binutils-gdb/gdb/testsuite/gdb.base/bp-permanent.exp

288 lines
8.6 KiB
Plaintext
Raw Normal View History

# Copyright (C) 2014-2015 Free Software Foundation, Inc.
fix skipping permanent breakpoints The gdb.arch/i386-bp_permanent.exp test is currently failing an assertion recently added: (gdb) stepi ../../src/gdb/infrun.c:2237: internal-error: resume: Assertion `sig != GDB_SIGNAL_0' failed. A problem internal to GDB has been detected, further debugging may prove unreliable. Quit this debugging session? (y or n) FAIL: gdb.arch/i386-bp_permanent.exp: Single stepping past permanent breakpoint. (GDB internal error) The assertion expects that the only reason we currently need to step a breakpoint instruction is when we have a signal to deliver. But when stepping a permanent breakpoint (with or without a signal) we also reach this code. The assertion is correct and the permanent breakpoints skipping code is wrong. Consider the case of the user doing "step/stepi" when stopped at a permanent breakpoint. GDB's `resume' calls the gdbarch_skip_permanent_breakpoint hook and then happily continues stepping: /* Normally, by the time we reach `resume', the breakpoints are either removed or inserted, as appropriate. The exception is if we're sitting at a permanent breakpoint; we need to step over it, but permanent breakpoints can't be removed. So we have to test for it here. */ if (breakpoint_here_p (aspace, pc) == permanent_breakpoint_here) { gdbarch_skip_permanent_breakpoint (gdbarch, regcache); } But since gdbarch_skip_permanent_breakpoint already advanced the PC manually, this ends up executing the instruction that is _after_ the breakpoint instruction. The user-visible result is that a single-step steps two instructions. The gdb.arch/i386-bp_permanent.exp test is actually ensuring that that's indeed how things work. It runs to an int3 instruction, does "stepi", and checks that "leave" was executed with that "stepi". Like this: (gdb) b *0x0804848c Breakpoint 2 at 0x804848c (gdb) c Continuing. Breakpoint 2, 0x0804848c in standard () (gdb) disassemble Dump of assembler code for function standard: 0x08048488 <+0>: push %ebp 0x08048489 <+1>: mov %esp,%ebp 0x0804848b <+3>: push %edi => 0x0804848c <+4>: int3 0x0804848d <+5>: leave 0x0804848e <+6>: ret 0x0804848f <+7>: nop (gdb) si 0x0804848e in standard () (gdb) disassemble Dump of assembler code for function standard: 0x08048488 <+0>: push %ebp 0x08048489 <+1>: mov %esp,%ebp 0x0804848b <+3>: push %edi 0x0804848c <+4>: int3 0x0804848d <+5>: leave => 0x0804848e <+6>: ret 0x0804848f <+7>: nop End of assembler dump. (gdb) One would instead expect that a stepi at 0x0804848c stops at 0x0804848d, _before_ the "leave" is executed. This commit changes GDB this way. Care is taken to make stepping into a signal handler when the step starts at a permanent breakpoint instruction work correctly. The patch adjusts gdb.arch/i386-bp_permanent.exp in this direction, and also makes it work on x86_64 (currently it only works on i*86). The patch also adds a new gdb.base/bp-permanent.exp test that exercises many different code paths related to stepping permanent breakpoints, including the stepping with signals cases. The test uses "hack/trick" to make it work on all (or most) platforms -- it doesn't really hard code a breakpoint instruction. Tested on x86_64 Fedora 20, native and gdbserver. gdb/ 2014-11-12 Pedro Alves <palves@redhat.com> * infrun.c (resume): Clear the thread's 'stepped_breakpoint' flag. Rewrite stepping over a permanent breakpoint. (thread_still_needs_step_over, proceed): Don't set stepping_over_breakpoint for permanent breakpoints. (handle_signal_stop): Don't clear stepped_breakpoint. Also pull single-step breakpoints out of the target on hardware step targets. (process_event_stop_test): If stepping a permanent breakpoint doesn't hit the step-resume breakpoint, delete the step-resume breakpoint. (switch_back_to_stepped_thread): Also check if the stepped thread has advanced already on hardware step targets. (currently_stepping): Return true if the thread stepped a breakpoint. gdb/testsuite/ 2014-11-12 Pedro Alves <palves@redhat.com> * gdb.arch/i386-bp_permanent.c: New file. * gdb.arch/i386-bp_permanent.exp: Don't skip on x86_64. (srcfile): Set to i386-bp_permanent.c. (top level): Adjust to work in both 32-bit and 64-bit modes. Test that stepi does not execute the 'leave' instruction, instead of testing it does execute. * gdb.base/bp-permanent.c: New file. * gdb.base/bp-permanent.exp: New file.
