Re: [RFA] Testsuite addition for x86 linux GDB and SIGALRM fix

[Andrew Cagney <ac131313 at cygnus dot com>]

Anyone?


	Andrew

> Here's the test I promised Andrew a while ago for the fix for the
> problem reported by Jonathan Larmour:
> 
>    http://sourceware.cygnus.com/ml/gdb/2000-q1/msg00803.html
> 
> The fix has already been checked in, the problem is still mentioned in
> the TODO file (let's keep it there until this test has been added).
> 
> I verified that some of these tests (the "stepi" and "nexti" tests)
> do fail without my fix to infrun.c.
> 
> I'm not sure to what extent the use of setitimer() is portable.
> However, it is hard to come up with a test that doesn't use it.
> 
> 
> 2000-05-20  Mark Kettenis  <kettenis@gnu.org>
> 
> Add tests for stepping with pending signals.
> 	* gdb.base/step-alarm.exp: New file.
> 	* gdb.base/step-alarm.c: New file.
> 
> 
> --- /dev/null	Thu Feb 19 16:30:24 1998
> +++ testsuite/gdb.base/step-alarm.exp	Sat May 20 01:09:56 2000
> @@ -0,0 +1,226 @@
> +# Copyright (C) 1997, 1998, 1999, 2000 Free Software Foundation, Inc.
> +
> +# This program is free software; you can redistribute it and/or modify
> +# it under the terms of the GNU General Public License as published by
> +# the Free Software Foundation; either version 2 of the License, or
> +# (at your option) any later version.
> +# 
> +# This program is distributed in the hope that it will be useful,
> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> +# GNU General Public License for more details.
> +# 
> +# You should have received a copy of the GNU General Public License
> +# along with this program; if not, write to the Free Software
> +# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.  */
> +
> +# Please email any bugs, comments, and/or additions to this file to:
> +# bug-gdb@gnu.org
> +
> +# use this to debug:
> +#
> +#log_user 1
> +
> +# step-alarm.exp -- Expect script to test stepping in gdb
> +#                   with a pending signals
> +
> +# Most of this script is copied over from step-test.exp.  We run (almost)
> +# the same tests, except that we set a timer and install a SIGALRM signal
> +# handler.  The addition of these tests was prompted by the following fix:
> +#
> +# 2000-05-01  Mark Kettenis  <kettenis@gnu.org>
> +#
> +#	* infrun.c (handle_inferior_event): Add missing call to keep_going
> +#	and missing return when handling an ordinary signal from the
> +#	inferior.
> +#
> +# for a problem where "stepi" didn't make any progress if a signal
> +# was pending. 
> +
> +if $tracelevel then {
> +    strace $tracelevel
> +}
> +
> +set testfile step-alarm
> +set srcfile ${srcdir}/${subdir}/${testfile}.c
> +set binfile ${objdir}/${subdir}/${testfile}
> +
> +remote_exec build "rm -f ${binfile}"
> +if { [gdb_compile "${srcfile}" "${binfile}" executable {debug}] != "" } {
> +     gdb_suppress_entire_file "Testcase compile failed, so all tests in this file will automatically fail."
> +}
> +
> +gdb_exit
> +gdb_start
> +gdb_reinitialize_dir $srcdir/$subdir
> +gdb_load ${binfile}
> +
> +if ![runto_main] then {
> +   fail "Can't run to main"
> +   return 0
> +}
> +
> +# Make sure that the signal handler is installed, and the timer is set.
> +#
> +gdb_test "break [gdb_get_line_number "w = 0"]" \
> +         ".*Breakpoint.* at .*" \
> +         "set breakpoint after timer activation"
> +gdb_test "continue" \
> +	 ".*Breakpoint ${decimal},.*w = 0.*" \
> +	 "run until timer is activated"
> +
> +
> +# Set a breakpoint at line 57, if stepi then finish fails, we would
> +# run to the end of the program, which would mess up the rest of the tests.
