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

SHGetSpecialFolderPathA and CSIDL_FOLDER_MASK
by Chris Green 11 Oct '01

11 Oct '01

I've been looking at this function, and I am starting to suspect we have the wrong definition of CSIDL_FOLDER_MASK (currently 0x00ff). What I've been seeing (eg. installing MS Office 97) seems to imply that only the lower 6 bits, rather than 8, are related to the folder, and the others are something else. Does anyone else think that the mask should be 0x003f? regards Chris _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com

2 1

Re: Nasty Evil Memory Fragmentation fix
by Alexandre Julliard 11 Oct '01

11 Oct '01

Gavriel State <gav(a)transgaming.com> writes: > This still doesn't help us: as Ove points out, we still end up allocating > from newly created free blocks even if old blocks would suit us better. But you want to use new blocks in general. The exception is that you want to avoid using the uncommitted part at the end of the heap, and I agree we could do a special case for that. But it has nothing to do with the size of the block. > I've had a brief look at the code referenced by Adam Moss (found at > http://g.oswego.edu/dl/html/malloc.html) and it looks like it would > be a vast improvement over Wine's current Heap manager. It even has > Windows code that uses VirtualAlloc to reserve actual memory. Yes, that would certainly be better that tweaking the existing code. You may want to look at ptmalloc too, it's supposed to be better at multi-threading (at least to get some ideas, the license is not 100% compatible with ours). -- Alexandre Julliard julliard(a)winehq.com

1 0

Re: Nasty Evil Memory Fragmentation fix
by Alexandre Julliard 11 Oct '01

11 Oct '01

Ove Kaaven <ovehk(a)ping.uio.no> writes: > Well, after looking at some allocator algorithms, it seems to me there's > no reason to change neither the bin sizes nor the data structures. We just > have to change the free block selection algorithm - from my reading of the > code, right now it seems to always choose the *newest* (most recently > added) free block that's big enough. By changing it to always choose the > *oldest* free block instead, we'd already have a huge improvement. > (That's essentially the effect of Gav's patch, but only for big blocks, > but I suspect that could safely be done for *all* block sizes) I don't think that's a good idea. It's pretty bad for the cache and paging system to reuse oldest blocks first. -- Alexandre Julliard julliard(a)winehq.com

1 0

Re: wine/ include/win.h include/wine/server_protoc ...
by Gerard Patel 10 Oct '01

10 Oct '01

At 06:46 PM 10/10/2001 +0200, you wrote: >No the application does a SendMessage > >096ed3f8:Call user32.SendMessageA(ffffffff,0000c00c,00000000,00000000) ret=004141aa ffffffff = Broadcast message. Take a look at this /*********************************************************************** * broadcast_message_callback * * Helper callback for broadcasting messages. */ static BOOL CALLBACK broadcast_message_callback( HWND hwnd, LPARAM lparam ) { struct send_message_info *info = (struct send_message_info *)lparam; if (!(GetWindowLongW( hwnd, GWL_STYLE ) & (WS_POPUP|WS_CAPTION))) return TRUE; GetWindowLong does not work currently between processes. There are a lot to do for the whole of Apis to work between processes yes, but there is not so much remaining for this kind of installer to work IMO. Gerard

1 0

Buisness Apps
by Nick Hudson 10 Oct '01

10 Oct '01

Greetings, Has anyone got Quick Books Pro 2001 and ACT! 2000 to run on wine? In quick books I get this error when it goes to install it: fixme:ole:CoGetClassObject CLSCTX_LOCAL_SERVER not supported! and it says that the set up failed..... in ACT! 2001 I the install stalls out when I have to say what country I am in.... I have to ctrl+alt+F1 and do a killall to get rid of it. Also its not seeing my ethernet connection.... do I need to add anything to the config file to make this work right? I was trying to install IE over the web.... As far as I can tell it doesnt see that im on the network Thanks Nick Hudson

1 0

Re: wine/ include/win.h include/wine/server_protoc ...
by Alexandre Julliard 10 Oct '01

10 Oct '01

Uwe Bonnes <bon(a)elektron.ikp.physik.tu-darmstadt.de> writes: > The WIN_FindWndPtr message flooded the xterm during install, so I supected > soething still wrong... There are still a *lot* of functions that need to be adapted for handles belonging to other processes, so yes if you run multiple processes right now you'll most likely get flooded with messages. Working on it... -- Alexandre Julliard julliard(a)winehq.com

