Google Git
Sign in
gnu / dejagnu / refs/heads/origin / . / doc / user.xml
blob: 6af22323cacbf3e09fc2e931ee8ec24c4f75caa6 [file] [log] [blame]
<sect1 id="gettingup">
<title>Getting &dj; up and running</title>
<para>This chapter was originally written by Niklaus Giger
(ngiger@mus.ch) because he lost a week to figure out how &dj; works
and how to write a first test.
</para>
<para>Follow these instructions as closely a possible in order get a
good insight into how &dj; works, else you might run into a lot of
subtle problems. You have been warned.</para>
<para>It should be no big problems installing &dj; using your
package manager or from the source code. Under a Debian/GNU/Linux
systems just type (as root) <programlisting>apt-get install
dejagnu</programlisting>. These examples were run on a primary machine
with a AMD K6 and a Mac Powerbook G3 serving as a remote
target.</para>
<para> The tests for Windows were run under Windows using the actual
Cygwin version (1.3.x as of October 2001). Its target system was a PPC
embedded system running vxWorks. </para>
<sect2>
<title>Test your installation</title>
<para>Create a new user called "dgt" (&dj;Test), which uses bash as
it login shell. PS1 must be set to '\u:\w\$ ' in its ~/.bashrc. Login
as this user, create an empty directory and change the working
directory to it. e.g</para>
<programlisting>
dgt:~$ mkdir ~/dejagnu.test
dgt:~$ cd ~/dejagnu.test
</programlisting>
<para>Now you are ready to test &dj;'s main program called
runtest. The expecteted output is shown</para>
<example>
<title>Runtest output in a empty directory
</title>
<programlisting>
dgt:~/dejagnu.test$ runtest
WARNING: Couldn't find the global config file.
WARNING: No tool specified Test
Run By dgt on Sun Nov 25 17:07:03 2001 Native configuration is i586-pc-linux-gnu
=== tests ===
Schedule of variations: unix
Running target unix Using /usr/share/dejagnu/baseboards/unix.exp as board description file for target.
Using /usr/share/dejagnu/config/unix.exp as generic interface file for target.
ERROR: Couldn't find tool config file for unix.
=== Summary ===</programlisting>
<para>We will show you later how to get rid of all the WARNING- and
ERROR-messages. The files testrun.sum and testrun.log have been
created, which do not interest us at this point. Let's remove
them.</para>
<programlisting>:~/dejagnu.test$ rm testrun.sum testrun.log
</programlisting>
</example>
<sect3>
<title>Windows</title>
<para>On Windows systems &dj; is part of a port of a lot of Unix
tools to the Windows OS, called Cygwin. Cygwin may be downloaded and
installed from a mirror of http://www.cygwin.com/. All examples were
also run on Windows. If nothing is said, you can assume that you
should get the same output as on a Unix system.</para>
<para>You will need a telnet daemon if you want to use a Windows box
as a remote target. There seems to be a freeware telnet daemon at
http://www.fictional.net/.</para>
</sect3>
<sect3>
<title>Getting the source code for the calc example</title>
<para>If you are running a Debian distribution you can find the
examples under /usr/share/doc/dejagnu/examples. These examples seem to
be missing in Red Hat's RPM. In this case download the sources of
&dj; and adjust the pathes to the &dj; examples
accordingly.</para>
</sect3>
</sect2>
<sect2>
<title>Create a minimal project, e.g. calc</title>
<para>In this section you will to start a small project,
using the sample application calc, which is part of your &dj;
distribution</para>
<sect3><title>A simple project without the GNU autotools</title>
<para>The runtest program can be run standalone. All the
autoconf/automake support is just cause those programs are commonly
used for other GNU applications. The key to running runtest standalone
is having the local site.exp file setup correctly, which automake
does.</para>
<para>The generated site.exp should like like:</para>
<programlisting>
set tool calc
set srcdir .
set objdir /home/dgt/dejagnu.test
</programlisting></sect3>
<sect3>
<title>Using autoconf/autoheader/automake</title>
<para>We have to prepare some input file in order to run autocon and
automake. There is book &quot;GNU autoconf, automake and
libtool&quot; by Garry V. Vaughan, et al. NewRider, ISBN
1-57870-190-2 which describes this process thoroughly.</para>
<para>From the calc example distributed with the &dj; documentation
you should copy the program file itself (calc.c) and some additional
files, which you might examine a little bit close to derive their
meanings.</para>
<programlisting>
dgt:~/dejagnu.test$ cp -r /usr/share/doc/dejagnu/examples/calc/\
{configure.in,Makefile.am,calc.c,testsuite} .
</programlisting>
<para>In Makemake.am note the presence of the AUTOMAKE_OPTIONS = dejagnu. This option is needed.</para>
<para>Run aclocal to generate aclocal.m4, which is a collection of
macros needed by configure.in</para>
<programlisting>
dgt:~/dejagnu.test$ aclocal
</programlisting>
<para>autoconf is another part of the auto-tools. Run it to generate
configure based on information contained in configure.in.</para>
<programlisting>
dgt:~/dejagnu.test$ autoconf
</programlisting>
<para>autoheader is another part of the auto-tools.
Run it to generate calc.h.in. </para>
<programlisting>
dgt:~/dejagnu.test$ autoheader
</programlisting>
<para>The Makefile.am of this example was developed as port of the &dj;
distribution.
Adapt Makefile.am for this test. Replace the line
&quot;#noinst_PROGRAMS = calc&quot; to
&quot;bin_PROGRAMS = calc&quot;.
Change the RUNTESTDEFAULTFLAGS from
&quot;$$srcdir/testsuite&quot; to
&quot;./testsuite&quot;.</para>
<para>Running automake at this point contains a series of warning in
its output as shown in the following example:</para>
<example>
<title>Sample output of automake with missing files</title>
<programlisting>
dgt:~/dejagnu.test$ automake --add-missing
automake: configure.in: installing `./install-sh'
automake: configure.in: installing `./mkinstalldirs'
automake: configure.in: installing `./missing'
automake: Makefile.am: installing `./INSTALL'
automake: Makefile.am: required file `./NEWS' not found
automake: Makefile.am: required file `./README' not found
automake: Makefile.am: installing `./COPYING'
automake: Makefile.am: required file `./AUTHORS' not found
automake: Makefile.am: required file `./ChangeLog' not found
configure.in: 4: required file `./calc.h.in' not found
Makefile.am:6: required directory ./doc does not exist
</programlisting>
</example>
<para>Create a empty directory doc and empty files
INSTALL, NEWS, README, AUTHORS, ChangeLog and COPYING.
The default COPYING will point to the GNU Public License (GPL).
In a real project it would be time to add some meaningfull text in each file.
</para>
<para>Adapt calc to your environment by calling configure.</para>
<example>
<title>Sample output of configure
</title>
<programlisting>
dgt:~/dejagnu.test$ ./configure
creating cache ./config.cache
checking whether to enable maintainer-specific portions of Makefiles... no
checking for a BSD compatible install... /usr/bin/install -c
checking whether build environment is sane... yes
checking whether make sets ${MAKE}... yes
checking for working aclocal... found
checking for working autoconf... found
checking for working automake... found
checking for working autoheader... found
checking for working makeinfo... found
checking for gcc... gcc checking whether the C compiler (gcc ) works... yes
checking whether the C compiler (gcc ) is a cross-compiler... no
checking whether we are using GNU C... yes
checking whether gcc accepts -g... yes
checking for a BSD compatible install... /usr/bin/install -c
checking how to run the C preprocessor... gcc -E
checking for stdlib.h... yes
checking for strcmp... yes
updating cache ./config.cache
creating ./config.status
creating Makefile creating calc.h
</programlisting>
</example>
<para>If you are familiar with GNU software,
this output should not contain any surprise to you.
Any errors should be easy to fix for such a simple program.</para>
<para>Build the calc executable:</para>
<example>
<title>Sample output building calc
</title>
<programlisting>
dgt:~/dejagnu.test$ make
gcc -DHAVE_CONFIG_H -I. -I. -I. -g -O2 -c calc.c
gcc -g -O2 -o calc calc.o
</programlisting>
</example>
<para>You prepared a few files and then called some
commands. Respecting the right order assures a automatic and correctly
compiled calc program. The following example resumes the correct
order.</para>
<example>
<title>Creating the calc program using the GNU autotools</title>
<programlisting>
dgt:~/dejagnu.test$ aclocal
dgt:~/dejagnu.test$ autoconf
dgt:~/dejagnu.test$ autoheader
dgt:~/dejagnu.test$ automake --add-missing
dgt:~/dejagnu.test$ ./configure
dgt:~/dejagnu.test$ make
</programlisting>
</example>
<para>Play with calc and verify whether it works correctly.
A sample session might look like this:</para>
<programlisting>
dgt:~/dejagnu.test$ ./calc
calc: version
Version: 1.1
calc:<emphasis> </emphasis>add 3 4
7
calc: multiply 3 4<emphasis> </emphasis>
12
calc: multiply 2 4<emphasis> </emphasis>
12
calc: quit
</programlisting>
<para>Look at the intentional bug that 2 times 4 equals 12.</para>
<para>The tests run by &dj; need a file called site.exp,
which is automatically generated if we call &quot;make
site.exp&quot;. This was the purpose of the &quot;AUTOMAKE_OPTIONS =
dejagnu&quot; in Makefile.am.</para>
<example>
<title>Sample output generating a site.exp</title>
<programlisting>
dgt: make site.exp
dgt:~/dejagnu.test$ make site.exp
Making a new site.exp file...
</programlisting>
</example>
</sect3>
</sect2>
<sect2>
<title>Our first automated tests</title>
<sect3><title>Running the test for the calc example</title>
<para>Now we are ready to call the automated tests </para>
<example>
<title>Sample output of runtest in a configured directory</title>
<programlisting>
dgt:~/dejagnu.test$ make check
make check-DEJAGNU
make[1]: Entering directory `/home/dgt/dejagnu.test' srcdir=`cd . &amp;&amp; pwd`; export srcdir; \
EXPECT=expect; export EXPECT; \ runtest=runtest; \
if /bin/sh -c "$runtest --version" > /dev/null 2>&amp;1; then \
$runtest --tool calc CALC=`pwd`/calc --srcdir ./testsuite ; \
else echo "WARNING: could not find \`runtest'" 1>&amp;2; :;\
fi
WARNING: Couldn't find the global config file.
WARNING: Couldn't find tool init file
Test Run By dgt on Sun Nov 25 21:42:21 2001
Native configuration is i586-pc-linux-gnu
=== calc tests ===
Schedule of variations:
unix
Running target unix
Using /usr/share/dejagnu/baseboards/unix.exp as board description file for target.
Using /usr/share/dejagnu/config/unix.exp as generic interface file for target.
Using ./testsuite/config/unix.exp as tool-and-target-specific interface file.
Running ./testsuite/calc.test/calc.exp ...
FAIL: multiply2 (bad match)
=== calc Summary ===
# of expected passes 5
# of unexpected failures 1
/home/Dgt/dejagnu.test/calc version Version: 1.1
make[1]: *** [check-DEJAGNU] Fehler 1
make[1]: Leaving directory `/home/Dgt/dejagnu.test' make: *** [check-am] Fehler 2
</programlisting>
</example>
<para>Did you see the line &quot;FAIL:&quot;? The test cases for calc catch the bug in the calc.c file. Fix the error in calc.c later as the following examples assume a unchanged calc.c.</para>
<para>Examine the output files calc.sum and calc.log. Try to
understand the testcases written in
~/dejagnu.test/testsuite/calc.test/calc.exp. To understand Expect you
might take a look at the book &quot;Exploring Expect&quot;, which is
an excellent resource for learning and using Expect. (Pub: O'Reilly,
ISBN 1-56592-090-2) The book contains hundreds of examples and also
includes a tutorial on Tcl. Exploring Expect is 602 pages
long.</para>
</sect3>
<sect3><title>The various config files or how to avoid warnings</title>
<para>&dj; may be customized by each user. It first searches for a
file called ~/.dejagnurc. Create the file ~/.dejagnurc and insert the
following line:</para>
<programlisting>
puts "I am ~/.dejagnurc"
</programlisting>
<para>Rerun make check. Test whether the output contains "I am ~/.dejagnurc".
Create ~/my_dejagnu.exp and insert the following line:</para>
<programlisting>
puts "I am ~/my_dejagnu.exp"
</programlisting>
<para>In a Bash-Shell enter</para>
<programlisting>
dgt:~/dejagnu.test$ export DEJAGNU=~/my_dejagnu.exp
</programlisting>
<para>Run &quot;make check&quot; again. The output should not contain
&quot;WARNING: Couldn't find the global config file.&quot;.
Create the sub-directory lib. Create the file &quot;calc.exp&quot; in it and insert the following line:</para>
<programlisting>
puts "I am lib/calc.exp"
</programlisting>
<para>The last warning &quot;WARNING: Couldn't find tool init file&quot;
should not be part of the output of make check.
