wine/documentation/debugger.sgml
2000-12-13 21:52:37 +00:00

1603 lines
62 KiB
Text

<chapter id="debugger">
<title>Debugging Wine</title>
<sect1 id="dbg-intro">
<title>Introduction</title>
<para>
Written by &name-eric-pouech; <email>&email-eric-pouech;</email>
(Last updated: 6/14/2000)
</para>
<para>
(Extracted from <filename>wine/documentation/winedbg</filename>)
</para>
<sect2>
<title>Processes and threads: in underlying OS and in Windows</title>
<para>
Before going into the depths of debugging in Wine, here's
a small overview of process and thread handling in Wine.
It has to be clear that there are two different beasts:
processes/threads from the Unix point of view and
processes/threads from a Windows point of view.
</para>
<para>
Each Windows' thread is implemented as a Unix process (under
Linux using the <function>clone</function> syscall), meaning
that all threads of a same Windows' process share the same
(unix) address space.
</para>
<para>
In the following:
</para>
<itemizedlist>
<listitem>
<para><varname>W-process</varname> means a process in Windows' terminology</para>
</listitem>
<listitem>
<para><varname>U-process</varname> means a process in Unix' terminology</para>
</listitem>
<listitem>
<para><varname>W-thread</varname> means a thread in Windows' terminology</para>
</listitem>
</itemizedlist>
<para>
A <varname>W-process</varname> is made of one or several
<varname>W-threads</varname>. Each
<varname>W-thread</varname> is mapped to one and only one
<varname>U-process</varname>. All
<varname>U-processes</varname> of a same
<varname>W-process</varname> share the same address space.
</para>
<para>
Each Unix process can be identified by two values:
</para>
<itemizedlist>
<listitem>
<para>the Unix process id (<varname>upid</varname> in the following)</para>
</listitem>
<listitem>
<para>the Windows's thread id (<varname>tid</varname>)</para>
</listitem>
</itemizedlist>
<para>
Each Windows' process has also a Windows' process id
(<varname>wpid</varname> in the following). It must be clear
that <varname>upid</varname> and <varname>wpid</varname> are
different and shall not be used instead of the other.
</para>
<para>
<varname>Wpid</varname> and <varname>tid</varname> are
defined (Windows) system wide. They must not be confused
with process or thread handles which, as any handle, is an
indirection to a system object (in this case process or
thread). A same process can have several different handles
on the same kernel object. The handles can be defined as
local (the values is only valid in a process), or system
wide (the same handle can be used by any
<varname>W-process</varname>).
</para>
</sect2>
<sect2>
<title>Wine, debugging and WineDbg</title>
<para>
When talking of debugging in Wine, there are at least two
levels to think of:
</para>
<itemizedlist>
<listitem>
<para>the Windows' debugging API.</para>
</listitem>
<listitem>
<para>the Wine integrated debugger, dubbed
<command>WineDbg</command>.</para>
</listitem>
</itemizedlist>
<para>
Wine implements most of the Windows' debugging API (the
part in KERNEL32, not the one in
<filename>IMAGEHLP.DLL</filename>), and allows any program
(emulated or Winelib) using that API to debug a
<varname>W-process</varname>.
</para>
<para>
<command>WineDbg</command> is a Winelib application making
use of this API to allow debugging both any Wine or Winelib
applications as well as Wine itself (kernel and all DLLs).
</para>
</sect2>
</sect1>
<sect1 id="dbg-modes">
<title>WineDbg's modes of invocation</title>
<sect2>
<title>Starting a process</title>
<para>
Any application (either a Windows' native executable, or a
Winelib application) can be run through
<command>WineDbg</command>. Command line options and tricks
are the same as for wine:
</para>
<screen>
winedbg telnet.exe
winedbg "hl.exe -windowed"
</screen>
</sect2>
<sect2>
<title>Attaching</title>
<para>
<command>WineDbg</command> can also be launched without any
command line argument: <command>WineDbg</command> is started
without any attached process. You can get a list of running
<varname>W-processes</varname> (and their
<varname>wpid</varname>'s) using the <command>walk
process</command> command, and then, with the
<command>attach</command> command, pick up the
<varname>wpid</varname> of the <varname>W-process</varname>
you want to debug. This is (for now) a neat feature for the
following reasons:
</para>
<itemizedlist>
<listitem>
<para>you can debug an already started application</para>
</listitem>
</itemizedlist>
</sect2>
<sect2 id="dbg-on-exception">
<title id="dbg-exception-title">On exception</title>
<para>
When something goes wrong, Windows tracks this as an
exception. Exceptions exist for segmentation violation,
stack overflow, division by zero...
</para>
<para>
When an exception occurs, Wine checks if the <varname>W-process</varname> is
debugged. If so, the exception event is sent to the
debugger, which takes care of it: end of the story. This
mechanism is part of the standard Windows' debugging API.
</para>
<para>
If the <varname>W-process</varname> is not debugged, Wine
tries to launch a debugger. This debugger (normally
<command>WineDbg</command>, see III Configuration for more
details), at startup, attaches to the
<varname>W-process</varname> which generated the exception
event. In this case, you are able to look at the causes of
the exception, and either fix the causes (and continue
further the execution) or dig deeper to understand what went
wrong.
</para>
<para>
If <command>WineDbg</command> is the standard debugger, the
<command>pass</command> and <command>cont</command> commands
are the two ways to let the process go further for the
handling of the exception event.
</para>
<para>
To be more precise on the way Wine (and Windows) generates
exception events, when a fault occurs (segmentation
violation, stack overflow...), the event is first sent to
the debugger (this is known as a first chance exception).
The debugger can give two answers:
</para>
<variablelist>
<varlistentry>
<term>continue:</term>
<listitem>
<para>
the debugger had the ability to correct what's
generated the exception, and is now able to continue
process execution.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term>pass:</term>
<listitem>
<para>
the debugger couldn't correct the cause of the
first chance exception. Wine will now try to walk
the list of exception handlers to see if one of them
can handle the exception. If no exception handler is
found, the exception is sent once again to the
debugger to indicate the failure of the exception
handling.
</para>
</listitem>
</varlistentry>
</variablelist>
<note>
<para>
since some of Wine's code uses exceptions and
<function>try/catch</function> blocks to provide some
functionality, <command>WineDbg</command> can be entered
in such cases with segv exceptions. This happens, for
example, with <function>IsBadReadPtr</function> function.
