus you've already done the work and it's wire compatible (I think) without the attribute, so you can remove the comment about it. -- Rob Shearman

1 0

Re: services.exe/advapi32[3/7]: start a local RPC server
by Robert Shearman 17 Mar '08

17 Mar '08

Mikołaj Zalewski wrote: > @@ -231,6 +247,101 @@ static inline LPWSTR SERV_dupmulti(LPCSTR str) > } > > /****************************************************************************** > + * RPC connection with servies.exe > + */ > + > +static BOOL check_services_exe() > +{ > Put "void" in the brackets here and in many other places. > +static SvcCtlRpcHandle connect_to_server(LPCWSTR server) > +{ > + WCHAR transport[] = SVCCTL_TRANSPORT; > + WCHAR endpoint[] = SVCCTL_ENDPOINT; > + LPWSTR server_copy = NULL; > + RPC_WSTR binding_str; > + RPC_STATUS status; > + SvcCtlRpcHandle rpc_handle; > + > + /* unlike Windows we start services.exe on demand. We start it always as > + * checking if this is our address can be tricky */ > + if (!check_services_exe()) > + return NULL; > + > + if (server) /* parameters of RpcStringBindingComposeW are not const */ > + { > + server_copy = HeapAlloc(GetProcessHeap(), 0, sizeof(WCHAR)*(strlenW(server) + 1)); > + strcpyW(server_copy, server); > + } > RpcStringBindingComposeW doesn't change server if you pass it in, so there is no need to make a copy. > + status = RpcStringBindingComposeW(NULL, transport, server_copy, endpoint, NULL, &binding_str); > + HeapFree(GetProcessHeap(), 0, server_copy); > + if (status != RPC_S_OK) > + { > + ERR("RpcStringBindingComposeW failed (%d)\n", (DWORD)status); > + return NULL; > + } > > diff --git a/include/wine/svcctl.idl b/include/wine/svcctl.idl > new file mode 100644 > index 0000000..fefd12f > --- /dev/null > +++ b/include/wine/svcctl.idl > @@ -0,0 +1,68 @@ > +/* > + * svcctl interface definitions - exported by services.exe to access the > + * services database > + * > + * Copyright 2007 Google (Mikolaj Zalewski) > + * > + * This library is free software; you can redistribute it and/or > + * modify it under the terms of the GNU Lesser General Public > + * License as published by the Free Software Foundation; either > + * version 2.1 of the License, or (at your option) any later version. > + * > + * This library is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU > + * Lesser General Public License for more details. > + * > + * You should have received a copy of the GNU Lesser General Public > + * License along with this library; if not, write to the Free Software > + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA > + */ > + > +import "wtypes.idl"; > + > +/* > + * some defined for the C code > + */ > +cpp_quote("#define SVCCTL_TRANSPORT {'n','c','a','c','n','_','n','p',0}"); > +cpp_quote("#define SVCCTL_ENDPOINT {'\\\\','p','i','p','e','\\\\','s','v','c','c','t','l',0}"); > + > +/* Not the Windows event name - if needed the true one can be found in Inside Windows */ > +cpp_quote("#define SVCCTL_STARTED_EVENT (const WCHAR[]){'_','_','w','i','n','e','_','S','v','c','c','t','l','S','t','a','r','t','e','d',0}"); > + > + > +/* Based on the Samba IDL. Some functions are compatible with Windows but some > + * aren't and many are missing (thus function numbers are different) so we > + * don't use the Windows uuid which is 367abb81-9844-35f1-ad32-98f038001003 . > + */ > +[ uuid(78e2d9e8-d2a8-40d8-9bbe-7328cc5a9c32), > + version(2.0), > + pointer_default(unique), > + endpoint("ncacn_np:[\\pipe\\svcctl]"), > + helpstring("Service Control") > +] interface svcctl > +{ > + typedef handle_t SvcCtlRpcHandle; > + > + typedef struct _POLICY_HANDLE > + { > + DWORD dwHandleType; > + GUID uuid; > + } POLICY_HANDLE; > There's no need to use Samba's concept of "POLICY_HANDLE". We fully support the DCE/RPC context_handle type which is what this actually is. You can find out more about how to use context handles here: http://msdn2.microsoft.com/en-us/library/aa373605(VS.85).aspx > + > + /* Compatible with Windows function 0x00 */ > + DWORD svcctl_CloseServiceHandle( > + [in] SvcCtlRpcHandle rpc_handle, > + [in,out] POLICY_HANDLE *handle > + ); > There's no need for the rpc_handle here once you make the second parameter into a context handle. > + > + /* Compatible with Windows function 0x0f */ > + DWORD svcctl_OpenSCManagerW( > + SvcCtlRpcHandle rpc_handle, > + [in,unique] LPCWSTR MachineName, > + [in,unique] LPCWSTR DatabaseName, > + [in] DWORD dwAccessMask, > + [out] POLICY_HANDLE *handle > + ); > + > +} > -- Rob Shearman