> +  
> +# Vanilla step/next
> +#
> +gdb_test "next" ".*${decimal}.*x = 1;.*" "next 1"
> +gdb_test "step" ".*${decimal}.*y = 2;.*" "step 1"
> +
> +# With count
> +#
> +gdb_test "next 2" ".*${decimal}.*w = w.*2;.*" "next 2"
> +gdb_test "step 3" ".*${decimal}.*z = z.*5;.*" "step 3"
> +gdb_test "next" ".*${decimal}.*callee.*OVER.*" "next 3"
> +
> +# Step over call
> +#
> +gdb_test "next" ".*${decimal}.*callee.*INTO.*" "next over"
> +
> +# Step into call
> +#
> +gdb_test "step" ".*${decimal}.*myglob.*"   "step into"
> +
> +# Step out of call
> +#
> +# I wonder if this is really portable.  Are there any caller-saves
> +# platforms, on which `finish' will return you to some kind of pop
> +# instruction, which is attributed to the line containing the function
> +# call?
> +
> +# On PA64, we end up at a different instruction than PA32.
> +# On IA-64, we also end up on callee instead of on the next line due
> +# to the restoration of the global pointer (which is a caller-save).
> +if { [istarget "hppa2.0w-hp-hpux*"] || [istarget "ia64-*-*"]} {
> +    send_gdb "finish\n"
> +    gdb_expect {
> +        -re ".*${decimal}.*a.*5.*= a.*3.*$gdb_prompt $" { pass "step out 1" }
> +        -re ".*${decimal}.*callee.*INTO.*$gdb_prompt $" { pass "step out 2" }
> +        timeout { fail "step out" }
> +    }
> +} else {
> +    gdb_test "finish" ".*${decimal}.*a.*5.*= a.*3.*" "step out"
> +}
> +
> +### Testing nexti and stepi.
> +###
> +### test_i NAME COMMAND HERE THERE
> +###
> +### Send COMMAND to gdb over and over, while the output matches the
> +### regexp HERE, followed by the gdb prompt.  Pass if the output
> +### eventually matches the regexp THERE, followed by the gdb prompt;
> +### fail if we have to iterate more than a hundred times, we time out
> +### talking to gdb, or we get output which is neither HERE nor THERE.  :)
> +###
> +### Use NAME as the name of the test.
> +###
> +### The exact regexps used are "$HERE.*$gdb_prompt $"
> +###                        and "$THERE.*$gdb_prompt $"
> +### 
> +proc test_i {name command here there} {
> +    global gdb_prompt
> +    
> +    set i 0
> +    while 1 {
> +	send_gdb "${command}\n"
> +	gdb_expect {
> +	    -re "$here.*$gdb_prompt $" {
> +		# Okay, we're still on the same line.  Just step again.
> +	    }
> +	    -re "$there.*$gdb_prompt $" {
> +		# We've reached the next line.  Rah.
> +		pass "$name"
> +		return
> +	    }
> +	    -re "$gdb_prompt $" {
> +		# We got something else.  Fail.
> +		fail "$name"
> +		return
> +	    }
> +	    timeout {
> +		fail "$name (timeout)"
> +		return
> +	    }
> +	}
> +
> +	# Have we gone for too many steps without seeing any progress?
> +	if {[incr i] >= 100} {
> +	    fail "$name (no progress after 100 steps)"
> +	    return
> +	}
> +    }
> +}
> +
> +test_i "stepi to next line" "stepi" \
> +       ".*${decimal}.*a.*5.* = a.*3" \
> +       ".*${decimal}.*callee.*STEPI"
> +
> +test_i "stepi into function" "stepi" \
> +       ".*${decimal}.*callee.*STEPI" \
> +       ".*callee \\(\\) at .*step-alarm\\.c" 
> +
> +# Continue to step until we reach the function's body.  This makes it
> +# more likely that we've actually completed the prologue, so "finish"
> +# will work.
