- Wine-Devel - list.winehq.org

Re: Threading on Solaris
by Gregg Mattinson July 25, 2002

July 25, 2002

Hey everyone. I've been trying to update threads by following Francois' directions, and I think I may have come across a bug. Can someone help me out here... Why does the SYSLEVEL structure have a CRITICAL_SECTION inside it? Shouldn't a mutex be used instead? According to MSDN, the function EnterCriticalSection has the following property: If a thread terminates while it has ownership of a critical section, the state of the critical section is undefined. Also, WaitForSingleObject returns WAIT_ABANDONED if: the specified object is a mutex object that was not released by the thread that owned the mutex object before the owning thread terminated. Ownership of the mutex object is granted to the calling thread, and the mutex is set to nonsignaled. So, if a thread _EnterSysLevel and is then terminated, no other threads can enter that sys level. This seems wrong to me... Should I replace the CRITICAL_SECTION with a mutex? Any objections? Or, should EnterCriticalSection be fixed so that it has similar behaviour to WaitForSingleObject, instead of "undefined" behaviour (which is currently to block forever) Let me know what you guys think, Gregg Mattinson Co-op Developer Sun Microsystems of Canada >From: Gregg Mattinson <gm138242(a)opcom-mail.canada.sun.com> >Subject: Re: Threading on Solaris >To: wine-devel(a)winehq.com >Cc: fgouget(a)codeweavers.com >Date: Tue, 23 Jul 2002 07:21:46 -0400 (EDT) > >Threads on Solaris are a big issue for Sun, so I will start tackling this >problem today. Thanks for telling me all the issues Francois. I have a >hack-ish version implemented, and now I can create a real solution to the >problem. > >Gregg Mattinson >Co-op Developer >Sun Microsystems of Canada > >>From: Francois Gouget <fgouget(a)codeweavers.com> >>To: wine-devel(a)winehq.com >>CC: Gregg Mattinson <gm138242(a)opcom-mail.canada.sun.com> >>Subject: Threading on Solaris >>Date: Tue, 23 Jul 2002 00:35:09 -0700 >> >>Hi, >> >> >>Some time ago I did some tests of Wine on Solaris x86 and found that >>there is a problem with the way we handle threads due to the differences >>between the Linux and Solaris threading models. These differences are >>also relevant for Solaris Sparc and probably many other Unices. >> >>Unfortunately I don't have time to work on it anymore so I created a bug >>report in Bugzilla where I tried to dump the relevant information, >>including a patch to illustrate the issues and the start of a fix. I >>hope it will be useful to someone... maybe Gregg who has been doing a >>lot of Solaris work recently. >> >>http://bugs.winehq.com/show_bug.cgi?id=904 >> >> >>-- >>Francois Gouget >>fgouget(a)codeweavers.com >

2 1

Re: dos console apps question
by dank＠kegel.com July 25, 2002

July 25, 2002

Jukka wrote: > Yes, many DOS console apps don't work under plain Wine > because DOS emulation does not create console for the > applications. Some applications actually make Wine crash > because the application gets illegal coordinates when > GetConsoleScreenBufferInfo fails. > > The correct place for console allocation would likely > be VGA_SetAlphaMode; call to normal Windows AllocConsole() > function should be enough, no server calls are needed. > Additionally, if DOS console application tries to call a > routine that requires console, VGA_SetAlphaMode should be called > in order to initialize console (see VGA_SetCursorPos). > > I have been thinking doing this myself, but I have always > ended up being uncertain about how this change is going to affect > piping input and output of DOS programs that use only > stream IO and not console IO. Sounds like we need some regression tests that run DOS progams like 'find' and 'more', and make sure they work properly. I myself use the sourcesafe commandline tools, which are of the sort that just pipe characters to stdout and from stdin, and was horrified to discover that sometime after 9/2001, wine stopped running them correctly -- it started putting stdout in raw mode (!), which caused all line-oriented reads from stdin to fail. It also did curses initialization (!!), which made things look very messy indeed. I'm still running my snapshot from around 9/2001, but if I ever get up the energy, maybe I'll write a regression test to make sure e.g. DOS "more" works correctly. - Dan

1 0

Re: dos console apps question
by Jukka Heinonen July 25, 2002

July 25, 2002

Yes, many DOS console apps don't work under plain Wine because DOS emulation does not create console for the applications. Some applications actually make Wine crash because the application gets illegal coordinates when GetConsoleScreenBufferInfo fails. The correct place for console allocation would likely be VGA_SetAlphaMode; call to normal Windows AllocConsole() function should be enough, no server calls are needed. Additionally, if DOS console application tries to call a routine that requires console, VGA_SetAlphaMode should be called in order to initialize console (see VGA_SetCursorPos). I have been thinking doing this myself, but I have always ended up being uncertain about how this change is going to affect piping input and output of DOS programs that use only stream IO and not console IO. Anyway, I don't have time for this stuff so I would appreciate if you could work on this. -- Jukka Heinonen <http://www.iki.fi/jhei/>

