This is the mail archive of the gdb-patches@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [RFA] breakpoint.c: Avoid double freeing in breakpoint_re_set_one


> > Anyway, I've committed the attached (slightly tweaked).
> 
> Doh! Try ...

I was going to say that was quite an editing job!

> You wouldn't have a test case - at least something that causes this code 
> to go through the motions?  That way on systems with pedantic memory 
> managers there'd bee some sort of error.

As you suggest, this is a little hard to test.  I am not too familiar with
the testsuite innards yet.  Here is a draft of a possible test.  Comments?

Paul

---------------------------- gdb.base/badfree.exp --------------------
# Copyright 2004 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@prep.ai.mit.edu

# Author: Paul N. Hilfinger (Hilfinger@gnat.com)

# Test that GDB cleans up properly after errors that result when a 
# breakpoint is reset. 

if $tracelevel then {
	strace $tracelevel
}

# IDT/SIM apparently doesn't have enough file descriptors to allow the
# problem checked by this test to occur.
if [istarget "mips-idt-*"] {
    return 0;
}

set testfile "badfree"
set srcfile ${testfile}.c
set binfile ${objdir}/${subdir}/${testfile}

if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug additional_flags=-DVARIABLE=var1}] != "" } {
    gdb_suppress_entire_file "Testcase compile failed, so all tests in this file will automatically fail."
}

set oldtimeout $timeout
set timeout 10
verbose "Timeout is now 10 seconds" 2

proc expect_to_stop_here { ident } {
    global gdb_prompt
    global decimal

    # the "at foo.c:36" output we get with -g.
    # the "in func" output we get without -g.
    gdb_expect 30 {
	-re "Breakpoint \[0-9\]*, stop_here .*$gdb_prompt $" { 
	    return 1
	}
	-re "$gdb_prompt $" { 
	    fail "running to stop_here $ident"
	    return 0
	}
	timeout { 
	    fail "running to stop_here $ident (timeout)"
	    return 0
	}
    }
    return 1
}

gdb_exit
gdb_start
gdb_reinitialize_dir $srcdir/$subdir
gdb_load ${binfile}

gdb_test "break stop_here if (var1 == 42)\n" \
    "Breakpoint.*at.* file .*$srcfile, line.*" \
    "setting conditional breakpoint on function"
gdb_run_cmd

expect_to_stop_here "first time"
    
gdb_continue_to_end "breakpoint first time through"

# Now we recompile the executable, but without a variable named "var1", first
# waiting to insure that even on fast machines, the file modification times
# are distinct. This will force GDB to reload the file on the 
# next "run" command causing an error when GDB tries to tries to reset 
# the breakpoint.

sleep 2
if  { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug additional_flags=-DVARIABLE=var2}] != "" } {

# Complication: Since GDB generally holds an open file descriptor on the 
# executable at this point, there are some systems in which the 
# re-compilation will fail. In such cases, we'll consider the test 
# (vacuously) passed providing that re-running it succeeds as before.

    gdb_run_cmd
    expect_to_stop_here "after re-compile fails"
    gdb_continue_to_end "after re-compile fails"

} else {

    send_gdb "run\n"
    gdb_expect {
	-re ".* has changed; .*Error in re-setting .*No symbol .var1..*Program exited normally.*"\
		{ pass "running with invalidated bpt condition after executable changes" }
	timeout {
	    fail "(timeout) running with invalidated bpt condition after executable changes" }
    }

}

---------------------------- gdb.base/badfree.c --------------------
/*
 *   Test that GDB cleans up properly after errors that result when a 
 * breakpoint is reset. 
 */

/* VARIABLE is a macro defined on the compiler command line. */

#include <stdlib.h>

int VARIABLE = 42;

void stop_here ()
{
  VARIABLE *= 2;
}

main ()
{
  stop_here ();
  exit (0);
}


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]