tut

unix v7是最后一个广泛发布的研究型UNIX版本

the message
.P1
	a.out:running
.P2
.PP
The UNIX quit and interrupt signals
act on ADB itself rather than on the program being debugged.
If such a signal occurs then the program being debugged is stopped and control is returned to ADB.
The signal is saved by ADB and is passed on to the test program if:
.P1
	:c
.P2
is typed.
This can be useful when testing interrupt
handling routines.
The signal is not passed on to the test program if:
.P1
	:c  0
.P2
is typed.
.PP
Now let us reset the breakpoint at
.ul
settab
and display the instructions located there when we reach the breakpoint.
This is accomplished by:
.P1
	settab+4:b  settab,5?ia  \fR*
.P2
.FS
* Owing to a bug in early versions of ADB (including the
version distributed in Generic 3 UNIX) these statements
must be written as:
.br
.in 1i
\fBsettab+4:b	settab,5?ia;0\fR
.ft B
.br
getc+4,3:b	main.c?C;0
.br
settab+4:b	settab,5?ia; ptab/o;0
.br
.ft R
.in -1i
Note that \fB;0\fR will set dot to zero and stop at the breakpoint.
.FE
It is also possible to execute the ADB requests for each occurrence of the breakpoint but
only stop after the third occurrence by typing:
.P1
	getc+4,3:b  main.c?C  \fR*
.P2
This request will print the local variable 
.ul
c
in the function 
.ul
main
at each occurrence of the breakpoint.
The semicolon is used to separate multiple ADB requests on a single line.
.PP
Warning:
setting a breakpoint causes the value of dot to be changed;
executing the program under ADB does not change dot.
Therefore:
.P1
	settab+4:b  .,5?ia
	fopen+4:b
.P2
will print the last thing dot was set to
(in the example \fIfopen+4\fP)
.ul
not
the current location (\fIsettab+4\fP)
at which the program is executing.
.PP
A breakpoint can be overwritten without first deleting the old breakpoint.
For example:
.P1
	settab+4:b  settab,5?ia; ptab/o  \fR*
.P2
could be entered after typing the above requests.
.PP
Now the display of breakpoints:
.P1
	$b
.P2
shows the above request for the
.ul
settab
breakpoint.
When the breakpoint at
.ul
settab
is encountered the ADB requests are executed.
Note that the location at
.ul
settab+4
has been changed to plant the breakpoint;
all the other locations match their original value.
.PP
Using the functions,
.ul
f, g
and 
.ul
h
shown in Figure 3,
we can follow the execution of each function by planting non-stopping
breakpoints.
We call ADB with the executable program of Figure 3 as follows:
.P1
	adb ex3 \-
.P2
Suppose we enter the following breakpoints:
.P1
	h+4:b	hcnt/d;  h.hi/;  h.hr/
	g+4:b	gcnt/d;  g.gi/;  g.gr/
	f+4:b	fcnt/d;  f.fi/;  f.fr/
	:r
.P2
Each request line indicates that the variables are printed in decimal
(by the specification \fBd\fR).
Since the format is not changed, the \fBd\fR can be left off all but
the first request.
.PP
The output in Figure 7 illustrates two points.
First, the ADB requests in the breakpoint line are not
examined until the program under
test is run.
That means any errors in those ADB requests is not detected until run time.
At the location of the error ADB stops running the program.
.PP
The second point is the way ADB handles register variables.
ADB uses the symbol table to address variables.
Register variables, like \fIf.fr\fR above, have pointers to uninitialized
places on the stack.
Therefore the message "symbol not found".
.PP
Another way of getting at the data in this example is to print
the variables used in the call as:
.P1
	f+4:b	fcnt/d;  f.a/;  f.b/;  f.fi/
	g+4:b	gcnt/d;  g.p/;  g.q/;  g.gi/
	:c
.P2
The operator / was used instead of ?
to read values from the \fIcore\fP file.
The output for each function, as shown in Figure 7, has the same format.
For the function \fIf\fP, for example, it shows the name and value of the
.ul
external
variable
.ul
fcnt.
It also shows the address on the stack and value of the
variables
.ul
a, b
and
.ul
fi.
.PP
Notice that the addresses on the stack will continue to decrease
until no address space is left for program execution
at which time (after many pages of output)
the program under test aborts.