1 0

Re: Procite and Wine
by Duane Clark July 25, 2002

July 25, 2002

On c.e.m.w Uwe Bonnes wrote: > Duane Clark <junkmail(a)junkmail.com> wrote: > : This is a multi-part message in MIME format. > : --------------040608050206010706050208 > : Content-Type: text/plain; charset=us-ascii; format=flowed > : Content-Transfer-Encoding: 7bit > > : Duane Clark wrote: > :> ... > :> Try this simple patch and see if it fixes it: > :> ... > > : Hmmm, that came out kind of poorly. Here it is as an attachment. > > This makes the the java application ChipView from the Xilinx webpack much > more usable (beside all those deficencies you expect form a java > application: delay and exceptions..) > > Please discuss (again?) on wine-devel. > > Bye Okay (switching from c.e.m.w to wine-devel), this had not yet been discussed on wine-devel, nor had I submitted the patch. It was one of those things I did awhile ago, and was going to look into more, but never got around to. I found this problem on two applications, and now it is apparently showing up on two more. On c.e.m.w Dong Wook Ko wrote: > Does anyone know where I should look for in order to use procite under wine? > > I got procite installed, and it sort of runs, but I can see it's > flickering very quickly, and my console is flooded with the same error > message: > > err:msg:DispatchMessageA BeginPaint not called on WM_PAINT for hwnd 30a4! > > And I cannnot open individual index of the bibs. > > Any help would be grateful. > > ps. If anyone could tell me how to export procite into bibtex, I won't > have to bother with this... > I have seen this in two other programs. The problem appears to be here, at about line 779 in windows/painting.c: else /* entire window or client depending on RDW_FRAME */ { if( flags & RDW_FRAME ) hRgn = 1; else { GETCLIENTRECTW( wndPtr, r2 ); hRgn = CreateRectRgnIndirect( &r2 ); } } Reading the MS docs, it appears to me that in this case RDW_FRAME should be ignored. What happens is the program has an area that needs to be repainted, but this part of the code causes only a portion to be repainted, leaving a portion of the screen still "invalidated". This causes another paint request, which since nothing has changed, still does not result in the correct repainting. And so it is stuck in an indefinite loop. This rather simple patch appears to fix things, but I did not submit it because it did not completely clear up the screen corruption (for example testing on Kaleidegraph, for which a free download is available somewhere...), though it did stop this bug: Index: windows/painting.c =================================================================== RCS file: /home/wine/wine/windows/painting.c,v retrieving revision 1.71 diff -u -r1.71 painting.c --- windows/painting.c 5 Jul 2002 01:23:31 -0000 1.71 +++ windows/painting.c 25 Jul 2002 00:44:15 -0000 @@ -779,15 +779,9 @@ OffsetRect( &r2, pt.x, pt.y ); hRgn = CreateRectRgnIndirect( &r2 ); } - else /* entire window or client depending on RDW_FRAME */ + else /* entire window */ { - if( flags & RDW_FRAME ) hRgn = 1; - else - { - GETCLIENTRECTW( wndPtr, r2 ); - hRgn = CreateRectRgnIndirect( &r2 ); - } } }

1 1

Re: Get number of bytes available in PeekNamedPipe (take 2)
by Alexandre Julliard July 25, 2002

July 25, 2002

Uwe Bonnes <bon(a)elektron.ikp.physik.tu-darmstadt.de> writes: > As I don't have a test target, I don't feel yet like filling lpvBuffer and I > don't yet have an idea how to do that. I looked at the kernel sources and > lseek unconditionally returns error on pipe. So how can we read the pipe and > get the filepointer back to the original position afterwards? You can't use lseek, you have to make the pipe a socket pair and use recv(MSG_PEEK). > wine/server/protocol.def > Define a FD_TYPE_PIPE > wine/server/file.c: get_handle_fd > On type FD_TYPE_PIPE check other end of type If this is really necessary it should be done on the client side. I'm not sure I see the point though. -- Alexandre Julliard julliard(a)winehq.com

2 2

DTR / DTS vs RTS/CTS vs Wine
by List vmn July 25, 2002

July 25, 2002

Hi, I am testing a serial printer driver that uses DTR/DTS hardware flow control on wine. While testing the driver running winedbg everything worked, not so when I ran the program directly from wine. I sent this problem to the printer manufacturer and they put a data scope on the serial line. It seems that the DTR/DTS signal was not being caught by the computer and therefore the computer was sending info over the serial port when the printer was not ready. I read on the net that the linux kernel does not implement DTR/DTS but only RTS/CTS hardware flow control. I also read that it is usual for cables to be built that connect the RTS/CTS pins to the DTR/DTS pins therefore tricking the computer and printer [the computer understands RTS/CTS and the printer DTR/DTS ] and getting good results. The output of --debugmsg +comm after SetCommState was called was ~IXONN ~IXOFF. Can anybody indicate that If I build the correct cable and ask for RTS/CTS with SetCommState , that the serial io will work? Any limitation of wine with RTS/CTS? Is it correct that windows does not do ANY hardware flow control? If so, how can this app work under windows since it only uses standard SetCommState, ReadFile, WriteFile calls... Thanks, Miguel Feitosa