2014-11-12 18:10:49 +08:00
# 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 file is part of the gdb testsuite.
# Test stepping over permanent breakpoints.
standard_testfile
set options { debug }
if { ![target_info exists gdb,nosignals] } {
lappend options "additional_flags=-DSIGNALS"
}
if {[build_executable "failed to prepare" $testfile $srcfile $options]} {
return -1
}
set line_bp [gdb_get_line_number "write permanent bp"]
# The test proper. ALWAYS_INSERTED indicates whether testing in
# "breakpoint always-inserted" mode. If SW_WATCHPOINT is true, set a
# software watchpoint, which forces constantly single-stepping, and
# exercises stepping the permanent breakpoint while delivering a
# signal at the same time.
proc test {always_inserted sw_watchpoint} {
global line_bp
global hex decimal
global gdb_prompt
global srcfile binfile
clean_restart $binfile
if ![runto_main] then {
return -1
}
gdb_test "set breakpoint always-inserted $always_inserted"
if {$sw_watchpoint} {
# Watching a convenience variable forces a software
# watchpoint.
gdb_test "watch \$dummy_convenience" "Watchpoint .*"
}
set address_bp ""
set address_after_bp ""
with_test_prefix "setup" {
# Set a breakpoint where we'll manually plant a permanent
# breakpoint.
set test "set probe breakpoint"
gdb_test_multiple "break $line_bp" $test {
-re "Breakpoint .* at ($hex).*$gdb_prompt $" {
set address_bp $expect_out(1,string)
pass $test
}
}
if {$address_bp == ""} {
return
}
# Get the size of the instruction where the breakpoint will
# manually inserted.
set test "get size of instruction"
gdb_test_multiple "x/2i $address_bp" $test {
-re ".*$hex <test\\+$decimal>:\[^\r\n\]+\r\n\[ \]+($hex).*\.\r\n$gdb_prompt $" {
set address_after_bp $expect_out(1,string)
pass $test
}
}
if {$address_after_bp == ""} {
return
}
# Write address range where the breakpoint is inserted to the
# corresponding variables in the inferior.
gdb_test "p /x addr_bp = $address_bp" " = $address_bp" \
"write addr_bp"
gdb_test "p /x addr_after_bp = $address_after_bp" " = $address_after_bp" \
"write addr_after_bp"
# Run the "setup" function in the inferior. This memcpy's the
# breakpoint instruction to a buffer in the inferior.
gdb_test "next" "test_basics \\(\\).*" "next over setup"
fix skipping permanent breakpoints The gdb.arch/i386-bp_permanent.exp test is currently failing an assertion recently added: (gdb) stepi ../../src/gdb/infrun.c:2237: internal-error: resume: Assertion `sig != GDB_SIGNAL_0' failed. A problem internal to GDB has been detected, further debugging may prove unreliable. Quit this debugging session? (y or n) FAIL: gdb.arch/i386-bp_permanent.exp: Single stepping past permanent breakpoint. (GDB internal error) The assertion expects that the only reason we currently need to step a breakpoint instruction is when we have a signal to deliver. But when stepping a permanent breakpoint (with or without a signal) we also reach this code. The assertion is correct and the permanent breakpoints skipping code is wrong. Consider the case of the user doing "step/stepi" when stopped at a permanent breakpoint. GDB's `resume' calls the gdbarch_skip_permanent_breakpoint hook and then happily continues stepping: /* Normally, by the time we reach `resume', the breakpoints are either removed or inserted, as appropriate. The exception is if we're sitting at a permanent breakpoint; we need to step over it, but permanent breakpoints can't be removed. So