wine-devel

wine-devel@winehq.org

8 participants
84486 discussions

RtlAllocateHeap crash
by Dimitrie O. Paun 21 Oct '02

21 Oct '02

While I was playing around with Newsbin Pro, I stumbled over this: Unhandled exception: page fault on write access to 0x0000000c in 32-bit code (0x400c2eaa). In 32-bit mode. 0x400c2eaa (RtlAllocateHeap+0xda [heap.c:1039] in libntdll.dll.so): movl %eax,0xc(%edx) 1040 pArena->prev->next = pArena->next; Is it worth investigating? -- Dimi.

1 0

Re: WebBrowser
by Dimitrie O. Paun 21 Oct '02

21 Oct '02

On October 19, 2002 03:27 am, Malte Starostik wrote: > ChangeLog > * Fixed IWebBrowser::Navigate() parameter type > * Added IWebBrowserApp and related typedefs BTW, any taker on implementing IWebBrowser & Co. in terms of the Gecko API, so we can get Mozilla working through Wine? -- Dimi.

3 4

Re: glibc 2.3 pthread patch
by Alexandre Julliard 20 Oct '02

20 Oct '02

Ove Kaaven <ovehk(a)ping.uio.no> writes: > Now that glibc 2.3 is available for those bleeding-edge folks on debian > unstable, it seems that Wines compiled on 2.2 won't run on 2.3 systems. > The glibc folks did warn us about this a while ago though, so here's a > patch that implements their suggestion for making wine compatible with as > many glibc versions as possible. Perhaps Alexandre wants to use his > WINE_GET_SONAME stuff to detect the name of the libc at configure time > instead of at runtime like I do, though. Funny, I was working on the same thing right now... But do you really need to fetch the real libc handle? It seems to me RTLD_NEXT should be enough. Here's how I did it: Index: scheduler/pthread.c =================================================================== RCS file: /opt/cvs-commit/wine/scheduler/pthread.c,v retrieving revision 1.26 diff -u -r1.26 pthread.c --- scheduler/pthread.c 12 Sep 2002 22:07:06 -0000 1.26 +++ scheduler/pthread.c 20 Oct 2002 21:45:13 -0000 @@ -38,13 +38,6 @@ #include "thread.h" #include "winternl.h" -static int init_done; - -void PTHREAD_init_done(void) -{ - init_done = 1; -} - /* Currently this probably works only for glibc2, * which checks for the presence of double-underscore-prepended * pthread primitives, and use them if available. @@ -62,28 +55,18 @@ asm(".globl " PSTR(alias) "\n" \ "\t.set " PSTR(alias) "," PSTR(orig)) -/* strong_alias does not work on external symbols (.o format limitation?), - * so for those, we need to use the pogo stick */ -#if defined(__i386__) && !defined(__PIC__) -/* FIXME: PIC */ -#define jump_alias(orig, alias) __ASM_GLOBAL_FUNC( alias, "jmp " PSTR(orig)) -#endif +static int init_done; -/* get necessary libc symbols */ -#if (__GLIBC__ == 2) && (__GLIBC_MINOR__ >= 1) && defined(HAVE___LIBC_FORK) -#define LIBC_FORK __libc_fork -#define PTHREAD_FORK __fork -#define ALIAS_FORK -#else -#define LIBC_FORK __fork -#define PTHREAD_FORK fork -#endif -extern pid_t LIBC_FORK(void); +static pid_t (*libc_fork)(void); +static int (*libc_sigaction)(int signum, const struct sigaction *act, struct sigaction *oldact); + +void PTHREAD_init_done(void) +{ + init_done = 1; + if (!libc_fork) libc_fork = dlsym( RTLD_NEXT, "fork" ); + if (!libc_sigaction) libc_sigaction = dlsym( RTLD_NEXT, "sigaction" ); +} -#define LIBC_SIGACTION __sigaction -extern int LIBC_SIGACTION(int signum, - const struct sigaction *act, - struct sigaction *oldact); /* NOTE: This is a truly extremely incredibly ugly hack! * But it does seem to work... */ @@ -255,15 +238,20 @@ } strong_alias(__pthread_atfork, pthread_atfork); -pid_t PTHREAD_FORK(void) +pid_t __fork(void) { pid_t pid; int i; + if (!libc_fork) + { + libc_fork = dlsym( RTLD_NEXT, "fork" ); + assert( libc_fork ); + } EnterCriticalSection( &atfork_section ); /* prepare handlers are called in reverse insertion order */ for (i = atfork_count - 1; i >= 0; i--) if (atfork_prepare[i]) atfork_prepare[i](); - if (!(pid = LIBC_FORK())) + if (!(pid = libc_fork())) { InitializeCriticalSection( &atfork_section ); for (i = 0; i < atfork_count; i++) if (atfork_child[i]) atfork_child[i](); @@ -275,9 +263,7 @@ } return pid; } -#ifdef ALIAS_FORK -strong_alias(PTHREAD_FORK, fork); -#endif +strong_alias(__fork, fork); /***** MUTEXES *****/ @@ -658,13 +644,20 @@ /***** ANTI-OVERRIDES *****/ /* pthreads tries to override these, point them back to libc */ -#ifdef jump_alias -jump_alias(LIBC_SIGACTION, sigaction); -#else int sigaction(int signum, const struct sigaction *act, struct sigaction *oldact) { - return LIBC_SIGACTION(signum, act, oldact); + if (!libc_sigaction) + { + libc_sigaction = dlsym( RTLD_NEXT, "sigaction" ); + assert( libc_sigaction ); + } + return libc_sigaction(signum, act, oldact); +} + +#else /* __GLIBC__ */ + +void PTHREAD_init_done(void) +{ } -#endif #endif /* __GLIBC__ */ -- Alexandre Julliard julliard(a)winehq.com

