May 2004 - Wine-Devel - list.winehq.org

text edit box not painting text ( on form startup )
by Doug Herbert March 9, 2005

March 9, 2005

( I have posted a similiar question in wine-users, don't flame me for doing so, just thinking about this some more, perhaps it may be more a developer question ) Ok, I have a windows app, that runs under wine fine - not quite. This app has a form, with many text field edit boxes on. Quite often these edit boxes already have text values, ie. they are not empty - there is a database behind the form. Anyway, the form shows on the screen, but the text within the edit fields is invisible, until you activate each edit box component. When you leave one edit box to the next, the text remains visible. It seems like the repainting is not working, on initial startup of the form. As a way to debug this database app ( I don't have the source code for it ) I wrote a real basic form app in Visual Studio, with two edit boxes, with data in each. Now this basic app shows the text values immediately on startup. So there must be something different that the database app is not doing. I then ran wine using 'wine --debugmsg +edit programname' for both app's. I see both logging 'Creating ANSI edit control, style = xxxxx' but the style reference is different between apps - I am not sure if this the cause or not. The problem app reports style = 54011104 and the basic app shows style = 50010080. Can somebody explain what this style reference means and how do I force an app to use a certain style of edit box ? What other wine class debugmsg types should I use to narrow down the source of this problem ? Is there anyway I can check to see how the database app is repainting text within the text edit components ? Regards Doug Herbert. __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com

3 4

Wine bug with delphi Menus
by Robert van Herk Sept. 5, 2004

Sept. 5, 2004

Hi people, I think I discovered a wine bug. In a delphi program, changing the visibility of a menuitem (TMenuItem) causes the whole menu of that form to stop working, when the program is run under Wine. Behind the screens, the setVisible of a menu item calls MenuChanged. This seems to put the menu in such a state that, in wine, all menu components stop doing anything. I've written a small application (455 kb EXE) to show what goes wrong. Would anybody be interested to help me in solving this bug? Would a message trace from Winsight (for example) be of any help? Greetings, Robert

2 1

Precompiled headers?
by Shachar Shemesh July 5, 2004

July 5, 2004

Hi, Is anyone considering putting precompiled headers support into the wine build system? http://gcc.gnu.org/onlinedocs/gcc/Precompiled-Headers.html#Precompiled%20He… -- Shachar Shemesh Lingnu Open Source Consulting http://www.lingnu.com/

2 1

General serial communications issues
by David Purucker June 14, 2004

June 14, 2004

Hello, I have tried several terminal programs (Hyperterm and TeraTerm) under several builds of Wine including Wine-20040505 and they seem to have the same generic problem. They all exhibit the same situation, DATA seems to go OUT, but never seems to come IN. Several other people have experienced the same thing, as seen in this bug report: http://bugs.winehq.org/show_bug.cgi?id=2115 By the way, TeraTerm does work fine in TCP/IP mode, so it's definitely a serial problem. It's freeware available here: http://hp.vector.co.jp/authors/VA002416/teraterm.html It appears to be a fundamental problem in the serial/file code, which I understand is now under the ntdll tree. Obviously, I don't know much about it, but there is definitely something wrong. I was wondering if any developers were aware of this problem, as the decreased general usage of serial communications may have caused this to slip under the carpet, so to speak. I look forward to hearing any ideas, work-arounds, or patches. Thanks _________________________________________________________________ Stop worrying about overloading your inbox - get MSN Hotmail Extra Storage! http://join.msn.click-url.com/go/onm00200362ave/direct/01/

7 20

[REGRESSION] Photoshop 7.0 works before wine-cvs-20040130 16:??:?? CST (bug #2017)
by Luca Capello June 13, 2004

June 13, 2004

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello, as reported on bug #2017 [1], Adobe Photoshop 7.0 stops working after wine-20031212 (in the majority of cases). This happens on Debian unstable (where I am) as well as on other distro. As suggested by Tony Lambregts, I made a regression testing to find the patch that broke this. The problem is between cvs-20040130-165720 and cvs-20040130-165721. The 'cvs -d ...' command gives me updates for only 3 files: ===== U dlls/kernel/Makefile.in U dlls/kernel/actctx.c U dlls/kernel/kernel32.spec ===== so I think the problem should be in these files, am I right? If it's so, well, the patch that broke Photoshop 7.0 working is the following [2] - - Jon Griffiths <jon_p_griffiths(a)yahoo.com> Add stubs for the Activation Context API (XP+). Please fix this problem, now that you know the cause. Thx, bye, Gismo / Luca [1] http://bugs.winehq.org/show_bug.cgi?id=2017 [2] http://www.winehq.org/hypermail/wine-cvs/2004/01/0394.html -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) Comment: Using GnuPG with Debian - http://enigmail.mozdev.org iD8DBQFAu1FuVAp7Xm10JmkRAgWIAJ90THCdqfIu6/uLGyurqrLfNNWEkACeIAze ALi4ZeOtMTEEDp8ETWu4G4c= =/pZb -----END PGP SIGNATURE-----

6 25

Request for winetesting volunteers
by Chris Morgan June 3, 2004

June 3, 2004

Thanks to Brian Vincent, Paul Millar, Dimi Paun, Ferenc Wagner and Kevin Koltzau the automated winetest suite is ready for more widespread testing. Kevin and Paul build winetests each night if there were cvs commits since the last time the tests were built. The winetests are published to a script on winehq that Brian and Dimi wrote. The winrash service that I wrote sends its current state to the service and receives a script to process that may direct it to check in later, download a winetest and run it or upgrade itself. Thanks to Ferenc the winetest results are sent back to winehq and can be accessed via http://test.winehq.org/data/ Pretty formatting of the results is coming soon ;-) We need volunteers that are willing to particpate in the winetesting process by running the winrash service on their computers. The winrash installer can be downloaded from: http://prdownloads.sourceforge.net/winrash/winrash-0007-chris-msvc.exe?down… You can find out more about winrash at http://winrash.sf.net The installer will install winrash as a service and run Wine tests as required. It is IMPORTANT that you set the client id correctly. Please use your first and last name WITHOUT spaces, like JohnSmith. Using the default client id will cause your winetest results to be ignored. The service is smart enough to autoupgrade as new versions are released. To uninstall just go to the Add/Remove programs and click 'remove' as you do for many windows programs. There is the ability to restrict the times when winrash runs winetests but this requires manually adding some keys to the registry. I haven't finished testing this feature but if someone wants it I can see about getting it polished up and in a release early this coming week. We currently need testers for the whole range of windows versions, win95/98/me, nt3/4, win2k, xp and 2003. Please reply if you want to help out so that we know what kind of coverage we can expect. The more the merrier, so please sign up even if others have done so for your version of Windows. Chris

3 5

wine
by Mike June 1, 2004

June 1, 2004

m tring to play a game using wine does anybody no why i receive this message and if so how to fix? Thanks Mike > Exception at EIP=0081F2B4: ACCESS VIOLATION > While attempting to read memory at 00000000 > ----------------------------------------------------- > EAX=00000000 EBX=00000001 ECX=0000013C EDX=41796208 > ESI=000006D7 EDI=407CFD7C EBP=00000000 ESP=407CFC30 >

2 1

Current Wine CVS status
by Dmitry Timoshkov June 1, 2004

June 1, 2004

Hello all (mostly Alexandre :-)), running today's ~/wine/wine under an old RH 6.2 install I get the following: Usage: ld.so [OPTION]... EXECUTABLE-FILE [ARGS-FOR-PROGRAM...] You have invoked `ld.so', the helper program for shared library executables. This program usually lives in the file `/lib/ld.so', and special directives in executable files using ELF shared libraries tell the system's program loader to load the helper program from this file. This helper program loads the shared libraries needed by the program executable, prepares the program to run, and runs it. You may invoke this helper program directly from the command line to load and run an ELF executable file; this is like executing that file itself, but always uses this helper program from the file you specified, instead of the helper program file specified in the executable file you run. This is mostly of use for maintainers to test new versions of this helper program; chances are you did not intend to run this program. --list list all dependencies and how they are resolved --verify verify that given object really is a dynamically linked object we get handle --library-path PATH use given PATH instead of content of the environment variable LD_LIBRARY_PATH --inhibit-rpath LIST ignore RPATH information in object names in LIST -- Dmitry.

2 2

broken mingw build
by Kevin Koltzau June 1, 2004

June 1, 2004

This patch http://cvs.winehq.org/patch.py?id=12495 seems to have broken the mingw build, currently getting i386-mingw32msvc-gcc -c -I. -I. -I../../include -I../../include -D__WINESRC__ -D_REENTRANT -Wall -pipe -fno-strength-reduce -mpreferred-stack-boundary=2 -fno-strict-aliasing -gstabs+ -Wpointer-arith -g -O2 -o interlocked.o interlocked.c {standard input}: Assembler messages: {standard input}:124: Error: unknown pseudo-op: `.previous' {standard input}:135: Error: unknown pseudo-op: `.previous' {standard input}:145: Error: unknown pseudo-op: `.previous' {standard input}:155: Error: unknown pseudo-op: `.previous' {standard input}:165: Error: unknown pseudo-op: `.previous' make[2]: *** [interlocked.o] Error 1

3 2

Re: [DInput] Fix RailroadTycoon
by Christian Costa May 31, 2004

May 31, 2004

Lionel Ulmer wrote: >Sorry for the huge diff, but I got sick when I saw my previous 2-space >indenting style, so I had to re-indent everything... For people who want to >do a code review, I could send them a 'diff -w' version :-) > Now, you're proposing, that would be cool. :-) > >And before anyone complains, yeah, I should have sent two patches (one >cosmetic, one with the real meat in it) :-) > > Lionel > >Changelog: > - some TRACEing improvements > - fix mouse Acquire which switched right and middle buttons > - reset the mouse cursor at its original location on Unacquiring the device > > > > > Bye, Christian

2 2

Making a system DLL unavailable (problem with COMMCTRL)
by Samuel Audet May 31, 2004

May 31, 2004

Hello, I have a Windows 3.1 application here that uses the OWL library. I have problems running it properly, and I have found the following: http://www.xploiter.com/programming/c/borland/3031.html Looking at what wine outputs when I use "WINEDEBUG=dll" or loaddll, I see that my program tries to load commctrl.dll and always succeeds. The problem is, I want it not to succeed, since AFAIK, 16 bit COMMCTRL functionality is not implemented in WINE, and the OWL library has a backup plan where it emulates some of the controls (like described on the site above) when it can't find COMMCTRL, which is what it does in Windows 2000 or XP for example. I have confirmed that the COMMCTRL is at the source of the problem by using one from a Windows 98 installation. Dialogs that are not showing using WINE's builtin COMMCTRL are now showing (although they are having sizing problems... anybody might know why?). Is there like a "disable" option in the [DllOverrides] section? Or some other way to disable a system DLL? I can't seem to just be able to "remove" commctrl.dll since even 32 bits builtin DLL (in this case comctl32.dll) require the 16 bit counterpart. In this case, when comdlg32 tries to load, it loads comctl32 and fails because it can't load commctrl. I don't think it's a good idea to make 32 bits DLL be dependent of their 16 bit counterpart. Thank for any input! -- Samuel Audet <guardia(a)step.polymtl.ca> http://www.step.polymtl.ca/~guardia/

3 5

Using objdump
by James Courtier-Dutton May 31, 2004

May 31, 2004

I have used objdump in the past to disassemble and check the code that is generated in a .so or .o file. How do I modify objdump so I can get it do the same with win32 .exe and .dll files ? Cheers James

4 4

Crash due to _fini
by Rajeev Bansal May 31, 2004

May 31, 2004

Hi i am trying to link wine library(msvcrt.dll.so) with my code but i getting segv in _fini call of msvcrt library Please tell me a way so that i can use wine libraries directly without winegcc or specs file on using winegcc for code which contains new and delete operator it gives undefined reference to following symbols operator new(unsigned) operator delete[](void*) _Unwind_Resume __gxx_personality_v0 please help me to resolve above symbols waiting for response Thanks Rajeev Test Program : #include <dlfcn.h> #include <direct.h> #include <math.h> #include <stdio.h> main() { char *a; char *b; double d; void *mod, *ker; const char *error; mod=dlopen("msvcrt.dll.so",RTLD_LAZY); ker=dlopen("kernel32.dll.so",RTLD_LAZY); if(mod) { if ((error = dlerror())) { printf("Couldn't find getcwd: %s\n", error); } if(dlsym(mod, "getcwd")){ printf("found"); b=getcwd(a,20); } dlclose(mod); dlclose(ker); } //end if return (0); } //end main Build command : g++ -g -Wall -c -I/home/timex/wine/wine-20040505/include -I. -I/home/timex/wine/wine-20040505/include/msvcrt test.cpp g++ -g -ldl /home/timex/wine/wine-20040505/dlls/msvcrt.dll.so test.o Debugger output [me(a)cvs TUCP]$ gdb -n GNU gdb Red Hat Linux (5.2.1-4) Copyright 2002 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i386-redhat-linux". (gdb) file ./a.out Reading symbols from ./a.out...done. (gdb) break test.cpp:10 Breakpoint 1 at 0x80486cc: file test.cpp, line 10. (gdb) run Starting program: /home/a.out Breakpoint 1, main () at test.cpp:14 14 mod=dlopen("msvcrt.dll.so",RTLD_LAZY); (gdb) next 15 ker=dlopen("kernel32.dll.so",RTLD_LAZY); (gdb) disp mod 1: mod = (void *) 0x40013578 (gdb) disp ker 2: ker = (void *) 0x4026cc40 (gdb) next 16 if(mod) 2: ker = (void *) 0x8049b60 1: mod = (void *) 0x40013578 (gdb) next 19 if(dlsym(mod, "getcwd")) 2: ker = (void *) 0x8049b60 1: mod = (void *) 0x40013578 (gdb) next 20 {printf("found"); 2: ker = (void *) 0x8049b60 1: mod = (void *) 0x40013578 (gdb) next 21 b=getcwd(a,20); 2: ker = (void *) 0x8049b60 1: mod = (void *) 0x40013578 (gdb) step getcwd (buf=0x4026cc40 "T\v\022", size=20) at direct.h:73 73 static inline char* getcwd(char * buf, int size) { return _getcwd(buf, size); } (gdb) step _getcwd (buf=0x4026cc40 "T\v\022", size=20) at dir.c:424 424 int dir_len = GetCurrentDirectoryA(MAX_PATH,dir); Current language: auto; currently c (gdb) step 422 { (gdb) step 424 int dir_len = GetCurrentDirectoryA(MAX_PATH,dir); (gdb) step 422 { (gdb) step 424 int dir_len = GetCurrentDirectoryA(MAX_PATH,dir); (gdb) step Program received signal SIGTRAP, Trace/breakpoint trap. 0x40062179 in _fini () from /home/timex/wine/wine-20040505/dlls/msvcrt.dll.so (gdb) bt #0 0x40062179 in _fini () from /home/timex/wine/wine-20040505/dlls/msvcrt.dll.so #1 0x08048790 in getcwd (buf=0x4026cc40 "T\v\022", size=20) at direct.h:73 #2 0x08048730 in main () at test.cpp:21 #3 0x40161b84 in __libc_start_main () from /lib/libc.so.6 (gdb) --- Outgoing mail is certified Virus Free. Checked by AVG anti-virus system (http://www.grisoft.com) Version: 6.0.642 / Virus Database: 410 - Release Date: 3/24/2004

