- Wine-Devel - list.winehq.org

ntoskrnl and safedisk
by Ivan Leo Puoti April 11, 2005

April 11, 2005

I won't have much time to work on this stuff for a while, so I thought I might as well post it here. I've got a ntoskrnl.exe that can load, unload and initialise drivers successfully (Well, I've only tried with a few builds of safedisk but that doesn't matter), it can also dispatch io requests received from DispatchDeviceControl to the driver. However safedisk fails by returning STATUS_UNSUCCESSFUL in the IRP's IoStatus.Status member, I haven't been able to debug this because winedbg refuses to step trough the code properly, however I will try again when I have time to do so. In the mean time you can download and build this stuff yourself, the patch that teaches wine to interact with ntoskrnl is at http://spazioinwind.libero.it/nonsolomicrosoft/public/patch.diff , ntoskrnl itself is here http://spazioinwind.libero.it/nonsolomicrosoft/public/ntoskrnl.tar.bz you'll also need http://spazioinwind.libero.it/nonsolomicrosoft/public/ntoskrnl_protocol.h , put it in include/wine. A full ./configure && make depend && make is needed to build this stuff into wine after patching. Comments/ideas/fixes welcome. Note that ntoskrnl isn't thread safe ATM and probably can't handle more than one driver at any one time. Ivan.

1 0

Alexandre is back!
by Jakob Eriksson April 11, 2005

April 11, 2005

Send in more patches, everybody! :-)

2 1

Re: lostwages/templates/en status_ui.template
by Steven Edwards April 11, 2005

April 11, 2005

Hi, --- Mike McCormack <mike(a)codeweavers.com> wrote: > Actually, I'd be interested to see how much of Gecko we'd need to > import/port so that Wine could render HTML. It seems pretty obvious to > me from my experiments with the Mozilla Active X control that nobody is > using it as it is. The ReactOS guys are using the Wine code mostly as is. There is still some issue preventing Mozilla from drawing so Martin Fuchs has made a lightwieght Webbrowser that is a wrapper around the currently implementation that installs the Control. It would be nice to not have to download any additional stuff to have Gecko support but I don't think its worth the massive amount of trouble to import plus I doubt Julliard would ever let it in the Wine tree anyway. Thanks Steven __________________________________ Do you Yahoo!? Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/

1 0

Re: Re2: Pushing up data through the stream in quartz
by Robert Shearman April 11, 2005

April 11, 2005

Maarten Lankhorst wrote: > HRESULT Capture_Run(CaptureBox * capBox, FILTER_STATE *state) > { > HRESULT hr; > IMediaSample *pSample = NULL; > LPBYTE pointer; > hr = OutputPin_GetDeliveryBuffer((OutputPin *)capBox->pOut, > &pSample, NULL, NULL, 0); > TRACE("Meat 1: %lx - %p\n", hr, pSample); > hr = IMediaSample_GetPointer(pSample, &pointer); > TRACE("Meat 2\n"); > IMediaSample_SetActualDataLength(pSample, 320*240*3); > TRACE("Meat 3\n"); > ZeroMemory(pointer, 320*240*3); > TRACE("Meat 4\n"); > OutputPin_SendSample((OutputPin *)capBox->pOut, pSample); > FIXME("%p -> (%p) stub\n", capBox, state); > return S_OK; > } > > I get E_VFW_NOT_COMMITTED after trying GetDeliveryBuffer, what's the > cause? > This error is due to a memory allocator needing to be committed before it can be used (see dlls/quartz/memallocator.c). You should call IMemAllocator_Commit in the Run implementation of your filter and then push one sample downstream. IIRC, you should also call IMemAllocator_Decommit in the Stop implementation of your filter, but this isn't done for the implementation I referred to (dlls/quartz/transform.c). Rob

2 1

real SetProcessClass and SetThreadPriority support
by Robert Reif April 11, 2005

April 11, 2005

Are there any plans or is anyone working on mapping Windows SetProcessClass and SetThreadPriority support to linux process priorities on kernels that support CAP_SYS_NICE?

5 8

Re: Reduce __attribute__(x) to nothing if __GNUC__ is unset
by Alexandre Julliard April 11, 2005

April 11, 2005

Vincent Béron <vberon(a)mecano.gme.usherb.ca> writes: > This lets us get rid of various #ifdef __GNUC__, as well as use > DECLSPEC_NORETURN at least everywhere winnt.h is included, without fear > of conflicts. This doesn't really help, because we want the code to build with the Microsoft headers too, so you still can't use __attribute__ without ifdefs. And really the use of __attribute__ should be kept to a minimum, your other patches are adding it in a lot of places where it's not necessary. -- Alexandre Julliard julliard(a)winehq.org