2 2

Known listview problems
by Dimitrie O. Paun 20 Oct '02

20 Oct '02

Folks, Things look pretty good on the listview front. In fact, I am kindda happy the way things are, as the S-series patches were rather tricky at points... Currently, I know of the following problems: -- Newsbin dies swiftly (reported my Rein Klazes) -- FTPCommander seems to experience some listview-related faults (reported by Carlos Lozano) -- Xnews has some minor focus rectangle problems (discovered by myself :)) -- IDAdemo _may_ have some minor drawing glitches (ditto) At this point, I would like to ask everybody that is aware of a listview problem to report it. I might not be able to put as much time, and effort into listview, as I did lately, so the more of a complete picture I have now, the better the chances that I'm gonna fix it. Next on my TODO: -- Fix Newsbin's crash (BTW, Rein I need a new trace with the latest CVS, if possible) -- Store per-column info into the listview (see the second status update for details) If you feel that there is a particular are of the listview that needs attention, let me know so I can work it into my TODO, time permitting. -- Dimi.

3 6

configuration
by Eric Pouech 20 Oct '02

20 Oct '02

I dunno if Alexandre will like this patch, but - when compiling with --disable-win16 turned on - the glue generation is based on the C_SRCS16 files as well as the C_SRCS - this, of course, brings in lots of unneed imports (triggered by the non needed glue code) the attached patch, when wine compilation is configured with --disable-win16, removes all 16 bit .c files from the inspection for glue generation this might not be sufficient: - the .glue.c file is generated in any case - however, it should compile as an empty compilation unit A+ Index: configure.ac =================================================================== RCS file: /home/cvs/cvsroot/wine/wine/configure.ac,v retrieving revision 1.84 diff -u -r1.84 configure.ac --- configure.ac 10 Oct 2002 17:54:28 -0000 1.84 +++ configure.ac 20 Oct 2002 17:43:53 -0000 @@ -25,10 +25,12 @@ AC_SUBST(OPTIONS) AC_SUBST(WIN16_FILES,"\$(WIN16_FILES)") AC_SUBST(WIN16_INSTALL,"\$(WIN16_INSTALL)") +AC_SUBST(C_SRCS16,"\$(C_SRCS16)") if test "x$enable_win16" = "xno" then WIN16_FILES="" WIN16_INSTALL="" + C_SRCS16="" fi if test "x$enable_debug" = "xno" then Index: dlls/Makedll.rules.in =================================================================== RCS file: /home/cvs/cvsroot/wine/wine/dlls/Makedll.rules.in,v retrieving revision 1.36 diff -u -r1.36 Makedll.rules.in --- dlls/Makedll.rules.in 2 Oct 2002 02:34:10 -0000 1.36 +++ dlls/Makedll.rules.in 20 Oct 2002 17:42:51 -0000 @@ -37,7 +37,7 @@ # Rule for 16-bit glue $(MODULE).glue.c: $(C_SRCS) $(C_SRCS16) $(WINEBUILD) - $(LDPATH) $(WINEBUILD) $(DEFS) -o $@ -glue -C$(SRCDIR) $(C_SRCS) $(C_SRCS16) + $(LDPATH) $(WINEBUILD) $(DEFS) -o $@ -glue -C$(SRCDIR) $(C_SRCS) @C_SRCS16@ # Rules for .dll files

2 1

Listview regression
by David D. Hagood 20 Oct '02

20 Oct '02

I've just applied/built/installed/tested the U series patches, and I have 2 regressions in the AAA MapNGo program: First, the first column of the directions window (which shows the type of road) is missing. Second, when I do the "Along the way", the program cores out. The log, and a screen cap showing (or rather, NOT showing) the missing field in the directions window, are in a tarball: http://sktc.net/~wowbagger/listview3.tar.gz