Create the directory ~/boards. Create the file ~/boards/standard.exp and insert the following line:</para>
<programlisting>
puts "I am boards/standard.exp"
</programlisting>
<para>If the variable DEJAGNU is still not empty then the (abbreviated) output of &quot;make check&quot; should look like this:</para>
<example>
<title>Sample output of runtest with the usual configuration files</title>
<programlisting>
dgt:~/dejagnu.test$ make check
&lt;...&gt;
fi
I am ~/.dejagnurc
I am ~/my_dejagnu.exp
I am lib/calc.exp
Test Run By dgt on Sun Nov 25 22:19:14 2001
Native configuration is i586-pc-linux-gnu
=== calc tests ===
Using /home/Dgt/boards/standard.exp as standard board description\
file for build.
I am ~/boards/standard.exp
Using /home/Dgt/boards/standard.exp as standard board description\
file for host.
I am ~/boards/standard.exp
Schedule of variations:
unix
Running target unix
Using /home/Dgt/boards/standard.exp as standard board description\
file for target.
I am ~/boards/standard.exp
Using /usr/share/dejagnu/baseboards/unix.exp as board description file\
for target.
&lt;...&gt;
</programlisting>
</example>
<para>It is up to you to decide when and where to use any of the above
mentioned config files for customizing.
This chapters showed you where and in which order the different config
files are run.</para>
</sect3>
<sect3>
<title>When trouble strikes</title>
<para>Calling runtest with the '-v'-flag shows you in even more details which files are searched in which order. Passing it several times gives more and more details. </para>
<example>
<title>Displaying details about runtest execution</title>
<programlisting>
runtest -v -v -v --tool calc CALC=`pwd`/calc --srcdir ./testsuite
</programlisting>
</example>
<para>Calling runtest with the '--debug'-flag logs a lot of details to dbg.log where you can analyse it afterwards. </para>
<para>In all test cases you can temporary adjust the verbosity of information by adding the following Tcl-command to any tcl file that gets loaded by
dejagnu, for instance, ~/.dejagnurc:</para>
<programlisting>
set verbose 9
</programlisting>
</sect3>
<sect3>
<title>Testing &quot;Hello world&quot; locally</title>
<para>This test checks, whether the built-in shell command &quot;echo Hello world&quot;
will really write &quot;Hello world&quot; on the console.
Create the file ~/dejagnu.test/testsuite/calc.test/local_echo.exp.
It should contain the following lines</para>
<example>
<title>A first (local) test case</title>
<programlisting>
set test "Local Hello World"
send "echo Hello World"
expect {
-re "Hello World" { pass "$test" }
}
</programlisting>
</example>
<para>Run runtest again and verify the output &quot;calc.log&quot;</para>
</sect3>
</sect2>
<sect2>
<title>A first remote test</title>
<para>Testing remote targets is a lot trickier especially if you are using an
embedded target
which has no built in support for things like a compiler, ftp server or a Bash-shell.
Before you can test calc on a remote target you have to acquire a few basics skills.</para>
<sect3>
<title>Setup telnet to your own host</title>
<para>The easiest remote host is usually the host you are working on.
In this example we will use telnet to login in your own workstation.
For security reason you should never have a telnet deamon running on
machine connected on the internet, as password and usernames are transmitted
in clear text.
We assume you know how to setup your machine for a telnet daemon.</para>
<para>Next try whether you may login in your own host by issuing the
command &quot;telnet localhost.1&quot;. In order to be able to
distinguish between a normal session an a telnet login add the following lines to /home/dgt/.bashrc.</para>
<programlisting>
if [ "$REMOTEHOST" ]
then
PS1='remote:\w\$ '
fi
</programlisting>
<para>Now on the machine a &quot;remote&quot; login looks like this:</para>
<example>
<title>Sample log of a telnet login to localhost</title>
<programlisting>
dgt:~/dejagnu.test$ telnet localhost
Trying 127.0.0.1...
Connected to 127.0.0.1.
Escape character is '^]'.
Debian GNU/Linux testing/unstable Linux
K6Linux login: dgt
Password:
Last login: Sun Nov 25 22:46:34 2001 from localhost on pts/4
Linux K6Linux 2.4.14 #1 Fre Nov 16 19:28:25 CET 2001 i586 unknown
No mail.
remote:~$ exit
logout
Connection closed by foreign host.
</programlisting>
</example>
</sect3>
<sect3>
<title>A test case for login via telnet</title>
<para>In order to define a correct setup we have add a line containing
&quot;set target unix&quot; either to ~/.dejagnurc or to ~/my_dejagnu.exp.
In ~/boards/standard.exp add the following four lines to define a few patterns for the &dj; telnet login procedure.</para>
<example>
<title>Defining a remote target board</title>
<programlisting>
set_board_info shell_prompt "remote:"
set_board_info telnet_username "dgt"
set_board_info telnet_password "top_secret"
set_board_info hostname "localhost"
</programlisting>
</example>
<para>As &dj; will be parsing the telnet session output for some well
known pattern the output there are a lot of things that can go wrong.
If you have any problems verify your setup:</para>
<itemizedlist>
<listitem>
<para>Is <filename>/etc/motd</filename> empty?</para></listitem>
<listitem>
<para>Is <filename>/etc/issue.net</filename> empty?</para></listitem>
<listitem>
<para>Exists a empty <filename>~/.hushlogin</filename>?</para></listitem>
<listitem>
<para>The LANG environment variable must be either empty or set to &quot;C&quot;. </para></listitem>
</itemizedlist>
<para>To test the login via telnet write a sample test case.
Create the file ~/dejagnu.test/testsuite/calc.test/remote_echo.exp and
add the following few lines:</para>
<example>
<title>&dj; script for logging in into a remote target</title>
<programlisting>
puts "this is remote_echo.exp target for $target "
target_info $target
#set verbose 9
set shell_id [remote_open $target]
set test "Remote login to $target"
#set verbose 0
puts "Spawn id for remote shell is $shell_id"
if { $shell_id > 0 } {
pass "$test"
} else {
fail "Remote open to $target"
}
</programlisting>
</example>
<para>In the runtest output you should find something like:</para>
<programlisting>
Running ./testsuite/calc.test/local_echo.exp ...
Running ./testsuite/calc.test/remote_echoo.exp ...
this is remote_echo.exp target is unix
Spawn id for remote shell is exp7
</programlisting>
<para>Have again a look at calc.log to get a feeling how &dj; and expect
parse the input. </para></sect3>
<sect3>
<title>Remote testing &quot;Hello world&quot;</title>
<para>Next you will transform the above &quot;hello world&quot; example to
its remote equivalent.
This can be done by adding the following lines to our file remote_echo.exp.</para>
<example>
<title>A first (local) remote "Hello world" test</title>
<programlisting>
set test "Remote_send Hello World"
set status [remote_send $target "echo \"Hello\" \"World\"\n" ]
pass "$test"
set test "Remote_expect Hello World"
remote_expect $target 5 {
-re "Hello World" { pass "$test" }
}
</programlisting>
</example>
<para>Call make check. The output should contain
&quot;# of expected passes 9&quot; and &quot;# of unexpected failures 1&quot;.</para>
<para>Have a look at the procedures in /usr/share/dejagnu/remote.exp to have an overview of the offered procedures and their features. </para>
<para>Now setup a real target.
In the following example we assume as target a PowerBook running Debian.
As above add a test user "dgt", install telnet and FTP servers.
In order to distinguish it from the host add the line
<programlisting>PS1='test:>'</programlisting> to /home/dgt/.bash_profile.
Also add a corresponding entry "powerbook" to /etc/hosts and verify that you
are able to ping, telnet and ftp to the target "powerbook".</para>
<para>In order to let runtest run its test on the "powerbook" target change the following lines in ~/boards/standard.exp:</para>
<example>
<title>Board definition for a remote target</title>
<programlisting>
set_board_info protocol "telnet"
set_board_info telnet_username "dgt"
set_board_info telnet_password "top_secret"
set_board_info shell_prompt "test:> "
set_board_info hostname "powerbook"
</programlisting>
</example>
<para>Now call runtest again with the same arguments and verify whether all went okay by taking a close look at calc.log.</para>
</sect3>
<sect3>
<title>Transferring files from/to the target</title>
<para>A simple procedure like this will do the job for you:</para>
<example>
<title>Test script to transfer a file to a remote target</title>
<programlisting>
set test "Remote_download"
puts "Running Remote_download"
# set verbose 9
set remfile /home/dgt/dejagnu2
set status [remote_download $target /home/dgt/.dejagnurc $remfile]
if { "$status" == "" } {
fail "Remote download to $remfile on $target"
} else {
pass "$test"
}
puts "status of remote_download ist $status"
# set verbose 0
</programlisting>
</example>
<para>After running runtest again, check whether the file dejagnu2 exists on the target.
This example will only work if the rcp command works with your target.
If you have a working FTP-server on the target you can use it by adding the
following lines to ~/boards/standard.exp:</para>
<example>
<title>Defining a board to use FTP as file transport</title>
<programlisting>
set_board_info file_transfer "ftp"
set_board_info ftp_username "dgt"
set_board_info ftp_password "1234"
</programlisting>
</example>
</sect3>
<sect3>
<title>Preparing for crosscompilation</title>
<para>For crosscompiling you need working binutils, gcc and a base library like
libc or glib for your target.
It is beyond the scope of this document to describe how to get it working.
The following examples assume a cross compiler for PowerPC which is called linux-powerpc-gcc.
</para>
<para>Add AC_CANONICAL_TARGET in dejagnu.test/configure.in at the following location. Copy config.guess from /usr/share/automake to dejagnu.test.</para>
<programlisting>
AM_CONFIG_HEADER(calc.h)
AC_CANONICAL_TARGET([])
AM_INIT_AUTOMAKE(calc, 1.1)
</programlisting>
<para>You need to run automake 2.5 or later.
Depending on your installation calling autoconf2.5 instead of autoconf is not needed.
The sequence to regenerate all files is:</para>
<example>
<title>Using autotools for cross development</title>
<programlisting>
$ autoconf2.5
$ autoheader
$ automake
$ ./configure --host=powerpc-linux --target=powerpc-linux
configure: WARNING: If you wanted to set the --build type, don't use --host.
If a cross compiler is detected then cross compile mode will be used.
checking build system type... ./config.guess: ./config.guess: No such file or directory
configure: error: cannot guess build type; you must specify one
$ cp /usr/share/automake/config.guess .
$ ./configure --host=powerpc-linux --target=powerpc-linux
configure: WARNING: If you wanted to set the --build type, don't use --host.
If a cross compiler is detected then cross compile mode will be used. \
checking build system type... i586-pc-linux-gnu
checking host system type... powerpc-unknown-linux-gnu
&lt;...&gt;
checking whether we are cross compiling... yes
&lt;...&gt;
Configuration:
Source code location: .
C Compiler: powerpc-linux-gcc
C Compiler flags: -g -O2
</programlisting>
</example>
<para>Everything should be ready to recompile for the target:</para>
<programlisting>$ make
powerpc-linux-gcc -DHAVE_CONFIG_H -I. -I. -I. -g -O2 -c calc.c
powerpc-linux-gcc -g -O2 -o calc calc.o
</programlisting>
</sect3>
<sect3>
<title>Remote testing of calc</title>
<para>Not yet written, as I have problem getting libc6-dev-powerpc to work. Probably I first have to build my cross compiler. </para>
</sect3>
<sect3>
<title>Using Windows as host and vxWorks as target</title>
<para>A more thorough walk-through will be written in a few weeks.</para>
<para>In order to test the vxWorks as a target I changed boards/standards.exp to reflect my settings (IP, username, password). Then I reconfigured vxWorks to include a FTP and telnet server (using the same username/password combination ad in boards/standard.exp).</para>
<para>With this setup and some minor modification (e.g. replacing echo by printf) in my test cases I could test my vxWorks system. It sure does not seem to be a correct setup by &dj; standard. For instance, it still loading /usr/share/dejagnu/baseboards/unix.exp instead of vxWorks. In any case I found that (at least under Windows) I did not find out how the command line would let me override settings in my personal config files.</para>
</sect3>
</sect2>
</sect1>
<sect1 id="runningtests">
<title>Running Tests</title>
<para>There are two ways to execute a testsuite. The most
common way is when there is existing support in the
<filename>Makefile</filename>. This support consists of a
<emphasis>check</emphasis> target. The other way is to execute the
<command>runtest</command> program directly. To run
<command>runtest</command> directcly from the command line requires
either all the correct options, or the <xref linkend="local"/> must be setup
correctly.</para>
<sect2 id="makecheck" xreflabel="Make Check">
<title>Make check</title>
<para>To run tests from an existing collection, first use
<command>configure</command> as usual to set up the
build directory. Then try typing:</para>
<screen>
make check
</screen>
<para>If the <emphasis>check</emphasis> target exists, it
usually saves you some trouble. For instance, it can set up any
auxiliary programs or other files needed by the tests. The most
common file the check builds is the
<emphasis>site.exp</emphasis>. The site.exp file contains
various variables that &dj; used to dertermine the
configuration of the program being tested. This is mostly for
supporting remote testing.</para>
<para>The <emphasis>check</emphasis> target is supported by GNU
<productname>Automake</productname>. To have &dj; support added to your
generated <filename>Makefile.in</filename>, just add the keyword
dejagnu to the AUTOMAKE_OPTIONS variable in your
<filename>Makefile.am</filename> file.</para>
<para>Once you have run <emphasis>make check</emphasis> to build
any auxiliary files, you can invoke the test driver
<command>runtest</command> directly to repeat the tests.