we have to test for it here. */ if (breakpoint_here_p (aspace, pc) == permanent_breakpoint_here) { gdbarch_skip_permanent_breakpoint (gdbarch, regcache); } But since gdbarch_skip_permanent_breakpoint already advanced the PC manually, this ends up executing the instruction that is _after_ the breakpoint instruction. The user-visible result is that a single-step steps two instructions. The gdb.arch/i386-bp_permanent.exp test is actually ensuring that that's indeed how things work. It runs to an int3 instruction, does "stepi", and checks that "leave" was executed with that "stepi". Like this: (gdb) b *0x0804848c Breakpoint 2 at 0x804848c (gdb) c Continuing. Breakpoint 2, 0x0804848c in standard () (gdb) disassemble Dump of assembler code for function standard: 0x08048488 <+0>: push %ebp 0x08048489 <+1>: mov %esp,%ebp 0x0804848b <+3>: push %edi => 0x0804848c <+4>: int3 0x0804848d <+5>: leave 0x0804848e <+6>: ret 0x0804848f <+7>: nop (gdb) si 0x0804848e in standard () (gdb) disassemble Dump of assembler code for function standard: 0x08048488 <+0>: push %ebp 0x08048489 <+1>: mov %esp,%ebp 0x0804848b <+3>: push %edi 0x0804848c <+4>: int3 0x0804848d <+5>: leave => 0x0804848e <+6>: ret 0x0804848f <+7>: nop End of assembler dump. (gdb) One would instead expect that a stepi at 0x0804848c stops at 0x0804848d, _before_ the "leave" is executed. This commit changes GDB this way. Care is taken to make stepping into a signal handler when the step starts at a permanent breakpoint instruction work correctly. The patch adjusts gdb.arch/i386-bp_permanent.exp in this direction, and also makes it work on x86_64 (currently it only works on i*86). The patch also adds a new gdb.base/bp-permanent.exp test that exercises many different code paths related to stepping permanent breakpoints, including the stepping with signals cases. The test uses "hack/trick" to make it work on all (or most) platforms -- it doesn't really hard code a breakpoint instruction. Tested on x86_64 Fedora 20, native and gdbserver. gdb/ 2014-11-12 Pedro Alves <palves@redhat.com> * infrun.c (resume): Clear the thread's 'stepped_breakpoint' flag. Rewrite stepping over a permanent breakpoint. (thread_still_needs_step_over, proceed): Don't set stepping_over_breakpoint for permanent breakpoints. (handle_signal_stop): Don't clear stepped_breakpoint. Also pull single-step breakpoints out of the target on hardware step targets. (process_event_stop_test): If stepping a permanent breakpoint doesn't hit the step-resume breakpoint, delete the step-resume breakpoint. (switch_back_to_stepped_thread): Also check if the stepped thread has advanced already on hardware step targets. (currently_stepping): Return true if the thread stepped a breakpoint. gdb/testsuite/ 2014-11-12 Pedro Alves <palves@redhat.com> * gdb.arch/i386-bp_permanent.c: New file. * gdb.arch/i386-bp_permanent.exp: Don't skip on x86_64. (srcfile): Set to i386-bp_permanent.c. (top level): Adjust to work in both 32-bit and 64-bit modes. Test that stepi does not execute the 'leave' instruction, instead of testing it does execute. * gdb.base/bp-permanent.c: New file. * gdb.base/bp-permanent.exp: New file.
2014-11-12 18:10:49 +08:00
delete_breakpoints
# We now have the breakpoint instruction stored in 'buffer'. Poke it
# to memory manually.
set count [expr $address_after_bp - $address_bp]
for {set i 0} {$i < $count} {incr i} {
set test "p /x addr_bp\[$i\] = buffer\[$i\]"