In that case, the <command>pass</command> command shall be
used, to let the handling of the exception to be done by
the <function>catch</function> block in
<function>IsBadReadPtr</function>.
</para>
</note>
</sect2>
<sect2>
<title>Quitting</title>
<para>
Unfortunately, Windows doesn't provide a detach kind of API,
meaning that once you started debugging a process, you must
do so until the process dies. Killing (or stopping/aborting)
the debugger will also kill the debugged process. This will
be true for any Windows' debugging API compliant debugger,
starting with <command>WineDbg</command>.
</para>
</sect2>
</sect1>
<sect1 id="wine-debugger">
<title>Using the Wine Debugger</title>
<para>
Written by &name-marcus-meissner; <email>&email-marcus-meissner;</email>,
additions welcome.
</para>
<para>
(Extracted from <filename>wine/documentation/debugging</filename>)
</para>
<para>
This file describes where to start debugging Wine. If at any
point you get stuck and want to ask for help, please read the
file <filename>documentation/bugreports</filename> for
information on how to write useful bug reports.
</para>
<sect2>
<title>Crashes</title>
<para>
These usually show up like this:
</para>
<screen>
|Unexpected Windows program segfault - opcode = 8b
|Segmentation fault in Windows program 1b7:c41.
|Loading symbols from ELF file /root/wine/wine...
|....more Loading symbols from ...
|In 16 bit mode.
|Register dump:
| CS:01b7 SS:016f DS:0287 ES:0000
| IP:0c41 SP:878a BP:8796 FLAGS:0246
| AX:811e BX:0000 CX:0000 DX:0000 SI:0001 DI:ffff
|Stack dump:
|0x016f:0x878a: 0001 016f ffed 0000 0000 0287 890b 1e5b
|0x016f:0x879a: 01b7 0001 000d 1050 08b7 016f 0001 000d
|0x016f:0x87aa: 000a 0003 0004 0000 0007 0007 0190 0000
|0x016f:0x87ba:
|
|0050: sel=0287 base=40211d30 limit=0b93f (bytes) 16-bit rw-
|Backtrace:
|0 0x01b7:0x0c41 (PXSRV_FONGETFACENAME+0x7c)
|1 0x01b7:0x1e5b (PXSRV_FONPUTCATFONT+0x2cd)
|2 0x01a7:0x05aa
|3 0x01b7:0x0768 (PXSRV_FONINITFONTS+0x81)
|4 0x014f:0x03ed (PDOXWIN_@SQLCURCB$Q6CBTYPEULN8CBSCTYPE+0x1b1)
|5 0x013f:0x00ac
|
|0x01b7:0x0c41 (PXSRV_FONGETFACENAME+0x7c): movw %es:0x38(%bx),%dx
</screen>
<para>
Steps to debug a crash. You may stop at any step, but please
report the bug and provide as much of the information
gathered to the newsgroup or the relevant developer as
feasible.
</para>
<orderedlist>
<listitem>
<para>
Get the reason for the crash. This is usually an access to
an invalid selector, an access to an out of range address
in a valid selector, popping a segmentregister from the
stack or the like. When reporting a crash, report this
<emphasis>whole</emphasis> crashdump even if it doesn't
make sense to you.
</para>
<para>
(In this case it is access to an invalid selector, for
<systemitem>%es</systemitem> is <literal>0000</literal>, as
seen in the register dump).
</para>
</listitem>
<listitem>
<para>
Determine the cause of the crash. Since this is usually
a primary/secondary reaction to a failed or misbehaving
Wine function, rerun Wine with <parameter>-debugmsg
+relay</parameter> added to the commandline. This will
generate quite a lot of output, but usually the reason is
located in the last call(s). Those lines usually look like
this:
</para>
<screen>
|Call KERNEL.90: LSTRLEN(0227:0692 "text") ret=01e7:2ce7 ds=0227
^^^^^^^^^ ^ ^^^^^^^^^ ^^^^^^ ^^^^^^^^^ ^^^^
| | | | | |Datasegment
| | | | |Return address
| | | |textual parameter
| | |
| | |Argument(s). This one is a win16 segmented pointer.
| |Function called.
|The module, the function is called in. In this case it is KERNEL.
|Ret KERNEL.90: LSTRLEN() retval=0x0004 ret=01e7:2ce7 ds=0227
^^^^^^
|Returnvalue is 16 bit and has the value 4.
</screen>
</listitem>
<listitem>
<para>
If you have found a misbehaving function, try to find out
why it misbehaves. Find the function in the source code.
Try to make sense of the arguments passed. Usually there is
a <function>TRACE(&lt;channel>,"(...)\n");</function> at
the beginning of the function. Rerun wine with
<parameter>-debugmsg +xyz,+relay</parameter> added to the
commandline.
</para>
</listitem>
<listitem>
<para>
Additional information on how to debug using the internal
debugger can be found in
<filename>debugger/README</filename>.
</para>
</listitem>
<listitem>
<para>
If this information isn't clear enough or if you want to
know more about what's happening in the function itself,
try running wine with <parameter>-debugmsg
+all</parameter>, which dumps ALL included debug
information in wine.
</para>
</listitem>
<listitem>
<para>
If even that isn't enough, add more debug output for
yourself into the functions you find relevant. See
<filename>documentation/debug-msgs</filename>. You might
also try to run the program in <command>gdb</command>
instead of using the WINE-debugger. If you do that, use
<parameter>handle SIGSEGV nostop noprint</parameter> to
disable the handling of seg faults inside
<command>gdb</command> (needed for Win16). If you don't use
the <parameter>--desktop</parameter> or
<parameter>--managed</parameter> option, start the WINE
process with <parameter>--sync</parameter>, or chances are
good to get X into an unusable state.
</para>
</listitem>
<listitem>
<para>
You can also set a breakpoint for that function. Start wine
with the <parameter>--debug</parameter> option added to the
commandline. After loading the executable wine will enter
the internal debugger. Use <parameter>break
KERNEL_LSTRLEN</parameter> (replace by function you want
to debug, CASE IS RELEVANT) to set a breakpoint. Then use
<command>continue</command> to start normal
program-execution. Wine will stop if it reaches the
breakpoint. If the program isn't yet at the crashing call
of that function, use <command>continue</command> again
until you are about to enter that function. You may now
proceed with single-stepping the function until you reach
the point of crash. Use the other debugger commands to
print registers and the like.