You will also have to execute <command>runtest</command>
directly for test collections with no
<emphasis>check</emphasis> target in the
<filename>Makefile</filename>.</para>
</sect2>
<sect2 id="runtest" xreflabel="Runtest">
<title>Runtest</title>
<para><command>runtest</command> is the executable test driver
for &dj;. You can specify two kinds of things on the
<command>runtest</command> command line: command line options,
and Tcl variables for the test scripts. The options are listed
alphabetically below.</para>
<para><command>runtest</command> returns an exit code of
<emphasis>1</emphasis> if any test has an unexpected result; otherwise
(if all tests pass or fail as expected) it returns <emphasis>0</emphasis>
as the exit code.</para>
<sect3 id="outputs" xreflabel="Output States">
<title>Output States</title>
<para><filename>runtest</filename> flags the outcome of each
test as one of these cases. <xref linkend="posix"/> for a
discussion of how POSIX specifies the meanings of these
cases.</para>
<variablelist>
<varlistentry>
<term>PASS</term>
<listitem><para>The most desirable outcome: the test succeeded, and
was expected to succeed.</para></listitem>
</varlistentry>
<varlistentry>
<term>XPASS</term>
<listitem><para>A pleasant kind of failure: a test was expected to
fail, but succeeded. This may indicate progress; inspect the test
case to determine whether you should amend it to stop expecting
failure.</para></listitem>
</varlistentry>
<varlistentry>
<term>FAIL</term>
<listitem><para>A test failed, although it was expected to succeed.
This may indicate regress; inspect the test case and the failing
software to locate the bug.</para></listitem>
</varlistentry>
<varlistentry>
<term>XFAIL</term>
<listitem><para>A test failed, but it was expected to fail. This
result indicates no change in a known bug. If a test fails because
the operating system where the test runs lacks some facility required
by the test, the outcome is <emphasis>UNSUPPORTED</emphasis>
instead.</para></listitem>
</varlistentry>
<varlistentry>
<term>UNRESOLVED</term>
<listitem><para>Output from a test requires manual inspection; the
testsuite could not automatically determine the outcome. For
example, your tests can report this outcome is when a test does not
complete as expected.</para></listitem>
</varlistentry>
<varlistentry>
<term>UNTESTED</term>
<listitem><para>A test case is not yet complete, and in particular
cannot yet produce a <emphasis>PASS</emphasis> or
<emphasis>FAIL</emphasis>. You can also use this outcome in dummy
``tests'' that note explicitly the absence of a real test case for a
particular property.</para></listitem>
</varlistentry>
<varlistentry>
<term>UNSUPPORTED</term>
<listitem><para>A test depends on a conditionally available feature
that does not exist (in the configured testing environment). For
example, you can use this outcome to report on a test case that does
not work on a particular target because its operating system support
does not include a required subroutine.</para></listitem>
</varlistentry>
</variablelist>
<para>runtest may also display the following messages:</para>
<variablelist>
<varlistentry>
<term>ERROR</term>
<listitem><para>Indicates a major problem (detected by the test case
itself) in running the test. This is usually an unrecoverable error,
such as a missing file or loss of communication to the target. (POSIX
testsuites should not emit this message; use
<emphasis>UNSUPPORTED</emphasis>, <emphasis>UNTESTED</emphasis>, or
<emphasis>UNRESOLVED</emphasis> instead, as
appropriate.)</para></listitem>
</varlistentry>
<varlistentry>
<term>WARNING</term>
<listitem><para>Indicates a possible problem in running the
test. Usually warnings correspond to recoverable errors, or display
an important message about the following tests.</para></listitem>
</varlistentry>
<varlistentry>
<term>NOTE</term>
<listitem><para>An informational message about the test
case.</para></listitem>
</varlistentry>
</variablelist>
</sect3>
<sect3 id="invoking" xreflabel="Invoking Runtest">
<title>Invoking Runtest</title>
<para>This is the full set of command line options that
<filename>runtest</filename> recognizes. Arguments may be
abbreviated to the shortest unique string.</para>
<variablelist>
<varlistentry>
<term><option>--all</option> (-a)</term>
<listitem><para>Display all test output. By default,
<emphasis>runtest</emphasis> shows only the output of tests that
produce unexpected results; that is, tests with status
<emphasis>FAIL</emphasis> (unexpected failure),
<emphasis>XPASS</emphasis> (unexpected success), or
<emphasis>ERROR</emphasis> (a severe error in the test case
itself). Specify <emphasis>--all</emphasis> to see output for tests
with status <emphasis>PASS</emphasis> (success, as expected)
<emphasis>XFAIL</emphasis> (failure, as expected), or
<emphasis>WARNING</emphasis> (minor error in the test case
itself).</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--build [string]</option></term>
<listitem><para><emphasis>string</emphasis> is a full configuration
``triple'' name as used by <command>configure</command>. This
is the type of machine &dj; and the tools to be tested are built
on. For a normal cross this is the same as the host, but for a
canadian cross, they are seperate.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--host [string]</option></term>
<listitem><para><symbol>string</symbol> is a full configuration
``triple'' name as used by <emphasis>configure</emphasis>. Use this
option to override the default string recorded by your
configuration's choice of host. This choice does not change how
anything is actually configured unless --build is also specified; it
affects <emphasis>only</emphasis> &dj; procedures that compare the
host string with particular values. The procedures
<emphasis>ishost</emphasis>, <emphasis>istarget</emphasis>,
<emphasis>isnative</emphasis>, and <emphasis>setup</emphasis>xfail}
are affected by <emphasis>--host</emphasis>. In this usage,
<emphasis>host</emphasis> refers to the machine that the tests are to
be run on, which may not be the same as the
<emphasis>build</emphasis> machine. If <emphasis>--build</emphasis>
is also specified, then <emphasis>--host</emphasis> refers to the
machine that the tests wil, be run on, not the machine &dj; is run
on.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--host_board [name]</option></term>
<listitem><para>The host board to use.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--target [string]</option></term>
<listitem><para>Use this option to override the default setting
(running native tests). <emphasis>string</emphasis> is a full
configuration ``triple'' name of the form
<emphasis>cpu-vendor-os</emphasis> as used by
<command>configure</command>. This option changes the
configuration <emphasis>runtest</emphasis> uses for the default tool
names, and other setup information.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--debug</option> (-de)</term>
<listitem><para>Turns on the <emphasis>expect</emphasis> internal
debugging output. Debugging output is displayed as part of the
<emphasis>runtest</emphasis> output, and logged to a file called
<filename>dbg.log</filename>. The extra debugging output does
<emphasis>not</emphasis> appear on standard output, unless the
verbose level is greater than 2 (for instance, to see debug output
immediately, specify <emphasis>--debug</emphasis>-v -v}). The
debugging output shows all attempts at matching the test output of
the tool with the scripted patterns describing expected output. The
output generated with <emphasis>--strace</emphasis> also goes into
<filename>dbg.log</filename>.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--help</option> (-he)</term>
<listitem><para>Prints out a short summary of the
<emphasis>runtest</emphasis> options, then exits (even if you also
specify other options).</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--ignore [name(s)] </option></term>
<listitem><para>The names of specific tests to
ignore.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--objdir [path]</option></term>
<listitem><para>Use <emphasis>path</emphasis> as the top directory
containing any auxiliary compiled test code. This defaults to
<filename>.</filename>. Use this option to locate pre-compiled test
code. You can normally prepare any auxiliary files needed with
<emphasis>make</emphasis>.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--outdir [path]</option></term>
<listitem><para>Write output logs in directory
<filename>path</filename>. The default is <emphasis>.},
the</emphasis> directory where you start
<emphasis>runtest</emphasis>. This option affects only the summary
and the detailed log files
<filename>tool.sum</filename> and
<filename>tool.log</filename>. The &dj; debug
log <filename>dbg.log</filename> always appears (when requested) in
the local directory.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--reboot [name]</option></term>
<listitem><para>Reboot the target board when
<emphasis>runtest</emphasis> initializes. Usually, when running tests
on a separate target board, it is safer to reboot the target to be
certain of its state. However, when developing test scripts,
rebooting takes a lot of time.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--srcdir [path]</option></term>
<listitem><para>Use <filename>path</filename> as the top directory
for test scripts to run. <emphasis>runtest</emphasis> looks in this
directory for any subdirectory whose name begins with the toolname
(specified with <emphasis>--tool</emphasis>). For instance, with
<emphasis>--tool</emphasis>gdb}, <emphasis>runtest</emphasis> uses
tests in subdirectories <filename>gdb.*</filename> (with the usual
shell-like filename expansion). If you do not use
<emphasis>--srcdir</emphasis>, <emphasis>runtest</emphasis> looks for
test directories under the current working
directory.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--strace [number]</option></term>
<listitem><para>Turn on internal tracing for
<emphasis>expect</emphasis>, to n levels deep. By adjusting the
level, you can control the extent to which your output expands
multi-level Tcl statements. This allows you to ignore some levels of
<emphasis>case</emphasis> or <emphasis>if</emphasis> statements.
Each procedure call or control structure counts as one ``level''. The
output is recorded in the same file, <filename>dbg.log</filename>,
used for output from <emphasis>--debug</emphasis>.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--connect [program]</option></term>
<listitem><para>Connect to a target testing environment as specified
by <emphasis>type</emphasis>, if the target is not the computer
running <emphasis>runtest</emphasis>. For example, use
<emphasis>--connect</emphasis> to change the program used to connect
to a ``bare board'' boot monitor. The choices for
<emphasis>type</emphasis> in the &dj; 1.4 distribution are
<emphasis>rlogin</emphasis>, <emphasis>telnet</emphasis>,
<emphasis>rsh</emphasis>, <emphasis>tip</emphasis>,
<emphasis>kermit</emphasis>, and <emphasis>mondfe</emphasis>.</para>
<para>The default for this option depends on the configuration most
convenient communication method available, but often other
alternatives work as well; you may find it useful to try alternative
connect methods if you suspect a communication problem with your
testing target.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--baud [number]</option></term>
<listitem><para>Set the default baud rate to something other than
9600. (Some serial interface programs, like <emphasis>tip</emphasis>,
use a separate initialization file instead of this
value.)</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--target_board [name(s)]</option></term>
<listitem><para>The list of target boards to run tests
on.</para></listitem>
</varlistentry>
<varlistentry id="tool-opt">
<term><option>--tool[name(s)]</option></term>
<listitem><para>Specifies which testsuite to run, and what
initialization module to use. <option>--tool</option> is used
<emphasis>only</emphasis> for these two purposes. It is
<emphasis>not</emphasis> used to name the executable program to
test. Executable tool names (and paths) are recorded in
<filename>site.exp</filename> and you can override them by specifying
Tcl variables on the command line.</para>
<para>For example, including "<option>--tool</option> gcc" on the
<emphasis>runtest</emphasis> command line runs tests from all test
subdirectories whose names match <filename>gcc.*</filename>, and uses
one of the initialization modules named
<filename>config/*-gcc.exp</filename>. To specify the name of the
compiler (perhaps as an alternative path to what
<emphasis>runtest</emphasis> would use by default), use
<emphasis>GCC=binname</emphasis> on the <emphasis>runtest</emphasis>
command line.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--tool_exec [name]</option></term>
<listitem><para>The path to the tool executable to
test.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--tool_opts [options]</option></term>
<listitem><para>A list of additional options to pass to the
tool.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--verbose</option> (-v)</term>
<listitem><para>Turns on more output. Repeating this option increases
the amount of output displayed. Level one (<emphasis>-v</emphasis>)
is simply test output. Level two (<emphasis>-v</emphasis>-v}) shows
messages on options, configuration, and process control. Verbose
messages appear in the detailed (<filename>*.log</filename>) log
file, but not in the summary (<filename>*.sum</filename>) log
file.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--version</option> (-V)</term>
<listitem><para>Prints out the version numbers of &dj;,
<emphasis>expect</emphasis> and Tcl, and exits without running any
tests.</para></listitem>
</varlistentry>
<varlistentry>
<term><option>--D[0-1]</option></term>
<listitem><para>Start the internal Tcl debugger. The Tcl debugger
supports breakpoints, single stepping, and other common debugging
activities. See the document "Debugger for Tcl Applications" by Don
Libes. (Distributed in PostScript form with
<emphasis>expect</emphasis> as the file
<filename>expect/tcl-debug.ps.</filename>. If you specify
<emphasis>-D1</emphasis>, the <emphasis>expect</emphasis> shell stops
at a breakpoint as soon as &dj; invokes it. If you specify
<emphasis>-D0</emphasis>, &dj; starts as usual, but you can enter
the debugger by sending an interrupt (e.g. by typing
<keycombo><keycap>C</keycap><keycap>c</keycap></keycombo>).