> +test_i "stepi into function's first source line" "stepi" \
> +	".*${decimal}.*\\{" \
> +	".*${decimal}.*myglob"
> +
> +# Have to be careful here, if the finish does not work, 
> +# then we may run to the end of the program, which
> +# will cause erroneous failures in the rest of the tests
> +send_gdb "finish\n"
> +gdb_expect {
> +  -re ".*(Program received|Program exited).*$gdb_prompt $" {
> +    # Oops... We ran to the end of the program...  Better reset     
> +    if {![runto_main]} then {
> +      fail "Can't run to main"
> +      return 0
> +    }
> +    if {![runto step-alarm.c:57]} {
> +      fail "Can't run to line 57"
> +      return 0
> +    }
> +    fail "stepi: finish call"
> +  }
> +  -re ".*${decimal}.*callee.*NEXTI.*$gdb_prompt $" {
> +    pass "stepi: finish call"
> +  }
> +  -re ".*${decimal}.*callee.*STEPI.*$gdb_prompt $" {
> +    # On PA64, we end up at a different instruction than PA32.
> +    # On IA-64, we end up on callee instead of on the following line due
> +    # to the restoration of the global pointer.
> +    if { [istarget "hppa2.0w-hp-hpux*"] || [istarget "ia64-*-*"] } {
> +        pass "stepi: finish call 2"
> +    } else {
> +	fail "stepi: finish call 2"
> +        return
> +    }
> +  }
> +  -re "$gdb_prompt $" {
> +    # We got something else.  Fail.
> +    fail "stepi: finish call"
> +    return
> +  }
> +  timeout {
> +    fail "stepi: finish call (timeout)"
> +    return
> +  }
> +}
> +
> +test_i "nexti over function" "nexti" \
> +       ".*${decimal}.*callee.*NEXTI" \
> +       ".*${decimal}.*y = w \\+ z;"
> +
> +return 0
> --- /dev/null	Thu Feb 19 16:30:24 1998
> +++ testsuite/gdb.base/step-alarm.c	Sat May 20 01:01:21 2000
> @@ -0,0 +1,65 @@
> +#include <signal.h>
> +#include <stdio.h>
> +#include <sys/time.h>
> +
> +/* Test stepping with a pending signal.  */
> +
> +int myglob = 0;
> +
> +int callee (void)
> +{
> +  myglob++;
> +  return 0;
> +}
> +
> +static void
> +handler (int signum)
> +{
> +}
> +
> +int
> +main (void)
> +{
> +  struct sigaction sa;
> +  struct itimerval it;
> +  int w, x, y, z;
> +  int a[10];
> +
> +  sa.sa_handler = handler;
> +  sigemptyset (&sa.sa_mask);
> +  sa.sa_flags = 0;
> +  sigaction (SIGALRM, &sa, NULL);
> +
> +  it.it_interval.tv_usec = 5000;
> +  it.it_interval.tv_sec = 0;
> +  it.it_value.tv_usec = 5000;
> +  it.it_value.tv_sec = 0;
> +  setitimer (ITIMER_REAL, &it, NULL);
> +
> +  /* Test "next" and "step" */
> +  w = 0;
> +  x = 1;
> +  y = 2;
> +  z = 3;
> +  w = w + 2;
> +  x = x + 3;
> +  y = y + 4;
> +  z = z + 5;
> +
> +  /* Test that "next" goes over a call */
> +  callee(); /* OVER */
> +
> +  /* Test that "step" doesn't */
> +  callee(); /* INTO */
> +
> +  /* Test "stepi"  */
> +  a[5] = a[3] - a[4];
> +  callee(); /* STEPI */
> +
> +  /* Test "nexti" */
> +  callee(); /* NEXTI */
> +  
> +  y = w + z;
> +
> +  exit (0);
> +}
> 
>