</para>
</listitem>
</orderedlist>
</sect2>
<sect2>
<title>Program hangs, nothing happens</title>
<para>
Switch to UNIX shell, get the process-ID using <command>ps -a |
grep wine</command>, and do a <command>kill -HUP
&lt;pid></command> (without the &lt; and >). Wine will
then enter its internal debugger and you can proceed as
explained above. Also, you can use
<parameter>--debug</parameter> switch and then you can get into
internal debugger by pressing
<keycombo><keycap>Ctrl</keycap><keycap>C</keycap></keycombo> in
the terminal where you run Wine.
</para>
</sect2>
<sect2>
<title>Program reports an error with a Messagebox</title>
<para>
Sometimes programs are reporting failure using more or
less nondescript messageboxes. We can debug this using the
same method as Crashes, but there is one problem... For
setting up a message box the program also calls Wine
producing huge chunks of debug code.
</para>
<para>
Since the failure happens usually directly before setting up
the Messagebox you can start wine with
<parameter>--debug</parameter> added to the commandline, set a
breakpoint at <function>MessageBoxA</function> (called by win16
and win32 programs) and proceed with
<command>continue</command>. With <parameter>--debugmsg
+all</parameter> Wine will now stop directly before setting
up the Messagebox. Proceed as explained above.
</para>
<para>
You can also run wine using <command>wine -debugmsg +relay
program.exe 2>&1 | less -i</command> and in
<command>less</command> search for <quote>MessageBox</quote>.
</para>
</sect2>
<sect2>
<title>Disassembling programs:</title>
<para>
You may also try to disassemble the offending program to
check for undocumented features and/or use of them.
</para>
<para>
The best, freely available, disassembler for Win16 programs is
<application>Windows Codeback</application>, archivename
<filename>wcbxxx.zip</filename>, which usually can be found in
the <filename>Cica-Mirror</filename> subdirectory on the WINE
ftpsites. (See <filename>ANNOUNCE</filename>).
</para>
<para>
Disassembling win32 programs is possible using
<application>Windows Disassembler 32</application>, archivename
something like <filename>w32dsm87.zip</filename> (or similar)
on <systemitem class="systemname">ftp.winsite.com</systemitem>
and mirrors. The shareware version does not allow saving of
disassembly listings. You can also use the newer (and in the
full version better) <application>Interactive
Disassembler</application> (IDA) from the ftp sites mentioned
at the end of the document. Understanding disassembled code is
mostly a question of exercise.
</para>
<para>
Most code out there uses standard C function entries (for it
is usually written in C). Win16 function entries usually
look like that:
</para>
<programlisting>
push bp
mov bp, sp
... function code ..
retf XXXX &lt;--------- XXXX is number of bytes of arguments
</programlisting>
<para>
This is a <function>FAR</function> function with no local
storage. The arguments usually start at
<literal>[bp+6]</literal> with increasing offsets. Note, that
<literal>[bp+6]</literal> belongs to the
<emphasis>rightmost</emphasis> argument, for exported win16
functions use the PASCAL calling convention. So, if we use
<function>strcmp(a,b)</function> with <parameter>a</parameter>
and <parameter>b</parameter> both 32 bit variables
<parameter>b</parameter> would be at <literal>[bp+6]</literal>
and <parameter>a</parameter> at <literal>[bp+10]</literal>.
</para>
<para>
Most functions make also use of local storage in the stackframe:
</para>
<programlisting>
enter 0086, 00
... function code ...
leave
retf XXXX
</programlisting>
<para>
This does mostly the same as above, but also adds
<literal>0x86</literal> bytes of stackstorage, which is
accessed using <literal>[bp-xx]</literal>. Before calling a
function, arguments are pushed on the stack using something
like this:
</para>
<programlisting>
push word ptr [bp-02] &lt;- will be at [bp+8]
push di &lt;- will be at [bp+6]
call KERNEL.LSTRLEN
</programlisting>
<para>
Here first the selector and then the offset to the passed
string are pushed.
</para>
</sect2>
<sect2>
<title>Sample debugging session:</title>
<para>
Let's debug the infamous Word <filename>SHARE.EXE</filename>
messagebox:
</para>
<screen>
|marcus@jet $ wine winword.exe
| +---------------------------------------------+
| | ! You must leave Windows and load SHARE.EXE|
| | before starting Word. |
| +---------------------------------------------+
</screen>
<screen>
|marcus@jet $ wine winword.exe -debugmsg +relay -debug
|CallTo32(wndproc=0x40065bc0,hwnd=000001ac,msg=00000081,wp=00000000,lp=00000000)
|Win16 task 'winword': Breakpoint 1 at 0x01d7:0x001a
|CallTo16(func=0127:0070,ds=0927)
|Call WPROCS.24: TASK_RESCHEDULE() ret=00b7:1456 ds=0927
|Ret WPROCS.24: TASK_RESCHEDULE() retval=0x8672 ret=00b7:1456 ds=0927
|CallTo16(func=01d7:001a,ds=0927)
| AX=0000 BX=3cb4 CX=1f40 DX=0000 SI=0000 DI=0927 BP=0000 ES=11f7
|Loading symbols: /home/marcus/wine/wine...
|Stopped on breakpoint 1 at 0x01d7:0x001a
|In 16 bit mode.
|Wine-dbg>break MessageBoxA &lt;---- Set Breakpoint
|Breakpoint 2 at 0x40189100 (MessageBoxA [msgbox.c:190])
|Wine-dbg>c &lt;---- Continue
|Call KERNEL.91: INITTASK() ret=0157:0022 ds=08a7
| AX=0000 BX=3cb4 CX=1f40 DX=0000 SI=0000 DI=08a7 ES=11d7 EFL=00000286
|CallTo16(func=090f:085c,ds=0dcf,0x0000,0x0000,0x0000,0x0000,0x0800,0x0000,0x0000,0x0dcf)
|... &lt;----- Much debugoutput
|Call KERNEL.136: GETDRIVETYPE(0x0000) ret=060f:097b ds=0927
^^^^^^ Drive 0 (A:)
|Ret KERNEL.136: GETDRIVETYPE() retval=0x0002 ret=060f:097b ds=0927
^^^^^^ DRIVE_REMOVEABLE
(It is a floppy diskdrive.)