2 1

segfault fault at startup
by Florian Goth May 31, 2004

May 31, 2004

H! I have the same Problem. I get a segfault, when I do "wine myapp" and when I do "wine-pthread" I'm using a SuSE System ( ~ 9.0 ) with kernel 2.6.6 and gcc 3.4.0 this is the gdb output: (gdb) set follow-fork-mode child (gdb) r Starting program: /usr/local/bin/wine [New Thread 16384 (LWP 5206)] [New Thread 32769 (LWP 5209)] [New Thread 16386 (LWP 5210)] Cannot find user-level thread for LWP 5206: generic error (gdb) this is some of the last strace output I get: t_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 writev(4, [{"$\0\0\0002\0\0\0\0\0\0\0\0\0\0\200\0\0\0\0\1\0\0\0\1\0"..., 64}, {"/home/sifff/.wine/dosdevices/c:/"..., 50}], 2) = 114 read(5, "\0\0\0\0\0\0\0\0000\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 rt_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 write(4, "G\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0A\0\0\0\5\0\17\0"..., 64) = 64 read(5, "\0\0\0\0\0\0\0\0004\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 rt_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 write(4, "I\0\0\0\0\0\0\0\0\0\0\0004\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 read(5, "\0\0\0\0\0\0\0\0\0\0\0\0\0000M\0A\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 rt_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 write(4, "&\0\0\0\0\0\0\0\0\0\0\0004\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 read(5, "\0\0\0\0\0\0\0\0\377\377\377\377\1\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 recvmsg(3, {msg_name(0)=NULL, msg_iov(1)=[{"4\0\0\0", 4}], msg_controllen=16, msg_control=0xbfffab68, , msg_flags=0}, 0) = 4 fcntl64(9, F_SETFD, FD_CLOEXEC) = 0 rt_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 write(4, "\26\0\0\0\0\0\0\0\0\0\0\0004\0\0\0\0\0\0\0\0\0\0\0\t\0"..., 64) = 64 read(5, "\0\0\0\0\0\0\0\0\0\0\0\0\t\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 dup(9) = 10 rt_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 write(4, "&\0\0\0\0\0\0\0\0\0\0\0004\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 read(5, "\0\0\0\0\0\0\0\0\t\0\0\0\1\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 dup(9) = 11 fstat64(11, {st_mode=S_IFREG|0755, st_size=5054496, ...}) = 0 fstatfs(11, {f_type="MSDOS_SUPER_MAGIC", f_bsize=4096, f_blocks=358741, f_bfree=57991, f_files=0, f_ffree=0, f_namelen=260}) = 0 close(11) = 0 mmap2(NULL, 5124096, PROT_READ, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40860000 munmap(0x40d33000, 65536) = 0 mmap2(0x40860000, 5058560, PROT_READ, MAP_PRIVATE|MAP_FIXED, 10, 0) = 0x40860000 rt_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 write(4, "\27\0\0\0\0\0\0\0\0\0\0\0\377\377\377\3774\0\0\0\377\377"..., 64) = 64 read(5, "\0\0\0\0\0\0\0\0008\0\0\0\377\377\377\377\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 close(10) = 0 rt_sigprocmask(SIG_BLOCK, [HUP INT USR1 USR2 ALRM IO], [RTMIN], 8) = 0 write(4, "\25\0\0\0\0\0\0\0\0\0\0\0004\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 read(5, "\0\0\0\0\0\0\0\0\t\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"..., 64) = 64 rt_sigprocmask(SIG_SETMASK, [RTMIN], NULL, 8) = 0 close(9) = 0 getpid() = 5154 open("/tmp/reg14220000.tmp", O_WRONLY|O_CREAT|O_EXCL|O_LARGEFILE, 0666) = 9 fcntl64(9, F_GETFL) = 0x8001 (flags O_WRONLY|O_LARGEFILE) fstat64(9, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0 mmap2(NULL, 131072, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40d33000 _llseek(9, 0, [0], SEEK_CUR) = 0 brk(0) = 0x7800a000 brk(0x7800b000) = 0x7800b000 write(9, "WINE REGISTRY Version 2\n[MIME\\\\D"..., 131072) = 131072 write(9, "trolSet\\\\Control\\\\DeviceClasses\\"..., 131072) = 131072 write(9, "&PID_0095&INST_0\"\n\"RefCount\"=dwo"..., 131072) = 131072 write(9, "ixer f\\xfcr Aureon 5.1 Fun\"\n\n[Sy"..., 131072) = 131072 write(9, "0,00,00,00,00,00,00,00,00,00,00,"..., 131072) = 131072 write(9, "stem\\\\CurrentControlSet\\\\Service"..., 131072) = 131072 brk(0) = 0x7800b000 brk(0x7800c000) = 0x7800c000 brk(0) = 0x7800c000 brk(0x7800d000) = 0x7800d000 brk(0) = 0x7800d000 brk(0x7800e000) = 0x7800e000 brk(0) = 0x7800e000 brk(0x7800f000) = 0x7800f000 brk(0) = 0x7800f000 brk(0x78010000) = 0x78010000 --- SIGSEGV (Segmentation fault) @ 0 (0) --- +++ killed by SIGSEGV +++ Thanks, Flo

1 0

segfault fault at startup
by terra＠gnome.org May 31, 2004

May 31, 2004

A newly built wine segfaults for me. No args: welinder(a)anemone:~/wine> ./wine Segmentation fault No core. gdb just reports: welinder(a)anemone:~/wine> LD_LIBRARY_PATH=/home/welinder/wine/libs/wine gdb loader/wine ... (gdb) r Starting program: /home/welinder/wine/loader/wine [New Thread 16384 (LWP 28622)] [New Thread 32769 (LWP 28624)] [New Thread 16386 (LWP 28625)] Cannot find user-level thread for LWP 28622: generic error This is with a suse 9.0 system. Ideas? Morten

4 4

Re: segfault fault at startup
by terra＠gnome.org May 30, 2004

May 30, 2004

No change with 'set follow-fork-mode child', or even '... ask'. With strace, I get... strace -i $WINELOADER [...] [40113d71] close(9) = 0 [400f0c37] getpid() = 30133 [40113ce3] open("/tmp/reg75b50000.tmp", O_WRONLY|O_CREAT|O_EXCL|O_LARGEFILE, 0666) = 9 [401140b8] fcntl64(9, F_GETFL) = 0x8001 (flags O_WRONLY|O_LARGEFILE) [40112dcb] fstat64(9, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0 [4011e6ed] old_mmap(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0x40032000 [40121ce7] _llseek(9, 0, [0], SEEK_CUR) = 0 [400b7fdc] --- SIGSEGV (Segmentation fault) @ 0 (0) --- upeek: ptrace(PTRACE_PEEKUSER,30133,48,0): No such process [????????] +++ killed by SIGSEGV +++ Hmm... welinder(a)anemone:~/wine> printenv | grep WINE WINESERVER=/home/welinder/wine/server/wineserver WINELOADER=/home/welinder/wine/loader/wine WINEDLLPATH=/home/welinder/wine/dlls:/home/welinder/wine/programs:/home/welinder/wine/dlls:/home/welinder/wine/programs welinder(a)anemone:~/wine> gdb $WINELOADER GNU gdb 5.3.92 Copyright 2003 Free Software Foundation, Inc. GDB is free software, covered by the GNU General Public License, and you are welcome to change it and/or distribute copies of it under certain conditions. Type "show copying" to see the conditions. There is absolutely no warranty for GDB. Type "show warranty" for details. This GDB was configured as "i586-suse-linux"... (gdb) tb main Breakpoint 1 at 0x8048857: file glibc.c, line 76. (gdb) r Starting program: /home/welinder/wine/loader/wine [New Thread 16384 (LWP 30150)] [Switching to Thread 16384 (LWP 30150)] main (argc=1, argv=0xbffff414) at glibc.c:76 76 const char *loader = getenv( "WINELOADER" ); (gdb) info symbol 0x400b7fdc nl_langinfo + 60 in section .text (I'm not sure I believe that) Morten

1 0

Help with winedbg
by Shachar Shemesh May 30, 2004

May 30, 2004

Hi, I'm trying to get a Visual C compiled program to run on Wine. I have the sources and full debug info for the PE part of the program. Obviously, however, it was not compiled on the same machine on which it is running. To make things even more interesting, I would rather use some GUI frontend for the debugger. As far as I can tell, I have two options to go about this: 1. Use Visual Studio's remote debugging 2. Use winedbg as a gdb backend for ddd Option 1 fails miserably. I suspect the project is just too complex for Visual Studio to handle over the wire. When I try to do that, the Visual Studio front end (the one running on a Windows machine) crashes. This leaves me with just option 2. When I try to run it like that (winedbg --gdb --no-start), winedbg starts ok. When I hook ddd to it, however, winedbg complains about a missing source file (one that is not in the Windows source, btw). In the mean while, ddd claims to time out on the connection, and closes down. Am I missing something? Shachar -- Shachar Shemesh Lingnu Open Source Consulting http://www.lingnu.com/

2 7

Tainted code in User32?
by Steven Edwards May 30, 2004

May 30, 2004

Hello this line: "However, disassembling NT implementation (WIN32K.SYS) reveals that........." in http://source.winehq.org/source/windows/win.c#L845 was introduced in to winehq by the following patch: http://cvs.winehq.com/cvsweb/wine/windows/win.c.diff?r1=1.61&r2=1.62 Does this not violate a clean rooming the implementation? The ReactOS code is a derived work of the Wine code in the case and if soon then we have to remove it. Thanks Steven __________________________________ Do you Yahoo!? Friends. Fun. Try the all-new Yahoo! Messenger. http://messenger.yahoo.com/

4 3

Re: EnumDateFormats patch
by Alexandre Julliard May 29, 2004

May 29, 2004

Wililam <computerman413(a)ezrs.com> writes: > Here is the ChangeLog entry: > > * dlls/kernel/lcformat.c > William Lahti <computerman413(a)ezrs.com> > > - Implemented the EnumDateFormatsW function and vastly improved upon the > EnumDateFormatsA function. This stuff clearly needs to be stored as data in the nls resource files, not in the code. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: [Fwd: winetest: copy and submit manually]
by Dimitrie O. Paun May 29, 2004

May 29, 2004

On Fri, May 28, 2004 at 07:42:54AM -0400, Robert Reif wrote: > It appears that the locale.c test on line 1866 and 1867 fail a lot > (0xffff?). Can you try to see why this happens on your box? > File too big (# > 1 MB), copy and submit manually > > where # is around 8 meg. Yeah, there's no point in submitting such a large file, it will be rejected by the server. The message about submitting manually is confusing, it should be 'this is a problem, please report (and investigate, if at all possible :))' -- Dimi.

2 1

Re: Print thread ID in 16-bit snoop traces in the noargs case
by Rein Klazes May 29, 2004

May 29, 2004

On Fri May 28 23:42:14 BST 2004, you wrote: > Mike Hearn <mh(a)codeweavers.com> > Print thread ID in 16-bit snoop traces in the noargs case Too late: http://cvs.winehq.org/patch.py?id=12478 Rein. -- Rein Klazes rklazes(a)xs4all.nl

2 1

Upcoming breakage warning
by Mike Hearn May 29, 2004

May 29, 2004

Hi, On Fedora Core 2, the kernel is compiled with the 4G/4G VM split option enabled, which can prevent us from performing a correct emulation. Win32 apps are apparently built with the assumption that they will not be allocated addresses beyond the 3G boundary. The solution therefore is to perform a similar trick to the preloader and reserve the address space before the emulation starts so mmap will never use addresses in that area. This probably won't be combined with the preloader itself though, as Alexandre says it can be done after Wine gets control (what about DSOs mapped >3gig?) and so it can be done more portably. Up until now this issue has only affected Solaris and therefore nobody fixed it. There is currently no known workaround for this problem short of recompiling your kernel. The error looks like this: wine: unable to create process heap Alexandre diagnosed this problem so I'm sure he'll correct me if this email contains any mistakes. thanks -mike

4 5

Re: wine/ loader/main.c loader/glibc.c loader/Make ...
by Mike Hearn May 29, 2004

May 29, 2004

On Fri, 28 May 2004 15:59:23 -0500, Alexandre Julliard wrote: > Log message: > Initial version of the Wine preloader, used to reserve memory > areas at startup. Based on the work of Mike McCormack. For those who haven't been following the execshield/prelink saga, this commit is the result of the various prototypes and patches that were posted to wine-devel since we got back from WineConf. The result of it should be that we no longer need to switch off execshield nor prelink in order to run Wine reliably, also the huge BSS array hack is also no longer needed. Binaries of every size should work. Combined with the patch Alexandre checked in a few days ago to do an iterative binary search of the upper 2gig of address space, we should now also be functional when the 4G/4G VM split is enabled. It could use some testing! Please update to latest CVS and run wine as you would normally. If you are on a prelink/execshield enabled system, make sure they are both switched on and that prelink is using exec-shield mode (should be by default). If you observe any odd breakage please let us know. keep on truckin' -mike

3 3

Re: Fix CoMarshalInterThreadInterfaceInStream
by Mike Hearn May 28, 2004

May 28, 2004

( again wishing pan respected reply-to fields ... ) On Sat, 29 May 2004 00:33:24 +0100, Robert Shearman wrote: > This patch fixes CoMarshalInterThreadInterfaceInStream and > CoGetInterfaceAndReleaseStream to match what MSDN says - calling > CoMarshalInterface and CoUnmarshalInterface respectively. Without this > patch a particular InstallShield installer would just hang. With this > patch it gets a bit further (to an unimplemented function). Calling CoMarshalInterface in this case doesn't actually do much, but if it helps InstallShield OK. If you are interested in this stuff I'd say trying to merge the TransGaming patch would be a good start. This includes a lot of what you've done lately such as proper apartments, proper inter-thread marshalling and so on. Otherwise the codebase drift is going to get even worse :( thanks -mike

1 0

Re: EnableWindow
by Alexandre Julliard May 28, 2004

May 28, 2004

Ulrich Czekalla <ulrich.czekalla(a)utoronto.ca> writes: > ChangeLog: > Ulrich Czekalla <ulrich(a)codeweavers.com> > EnableWindow should not remove the focus of child windows Thanks for providing a test, but could you please merge it with the existing win.c and/or msg.c tests? We can't create a new file for every API being tested. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Unsupported feature in SHFileOperationW
by Phillip.Ezolt＠hp.com May 28, 2004

May 28, 2004

Hi all, Here is patch for the SHFileOperationW problem. (Which I originally sent out in January. ) It isn't perfect, and it fixes the problem for my application. (which happens to be the same as Lionel's) According to Lionel the fix is better than what is there today, but it falls short of windows behavior in the following manner (for the record): > But it's not 100 % correct as, for example, it does not work if one needs to > create more than one level of directories (which is working fine in Windows, > I just checked it). Moreover, there should be a message box (if not disabled > like in RollerCoasterTycoon) asking if the users wants or not to create the > directory (I checked this on Windows too). --Phil Hewlett-Packard: High Performance Technical Computing/Visualization --------------------------------------------------------------------------- Phillip.Ezolt(a)hp.com Performance/Development ---------- Forwarded message ---------- Date: Fri, 30 Jan 2004 12:54:48 -0500 (EST) From: Phillip Ezolt <phillip.ezolt(a)hp.com> To: wine-patches(a)winehq.org Subject: SHFileOperationW: Create subdir on Copy Hi All, Here's a small fix. Comments/questions are welcomed. Summary ------- This patch allows SHFileOperationW to automatically create subdirectories if they don't exists. (The Microsoft documentation states that this is correct.) This patch fixes bug #1954. I have verified that it works with Today's CVS. (January 30,2004) I ran into this problem with the Rollercoaster Tycoon demo installer. It can successfully copy files to this newly created directory, but coping into any (currently non-existant) directories below it fail. According to MS: (http://msdn.microsoft.com/library/default.asp?url=/library/en-us/shellcc/pl…) " Copy and Move operations can specify destination directories that do not exist and the system will attempt to create them. " (See footnote #1 for more info) For example, this succeeds: trace:shell:SHNotifyCopyFileW (L"E:\\WZS21db.tmp\\English\\rct.exe" L"C:\\Program Files\\Hasbro Interactive\\RollerCoaster Tycoon Demo\\rct.exe" ) But this fails: trace:shell:SHNotifyCopyFileW (L"E:\\WZS21db.tmp\\Saved Games\\001" L"C:\\Program Files\\Hasbro Interactive\\RollerCoaster Tycoon Demo\\Saved Games\\001" ) The directory "Saved Games" should be created before "001" is copied into it. Instead, the call fails. --Phil Hewlett-Packard: High Performance Technical Computing/Visualization --------------------------------------------------------------------------- Phillip.Ezolt(a)hp.com Performance/Development

1 0

Re: wine/ windows/win.c server/user.c
by Uwe Bonnes May 28, 2004

May 28, 2004

>>>>> "Alexandre" == Alexandre Julliard <julliard(a)winehq.org> writes: Alexandre> ChangeSet ID: 12496 CVSROOT: /opt/cvs-commit Module name: Alexandre> wine Changes by: julliard(a)wine.codeweavers.com 2004/05/28 Alexandre> 14:35:37 Alexandre> Modified files: windows : win.c server : user.c Alexandre> Log message: Only allocate even-numbered window handles, MS Alexandre> Project depends on that (found by Dmitry Timoshkov). Alexandre> Patch: http://cvs.winehq.org/patch.py?id=12496 When Opera sued MS for having non compatible Webpages and won, we should consider the same for this stupid thing :-) -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ----------

1 0

Re: Add documentation on the address space layout in Wine
by Mike Hearn May 28, 2004

May 28, 2004

On Fri, 28 May 2004 09:10:22 +0200, Pouech Eric DMI AEI CAEN wrote: > Hi Mike > I've started to revisit the architecture.sgml file, and started to write > also about this matters (and a few others). For coherency reasons, if > you don't mind, I'll reintegrate your bits in the architecture.sgml file > (what you wrote covers also process image loading...). Sure, go ahead and do what you want with it. Do you know when you'll have something to submit? Also what else have you written about? I'm keen on getting as much of this low level technical detail down in writing so developers new to Wine stand a chance of understanding the scarier bits of the code. I was thinking of tackling DCOM next. thanks -mike

2 1

Re: Unsupported feature in SHFileOperationW
by fenix＠club-internet.fr May 28, 2004

May 28, 2004

LOL I have seen (and sent a patch) a problem on FindExecutable who seems equivalent: trying to create a Link to an executable while executable not already present. Installers are funny :) Regards, Raphael ----Message d'origine---- >De: Lionel Ulmer <lionel.ulmer(a)free.fr> >A: wine-devel(a)winehq.org >Sujet: Unsupported feature in SHFileOperationW >Date: Fri, 28 May 2004 18:02:50 +0200 > >Hi all, > >A game demo installer uses SHFileOperationA to copy files over... And it >expects also the destination directories to be created when the copy is >done. > >Basically, it gives a list of fully qualified source files and their >corresponding fully qualified destination files and cann the >SHFileOperationA function with the 'FO_COPY' function and >'FOF_MULTIDESTFILES FOF_NOCONFIRMATION FOF_NOCONFIRMMKDIR' flags. > >This seems to not be handled properly in the current Wine tree as it fails >as soon as it needs to create the directory on the destination files. I >tested this on Windows and it works just fine (with nice MessageBoxes if one >ommits the 'NOCONFIRMMKDIR' flag :-) ). > >Is anyone (Martin, Ge ?) currently working on Shell32 who has an updated >tree with fixes in the Shell file operations ? > >If not, I will try to fix it on my own (be aware though that as I find the >style of coding used in this file particularely hard to decipher, I may do a >'NIH' on it and rewrite most of it to fit my tastes :-) ). > > Lionel > >-- > Lionel Ulmer - http://www.bbrox.org/ > >