1 0

Re: wine/ include/win.h include/wine/server_protoc ...
by Alexandre Julliard 10 Oct '01

10 Oct '01

Uwe Bonnes <bon(a)elektron.ikp.physik.tu-darmstadt.de> writes: > No the application does a SendMessage > > 096ed3f8:Call user32.SendMessageA(ffffffff,0000c00c,00000000,00000000) ret=004141aa > err:win:WIN_FindWndPtr window 1002a belongs to other process > err:win:WIN_FindWndPtr window 10021 belongs to other process > err:win:WIN_FindWndPtr window 10029 belongs to other process > 096ed3f8:Ret user32.SendMessageA() retval=00000001 ret=004141aa > 096ed3f8:Call user32.SendMessageA(ffffffff,0000c00d,00000000,00000000) ret=004141bc > err:win:WIN_FindWndPtr window 1002a belongs to other process > err:win:WIN_FindWndPtr window 10021 belongs to other process > err:win:WIN_FindWndPtr window 10029 belongs to other process > 096ed3f8:Ret user32.SendMessageA() retval=00000001 ret=004141bc > > 096ed3f8 is a process created with Winexec, so I expect it to be a > sibling. Why isn't the WindowsFunction of at least that process called? It will be called only if the process has created a top-level window. Could you check if this is the case? -- Alexandre Julliard julliard(a)winehq.com

2 1

Re: CommandLineToArgvW fix
by Eric Pouech 10 Oct '01

10 Oct '01

just a last question (that I didn't check) what if the process name contains spaces like: C:\Program Files\foo\bar.exe what does Windows do ? returns in argv[0] "C:\Program" or "C:\Program Files". The later would also require testing of existence of the directories (doing a more semantical approach while current implementation is purely syntaxical) A+ > > > Here's another version that uses GlobalLock... > I also retested in Windows and I believe the comment about spaces was > wrong. > > > -- > Francois Gouget fgouget(a)free.fr http://fgouget.free.fr/ > It really galls me that most of the computer power in the world > is wasted on screen savers. > Chris Caldwell from the GIMPS project > http://www.mersenne.org/prime.htm > --------------- Eric Pouech (http://perso.wanadoo.fr/eric.pouech/) The future will be better tomorrow, Vice President Dan Quayle __________________________________________________________ Faites un voeu et puis Voila ! www.voila.fr Pendant tout l'été, gagnez une Ford Ka sur Voila Gratter ! http://cybergrattage.voila.fr/voila

2 1

Re: wine/ include/win.h include/wine/server_protoc ...
by Gerard Patel 10 Oct '01

10 Oct '01

At 04:55 PM 10/10/2001 +0200, you wrote: . > Alexandre> Implemented inter-process FindWindow(). > >Shouldn't this avoid errors like these: >err:win:WIN_FindWndPtr window 1002a belongs to other process >err:win:WIN_FindWndPtr window 10021 belongs to other process >err:win:WIN_FindWndPtr window 10029 belongs to other process Handles of other process windows can be returned by FindWindow, yes. But current Wine returns them also in enumeration functions, like EnumWindows. A functions like SetCursor, when trying to set the cursor for all desktop children, triggers this kind of message. The error message shows that Wine development is not complete because some function fails between processes (in SetCursor case, it is X11DRV_get_whole_window), not necessarily a problem in FindWindow. Try to lookup the origin for the handles : it is a problem only if the app get them from a FindWindow call. Gerard

2 1

Re: Compiling WINE on solaris 7
by Jeremy White 10 Oct '01

10 Oct '01

> > First I thought this was a problem caused by the SUN as, but then I > installed the GNU as, and typed > "make clean; rm config.cache; configure && make depend && make" > but the problem remained. I suspect that the problem is that gcc is still invoking 'as' instead of 'gas'. My understanding is the only clean way to get gcc to use the GNU binutils is to rebuild it with such support. What we've done is very hackish, but seems to work - place symlinks called 'as', 'cc', and 'ld' that point to the respective GNU tools ahead of the 'normal' SUN tools in the PATH. HTH Jer

1 0