|Call KERNEL.136: GETDRIVETYPE(0x0001) ret=060f:097b ds=0927
^^^^^^ Drive 1 (B:)
|Ret KERNEL.136: GETDRIVETYPE() retval=0x0000 ret=060f:097b ds=0927
^^^^^^ DRIVE_CANNOTDETERMINE
(I don't have drive B: assigned)
|Call KERNEL.136: GETDRIVETYPE(0x0002) ret=060f:097b ds=0927
^^^^^^^ Drive 2 (C:)
|Ret KERNEL.136: GETDRIVETYPE() retval=0x0003 ret=060f:097b ds=0927
^^^^^^ DRIVE_FIXED
(specified as a harddisk)
|Call KERNEL.97: GETTEMPFILENAME(0x00c3,0x09278364"doc",0x0000,0927:8248) ret=060f:09b1 ds=0927
^^^^^^ ^^^^^ ^^^^^^^^^
| | |buffer for fname
| |temporary name ~docXXXX.tmp
|Force use of Drive C:.
|Warning: GetTempFileName returns 'C:~doc9281.tmp', which doesn't seem to be writeable.
|Please check your configuration file if this generates a failure.
</screen>
<para>
Whoops, it even detects that something is wrong!
</para>
<screen>
|Ret KERNEL.97: GETTEMPFILENAME() retval=0x9281 ret=060f:09b1 ds=0927
^^^^^^ Temporary storage ID
|Call KERNEL.74: OPENFILE(0x09278248"C:~doc9281.tmp",0927:82da,0x1012) ret=060f:09d8 ds=0927
^^^^^^^^^^^^^^^^ ^^^^^^^^^ ^^^^^^^
|filename |OFSTRUCT |open mode:
OF_CREATE|OF_SHARE_EXCLUSIVE|OF_READWRITE
</screen>
<para>
This fails, since my <medialabel>C:</medialabel> drive is in
this case mounted readonly.
</para>
<screen>
|Ret KERNEL.74: OPENFILE() retval=0xffff ret=060f:09d8 ds=0927
^^^^^^ HFILE_ERROR16, yes, it failed.
|Call USER.1: MESSAGEBOX(0x0000,0x09278376"Sie mussen Windows verlassen und SHARE.EXE laden bevor Sie Word starten.",0x00000000,0x1030) ret=060f:084f ds=0927
</screen>
<para>
And MessageBox'ed.
</para>
<screen>
|Stopped on breakpoint 2 at 0x40189100 (MessageBoxA [msgbox.c:190])
|190 { &lt;- the sourceline
In 32 bit mode.
Wine-dbg>
</screen>
<para>
The code seems to find a writeable harddisk and tries to create
a file there. To work around this bug, you can define
<medialabel>C:</medialabel> as a networkdrive, which is ignored
by the code above.
</para>
</sect2>
<sect2>
<title>Debugging Tips</title>
<para>
Here are some useful debugging tips, added by Andreas Mohr:
</para>
<itemizedlist>
<listitem>
<para>
If you have a program crashing at such an early loader phase that you can't
use the Wine debugger normally, but Wine already executes the program's
start code, then you may use a special trick. You should do a
<programlisting>
wine --debugmsg +relay program
</programlisting>
to get a listing of the functions the program calls in its start function.
Now you do a
<programlisting>
wine --debug winfile.exe
</programlisting>
</para>
<para>
This way, you get into <command>Wine-dbg</command>. Now you
can set a breakpoint on any function the program calls in
the start function and just type <userinput>c</userinput>
to bypass the eventual calls of Winfile to this function
until you are finally at the place where this function gets
called by the crashing start function. Now you can proceed
with your debugging as usual.
</para>
</listitem>
<listitem>
<para>
If you try to run a program and it quits after showing an error messagebox,
the problem can usually be identified in the return value of one of the
functions executed before <function>MessageBox()</function>.
That's why you should re-run the program with e.g.
<programlisting>
wine --debugmsg +relay &lt;program name> &>relmsg
</programlisting>
Then do a <command>more relmsg</command> and search for the
last occurrence of a call to the string "MESSAGEBOX". This is a line like
<programlisting>
Call USER.1: MESSAGEBOX(0x0000,0x01ff1246 "Runtime error 219 at 0004:1056.",0x00000000,0x1010) ret=01f7:2160 ds=01ff
</programlisting>
In my example the lines before the call to
<function>MessageBox()</function> look like that:
<programlisting>
Call KERNEL.96: FREELIBRARY(0x0347) ret=01cf:1033 ds=01ff
CallTo16(func=033f:0072,ds=01ff,0x0000)
Ret KERNEL.96: FREELIBRARY() retval=0x0001 ret=01cf:1033 ds=01ff
Call KERNEL.96: FREELIBRARY(0x036f) ret=01cf:1043 ds=01ff
CallTo16(func=0367:0072,ds=01ff,0x0000)
Ret KERNEL.96: FREELIBRARY() retval=0x0001 ret=01cf:1043 ds=01ff
Call KERNEL.96: FREELIBRARY(0x031f) ret=01cf:105c ds=01ff
CallTo16(func=0317:0072,ds=01ff,0x0000)
Ret KERNEL.96: FREELIBRARY() retval=0x0001 ret=01cf:105c ds=01ff
Call USER.171: WINHELP(0x02ac,0x01ff05b4 "COMET.HLP",0x0002,0x00000000) ret=01cf:1070 ds=01ff
CallTo16(func=0117:0080,ds=01ff)
Call WPROCS.24: TASK_RESCHEDULE() ret=00a7:0a2d ds=002b
Ret WPROCS.24: TASK_RESCHEDULE() retval=0x0000 ret=00a7:0a2d ds=002b
Ret USER.171: WINHELP() retval=0x0001 ret=01cf:1070 ds=01ff
Call KERNEL.96: FREELIBRARY(0x01be) ret=01df:3e29 ds=01ff
Ret KERNEL.96: FREELIBRARY() retval=0x0000 ret=01df:3e29 ds=01ff
Call KERNEL.52: FREEPROCINSTANCE(0x02cf00ba) ret=01f7:1460 ds=01ff
Ret KERNEL.52: FREEPROCINSTANCE() retval=0x0001 ret=01f7:1460 ds=01ff
Call USER.1: MESSAGEBOX(0x0000,0x01ff1246 "Runtime error 219 at 0004:1056.",0x00000000,0x1010) ret=01f7:2160 ds=01ff
</programlisting>
</para>
<para>
I think that the call to <function>MessageBox()</function>
in this example is <emphasis>not</emphasis> caused by a
wrong result value of some previously executed function
(it's happening quite often like that), but instead the
messagebox complains about a runtime error at
<literal>0x0004:0x1056</literal>.