</para></listitem>
</varlistentry>
<varlistentry>
<term><filename>testfile</filename>.exp[=arg(s)]</term>
<listitem><para>Specify the names of testsuites to run. By default,
<emphasis>runtest</emphasis> runs all tests for the tool, but you can
restrict it to particular testsuites by giving the names of the
<emphasis>.exp expect</emphasis> scripts that control
them. <emphasis>testsuite</emphasis>.exp may not include path
information; use plain filenames.</para></listitem>
</varlistentry>
<varlistentry>
<term><filename>testfile</filename>.exp="testfile1 ..."</term>
<listitem><para>Specify a subset of tests in a suite to run. For
compiler or assembler tests, which often use a single
<emphasis>.exp</emphasis> script covering many different source
files, this option allows you to further restrict the tests by
listing particular source files to compile. Some tools even support
wildcards here. The wildcards supported depend upon the tool, but
typically they are <emphasis>?</emphasis>, <emphasis>*</emphasis>,
and <emphasis>[chars]</emphasis>.</para></listitem>
</varlistentry>
<varlistentry>
<term><symbol>tclvar</symbol>=value</term>
<listitem><para>You can define Tcl variables for use by your test
scripts in the same style used with <emphasis>make</emphasis> for
environment variables. For example, <emphasis>runtest
GDB=gdb.old</emphasis> defines a variable called
<command>GDB</command>; when your scripts refer to
<symbol>$GDB</symbol> in this run, they use the value
<emphasis>gdb.old</emphasis>.</para>
<para>The default Tcl variables used for most tools are defined in
the main &dj; <emphasis>Makefile</emphasis>; their values are
captured in the <filename>site.exp</filename> file.</para></listitem>
</varlistentry>
</variablelist>
</sect3>
<sect3 id="common" xreflabel="Common Operations">
<title>Common Options</title>
<para>Typically, you don't need must to use any command-line options.
<option>--tool</option> used is only required when there are more than
one testsuite in the same directory. The default options are in the
local site.exp file, created by "make site.exp".</para>
<para>For example, if the directory <filename>gdb/testsuite</filename>
contains a collection of &dj; tests for GDB, you can run them like
this:</para>
<screen>
eg$ cd gdb/testsuite
eg$ runtest --tool gdb
</screen>
<para>Test output follows, ending with:</para>
<screen>
=== gdb Summary ===
# of expected passes 508
# of expected failures 103
/usr/latest/bin/gdb version 4.14.4 -nx
</screen>
<para>You can use the option <emphasis>--srcdir</emphasis> to point to
some other directory containing a collection of tests:</para>
<screen>
eg$ runtest--srcdir /devo/gdb/testsuite
</screen>
<para>By default, <command>runtest</command> prints only the
names of the tests it runs, output from any tests that have unexpected
results, and a summary showing how many tests passed and how many
failed. To display output from all tests (whether or not they behave
as expected), use the <emphasis>--all</emphasis> option. For more
verbose output about processes being run, communication, and so on, use
<emphasis>--verbose</emphasis>. To see even more output, use multiple
<emphasis>--verbose</emphasis> options. for a more detailed explanation
of each <command>runtest</command> option.</para>
<para>Test output goes into two files in your current directory:
summary output in <filename>tool.sum</filename>,
and detailed output in <filename>
tool.log</filename>. (<emphasis>tool</emphasis>
refers to the collection of tests; for example, after a run with
<emphasis>--tool</emphasis> gdb, look for output files
<filename>gdb.sum</filename> and
<filename>gdb.log</filename>.)</para>
</sect3>
</sect2>
<sect2 id="outputfiles" xreflabel="Output Files">
<title>The files &dj; produces.</title>
<para>&dj; always writes two kinds of output files: summary
logs and detailed logs. The contents of both of these are
determined by your tests.</para>
<para>For troubleshooting, a third kind of output file is useful:
use <option>--debug</option> to request an output file showing
details of what <productname>Expect</productname> is doing
internally.</para>
<sect3 id="sum" xreflabel="Summary File">
<title>Summary File</title>
<para>&dj; always produces a summary output file
<filename>tool.sum</filename>. This summary shows the names of
all test files run; for each test file, one line of output from
each <command>pass</command> command (showing status
<emphasis>PASS</emphasis> or <emphasis>XPASS</emphasis>) or
<command>fail</command> command (status
<emphasis>FAIL</emphasis> or <emphasis>XFAIL</emphasis>);
trailing summary statistics that count passing and failing tests
(expected and unexpected); and the full pathname and version
number of the tool tested. (All possible outcomes, and all
errors, are always reflected in the summary output file,
regardless of whether or not you specify
<option>--all</option>.)</para>
<para>If any of your tests use the procedures
<command>unresolved</command>, <command>unsupported</command>,
or <command>runtested</command>, the summary output also
tabulates the corresponding outcomes.</para>
<para>For example, after <command>runtest --tool
binutils</command>, look for a summary log in
<filename>binutils.sum</filename>. Normally, &dj; writes this
file in your current working directory; use the
<option>--outdir</option> option to select a different
directory.</para>
<example>
<title>Here is a short sample summary log</title>
<screen>
Test Run By rob on Mon May 25 21:40:57 PDT 1992
=== gdb tests ===
Running ./gdb.t00/echo.exp ...
PASS: Echo test
Running ./gdb.all/help.exp ...
PASS: help add-symbol-file
PASS: help aliases
PASS: help breakpoint "bre" abbreviation
FAIL: help run "r" abbreviation
Running ./gdb.t10/crossload.exp ...
PASS: m68k-elf (elf-big) explicit format; loaded
XFAIL: mips-ecoff (ecoff-bigmips) "ptype v_signed_char" signed C types
=== gdb Summary ===
# of expected passes 5
# of expected failures 1
# of unexpected failures 1
/usr/latest/bin/gdb version 4.6.5 -q
</screen>
</example>
</sect3>
<sect3 id="log" xreflabel="Log File">
<title>Log File</title>
<para>&dj; also saves a detailed log file
<filename>tool.log</filename>, showing any output generated by
tests as well as the summary output. For example, after
<command>runtest --tool binutils</command>, look for a detailed
log in <filename>binutils.log</filename>. Normally, &dj;
writes this file in your current working directory; use the
<option>--outdir</option> option to select a different
directory.</para>
<example>
<title>Here is a brief example showing a detailed log for
<productname>G++</productname> tests</title>
<screen>
Test Run By rob on Mon May 25 21:40:43 PDT 1992
=== g++ tests ===
--- Running ./g++.other/t01-1.exp ---
PASS: operate delete
--- Running ./g++.other/t01-2.exp ---
FAIL: i960 bug EOF
p0000646.C: In function `int warn_return_1 ()':
p0000646.C:109: warning: control reaches end of non-void function
p0000646.C: In function `int warn_return_arg (int)':
p0000646.C:117: warning: control reaches end of non-void function
p0000646.C: In function `int warn_return_sum (int, int)':
p0000646.C:125: warning: control reaches end of non-void function
p0000646.C: In function `struct foo warn_return_foo ()':
p0000646.C:132: warning: control reaches end of non-void function
--- Running ./g++.other/t01-4.exp ---
FAIL: abort
900403_04.C:8: zero width for bit-field `foo'
--- Running ./g++.other/t01-3.exp ---
FAIL: segment violation
900519_12.C:9: parse error before `;'
900519_12.C:12: Segmentation violation
/usr/latest/bin/gcc: Internal compiler error: program cc1plus got fatal signal
=== g++ Summary ===
# of expected passes 1
# of expected failures 3
/usr/latest/bin/g++ version cygnus-2.0.1
</screen>
</example>
</sect3>
<sect3 id="debugfile" xreflabel="Debug Log File">
<title>Debug Log File</title>
<para>With the <option>--debug</option> option, you can request
a log file showing the output from
<productname>Expect</productname> itself, running in debugging
mode. This file (<filename>dbg.log</filename>, in the directory
where you start <command>runtest</command>) shows each pattern
<productname>Expect</productname> considers in analyzing test
output.</para>
<para>This file reflects each <command>send</command> command,
showing the string sent as input to the tool under test; and
each <productname>Expect</productname> command, showing each
pattern it compares with the tool output.</para>
<example>
<title>The log messages begin with a message of the form</title>
<screen>
expect: does {<symbol>tool output</symbol>} (spawn_id <symbol>n</symbol>)
match pattern {<emphasis>expected pattern</emphasis>}?
</screen>
</example>
<para>For every unsuccessful match,
<productname>Expect</productname> issues a
<emphasis>no</emphasis> after this message; if other patterns
are specified for the same <productname>Expect</productname>
command, they are reflected also, but without the first part of
the message (<emphasis>expect... match pattern</emphasis>).</para>
<para>When <productname>Expect</productname> finds a match, the
log for the successful match ends with <emphasis>yes</emphasis>,
followed by a record of the <productname>Expect</productname>
variables set to describe a successful match.</para>
<example>
<title>Here is an excerpt from the debugging log for a
<productname>GDB</productname> test:</title>
<screen>
send: sent {break gdbme.c:34\n} to spawn id 6
expect: does {} (spawn_id 6) match pattern {Breakpoint.*at.* file
gdbme.c, line 34.*\(gdb\) $}? no
{.*\(gdb\) $}? no
expect: does {} (spawn_id 0) match pattern {return} ? no
{\(y or n\) }? no
{buffer_full}? no
{virtual}? no
{memory}? no
{exhausted}? no
{Undefined}? no
{command}? no
break gdbme.c:34
Breakpoint 8 at 0x23d8: file gdbme.c, line 34.
(gdb) expect: does {break gdbme.c:34\r\nBreakpoint 8 at 0x23d8:
file gdbme.c, line 34.\r\n(gdb) } (spawn_id 6) match pattern
{Breakpoint.*at.* file gdbme.c, line 34.*\(gdb\) $}? yes
expect: set expect_out(0,start) {18}
expect: set expect_out(0,end) {71}
expect: set expect_out(0,string) {Breakpoint 8 at 0x23d8: file
gdbme.c, line 34.\r\n(gdb) }
epect: set expect_out(spawn_id) {6}
expect: set expect_out(buffer) {break gdbme.c:34\r\nBreakpoint 8
at 0x23d8: file gdbme.c, line 34.\r\n(gdb) }
PASS: 70 0 breakpoint line number in file
</screen>
</example>
<para>This example exhibits three properties of
<productname>Expect</productname> and
<productname>&dj;</productname> that might be surprising at
first glance:</para>
<itemizedlist mark="bullet">
<listitem><para>Empty output for the first attempted match. The
first set of attempted matches shown ran against the output
<emphasis>{}</emphasis> --- that is, no
output. <productname>Expect</productname> begins
attempting to match the patterns supplied immediately; often,
the first pass is against incomplete output (or completely
before all output, as in this case).</para></listitem>
<listitem><para>Interspersed tool output. The beginning of
the log entry for the second attempted match may be hard to
spot: this is because the prompt <emphasis>{(gdb) }</emphasis>
appears on the same line, just before the
<emphasis>expect:</emphasis> that marks the beginning of the
log entry.</para></listitem>
<listitem><para>Fail-safe patterns. Many of the patterns
tested are fail-safe patterns provided by
<productname>GDB</productname> testing utilities, to reduce
possible indeterminacy. It is useful to anticipate potential
variations caused by extreme system conditions
(<productname>GDB</productname> might issue the message
<emphasis>virtual memory exhausted</emphasis> in rare
circumstances), or by changes in the tested program
(<emphasis>Undefined command</emphasis> is the likeliest
outcome if the name of a tested command changes).</para>
<para>The pattern <emphasis>{return}</emphasis> is a
particularly interesting fail-safe to notice; it checks for an
unexpected <keycap>RET</keycap> prompt. This may happen,
for example, if the tested tool can filter output through a
pager.</para>
<para>These fail-safe patterns (like the debugging log itself)
are primarily useful while developing test scripts. Use the
<command>error</command> procedure to make the actions for
fail-safe patterns produce messages starting with
<emphasis>ERROR</emphasis> on standard output, and in the
detailed log file.</para></listitem>
</itemizedlist>
</sect3>
</sect2>
</sect1>
<sect1 id="Customizing" xreflabel="Customizing DejaGnu">
<title>Customizing &dj;</title>
<para>The site configuration file, <filename>site.exp</filename>,
captures configuration-dependent values and propagates them to the
&dj; test environment using Tcl variables. This ties the
&dj; test scripts into the <command>configure</command> and
<command>make</command> programs. If this file is setup correctly,
it is possible to execute a testsuite merely by typing
<command>runtest</command>.</para>
<para>&dj; supports two <filename>site.exp</filename>
files. The multiple instances of <filename>site.exp</filename> are
loaded in a fixed order built into &dj;. The first file loaded
is the local file <filename>site.exp</filename>, and then the
optional global <filename>site.exp</filename> file as
pointed to by the <symbol>DEJAGNU</symbol> environment
variable.</para>
<para>There is an optional <emphasis>master</emphasis>
<filename>site.exp</filename>, capturing configuration values that
apply to &dj; across the board, in each configuration-specific
subdirectory of the &dj; library directory.