gdb_test_multiple $test $test {
-re "Cannot access memory at address $hex.*$gdb_prompt $" {
# Some targets (QEMU for one) will disallow writes to the
# .text section under certain circumstances. It is no use
# continuing with the test at this point. Just return.
unsupported "Cannot modify memory"
return
}
-re " = .*$gdb_prompt $" {
pass $test
}
}
fix skipping permanent breakpoints The gdb.arch/i386-bp_permanent.exp test is currently failing an assertion recently added: (gdb) stepi ../../src/gdb/infrun.c:2237: internal-error: resume: Assertion `sig != GDB_SIGNAL_0' failed. A problem internal to GDB has been detected, further debugging may prove unreliable. Quit this debugging session? (y or n) FAIL: gdb.arch/i386-bp_permanent.exp: Single stepping past permanent breakpoint. (GDB internal error) The assertion expects that the only reason we currently need to step a breakpoint instruction is when we have a signal to deliver. But when stepping a permanent breakpoint (with or without a signal) we also reach this code. The assertion is correct and the permanent breakpoints skipping code is wrong. Consider the case of the user doing "step/stepi" when stopped at a permanent breakpoint. GDB's `resume' calls the gdbarch_skip_permanent_breakpoint hook and then happily continues stepping: /* Normally, by the time we reach `resume', the breakpoints are either removed or inserted, as appropriate. The exception is if we're sitting at a permanent breakpoint; we need to step over it, but permanent breakpoints can't be removed. So we have to test for it here. */ if (breakpoint_here_p (aspace, pc) == permanent_breakpoint_here) { gdbarch_skip_permanent_breakpoint (gdbarch, regcache); } But since gdbarch_skip_permanent_breakpoint already advanced the PC manually, this ends up executing the instruction that is _after_ the breakpoint instruction. The user-visible result is that a single-step steps two instructions. The gdb.arch/i386-bp_permanent.exp test is actually ensuring that that's indeed how things work. It runs to an int3 instruction, does "stepi", and checks that "leave" was executed with that "stepi". Like this: (gdb) b *0x0804848c Breakpoint 2 at 0x804848c (gdb) c Continuing. Breakpoint 2, 0x0804848c in standard () (gdb) disassemble Dump of assembler code for function standard: 0x08048488 <+0>: push %ebp 0x08048489 <+1>: mov %esp,%ebp 0x0804848b <+3>: push %edi => 0x0804848c <+4>: int3 0x0804848d <+5>: leave 0x0804848e <+6>: ret 0x0804848f <+7>: nop (gdb) si 0x0804848e in standard () (gdb) disassemble Dump of assembler code for function standard: 0x08048488 <+0>: push %ebp 0x08048489 <+1>: mov %esp,%ebp 0x0804848b <+3>: push %edi 0x0804848c <+4>: int3 0x0804848d <+5>: leave => 0x0804848e <+6>: ret 0x0804848f <+7>: nop End of assembler dump. (gdb) One would instead expect that a stepi at 0x0804848c stops at 0x0804848d, _before_ the "leave" is executed. This commit changes GDB this way. Care is taken to make stepping into a signal handler when the step starts at a permanent breakpoint instruction work correctly. The patch adjusts gdb.arch/i386-bp_permanent.exp in this direction, and also makes it work on x86_64 (currently it only works on i*86). The patch also adds a new gdb.base/bp-permanent.exp test that exercises many different code paths related to stepping permanent breakpoints, including the stepping with signals cases. The test uses "hack/trick" to make it work on all (or most) platforms -- it doesn't really hard code a breakpoint instruction. Tested on x86_64 Fedora 20, native and gdbserver. gdb/ 2014-11-12 Pedro Alves <palves@redhat.com> * infrun.c (resume): Clear the thread's 'stepped_breakpoint' flag. Rewrite stepping over a permanent breakpoint. (thread_still_needs_step_over, proceed): Don't set stepping_over_breakpoint for permanent breakpoints. (handle_signal_stop): Don't clear stepped_breakpoint. Also pull single-step breakpoints out of the target on hardware step targets. (process_event_stop_test): If stepping a permanent breakpoint doesn't hit the step-resume breakpoint, delete the step-resume breakpoint. (switch_back_to_stepped_thread): Also check if the stepped thread has advanced already on hardware step targets. (currently_stepping): Return true if the thread stepped a breakpoint. gdb/testsuite/ 2014-11-12 Pedro Alves <palves@redhat.com> * gdb.arch/i386-bp_permanent.c: New file. * gdb.arch/i386-bp_permanent.exp: Don't skip on x86_64. (srcfile): Set to i386-bp_permanent.c. (top level): Adjust to work in both 32-bit and 64-bit modes. Test that stepi does not execute the 'leave' instruction, instead of testing it does execute. * gdb.base/bp-permanent.c: New file. * gdb.base/bp-permanent.exp: New file.
2014-11-12 18:10:49 +08:00
}
}
with_test_prefix "basics" {
# Run to the permanent breakpoint, just to make sure we've inserted it
# correctly.
gdb_test "continue" "Program received signal SIGTRAP.*" \
"permanent breakpoint causes random signal"
# Now set a breakpoint on top, thus creating a permanent breakpoint.
gdb_breakpoint "$line_bp"
# Depending on whether this is a decr_pc_after_break arch, the PC will
# be either pointing at the permanent breakpoint address, or just
# after. Set the GDB breakpoint on top, and continue, twice. At
# least once, GDB will need to step-over the permanent breakpoint.
gdb_test "continue" "Breakpoint .*" "stop at permanent breakpoint"
gdb_test "p \$prev_counter = counter" " = $decimal"
gdb_test "continue" "Breakpoint .*" "stop at permanent breakpoint twice"