A display with names would be produced by requests like the following:
.P1
	f+4:b	fcnt/d;  f.a/"a="d;  f.b/"b="d;  f.fi/"fi="d
.P2
In this format the quoted string is printed literally and the \fBd\fP
produces a decimal display of the variables.
The results are shown in Figure 7.
.NH 2
Other Breakpoint Facilities
.LP
.IP \(bu 4
Arguments and change of standard input and output are passed to a program as:
.P1
	:r  arg1  arg2 ... <infile  >outfile
.P2
This request
kills any existing program under test and
starts the
.ul
a.out
afresh.
.IP \(bu
The program being debugged can be single stepped
by:
.P1
	:s
.P2
If necessary, this request will start up the program being
debugged and stop after executing
the first instruction.
.IP \(bu
ADB allows a program to be entered at a specific address
by typing:
.P1
	address:r
.P2
.IP \(bu
The count field can be used to skip the first \fIn\fR breakpoints as:
.P1
	,n:r
.P2
The request:
.P1
	,n:c
.P2
may also be used for skipping the first \fIn\fR breakpoints
when continuing a program.
.sp
.IP \(bu
A program can be continued at an address different from the breakpoint by:
.P1
	address:c
.P2
.IP \(bu
The program being debugged runs as a separate process and can be killed by:
.P1
	:k
.P2
.LP
.NH
Maps
.PP
UNIX supports several executable file formats.  These are used to tell
the loader how to load  the program file.  File type 407
is the most common and is generated by a C compiler invocation such as
\fBcc pgm.c\fP.
A 410 file is produced by a C compiler command of the form \fBcc -n pgm.c\fP,
whereas a 411 file is produced by \fBcc -i pgm.c\fP.
ADB interprets these different file formats and
provides access to the different segments through a set of maps (see Figure 8).
To print the maps type:
.P1
	$m
.P2
.PP
In 407 files, both text (instructions) and data are intermixed.
This makes it impossible for ADB to differentiate data from
instructions and some of the printed symbolic addresses look incorrect;
for example, printing data addresses as offsets from routines.
.PP
In 410 files (shared text), the instructions are separated from data and
\fB?*\fR accesses the data part of the \fIa.out\fP file.
The \fB?* \fP request tells ADB to use the second part of the
map in the
.ul
a.out
file.
Accessing data in the \fIcore\fP file shows
the data after it was modified by the execution of the program.
Notice also that the data segment may have grown during
program execution.
.PP
In 411 files (separated I & D space), the
instructions and data are also separated.
However, in this
case, since data is mapped through a separate set of segmentation
registers, the base of the data segment is also relative to address zero.
In this case since the addresses overlap it is necessary to use
the \fB?*\fR operator to access the data space of the \fIa.out\fP file.
In both 410 and 411 files the corresponding
core file does not contain the program text.
.PP
Figure 9 shows the display of three maps
for the same program linked as a 407, 410, 411 respectively.
The b, e, and f fields are used by ADB to map
addresses into file addresses.
The "f1" field is the
length of the header at the beginning of the file (020 bytes
for an \fIa.out\fP file and 02000 bytes for a \fIcore\fP file).
The "f2" field is the displacement from the beginning of the file to the data.
For a 407 file with mixed text and data this is the
same as the length of the header; for 410 and 411 files this
is the length of the header plus the size of the text portion.
.PP
The "b" and "e" fields are the starting and ending locations
for a segment.
Given an address, A, the location in
the file (either \fIa.out\fP or \fIcore\fP) is calculated as:
.P1
	b1\(<=A\(<=e1 =\h'-.5m'> file address = (A\-b1)+f1
	b2\(<=A\(<=e2 =\h'-.5m'> file address = (A\-b2)+f2
.P2
A user can access locations by using the ADB defined variables.
The \fB$v\fR request prints the variables initialized by ADB:
.P1
	b	base address of data segment
	d	length of the data segment
	s	length of the stack
	t	length of the text
	m	execution type (407,410,411)
.P2
.PP
In Figure 9 those variables not present are zero.
Use can be made of these variables by expressions such as:
.P1
	<b
.P2
in the address field.
Similarly the value of the variable can be changed by an assignment request
such as:
.P1
	02000>b
.P2
that sets \fBb\fP to octal 2000.
These variables are useful to know if the file under examination
is an executable or \fIcore\fP image file.
.PP
ADB reads the header of the \fIcore\fP image file to find the
values for these variables.
If the second file specified does not
seem to be a \fIcore\fP file, or if it is missing then the header of
the executable file is used instead.