<command>runtest</command> loads these values first. The master
<filename>site.exp</filename> contains the default values for all
targets and hosts supported by &dj;. This master file is
identified by setting the environment variable
<symbol>DEJAGNU</symbol> to the name of the file. This is also
refered to as the ``global'' config file.</para>
<para>Any directory containing a configured testsuite also has a
local <filename>site.exp</filename>, capturing configuration values
specific to the tool under test. Since <command>runtest</command>
loads these values last, the individual test configuration can
either rely on and use, or override, any of the global values from
the global <filename>site.exp</filename> file.</para>
<para>You can usually generate or update the testsuite's local
<filename>site.exp</filename> by typing <command>make
site.exp</command> in the testsuite directory, after the test
suite is configured.</para>
<para>You can also have a file in your home directory called
<filename>.dejagnurc</filename>. This gets loaded first before the
other config files. Usually this is used for personal stuff, like
setting the <symbol>all_flag</symbol> so all the output gets
printed, or your own verbosity levels. This file is usually
restricted to setting command line options.</para>
<para>You can further override the default values in a
user-editable section of any <filename>site.exp</filename>, or by
setting variables on the <command>runtest</command> command
line.</para>
<sect2 id="local" xreflabel="Local Config File">
<title>Local Config File</title>
<para>It is usually more convenient to keep these <emphasis>manual
overrides</emphasis> in the <filename>site.exp</filename>
local to each test directory, rather than in the global
<filename>site.exp</filename> in the installed &dj;
library. This file is mostly for supplying tool specific info
that is required by the testsuite.</para>
<para>All local <filename>site.exp</filename> files have
two sections, separated by comment text. The first section is
the part that is generated by <command>make</command>. It is
essentially a collection of Tcl variable definitions based on
<filename>Makefile</filename> environment variables. Since they
are generated by <command>make</command>, they contain the
values as specified by <command>configure</command>. (You can
also customize these values by using the <option>--site</option>
option to <command>configure</command>.) In particular, this
section contains the <filename>Makefile</filename>
variables for host and target configuration data. Do not edit
this first section; if you do, your changes are replaced next
time you run <command>make</command>.</para>
<example>
<title>The first section starts with</title>
<programlisting>
## these variables are automatically generated by make ##
# Do not edit here. If you wish to override these values
# add them to the last section
</programlisting>
</example>
<para>In the second section, you can override any default values
(locally to &dj;) for all the variables. The second section
can also contain your preferred defaults for all the command
line options to <command>runtest</command>. This allows you to
easily customize <command>runtest</command> for your preferences
in each configured test-suite tree, so that you need not type
options repeatedly on the command line. (The second section may
also be empty, if you do not wish to override any defaults.)</para>
<example>
<title>The first section ends with this line</title>
<programlisting>
## All variables above are generated by configure. Do Not Edit ##
</programlisting>
</example>
<para>You can make any changes under this line. If you wish to
redefine a variable in the top section, then just put a
duplicate value in this second section. Usually the values
defined in this config file are related to the configuration of
the test run. This is the ideal place to set the variables
<symbol>host_triplet</symbol>, <symbol>build_triplet</symbol>,
<symbol>target_triplet</symbol>. All other variables are tool
dependant, i.e., for testing a compiler, the value for
<symbol>CC</symbol> might be set to a freshly built binary, as
opposed to one in the user's path.</para>
<para>Here's an example local site.exp file, as used for
<productname>GCC/G++</productname> testing.</para>
<example>
<title>Local Config File</title>
<programlisting>
## these variables are automatically generated by make ##
# Do not edit here. If you wish to override these values
# add them to the last section
set rootme "/build/devo-builds/i586-pc-linux-gnulibc1/gcc"
set host_triplet i586-pc-linux-gnulibc1
set build_triplet i586-pc-linux-gnulibc1
set target_triplet i586-pc-linux-gnulibc1
set target_alias i586-pc-linux-gnulibc1
set CFLAGS ""
set CXXFLAGS "-isystem /build/devo-builds/i586-pc-linux-gnulibc1/gcc/../libio -isystem $srcdir/../libg++/src -isystem $srcdir/../libio -isystem $srcdir/../libstdc++ -isystem $srcdir/../libstdc++/stl -L/build/devo-builds/i586-pc-linux-gnulibc1/gcc/../libg++ -L/build/devo-builds/i586-pc-linux-gnulibc1/gcc/../libstdc++"
append LDFLAGS " -L/build/devo-builds/i586-pc-linux-gnulibc1/gcc/../ld"
set tmpdir /build/devo-builds/i586-pc-linux-gnulibc1/gcc/testsuite
set srcdir "${srcdir}/testsuite"
## All variables above are generated by configure. Do Not Edit ##
</programlisting>
</example>
<para>This file defines the required fields for a local config
file, namely the three config triplets, and the srcdir. It also
defines several other Tcl variables that are used exclusivly by
the GCC testsuite. For most test cases, the CXXFLAGS and LDFLAGS
are supplied by &dj; itself for cross testing, but to test a
compiler, GCC needs to manipulate these itself.</para>
</sect2>
<sect2 id="global" xreflabel="Global Config File">
<title>Global Config File</title>
<para>The master config file is where all the target specific
config variables for a whole site get set. The idea is
that for a centralized testing lab where people have to share a
target between multiple developers. There are settings for both
remote targets and remote hosts. Here's an example of a Master
Config File (also called the Global config file) for a
<emphasis>canadian cross</emphasis>. A canadian cross is when
you build and test a cross compiler on a machine other than the
one it's to be hosted on.</para>
<para>Here we have the config settings for our California
office. Note that all config values are site dependant. Here we
have two sets of values that we use for testing m68k-aout cross
compilers. As both of these target boards has a different
debugging protocol, we test on both of them in sequence.</para>
<example>
<title>Global Config file</title>
<programlisting>
# Make sure we look in the right place for the board description files.
if ![info exists boards_dir] {
set boards_dir {}
}
lappend boards_dir "/nfs/cygint/s1/cygnus/dejagnu/boards"
verbose "Global Config File: target_triplet is $target_triplet" 2
global target_list
case "$target_triplet" in {
{ "native" } {
set target_list "unix"
}
{ "sparc64-*elf" } {
set target_list "sparc64-sim"
}
{ "mips-*elf" } {
set target_list "mips-sim wilma barney"
}
{ "mips-lsi-elf" } {
set target_list "mips-lsi-sim{,soft-float,el}"
}
{ "sh-*hms" } {
set target_list { "sh-hms-sim" "bloozy" }
}
}
</programlisting>
</example>
<para>In this case, we have support for several cross compilers,
that all run on this host. For testing on operating systems that
don't support Expect, &dj; can be run on the local build
machine, and it can connect to the remote host and run all the
tests for this cross compiler on that host. All the remote OS
requires is a working telnetd.</para>
<para>As you can see, all one does is set the variable
<symbol>target_list</symbol> to the list of targets and options to
test. The simple settings, like for
<emphasis>sparc64-elf</emphasis> only require setting the name of
the single board config file. The <emphasis>mips-elf</emphasis>
target is more complicated. Here it sets the list to three target
boards. One is the default mips target, and both
<emphasis>wilma</emphasis> <emphasis>barney</emphasis> are
symbolic names for other mips boards. Symbolic names are covered
in the <xref linkend="addboard"/> chapter. The more complicated
example is the one for <emphasis>mips-lsi-elf</emphasis>. This one
runs the tests with multiple iterations using all possible
combinations of the <option>--soft-float</option> and the
<option>--el</option> (little endian) option. Needless to say,
this last feature is mostly compiler specific.</para>
</sect2>
<sect2 id="boardconfig" xreflabel="Board Config File">
<title>Board Config File</title>
<para>The board config file is where board specfic config data
is stored. A board config file contains all the higher-level
configuration settings. There is a rough inheritance scheme, where it is
possible to base a new board description file on an existing one. There
are also collections of custom procedures for common environments. For
more information on adding a new board config file, go to the <xref
linkend="addboard"/> chapter. </para>
<para>An example board config file for a GNU simulator is as
follows. <function>set_board_info</function> is a procedure that sets the
field name to the specified value. The procedures in square brackets
<emphasis>[]</emphasis> are <emphasis>helper procedures</emphasis>. Thes
are used to find parts of a tool chain required to build an executable
image that may reside in various locations. This is mostly of use for
when the startup code, the standard C lobraries, or the tool chain itself
is part of your build tree.</para>
<example>
<title>Board Config File</title>
<programlisting>
# This is a list of toolchains that are supported on this board.
set_board_info target_install {sparc64-elf}
# Load the generic configuration for this board. This will define any
# routines needed by the tool to communicate with the board.
load_generic_config "sim"
# We need this for find_gcc and *_include_flags/*_link_flags.
load_base_board_description "basic-sim"
# Use long64 by default.
process_multilib_options "long64"
setup_sim sparc64
# We only support newlib on this target. We assume that all multilib
# options have been specified before we get here.
set_board_info compiler "[find_gcc]"
set_board_info cflags "[libgloss_include_flags] [newlib_include_flags]"
set_board_info ldflags "[libgloss_link_flags] [newlib_link_flags]"
# No linker script.
set_board_info ldscript "";
# Used by a few gcc.c-torture testcases to delimit how large the
# stack can be.
set_board_info gcc,stack_size 16384
# The simulator doesn't return exit statuses and we need to indicate this
# the standard GCC wrapper will work with this target.
set_board_info needs_status_wrapper 1
# We can't pass arguments to programs.
set_board_info noargs 1
</programlisting>
</example>
<para>There are five helper procedures used in this example. The first
one, <function>find gcc</function> looks for a copy of the GNU compiler in
your build tree, or it uses the one in your path. This will also return
the proper transformed name for a cross compiler if you whole build tree
is configured for one. The next helper procedures are
<function>libgloss_include_flags</function> &amp;
<function>libgloss_link_flags</function>. These return the proper flags to
compiler and link an executable image using <xref
linkend="libgloss"/>, the GNU BSP (Board Support Package). The final
procedures are <function>newlib_include_flag</function> &amp;
<function>newlib_include_flag</function>. These find the Newlib C
library, which is a reentrant standard C library for embedded systems
comprising of non GPL'd code.</para>
</sect2>
<sect2 id="releng" xreflabel="Remote Host Testing">
<title>Remote Host Testing</title>
<note><para>Thanks to Dj Delorie for the original paper that
this section is based on.</para></note>
<para>&dj; also supports running the tests on a remote
host. To set this up, the remote host needs an ftp server, and a
telnet server. Currently foreign operating systems used as
remote hosts are VxWorks, VRTX, DOS/Windows 3.1, MacOS and Windows.</para>
<para>The recommended source for a Windows-based FTP
server is to get IIS (either IIS 1 or Personal Web Server) from
<ulink
url="http://www.microsoft.com">http://www.microsoft.com</ulink>.
When you install it, make sure you install the FTP server - it's
not selected by default. Go into the IIS manager and change the
FTP server so that it does not allow anonymous FTP. Set the home
directory to the root directory (i.e. c:\) of a suitable
drive. Allow writing via FTP.</para>
<para>It will create an account like IUSR_FOOBAR where foobar is
the name of your machine. Go into the user editor and give that
account a password that you don't mind hanging around in the
clear (i.e. not the same as your admin or personal
passwords). Also, add it to all the various permission groups.</para>
<para>You'll also need a telnet server. For Windows, go
to the <ulink url="http://ataman.com">Ataman</ulink> web site,
pick up the Ataman Remote Logon Services for Windows, and
install it. You can get started on the eval period anyway. Add
IUSR_FOOBAR to the list of allowed users, set the HOME directory
to be the same as the FTP default directory. Change the Mode
prompt to simple.</para>
<para>Ok, now you need to pick a directory name to do all the
testing in. For the sake of this example, we'll call it piggy
(i.e. c:\piggy). Create this directory.</para>
<para>You'll need a unix machine. Create a directory for the
scripts you'll need. For this example, we'll use
/usr/local/swamp/testing. You'll need to have a source tree
somewhere, say /usr/src/devo. Now, copy some files from
releng's area in SV to your machine:</para>
<example>
<title>Remote host setup</title>
<screen>
cd /usr/local/swamp/testing
mkdir boards
scp darkstar.welcomehome.org:/dejagnu/cst/bin/MkTestDir .
scp darkstar.welcomehome.org:/dejagnu/site.exp .
