October 2001 - Wine-devel - list.winehq.org

Re: Page faults and instruction emulation
by Alexandre Julliard 19 Oct '01

19 Oct '01

Ove Kaaven <ovehk(a)ping.uio.no> writes: > Well, perhaps it could depend on the windows version somehow... No, you really don't want to have the layout of core data structures depend on the version. > Maybe... I was just concerned whether you'd like Wine to be subject to > lawsuits or something (reverse engineering, trade secrets, DMCA, > whatever... not that the kernel driver is really complex, though). > Also, I was afraid some games might refuse to work with --winver nt40 I don't think lawsuits should be a problem, as long as we make a reasonable effort to implement the correct functionality and not just fake it. Of course there's always a risk that someone will sue, but these days the risk exists as soon as you release 3 lines of code. We really should move towards an NT-like architecture as much as possible; this is where Windows is going too, and it's the only one we have a chance of making work 100% correctly. We'll never be able to implement the full brokenness of the Win95 "architecture" anyway, so it's better to concentrate on the things that we can make work right. -- Alexandre Julliard julliard(a)winehq.com

1 0

Re: File media type
by Alexandre Julliard 19 Oct '01

19 Oct '01

Ove Kaaven <ovehk(a)ping.uio.no> writes: > But since to do this, MapViewOfFile needs to know the media type, and the > file name is only known in CreateFile, I apparently need to store the > media type in the wineserver object to accomplish this. > > Would changing the wineserver protocol in this way be a good idea? It's certainly a possibility. It would be nice to be able to tell directly from the file descriptor, so that we wouldn't depend on the user getting the drive specifications right. But I can't think of a Unix function that would allow doing this cleanly. -- Alexandre Julliard julliard(a)winehq.com

3 2

Re: wine/port.h #include fixes
by Roger Fujii 18 Oct '01

18 Oct '01

> It's slowly coming around (Francois recent patches are key), > but we still have some fairly serious issues. Is this in? As of 10/17, it still blows up in NtCurrentTeb(). What does this function do? - is it trying to use a protected instruction? (in x86 instruction knowledge is pretty dated). > You've hit the first issue - being sure to use the > GNU tool chain. yup... want to try tackling that once I get the program to at least come up... > Doing a cvs update and trying again (after, of course, > Alexandre commits Francois' recent patches) may > bring you more joy. a bold statement.. :) I presume that you don't have to install it to test it out, right? (so long as I have LD_LIBRARY_PATH set to dlls:.). I take it that someone does have something that works on solx86? -r

3 6

Re: wine/port.h #include fixes
by Alexandre Julliard 18 Oct '01

18 Oct '01

Roger Fujii <rmf(a)lookhere.com> writes: > well, I think more to the point is that something says what > steps need to be done to run a specific version. The current > instructions just say to set LD_LIBRARY_PATH, which isn't > really correct. So, just for the record, all you have to do > is set LD_LIBRARY_PATH to srcdir:srcdir/dlls and WINESERVER > to srcdir/server/wineserver? Or do you need access to things > like winebuild? To make absolutely sure you don't pick anything from the configured prefix directory, you have to set WINESERVER, WINELOADER, WINEDLLPATH and LD_LIBRARY_PATH. -- Alexandre Julliard julliard(a)winehq.com

1 0

Re: Page faults and instruction emulation
by Alexandre Julliard 18 Oct '01

18 Oct '01

Ove Kaaven <ovehk(a)ping.uio.no> writes: > I know of two Wine architecture issues that must be resolved before the > anti-debugger check can succeed in win95 mode. > > 1. it checks that teb->debug_context (TEB offset 0x20) is zero. In Wine, > it isn't (the NT pid is stored there), so this check fails. > > Is it possible to change Wine's TEB structure to fix this? It would probably break some of the NT native dlls. > 2. it messes with the IDT. It uses the sidt instruction and tries to > change the INT5 vector (BOUND exception vector), later to try to execute a > BOUND and expect this vector to be branched to, all protected with > exception handlers. Now, this would actually be possible for Wine to > emulate, even though sidt is not a privileged instruction. The idea is > that sidt returns an address to kernel space, where all access is > prohibited, hence causing page faults. Wine could trap these, and then > emulate the faulting instruction by operating on a fake IDT instead. And > then the segfault caused by BOUND could then look up the fake IDT entry, > and call the vector therein. > > But to emulate an instruction from Wine's pre-exception page fault handler > (VIRTUAL_HandleFault()), the current context must be passed along to it, > so VIRTUAL_HandleFault and all Wine handlers that can be registered with > VIRTUAL_SetFaultHandler must get an extra parameter. Is this too > objectionable to do? The extra parameter is not that bad, but having to do full emulation of all instructions accessing memory, plus the complete IDT handling, is going to lead to major ugliness. Wouldn't it be possible to somehow emulate the kernel driver to make the NT version work? That would be much cleaner. -- Alexandre Julliard julliard(a)winehq.com