.NH
Advanced Usage
.PP
It is possible with ADB to combine formatting requests
to provide elaborate displays.
Below are several examples.
.NH 2
Formatted dump
.PP
The line:
.P1
	<b,\-1/4o4^8Cn
.P2
prints 4 octal words followed by their ASCII interpretation
from the data space of the core image file.
Broken down, the various request pieces mean:
.sp
.in 1.7i
.ta .7i
.ti -.7i
<b	The base address of the data segment.
.sp
.ti -.7i
<b,\-1	Print from the base address to the end of file.
A negative count is used here and elsewhere to loop indefinitely
or until some error condition (like end of file) is detected.
.sp
.ti -1.7i
The format \fB4o4^8Cn\fR is broken down as follows:
.sp
.ti -.7i
4o	Print 4 octal locations.
.sp
.ti -.7i
4^	Backup the current address 4 locations (to the original start of the field).
.sp
.ti -.7i
8C	Print 8 consecutive characters using an escape convention;
each character in the range 0 to 037 is printed as @ followed by the corresponding character in the range 0140 to 0177.
An @ is printed as @@.
.sp
.ti -.7i
n	Print a newline.
.in -1.7i
.fi
.sp
.PP
The request:
.P1
	<b,<d/4o4^8Cn
.P2
could have been used instead to allow the printing to stop
at the end of the data segment (<d provides the data segment size in bytes).
.PP
The formatting requests can be combined with ADB's ability
to read in a script to produce a core image dump script.
ADB is invoked as:
.P1
	adb a.out core < dump
.P2
to read in a script file,
.ul
dump,
of requests.
An example of such a script is:
.P1
	120$w
	4095$s
	$v
	=3n
	$m
	=3n"C Stack Backtrace"
	$C
	=3n"C External Variables"
	$e
	=3n"Registers"
	$r
	0$s
	=3n"Data Segment"
	<b,\-1/8ona
.P2
.PP
The request \fB120$w\fP sets the width of the output to
120 characters
(normally, the width is 80 characters).
ADB attempts to print addresses as:
.P1
	symbol + offset
.P2
The request \fB4095$s\fP increases the maximum permissible offset
to the nearest symbolic address from 255 (default) to 4095.
The request \fB=\fP can be used to print literal strings.
Thus,
headings are provided in this
.ul
dump
program
with requests of the form:
.P1
	=3n"C Stack Backtrace"
.P2
that spaces three lines and prints the literal
string.
The request \fB$v\fP prints all non-zero ADB variables (see Figure 8).
The request
\fB0$s\fP
sets the maximum offset for symbol matches to zero thus
suppressing the printing of symbolic labels in favor
of octal values.
Note that this is only done for the printing of the data segment.
The request:
.P1
	<b,\-1/8ona
.P2
prints a dump from the base of the data segment to the end of file
with an octal address field and eight octal numbers per line.
.PP
Figure 11 shows the results of some formatting requests
on the C program of Figure 10.
.NH 2
Directory Dump
.PP
As another illustration (Figure 12) consider a set of requests to dump
the contents of a directory (which is made up
of an integer \fIinumber\fP followed by a 14 character name):
.P1
	adb dir \-
	=n8t"Inum"8t"Name"
	0,\-1? u8t14cn
.P2
In this example, the \fBu\fP prints the \fIinumber\fP as an unsigned decimal integer,
the \fB8t\fP means that ADB will space to the next
multiple of 8 on the output line, and the \fB14c\fP prints the 14 character file name.
.NH 2
Ilist Dump
.PP
Similarly the contents of the \fIilist\fP of a file system, (e.g. /dev/src,
on UNIX systems distributed by the UNIX Support Group;
see UNIX Programmer's
Manual Section V) could be dumped with the following set of 
requests:
.P1
	adb /dev/src \-
	02000>b
	?m <b
	<b,\-1?"flags"8ton"links,uid,gid"8t3bn",size"8tbrdn"addr"8t8un"times"8t2Y2na
.P2
In this example the value of the base for the map was changed 
to 02000 (by saying \fB?m<b\fR) since that is the start of an \fIilist\fP within a file system.
An artifice (\fBbrd\fP above) was used to print the 24 bit size field
as a byte, a space, and a decimal integer.
The last access time and last modify time are printed with the
\fB2Y\fR
operator.
Figure 12 shows portions of these requests as applied to a directory
and file system.