</para>
<para>
As the segment value of the address is only
<literal>4</literal>, I think that that is only an internal
program value. But the offset address reveals something
quite interesting: Offset <literal>1056</literal> is
<emphasis>very</emphasis> close to the return address of
<function>FREELIBRARY()</function>:
<programlisting>
Call KERNEL.96: FREELIBRARY(0x031f) ret=01cf:105c ds=01ff
^^^^
</programlisting>
</para>
<para>
Provided that segment <literal>0x0004</literal> is indeed segment
<literal>0x1cf</literal>, we now we can use IDA (available at
<ulink url="ftp://ftp.uni-koeln.de/pc/msdos/programming/assembler/ida35bx.zip">
ftp://ftp.uni-koeln.de/pc/msdos/programming/assembler/ida35bx.zip</ulink>) to
disassemble the part that caused the error. We just have to find the address of
the call to <function>FreeLibrary()</function>. Some lines before that the
runtime error occurred. But be careful! In some cases you don't have to
disassemble the main program, but instead some DLL called by it in order to find
the correct place where the runtime error occurred. That can be determined by
finding the origin of the segment value (in this case <literal>0x1cf</literal>).
</para>
</listitem>
<listitem>
<para>
If you have created a relay file of some crashing
program and want to set a breakpoint at a certain
location which is not yet available as the program loads
the breakpoint's segment during execution, you may set a
breakpoint to <function>GetVersion16/32</function> as
those functions are called very often.
</para>
<para>
Then do a <userinput>c</userinput> until you are able to
set this breakpoint without error message.
</para>
</listitem>
<listitem>
<para>
Some useful programs:
</para>
<variablelist>
<varlistentry>
<term>
<application>IDA</application>:
<filename>
<ulink url="ftp://ftp.uni-koeln.de/pc/msdos/programming/assembler/ida35bx.zip">
ftp://ftp.uni-koeln.de/pc/msdos/programming/assembler/ida35bx.zip</ulink>
</filename>
</term>
<listitem>
<para>
<emphasis>Very</emphasis> good DOS disassembler ! It's badly needed
for debugging Wine sometimes.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term>
<application>XRAY</application>:
<filename>
<ulink url="ftp://ftp.th-darmstadt.de/pub/machines/ms-dos/SimTel/msdos/asmutil/xray15.zip">
ftp://ftp.th-darmstadt.de/pub/machines/ms-dos/SimTel/msdos/asmutil/xray15.zip</ulink>
</filename>
</term>
<listitem>
<para>
Traces DOS calls (Int 21h, DPMI, ...). Use it with
Windows to correct file management problems etc.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term>
<application>pedump</application>:
<filename>
<ulink url="http://oak.oakland.edu/pub/simtelnet/win95/prog/pedump.zip">
http://oak.oakland.edu/pub/simtelnet/win95/prog/pedump.zip</ulink>
</filename>
</term>
<listitem>
<para>
Dumps the imports and exports of a PE (Portable
Executable) DLL.
</para>
</listitem>
</varlistentry>
</variablelist>
</listitem>
</itemizedlist>
</sect2>
<sect2>
<title>Some basic debugger usages:</title>
<para>
After starting your program with
</para>
<screen>
wine -debug myprog.exe
</screen>
<para>
the program loads and you get a prompt at the program
starting point. Then you can set breakpoints:
</para>
<screen>
b RoutineName (by outine name) OR
b *0x812575 (by address)
</screen>
<para>
Then you hit <command>c</command> (continue) to run the
program. It stops at the breakpoint. You can type
</para>
<screen>
step (to step one line) OR
stepi (to step one machine instruction at a time;
here, it helps to know the basic 386
instruction set)
info reg (to see registers)
info stack (to see hex values in the stack)
info local (to see local variables)
list &lt;line number> (to list source code)
x &lt;variable name> (to examine a variable; only works if code
is not compiled with optimization)
x 0x4269978 (to examine a memory location)
? (help)
q (quit)
</screen>
<para>
By hitting <keycap>Enter</keycap>, you repeat the last
command.
</para>
</sect2>
</sect1>
<sect1 id="memory-addresses">
<title>Useful memory addresses</title>
<para>
Written by &name-andreas-mohr; <email>&email-andreas-mohr;</email>
</para>
<para>
Wine uses several different kinds of memory addresses.
</para>
<variablelist>
<varlistentry>
<term>
Win32/"normal" Wine addresses/Linux: linear addresses.
</term>
<listitem>
<para>
Linear addresses can be everything from 0x0 up to
0xffffffff. In Wine on Linux they are often around
e.g. 0x08000000, 0x00400000 (std. Win32 program load
address), 0x40000000. Every Win32 process has its own
private 4GB address space (that is, from 0x0 up to
0xffffffff).
</para>
</listitem>
</varlistentry>
<varlistentry>
<term>
Win16 "enhanced mode": segmented addresses.
</term>
<listitem>
<para>
These are the "normal" Win16 addresses, called SEGPTR.
They have a segment:offset notation, e.g. 0x01d7:0x0012.
The segment part usually is a "selector", which *always*
has the lowest 3 bits set. Some sample selectors are
0x1f7, 0x16f, 0x8f. If these bits are set except for
the lowest bit, as e.g. with 0x1f6,xi then it might be a
handle to global memory. Just set the lowest bit to get
the selector in these cases. A selector kind of
"points" to a certain linear (see above) base address.
It has more or less three important attributes: segment
base address, segment limit, segment access rights.
</para>
<para>
Example:
</para>
<para>
Selector 0x1f7 (0x40320000, 0x0000ffff, r-x) So 0x1f7
has a base address of 0x40320000, the segment's last
address is 0x4032ffff (limit 0xffff), and it's readable
and executable. So an address of 0x1f7:0x2300 would be
the linear address of 0x40322300.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term>
DOS/Win16 "standard mode"
</term>
<listitem>
<para>
They, too, have a segment:offset notation. But they are
completely different from "normal" Win16 addresses, as
they just represent at most 1MB of memory: The segment
part can be anything from 0 to 0xffff, and it's the same
with the offset part.