1 0

re: improve winetricks dotnet 1.1
by Dan Kegel 17 Mar '08

17 Mar '08

Louis wrote: > Hi, i wonder if it would be possible to improve > winetricks dotnet 1.1, so that it's actually able to run >.net-1.1 applications. Currently it only installs the framework, > but due to bugs the apps don't run. Atm i use pretty worthless > script below to get them running.. I'm kind of hoping James implements the missing bits in msi soon. James, think that will happen soon enough that there's no need to update winetricks? > Note, the dlls gdiplus and mscoree should be set to native only, not (native, builtin), so override_dll should look tiny bit different then in winetricks. Ah, good point. - Dan

4 7

UrlCanonicalize patch I submitted
by John Morris 17 Mar '08

17 Mar '08

Hi, Last week I submitted a patch to correct the behavior of the UrlCanonicalize{W,A} functions. I found that "Stylus Studio 2007" had problems opening files that had a space in the path name. Tracing through the code, I found that the UrlCanonicalizeW function was not removing the "%20" escaped space when the URL_UNESCAPE bit was set in the flags parameter. I wrote a small program that set the flags like stylus studio was doing and ran it on both wine and windows, and found that wine definitely didn't handle it the same as windows. I made a small change to dlls/shlwapi/url.c, as well as added a test case to dlls/shlwapi/tests/url.c. With that change I got the same result with my program on both windows and wine. This is only my second patch to wine, so please feel free to give me any feedback, and let me know if there is anything that would be deemed incorrect about the way I did this. Many Thanks, John Morris

1 0

Re: msi: Skip the endlines test if the table fails on import
by Paul Vriens 17 Mar '08

17 Mar '08

James Hawkins wrote: > Hi, > > Fixes 33 test failures on win9x. > > Changelog: > * Skip the endlines test if the table fails on import. > > dlls/msi/tests/db.c | 7 ++++++- > 1 files changed, 6 insertions(+), 1 deletions(-) > > > > ------------------------------------------------------------------------ > > But shouldn't the ok-call stay and be moved after the if? Or is there not much extra information when add_table_to_db fails? -- Cheers, Paul.

2 1

Running services in WINE?
by Christopher 17 Mar '08

17 Mar '08

I've been trying to get MozyHome working on WINE, and have made a little progress. However, it seems that the service which Mozy uses won't start. Searching a bit I found a thread from a year ago stating that services don't work in WINE, but that it was being worked on. Are services implemented in WINE now, or if they're not is someone working on it who I could help out? Christopher Berner

6 12

services.exe: series of patches - empty attachments.
by Kirill K. Smirnov 17 Mar '08

17 Mar '08

Hi, Empty patches are attached to every e-mail in this series. Size: 0 bytes http://www.winehq.org/pipermail/wine-patches/2008-March/051638.html http://www.winehq.org/pipermail/wine-patches/2008-March/051651.html http://www.winehq.org/pipermail/wine-patches/2008-March/051641.html ... -- Kirill

2 1