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

[Michael Snyder <msnyder at cygnus dot com>]

Andrew Cagney wrote:
> 
> Anyone?

I think the test is good.  Just a couple of comments:
1) There are a couple of instances of the following:
  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"
  }
I think that the "hppa2.0" clause should be used for all targets, 
because there are numerous targets on which a "finish" command will
end up on the same line as the function call, rather than on the 
next line.  I don't think it is important enough to try to 
distinguish which targets can expect this behavior -- instead, 
it is good enough to just accept either result.

2) Do the gnu coding standards apply to testsuite source files?
If so, "callee()" should be replaced by "callee ()".

Michael



> 
>         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);
> > +}
> >
> >