.NH 2
Converting values
.PP
ADB may be used to convert values from one representation to
another.
For example:
.P1
	072 = odx
.P2
will print
.P1
	072	58	#3a
.P2
which is the octal, decimal and hexadecimal representations
of 072 (octal).
The format is remembered so that typing
subsequent numbers will print them in the given formats.
Character values may be converted similarly, for example:
.P1
	'a' = co
.P2
prints
.P1
	a	0141
.P2
It may also be used to evaluate expressions but be
warned that all binary operators have
the same precedence which is lower than that for unary operators.
.NH
Patching
.PP
Patching files with ADB is accomplished with the 
.ul
write,
\fBw\fP or \fBW\fP, request (which is not like the \fIed\fP editor write command).
This is often used in conjunction with the 
.ul
locate,
\fBl\fP or \fBL\fP
request.
In general, the request syntax for \fBl\fP and \fBw\fP are similar as follows:
.P1
	?l value
.P2
The request \fBl\fP is used to match on two bytes, \fBL\fP is used for
four bytes.
The request \fBw\fP is used to write two bytes, whereas
\fBW\fP writes four bytes.
The \fBvalue\fP field in either 
.ul
locate
or
.ul
write
requests
is an expression.
Therefore, decimal and octal numbers, or character strings are supported.
.PP
In order to modify a file, ADB must be called as:
.P1
	adb \-w file1 file2
.P2
When called with this option, 
.ul
file1
and 
.ul
file2
are created if necessary and opened for both reading and writing.
.PP
For example, consider the C program shown in Figure 10.
We can change the word "This" to "The " in the executable file
for this program, \fIex7\fP, by using the following requests:
.P1
	adb \-w ex7 \-
	?l 'Th'
	?W 'The '
.P2
The request \fB?l\fP starts at dot and stops at the first match of "Th"
having set dot to the address of the location found.
Note the use of \fB?\fP to write to the 
.ul
a.out
file.
The form \fB?*\fP would have been used for a 411 file.
.PP
More frequently the 
request will be typed as:
.P1
	?l 'Th'; ?s
.P2
and locates the first occurrence of "Th" and print the entire string.
Execution of this ADB request will set dot to the address of the 
"Th" characters.
.PP
As another example of the utility of the patching facility,
consider a C program that has an internal logic flag.
The flag could be set by the user through ADB and the program run.
For example:
.P1
	adb a.out \-
	:s arg1 arg2
	flag/w 1
	:c
.P2
The \fB:s\fR request is normally used to single step through a process
or start a process in single step mode.
In this case it starts
.ul
a.out
as a subprocess
with arguments \fBarg1\fP and \fBarg2\fP.
If there is a subprocess running ADB writes to it rather than to the file
so the \fBw\fP request causes \fIflag\fP to be changed in the memory of the subprocess.
.NH
Anomalies
.PP
Below is a list of some strange things that users
should be aware of.
.IP 1.
Function calls and arguments are put on the stack by the C
save routine.
Putting breakpoints at the entry point to routines
means that the function appears not to have been called
when the
breakpoint occurs.
.IP 2.
When printing addresses, ADB uses
either text or data symbols from the \fIa.out\fP file.
This sometimes causes unexpected symbol names to be printed 
with data (e.g. \fIsavr5+022\fP).
This does not happen if
\fB?\fR is used for text (instructions)
and \fB/\fP for data.
.IP 3.
ADB cannot handle C register variables
in the most recently activated function.
.LP
.NH
Acknowledgements
.PP
The authors are grateful for the thoughtful comments
on how to organize this document
from R. B. Brandt, E. N. Pinson and B. A. Tague.
D. M. Ritchie made the system changes necessary to accommodate
tracing within ADB. He also participated in discussions 
during the writing of ADB.
His earlier work with DB and CDB led to many of the 
features found in ADB.
.SG MH-8234-JFM/1273-SRB-unix
.NH
References
.LP
.IP 1.
D. M. Ritchie and K. Thompson,
``The UNIX Time-Sharing System,''
CACM, July, 1974.
.IP 2.
B. W. Kernighan and D. M. Ritchie,
.ul
The C Programming Language,
Prentice-Hall, 1978.
.IP 3.
K. Thompson and D. M. Ritchie,
UNIX Programmer's Manual - 7th Edition,
1978.
.IP 4.
B. W. Kernighan and P. J. Plauger,
.ul
Software Tools,
Addison-Wesley, 1976.