</para>
<para>
Now the strange thing is the calculation that's behind
these addresses: Just calculate segment*16 + offset in
order to get a "linear DOS" address. So
e.g. 0x0f04:0x3628 results in 0xf040 + 0x3628 = 0x12668.
And the highest address you can get is 0xfffff (1MB), of
course. In Wine, this "linear DOS" address of 0x12668
has to be added to the linear base address of the
corresponding DOS memory allocated for dosmod in order
to get the true linear address of a DOS seg:offs
address. And make sure that you're doing this in the
correct process with the correct linear address space,
of course ;-)
</para>
</listitem>
</varlistentry>
</variablelist>
</sect1>
<sect1 id="dbg-config">
<title>Configuration</title>
<sect2>
<title>Registry configuration</title>
<para>
The Windows' debugging API uses a registry entry to know
which debugger to invoke when an unhandled exception occurs
(see <link endterm="dbg-exception-title"
linkend="dbg-on-exception"></link> for some details). Two
values in key
</para>
<programlisting>
"MACHINE\\Software\\Microsoft\\Windows NT\\CurrentVersion\\AeDebug"
</programlisting>
<para>
Determine the behavior:
</para>
<variablelist>
<varlistentry>
<term>Debugger:</term>
<listitem>
<para>
this is the command line used to launch the debugger
(it uses two <function>printf</function> formats
(<literal>%ld</literal>) to pass context dependent
information to the debugger). You should put here a
complete path to your debugger
(<command>WineDbg</command> can of course be used, but
any other Windows' debugging API aware debugger will
do).
The path to the debugger you chose to use must be reachable
via a DOS drive in the Wine config file !
</para>
</listitem>
</varlistentry>
<varlistentry>
<term>Auto:</term>
<listitem>
<para>
if this value is zero, a message box will ask the
user if he/she wishes to launch the debugger when an
unhandled exception occurs. Otherwise, the debugger
is automatically started.
</para>
</listitem>
</varlistentry>
</variablelist>
<para>
A regular Wine registry looks like:
</para>
<programlisting>
[MACHINE\\Software\\Microsoft\\Windows NT\\CurrentVersion\\AeDebug] 957636538
"Auto"=dword:00000001
"Debugger"="/usr/local/bin/winedbg %ld %ld"
</programlisting>
<note>
<title>Note 1</title>
<para>
creating this key is mandatory. Not doing so will not
fire the debugger when an exception occurs.
</para>
</note>
<note>
<title>Note 2</title>
<para>
<command>wineinstall</command> sets up this correctly.
However, due to some limitation of the registry installed,
if a previous Wine installation exists, it's safer to
remove the whole
</para>
<programlisting>
[MACHINE\\Software\\Microsoft\\Windows NT\\CurrentVersion\\AeDebug]
</programlisting>
<para>
key before running again <command>wineinstall</command> to
regenerate this key.
</para>
</note>
</sect2>
<sect2>
<title>WineDbg configuration</title>
<para>
<command>WineDbg</command> can be configured thru a number
of options. Those options are stored in the registry, on a
per user basis. The key is (in <emphasis>my</emphasis> registry)
</para>
<programlisting>
[eric\\Software\\Wine\\WineDbg]
</programlisting>
<para>
Those options can be read/written while inside
<command>WineDbg</command>, as part of the debugger
expressions. To refer to one of these options, its name must
be prefixed by a <literal>$</literal> sign. For example,
</para>
<programlisting>
set $BreakAllThreadsStartup = 1
</programlisting>
<para>
sets the option <varname>BreakAllThreadsStartup</varname> to
<literal>TRUE</literal>.
</para>
<para>
All the options are read from the registry when
<command>WineDbg</command> starts (if no corresponding value
is found, a default value is used), and are written back to
the registry when <command>WineDbg</command> exits (hence,
all modifications to those options are automatically saved
when <command>WineDbg</command> terminates).
</para>
<para>
Here's the list of all options:
</para>
<sect3>
<title>Controlling when the debugger is entered</title>
<variablelist>
<varlistentry>
<term><varname>BreakAllThreadsStartup</varname></term>
<listitem>
<para>
Set to <literal>TRUE</literal> if at all threads
start-up the debugger stops set to
<literal>FALSE</literal> if only at the first thread
startup of a given process the debugger stops.
<literal>FALSE</literal> by default.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><varname>BreakOnCritSectTimeOut</varname></term>
<listitem>
<para>
Set to <literal>TRUE</literal> if the debugger stops
when a critical section times out (5 minutes);
<literal>TRUE</literal> by default.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><varname>BreakOnAttach</varname></term>
<listitem>
<para>
Set to <literal>TRUE</literal> if when
<command>WineDbg</command> attaches to an existing
process after an unhandled exception,
<command>WineDbg</command> shall be entered on the
first attach event. Since the attach event is
meaningless in the context of an exception event
(the next event which is the exception event is of
course relevant), that option is likely to be
<literal>FALSE</literal>.
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><varname>BreakOnFirstChance</varname></term>
<listitem>
<para>
An exception can generate two debug events. The
first one is passed to the debugger (known as a
first chance) just after the exception. The debugger
can then decides either to resume execution (see
<command>WineDbg</command>'s <command>cont</command>
command) or pass the exception up to the exception
handler chain in the program (if it exists)
(<command>WineDbg</command> implements this thru the
<command>pass</command> command). If none of the
exception handlers takes care of the exception, the
exception event is sent again to the debugger (known
as last chance exception). You cannot pass on a last
exception. When the
<varname>BreakOnFirstChance</varname> exception is
<literal>TRUE</literal>, then winedbg is entered for
both first and last chance execptions (to
<literal>FALSE</literal>, it's only entered for last
chance exceptions).
