[snip] That's what I thought. At some stage I'll submit my patch to vsti which makes this change there as well. It seems that the VST spec is not very specific about which functions can be called in what threads. Unfortunately not. I'm still getting the same segfault which I reported to this list on Thu, 27 May 2004 under the thread "jack_fst segfault". If there's anything I can do to help solve the jack_fst problems, please let me know. Thanks Carl

On Wed, 2 Jun 2004, Mark Knecht wrote: Hi Mark, I'm sure it's possible, but perhaps vsti doesn't support it. Shouldn't be a problem to add, though. I'll have a play with the demo some time. I've been meaning to try it anyway, as it sounds really cool. Cheers Carl

On Wed, Jun 02, 2004 at 04:09:35PM +0100, Chris Cannam wrote: i will have a go and implement this into galan. what is interresting is which thread calls these callbacks. if its the audiothread its really no problem to implement this. -- torben Hohn http://galan.sourceforge.net -- The graphical Audio language

On Wed, Jun 02, 2004 at 10:19:40AM -0700, Mark Knecht wrote: urgh... go down to an older version of wine. i will get this wine and fix libfst... hmm... strange the number of arguments for wine_init did not change. i am currently compiling wine and will try to build fst with it. your setup could be too recent. i will report back... -- torben Hohn http://galan.sourceforge.net -- The graphical Audio language

On Wed, Jun 02, 2004 at 11:21:29AM -0700, Mark Knecht wrote: how does /usr/local/include/wine/library.h declare wine_init ? here its extern void wine_init( int argc, char *argv[], char *error, int error_size ); and in the new version i dled its the same. this must be due to the patches you applied to wine. -- torben Hohn http://galan.sourceforge.net -- The graphical Audio language

On Thursday 03 June 2004 19:33, Mark Knecht wrote: my changed libwinelib.c looks like this: int SharedWineInit (void (*sighandler)(int,siginfo_t*,void*)) { unsigned char Error[2024]=""; char *WineArguments[] = {"sharedapp", LIBPATH "/libfst.so", NULL}; void* stackbase; size_t stacksize; void *ntdll; void *ntso; char ntdllpath[PATH_MAX+1]; char* dlerr; char* env[] = { NULL }; sharedwine_signal_handler = sighandler; if (setjmp (jump) == 0) { wine_init (2, WineArguments, env, Error, sizeof (Error)); ... hopefully this may work for you (at least it should compile). I did not get fst working for me. It always crashes in wine_init() for all tools (like fstconfig). The core shows an destroyed stack and does not give useful information (crash in lib function opendir()). I tried different winelib versions, compiler(gcc3.3/3.4), linux versions (2.4/2.6) with no luck so far :-( /werner

On Thursday 03 June 2004 20:52, torbenh(a)gmx.de wrote: its from cvs, last updated yesterday (03.06.2004). i tried again with wine-20040505. Calling fstconfig produces a core dump. Gdb says: (gdb) bt #0 __pthread_sighandler (signo=11, ctx= {gs = 0, __gsh = 0, fs = 51, __fsh = 0, es = 123, __esh = 0, ds = 123, __dsh = 0, edi = 131120, esi = 0, ebp = 1080816716, esp = 1080816708, ebx = 1075170628, edx = 0, ecx = 1, eax = 1082129376, trapno = 14, err = 4, eip = 1075232782, cs = 115, __csh = 0, eflags = 2163206, esp_at_signal = 1080816708, ss = 123, __ssh = 0, fpstate = 0x0, oldmask = 2147483648, cr2 = 1082129788}) at sighandler.c:28 #1 <signal handler called> #2 0x4016c00e in __pthread_internal_tsd_get (key=1077936127) at specific.c:225 #3 0x400a434e in __libc_calloc (n=1, elem_size=1077936127) at malloc.c:3585 #4 0x400d7813 in __opendir (name=0x20000 <Address 0x20000 out of bounds>) at ../sysdeps/unix/opendir.c:146 #5 0x40315820 in wine_nt_to_unix_file_name (nameW=0x406bfb58, unix_name_ret=0x406bfb50, disposition=1, check_case=0 '\0') at directory.c:873 #6 0x403164e0 in RtlDoesFileExists_U (file_name=0x40380750) at directory.c:1215 #7 0x40328273 in RtlDosSearchPath_U (paths=0x40380510, search=0x406bfe08, ext=0x0, buffer_size=64, buffer=0x406bfd2c, file_part=0x406bfc68) at path.c:428 #8 0x40320e29 in load_dll (load_path=0x40380510, libname=0x406bfe08, flags=0, pwm=0x406bfe00) at loader.c:1359 #9 0x40320093 in fixup_imports (wm=0x403806c8, load_path=0x40380510) at loader.c:384 #10 0x40322262 in LdrInitializeThunk (main_file=0x0, unknown2=0, unknown3=0, unknown4=0) at loader.c:1930 #11 0x4051296b in start_process (arg=0x0) at process.c:822 its essential the same as in the current wine cvs version. My system uses glibc 2.3.2, linux 2.6.7-rc1 (no nptl). /werner