1 0

stubbed powrprof.dll
by Benjamin Cutler April 11, 2005

April 11, 2005

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 This is my first attempt at contributing anything to the Wine tree, so if I've made any common 'newbie' mistakes, that's probably why. I've been experimenting with dx9wine to see if I can get Planetside working (that game is the only reason I'd even consider going back to windows), but the program requires one function from powrprof.dll, which, unless I'm blind, doesn't seem to exist in the wine tree at present. So I spent a few hours looking at the source tree and looking up info on the functions provided by powrprof.dll, and wrote up a header and a .c file that includes stubbed out functions so that applications that use powrprof will at least load (they of course might not run properly). Attached you'll find the result of that work, including a patch to configure.ac and dlls/Makefile.in, and a tarball containing dlls/powrprof/ and include/powrprof.h. Note that powrprof.h contains several struct and enum definitions that really belong in winnt.h, but I wasn't sure where in winnt.h I should place them, so I simply stuck them in powrprof.h for now. After compiling powrprof.dll.so and placing it in /usr/lib/dx9wine/lib/wine/, I was able to get the Planetside exe to load, but it promptly bombed out because it thought my graphics card was only 16 megs (it's 128, but I'm sure I just need to fix a config variable), so I think I'm on the right track. Thank you for your time. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.0 (GNU/Linux) Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org iD8DBQFCWJNy3ZWQ5WzMh1oRAupIAJ9tOd4krieCMN5xBm1ypbqfxk8m7gCgj3Rr 52ZhgBcks8jKFkuWF+mElT4= =EEgB -----END PGP SIGNATURE-----

3 7

Re: [Wine]Turining off Fixme Errors
by Brouard Nicolas April 11, 2005

April 11, 2005

Le dimanche 10 avril 2005 à 23:00 +0200, TimeFX a écrit : > WINEDEBUG=fixme-all wine program.exe > > see http://www.winehq.com/site/docs/wine-user/x1824#AEN1826 This page is outdated concerning the Wine Developer's Guide, and refers to a codeweaver no more existing page. See and the wrong URL: "For more details about debug channels, check out the The Wine Developer's Guide." Nicolas > ----- Original Message ----- > From: "Phillip Bruce" <pbbruce(a)mindspring.com> > To: "wine-users" <wine-users(a)winehq.com> > Sent: Sunday, April 10, 2005 10:21 PM > Subject: [Wine]Turining off Fixme Errors > > > > Hi, > > > > I hate asking for this as somone did provide me the answer > > earlier. But what is the winedebug syntax for turning off > > all fixme errors? > > > > Thanks. > > > > Phillip > _______________________________________________ > wine-users mailing list > wine-users(a)winehq.org > http://www.winehq.org/mailman/listinfo/wine-users -- Brouard Nicolas <nicolas.brouard(a)libertysurf.fr>

1 0

Mismatch in winternl.h for SYSTEM_*_INFORMATION structs
by Paul Vriens April 11, 2005

April 11, 2005

Hi, I'm busy fixing up NtQuerySystemInformation as this is needed for my 'Get Processor Explorer running' project. I've found that (at least) 2 structs are wrongly sized: /* System Information Class 0x03 */ typedef struct _SYSTEM_TIMEOFDAY_INFORMATION { #ifdef __WINESRC__ LARGE_INTEGER liKeBootTime; LARGE_INTEGER liKeSystemTime; LARGE_INTEGER liExpTimeZoneBias; ULONG uCurrentTimeZoneId; DWORD dwReserved; #else BYTE Reserved1[48]; #endif } SYSTEM_TIMEOFDAY_INFORMATION, *PSYSTEM_TIMEOFDAY_INFORMATION; /* was SYSTEM_TIME_INFORMATION */ In the __WINESRC__ case it's 32 bytes and 48 otherwise. The same is true for the _SYSTEM_PROCESSOR_PERFORMANCE_INFORMATION struct, where in the __WINESRC__ case it's 48 bytes and 44 bytes otherwise. So my questions are: 1. Is there a reason for this, or is it a bug. 2. Why do we have separate structs for the WINE and the NOT-WINE case. Is this because most of it is undocumented? Cheers, Paul Vriens.

1 0

LISTVIEW_SetItemState problem in FlashFXP (and other applications)
by Rolf Kalbermatter April 11, 2005

April 11, 2005

I was just discussing the problem with FlashFXP on Wine and did some Google search on this. It turned out that back in 2003 this problem was already noticed by one of the developers of FlashFXP and also somehow documented as follows: > http://forum.ru-board.com/topic.cgi?forum=5&topic=0132&start=240: > > 1. FlashFXP is no longer compatible with WINE, Switching to the OwnerData > listview has broken compatiblity, it's unlikely that WINE will be supported. > WINE needs to be made more compatible with FlashFXP. > |- After reviewing the wine source code I believe the problem is located here > http://source.winehq.org/source/dlls/comctl32/listview.c line 2971 & 2972 > 2971 for (i = nFirst; i <= nLast; i++) > 2972 LISTVIEW_SetItemState(infoPtr,i,&item); > The Problem: In OwnerData mode the OnData event is triggered when the > ItemState changes. However in the OnData event we set the item state. > > I suspect this causes an infinite loop. > This probably is very close to the actual problem. Looking at current listview.c in CVS I see following at about this location: > 2979 /* FIXME: this is not correct LVS_OWNERDATA > 2980 * setting the item states individually will generate > 2981 * a LVN_ITEMCHANGED notification for each one. Instead, > 2982 * we have to send a LVN_ODSTATECHANGED notification. > 2983 * See MSDN documentation for LVN_ITEMCHANGED. > 2984 */ > 2985 for (i = nFirst; i <= nLast; i++) > 2986 LISTVIEW_SetItemState(infoPtr,i,&item); Of course there are people here much more familiar with Common Controls than me, so I will probably not be able to do a quick fix to this myself. Just thought I would point out my findings so that others can confirm or reject the probable cause of the endless loop in FlashFXP and other applications. Rolf Kalbermatter

2 1