1 0

msvcmaker - fix building makefiles on Windows
by Steven Edwards May 28, 2004

May 28, 2004

Hello, I am not 100% sure if this is correct as I dont have a Windows box handy and dont know perl but this has been a known bug where msvcmaker did not properly build Makefiles when using Perl on Windows. A anonymous slashdot reader sent me this patch to fix this oversite. He saif to either attribute the changelog entry to Fred Patrik or to me. If someone has a Windows box handy could you test msvcmaker with this patch and see if it generates the makefiles properly now? Thanks Steven Changelog: Fix generation of MS_VC Makfiles on Windows __________________________________ Do you Yahoo!? Friends. Fun. Try the all-new Yahoo! Messenger. http://messenger.yahoo.com/

1 0

Unsupported feature in SHFileOperationW
by Lionel Ulmer May 28, 2004

May 28, 2004

Hi all, A game demo installer uses SHFileOperationA to copy files over... And it expects also the destination directories to be created when the copy is done. Basically, it gives a list of fully qualified source files and their corresponding fully qualified destination files and cann the SHFileOperationA function with the 'FO_COPY' function and 'FOF_MULTIDESTFILES FOF_NOCONFIRMATION FOF_NOCONFIRMMKDIR' flags. This seems to not be handled properly in the current Wine tree as it fails as soon as it needs to create the directory on the destination files. I tested this on Windows and it works just fine (with nice MessageBoxes if one ommits the 'NOCONFIRMMKDIR' flag :-) ). Is anyone (Martin, Ge ?) currently working on Shell32 who has an updated tree with fixes in the Shell file operations ? If not, I will try to fix it on my own (be aware though that as I find the style of coding used in this file particularely hard to decipher, I may do a 'NIH' on it and rewrite most of it to fit my tastes :-) ). Lionel -- Lionel Ulmer - http://www.bbrox.org/

1 0

focused code review
by Glen Kaukola May 28, 2004

May 28, 2004

Hi, So one project suggested on the website is to do a focused code review. In particular it talks about getting rid of inefficient uses of strcat. But that leaves me wondering what exactly constitutes an inefficient use of strcat? If anyone could provide an example I'd appreciate it. Thanks. Glen

2 1

Re: Default to 96dpi for the no config file case
by Rolf Kalbermatter May 28, 2004

May 28, 2004

>If you mean via DDC then I guess this should work in most places, but for >whatever reason it does not work here :( In particular, xdpyinfo says: > >screen #0: > dimensions: 1024x768 pixels (321x241 millimeters) > resolution: 81x81 dots per inch > >So it appears to know the screen size (I measured with a ruler and >321x241 is right), but it gets the DPI too small anyway. At 81 dpi the >text is hardly readable. But: 1024/321*25.4 = 81.026 768/241*25.4 = 80.943 Seems to me X has it very right. That Windows usually doesn't care about real resolutions and instead fudges with it (consider the Large Font setting in the Display Control Panel) is probably another issue. Or the metrics this function requires are not dots per inch but something else. Rolf Kalbermatter

1 0

Re: Default to 96dpi for the no config file case
by Rein Klazes May 28, 2004

May 28, 2004

On Thu, 27 May 2004 22:18:20 +0100, you wrote: > wineprefixcreate doesn't create a config file. This means we use a (wrong) > formula to decide what the DPI is supposed to be. Why do you mean (wrong) ? Do you not get the dpi values used by X? Or is it that is uses the X values but those are wrong? > It's wrong because > nobody tells the system how big their monitor is these days so it always > picks a DPI that is too small. Here X gets it from the monitor at start-up (via the logic of the video card). I did not realize that my situation is not normal. Somehow, I think setting a proper dpi should ideally be done in X. But in practice ... I agree with you that 95 dpi is about the minimum that is being used today. But I would keep this the minimum: if the dpi is set bigger then it is probably intentional. Either monitor magic or from a careful system administration. Rein. -- Rein Klazes rklazes(a)xs4all.nl

4 4

Re: Add documentation on the address space layout in Wine
by Pouech Eric DMI AEI CAEN May 28, 2004

May 28, 2004

Hi Mike I've started to revisit the architecture.sgml file, and started to write also about this matters (and a few others). For coherency reasons, if you don't mind, I'll reintegrate your bits in the architecture.sgml file (what you wrote covers also process image loading...). A+ > Message du 28/05/04 01:35 > De : "Mike Hearn" > A : wine-patches(a)winehq.com > Copie à : > Objet : Add documentation on the address space layout in Wine > Mike Hearn > Add documentation on the address space layout in Wine > > Generated from: > * mike(a)navi.cx--2004/wine--mainline--0.9--patch-25 > > --- /dev/null 2003-09-15 14:40:47.000000000 +0100 > +++ documentation/address-space.sgml 2004-05-28 00:26:06.000000000 +0100 > @@ -0,0 +1,175 @@ > + > + > + > + > + Every Win32 process in Wine has its own dedicated native process on the host system, and > + therefore its own address space. This section explores the layout of the Windows address space > + and how it is emulated. > + > + > + > + Firstly, a quick recap of how virtual memory works. Physical memory in RAM chips is split > + into frames, and the memory that each process sees is split > + into pages. Each process has its own 4 gigabytes of address space (4gig > + being the maximum space addressable with a 32 bit pointer). Pages can be mapped or unmapped: > + attempts to access an unmapped page cause an EXCEPTION_ACCESS_VIOLATION which has the > + easily recognizable code of 0xC0000005. Any page can be mapped to any frame, therefore you can > + have multiple addresses which actually "contain" the same memory. Pages can also be mapped to > + things like files or swap space, in which case accessing that page will cause a disk access to > + read the contents into a free frame. > + > + > + > + > + > + > + When a Win32 process starts, it does not have a clear address space to use as it pleases. Many pages > + are already mapped by the operating system. In particular, the EXE file itself and any DLLs it > + needs are mapped into memory, and space has been reserved for the stack and a couple of heaps > + (zones used to allocate memory to the app from). Some of these things need to be at a fixed > + address, and others can be placed anywhere. > + > + > + > + The EXE file itself is almost always mapped at address 0x400000 and up: indeed, most EXEs have > + their relocation records stripped which means they must be loaded at their base address and > + cannot be loaded at any other address. > + > + > + > + DLLs are internally much the same as EXE files but they have relocation records, which means > + that they can be mapped at any address in the address space. Remember we are not dealing with > + physical memory here, but rather virtual memory which is different for each > + process. Therefore OLEAUT32.DLL may be loaded at one address in one process, and a totally > + different one in another. Ensuring all the functions loaded into memory can find each other > + is the job of the Windows dynamic linker, which is a part of NTDLL. > + > + > + > + So, we have the EXE and its DLLs mapped into memory. Two other very important regions also > + exist: the stack and the process heap. The process heap is simply the equivalent of the libc > + malloc arena on UNIX: it's a region of memory managed by the OS which malloc/HeapAlloc > + partitions and hands out to the application. Windows applications can create several heaps but > + the process heap always exists. It's created as part of process initialization in > + dlls/ntdll/thread.c:thread_init(). > + > + > + > + There is another heap created as part of process startup, the so-called shared or system > + heap. This is an undocumented service that exists only on Windows 9x: it is implemented in > + Wine so native win9x DLLs can be used. The shared heap is unusual in that anything allocated > + from it will be visible in every other process. This heap is always created at the > + SYSTEM_HEAP_BASE address or 0x65430000 and defaults to a megabyte in size. > + > + > + > + So far we've assumed the entire 4 gigs of address space is available for the application. In > + fact that's not so: only the lower 2 gigs are available, the upper 2 gigs are on Windows NT > + used by the operating system and hold the kernel (from 0x80000000). Why is the kernel mapped > + into every address space? Mostly for performance: while it's possible to give the kernel its > + own address space too - this is what Ingo Molnars 4G/4G VM split patch does for Linux - it > + requires that every system call into the kernel switches address space. As that is a fairly > + expensive operation (requires flushing the translation lookaside buffers etc) and syscalls are > + made frequently it's best avoided by keeping the kernel mapped at a constant position in every > + processes address space. > + > + > + > + On Windows 9x, in fact only the upper gigabyte (0xC0000000 and up) is used by the kernel, the > + region from 2 to 3 gigs is a shared area used for loading system DLLs and for file > + mappings. The bottom 2 gigs on both NT and 9x are available for the programs memory allocation > + and stack. > + > + > + > + There are a few other magic locations. The bottom 64k of memory is deliberately left unmapped > + to catch null pointer dereferences. The region from 64k to 4mb are reserved for DOS > + compatibility and contain various DOS data structures. Finally, the address space also > + contains mappings for the Wine binary itself, any native libaries Wine is using, the glibc > + malloc arena and so on. > + > + > + > + > + > + > + > + > + Up until about the start of 2004, the Linux address space very much resembled the Windows 9x > + layout: the kernel sat in the top gigabyte, the bottom pages were unmapped to catch null > + pointer dereferences, and the rest was free. The kernels mmap algorithm was predictable: it > + would start by mapping files at low addresses and work up from there. > + > + > + > + The development of a series of new low level patches violated many of these assumptions, and > + resulted in Wine needing to force the Win32 address space layout upon the system. This > + section looks at why and how this is done. > + > + > + > + The exec-shield patch increases security by randomizing the kernels mmap algorithms. Rather > + than consistently choosing the same addresses given the same sequence of requests, the kernel > + will now choose randomized addresses. Because the Linux dynamic linker (ld-linux.so.2) loads > + DSOs into memory by using mmap, this means that DSOs are no longer loaded at predictable > + addresses, so making it harder to attack software by using buffer overflows. It also attempts > + to relocate certain binaries into a special low area of memory known as the ASCII armor so > + making it harder to jump into them when using string based attacks. > + > + > + > + Prelink is a technology that enhances startup times by precalculating ELF global offset > + tables then saving the results inside the native binaries themselves. By grid fitting each > + DSO into the address space, the dynamic linker does not have to perform as many relocations > + so allowing applications that heavily rely on dynamic linkage to be loaded into memory much > + quicker. Complex C++ applications such as Mozilla, OpenOffice and KDE can especially benefit > + from this technique. > + > + > + > + The 4G VM split patch was developed by Ingo Molnar. It gives the Linux kernel its own address > + space, thereby allowing processes to access the maximum addressable amount of memory on a > + 32-bit machine: 4 gigabytes. It allows people with lots of RAM to fully utilise that in any > + given process at the cost of performance: as mentioned previously the reason behind giving > + the kernel a part of each processes address space was to avoid the overhead of switching on > + each syscall. > + > + > + > + Each of these changes alter the address space in a way incompatible with Windows. Prelink and > + exec-shield mean that the libraries Wine uses can be placed at any point in the address > + space: typically this meant that a library was sitting in the region that the EXE you wanted > + to run had to be loaded (remember that unlike DLLs, EXE files cannot be moved around in > + memory). The 4G VM split means that programs could receive pointers to the top gigabyte of > + address space which some are not prepared for (they may store extra information in the high > + bits of a pointer, for instance). In particular, in combination with exec-shield this one is > + especially deadly as it's possible the process heap could be allocated beyond > + ADDRESS_SPACE_LIMIT which causes Wine initialization to fail. > + > + > + > + The solution to these problems is for Wine to reserve particular parts of the address space > + so that areas that we don't want the system to use will be avoided. We later on > + (re/de)allocate those areas as needed. One problem is that some of these mappings are put in > + place automatically by the dynamic linker: for instance any libraries that Wine > + is linked to (like libc, libwine, libpthread etc) will be mapped into memory before Wine even > + gets control. In order to solve that, Wine overrides the default ELF initialization sequence > + at a low level and reserves the needed areas by using direct syscalls into the kernel (ie > + without linking against any other code to do it) before restarting the standard > + initialization and letting the dynamic linker continue. This is referred to as the > + preloader and is found in ld-winepreload.so (loader/preloader.c) > + > + > + > + Once the usual ELF boot sequence has been completed, some native libraries may well have been > + mapped above the 3gig limit: however, this doesn't matter as 3G is a Windows limit, not a > + Linux limit. We still have to prevent the system from allocating anything else above there > + (like the heap or other DLLs) though so Wine performs a binary search over the upper gig of > + address space in order to iteratively fill in the holes with MAP_NORESERVE mappings so the > + address space is allocated but the memory to actually back it is not. This code can be found > + in libs/wine/mmap.c:reserve_area. > + > + > + > + > + > --- documentation/wine-devel.sgml > +++ documentation/wine-devel.sgml > @@ -13,6 +13,7 @@ > > > > + > > > > @@ -138,6 +139,7 @@ > &implementation; > &porting; > &consoles; > + &address-space; > &cvs-regression; > > > > >

1 0

[Fwd: winetest: copy and submit manually]
by Robert Reif May 27, 2004

May 27, 2004

oops, wrong list.

1 0

RE: FW: Edit Paste operation does not behave same as MS Windows
by Krishna Murthy May 27, 2004

May 27, 2004

This means it is sufficient to return FALSE in EDIT_MakeFit if the honor limit is not set and size exceeds the buffer limit? static BOOL EDIT_MakeFit(EDITSTATE *es, UINT size, BOOL honor_limit) { HLOCAL hNew32W; if ((!honor_limit) && (es->buffer_limit > 0) && (size > es->buffer_limit)) { EDIT_NOTIFY_PARENT(es, EN_MAXTEXT, "EN_MAXTEXT"); return FALSE; } ... } I will verify with this change and if works in all condition, will submit the patch. Thanks for your help, Krishna -----Original Message----- From: Alexandre Julliard [mailto:julliard(a)winehq.org] Sent: Thursday, May 27, 2004 5:12 PM To: Krishna Murthy Cc: wine-devel(a)winehq.org Subject: Re: FW: Edit Paste operation does not behave same as MS Windows Krishna Murthy <Krishna.Murthy(a)guptaworldwide.com> writes: > Actually the logic truncates the data which exceeded the buffer. If > the honor limit is set and the size exceeded, the EDIT_MakeFit will return TRUE. > The following code after the EDIT_MakeFit will try to take limit to > the buffer size. > > + /* Only replace or insert that many characters that can fit into the > + edit control. Do this only if honor_limit is TRUE */ > + if(honor_limit && strl > es->buffer_limit) > + strl = es->buffer_limit - (tl - (e-s)); That only takes into account the limit, not the current buffer size. It's very possible that you need to grow the buffer even when the limit is not exceeded. Also you only check the length of the inserted string against the limit, you should check the total resulting length instead. -- Alexandre Julliard julliard(a)winehq.org

1 0

RE: FW: Edit Paste operation does not behave same as MS Windows
by Krishna Murthy May 27, 2004

May 27, 2004

Actually the logic truncates the data which exceeded the buffer. If the honor limit is set and the size exceeded, the EDIT_MakeFit will return TRUE. The following code after the EDIT_MakeFit will try to take limit to the buffer size. + /* Only replace or insert that many characters that can fit into the + edit control. Do this only if honor_limit is TRUE */ + if(honor_limit && strl > es->buffer_limit) + strl = es->buffer_limit - (tl - (e-s)); Krishna -----Original Message----- From: Alexandre Julliard [mailto:julliard(a)winehq.org] Sent: Thursday, May 27, 2004 2:30 PM To: Krishna Murthy Cc: wine-devel(a)winehq.org Subject: Re: FW: Edit Paste operation does not behave same as MS Windows Krishna Murthy <Krishna.Murthy(a)guptaworldwide.com> writes: > - Change log Entry: > EDIT_MakeFit should return TRUE when the honor_limit is set and paste > buffer exceeds the limit. Also EDIT_EM_ReplaceSel should compute size > which could fit in the edit control by truncating the extra characters > to behave similar to MS Windows paste operation. You are bypassing the reallocation of the buffer when the limit is exceeded, but still adding the data to the buffer. I don't think this will work right. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: Crash and corrupted registry
by Alexandre Julliard May 27, 2004

May 27, 2004

Christian Costa <titan.costa(a)wanadoo.fr> writes: > This seems to be more a filesytem issue then because the file is in a > really bad shape. > I use reiserfs on a Mandrake 2.4.21-0.13mdk kernel. > It seems to happend on files that are open when the crash (freeze + > reboot) occurs. Yes, it very much looks like a filesystem issue, I prefer that <g> If you feel like trying to reproduce it, maybe you could try to add an fsync() in the saving code and see if this makes a difference. -- Alexandre Julliard julliard(a)winehq.org

3 3

Re: FW: Edit Paste operation does not behave same as MS Windows
by Alexandre Julliard May 27, 2004

May 27, 2004

Krishna Murthy <Krishna.Murthy(a)guptaworldwide.com> writes: > - Change log Entry: > EDIT_MakeFit should return TRUE when the honor_limit is set and paste buffer > exceeds the limit. Also EDIT_EM_ReplaceSel should compute size which could > fit in the edit control by truncating the extra characters to behave similar > to MS Windows paste operation. You are bypassing the reallocation of the buffer when the limit is exceeded, but still adding the data to the buffer. I don't think this will work right. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Repaint the whole nonclient window area on SetWindowPos(SWP_FRAMECHANGED)
by Alexandre Julliard May 27, 2004

May 27, 2004

Dmitry Timoshkov <dmitry(a)baikal.ru> writes: > SetWindowPos(SWP_FRAMECHANGED) causes the whole nonclient window area > to be repainted. > > Also, according to the tests Windows always repaints windows before > SetWindowPos has finished. > > This patch reduces number of message test failures from 56 to 47. The message order is wrong because we rely on X expose events to repaint the window. This is one of the reasons we need to go back to not using X windows for every window; but until that is done, your fix is going to cause every window to be repainted twice. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Add a test for accessing classes from another thread
by Alexandre Julliard May 27, 2004

May 27, 2004

Dmitry Timoshkov <dmitry(a)baikal.ru> writes: > somebody on the ReactOS mailing list mentioned that Windows stores class > information in the per thread list, not a per process or per module list. > That would effecively meant that accessing classes or creating windows > from other threads would fail. > > So, I decided to write a test for that. The test passes under Wine. Not for me: class.c:258: Test failed: GetLastError() should be set to ERROR_CLASS_HAS_WINDOWS class.c:258: Test failed: GetLastError() should be set to ERROR_CLASS_HAS_WINDOWS -- Alexandre Julliard julliard(a)winehq.org