2 10

Fw: Wine problem
by Leonardo Bocalon 20 Oct '02

20 Oct '02

Hello it´s me again , i tried the procedure that you told me to do, I downloaded the last version of the wine, i installed it and still not working the keyboard display. Trying to explain to you more specifically the problem I´m going to send you the log which was the problem!!(I compacted whit .lha). I also want to know how to register (use) the files virtual drivers ( VXD)! Thanks again for you atention. Leonardo Bocalon Sysmo Informática / Brasil ----- Original Message ----- From: "Andreas Mohr" <andi(a)rhlx01.fht-esslingen.de> To: "Leonardo Bocalon" <leonardo(a)sysmo.com.br> Cc: <wine-devel(a)winehq.com> Sent: Wednesday, October 16, 2002 4:16 AM Subject: Re: Wine problem > On Tue, Oct 15, 2002 at 01:50:19PM -0300, Leonardo Bocalon wrote: > > Ac wine programers !! > > I am from a software company and we have a problem, our software communicate with a keyboard which has a display this software is make using Delphi (by borland) and for cominicate with the keyboard we use assembler code , wile we run the software in windows it work perfectly but when we try to use it in Linux using the wine it dosen´t work (the keyboard display). > > > > Please a need a sulution for it !!!! > Please retry with latest Wine CVS if you haven't yet, then add a BugZilla report > describing the problem if it still persists. > Attach screenshots if useful and maybe already add some log file like: > wine --debugmsg +relay,+event,+key,+keyboard MyPrg.exe &>/tmp/winekey.log > > -- > Andreas Mohr Stauferstr. 6, D-71272 Renningen, Germany > Tel. +49 7159 800604 http://mohr.de.tt

5 4

Re: DPA_InsertPtr (A1)
by Dimitrie O. Paun 20 Oct '02

20 Oct '02

On October 20, 2002 12:33 pm, Paul Rupe wrote: > I was just looking at this same problem. Should we not also remove the > line hdpa->nItemCount++? DPA_SetPtr already does this (which is why we > needed the -1 you added), so I think we end up incrementing it twice. You are correct. I've actually sent the patch in about 5minutes ago ... :) > For comparison, here's the patch I was about to send (relative to the first > comctl32undoc.c patch): Yup, that's what we need to end up with. -- Dimi.

1 0

Re: InstallShield
by Jeff Smith 20 Oct '02

20 Oct '02

Already running builtin ole32/oleaut32/rpcrt4. I have stdole2.tlb and stdole32.tlb, but no tbl files. Am I missing something? Any further thoughts? BTW, the only command-line message I get is: "fixme:seh:check_resource_write Broken app is writing to the resource data, enabling work-around" - Thanks - Jeff S. > >>>>> "Jeff" == Jeff Smith <whydoubt(a)hotmail.com> writes: > > Jeff> Trying to install the Music Ace Demo (harmonicvision.com), I get > Jeff> stopped by a messagethat "The InstallShield Engine (iKernel.exe) > Jeff> could not be installed. (0x20)" > > Jeff> I think someone said DCOM98 needs to be installed for > Jeff> InstallShield to work properly, so I did, and nothing changed. > >Have native stdole*.tbl available in the system directory and run with >builtin ole32/oleaut32/rpcrt4. Has been told uqite often... > >Bye >-- >Uwe Bonnes bon(a)elektron.ikp.physik.tu-darmstadt.de _________________________________________________________________ Get a speedy connection with MSN Broadband. Join now! http://resourcecenter.msn.com/access/plans/freeactivation.asp

2 1

DOS/Win16 interrupt handling architecture
by Jukka Heinonen 20 Oct '02

20 Oct '02

I have been thinking about interrupt handling in Wine because it needs some reworking in order to get DPMI32 programs to work. I have formed a proposal for refactoring current interrupt handling code and I would like to get some comments about this scheme before I decide whether I want to work on it or not. So, my proposal is to move all interrupt handlers to winedos dll and have only single interrupt handler routine per interrupt, instead of having separate real and protected mode handlers. If an interrupt occurs in protected mode ntdll would just load winedos and call single function from winedos that simulates the interrupt. A list of some things pro and con the proposal: + Win32 programs don't care about interrupt handlers. It would seem reasonable to think that many Win16 programs don't need them either. Also, architectures not supporting DOS/Win16 probably don't want to have any interrupt code. + Real and protected mode interrupt handlers are heavily interconnected. For example, protected mode handlers already serve both real mode and protected mode interrupts. MSDN documentation also says that nearly all interrupts under Win16 work just like they do under DOS real mode. + Differences in handling interrupts depend more on whether we are running DOS or Windows process than on whether the interrupt was called from real or protected mode. + We would only have one place that handles interrupts. This should make it easier to maintain interrupt code. + DLL separation would be easier because we would require far fewer calls between winedos and ntdll. - Win16 programs using interrupts take a bit longer to start and they get all kinds of unnecessary stuff into their address space from winedos. - Restructuring stuff means lots of work and possibility for regression bugs in Win16 support. P.S. I have a good idea on what needs to be done to make Wine support DPMI32. I have actually got some 32-bit DOS extenders to almost work. However, this requires lots of interrupt related work and I would like to wait until I get enough feedback about this interrupt refactoring proposal in order to see whether to continue working using the current interrupt code. -- Jukka Heinonen <http://www.iki.fi/jhei/>

3 4

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

wine-devel