2 1

Re: wine/port.h #include fixes
by Alexandre Julliard 18 Oct '01

18 Oct '01

Fran�ois Gouget <fgouget(a)codeweavers.com> writes: > Should Wine really try the installed prefix first? It seems to me > that it should try it last instead. After all, if you change the path > you would expect your change to have an effect, no? No, you don't want changing the PATH suddenly to start breaking your applications. > Plus checking the > installed prefix first assumes that whichever version of wine is started > (e.g. /home/fgouget/bin/wine) is the same as that of > /usr/bin/wineserver! Yes of course, if you compiled it with /usr/bin as prefix it assumes that /usr/bin is where its wineserver is. If you don't want that you have to change the prefix, or set WINESERVER explicitly. -- Alexandre Julliard julliard(a)winehq.com

2 1

Page faults and instruction emulation
by Ove Kaaven 18 Oct '01

18 Oct '01

As we all know, some (kinda worthless) "copy protection" software exist out there that doesn't work under Wine. Some of these work under win9x and winNT by using a different set of techniques for each OS. The winNT code installs a kernel driver, so I'm mainly interested in the win95 code, which only messes with Win95 internals to check that no debugger is snooping, before proceeding. I know of two Wine architecture issues that must be resolved before the anti-debugger check can succeed in win95 mode. 1. it checks that teb->debug_context (TEB offset 0x20) is zero. In Wine, it isn't (the NT pid is stored there), so this check fails. Is it possible to change Wine's TEB structure to fix this? 2. it messes with the IDT. It uses the sidt instruction and tries to change the INT5 vector (BOUND exception vector), later to try to execute a BOUND and expect this vector to be branched to, all protected with exception handlers. Now, this would actually be possible for Wine to emulate, even though sidt is not a privileged instruction. The idea is that sidt returns an address to kernel space, where all access is prohibited, hence causing page faults. Wine could trap these, and then emulate the faulting instruction by operating on a fake IDT instead. And then the segfault caused by BOUND could then look up the fake IDT entry, and call the vector therein. But to emulate an instruction from Wine's pre-exception page fault handler (VIRTUAL_HandleFault()), the current context must be passed along to it, so VIRTUAL_HandleFault and all Wine handlers that can be registered with VIRTUAL_SetFaultHandler must get an extra parameter. Is this too objectionable to do? Alexandre?

1 0

File media type
by Ove Kaaven 18 Oct '01

18 Oct '01

I'm trying to think of a way to make multi-CD installers work under Wine. Right now, the hardest part is that when the installers tell you to change discs, the Setup.exe file mappings are still open. I'm thinking of making Wine not map the file, but read it straight into RAM, if the executable to be mapped is on removable media. But since to do this, MapViewOfFile needs to know the media type, and the file name is only known in CreateFile, I apparently need to store the media type in the wineserver object to accomplish this. Would changing the wineserver protocol in this way be a good idea? Index: server/protocol.def =================================================================== RCS file: /cvsroot/winex/wine/server/protocol.def,v retrieving revision 1.5 diff -u -r1.5 protocol.def --- server/protocol.def 2001/10/13 20:20:06 1.5 +++ server/protocol.def 2001/10/18 22:50:29 @@ -505,6 +505,7 @@ unsigned int sharing; /* sharing flags */ int create; /* file create action */ unsigned int attrs; /* file attributes for creation */ + int mediatype; /* media type file is on */ VARARG(filename,string); /* file name */ @REPLY handle_t handle; /* handle to the file */ @@ -527,6 +528,7 @@ @REPLY int fd; /* file descriptor */ int type; /* the type of file */ + int mediatype; /* media type file is on */ @END #define FD_TYPE_INVALID 0 #define FD_TYPE_DEFAULT 1

2 1

Re: Use the debugmsg for the pthread FIXME messages.
by Alexandre Julliard 18 Oct '01

18 Oct '01

"Medland, Bill" <Bill.Medland(a)accpac.com> writes: > I am fed up with the pthread fixmes; we ought at least to be able to turn > them off with debugmsg. This will crash and burn if glibc calls a pthread function before ntdll has been properly initialized. Besides, the messages are annoying on purpose: to motivate someone to implement the missing functions... -- Alexandre Julliard julliard(a)winehq.com

1 0

lsarpc.idl API
by Luke Kenneth Casson Leighton 18 Oct '01

18 Oct '01

Mail-Followup-To: tng-technical(a)samba-tng.org, lsa-dev(a)lists.dcerpc.net, freedce-dev(a)lists.dcerpc.net, samba-ntdom(a)samba.org, samba-technical(a)samba.org, wine-dev(a)winehq.com User-Agent: Mutt/1.2.4i X-original-sender: lkcl(a)samba-tng.org Precedence: bulk Reply-To: tng-technical(a)lists.dcerpc.org X-list: tng-technical thanks to matty for the lsarpc.idl file: it compiles in dceidl (slightly modified) and we're away! client-side header file and API is generated _automatically_ by dceidl, for immediate use with freedce libraries. porting of TNG's srv_lsa_nt.c can now take place if anyone's interested. volunteers, anyone? luke p.s. we also have winreg.idl so that means that the Reg* API is also automatically available for immediate use. if anyone's interested.

1 0