5 5

Re: wine/dlls/shell32 shellord.c shlexec.c
by Max July 25, 2002

July 25, 2002

Tony Lambregts wrote: > > It seems that one of the patches to shlexec.c has broken WinZip. The > following is a trace from WinZip from this morning. > > > fixme:toolbar:TOOLBAR_CheckStyle [10026] TBSTYLE_ALTDRAG not implemented > fixme:toolbar:TOOLBAR_CheckStyle [10026] TBSTYLE_ALTDRAG not implemented > fixme:listview:LISTVIEW_UnsupportedStyles !LVS_SHAREIMAGELISTS > err:shell:SHGetFileInfoA pidl is null! > fixme:listview:LISTVIEW_SetColumnOrderArray iCount 9 lpiArray 0x405b689c > err:shell:SHGetFileInfoA pidl is null! > > I have been trying to debug this for a while and it did not have this > problem yesterday. If you want to download WinZip here is the url > http://www.winzip.com/ Same problem, winzip is broken as explained; I tested it some days ago and it worked fine. Regards Max

1 0

Re: Dinput - Keyb Bug
by Kristoffer Ericson July 25, 2002

July 25, 2002

Greetings, I've been working with this problem for about 2 weeks now (play alot of tribes) and it's finally working with keyboard. I tried dinput about a week ago but couldn't make it work but as you said i should have used directx5 or 6 instead. Keyboard works fine but i still have other issues, to make the game fully functional. Dsound komplains somewhat and the sound sounds pitched (plays to fast, thus giving the smurf syndrom), i tried to use -dll dsound.dll but it says that it can't find the file. I know i've done right so i checked around on irc.openprojects.net and they said that not all -dll's can be used natively, is this right? Also i fail to get opengl to work, i simply can't select it in game. Any ideas? I've compiled with --enable-opengl and i've run a testprogram using opengl which works fine. Tribes running in fullscreen mode it feels slow and strange, but running it in window mode makes it feel fast and stable. Why? i tried to use native ddraw.dll but then the game won't even start, there's just a black screen. Thank you for all your support, considering i've been running around looking at homepages and chatting on irc when you all solved my problem within 1 day, THANKS(!). Other than that Tribes works fine. >From: Steven Edwards <steven_ed4153(a)yahoo.com> >To: Kristoffer Ericson <kristoffer_e1(a)hotmail.com>, wine-devel(a)winehq.com >Subject: Re: Dinput - Keyb Bug >Date: Tue, 23 Jul 2002 15:11:53 -0700 (PDT) > >You need to use the windows dinput.dll as ours is (still) not up to spec. I >used to play this game >all the time under linux till tribes2. I dont know if this will work with >the directx 8.1 but if >you can find a dinput.dll from dx5 or 6 it should work. > >I will add this information to the wine application database if its not >already there and if >someone working on directx can figure out what is going on that would rock >long term. > >Thanks >Steven > > >--- Kristoffer Ericson <kristoffer_e1(a)hotmail.com> wrote: > > I'm trying to get StarSiege Tribes (1) to work, and from what i gather >is it > > suffering from a quite common wine-gaming-problem (both wine & winex). > > > > The game starts smoothly enough but keyboard is dead. When i press any >key > > it will instantly drop out and quit. No error message or breaks. > > Considering the keyboard makes you move it basicly leaves you jumping up >and > > down shooting at everything that comes..=) Other than the keyboard error > > tribes runs fine. > > > > I daily run updates from winex & wine but been working at this problem >for > > about 2 weeks. > > > > I've read all the info i could find on the subject and from what i can > > gather is that tribes used to work on older wine. So can anyone give me >some > > futher info about where i could find older wine-archives or if the >dinput (i > > guess the problem is due to dinput) problem is about to be solvedŽ, >tweaks, > > patches, ideas.. > > > > With respect Ken > > > > > > > > _________________________________________________________________ > > Chat with friends online, try MSN Messenger: http://messenger.msn.com > > > > > > >__________________________________________________ >Do You Yahoo!? >Yahoo! Health - Feel better, live better >http://health.yahoo.com _________________________________________________________________ Join the worlds largest e-mail service with MSN Hotmail. http://www.hotmail.com

3 2

Calling a winelib program (wcmd) from inside wine with arguments
by Uwe Bonnes July 25, 2002

July 25, 2002

Hallo, how do I call a winelib programm with arguments from inside wine. I tried several appoaches in a first hack for Batch file support in wine, but I didn't succeed. It seemed that either the winelib programm insisted on the "--" divider or it tried to interpret it. Thanks -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

2 1

Re : Coded breakpoints
by Max July 24, 2002

July 24, 2002

Never mind, if WAS a stupid question :-) For who needs it too, just add __asm__("int $0x03\n\t") ; at the required point, maybe inside an huge if(....) Regards Max

4 3