1 0

atoi and undefined usage
by andrejoh＠c2i.net May 27, 2004

May 27, 2004

The atoi() function has undefined usage when trying to convert numbers outside the range of int. When parsing numbers >= 2^31, GNU libc seems to truncate the number to 2^31-1; however, with MSVC, such a number would wrap, and when cast to an unsigned integer, numbers within an unsigned int are preserved. Should Wine care about such undefined behaviour? Currently, it seems Wine uses the GNU libc implementation, giving a different result than when running in Windows. (In my case, an IP address was given as a string, and ended up as "127.255.255.255" when running in Wine. I've now fixed that specific bug in the source code, but with other programs, that might not be an option...) -- Try Anarchy Online http://www.anarchy-online.com/content/downloads/tryout/ Netiquette guidelines ftp://ftp.rfc-editor.org/in-notes/rfc1855.txt How to quote properly http://www.i-hate-computers.demon.co.uk/ Hvordan sitere riktig http://home.online.no/~vidaandr/news/OBSquoting.html

2 1

Re: test for GetAcceptLanguagesA
by Martin Fuchs May 27, 2004

May 27, 2004

On 27.05.2004 04:09:34 Alexandre Julliard wrote: > Stefan Leichter <Stefan.Leichter(a)camLine.com> writes: > > before removing the acsii calls from GetAcceptLanguagesW, i like to see > the > > results of the attached tests. This will hopefuly save me some work when > > rewriting the function. The tests passes on my Win2k system. > But they fail on Wine, so I can't apply the patch. I've seen some test cases wrapped in TODO sections, for example: todo_wine { SetWindowLong(hwnd1, GWL_EXSTYLE, GetWindowLong(hwnd1,GWL_EXSTYLE)|WS_EX_TOPMOST); ret = GetWindowLong(hwnd1,GWL_EXSTYLE)&WS_EX_TOPMOST? TRUE: FALSE; ok(!ret, "SetWindowExStyle(hwnd1, WS_EX_TOPMOST)\n"); } This prints a warning message about something yet to to in Wine, but the test doesn't fail. Can't you use this also for the patch in question? Regards, Martin

1 0

having problem in understanding debug channels
by zeeshaan umer May 27, 2004

May 27, 2004

hi, its zee here, i am a student and i have just came to know about wine 2-3 weeks ago. i have started studying it for exploration purpose. its quite a great effort of u people. really it is........ nways the problem i m having is that i m not clearly understanding debug channels. i actually was trying to be a part of developers of wine and i tried to implement an API but when i wrote and included it in the source it could not be compiled i got an error " invalid debug channel specification ' ZEE_DEFAULT_DEBUG_CHANNEL( );' " i de have define ZEE_DEFAULT_DEBUG_CHANNEL in debug.h but frankly i have not actually understood the complete philosphy of these debug channels that how they work and how are they defined properly. plz,.......anybody help me. i m a student so dont mind if it is really rediculous type of problem......:-) zeeshan xeshaan(a)yahoo.co.in Yahoo! India Matrimony: Find your partner online.

2 1

Re: Additional downloads for winehq
by Ivan May 27, 2004

May 27, 2004

OK, I don't see the utility of having a 3rd party download section just to list one download. But what about dcom95? We have a message in oleaut that tells the user to download dcom95, and it provides a link to sf, but we can't host dom on sf because it isn't open source, and people can download it free from m$ anyway. But Alexandre doesn't like long links, and the link to the dcom95 download page is just too long. It's http://www.microsoft.com/downloads/details.aspx?familyid=d7a4de78-81a9-4db7… and we want to keep the error log short. Ivan.

2 1

Re: Re: test for GetAcceptLanguagesA
by Pouech Eric DMI AEI CAEN May 27, 2004

May 27, 2004

> Its not a fault of the tests that they fail on wine, its a fault of > wine. For my understanding the test are created to show that either the > behaviour of wine is correct (when the test succeeds) or that something > left to do one wine (when a test fails)!? check out the todo_wine construct in winetest Alexandre point is not that you cannot add a test that currently fails in Wine. His point is that you cannot add a test that will make the test process fail. The todo_wine is done for that: 1/ it marks a test as failing on wine 2/ it will report it (but not make the test fail) 3/ it will also make the test process fail if the test succeeds (so that we can remove the todo_wine mark) A+

1 0

Re: test for GetAcceptLanguagesA
by Alexandre Julliard May 27, 2004

May 27, 2004

Stefan Leichter <Stefan.Leichter(a)camLine.com> writes: > before removing the acsii calls from GetAcceptLanguagesW, i like to see the > results of the attached tests. This will hopefuly save me some work when > rewriting the function. The tests passes on my Win2k system. But they fail on Wine, so I can't apply the patch. -- Alexandre Julliard julliard(a)winehq.org

4 3

[Janitorial]: Missing inversion im Bit manipulation
by Uwe Bonnes May 27, 2004

May 27, 2004

Changelog: wine/dlls/winmm/winmm.c: Add obvious missing inversion -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ---------- Index: wine/dlls/winmm/winmm.c =================================================================== RCS file: /home/wine/wine/dlls/winmm/winmm.c,v retrieving revision 1.33 diff -u -w -r1.33 winmm.c --- wine/dlls/winmm/winmm.c 27 Feb 2004 21:29:32 -0000 1.33 +++ wine/dlls/winmm/winmm.c 27 May 2004 07:56:13 -0000 @@ -1983,7 +1983,7 @@ lpMidiHdr = (LPMIDIHDR)msg->lParam; lpMidiHdr->lpNext = 0; lpMidiHdr->dwFlags |= MHDR_INQUEUE; - lpMidiHdr->dwFlags &= MHDR_DONE; + lpMidiHdr->dwFlags &= ~MHDR_DONE; lpMidiHdr->dwOffset = 0; break;

1 0

Re: Additional downloads for winehq
by Jeremy Newman May 26, 2004

May 26, 2004

I don't think these belong on the download page. That would confuse the users. IMO the download page should only contain links to download Wine. All this should go in a HowTo or in the docs. I know we have the 3rd party tools section there. I was opposed to that as well. We just cleaned up the download page, and you people want to clutter it up again! <g> Btw, your link to the download page in the second patch in wrong. it should be: http://www.winehq.org/download (we need to keep consistent on the .org thing) On Wed, 2004-05-26 at 18:09, Ivan wrote: > These downloads can be useful for lots of wine users. > > Ivan.

1 0

Re: [DMUSIC] IDirectMusicSegment8 loading of Wave Segments
by Rok Mandeljc May 26, 2004

May 26, 2004

> PS: > Rok, i have you have implemented IDirectMusicSegment8 interface on > IDirectMusicWaveImpl. > Why ? > What it your "sample" who permit to check this behavior ? > For testing my patch you only have, for example, to open a .wav file using > 3DAudio.exe sdk sample Hey, I'm sorry but I cannot really remember (have experienced computer crash and it will take some time for me to organise everything again). I'm not even sure if IDirectMusicSegment should be implemented for it, but i think my test was to create native IDirectMusicWave via CoCreateInstance and then try to QueryInterface it with all possible IIDs; I can't remember result, but I think it was supported (unless I made a mistake ;=) - Rok

2 1

Assume MAX_PATH length for pszDisplayName
by Uwe Bonnes May 26, 2004

May 26, 2004

Changelog: dlls/shell32/brsfolders.c: SHBrowseForFolderA LPBROWSEINFOA->pszDisplayName is assumed to have MAX_PATH legth Fixes a crash in Xilinx ISE when vreating a new project with a long path -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ---------- Index: wine/dlls/shell32/brsfolder.c =================================================================== RCS file: /home/wine/wine/dlls/shell32/brsfolder.c,v retrieving revision 1.51 diff -u -w -5 -r1.51 brsfolder.c --- wine/dlls/shell32/brsfolder.c 19 Apr 2004 20:12:14 -0000 1.51 +++ wine/dlls/shell32/brsfolder.c 26 May 2004 13:56:29 -0000 @@ -460,11 +460,12 @@ bi.hwndOwner = lpbi->hwndOwner; bi.pidlRoot = lpbi->pidlRoot; if (lpbi->pszDisplayName) { len = MultiByteToWideChar(CP_ACP, 0, lpbi->pszDisplayName, -1, NULL, 0); - bi.pszDisplayName = HeapAlloc(GetProcessHeap(), 0, len * sizeof(WCHAR)); + /*lpbi->pszDisplayName is assumed to be MAX_PATH (MSDN) */ + bi.pszDisplayName = HeapAlloc(GetProcessHeap(), 0, MAX_PATH * sizeof(WCHAR)); MultiByteToWideChar(CP_ACP, 0, lpbi->pszDisplayName, -1, bi.pszDisplayName, len); } else bi.pszDisplayName = NULL;

1 0

Type in msvcrt
by Uwe Bonnes May 26, 2004

May 26, 2004

Changelog: dlls/msvcrt/file.c: _setmode() Reset only the affected flag Make Xilinx ise not hang at startup Bye -- Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de Institut fuer Kernphysik Schlossgartenstrasse 9 64289 Darmstadt --------- Tel. 06151 162516 -------- Fax. 06151 164321 ---------- Index: wine/dlls/msvcrt/file.c =================================================================== RCS file: /home/wine/wine/dlls/msvcrt/file.c,v retrieving revision 1.67 diff -u -w -5 -r1.67 file.c --- wine/dlls/msvcrt/file.c 13 May 2004 00:23:25 -0000 1.67 +++ wine/dlls/msvcrt/file.c 26 May 2004 12:17:06 -0000 @@ -1437,11 +1448,11 @@ if (mode & (~(_O_TEXT|_O_BINARY))) FIXME("fd (%d) mode (0x%08x) unknown\n",fd,mode); if ((mode & _O_TEXT) == _O_TEXT) MSVCRT_fdesc[fd].xflag |= WX_TEXT; else - MSVCRT_fdesc[fd].xflag &= WX_TEXT; + MSVCRT_fdesc[fd].xflag &= ~WX_TEXT; return ret; } /********************************************************************* * _stati64 (MSVCRT.@)

1 0

Re: Re: [OLE] Better handling of References
by fenix＠club-internet.fr May 26, 2004

May 26, 2004

>On Wed, 26 May 2004 03:55:25 +0200, Raphael wrote: >> I'm trying to install Splinter Cell Demo (installshield). >> And i have seen: >> - one typelib was released while typeinfos (childs of typelib) are in use >> - the install "crash" (catched and returns an E_UNEXPECTED error) when it >> wants to call ITypeInfo_fnGetContainingTypeLib (who want to AddRef the >> freshly released TypeLib) > >Hi Raphael, Hi Mike >This has been a known problem for a long time now, with several attempts >to fix it, all of them have been incorrect. It seems your approach is a >new one and interesting, but I'm not sure this is how Windows does it. Well Soft References are used in many places in COM (for example D3D textures/surfaces use it on windows) >Does anybody know exactly what Windows does in this situation? It might be >worth writing a test case to find out. > >Objects in a semi-freed state seem like a slightly odd way to do things. > >> Now i have another crash later, but it seems a problem with an invoke method >> who failed :( >> If any ole/rpc expert can look at this problem i'll be very happy :) > >Well, if you could provide more detail (backtrace etc) on the crash it >would be good. I can't have a backtrace as i can't get winedbg working correctly on installshield programs :( But i'll send you the traces later (when i'll be back home). For what i have see it's a problem of ITypeInfo::Invoke when it should returns a Pointer to an IUnknown Variant. InstallShield complains about Objects References not set !! >A few comments on the patch: > >- --(This->ref); >+ /** escape decrement for soft_references */ >+ if (This->ref > 0) { >+ --(This->ref); >+ } else { >+ ERR("trying to Release again an ITypeLib2 in freing state waiting for ITypeInfo (%p) childs Releases\n", This); >+ } > >I don't remember whether this object is supposed to be thread safe or not, >but should this not be an InterlockedDecrement? Yes, but as the code of ITypeLib2/ITypeInfo don't seems thread safe at all ... (anyway i'll fix it) :) >I think the error could be better phrased as: > >ERR("Reference count on an ITypeLib2 object (%p) already zero: waiting for >ITypeInfo children to be released as well"); :) >or words to that effect. > >+ if (This->soft_ref > 0) { >+ ERR("trying to Release an ITypeLib2 while some ITypeInfo (%p) childs are actually with Refs > 1. So Dont free it now \n", This); >+ /** maybe scaning ITypeInfo list to dump ITypeInfos with Refs > 1 */ >+ return 1; >+ } > >Is it actually an error to do that? If you free a typelib while refs are >held on child typeinfos, is this a bug in the app or a legitimate thing to >do? If it's not a bug in the app we probably shouldn't use ERR here. Weel, we had problems like that on d3d8 (a long time ago) about ref counting. For me, seems more a wine bug (we have to found where missing AddRefs as to be done) than an app bug (new traces can help us to find what happened) >--- variant.c 30 Apr 2004 18:32:58 -0000 1.97 >+++ variant.c 26 May 2004 01:46:44 -0000 >@@ -605,6 +605,8 @@ > > hres = VARIANT_ValidateType(V_VT(pVarg)); > >+ if (V_VT(pVarg) == VT_EMPTY) return hres; >+ > if (SUCCEEDED(hres)) > { > if (!V_ISBYREF(pVarg)) > >This change looks OK though I'm not sure why it's necessary. Does the >Splinter Cell installer actually need this? No, i have only to no have a lot of debug traces (i have added on my tree) on Variant* Operations when clearing already empty Variants >thanks -mike > Regards, Raphael

2 1

Re: [OLE] Better handling of References
by Mike Hearn May 26, 2004

May 26, 2004

Oops, sending to wine-devel too (this is what I get for using a newsreader...) On Wed, 26 May 2004 03:55:25 +0200, Raphael wrote: > I'm trying to install Splinter Cell Demo (installshield). > And i have seen: > - one typelib was released while typeinfos (childs of typelib) are in use > - the install "crash" (catched and returns an E_UNEXPECTED error) when it > wants to call ITypeInfo_fnGetContainingTypeLib (who want to AddRef the > freshly released TypeLib) Hi Raphael, This has been a known problem for a long time now, with several attempts to fix it, all of them have been incorrect. It seems your approach is a new one and interesting, but I'm not sure this is how Windows does it. Does anybody know exactly what Windows does in this situation? It might be worth writing a test case to find out. Objects in a semi-freed state seem like a slightly odd way to do things. > Now i have another crash later, but it seems a problem with an invoke method > who failed :( > If any ole/rpc expert can look at this problem i'll be very happy :) Well, if you could provide more detail (backtrace etc) on the crash it would be good. A few comments on the patch: - --(This->ref); + /** escape decrement for soft_references */ + if (This->ref > 0) { + --(This->ref); + } else { + ERR("trying to Release again an ITypeLib2 in freing state waiting for ITypeInfo (%p) childs Releases\n", This); + } I don't remember whether this object is supposed to be thread safe or not, but should this not be an InterlockedDecrement? I think the error could be better phrased as: ERR("Reference count on an ITypeLib2 object (%p) already zero: waiting for ITypeInfo children to be released as well"); or words to that effect. + if (This->soft_ref > 0) { + ERR("trying to Release an ITypeLib2 while some ITypeInfo (%p) childs are actually with Refs > 1. So Dont free it now \n", This); + /** maybe scaning ITypeInfo list to dump ITypeInfos with Refs > 1 */ + return 1; + } Is it actually an error to do that? If you free a typelib while refs are held on child typeinfos, is this a bug in the app or a legitimate thing to do? If it's not a bug in the app we probably shouldn't use ERR here. ++(This->ref); + if (This->ref > 1) { + if (NULL != This->pTypeLib) ++(This->pTypeLib->soft_ref); + } + Ditto on thread safety. Please forgive me if this object is not free threaded. --- variant.c 30 Apr 2004 18:32:58 -0000 1.97 +++ variant.c 26 May 2004 01:46:44 -0000 @@ -605,6 +605,8 @@ hres = VARIANT_ValidateType(V_VT(pVarg)); + if (V_VT(pVarg) == VT_EMPTY) return hres; + if (SUCCEEDED(hres)) { if (!V_ISBYREF(pVarg)) This change looks OK though I'm not sure why it's necessary. Does the Splinter Cell installer actually need this? thanks -mike

1 0

CoRegisterMessageFilter
by Simon Harvey May 26, 2004

May 26, 2004

hello I was wondering if anybody was doing work on or was in the process of implementing CoRegisterMessageFilter in dlls/ole32/ole2.c ? I would like to take a stab at it but i wont if somebody else is going to do it. thanks simon Find local movie times and trailers on Yahoo! Movies. http://au.movies.yahoo.com

2 1

Bad wine today.
by Rein Klazes May 26, 2004

May 26, 2004

Hi, I wonder if I am the only one that sees this with any windows prog that I try: |$ wine sol |x11drv: Can't open display: |Please ensure that your X server is running and that $DISPLAY is set correctly. |$ (other X programs run fine). I did the usual things when in trouble: a make clean, fresh cvs update. That did not help. Debugmessages "+all", does not give any messages. Version win98 or win2k makes no difference. Reverting to the yesterdays cvs, makes wine work again. Any idea? Rein. -- Rein Klazes rklazes(a)xs4all.nl

2 2

Crash and corrupted registry
by Christian Costa May 25, 2004

May 25, 2004

Hi, This is the second time I had to reboot when by box is stuck while running a windows apps. In both case the registry had been corrupted. This is very annoying. Could we have the wineserver do a backup of registry files at startup so we can retreive the previous version if a corruption occurs? That would be very usefull. Thanks, Christian

2 2

Re: Bad wine today.
by Ivan May 25, 2004

May 25, 2004

Current CVS works for me. Ivan.