scp darkstar.welcomehome.org:/dejagnu/boards/useless98r2.exp boards/foobar.exp
export DEJAGNU=/usr/local/swamp/testing/site.exp
</screen>
</example>
<para>You must edit the boards/foobar.exp file to reflect your
machine; change the hostname (foobar.com), username
(iusr_foobar), password, and ftp_directory (c:/piggy) to match
what you selected.</para>
<para>Edit the global <filename> site.exp</filename> to reflect your
boards directory:</para>
<example>
<title>Add The Board Directory</title>
<programlisting>
lappend boards_dir "/usr/local/swamp/testing/boards"
</programlisting>
</example>
<para>Now run MkTestDir, which is in the contrib
directory. The first parameter is the toolchain prefix, the
second is the location of your devo tree. If you are testing a
cross compiler (ex: you have sh-hms-gcc.exe in your PATH on
the PC), do something like this:</para>
<example>
<title>Setup Cross Remote Testing</title>
<programlisting>
./MkTestDir sh-hms /usr/dejagnu/src/devo
</programlisting>
</example>
<para>If you are testing a native PC compiler (ex: you have
gcc.exe in your PATH on the PC), do this:</para>
<example>
<title>Setup Native Remote Testing</title>
<programlisting>
./MkTestDir '' /usr/dejagnu/src/devo
</programlisting>
</example>
<para>To test the setup, <command>ftp</command> to your PC
using the username (iusr_foobar) and password you selected. CD
to the test directory. Upload a file to the PC. Now telnet to
your PC using the same username and password. CD to the test
directory. Make sure the file is there. Type "set" and/or "gcc
-v" (or sh-hms-gcc -v) and make sure the default PATH contains
the installation you want to test.</para>
<example>
<title>Run Test Remotely</title>
<programlisting>
cd /usr/local/swamp/testing
make -k -w check RUNTESTFLAGS="--host_board foobar --target_board foobar -v -v" > check.out 2>&amp;1
</programlisting>
</example>
<para>To run a specific test, use a command like this (for
this example, you'd run this from the gcc directory that
MkTestDir created):</para>
<example>
<title>Run a Test Remotely</title>
<programlisting>
make check RUNTESTFLAGS="--host_board sloth --target_board sloth -v compile.exp=921202-1.c"
</programlisting>
</example>
<para>Note: if you are testing a cross-compiler, put in the
correct target board. You'll also have to download more .exp
files and modify them for your local configuration. The -v's
are optional.</para>
</sect2>
<sect2 id="configfile" xreflabel="Config File Values">
<title>Config File Values</title>
<para>&dj; uses a named array in Tcl to hold all the info for
each machine. In the case of a canadian cross, this means host
information as well as target information. The named array is
called <symbol>target_info</symbol>, and it has two indices. The
following fields are part of the array.</para>
<sect3 id="optiondefs" xreflabel="Option Variables">
<title>Command Line Option Variables</title>
<para>In the user editable second section of the <xref
linkend="personal"/> you can not only override the configuration
variables captured in the first section, but also specify
default values for all on the <command>runtest</command>
command line options. Save for <option>--debug</option>,
<option>--help</option>, and <option>--version</option>, each
command line option has an associated Tcl variable. Use the
Tcl <command>set</command> command to specify a new default
value (as for the configuration variables). The following
table describes the correspondence between command line
options and variables you can set in
<filename>site.exp</filename>. <xref linkend="invoking"/>, for
explanations of the command-line options.</para>
<para><table frame="all" rowsep="0" colsep="0">
<title>Tcl Variables For Command Line Options</title>
<tgroup cols="3" align="char" rowsep="1" colsep="0">
<thead><row>
<entry>runtest</entry><entry>Tcl</entry>
<entry>option</entry><entry>variable</entry><entry>description</entry>
</row></thead>
<tbody>
<row>
<entry>--all</entry>
<entry>all_flag</entry>
<entry>display all test results if set</entry>
</row>
<row>
<entry>--baud</entry>
<entry>baud</entry>
<entry>set the default baud rate to something other than
9600.</entry>
</row>
<row>
<entry>--connect</entry>
<entry>connectmode</entry>
<entry><command>rlogin</command>,
<command>telnet</command>, <command>rsh</command>,
<command>kermit</command>, <command>tip</command>, or
<command>mondfe</command></entry>
</row>
<row>
<entry>--outdir</entry>
<entry>outdir</entry>
<entry>directory for <filename>tool.sum</filename> and
<filename>tool.log.</filename></entry>
</row>
<row>
<entry>--objdir</entry>
<entry>objdir</entry>
<entry>directory for pre-compiled binaries</entry>
</row>
<row>
<entry>--reboot</entry>
<entry>reboot</entry>
<entry>reboot the target if set to
<emphasis>"1"</emphasis>; do not reboot if set to
<emphasis>"0"</emphasis> (the default).</entry>
</row>
<row>
<entry>--srcdir</entry>
<entry>srcdir</entry>
<entry>directory of test subdirectories</entry>
</row>
<row>
<entry>--strace</entry>
<entry>tracelevel</entry>
<entry>a number: Tcl trace depth</entry>
</row>
<row>
<entry>--tool</entry>
<entry>tool</entry>
<entry>name of tool to test; identifies init, test subdir</entry>
</row>
<row>
<entry>--verbose</entry>
<entry>verbose</entry>
<entry>verbosity level. As option, use multiple times; as
variable, set a number, 0 or greater.</entry>
</row>
<row>
<entry>--target</entry>
<entry>target_triplet</entry>
<entry>The canonical configuration string for the target.</entry>
</row>
<row>
<entry>--host</entry>
<entry>host_triplet</entry>
<entry>The canonical configuration string for the host.</entry>
</row>
<row>
<entry>--build</entry>
<entry>build_triplet</entry>
<entry>The canonical configuration string for the build
host.</entry>
</row>
<row>
<entry>--mail</entry>
<entry>address</entry>
<entry>Email the output log to the specified address.</entry>
</row>
</tbody>
</tgroup>
</table>
</para>
</sect3>
<sect3 id="personal" xreflabel="Personal Config File">
<title>Personal Config File</title>
<para>The personal config file is used to customize
<command>runtest's</command> behaviour for each person. It is
typically used to set the user prefered setting for verbosity,
and any experimental Tcl procedures. My personal
<filename>~/.dejagnurc</filename> file looks like:</para>
<example>
<title>Personal Config File</title>
<programlisting>
set all_flag 1
set RLOGIN /usr/ucb/rlogin
set RSH /usr/local/sbin/ssh
</programlisting>
</example>
<para>Here I set <symbol>all_flag</symbol> so I see all the test
cases that PASS along with the ones that FAIL. I also set
<symbol>RLOGIN</symbol> to the BSD version. I have
<productname>Kerberos</productname> installed, and when I rlogin
to a target board, it usually isn't supported. So I use the non
secure version rather than the default that's in my path. I also
set <symbol>RSH</symbol> to the <productname>SSH</productname>
secure shell, as rsh is mostly used to test unix
machines within a local network here.</para>
</sect3>
</sect2>
</sect1>
<sect1 id="Extending" xreflabel="Extending DejaGnu">
<title>Extending &dj;</title>
<sect2 id="addsuite" xreflabel="Adding a new Testsuite">
<title>Adding A New Testsuite</title>
<para>The testsuite for a new tool should always be located in that tools
source directory. &dj; require the directory be named
<filename>testsuite</filename>. Under this directory, the test cases go
in a subdirectory whose name begins with the tool name. For example, for
a tool named <emphasis>flubber</emphasis>, each subdirectory containing
testsuites must start with <emphasis>"flubber."</emphasis>.</para>
</sect2>
<sect2 id="addtool" xreflabel="Adding A New Tool">
<title>Adding A New Tool</title>
<para>In general, the best way to learn how to write (code or even prose)
is to read something similar. This principle applies to test cases and
to testsuites. Unfortunately, well-established testsuites have a way
of developing their own conventions: as test writers become more
experienced with &dj; and with Tcl, they accumulate more utilities,
and take advantage of more and more features of
<productname>Expect</productname> and <productname>Tcl</productname> in
general.</para>
<para>Inspecting such established testsuites may make the prospect of
creating an entirely new testsuite appear overwhelming. Nevertheless,
it is quite straightforward to get a new testsuite going.</para>
<para>There is one testsuite that is guaranteed not to grow more
elaborate over time: both it and the tool it tests were created expressly
to illustrate what it takes to get started with &dj;. The
<filename>example/</filename> directory of the &dj; distribution
contains both an interactive tool called <command>calc</command>, and a
testsuite for it. Reading this testsuite, and experimenting with it,
is a good way to supplement the information in this section. (Thanks to
Robert Lupton for creating calc and its testsuite---and also the first
version of this section of the manual!)</para>
<para>To help orient you further in this task, here is an outline of the
steps to begin building a testsuite for a program example.</para>
<itemizedlist mark="bullet">
<listitem><para>Create or select a directory to contain your new
collection of tests. Change into that directory (shown here as
<filename>testsuite</filename>):</para>
<para>Create a <filename>configure.in</filename> file in this directory,
to control configuration-dependent choices for your tests. So far as
&dj; is concerned, the important thing is to set a value for the
variable <symbol>target_abbrev</symbol>; this value is the link to the
init file you will write soon. (For simplicity, we assume the
environment is Unix, and use <emphasis>unix</emphasis> as the
value.)</para>
<para>What else is needed in <filename>configure.in</filename> depends on
the requirements of your tool, your intended test environments, and which
configure system you use. This example is a minimal configure.in for use
with <productname>GNU Autoconf</productname>. </para></listitem>
<listitem><para>Create <filename>Makefile.in</filename> (if you are using
Autoconf), or <filename>Makefile.am</filename>(if you are using
Automake), the source file used by configure to build your
<filename>Makefile</filename>. If you are using GNU Automake.just add the
keyword <emphasis>dejagnu</emphasis> to the
<emphasis>AUTOMAKE_OPTIONS</emphasis> variable in your
<filename>Makefile.am</filename> file. This will add all the Makefile
support needed to run &dj;, and support the <xref linkend="makecheck"/>
target.</para>
<para>You also need to include two targets important to &dj;:
<emphasis>check</emphasis>, to run the tests, and
<emphasis>site.exp</emphasis>, to set up the Tcl copies of
configuration-dependent values. This is called the <xref linkend="local"/>
The check target must run the <command>runtest</command> program to
execute the tests.</para>
<para>The <filename>site.exp</filename> target should usually set up
(among other things) the $tool variable for the name of your program. If
the local site.exp file is setup correctly, it is possible to execute the
tests by merely typing <command>runtest</command> on the command
line.</para>
<example>
<title>Sample Makefile.in Fragment</title>
<programlisting>
# Look for a local version of &dj;, otherwise use one in the path
RUNTEST = `if test -f $(top_srcdir)/../dejagnu/runtest; then \
echo $(top_srcdir) ../dejagnu/runtest; \
else \
echo runtest; \
fi`
# The flags to pass to runtest
RUNTESTFLAGS =
# Execute the tests
check: site.exp all
$(RUNTEST) $(RUNTESTFLAGS) \
--tool <symbol>${example}</symbol> --srcdir $(srcdir)
# Make the local config file
site.exp: ./config.status Makefile
@echo "Making a new config file..."
-@rm -f ./tmp?
@touch site.exp
-@mv site.exp site.bak
@echo "## these variables are automatically\
generated by make ##" > ./tmp0
@echo "# Do not edit here. If you wish to\
override these values" >> ./tmp0
@echo "# add them to the last section" >> ./tmp0
@echo "set host_os ${host_os}" >> ./tmp0
@echo "set host_alias ${host_alias}" >> ./tmp0
@echo "set host_cpu ${host_cpu}" >> ./tmp0
@echo "set host_vendor ${host_vendor}" >> ./tmp0
@echo "set target_os ${target_os}" >> ./tmp0
@echo "set target_alias ${target_alias}" >> ./tmp0
@echo "set target_cpu ${target_cpu}" >> ./tmp0
@echo "set target_vendor ${target_vendor}" >> ./tmp0
@echo "set host_triplet ${host_canonical}" >> ./tmp0
@echo "set target_triplet ${target_canonical}">>./tmp0
@echo "set tool binutils" >> ./tmp0
@echo "set srcdir ${srcdir}" >> ./tmp0
@echo "set objdir `pwd`" >> ./tmp0
@echo "set <symbol>${examplename}</symbol> <symbol>${example}</symbol>" >> ./tmp0
@echo "## All variables above are generated by\
configure. Do Not Edit ##" >> ./tmp0
@cat ./tmp0 > site.exp
@sed &lt; site.bak \
-e '1,/^## All variables above are.*##/ d' \
>> site.exp
-@rm -f ./tmp?
</programlisting>
</example>
</listitem>
<listitem><para>Create a directory (in <filename>testsuite</filename>)
called <filename>config</filename>. Make a <emphasis>Tool Init
File</emphasis> in this directory. Its name must start with the
<symbol>target_abbrev</symbol> value, or be named
<filename>default.exp</filename> so call it
<filename>config/unix.exp</filename> for our Unix based example. This
is the file that contains the target-dependent procedures.