</para>
</listitem>
</varlistentry>
</variablelist>
</sect3>
<sect3>
<title>Output handling</title>
<variablelist>
<varlistentry>
<term><varname>ConChannelMask</varname></term>
<listitem>
<para>
Mask of active debugger output channels on console
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><varname>StdChannelMask</varname></term>
<listitem>
<para>
Mask of active debugger output channels on <filename>stderr</filename>
</para>
</listitem>
</varlistentry>
<varlistentry>
<term><varname>UseXTerm</varname></term>
<listitem>
<para>
Set to <literal>TRUE</literal> if the debugger uses
its own <command>xterm</command> window for console
input/output. Set to <literal>FALSE</literal> if
the debugger uses the current Unix console for
input/output
</para>
</listitem>
</varlistentry>
</variablelist>
<para>
Those last 3 variables are jointly used in two generic ways:
</para>
<orderedlist>
<listitem>
<para>default</para>
<programlisting>
ConChannelMask = DBG_CHN_MESG (1)
StdChannelMask = 0
UseXTerm = 1
</programlisting>
<para>
In this case, all input/output goes into a specific
<command>xterm</command> window (but all debug
messages <function>TRACE</function>,
<function>WARN</function>... still goes to tty where
wine is run from).
</para>
</listitem>
<listitem>
<para>
to have all input/output go into the tty where Wine
was started from (to be used in a X11-free
environment)
</para>
<screen>
ConChannelMask = 0
StdChannelMask = DBG_CHN_MESG (1)
UseXTerm = 1
</screen>
</listitem>
</orderedlist>
<para>
Those variables also allow, for example for debugging
purposes, to use:
</para>
<screen>
ConChannelMask = 0xfff
StdChannelMask = 0xfff
UseXTerm = 1
</screen>
<para>
This allows to redirect all <function>WineDbg</function>
output to both tty Wine was started from, and
<command>xterm</command> debugging window. If Wine (or
<command>WineDbg</command>) was started with a redirection
of <filename>stdout</filename> and/or
<filename>stderr</filename> to a file (with for
example &gt;& shell redirect command), you'll get in that
file both outputs. It may be interesting to look in the
relay trace for specific values which the process segv'ed
on.
</para>
</sect3>
<sect3>
<title>Context information</title>
<variablelist>
<varlistentry>
<term><varname>ThreadId</varname></term>
<listitem>
<para>ID of the <varname>W-thread</varname> currently
examined by the debugger</para>
</listitem>
</varlistentry>
<varlistentry>
<term><varname>ProcessId</varname></term>
<listitem>
<para>ID of the <varname>W-thread</varname> currently
examined by the debugger</para>
</listitem>
</varlistentry>
<varlistentry>
<term>&lt;registers></term>
<listitem>
<para>All CPU registers are also available</para>
</listitem>
</varlistentry>
</variablelist>
<para>
The <varname>ThreadId</varname> and
<varname>ProcessId</varname> variables can be handy to set
conditional breakpoints on a given thread or process.
</para>
</sect3>
</sect2>
</sect1>
<sect1 id="dbg-commands">
<title>WineDbg Command Reference</title>
<sect2>
<title>Misc</title>
<screen>
abort aborts the debugger
quit exits the debugger
attach N attach to a W-process (N is its ID). IDs can be
obtained thru walk process command
</screen>
<screen>
help prints some help on the commands
help info prints some help on info commands
</screen>
<screen>
mode 16 switch to 16 bit mode
mode 32 switch to 32 bit mode
</screen>
</sect2>
<sect2>
<title>Flow control</title>
<screen>
cont continue execution until next breakpoint or exception.
pass pass the exception event up to the filter chain.
step continue execution until next C line of code (enters
function call)
next continue execution until next C line of code (doesn't
enter function call)
stepi execute next assembly instruction (enters function
call)
nexti execute next assembly instruction (doesn't enter
function call)
finish do nexti commands until current function is exited
</screen>
<para>
cont, step, next, stepi, nexti can be postfixed by a
number (N), meaning that the command must be executed N
times.
</para>
</sect2>
<sect2>
<title>Breakpoints, watch points</title>
<screen>
enable N enables (break|watch)point #N
disable N disables (break|watch)point #N
delete N deletes (break|watch)point #N
cond N removes any a existing condition to (break|watch)point N
cond N &lt;expr&gt; adds condition &lt;expr&gt; to (break|watch)point N. &lt;expr&gt;
will be evaluated each time the breakpoint is hit. If
the result is a zero value, the breakpoint isn't
triggered
break * N adds a breakpoint at address N
break &lt;id&gt; adds a breakpoint at the address of symbol &lt;id&gt;
break &lt;id&gt; N adds a breakpoint at the address of symbol &lt;id&gt; (N ?)
break N adds a breakpoint at line N of current source file
break adds a breakpoint at current $pc address
watch * N adds a watch command (on write) at address N (on 4 bytes)
watch &lt;id&gt; adds a watch command (on write) at the address of
symbol &lt;id&gt;
info break lists all (break|watch)points (with state)
</screen>
</sect2>
<sect2>
<title>Stack manipulation</title>
<screen>
bt print calling stack of current thread
up goes up one frame in current thread's stack
up N goes up N frames in current thread's stack
dn goes down one frame in current thread's stack
dn N goes down N frames in current thread's stack
frame N set N as the current frame
info local prints information on local variables for current
function
</screen>
</sect2>
<sect2>
<title>Directory & source file manipulation</title>
<screen>
show dir
dir &lt;pathname&gt;
dir
symbolfile &lt;pathname&gt;
</screen>
<screen>
list lists 10 source lines from current position
list - lists 10 source lines before current position
list N lists 10 source lines from line N in current file
list &lt;path&gt;:N lists 10 source lines from line N in file &lt;path&gt;
list &lt;id&gt; lists 10 source lines of function &lt;id&gt;
list * N lists 10 source lines from address N
</screen>
<para>
You can specify the end target (to change the 10 lines
value) using the ','. For example:
</para>
<screen>
list 123, 234 lists source lines from line 123 up to line 234 in
current file
list foo.c:1,56 lists source lines from line 1 up to 56 in file foo.c
</screen>
</sect2>
<sect2>
<title>Displaying</title>
<para>
A display is an expression that's evaluated and printed
after the execution of any <command>WineDbg</command>
command.