1 0

Re: additional download for winehq download page
by Ivan May 25, 2004

May 25, 2004

< And if a prog doesn't provide the dll it should have at least a readme to tell people where to get it. I can tell you out of experience that this isn't always true. What about the attached patches? Ivan.

1 0

Re: Strip comments from end of line in INI parser
by Alexandre Julliard May 25, 2004

May 25, 2004

Mike Hearn <mike(a)navi.cx> writes: > Mike Hearn <mike(a)navi.cx> > Strip comments and whitespace from end of line in PROFILE_CopyString That will make it impossible to have an ini entry containing a ';'. I really doubt that Windows does it this way; could you please provide a test case? -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: winegcc: Use $(CC) for building generated .spec.c
by Dimitrie O. Paun May 25, 2004

May 25, 2004

On Tue, May 25, 2004 at 12:56:27AM -0700, Jon Griffiths wrote: > Building and linking a C++ dll to a Winelib app currently fails using > winegcc as it will use $(CXX) to compile the .spec.c file. This means > it won't link correctly to libwine.so due to the name mangling. I > tried 2 solutions: changing winebuild to generate extern "C" {} > declarations where needed, and changing winegcc to use $(CC) instead. > Of the two, I think the latter is neater. The following patch > implements it. I thought that gcc will figure the langauge based on the extension. So g++ file.c would actually compile as a C file... But this may not work for other compilers. We could do it ourselves in compile(), choose $CC or $CXX based on the extension. However, I'm not sure it's better than your solution. In fact, it's a lot more risky. -- Dimi.

2 1

Proprietary Hardware and related stuff - Get them work with Linux
by Gianfranco Luceri May 25, 2004

May 25, 2004

Hello, I own a Matrox Mistyque with Rainbow Runner Studio and Rainbow Runner TV (not G-Series) add-on cards and there's no way of getting them all work with Linux due to the lack of documentation about a proprietary chip (MGA-something-I-cant-remember-any-more). The only supported OSes are Win95 (fully operational, not taking into account the Os flaws) and WinNT 4.0 (some "minor" features aren't available). Maybe WINE is the only way to get it (and similar hardware) work with Linux. Infact the driver and related software could be installed and executed with WINE, assuming that WINE could succesfully access the I/O port used by the hardware (and maybe linux is not even aware of that hardware!) Maybe I'll try that by myself and post the (maybe bad) results I'll get (at least a divorce from my wife). Any comments, hint, post wuold be appreciated. There's a lot o people owning a Rainbow Runner (not G-Series) or hardware with the same problem out there. So please help. Thank you. DonalDuck

1 0

mingw crossbuild broke
by Kevin Koltzau May 25, 2004

May 25, 2004

the mingw crossbuild broke last night with mmap.c: In function `wine_mmap_add_reserved_area': mmap.c:291: error: `PROT_NONE' undeclared (first use in this function) mmap.c:291: error: (Each undeclared identifier is reported only once mmap.c:291: error: for each function it appears in.) mmap.c:291: error: `MAP_FIXED' undeclared (first use in this function) should those defines be included in one of our headers?

1 0

failure report: picasa-1.6
by Dan Kegel May 25, 2004

May 25, 2004

I tried running the demo of Picasa from http://www.lifescapeinc.com/picasa/ under Wine-20040213 and Wine-20040505, but the installer just put up "Out of Memory" boxes :-( This progam uses some unusual installer app (it announces which one early on when you run it under Windows). Anyone seen this before? -- My technical stuff: http://kegel.com My politics: see http://www.misleader.org for examples of why I'm for regime change

4 8

Re: [DMUSIC] Make Unreal2 playable
by Dmitry Timoshkov May 25, 2004

May 25, 2004

"Raphael" <fenix(a)club-internet.fr> wrote: > @@ -131,6 +149,7 @@ > } > > static BOOL PostMessageToProcessMsgThread(IDirectMusicPerformance8Impl* This, UINT iMsg) { > + return FALSE; > if (FALSE == This->procThreadTicStarted && PROCESSMSG_EXIT != iMsg) { > This->procThread = CreateThread(NULL, 0, ProcessMsgThread, This, 0, &This->procThreadId); > if (NULL == This->procThread) return FALSE; Are you sure you want this change to be included? -- Dmitry.

1 0

Re: General serial communications issues
by David Purucker May 25, 2004

May 25, 2004

>>I have tried several terminal programs (Hyperterm and TeraTerm) under > > several builds of > > Wine including Wine-20040505 and they seem to have the same generic >problem. > >We are using serial ports under Wine for a current project, and they >work OK (apart from a few warnings). The machine I'm using doesn't >have X, so I haven't tried HyperTerminal or TeraTerm, but our >console-mode applications operate pretty much as expected. > >I'm using two patches applied to a recent CVS. The first is a patch >Eric Pouech recently posted: > > http://www.winehq.org/hypermail/wine-devel/2004/05/0297.html > >The second is a patch of my own to supress warnings when using a >pseudo-tty as a serial port: > > http://www.winehq.org/hypermail/wine-devel/2004/05/0278.html > >But our apps worked OK before that; they just generated some warnings. >Still, these patches might help, or at least cut down on the noise and >help you find the real error. > >TeraTerm has source available, so it should be one of the easier >applications to debug. > >----ScottG. Thanks Scott, for the prompt reply. I tried the patch posted by Eric Pouech and still no luck, the program still appears to receive nothing back. I can type directly into the terminal a ATDT command to dial a number and I get a connect, but never see a response on the screen. I have tested it on several different machines with different wine builds, but could someone please test this, just to make sure that I am not doing something trivially wrong. Thanks again _________________________________________________________________ Is your PC infected? Get a FREE online computer virus scan from McAfee® Security. http://clinic.mcafee.com/clinic/ibuy/campaign.asp?cid=3963

1 0

Re: additional download for winehq download page
by Ivan May 25, 2004

May 25, 2004

> This may be a fine thing to do for another project, but this is > not a Wine thing, sorry. OK, but at least the vb runtime and the win installer should be there. Nobody is implementing, or, AFAIK, planning to implement these parts of windows at the moment, and without them many programs simply have no hope of working with wine. We may also want to list dcom, as we even had it up on the sf page at one time. Ivan.