Fortunately, on Unix, most of them do not have to do very much in
order for <command>runtest</command> to run.</para>
<para>If the program being tested is not interactive, you can get
away with this minimal <filename>unix.exp</filename> to begin
with:</para>
<example>
<title>Simple Batch Program Tool Init File</title>
<programlisting>
proc foo_exit {} {}
proc foo_version {} {}
</programlisting>
</example>
<para>If the program being tested is interactive, however, you might
as well define a <emphasis>start</emphasis> routine and invoke it by
using an init file like this:</para>
<example>
<title>Simple Interactive Program Tool Init File</title>
<programlisting>
proc foo_exit {} {}
proc foo_version {} {}
proc foo_start {} {
global ${examplename}
spawn ${examplename}
expect {
-re "" {}
}
}
# Start the program running we want to test
foo_start
</programlisting>
</example>
</listitem>
<listitem><para>Create a directory whose name begins with your tool's
name, to contain tests. For example, if your tool's name is
<emphasis>gcc</emphasis>, then the directories all need to start with
<emphasis>"gcc."</emphasis>.</para></listitem>
<listitem><para>Create a sample test file. Its name must end with
<filename>.exp</filename>. You can use
<filename>first-try.exp</filename>. To begin with, just write there a
line of Tcl code to issue a message.</para>
<example>
<title>Testing A New Tool Config</title>
<programlisting>
send_user "Testing: one, two...\n"
</programlisting>
</example>
</listitem>
<listitem><para>Back in the <filename>testsuite</filename> (top
level) directory, run <command>configure</command>. Typically you do
this while in the build directory. You may have to specify more of a
path, if a suitable configure is not available in your execution
path.</para></listitem>
<listitem><para>e now ready to triumphantly type <command>make
check</command> or <command>runtest</command>. You should see
something like this:</para>
<example>
<title>Example Test Case Run</title>
<screen>
Test Run By rhl on Fri Jan 29 16:25:44 EST 1993
=== example tests ===
Running ./example.0/first-try.exp ...
Testing: one, two...
=== example Summary ===
</screen>
</example>
<para>There is no output in the summary, because so far the example
does not call any of the procedures that establish a test
outcome.</para></listitem>
<listitem><para>Write some real tests. For an interactive tool, you
should probably write a real exit routine in fairly short order. In
any case, you should also write a real version routine
soon. </para></listitem>
</itemizedlist>
</sect2>
<sect2 id="addtarget" xreflabel="Adding A New Target">
<title>Adding A New Target</title>
<para>&dj; has some additional requirements for target support, beyond
the general-purpose provisions of configure. &dj; must actively
communicate with the target, rather than simply generating or managing
code for the target architecture. Therefore, each tool requires an
initialization module for each target. For new targets, you must supply
a few Tcl procedures to adapt &dj; to the target. This permits
&dj; itself to remain target independent.</para>
<para>Usually the best way to write a new initialization module is to
edit an existing initialization module; some trial and error will be
required. If necessary, you can use the @samp{--debug} option to see what
is really going on.</para>
<para>When you code an initialization module, be generous in printing
information controlled by the <function>verbose</function>
procedure.</para>
<para>For cross targets, most of the work is in getting the
communications right. Communications code (for several situations
involving IP networks or serial lines) is available in a &dj; library
file.</para>
<para>If you suspect a communication problem, try running the connection
interactively from <productname>Expect</productname>. (There are three
ways of running <productname>Expect</productname> as an interactive
interpreter. You can run <productname>Expect</productname> with no
arguments, and control it completely interactively; or you can use
<command>expect -i</command> together with other command-line options and
arguments; or you can run the command <command>interpreter</command> from
any <productname>Expect</productname> procedure. Use
<command>return</command> to get back to the calling procedure (if any),
or <command>return -tcl</command> to make the calling procedure itself
return to its caller; use <command>exi</command>t or end-of-file to leave
Expect altogether.) Run the program whose name is recorded in
<symbol>$connectmode</symbol>, with the arguments in
<symbol>$targetname</symbol>, to establish a connection. You should at
least be able to get a prompt from any target that is physically
connected.</para>
</sect2>
<sect2 id="addboard" xreflabel="Adding A New Board">
<title>Adding A New Board</title>
<para>Adding a new board consists of creating a new board config
file. Examples are in
<filename>dejagnu/baseboards</filename>. Usually to make a new
board file, it's easiest to copy an existing one. It is also
possible to have your file be based on a
<emphasis>baseboard</emphasis> file with only one or two
changes needed. Typically, this can be as simple as just
changing the linker script. Once the new baseboard file is done,
add it to the <symbol>boards_DATA</symbol> list in the
<filename>dejagnu/baseboards/Makefile.am</filename>, and regenerate the
Makefile.in using automake. Then just rebuild and install &dj;. You
can test it by:</para>
<para>There is a crude inheritance scheme going on with board files, so
you can include one board file into another, The two main procedures used
to do this are <function>load_generic_config</function> and
<function>load_base_board_description</function>. The generic config file
contains other procedures used for a certain class of target. The
board description file is where the board specfic settings go. Commonly
there are similar target environments with just different
processors.</para>
<example>
<title>Testing a New Board Config File</title>
<screen>
make check RUNTESTFLAGS="--target_board=<emphasis>newboardfile</emphasis>".
</screen>
</example>
<para>Here's an example of a board config file. There are
several <emphasis>helper procedures</emphasis> used in this
example. A helper procedure is one that look for a tool of files
in commonly installed locations. These are mostly used when
testing in the build tree, because the executables to be tested
are in the same tree as the new dejagnu files. The helper
procedures are the ones in square braces
<emphasis>[]</emphasis>, which is the Tcl execution characters.</para>
<example>
<title>Example Board Config File</title>
<programlisting>
# Load the generic configuration for this board. This will define a basic
# set of routines needed by the tool to communicate with the board.
load_generic_config "sim"
# basic-sim.exp is a basic description for the standard Cygnus simulator.
load_base_board_description "basic-sim"
# The compiler used to build for this board. This has *nothing* to do
# with what compiler is tested if we're testing gcc.
set_board_info compiler "[find_gcc]"
# We only support newlib on this target.
# However, we include libgloss so we can find the linker scripts.
set_board_info cflags "[newlib_include_flags] [libgloss_include_flags]"
set_board_info ldflags "[newlib_link_flags]"
# No linker script for this board.
set_board_info ldscript "-Tsim.ld";
# The simulator doesn't return exit statuses and we need to indicate this.
set_board_info needs_status_wrapper 1
# Can't pass arguments to this target.
set_board_info noargs 1
# No signals.
set_board_info gdb,nosignals 1
# And it can't call functions.
set_board_info gdb,cannot_call_functions 1
</programlisting>
</example>
</sect2>
<sect2 id="boarddefs" xreflabel="Board File Values">
<title>Board Config File Values</title>
<para>These fields are all in the <symbol>board_info</symbol> These are
all set by using the <function>set_board_info</function> procedure. The
parameters are the field name, followed by the value to set the field
to.</para>
<para><table frame="all" rowsep="0" colsep="0">
<title>Common Board Info Fields</title>
<tgroup cols="3" align="char" rowsep="1" colsep="0">
<thead><row>
<entry>Field</entry>
<entry>Sample Value</entry>
<entry>Description</entry>
</row></thead>
<tbody>
<row>
<entry>compiler</entry>
<entry>"[find_gcc]"</entry>
<entry>The path to the compiler to use.</entry>
</row>
<row>
<entry>cflags</entry>
<entry>"-mca"</entry>
<entry>Compilation flags for the compiler.</entry>
</row>
<row>
<entry>ldflags</entry>
<entry>"[libgloss_link_flags] [newlib_link_flags]"</entry>
<entry>Linking flags for the compiler.</entry>
</row>
<row>
<entry>ldscript</entry>
<entry>"-Wl,-Tidt.ld"</entry>
<entry>The linker script to use when cross compiling.</entry>
</row>
<row>
<entry>libs</entry>
<entry>"-lgcc"</entry>
<entry>Any additional libraries to link in.</entry>
</row>
<row>
<entry>shell_prompt</entry>
<entry>"cygmon>"</entry>
<entry>The command prompt of the remote shell.</entry>
</row>
<row>
<entry>hex_startaddr</entry>
<entry>"0xa0020000"</entry>
<entry>The Starting address as a string.</entry>
</row>
<row>
<entry>start_addr</entry>
<entry>0xa0008000</entry>
<entry>The starting address as a value.</entry>
</row>
<row>
<entry>startaddr</entry>
<entry>"a0020000"</entry>
<entry></entry>
</row>
<row>
<entry>exit_statuses_bad</entry>
<entry>1</entry>
<entry>Whether there is an accurate exit status.</entry>
</row>
<row>
<entry>reboot_delay</entry>
<entry>10</entry>
<entry>The delay between power off and power on.</entry>
</row>
<row>
<entry>unreliable</entry>
<entry>1</entry>
<entry>Whether communication with the board is unreliable.</entry>
</row>
<row>
<entry>sim</entry>
<entry>[find_sim]</entry>
<entry>The path to the simulator to use.</entry>
</row>
<row>
<entry>objcopy</entry>
<entry>$tempfil</entry>
<entry>The path to the <command>objcopy</command> program.</entry>
</row>
<row>
<entry>support_libs</entry>
<entry>"${prefix_dir}/i386-coff/"</entry>
<entry>Support libraries needed for cross compiling.</entry>
</row>
<row>
<entry>addl_link_flags</entry>
<entry>"-N"</entry>
<entry>Additional link flags, rarely used.</entry>
</row>
</tbody>
</tgroup>
</table>
</para>
<para>These fields are used by the GCC and GDB tests, and are mostly
only useful to somewhat trying to debug a new board file for one of
these tools. Many of these are used only by a few testcases, and their
purpose is esoteric. These are listed with sample values as a guide to
better guessing if you need to change any of these.</para>
<para><table frame="all" rowsep="0" colsep="0">
<title>Board Info Fields For GCC &amp; GDB</title>
<tgroup cols="3" align="char" rowsep="1" colsep="0">
<thead><row>
<entry>Field</entry>
<entry>Sample Value</entry>
<entry>Description</entry>
</row></thead>
<tbody>
<row>
<entry>strip</entry>
<entry>$tempfile</entry>
<entry>Strip the executable of symbols.</entry>
</row>
<row>
<entry>gdb_load_offset</entry>
<entry>"0x40050000"</entry>
</row>
<row>
<entry>gdb_protocol</entry>
<entry>"remote"</entry>
<entry>The GDB debugging protocol to use.</entry>
</row>
<row>
<entry>gdb_sect_offset</entry>
<entry>"0x41000000";</entry>
</row>
<row>
<entry>gdb_stub_ldscript</entry>
<entry>"-Wl,-Teva-stub.ld"</entry>
<entry>The linker script to use with a GDB stub.</entry>
</row>
<row>
<entry>gdb_init_command</entry>
<entry>"set mipsfpu none"</entry>
</row>
<row>
<entry>gdb,cannot_call_functions</entry>
<entry>1</entry>
<entry>Whether GDB can call functions on the target,</entry>
</row>
<row>
<entry>gdb,noargs</entry>
<entry>1</entry>
<entry>Whether the target can take command line arguments.</entry>
</row>
<row>
<entry>gdb,nosignals</entry>
<entry>1</entry>
<entry>Whether there are signals on the target.</entry>
</row>
<row>
<entry>gdb,short_int</entry>
<entry>1</entry>
</row>
<row>
<entry>gdb,start_symbol</entry>
<entry>"_start";</entry>
<entry>The starting symbol in the executable.</entry>
</row>
<row>
<entry>gdb,target_sim_options</entry>
<entry>"-sparclite"</entry>
<entry>Special options to pass to the simulator.</entry>
</row>
<row>
<entry>gdb,timeout</entry>
<entry>540</entry>
<entry>Timeout value to use for remote communication.</entry>
</row>
<row>
<entry>gdb_init_command</entry>
<entry>"print/x \$fsr = 0x0"</entry>
</row>
<row>
<entry>gdb_load_offset</entry>
<entry>"0x12020000"</entry>
</row>
<row>
<entry>gdb_opts</entry>
<entry>"--command gdbinit"</entry>
</row>
<row>
<entry>gdb_prompt</entry>
<entry>"\\(gdb960\\)"</entry>
<entry>The prompt GDB is using.</entry>
</row>
<row>
<entry>gdb_run_command</entry>
<entry>"jump start"</entry>
</row>
<row>
<entry>gdb_stub_offset</entry>
<entry>"0x12010000"</entry>
</row>
<row>
<entry>use_gdb_stub</entry>
<entry>1</entry>
<entry>Whether to use a GDB stub.</entry>
</row>
<row>
<entry>use_vma_offset</entry>
<entry>1</entry>
</row>
<row>
<entry>wrap_m68k_aout</entry>
<entry>1</entry>
</row>
<row>
<entry>gcc,no_label_values</entry>
<entry>1</entry>
</row>
<row>
<entry>gcc,no_trampolines</entry>
<entry>1</entry>
</row>
<row>
<entry>gcc,no_varargs</entry>
<entry>1</entry>
</row>
<row>
<entry>gcc,stack_size</entry>
<entry>16384</entry>
<entry>Stack size to use with some GCC testcases.</entry>
</row>
<row>
<entry>ieee_multilib_flags</entry>
<entry>"-mieee";</entry>
</row>
<row>
<entry>is_simulator</entry>
<entry>1</entry>
</row>
<row>
<entry>needs_status_wrapper</entry>
<entry>1</entry>
</row>
<row>
<entry>no_double</entry>
<entry>1</entry>
</row>
<row>
<entry>no_long_long</entry>
<entry>1</entry>
</row>
<row>
<entry>noargs</entry>
<entry>1</entry>
</row>
<row>
<entry>nullstone,lib</entry>
<entry>"mips-clock.c"</entry>
</row>
<row>
<entry>nullstone,ticks_per_sec</entry>
<entry>3782018</entry>
</row>
<row>
<entry>sys_speed_value</entry>
<entry>200</entry>
</row>
<row>
<entry>target_install</entry>
<entry>{sh-hms}</entry>
</row>
</tbody>
</tgroup>
</table>
</para>
</sect2>
<sect2 id="writing" xreflabel="Writing A Test Case">
<title>Writing A Test Case</title>
<para>The easiest way to prepare a new test case is to base it
on an existing one for a similar situation. There are two major
categories of tests: batch or interactive. Batch oriented tests
are usually easier to write.</para>
<para>The GCC tests are a good example of batch oriented tests.