</para>
<screen>
display lists the active displays
info display (same as above command)
display &lt;expr&gt; adds a display for expression &lt;expr&gt;
display /fmt &lt;expr&gt; adds a display for expression &lt;expr&gt;. Printing
evaluated &lt;expr&gt; is done using the given format (see
print command for more on formats)
del display N deletes display #N
undisplay N (same as del display)
</screen>
</sect2>
<sect2>
<title>Disassembly</title>
<screen>
disas disassemble from current position
disas &lt;expr&gt; disassemble from address &lt;expr&gt;
disas &lt;expr&gt;,&lt;expr&gt;disassembles code between addresses specified by
the two &lt;expr&gt;
</screen>
</sect2>
<sect2>
<title>Information on Wine's internals</title>
<screen>
info class &lt;id&gt; prints information on Windows's class &lt;id&gt;
walk class lists all Windows' class registered in Wine
info share lists all the dynamic libraries loaded the debugged
program (including .so files, NE and PE DLLs)
info module N prints information on module of handle N
walk module lists all modules loaded by debugged program
info queue N prints information on Wine's queue N
walk queue lists all queues allocated in Wine
info regs prints the value of CPU register
info segment N prints information on segment N
info segment lists all allocated segments
info stack prints the values on top of the stack
info map lists all virtual mappings used by the debugged
program
info wnd N prints information of Window of handle N
walk wnd lists all the window hierarchy starting from the
desktop window
walk wnd N lists all the window hierarchy starting from the
window of handle N
walk process lists all w-processes in Wine session
walk thread lists all w-threads in Wine session
walk modref (no longer avail)
</screen>
</sect2>
<sect2>
<title>Memory (reading, writing, typing)</title>
<screen>
x &lt;expr&gt; examines memory at &lt;expr&gt; address
x /fmt &lt;expr&gt; examines memory at &lt;expr&gt; address using format /fmt
print &lt;expr&gt; prints the value of &lt;expr&gt; (possibly using its type)
print /fmt &lt;expr&gt; prints the value of &lt;expr&gt; (possibly using its
type)
set &lt;lval&gt;=&lt;expr&gt; writes the value of &lt;expr&gt; in &lt;lval&gt;
whatis &lt;expr&gt; prints the C type of expression &lt;expr&gt;
</screen>
<para>
<filename>/fmt</filename> is either <filename>/&lt;letter&gt;</filename> or
<filename>/&lt;count&gt;&lt;letter&gt;</filename> letter can be
</para>
<screen>
s =&gt; an ASCII string
u =&gt; an Unicode UTF16 string
i =&gt; instructions (disassemble)
x =&gt; 32 bit unsigned hexadecimal integer
d =&gt; 32 bit signed decimal integer
w =&gt; 16 bit unsigned hexadecimal integer
c =&gt; character (only printable 0x20-0x7f are actually
printed)
b =&gt; 8 bit unsigned hexadecimal integer
</screen>
</sect2>
</sect1>
<sect1 id="dbg-others">
<title>Other debuggers</title>
<sect2>
<title>Using other Unix debuggers</title>
<para>
You can also use other debuggers (like
<command>gdb</command>), but you must be aware of a few
items:
</para>
<para>
You need to attach the unix debugger to the correct unix
process (representing the correct windows thread) (you can
"guess" it from a <command>ps fax</command> for example:
When running the emulator, usually the first two
<varname>upids</varname> are for the Windows' application
running the desktop, the first thread of the application is
generally the third <varname>upid</varname>; when running a
Winelib program, the first thread of the application is
generally the first <varname>upid</varname>)
</para>
<note>
<para>
Even if latest <command>gdb</command> implements the
notion of threads, it won't work with Wine because the
thread abstraction used for implementing Windows' thread
is not 100% mapped onto the linux posix threads
implementation. It means that you'll have to spawn a
different <command>gdb</command> session for each Windows'
thread you wish to debug.
</para>
</note>
<!-- *** Extra content spliced in from article by Andreas Mohr *** -->
<para>
Following text written by &name-andreas-mohr; <email>&email-andreas-mohr;</email>
</para>
<para>
Here's how to get info about the current execution status of a
certain Wine process:
</para>
<para>
Change into your Wine source dir and enter:
</para>
<screen>
$ gdb wine
</screen>
<para>
Switch to another console and enter <command>ps ax | grep
wine</command> to find all wine processes. Inside
<command>gdb</command>, repeat for all Wine processes:
</para>
<screen>
(gdb) attach <userinput>PID</userinput>
</screen>
<para>
with <userinput>PID</userinput> being the process ID of one of
the Wine processes. Use
</para>
<screen>
(gdb) bt
</screen>
<para>
to get the backtrace of the current Wine process, i.e. the
function call history. That way you can find out what the
current process is doing right now. And then you can use
several times:
</para>
<screen>
(gdb) n
</screen>
<para>
or maybe even
</para>
<screen>
(gdb) b <userinput>SomeFunction</userinput>
</screen>
<para>
and
</para>
<screen>
(gdb) c
</screen>
<para>
to set a breakpoint at a certain function and continue up to
that function. Finally you can enter
</para>
<screen>
(gdb) detach
</screen>
<para>
to detach from the Wine process.
</para>
<!-- *** End of xtra content *** -->
</sect2>
<sect2>
<title>Using other Windows debuggers</title>
<para>
You can use any Windows' debugging API compliant debugger
with Wine. Some reports have been made of success with
VisualStudio debugger (in remote mode, only the hub runs
in Wine). GoVest fully runs in Wine.
</para>
</sect2>
<sect2>
<title>Main differences between winedbg and regular Unix debuggers</title>
<!-- FIXME: convert this into a table -->
<screen>
+----------------------------------+---------------------------------+
| WineDbg | gdb |
+----------------------------------+---------------------------------+
|WineDbg debugs a Windows' process:|gdb debugs a Windows' thread: |
|+ the various threads will be |+ a separate gdb session is |
| handled by the same WineDbg | needed for each thread of |
| session | Windows' process |
|+ a breakpoint will be triggered |+ a breakpoint will be triggered |
| for any thread of the w-process | only for the w-thread debugged |
+----------------------------------+---------------------------------+
|WineDbg supports debug information|gdb supports debug information |
|from: |from: |
|+ stabs (standard Unix format) |+ stabs (standard Unix format) |
|+ Microsoft's C, CodeView, .DBG | |
+----------------------------------+---------------------------------+
</screen>
</sect2>
</sect1>
<sect1 id="dbg-limits">
<title>Limitations</title>
<para>
16 bit processes are not supported (but calls to 16 bit code
in 32 bit applications are).
</para>
</sect1>
</chapter>
<!-- Keep this comment at the end of the file
Local variables:
mode: sgml
sgml-parent-document:("wine-doc.sgml" "set" "book" "part" "chapter" "")
End:
-->