2 1

exec-shield patch 7
by Mike McCormack May 25, 2004

May 25, 2004

Description: This patch allows platforms using exec-shield patch against the Linux kernel (eg. Fedora Core) to run Wine without disabling exec-shield. License: LGPL Changes in version 7: * fix bug that caused crash when CreateProcess was called * update to work with the new functions for reserving memory Changes in version 6: * include ulimit workaround * resolve conflict with recent patch to WineHQ CVS tip Changes in version 5: * reserve the shared heap area as well as the PE load area * allow multiple reserve areas in libwine Changes in version 4: * set the WINEPEEXEPATH environment variable correctly Changes in version 3: * reserve only the area required by the PE EXE if launched from CreateProcess (else just a default slab) * pass the reserved area to wine, so wine knows about it * unmap the preloader after it is no longer used Changes in verison 2: * replace the shared object loader instead of using a bootstrap exe Version 1: * create a bootstrap exe that reserves memory before loading wine-pthread or wine-kthread Index: dlls/kernel/process.c =================================================================== RCS file: /home/wine/wine/dlls/kernel/process.c,v retrieving revision 1.63 diff -u -r1.63 process.c --- dlls/kernel/process.c 25 May 2004 01:29:24 -0000 1.63 +++ dlls/kernel/process.c 25 May 2004 04:59:54 -0000 @@ -1257,7 +1257,8 @@ * * Build the environment of a new child process. */ -static char **build_envp( const WCHAR *envW ) +static char **build_envp( const WCHAR *envW, + const char *workdir, const char *unixpath ) { const WCHAR *p; char **envp; @@ -1270,7 +1271,7 @@ if (!(env = malloc( length ))) return NULL; WideCharToMultiByte( CP_UNIXCP, 0, envW, p - envW, env, length, NULL, NULL ); - count += 4; + count += 6; if ((envp = malloc( count * sizeof(*envp) ))) { @@ -1283,12 +1284,19 @@ if ((p = getenv("TMP"))) *envptr++ = alloc_env_string( "TMP=", p ); if ((p = getenv("HOME"))) *envptr++ = alloc_env_string( "HOME=", p ); if ((p = getenv("WINEPREFIX"))) *envptr++ = alloc_env_string( "WINEPREFIX=", p ); + + /* record the unix working directory */ + if (workdir) *envptr++ = alloc_env_string( "WINEWORKDIR=", workdir ); + if (workdir) *envptr++ = alloc_env_string( "WINEPEEXEPATH=", unixpath ); + /* now put the Windows environment strings */ for (p = env; *p; p += strlen(p) + 1) { if (is_special_env_var( p )) /* prefix it with "WINE" */ *envptr++ = alloc_env_string( "WINE", p ); else if (strncmp( p, "HOME=", 5 ) && + strncmp( p, "WINEWORKDIR=", 12 ) && + strncmp( p, "WINEPEEXEPATH=", 14 ) && strncmp( p, "WINEPREFIX=", 11 )) *envptr++ = p; } *envptr = 0; @@ -1319,7 +1327,7 @@ if (!(pid = fork())) /* child */ { char **argv = build_argv( cmdline, 0 ); - char **envp = build_envp( env ); + char **envp = build_envp( env, NULL, NULL ); close( fd[0] ); /* Reset signals that we previously set to SIG_IGN */ @@ -1354,14 +1362,8 @@ RTL_USER_PROCESS_PARAMETERS *params; UNICODE_STRING image_str, cmdline_str, curdir_str, desktop, title; NTSTATUS status; - WCHAR buffer[MAX_PATH]; - - if(!GetLongPathNameW( filename, buffer, MAX_PATH )) - lstrcpynW( buffer, filename, MAX_PATH ); - if(!GetFullPathNameW( buffer, MAX_PATH, buffer, NULL )) - lstrcpynW( buffer, filename, MAX_PATH ); - RtlInitUnicodeString( &image_str, buffer ); + RtlInitUnicodeString( &image_str, filename ); RtlInitUnicodeString( &cmdline_str, cmdline ); if (cur_dir) RtlInitUnicodeString( &curdir_str, cur_dir ); if (startup->lpDesktop) RtlInitUnicodeString( &desktop, startup->lpDesktop ); @@ -1415,10 +1417,16 @@ pid_t pid; int err; char dummy = 0; + WCHAR buffer[MAX_PATH]; if (!env) env = GetEnvironmentStringsW(); - if (!(params = create_user_params( filename, cmd_line, cur_dir, startup ))) + if(!GetLongPathNameW( filename, buffer, MAX_PATH )) + lstrcpynW( buffer, filename, MAX_PATH ); + if(!GetFullPathNameW( buffer, MAX_PATH, buffer, NULL )) + lstrcpynW( buffer, filename, MAX_PATH ); + + if (!(params = create_user_params( buffer, cmd_line, cur_dir, startup ))) return FALSE; /* create the synchronization pipes */ @@ -1444,7 +1452,20 @@ if (!(pid = fork())) /* child */ { char **argv = build_argv( cmd_line, 1 ); - char **envp = build_envp( env ); + char **envp = build_envp( env, NULL, NULL ); + char *unixpath; + + unixpath = wine_get_unix_file_name( buffer ); + + /* preseve the current directory, the preloader may forget it */ + if( !unixdir ) + { + char *dir = HeapAlloc( GetProcessHeap(), 0, MAX_PATH ); + getcwd( dir, MAX_PATH ); + unixdir = dir; + } + envp = build_envp( env, unixdir, unixpath ); + HeapFree( GetProcessHeap(), 0, unixpath ); close( startfd[1] ); close( execfd[0] ); @@ -1456,8 +1477,6 @@ /* Reset signals that we previously set to SIG_IGN */ signal( SIGPIPE, SIG_DFL ); signal( SIGCHLD, SIG_DFL ); - - if (unixdir) chdir(unixdir); if (argv && envp) { Index: libs/wine/config.c =================================================================== RCS file: /home/wine/wine/libs/wine/config.c,v retrieving revision 1.4 diff -u -r1.4 config.c --- libs/wine/config.c 11 May 2004 04:29:18 -0000 1.4 +++ libs/wine/config.c 25 May 2004 04:59:54 -0000 @@ -271,6 +271,23 @@ return user_name; } +/* run wine binary in it's own directory, so it can find it's preloader */ +int wine_cwd_execve( const char *name, char **argv, char **envp ) +{ + char *dir, *p; + + dir = strdup( name ); + p = strrchr( dir, '/' ); + if( p ) + { + *p++ = 0; + chdir( dir ); + name = p; + } + + return execve( name, argv, envp ); +} + /* exec a wine internal binary (either the wine loader or the wine server) */ /* if name is null, default to the name of the current binary */ void wine_exec_wine_binary( const char *name, char **argv, char **envp ) @@ -285,7 +302,7 @@ argv[0] = xmalloc( sizeof(BINDIR "/") + strlen(name) ); strcpy( argv[0], BINDIR "/" ); strcat( argv[0], name ); - execve( argv[0], argv, envp ); + wine_cwd_execve( argv[0], argv, envp ); free( argv[0] ); /* now try the path of argv0 of the current binary */ @@ -294,7 +311,7 @@ argv[0] = xmalloc( strlen(argv0_path) + strlen(name) + 1 ); strcpy( argv[0], argv0_path ); strcat( argv[0], name ); - execve( argv[0], argv, envp ); + wine_cwd_execve( argv[0], argv, envp ); free( argv[0] ); } @@ -311,7 +328,7 @@ memcpy( argv[0], pos, ptr - pos ); strcpy( argv[0] + (ptr - pos), "/" ); strcat( argv[0] + (ptr - pos), name ); - execve( argv[0], argv, envp ); + wine_cwd_execve( argv[0], argv, envp ); pos = ptr; } free( argv[0] ); Index: loader/Makefile.in =================================================================== RCS file: /home/wine/wine/loader/Makefile.in,v retrieving revision 1.14 diff -u -r1.14 Makefile.in --- loader/Makefile.in 22 Nov 2003 00:08:26 -0000 1.14 +++ loader/Makefile.in 25 May 2004 04:59:54 -0000 @@ -23,20 +23,30 @@ LIBPTHREAD = @LIBPTHREAD@ LDEXECFLAGS = @LDEXECFLAGS@ -wine-glibc: glibc.o Makefile.in +WINELDEXECFLAGS = $(LDEXECFLAGS),--dynamic-linker=$(WINELDSO) + +WINELDSO=ld-winepreload.so + +$(WINELDSO): preload.o Makefile.in + $(CC) -nostartfiles -lc -static -Wl,"-Ttext=0x7ff81000" -o $@ $< + +preload.c: preload.h main.h + +wine-glibc: glibc.o Makefile.in $(WINELDSO) $(CC) -o $@ $(LDEXECFLAGS) glibc.o $(LIBWINE) $(LIBPORT) $(LIBPTHREAD) $(EXTRALIBS) $(LDFLAGS) -wine-kthread: $(KTHREAD_OBJS) Makefile.in - $(CC) -o $@ $(LDEXECFLAGS) $(KTHREAD_OBJS) $(LIBWINE) $(LIBPORT) $(EXTRALIBS) $(LDFLAGS) +wine-kthread: $(KTHREAD_OBJS) Makefile.in $(WINELDSO) + $(CC) -o $@ $(WINELDEXECFLAGS) $(KTHREAD_OBJS) $(LIBWINE) $(LIBPORT) $(EXTRALIBS) $(LDFLAGS) -wine-pthread: $(PTHREAD_OBJS) Makefile.in - $(CC) -o $@ $(LDEXECFLAGS) $(PTHREAD_OBJS) $(LIBWINE) $(LIBPORT) $(LIBPTHREAD) $(EXTRALIBS) $(LDFLAGS) +wine-pthread: $(PTHREAD_OBJS) Makefile.in $(WINELDSO) + $(CC) -o $@ $(WINELDEXECFLAGS) $(PTHREAD_OBJS) $(LIBWINE) $(LIBPORT) $(LIBPTHREAD) $(EXTRALIBS) $(LDFLAGS) $(MODULE): $(MAIN_BINARY) $(RM) $(MODULE) && $(LN_S) $(MAIN_BINARY) $(MODULE) install:: $(WINE_BINARIES) $(MKINSTALLDIRS) $(bindir) + $(INSTALL_PROGRAM) $(WINELDSO) $(bindir); \ for f in $(WINE_BINARIES); do \ if [ "$(MAIN_BINARY)" = "$$f" ]; \ then $(INSTALL_PROGRAM) $$f $(bindir)/$(MODULE); \ @@ -48,6 +58,6 @@ $(RM) $(WINE_BINARIES:%=$(bindir)/%) $(bindir)/$(MODULE) clean:: - $(RM) $(WINE_BINARIES) $(MODULE) + $(RM) $(WINE_BINARIES) $(MODULE) $(WINELDSO) ### Dependencies: Index: loader/glibc.c =================================================================== RCS file: /home/wine/wine/loader/glibc.c,v retrieving revision 1.3 diff -u -r1.3 glibc.c --- loader/glibc.c 20 Feb 2004 20:19:23 -0000 1.3 +++ loader/glibc.c 25 May 2004 04:59:54 -0000 @@ -29,6 +29,7 @@ #ifdef HAVE_PTHREAD_H # include <pthread.h> #endif +#include <sys/resource.h> #include "wine/library.h" @@ -75,6 +76,23 @@ { const char *loader = getenv( "WINELOADER" ); const char *threads = get_threading(); + char cwd[MAX_PATH], *cwd_var; + + /* hack to work-around broken address space layout in RHEL3 */ + struct rlimit limit; + getrlimit( RLIMIT_STACK, &limit ); + limit.rlim_cur = limit.rlim_max; + setrlimit( RLIMIT_STACK, &limit ); + + if( !getcwd( cwd, sizeof cwd ) ) + { + fprintf( stderr, "wine: could get current directory\n" ); + exit(1); + } + cwd_var = xmalloc( sizeof("WINEWORKDIR=") + strlen(cwd) ); + strcpy( cwd_var, "WINEWORKDIR=" ); + strcat( cwd_var, cwd ); + putenv( cwd_var ); if (loader) { @@ -86,6 +104,9 @@ new_name = xmalloc( (path - loader) + strlen(threads) + 1 ); memcpy( new_name, loader, path - loader ); + new_name[ path - loader ] = 0; + if( new_name[0] ) + chdir( new_name ); strcpy( new_name + (path - loader), threads ); /* update WINELOADER with the new name */ Index: loader/main.c =================================================================== RCS file: /home/wine/wine/loader/main.c,v retrieving revision 1.91 diff -u -r1.91 main.c --- loader/main.c 25 May 2004 01:29:24 -0000 1.91 +++ loader/main.c 25 May 2004 04:59:54 -0000 @@ -20,14 +20,52 @@ #include <stdio.h> #include <stdlib.h> +#include <unistd.h> +#include <sys/mman.h> #include "wine/library.h" +#include "main.h" + +/* this variable is set by the wine-specific shared object loader */ +struct wine_preload_info *wine_preload_vars; + /********************************************************************** * main */ int main( int argc, char *argv[] ) { - char error[1024]; + char error[1024], *dir; + + if( wine_preload_vars ) + { +#if 0 + printf("vars(*%p) = %p -> (%p, %08x) (%p, %08x) (%p, %08x)\n", + &wine_preload_vars, wine_preload_vars, + wine_preload_vars->pe_address, + wine_preload_vars->pe_size, + wine_preload_vars->heap_address, + wine_preload_vars->heap_size, + wine_preload_vars->preload_base, + wine_preload_vars->preload_size ); +#endif + + /* let wine know which area was mapped for the PE EXE */ + wine_mmap_add_reserved_area( wine_preload_vars->pe_address, + wine_preload_vars->pe_size); + + /* let wine know which area was mapped for the system heap */ + wine_mmap_add_reserved_area( wine_preload_vars->heap_address, + wine_preload_vars->heap_size); + + /* unmap the preloader itself */ + munmap( wine_preload_vars->preload_base, + wine_preload_vars->preload_size ); + wine_preload_vars = NULL; + } + + dir = getenv( "WINEWORKDIR" ); + if( dir ) + chdir( dir ); wine_init( argc, argv, error, sizeof(error) ); fprintf( stderr, "wine: failed to initialize: %s\n", error ); --- /dev/null 1994-07-18 08:46:18.000000000 +0900 +++ loader/main.h 2004-04-15 17:29:17.000000000 +0900 @@ -0,0 +1,10 @@ + +struct wine_preload_info +{ + void *pe_address; + size_t pe_size; + void *heap_address; + size_t heap_size; + void *preload_base; + size_t preload_size; +}; --- /dev/null 1994-07-18 08:46:18.000000000 +0900 +++ loader/preload.h 2004-04-15 17:29:17.000000000 +0900 @@ -0,0 +1,51 @@ +/* + mini loader for ld.so - x86 Linux 2.4.x+ kernels only + + Copyright 2004 Mike McCormack for Codeweavers + + Copyright (C) 1995,96,97,98,99,2000,2001,2002 Free Software Foundation, Inc. + + The GNU C 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. + + The GNU C 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 the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. + */ + +#define LDSO_NAME "/lib/ld-linux.so.2" + +/* ELF definitions */ +#define ELF_PREFERRED_ADDRESS(loader, maplength, mapstartpref) (mapstartpref) +#define ELF_FIXED_ADDRESS(loader, mapstart) ((void) 0) + +#define __ELF_NATIVE_CLASS 32 +#define MAP_BASE_ADDR(l) 0 +#define MAP_COPY MAP_PRIVATE +#define ANONFD -1 + +#define ElfW(type) _ElfW (Elf, __ELF_NATIVE_CLASS, type) +#define _ElfW(e,w,t) _ElfW_1 (e, w, _##t) +#define _ElfW_1(e,w,t) e##w##t + +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <unistd.h> +#include <sys/mman.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> + +#define GL(x) x + +extern unsigned int GL(dl_pagesize); + --- /dev/null 1994-07-18 08:46:18.000000000 +0900 +++ loader/preload.c 2004-05-25 15:04:52.000000000 +0900 @@ -0,0 +1,750 @@ +/* + mini loader for ld.so - x86 Linux 2.4.x+ kernels only + + Copyright 2004 Mike McCormack for Codeweavers + + Copyright (C) 1995,96,97,98,99,2000,2001,2002 Free Software Foundation, Inc. + + The GNU C 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. + + The GNU C 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 the GNU C Library; if not, write to the Free + Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA + 02111-1307 USA. + */ + +/* + * Design notes + * + * The goal of this program is to be a workaround for exec-shield, as used + * by the Linux kernel distributed with Fedora Core and other distros. + * + * To do this, we implement our own shared object loader that reserves memory + * that is important to Wine, and then loads the the standard shared object + * loader (ELF interpreter), usually ld-linux.so.2 + * + * This program will be invoked by linking it as the shared object loader + * using the link time flags --shared-loader=ld-winepreload.so . Since we + * are not using an absolute path, an executable using this shared object + * loader must be invoked from the directory that contains this loader. + * + * We will try to set up the stack an memory area so that the program that + * loads after us (eg. the wine binary) never knows we were here, except that + * areas of memory it needs are already magically reserved. + * + * The following memory areas are important to Wine: + * 0x00000000 - the DOS area + * 0x00400000 - the standard Win32 load address + * 0x65430000 - the shared heap + * + * If this program is used as the shared object loader. the only difference + * that the loaded programs should see is that this loader will be mapped + * into memory when it starts. + * + * Limitations: + * * using a non-absolution path for the shared object loader is not standard + */ + +/* + * References (things I consulted to understand how ELF loading works): + * + * glibc 2.3.2 elf/dl-load.c + * http://www.gnu.org/directory/glibc.html + * + * Linux 2.6.4 fs/binfmt_elf.c + * ftp://ftp.kernel.org/pub/linux/kernel/v2.6/linux-2.6.4.tar.bz2 + * + * Userland exec, by <grugq(a)hcunix.net> + * http://cert.uni-stuttgart.de/archive/bugtraq/2004/01/msg00002.html + * + * The ELF specification: + * http://www.linuxbase.org/spec/booksets/LSB-Embedded/LSB-Embedded/book387.ht… + */ + +#include "preload.h" +#include "main.h" +#include <elf.h> + +/* the following headers are needed to get the definition for IMAGE_NT_HEADERS */ +#include <stdarg.h> +#include "windef.h" +#include "winbase.h" + +/* debugging */ +#undef DUMP_SEGMENTS +#undef DUMP_ARG +#undef DUMP_AUX_INFO +#undef DUMP_SYMS +#undef DUMP_PERESERVE + +/* older systems may not define these */ +#ifndef PT_TLS +#define PT_TLS 7 +#endif +#ifndef PT_GNU_EH_FRAME +#define PT_GNU_EH_FRAME 0x6474e550 +#endif +#ifndef PT_GNU_STACK +#define PT_GNU_STACK 0x6474e551 +#endif +#ifndef PT_GNU_RELRO +#define PT_GNU_RELRO 0x6474e552 +#endif + +unsigned int GL(dl_pagesize) = 0x0000; +extern const unsigned char _end[]; + +struct link_map { + ElfW(Addr) l_addr; + ElfW(Dyn) *l_ld; + const ElfW(Phdr) *l_phdr; + ElfW(Addr) l_entry; + ElfW(Half) l_ldnum; + ElfW(Half) l_phnum; + ElfW(Addr) l_map_start, l_map_end; +}; + +/* + * wld_printf - just the basics + * + * %x prints a hex number + * %s prints a string + */ +void wld_sprintf(char *str, char *fmt, int *args ) +{ + char *p = fmt; + + while( *p ) + { + if( *p == '%' ) + { + p++; + if( *p == 'x' ) + { + int ch, i, x = *++args; + for(i=7; i>=0; i--) + { + ch = (x>>(i*4))&0xf; + ch += '0'; + if(ch>'9') + ch+=('A'-10-'0'); + *str++ = ch; + } + } + else if( *p == 's' ) + { + char *s = (char*)*++args; + while(*s) + *str++ = *s++; + } + else if( *p == 0 ) + break; + p++; + } + *str++ = *p++; + } + *str = 0; +} + +int wld_strlen(char *str) +{ + int n = 0; + while( str[n] ) + n++; + return n; +} + +void wld_printf(char *fmt, ... ) +{ + char buffer[100]; + int n; + + buffer[0]=0; + wld_sprintf(buffer, fmt, (int*)&fmt ); + n = wld_strlen( buffer ); + write(2, buffer, n); +} + +void wld_strncpy(char *dest, char *src, int count ) +{ + while( (*dest++ = *src++) && --count ) + ; + *dest = 0; +} + +void wld_memset( unsigned char *buf, int val, size_t len ) +{ + while( len-- ) + *buf++=val; +} + +#ifdef DUMP_AUX_INFO +/* + * Dump interesting bits of the ELF auxv_t structure that is passed + * as the 4th parameter to the _start function + */ +static void wld_dump_auxilary( ElfW(auxv_t) *av ) +{ + for ( ; av->a_type != AT_NULL; av++) + switch (av->a_type) + { + case AT_PAGESZ: + wld_printf("AT_PAGESZ = %x\n",av->a_un.a_val); + break; + case AT_PHDR: + wld_printf("AT_PHDR = %x\n",av->a_un.a_ptr); + break; + case AT_PHNUM: + wld_printf("AT_PHNUM = %x\n",av->a_un.a_val); + break; + case AT_ENTRY: + wld_printf("AT_ENTRY = %x\n",av->a_un.a_val); + break; + case AT_BASE: + wld_printf("AT_BASE = %x\n",av->a_un.a_val); + break; + } +} +#endif + +/* + * wld_set_auxilary + * + * Set a field of the auxillary structure + */ +void wld_set_auxilary( ElfW(auxv_t) *av, int type, long int val ) +{ + for ( ; av->a_type != AT_NULL; av++) + if( av->a_type == type ) + av->a_un.a_val = val; +} + +/* + * wld_get_auxilary + * + * Get a field of the auxillary structure + */ +int wld_get_auxilary( ElfW(auxv_t) *av, int type, int *val ) +{ + for ( ; av->a_type != AT_NULL; av++) + if( av->a_type == type ) + { + *val = av->a_un.a_val; + return 1; + } + return 0; +} + +/* + * wld_map_so + * + * modelled after _dl_map_object_from_fd() from glibc-2.3.1/elf/dl-load.c + * + * This function maps the segments from an ELF object, and optionally + * stores information about the mapping into the auxv_t structure. + */ +void* wld_map_so( int fd, void* buf, size_t buflen, struct link_map *l) +{ + ElfW(Ehdr) *header = buf; + ElfW(Phdr) *phdr, *ph; + /* Scan the program header table, collecting its load commands. */ + struct loadcmd + { + ElfW(Addr) mapstart, mapend, dataend, allocend; + off_t mapoff; + int prot; + } loadcmds[header->e_phnum], *c; + size_t nloadcmds = 0, maplength; + + header = buf; + phdr = (void*) (((unsigned char*)buf) + header->e_phoff); + + if( ( header->e_ident[0] != 0x7f ) || + ( header->e_ident[1] != 'E' ) || + ( header->e_ident[2] != 'L' ) || + ( header->e_ident[3] != 'F' ) ) + { + wld_printf("Not an ELF binary... don't know how to load it\n"); + return NULL; + } + + if( header->e_machine != EM_386 ) + { + wld_printf("Not an i386 ELF binary... don't know how to load it\n"); + return NULL; + } + + maplength = header->e_phnum * sizeof (ElfW(Phdr)); + if (header->e_phoff + maplength > (size_t) buflen) + { + wld_printf("oops... not enough space for ELF headers\n"); + exit(1); + } + + l->l_ld = 0; + l->l_addr = 0; + l->l_phnum = header->e_phnum; + l->l_entry = header->e_entry; + + for (ph = phdr; ph < &phdr[l->l_phnum]; ++ph) + { + +#ifdef DUMP_SEGMENTS + wld_printf( "ph = %x\n", ph ); + wld_printf( " p_type = %x\n", ph->p_type ); + wld_printf( " p_flags = %x\n", ph->p_flags ); + wld_printf( " p_offset = %x\n", ph->p_offset ); + wld_printf( " p_vaddr = %x\n", ph->p_vaddr ); + wld_printf( " p_paddr = %x\n", ph->p_paddr ); + wld_printf( " p_filesz = %x\n", ph->p_filesz ); + wld_printf( " p_memsz = %x\n", ph->p_memsz ); + wld_printf( " p_align = %x\n", ph->p_align ); +#endif + + switch (ph->p_type) + { + /* These entries tell us where to find things once the file's + segments are mapped in. We record the addresses it says + verbatim, and later correct for the run-time load address. */ + case PT_DYNAMIC: + l->l_ld = (void *) ph->p_vaddr; + l->l_ldnum = ph->p_memsz / sizeof (Elf32_Dyn); + break; + + case PT_PHDR: + l->l_phdr = (void *) ph->p_vaddr; + break; + + case PT_TLS: + /* + * We don't need to set anything up because we're + * emulating the kernel, not ld-linux.so.2 + * The ELF loader will set up the TLS data itself. + */ + break; + + case PT_LOAD: + { + if ((ph->p_align & (GL(dl_pagesize) - 1)) != 0) + { + wld_printf("ELF load command alignment not page-aligned"); + return NULL; + } + if (((ph->p_vaddr - ph->p_offset) & (ph->p_align - 1)) != 0) + { + wld_printf("ELF load command address/offset not properly aligned"); + return NULL; + } + + c = &loadcmds[nloadcmds++]; + c->mapstart = ph->p_vaddr & ~(ph->p_align - 1); + c->mapend = ((ph->p_vaddr + ph->p_filesz + GL(dl_pagesize) - 1) + & ~(GL(dl_pagesize) - 1)); + c->dataend = ph->p_vaddr + ph->p_filesz; + c->allocend = ph->p_vaddr + ph->p_memsz; + c->mapoff = ph->p_offset & ~(ph->p_align - 1); + + c->prot = 0; + if (ph->p_flags & PF_R) + c->prot |= PROT_READ; + if (ph->p_flags & PF_W) + c->prot |= PROT_WRITE; + if (ph->p_flags & PF_X) + c->prot |= PROT_EXEC; + } + break; + + case PT_INTERP: + break; + + case PT_SHLIB: + case PT_NOTE: + case PT_GNU_EH_FRAME: + case PT_GNU_STACK: + case PT_GNU_RELRO: + break; + + default: + /* wld_printf("warning: unknown header type %x\n", ph->p_type); */ + break; + } + } + + /* Now process the load commands and map segments into memory. */ + c = loadcmds; + + /* Length of the sections to be loaded. */ + maplength = loadcmds[nloadcmds - 1].allocend - c->mapstart; + + if( header->e_type == ET_DYN ) + { + ElfW(Addr) mappref; + mappref = (ELF_PREFERRED_ADDRESS (loader, maplength, c->mapstart) + - MAP_BASE_ADDR (l)); + + /* Remember which part of the address space this object uses. */ + l->l_map_start = (ElfW(Addr)) mmap ((void *) mappref, maplength, + c->prot, MAP_COPY | MAP_FILE, + fd, c->mapoff); + /* wld_printf("set : offset = %x\n", c->mapoff); */ + /* wld_printf("l->l_map_start = %x\n", l->l_map_start); */ + + l->l_map_end = l->l_map_start + maplength; + l->l_addr = l->l_map_start - c->mapstart; + + mprotect ((caddr_t) (l->l_addr + c->mapend), + loadcmds[nloadcmds - 1].allocend - c->mapend, + PROT_NONE); + goto postmap; + } + else + { + ELF_FIXED_ADDRESS (loader, c->mapstart); + } + + /* Remember which part of the address space this object uses. */ + l->l_map_start = c->mapstart + l->l_addr; + l->l_map_end = l->l_map_start + maplength; + + while (c < &loadcmds[nloadcmds]) + { + if (c->mapend > c->mapstart) + /* Map the segment contents from the file. */ + mmap ((void *) (l->l_addr + c->mapstart), + c->mapend - c->mapstart, c->prot, + MAP_FIXED | MAP_COPY | MAP_FILE, fd, c->mapoff); + + postmap: + if (l->l_phdr == 0 + && (ElfW(Off)) c->mapoff <= header->e_phoff + && ((size_t) (c->mapend - c->mapstart + c->mapoff) + >= header->e_phoff + header->e_phnum * sizeof (ElfW(Phdr)))) + /* Found the program header in this segment. */ + l->l_phdr = (void *) (c->mapstart + header->e_phoff - c->mapoff); + + if (c->allocend > c->dataend) + { + /* Extra zero pages should appear at the end of this segment, + after the data mapped from the file. */ + ElfW(Addr) zero, zeroend, zeropage; + + zero = l->l_addr + c->dataend; + zeroend = l->l_addr + c->allocend; + zeropage = ((zero + GL(dl_pagesize) - 1) + & ~(GL(dl_pagesize) - 1)); + + /* + * This is different to the dl-load load... + * ld-linux.so.2 relies on the whole page being zero'ed + */ + zeroend = (zeroend +GL(dl_pagesize) - 1) & ~(GL(dl_pagesize) - 1); + + if (zeroend < zeropage) + { + /* All the extra data is in the last page of the segment. + We can just zero it. */ + zeropage = zeroend; + } + + if (zeropage > zero) + { + /* Zero the final part of the last page of the segment. */ + if ((c->prot & PROT_WRITE) == 0) + { + /* Dag nab it. */ + mprotect ((caddr_t) (zero & ~(GL(dl_pagesize) - 1)), + GL(dl_pagesize), + c->prot|PROT_WRITE); + } + memset ((void *) zero, '\0', zeropage - zero); + if ((c->prot & PROT_WRITE) == 0) + mprotect ((caddr_t) (zero & ~(GL(dl_pagesize) - 1)), + GL(dl_pagesize), c->prot); + } + + if (zeroend > zeropage) + { + /* Map the remaining zero pages in from the zero fill FD. */ + caddr_t mapat; + mapat = mmap ((caddr_t) zeropage, zeroend - zeropage, + c->prot, MAP_ANON|MAP_PRIVATE|MAP_FIXED, + ANONFD, 0); + } + } + + ++c; + } + + if (l->l_phdr == NULL) + { + wld_printf("no program header\n"); + exit(1); + } + + (ElfW(Addr)) l->l_phdr += l->l_addr; + (ElfW(Addr)) l->l_entry += l->l_addr; + + return (void*)l->l_entry; +} + +#define ROUND_SIZE(addr,size) \ + (((int)(size) + ((int)(addr) & page_mask) + page_mask) & ~page_mask) + +/* + * wld_reserve_wine_memory + * + * This is the reason we are here. + * We want to reserve address space so that the kernel doesn't + * map stuff there. We do this by getting in first before + * anything else is mapped (except this program) and reserve + * the memory we want. + */ +void wld_reserve_wine_memory( char *pe_file, void **base, size_t *size ) +{ + IMAGE_DOS_HEADER dos; + IMAGE_NT_HEADERS nt; + int unix_fd, page_mask = GL(dl_pagesize) - 1; + + if( !pe_file ) + return; + + unix_fd = open( pe_file, O_RDONLY, 0); + if( unix_fd < 0 ) + return; + + /* + * The following code is borrowed from + * get_image_params() in wine/server/mapping.c + */ + if (read( unix_fd, &dos, sizeof(dos) ) != sizeof(dos)) goto error; + if (dos.e_magic != IMAGE_DOS_SIGNATURE) goto error; + if (lseek( unix_fd, dos.e_lfanew, SEEK_SET ) == -1) goto error; + + if (read( unix_fd, &nt.Signature, sizeof(nt.Signature) ) != sizeof(nt.Signature)) goto error; + if (nt.Signature != IMAGE_NT_SIGNATURE) goto error; + if (read( unix_fd, &nt.FileHeader, sizeof(nt.FileHeader) ) != sizeof(nt.FileHeader)) goto error; + /* zero out Optional header in the case it's not present or partial */ + memset(&nt.OptionalHeader, 0, sizeof(nt.OptionalHeader)); + if (read( unix_fd, &nt.OptionalHeader, nt.FileHeader.SizeOfOptionalHeader) != nt.FileHeader.SizeOfOptionalHeader) goto error; + + *size = ROUND_SIZE( 0, nt.OptionalHeader.SizeOfImage ); + *base = (void *)nt.OptionalHeader.ImageBase; + +#ifdef DUMP_PERESERVE + wld_printf( "file %s\n", pe_file ); + wld_printf( "parsed PE image... base %x size %x\n", *base, *size ); +#endif + +error: + close( unix_fd ); +} + +/* + * Find a symbol in the symbol table of the executable loaded + */ +void *wld_find_symbol( const ElfW(Phdr) *phdr, int num, char *var ) +{ + const ElfW(Dyn) *dyn = NULL; + void *ptr = NULL; + const ElfW(Phdr) *ph; + const ElfW(Sym) *symtab = NULL; + const char *strings = NULL; + + /* check the values */ +#ifdef DUMP_SYMS + wld_printf("%x %x\n", phdr, num ); +#endif + if( ( phdr == NULL ) || ( num == 0 ) ) + { + wld_printf("could not find PT_DYNAMIC header entry\n"); + return NULL; + } + + /* parse the (already loaded) ELF executable's header */ + for (ph = phdr; ph < &phdr[num]; ++ph) + { + if( PT_DYNAMIC == ph->p_type ) + { + dyn = (void *) ph->p_vaddr; + num = ph->p_memsz / sizeof (Elf32_Dyn); + break; + } + } + if( !dyn ) + { + wld_printf("No PT_DYNAMIC section\n"); + return ptr; + } + + while( dyn->d_tag ) + { + if( dyn->d_tag == DT_STRTAB ) + strings = (const char*) dyn->d_un.d_ptr; + if( dyn->d_tag == DT_SYMTAB ) + symtab = (const ElfW(Sym) *)dyn->d_un.d_ptr; +#ifdef DUMP_SYMS + wld_printf("%x %x\n", dyn->d_tag, dyn->d_un.d_ptr ); +#endif + dyn++; + } + + if( (!symtab) || (!strings) ) + { + wld_printf("No symbol table found\n"); + return ptr; + } + + /* FIXME: this crashes if it doesn't find the symbol we want */ + do + { + if( ( ELF32_ST_BIND(symtab->st_info) == STT_OBJECT ) && + ( 0 == strcmp( strings+symtab->st_name, var ) ) ) + { + ptr = (void*)symtab->st_value; + break; + } + symtab++; + } + while( symtab->st_name ); + +#ifdef DUMP_SYMS + wld_printf("Found %s -> %x\n", strings+symtab->st_name, symtab->st_value ); +#endif + + return ptr; +} + +/* + * wld_start + * + * Repeat the actions the kernel would do when loading a dynamically linked .so + * Load the ELF interperter ld-linux.so and then binary itself. + * + * Note, we assume that the binary is a dynamically linked ELF shared object + * and that ld-linux.so is its ELF interpreter. + */ +void* wld_start( int argc ) +{ + int fd, phnum = 0; + unsigned char buf[0x800]; + void *start; + char **argv = (char**) &argc, **p, *pe_file = NULL; + struct link_map map; + ElfW(auxv_t)* av; + ElfW(Phdr) *elf_phdr = NULL; + struct wine_preload_info **wine_preload_vars; + static struct wine_preload_info vars; + + memset( &map, 0, sizeof map ); + + /* skip over the parameters */ + p = ++argv; +#ifdef DUMP_ARG + { + int i; + for( i = 0; i<argc; i++ ) + wld_printf("argv[%x] = %s\n", i, *p++); + if(*p++) + wld_printf("argv[argc] not NULL!!\n"); + } +#else + p+=(argc+1); +#endif + + /* skip over the environment, find the name of the PE file */ + while( *p ) + { + const char pefile[] = "WINEPEEXEPATH="; + if( 0 == memcmp( pefile, *p, sizeof pefile - 1 ) ) + pe_file = (*p) + sizeof pefile - 1 ; + p++; + } + + av = (ElfW(auxv_t)*) (p+1); + wld_get_auxilary( av, AT_PAGESZ, &dl_pagesize ); +#ifdef DUMP_AUX_INFO + wld_dump_auxilary( av ); +#endif + + /* reserve memory that Wine needs */ + vars.pe_size = 0x01000000; + vars.pe_address = (void*)0x00400000; + wld_reserve_wine_memory( pe_file, &vars.pe_address, &vars.pe_size ); + mmap( vars.pe_address, vars.pe_size, PROT_NONE, + MAP_FIXED|MAP_PRIVATE|MAP_ANON, -1, 0); + + /* reserve the memory that wine needs for the global heap */ + vars.heap_size = 0x400000; + vars.heap_address = (void*)0x65430000; + mmap( vars.heap_address, vars.heap_size, PROT_NONE, + MAP_FIXED|MAP_PRIVATE|MAP_ANON, -1, 0); + + /* load the original dynamic loader itself */ + fd = open( LDSO_NAME, O_RDONLY, 0); + if( fd < 0 ) + { + wld_printf( "error: fd = %x\n", fd); + exit(1); + } + read( fd, buf, sizeof buf ); + start = wld_map_so( fd, buf, sizeof buf, &map ); + if( !start ) + exit(0); + close( fd ); + + /* + * Pass information to the wine loader in a static variable + * so it doesn't get destroyed after we leave this stack frame + */ + wld_get_auxilary( av, AT_PHDR, (int*)&elf_phdr ); + wld_get_auxilary( av, AT_PHNUM, (int*)&phnum ); + wine_preload_vars = wld_find_symbol( elf_phdr, phnum, "wine_preload_vars" ); + if( wine_preload_vars ) + { + unsigned char *base = (void*)0x7ff81000; /* FIXME - define properly */ + vars.preload_base = (void*) base; + vars.preload_size = (size_t) (_end - base); + *wine_preload_vars = &vars; + } + + wld_set_auxilary( av, AT_BASE, map.l_addr ); + +#ifdef DUMP_AUX_INFO + wld_printf("New auxillary info:\n"); + wld_dump_auxilary( av ); + wld_printf("jumping to %x\n", start); +#endif + + return start; +} + +/* + * The _start function is the entry and exit point of this program + * + * It calls wld_start, passing a pointer to the args it receives + * then jumps to the address wld_start returns after removing the + * first argv[] value, and decrementing argc + */ +void _start(void); +__asm ( + ".align 4\n" + "\t.global _start\n" + "\t.type _start,@function\n" +"_start:\n" + "\tcall wld_start\n" + "\tpush %eax\n" + "\txor %eax,%eax\n" + "\txor %ebx,%ebx\n" + "\txor %ecx,%ecx\n" + "\txor %edx,%edx\n" + "\tret\n" +); +

