1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
|
# Copyright 2003, 2007, 2008, 2009 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 3 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, see <http://www.gnu.org/licenses/>.
# This is a regression test for the following bug, as of 2003-12-12:
#
# Set a breakpoint which will be hit many times. Attach a complex set
# of commands to it, including a "continue" command. Run the program,
# so that the breakpoint is hit, its commands get executed, and the
# program continues and hits the breakpoint again. You will see
# messages like "warning: Invalid control type in command structure.",
# or maybe GDB will crash.
#
# When the breakpoint is hit, bpstat_stop_status copies the
# breakpoint's command tree to the bpstat. bpstat_do_actions then
# calls execute_control_command to run the commands. The 'continue'
# command invokes the following chain of calls:
#
# continue_command
# -> clear_proceed_status
# -> bpstat_clear
# -> free_command_lines
# -> frees the commands we are currently running.
#
# When control does eventually return to execute_control_command, GDB
# continues to walk the tree of freed command nodes, resulting in the
# error messages and / or crashes.
#
# Since this bug depends on storage being reused between the time that
# we continue and the time that we fall back to bpstat_do_actions, the
# reproduction recipe is more delicate than I would like. I welcome
# suggestions for improving this.
set prms_id 0
set bug_id 0
set testfile "freebpcmd"
set srcfile ${testfile}.c
set binfile ${objdir}/${subdir}/${testfile}
if { [gdb_compile "${srcdir}/${subdir}/${srcfile}" "${binfile}" executable {debug}] != "" } {
untested freebpcmd.exp
return -1
}
gdb_exit
gdb_start
gdb_reinitialize_dir $srcdir/$subdir
gdb_load ${binfile}
gdb_test "break ${srcfile}:[gdb_get_line_number "euphonium"]" "" "set breakpoint"
# The goal of all this is to make sure that there's plenty of memory
# churn, and different amounts of it each time the inferior stops;
# this seems to make GDB crash more reliably.
set lines {{if (i%2) == 0}
{echo "even "}
{print i}
{else}
{echo "odd "}
{print i}
{end}
{set variable $foo = 0}
{set variable $j = 0}
{while $j < i}
{set variable $foo += $j}
{set variable $j++}
{end}
{print $foo}
{if i != 40}
{c}
{end}
{end}}
send_gdb "commands\n"
for {set i 0} {$i < [llength $lines]} {incr i} {
gdb_expect {
-re ".*>" {
send_gdb "[lindex $lines $i]\n"
}
-re "$gdb_prompt $" {
set reason "got top-level prompt early"
break
}
timeout {
set reason "timeout"
break
}
}
}
if {$i >= [llength $lines]} {
pass "send breakpoint commands"
} else {
fail "send breakpoint commands ($reason)"
}
gdb_run_cmd
set timeout 120
gdb_test_multiple "" "run program with breakpoint commands" {
-re "warning: Invalid control type in command structure" {
kfail "gdb/1489" "run program with breakpoint commands"
}
-re "$gdb_prompt $" {
pass "run program with breakpoint commands"
}
eof {
kfail "gdb/1489" "run program with breakpoint commands (GDB died)"
}
}
|