# Check that indeed the continue made progress, instead of re-trapping
# without advancing.
gdb_test "p counter - \$prev_counter" " = 1"
gdb_test "info breakpoints" \
"breakpoint.*keep.*y.*$hex.*in test at .*$srcfile:$line_bp.*already hit 2 times.*" \
"info breakpoints show enabled breakpoint"
gdb_test "disable \$bpnum"
gdb_test "commands\nset \$commands_ran = 1\nend" "" \
"set breakpoint commands"
gdb_test "info breakpoints" \
"breakpoint.*keep.*n.*$hex.*in test at .*$srcfile:$line_bp.*already hit 2 times.*" \
"info breakpoints shows disabled breakpoint"
# Run to the permanent breakpoint again. This time, since it's
# disabled, it should act as if we hadn't created it in the first
# place. IOW, we should get a random signal, and, the breakpoint's
# command should not run.
gdb_test "continue" "Program received signal SIGTRAP.*" \
"disabled permanent breakpoint doesn't explain stop"
gdb_test "info breakpoints" \
"breakpoint.*keep.*n.*$hex.*in test at .*$srcfile:$line_bp.*already hit 2 times.*" \
"info breakpoints still shows same number of hits"
gdb_test "print \$commands_ran" " = void" \
"breakpoint commands didn't run"
# Reenable the breakpoint, and check that it gets hit and accounted
# for this time.
gdb_test "enable \$bpnum" "" "reenable breakpoint"
gdb_test "continue" "Breakpoint .*" \
"stop at permanent breakpoint thrice"
gdb_test "info breakpoints" \
"breakpoint.*keep.*y.*$hex.*in test at .*$srcfile:$line_bp.*already hit 3 times.*" \
"info breakpoints shows one more hit"
gdb_test "print \$commands_ran" " = 1" "breakpoint commands ran"
# Check that stepi advances only past the permanent breakpoint, and
# not a single instruction more.
gdb_test "stepi" "after permanent bp .*" \
"single-step past permanent breakpoint"
}
with_test_prefix "next trips on permanent bp" {
delete_breakpoints
gdb_breakpoint "test_next"
gdb_continue_to_breakpoint "test_next"
gdb_breakpoint "$line_bp"
gdb_test "condition \$bpnum 0"
gdb_test "next" "after next .*"
}
if ![target_info exists gdb,nosignals] {
with_test_prefix "continue trips on nested permanent bp" {
delete_breakpoints
gdb_breakpoint "test_signal_nested"
gdb_continue_to_breakpoint "test_signal_nested"
gdb_breakpoint "$line_bp"
gdb_continue_to_breakpoint "permanent bp"
gdb_test "condition \$bpnum 0"
# Let SIGALRM trigger.
sleep 2
# We're now stopped at a permanent breakpoint, with a
# signal pending.
gdb_breakpoint "test_signal_nested_done"
gdb_continue_to_breakpoint "test_signal_nested_done"
# Ensure that the handler did run. There's one call to
# test in the mainline code, and another in the signal
# handler.
gdb_test "p counter" " = 2"
}
if [can_single_step_to_signal_handler] {
with_test_prefix "stepi signal with handler" {
delete_breakpoints
gdb_breakpoint "test_signal_with_handler"
gdb_continue_to_breakpoint "test_signal_with_handler"
gdb_breakpoint "$line_bp"
gdb_test "continue" "Breakpoint .*" "stop at permanent breakpoint"
gdb_test "queue-signal SIGUSR1"
set test "single-step to handler"
gdb_test_multiple "stepi" $test {
-re "Program received signal SIGTRAP.*$gdb_prompt $" {
fail $test
}
-re "handler .*$gdb_prompt $" {
pass $test
}
}
# Check that the mainline PC points at the permanent
# breakpoint.
gdb_test "up 2" "test .*" "up to mainline code"
gdb_test "p /x \$pc" " = $address_bp" \
"mainline pc points at permanent breakpoint"
gdb_test "continue" "Breakpoint .*" \
"stop at permanent breakpoint, out of handler"
}
with_test_prefix "stepi signal with no handler" {
gdb_breakpoint "test_signal_no_handler"
gdb_continue_to_breakpoint "test_signal_no_handler"
gdb_test "continue" "Breakpoint .*" "stop at permanent breakpoint"
gdb_test "queue-signal SIGUSR1"
gdb_test "stepi" "after permanent bp .*" \
"single-step past permanent breakpoint"
}
}
}
}
foreach always_inserted {off on} {
foreach sw_watchpoint {0 1} {
with_test_prefix "always_inserted=$always_inserted, sw_watchpoint=$sw_watchpoint" {
test $always_inserted $sw_watchpoint
}
}
}