1 0

Re: Resend - Unitialised EDITSTATE
by Alexandre Julliard May 24, 2004

May 24, 2004

jeffl <jeffl(a)defcen.gov.au> writes: > People, when trying to run DebugApiSpy.exe in encounter a divide by > zero. It seems there is a circumstance where the EDITSTATE block can be > unitialised. I have not found this yet and I am not sure how to track > it down. This patch fixes the instance where the problem occurs for me. That's not a fix, that's just hiding the problem. You need to figure out why the value was not initialized at this point, it's probably a message order thing. -- Alexandre Julliard julliard(a)winehq.org

1 0

Calendars
by William Lahti May 24, 2004

May 24, 2004

I am trying to implement EnumCalendarInfo and would like to know if any locales have multiple calendars, where the settings for the extra calendars would be in the nls file, and if any calendars have their own seperate file.

3 3

additional download for winehq download page
by Ivan May 24, 2004

May 24, 2004

This patch adds lots of download to the winehq download page. Sending only to wine-devel for now, as I'm not sure they are all needed, but they could all be useful in some situation or another. Comment welcome. Ivan. p.s. Maybe we should have a download page just for m$ components?

3 2

Re: additional download for winehq download page
by Ivan May 24, 2004

May 24, 2004

> The link to the images is broken, it is winedows.gif when it should be > windows.gif. Yes, I've just noticed that. Ivan.

1 0

Re: Check that at least one drive is defined on startup
by Alexandre Julliard May 24, 2004

May 24, 2004

Mike Hearn <mh(a)codeweavers.com> writes: > Mike Hearn <mh(a)codeweavers.com> > Add a check to process_init() to ensure at least one drive is defined, > as things break if that isn't the case. There is no need to define the Drives registry key, everything should work fine without it. If something doesn't, that's a bug. -- Alexandre Julliard julliard(a)winehq.org

2 3

Re: Split the dsound test
by Robert Reif May 24, 2004

May 24, 2004

Francois Gouget wrote: > > You will notice the following problems: > * the test takes awfully long (>6 minutes here). It loops through the > known formats and for each of them it plays, in turn, the regular > test, a test through a duplicated buffer, with a listener, etc. > * as a result of the above, a user listening to the test does not > know what to expect, which makes it hard to know whether the sound was > played correctly. The corresponding Microsoft test takes hours to complete. If time is an issue, then the test should be split up into a quick test and an exhaustive test. Also, volume and pan tests on primary and secondary buffers is not the same thing as 3D. They just happen to be similar now because we don't support hardware acceleration yet.

2 2

Re: Split the dsound test
by Robert Reif May 24, 2004

May 24, 2004

Francois Gouget wrote: >+ >+#define NB_FORMATS 24 >+extern const unsigned int formats[NB_FORMATS][3]; >+ > > Please don't do this. It is desirable to be able to add or comment out formats easily when using this test as a debugging and development tool.

1 0

failure report: Astrum installer
by Dan Kegel May 23, 2004

May 23, 2004

Looking through the debug log for my failed attempt to install picasa, I saw that the installer it uses is called Astrum; see http://www.thraexsoftware.com/aiw/ A free demo of that installer is downloadable from http://www.thraexsoftware.com/download/aiw.exe Predictably, it doesn't run under Wine, either; it also puts up a dialog box at the beginning (which unhelpfully doesn't show any text), and if you click "ok", it just puts up the same dialog box again. I tried for a while to figure out WINEDEBUG settings that would provide some illumination about what's wrong. Best I came up with was WINEDEBUG=warn+all,+dialog,+text wine aiw.exe > log 2>&1 in hopes I could tell what the missing text was, but no dice. Given that this is a free-ish installer, and more apps might start using it than just Picasa, maybe it's worth looking into... - Dan -- My technical stuff: http://kegel.com My politics: see http://www.misleader.org for examples of why I'm for regime change

1 0

Re: [ros-cvs] CVS Update: reactos
by Steven Edwards May 23, 2004

May 23, 2004

Hello Dmitry, --- Dmitry Timoshkov <dmitry(a)baikal.ru> wrote: > Yes, Wine is not perfect. No offence intended, but when ReactOS will > be able > to run at least MS Office XP then you could start comparing our > implementations > and state that Wine is not perfect in comparison. Right now I'm > afraid that > comparison is not on the ReactOS side, and just not using Wine source > is plain > wrong and just a big mistake. I agree with you and wish we could share more. I discussed abstracting WINE user32/gdi32 and creating a Win32k.sys implementation for Wine so we could try to hide the x11drv behind it but at the time Alexandre does not seem very interested in the idea. The way it stands right now while WINE is more correct in its functionality its implementation is not 100% correct for the design we are trying to replicate. If we can change this in WINE, then sure lets share even more and stop duplicating work. Thanks Steven __________________________________ Do you Yahoo!? Yahoo! Domains � Claim yours for only $14.70/year http://smallbusiness.promotions.yahoo.com/offer

2 1

Re: Fix elf_read_wine_loader_dbg_info() declaration for the !__ELF__ case
by Eric Pouech May 23, 2004

May 23, 2004

> -unsigned long elf_read_wine_loader_dbg_info(struct process* pcs) > +BOOL elf_read_wine_loader_dbg_info(struct process* pcs) > { > return -1; > } we should return FALSE too

1 0

tlhelp32 problem
by andrejoh＠c2i.net May 23, 2004

May 23, 2004

Hi! While trying to get Anarchy Online working with Wine, I discovered that dynamically getting symbols from the running program does not work properly in Wine. The attached example code shows the problem; when running in Windows, I get this output The mangled name is "?Instantiate(a)MyTestClass@@SAPAV1(a)PAVObjectArchive_c@fun@@@Z" Searching in "ntdll.dll" Searching in "kernel32.dll" Searching in "Factory.dll" Searching in "MSVCP60D.dll" Searching in "MSVCRTD.dll" Searching in "Test.dll" Found the function at 00321023 Creating a new object! The returned text is "Found the right one!" whereas when running this in Wine, this is the output [andrej(a)rivendell tst]$ PATH=$PATH:$PWD wine tst The mangled name is "?Instantiate(a)MyTestClass@@SAPAV1(a)PAVObjectArchive_c@fun@@@Z" Searching in "" Searching in "" Searching in "" Searching in "" Searching in "" Searching in "" Searching in "" Searching in "" The returned text is "<function for "MyTestClass" not found>" If the test for a ".dll" ending is removed (as can be seen in the output, the name returned is empty, so it will always fail), this is the output [andrej(a)rivendell tst]$ PATH=$PATH:$PWD wine tst The mangled name is "?Instantiate(a)MyTestClass@@SAPAV1(a)PAVObjectArchive_c@fun@@@Z" Searching in "" Searching in "" Found the function at 0x408a9023 Creating a new object! The returned text is "Found the right one!" (If anybody needs the compiled code, and/or MSDS project files, please contact me directly.) -- Try Anarchy Online http://www.anarchy-online.com/content/downloads/tryout/ Netiquette guidelines ftp://ftp.rfc-editor.org/in-notes/rfc1855.txt How to quote properly http://www.i-hate-computers.demon.co.uk/ Hvordan sitere riktig http://home.online.no/~vidaandr/news/OBSquoting.html

2 3

great improvements!
by Marius Andreiana May 23, 2004

May 23, 2004

Hi I've been using wine for a long time (without any MS Windows dll). I've upgraded it to the latest version rpm once I installed Fedora Core 2 and I've happily noted these improvements: * /etc/wine for global config options * automatic creation of ~/.wine (drives, C:) There are still some fedora core issues, already in bugzilla. As the roadmap says, would be nice to have stable releases which focus on running by default (without any tweaks) common software, such as IE and Photoshop. (like crossweaver office) Thanks! -- Marius Andreiana Galuna - Solutii Linux in Romania http://www.galuna.ro