On Friday 04 June 2004 14:18, torbenh(a)gmx.de wrote: i do not remember the configure options for my glibc 2.3.3. I replaced it this weekend with a brand new version 2.3.4 (nptl). And it now works! So i believe it was a glibc problem on my system. Googling around i found the following information about glibc and %gs usage in the thread implementation: (http://www-124.ibm.com/developerworks/oss/jikesrvm/userguide/HTML/faq_build…): "In order for our m-to-n threading scheme to work, we need to be able to swap stacks for a single pthread. This only works if glibc is built with an option to use the GS segment register to address certain pthread-specific state. Otherwise, glibc attempts to use stack-pointer arithmetic to find pthread-specific state, which fails once we bash the stack. On Red Hat? Linux and SuSE® Linux, glibc is built with this option by default. On Debian® GNU/Linux, as of this writing (6 May 2004), the version of x86 libc6 in the Debian Woody distribution does not use GS-segment addressing. Debian/Sarge will use GS-segment addressing on 2.6 Linux kernels, but will not use it on 2.4 and 2.2 kernels." /werner

torbenh(a)gmx.de wrote: I'm still broken. I had unmerged wine earlier mark@flash mark $ qpkg -I | grep wine mark@flash mark $ I removed (rm -rf) my existing wine-cvs directory and then downloaded a new copy from WineHQ's cvs server using: mark@flash mark $ export CVSROOT=:pserver:cvs@cvs.winehq.org:/home/wine mark@flash mark $ cvs login cvs mark@flash mark $ cvs -z 0 checkout wine mark@flash mark $ cd wine/ mark@flash mark $ tools/wineinstall Wine builds and gets installed. My directory is new: mark@flash mark $ cd /usr/local/include/ mark@flash include $ ls -la total 24 drwxr-xr-x 5 root root 4096 Jun 3 11:49 . drwxr-xr-x 9 root root 4096 May 26 09:37 .. drwxr-xr-x 2 root root 4096 May 15 17:02 lscp drwxr-xr-x 2 mark users 4096 May 20 15:30 vst -rw-r--r-- 1 root root 3174 May 26 13:09 vstlib.h drwxr-xr-x 4 root root 4096 Jun 3 11:49 wine library.h is new: mark@flash wine $ ls -la total 64 drwxr-xr-x 4 root root 4096 Jun 3 11:49 . drwxr-xr-x 5 root root 4096 Jun 3 11:49 .. -rw-r--r-- 1 root root 8467 Jun 3 11:49 debug.h -rw-r--r-- 1 root root 6238 Jun 3 11:49 exception.h -rw-r--r-- 1 root root 8572 Jun 3 11:49 library.h drwxr-xr-x 3 root root 4096 Jun 3 11:49 msvcrt -rw-r--r-- 1 root root 9541 Jun 3 11:49 unicode.h drwxr-xr-x 2 root root 8192 Jun 3 11:49 windows mark@flash wine $ The contents are wrong: extern void wine_init( int argc, char *argv[], char *envp[], char *error, int er ror_size ); What am I doing wrong? Me too! Thanks, Mark

torbenh(a)gmx.de wrote: <SNIP> <SNIP> Torben, Please use it where? I think I am using wine-20040505 mark@flash mark $ wine --version Wine 20040505 mark@flash Do you want me to do something different when checking out Wine from CVS? If so, please give me the instructions and I'll give it a try. Thanks, Mark

torbenh(a)gmx.de wrote: Yes, the guys on the Wine Users lists told me much the same thing yesterday, but they also showed me that doing cvs checkout -r Wine-20040505 wine will get me the official revision that you wanted me to use. There's also the ability to get it by date: cvs update -D "2004-05-05 20:00:00 CDT" if you know the date and time you are looking for. THE GOOD NEWS: jack_fst now compiles as you said it would. (I got a little help from Jesse off list on the VST .h file. Thanks Jesse! I also think Frank's instructions are probably good too.) THE BAD NEWS: Since it was a new Wine version I had to reinstall Groove Agent before I could try it out again. Unfortunately GS does not install with wine-20040505, but does install with the most up to date CVS version. I think part of this may be the Wine config file so I'll look at that more today, but as of right now GA doesn't install so I need to figure out how to get around that. Thanks for your help! Cheers, Mark

On Wed, Jun 02, 2004 at 02:57:43PM +0100, Carl Hetherington wrote: right... such things are discussed on vst-plugins list.... :( the reply said "you have them" but i get the same output from my libc. this looks like a tls problem. i would try to change the libc version. try to get a libc without tls or a new gcc. but i am not sure. i had similar thread and wine related problems with libfst. we override the threading implementation wine uses in libfst. so it could be that we dont support this tls stuff properly. i have different things to do right now, so i wont dig in this threading code. can you trace the location where the segfault is happening ? is it in fst_init ? or later ? -- torben Hohn http://galan.sourceforge.net -- The graphical Audio language