All GCC tests consist primarily of a call to a single common
procedure, Since all the tests either have no output, or only
have a few warning messages when successfully compiled. Any
non-warning output is a test failure. All the C code needed is
kept in the test directory. The test driver, written in Tcl,
need only get a listing of all the C files in the directory, and
compile them all using a generic procedure. This procedure and a
few others supporting for these tests are kept in the library
module <filename>lib/c-torture.exp</filename> in the GCC test
suite. Most tests of this kind use very few
<productname>expect</productname> features, and are coded almost
purely in Tcl.</para>
<para>Writing the complete suite of C tests, then, consisted of
these steps:</para>
<itemizedlist mark="bullet">
<listitem><para>Copying all the C code into the test directory.
These tests were based on the C-torture test created by Torbjorn
Granlund (on behalf of the Free Software Foundation) for GCC
development.</para></listitem>
<listitem><para>Writing (and debugging) the generic Tcl procedures for
compilation.</para></listitem>
<listitem><para>Writing the simple test driver: its main task is to
search the directory (using the Tcl procedure
<emphasis>glob</emphasis> for filename expansion with wildcards)
and call a Tcl procedure with each filename. It also checks for
a few errors from the testing procedure.</para></listitem>
</itemizedlist>
<para>Testing interactive programs is intrinsically more
complex. Tests for most interactive programs require some trial
and error before they are complete.</para>
<para>However, some interactive programs can be tested in a
simple fashion reminiscent of batch tests. For example, prior
to the creation of &dj;, the GDB distribution already
included a wide-ranging testing procedure. This procedure was
very robust, and had already undergone much more debugging and
error checking than many recent &dj; test cases.
Accordingly, the best approach was simply to encapsulate the
existing GDB tests, for reporting purposes. Thereafter, new GDB
tests built up a family of Tcl procedures specialized for GDB
testing.</para>
</sect2>
<sect2 id="debugging" xreflabel="Debugging A Test Case">
<title>Debugging A Test Case</title>
<para>These are the kinds of debugging information available
from &dj;:</para>
<itemizedlist mark="bullet">
<listitem><para>Output controlled by test scripts themselves,
explicitly allowed for by the test author. This kind of
debugging output appears in the detailed output recorded in the
&dj; log file. To do the same for new tests, use the
<command>verbose</command> procedure (which in turn uses the
variable also called <emphasis>verbose</emphasis>) to control
how much output to generate. This will make it easier for other
people running the test to debug it if necessary. Whenever
possible, if <emphasis>$verbose</emphasis> is
<emphasis>0</emphasis>, there should be no output other than the
output from <emphasis>pass</emphasis>,
<emphasis>fail</emphasis>, <emphasis>error</emphasis>, and
<emphasis>warning</emphasis>. Then, to whatever extent is
appropriate for the particular test, allow successively higher
values of <emphasis>$verbose</emphasis> to generate more
information. Be kind to other programmers who use your tests:
provide for a lot of debugging information.</para></listitem>
<listitem><para>Output from the internal debugging functions of
Tcl and <productname>Expect</productname>. There is a command
line options for each; both forms of debugging output are
recorded in the file <filename>dbg.log</filename> in the current
directory.</para>
<para>Use <option>--debug</option> for information from the
expect level; it generates displays of the expect attempts to
match the tool output with the patterns specified. This output
can be very helpful while developing test scripts, since it
shows precisely the characters received. Iterating between the
latest attempt at a new test script and the corresponding
<filename>dbg.log</filename> can allow you to create the final
patterns by ``cut and paste''. This is sometimes the best way
to write a test case.</para></listitem>
<listitem><para>Use <option>--strace</option> to see more
detail at the Tcl level; this shows how Tcl procedure
definitions expand, as they execute. The associated number
controls the depth of definitions expanded.</para></listitem>
<listitem><para>Finally, if the value of
<emphasis>verbose</emphasis> is 3 or greater,&dj; turns on
the expect command <command>log_user</command>. This command
prints all expect actions to the expect standard output, to the
detailed log file, and (if <option>--debug</option> is on) to
<filename>dbg.log</filename>.</para></listitem>
</itemizedlist>
</sect2>
<sect2 id="adding" xreflabel="Adding A Test Case To A Testsuite">
<title>Adding A Test Case To A Testsuite.</title>
<para>There are two slightly different ways to add a test
case. One is to add the test case to an existing directory. The
other is to create a new directory to hold your test. The
existing test directories represent several styles of testing,
all of which are slightly different; examine the directories for
the tool of interest to see which (if any) is most suitable.</para>
<para>Adding a GCC test can be very simple: just add the C code
to any directory beginning with <filename>gcc</filename>. and it
runs on the next <programlisting>runtest --tool
gcc</programlisting>.</para>
<para>To add a test to GDB, first add any source code you will
need to the test directory. Then you can either create a new
expect file, or add your test to an existing one (any
file with a <emphasis>.exp</emphasis> suffix). Creating a new
.exp file is probably a better idea if the test is significantly
different from existing tests. Adding it as a separate file also
makes upgrading easier. If the C code has to be already compiled
before the test will run, then you'll have to add it to the
<filename>Makefile.in</filename> file for that test directory,
then run <command>configure</command> and
<command>make</command>.</para>
<para>Adding a test by creating a new directory is very
similar:</para>
<itemizedlist mark="bullet">
<listitem><para>Create the new directory. All subdirectory names
begin with the name of the tool to test; e.g. G++ tests might be
in a directory called <filename>g++.other</filename>. There can
be multiple test directories that start with the same tool name
(such as <emphasis>g++</emphasis>).</para></listitem>
<listitem><para>Add the new directory name to the
<symbol>configdirs</symbol> definition in the
<filename>configure.in</filename> file for the testsuite
directory. This way when <command>make</command> and
<command>configure</command> next run, they include the new
directory.</para></listitem>
<listitem><para>Add the new test case to the directory, as
above. </para></listitem>
<listitem><para>To add support in the new directory for
configure and make, you must also create a
<filename>Makefile.in</filename> and a
<filename>configure.in</filename>.</para></listitem>
</itemizedlist>
</sect2>
<sect2 id="hints" xreflabel="Hints On Writing A Test Case">
<title>Hints On Writing A Test Case</title>
<para>It is safest to write patterns that match all the output
generated by the tested program; this is called closure.
If a pattern does not match the entire output, any output that
remains will be examined by the next <command>expect</command>
command. In this situation, the precise boundary that determines
which <command>expect</command> command sees what is very
sensitive to timing between the Expect task and the task running
the tested tool. As a result, the test may sometimes appear to
work, but is likely to have unpredictable results. (This problem
is particularly likely for interactive tools, but can also
affect batch tools---especially for tests that take a long time
to finish.) The best way to ensure closure is to use the
<option>-re</option> option for the <command>expect</command>
command to write the pattern as a full regular expressions; then
you can match the end of output using a <emphasis>$</emphasis>.
It is also a good idea to write patterns that match all
available output by using <emphasis>.*\</emphasis> after the
text of interest; this will also match any intervening blank
lines. Sometimes an alternative is to match end of line using
<emphasis>\r</emphasis> or <emphasis>\n</emphasis>, but this is
usually too dependent on terminal settings.</para>
<para>Always escape punctuation, such as <emphasis>(</emphasis>
or <emphasis>&quot;</emphasis>, in your patterns; for example, write
<emphasis>\(</emphasis>. If you forget to escape punctuation,
you will usually see an error message like <programlisting>extra
characters after close-quote.</programlisting></para>
<para>If you have trouble understanding why a pattern does not
match the program output, try using the <option>--debug</option>
option to <command>runtest</command>, and examine the debug log
carefully.</para>
<para>Be careful not to neglect output generated by setup rather
than by the interesting parts of a test case. For example,
while testing GDB, I issue a send <emphasis>set height
0\n</emphasis> command. The purpose is simply to make sure GDB
never calls a paging program. The <emphasis>set
height</emphasis> command in GDB does not generate any
output; but running any command makes GDB issue a new
<emphasis>(gdb) </emphasis> prompt. If there were no
<command>expect</command> command to match this prompt, the
output <emphasis>(gdb) </emphasis> begins the text seen by the
next <command>expect</command> command---which might make that
pattern fail to match.</para>
<para>To preserve basic sanity, I also recommended that no test
ever pass if there was any kind of problem in the test case. To
take an extreme case, tests that pass even when the tool will
not spawn are misleading. Ideally, a test in this sort of
situation should not fail either. Instead, print an error
message by calling one of the &dj; procedures
<command>error</command> or <command>warning</command>.</para>
</sect2>
<sect2 id="tvariables" xreflabel="Test Case Variables">
<title>Special variables used by test cases.</title>
<para>There are special variables used by test cases. These contain
other information from &dj;. Your test cases can use these variables,
with conventional meanings (as well as the variables saved in
<filename>site.exp</filename>. You can use the value of these variables,
but they should never be changed.</para>
<variablelist>
<varlistentry>
<term>$prms_id</term>
<listitem><para>The tracking system (e.g. GNATS) number identifying
a corresponding bugreport. (<emphasis>0</emphasis>} if you do not
specify it in the test script.)</para></listitem>
</varlistentry>
<varlistentry>
<term>$item bug_id</term>
<listitem><para>An optional bug id; may reflect a bug
identification from another organization. (<emphasis>0</emphasis>
if you do not specify it.)</para></listitem>
</varlistentry>
<varlistentry>
<term>$subdir</term>
<listitem><para>The subdirectory for the current test
case.</para></listitem>
</varlistentry>
<varlistentry>
<term>$expect_out(buffer)</term>
<listitem><para>The output from the last command. This is an
internal variable set by Expect. More information can be found in
the Expect manual.</para></listitem>
</varlistentry>
<varlistentry>
<term>$exec_output</term>
<listitem><para>This is the output from a
<function>${tool}_load</function> command. This only applies to
tools like GCC and GAS which produce an object file that must in
turn be executed to complete a test.</para></listitem>
</varlistentry>
<varlistentry>
<term>$comp_output</term>
<listitem><para>This is the output from a
<function>${tool}_start</function> command. This is conventionally
used for batch oriented programs, like GCC and GAS, that may
produce interesting output (warnings, errors) without further
interaction.</para></listitem>
</varlistentry>
</variablelist>
</sect2>
</sect1>
<sect1 id="unit" xreflabel="Unit Testing">
<title>Unit Testing</title>
<sect2 id="unittest" xreflabel="What Is Unit Testing ?">
<title>What Is Unit Testing ?</title>
<para>Most regression testing as done by &dj; is system
testing. This is the complete application is tested all at
once. Unit testing is for testing single files, or small
libraries. In this case, each file is linked with a test case in
C or C++, and each function or class and method is tested in
series, with the test case having to check private data or
global variables to see if the function or method worked.</para>
<para>This works particularly well for testing APIs and at level
where it is easier to debug them, than by needing to trace through
the entire appication. Also if there is a specification for the
API to be tested, the testcase can also function as a compliance
test.</para>
</sect2>
<sect2 id="djh" xreflabel="The dejagnu.h Header File">
<title>The dejagnu.h Header File</title>
<para>&dj; uses a single header file to assist in unit
testing. As this file also produces it's one test state output,
it can be run standalone, which is very useful for testing on
embedded systems. This header file has a C and C++ API for the
test states, with simple totals, and standardized
output. Because the output has been standardized, &dj; can be
made to work with this test case, without writing almost any
Tcl. The library module, dejagnu.exp, will look for the output
messages, and then merge them into &dj;'s.</para>
</sect2>
</sect1>
<!-- Keep this comment at the end of the file
Local variables:
mode: sgml
sgml-omittag:t
sgml-shorttag:t
sgml-namecase-general:t
sgml-general-insert-case:lower
sgml-minimize-attributes:nil
sgml-always-quote-attributes:t
sgml-indent-step:1
sgml-indent-data:nil
sgml-parent-document:nil
sgml-exposed-tags:nil
sgml-local-catalogs:nil
sgml-local-ecat-files:nil
End:
-->