1 0

Unhandled exception
by RT May 22, 2004

May 22, 2004

I received the below message on FreeBSD 5.2.1-RELEASE, Wine 20040505 while playing Total Annihilation (non-demo). err:seh:EXC_DefaultHandling Unhandled exception code c0000005 flags 0 addr 0x0 #0 wine_get_thread_data () at pthread.c:534 534 return NtCurrentTeb()->pthread_data; (gdb) backtrace #0 wine_get_thread_data () at pthread.c:534 #1 0x00111e09 in __pthread_thread_self () at kthread.c:106 #2 0x00111e38 in __errno_location () at kthread.c:121 #3 0x00111e48 in __error () at kthread.c:124 #4 0x2020e406 in .cerror () from /lib/libc.so.5 #5 0x00111f19 in cleanup_thread () at kthread.c:198 -- Rod Taylor <rbt [at] rbt [dot] ca> Build A Brighter Lamp :: Linux Apache {middleware} PostgreSQL PGP Key: http://www.rbt.ca/signature.asc

1 0

Re: Building wine on darwin (resend) (3/6)
by Alexandre Julliard May 22, 2004

May 22, 2004

emmanuel maillard <e.rsz(a)libertysurf.fr> writes: > changelog > Add darwin/ppc support in THREAD_InitStack I'm not sure what the problem is, but the fix doesn't look right at all. Tib.StackBase should point to the base of the stack as it has been allocated, no matter how it's used later on. Most likely you need to fix wine_switch_to_stack or something like that. -- Alexandre Julliard julliard(a)winehq.org

4 4

Fwd: TGMarlett
by David Hammerton May 22, 2004

May 22, 2004

Thanks Steven, However - just a thought here.. Do we want to simply work with the '.ttf' file, or should we keep the 'source' in CVS also? The 'source' file is a '.sfd' file created with PfaEdit [http://pfaedit.sourceforge.net/oldindex.html] now called FontForge [http://fontforge.sourceforge.net/] I'm not really familiar with what extra data the '.sfd' file may have compared to the '.ttf' that may be useful. Is anyone familiar with this stuff? Is the '.sfd' pretty much useless when we have the '.ttf'? Either way, both can be found here (although the '.ttf' Steven sent is probably better, as it contains those fixes he mentions): http://craz.net/files/ Cheers, David Begin forwarded message: > From: Steven Edwards <steven_ed4153(a)yahoo.com> > Date: May 19, 2004 9:21:18 AM GMT+10:00 > To: wine-patches(a)winehq.com > Subject: TGMarlett > Reply-To: wine-devel(a)winehq.org > > Hello, > This is a update to the Marlett replacement developed by TransGamming. > It is needed to get Steam working on Wine. Sylvain Petreolle has hacked > it a bit so it will now load on Windows as well as ReactOS and Linux > with FreeType. The older TGMarlett did not work properly on WindowsXP. > > Thanks > Steven > > > > > __________________________________ > Do you Yahoo!? > SBC Yahoo! - Internet access at a great low price. > http://promo.yahoo.com/sbc/ > -- David Hammerton david(a)craz.net http://crazney.net/ Welfare *NOT* Warfare

1 0

symbol hiding - with gcc -fvisibility
by Guido Draheim May 21, 2004

May 21, 2004

sorry, just saw the note on WWN. I did hope that there would be someone to point to the gcc visibility stuff already - but it seems that people are more concerned to exchange their believes and politics rather showing off some engineering stuff. * preface: it's not in the file format - it's in the tools. The name ELF stands for Extensible Linker Format, it allows to invent all kind of new sections, and expand the symbol table with many interesting attributes that can be checked up by a toolchain. Well, "IF" the tools check for them. * gcc -fvisibility The gcc -fvisibility exists since 3.1 and is supported by binutils. If you read up some web reports, it seems to have been invented for hiding stuff not quite unlike what the original poster wanted. The icc supports all -fvisibility options as well afair. * the real problem traditionally, one can kill local systems which are usable within one file only, i.e. "static". Pretty 20year old unix stuff. The other symbols are referencable outside (in .o) and the standard shard linker will put all of them into the shared library symbol export table. - But now we want two different symbol flavours, one exported to the same library, the others for the rest of the world. Perhaps use a linker script to decide later about symbol visibility. * example solution gcc supports the universal attribute syntax, and it does now know about more symbol flavours as to their visibility. Probably you want to have "hidden". Attached are two simple test*.c files and a makefile. The final sharedlib symbol table contains only test3 after stripping. - using a linker script instead of in-source __attribute__ is left as an excercise to the reader. * more hints search freshmeat - there are ELF tools to staticlink, to obfuscate, and more. It's all in the tools. There is only one problem with opensource: many programmers do not need to hide symbols to the extreme, so they did not write such tools so far. Not yet. There are some real ELF-educated people out there - hire one, and let him write the tool you need. The ELF format is well-defined and simple enough to allow everything and quite easily. Do ut des. -- cheers, guido http://google.de/search?q=guidod extern void test2(void); void __attribute__((visibility("hidden"))) test1(void) { test2(); //__attribute__((visiblity("internal"))) } extern void test1(void); void __attribute__((visibility("hidden"))) test2(void) { test1(); } void test3(void) { test2(); } test.so : test1.c test2.c Makefile gcc -Wl,-x -shared -o $@ test1.c test2.c strip -x $@ objdump -t $@ | grep "test[123]" gcc -Wl,-x -shared -o test.so test1.c test2.c strip -x test.so objdump -t test.so | grep "test[123]" 00000631 g F .text 0000000d test3

3 3

WINE filesystem idea
by Jonathan Wilson May 21, 2004

May 21, 2004

Basicly, when emulated windows version is set to NT4, 2000, XP or 2003, WINE should pretend to apps that its running on NTFS. If running as 95,98 or ME, it should pretend to apps that its running on FAT32. Or better yet, add an option for this (with my suggestions being the default) Just an idea :)

2 1

Wine on AMD64 (heap problem)
by Maurizio Monge May 21, 2004

May 21, 2004

Hello, i compiled wine cvs on my amd64 without many problem (i had to set PATH to point to a directory with wrappers for gcc,ld and as that would call respectively "gcc -m32","ld -melf_i386","as --32", it would be nice if wine would support natively this or would support LDFLAGS=-melf_i368.., but this is not the point). When i started wine i got: wine: failed to create the process heap The same wine executable was running ok on my 32bits distro. I read that this happens on fedora2 too because of nx stack, and i think that, even if i do not have SeLinux installed, the stack is not executable by default on amd64... Do you have any idea of how this could be fixed/hacked? Thanx Maurizio I am not subscribed to wine-devel, so please CC monge(at)sns(dot)it

4 7

Regression with Blizzard patching software
by Aneurin Price May 21, 2004

May 21, 2004

I recently discovered that the software used by Blizzard to patch Diablo 2 and Warcraft 3 (and possibly the same version used with other games), has stopped working. It starts do do something, but before actually applying the patch it crashes out, with the final error message being: Unhandled exception: page fault on write access to 0x00000000 in 32-bit code (0x408ba2e0). In 32-bit mode. 0x408ba2e0 (DIALOG_CreateIndirect+0x300 [dialog.c:636] in USER32.DLL): movl $0x0,0x0(%eax) 139 return TRUE; Wine-dbg>WineDbg terminated on pid a The whole of the scrollback buffer is filled with the usual messages about loading debug information. I could run it with script to see what's happening before that, but I don't think it's really necessary: Since this *used* to work, I've gone through some CVS regression testing and determined that it stopped working sometime between 2004-04-19 22:00:00 and 2004-04-19 22:03:00 CST, so I'm willing to bet that this is what changed: http://www.winehq.org/hypermail/wine-cvs/2004/04/0272.html Since I have no understanding of what's going on here I really coudn't suggest a fix, but I'm happy to provide any further information I can and thought this would be worth pointing out. Aneurin Price

4 3

Re: loader fix for dbghelp
by Alexandre Julliard May 20, 2004

May 20, 2004

Eric Pouech <pouech-eric(a)wanadoo.fr> writes: > I fixed that by sending (in ntdll loader code) the DLL load event a > bit later in the loading processus, when we're out of the dlopen call. This will break dlls that are loading implicitly by ELF loading. That stuff really has to be done in the constructor. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: Empty (Stubs in spec only) d3drm
by Ivan Leo Murray-Smith May 20, 2004

May 20, 2004

What app do you know about that uses this dll and works with the native version? All the DX7SDK samples fail with the native version. Ivan.

1 0

Fedora Core
by gslink May 20, 2004

May 20, 2004

I notice that a problem has been detected with Fedora Core 2 but has anyone checked Core 3? That is the one in directory 1.92.

1 0

x - D3D Errors while starting
by Florian Goth May 20, 2004

May 20, 2004

I tried to start The Game x and it bailed out with a lot of fixme's and some errors like these: err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_ZVISIBLE (0000001e) value : 00000000 ! The game uses DiectX6 There's a demo available at http://www.egosoft.com/x/download/index.htm This is the complete log: ixme:ddraw:Main_DirectDraw_SetCooperativeLevel (0x40434ab0)->(00020022,00000013) fixme:x11drv:X11DRV_desktop_SetCurrentMode Cannot change screen BPP from 32 to 16 fixme:ddraw:DIB_DirectDrawSurface_Blt dwFlags DDBLT_WAIT and/or DDBLT_ASYNC: can't handle right now. fixme:ole:CoCreateInstance no instance created for interface {279afa83-4981-11ce-a521-0020af0be560} of class {d8f1eee0-f634-11cf-8700-00a0245d918b}, hres is 0x8000ffff fixme:dsound:IDirectSoundImpl_SetCooperativeLevel level=DSSCL_PRIORITY not fully supported err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_LINEPATTERN (0000000a) value : 00000000 ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_MONOENABLE (0000000b) value : 00000000 ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_ROP2 (0000000c) value : 0000000d ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_PLANEMASK (0000000d) value : ffffffff ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_LASTPIXEL (00000010) value : 00000001 ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_ZVISIBLE (0000001e) value : 00000000 ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_EDGEANTIALIAS (00000028) value : 00000000 ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_RANGEFOGENABLE (00000030) value : 00000000 ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_EXTENTS (0000008a) value : 00000000 ! err:ddraw:set_render_state Unhandled dwRenderStateType D3DRENDERSTATE_VERTEXBLEND (00000097) value : 00000000 ! fixme:amstream:AM_create ((nil),0x4095f5ec) fixme:amstream:IAMMultiMediaStreamImpl_QueryInterface (0x404ddbf0/0x404ddbf0)->({bebe595c-9a6f-11d0-8fde-00c04fd9189d},0x439644d0) fixme:amstream:IAMMultiMediaStreamImpl_AddRef (0x404ddbf0/0x404ddbf0) fixme:amstream:IAMMultiMediaStreamImpl_Release (0x404ddbf0/0x404ddbf0) fixme:amstream:IAMMultiMediaStreamImpl_Initialize (0x404ddbf0/0x404ddbf0)->(0,1,(nil)) stub! fixme:amstream:IAMMultiMediaStreamImpl_AddMediaStream (0x404ddbf0/0x404ddbf0)->((nil),0x495398,0,(nil)) stub! fixme:amstream:IAMMultiMediaStreamImpl_GetFilterGraph (0x404ddbf0/0x404ddbf0)->(0x439645e0) stub! fixme:quartz:QUARTZ_DllGetClassObject {4a2286e0-7bef-11ce-9bd9-0000e202599c}: no class found. fixme:ole:CoCreateInstance no classfactory created for CLSID {4a2286e0-7bef-11ce-9bd9-0000e202599c}, hres is 0x80040111 fixme:amstream:IAMMultiMediaStreamImpl_OpenFile (0x404ddbf0/0x404ddbf0)->(0x4095f694,0) stub! fixme:amstream:IAMMultiMediaStreamImpl_GetMediaStream (0x404ddbf0/0x404ddbf0)->(0x495398,0x439644e4) stub! wine: Unhandled exception (thread 0009), starting debugger... err:seh:start_debugger Couldn't start debugger ("debugger/winedbg 8 224") (2) Read the Wine Developers Guide on how to set up winedbg or another debugger fixme:winmm:MMDRV_Exit Closing while ll-driver open

2 1

crosstest seems to be broken
by Kevin Koltzau May 19, 2004

May 19, 2004

I noticed my crosstest build last night failed on module.cross.o(.text+0x713): In function `testGetModuleFileName': /home/wine/wine/dlls/kernel/tests/module.c:63: undefined reference to `wine_dbgstr_wn' module.cross.o(.text+0x727):/home/wine/wine/dlls/kernel/tests/module.c:63: undefined reference to `wine_dbgstr_an' looks like this patch introduced the problem http://cvs.winehq.org/patch.py?id=12409 I won't have time to work on this in the next few days..so if someone could take a look at this I'd be most appreciative

1 0

Bug in EndDialog
by Santosh Siddheshwar May 19, 2004

May 19, 2004

Hi, I think there may be a bug in EndDialog. Attached is a simple app which work fines on Windows but hangs on Linux. Follow the steps below to reproduce the problem. 1. Create the Sample2.exe from the attached project. 2. Run Sample2.exe 3. Click on Help | About menu item 4. Click on the OK button 5. App Hangs. Any clues? Regards Santosh Siddheshwar <<Sample2.zip>> ********************************************************************* Disclaimer: The information in this e-mail and any attachments is confidential / privileged. It is intended solely for the addressee or addressees. If you are not the addressee indicated in this message, you may not copy or deliver this message to anyone. In such case, you should destroy this message and kindly notify the sender by reply email. Please advise immediately if you or your employer does not consent to Internet email for messages of this kind. *********************************************************************

1 0

some ideas for new janitorial projects
by Eric Pouech May 19, 2004

May 19, 2004

I've been toying with some gcc warnings lately, and this could generate some new ideas for janitorial tasks. For example: o -Wmissing-declarations should point lots of missing APIs declarations (I've done it in some cases, and I'm in the process of submitting some fixes). We cannot turn this warning always on because there are lots of cases where we have global functions without prototypes (most than 2/3 of them are made out of the 16bit entry points) o -Wcast-qual should point out all the (const bla*) => (foo*) casts. We use quite a bit this, and in some cases it would prevent some errors. These are the main cases pointed out by this warning: 1/ const void* ptr; and then reading a value from ptr (word...) like *(WORD*)ptr which should be written *(const WORD*). The warning in this case is harmless. 2/ const void* ptr; and then setting it to another const pointer: const WORD* pw = (WORD*)ptr; which should be written pw = (const WORD*)ptr;. This warning is harmless if pw is really defined as const, in some cases it isn't and this should be fixed. 3/ const void* ptr; and then setting it to a pointer to a pointer (used a lot for qsort/bsearch... callbacks), when dealing with arrays of pointers. Here again, what's const is the first pointer, so const foo* f = *(const foo**)ptr is wrong, it should be const foo* f = (const foo* const*)ptr; This could be harmfull if not declared properly. Unfortunately, we cannot turn this warning on all the time because some C functions implementation would always trigger it (strstr for example), unless we use intergral values (not pointer) to cast from the const char* to the returned char*), and this is uglier IMO than the warning we try to avoid. Some others warnings could be used as well. Trying also the Intel compiler gave lots of interesting warnings (and a tons of not so usefull too). A+

2 1

Re: RE: [DMUSIC] Happy Happy :)
by fenix＠club-internet.fr May 19, 2004

May 19, 2004

>Raphael wrote: >> >> Hi, >> >> I'm glad to announce that now (with this little patch) Unreal2 >> is working >> and playable (it's only slow and i havent seen graphics problems for now). >> >> Now play :) >> >> Changelog: >> - better handling of User IDirectMusicTools > >You don't seem to free the items of the list anywhere, so this patch will >cause a memory leak. Yes, i know that but the DMPerformance (and the default Graph assiocated) are sessions objects (they are only deleted on program exit) Most of the dmusic objects must be cleaned (released) at exit, It will be done when dmusic will be more complete >Rob Regards, Raphael

1 0

[DMUSIC] Happy Happy :)
by Raphael May 19, 2004

May 19, 2004

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, I'm glad to announce that now (with this little patch) Unreal2 is working and playable (it's only slow and i havent seen graphics problems for now). Now play :) Changelog: - better handling of User IDirectMusicTools Regards, Raphael -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) iD8DBQFAqxTCp7NA3AmQTU4RAoEgAJ48qn95odkW2085hZ+6FQm5y+6kYgCfQSPY KLHJE3RjMDyZmFcDDdZsQ/s= =OP9b -----END PGP SIGNATURE-----

2 1

Re: Exporting symbols
by Luchezar Belev May 18, 2004

May 18, 2004

-- Dan Timis wrote: >Is there any way to export only the functions that are called by the >code we link to, without exposing the internal workings? Ideally we >would like to be able to provide a static library created with ar or >ranlib, link that library with other object files to create a .exe.so >for wine, but when you do "nm" on the resulting .exe.so you would not >see all the inner workings of our library. > >Is that possible? try the "-s" linker option Lucho. __________________________________ Do you Yahoo!? SBC Yahoo! - Internet access at a great low price. http://promo.yahoo.com/sbc/

1 0

Re: [Resend] wininet patch
by Alexandre Julliard May 18, 2004

May 18, 2004

Tim Hentenaar <tth(a)one.net> writes: > /*********************************************************************** > * FtpGetCurrentDirectoryW (WININET.@) > * > - * Retrieves the current directory > + * Retrieves the current directory (unicode) > + * Implemented 5/17/2004 Tim Hentenaar <tth(a)one.net> Please don't add changelog information in the code, that's what the CVS log and the Changelog file are for. -- Alexandre Julliard julliard(a)winehq.org

3 2

E-mail Virus
by Abby Ricart May 18, 2004

May 18, 2004

Is it just me, or is some one posing as julliard(a)winehq.org and dmitry(a)baikal.ru sending e-mail virii into the mailing list? My ISP filtered it before I could get it, and it's a Windows virus so I doubt it could do very much damage. The virus is described as W32.Beagle.X(a)mm. Someone is being rather childish. The subject lines were: "Encrypted Document" "Hidden Message" and "RE: Text Message"

4 3