binutils-gdb/gdb/testsuite/gdb.base/interrupt.exp
Pedro Alves 4ef9fb2f07 Fix gdb.base/interrupt.exp racy fail against gdbserver
Currently, against gdbserver, interrupt.exp occasionaly fails like
this:

 ERROR: Process no longer exists
 UNRESOLVED: gdb.base/interrupt.exp: send end of file

The problem is that we see gdbserver exiting before we match gdb's
output:

 expect: does "\r\n\r\nChild exited with status 0\r\nGDBserver exiting\r\n" (spawn_id exp8) match regular expression "end of file"? Gate "end of file"? gate=no
 expect: read eof
 expect: set expect_out(spawn_id) "exp8"
 expect: set expect_out(buffer) "\r\n\r\nChild exited with status 0\r\nGDBserver exiting\r\n"

Fix this by removing $inferior_spawn_id from the set of spawn ids
expect is watching as soon as we see the "end of file" string out of
the inferior spawn id, using an indirect spawn id list.

Tested on x86-64 Fedora 20, native and gdbserver (both target remote
and extended-remote).

gdb/testsuite/ChangeLog:
2015-04-23  Pedro Alves  <palves@redhat.com>

	* gdb.base/interrupt.exp: Use an indirect spawn id list holding
	$inferior_spawn_id instead of $inferior_spawn_id directly.  On
	"end of file", remove $inferior_spawn_id from the indirect list.
2015-04-23 15:48:27 +01:00

239 lines
6.7 KiB
Plaintext

# Copyright 1994-2015 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/>.
if [target_info exists gdb,nointerrupts] {
verbose "Skipping interrupt.exp because of nointerrupts."
continue
}
standard_testfile
set options { debug }
if { ! [target_info exists gdb,nosignals] } {
lappend options "additional_flags=-DSIGNALS"
}
if {[build_executable $testfile.exp $testfile $srcfile $options] == -1} {
untested $testfile.exp
return -1
}
gdb_start
if ![file exists $binfile] then {
perror "$binfile does not exist."
return 0
} else {
gdb_reinitialize_dir $srcdir/$subdir
gdb_load $binfile
# Hope this is unix :-)
gdb_test "shell stty intr '^C'" ".*" \
"set interrupt character in interrupt.exp"
if [runto_main] then {
global inferior_spawn_id gdb_spawn_id
if {[target_info exists gdb,noinferiorio] && $inferior_spawn_id == $gdb_spawn_id} {
verbose "Skipping interrupt.exp because of noinferiorio."
return
}
set msg "process is alive"
gdb_test_multiple "continue" $msg {
-i "$inferior_spawn_id" -re "talk to me baby\r\n" {
pass $msg
}
}
# This should appear twice, once for the echo and once for the
# program's output.
set msg "child process ate our char"
send_inferior "a\n"
gdb_test_multiple "" $msg {
-i "$inferior_spawn_id" -re "^a\r\na\r\n$" {
pass $msg
}
}
# Wait until the program is in the read system call again.
sleep 2
# Cntrl-c may fail for simulator targets running on a BSD host.
# This is the result of a combination of the read syscall
# being restarted and gdb capturing the cntrl-c signal.
# Cntrl-c may fail for simulator targets on slow hosts.
# This is because there is a race condition between entering
# the read and delivering the cntrl-c.
send_gdb "\003"
set msg "send_gdb control C"
gdb_test_multiple "" $msg {
-re "Program received signal SIGINT.*$gdb_prompt $" {
pass $msg
}
}
set msg "call function when asleep"
send_gdb "p func1 ()\n"
gdb_test_multiple "" $msg {
-re " = 4.*$gdb_prompt $" {
pass $msg
}
-re ".*Program received signal SIG(SEGV|ILL).*$gdb_prompt $" {
setup_xfail "i*86-pc-linux*-gnu*"
fail "child died when we called func1, skipped rest of tests"
return
}
-re "$gdb_prompt $" {
# On HPUX-11.0 'send "p func1 ()"' above
# terminates the program. A defect is pending on this
# issue [defect #DTS CHFts24203]. Hence calling setup_xfail
# below.
setup_xfail "hppa*-*-*11*" CHFts24203
fail "call function when asleep (wrong output)"
}
default {
# This fail probably happens whenever we use /proc (we
# don't use PRSABORT), but apparently also happens on
# other machines as well.
setup_xfail "sparc*-*-solaris2*"
setup_xfail "i*86-*-solaris2*"
setup_xfail "*-*-sysv4*"
setup_xfail "vax-*-*"
setup_xfail "alpha-*-*"
setup_xfail "*-*-*bsd*"
setup_xfail "*-*-hpux*"
setup_xfail "*-*-*lynx*"
fail "$msg (stays asleep)"
# Send the inferior a newline to wake it up.
send_inferior "\n"
gdb_test "" " = 4" "call function after waking it"
}
}
# Now try calling the function again.
gdb_test "p func1 ()" " = 4" "call function a second time"
# And the program should still be doing the same thing.
# The optional trailing \r\n is in case we sent a newline above
# to wake the program, in which case the program now sends it
# back. We check for it either here or in the next gdb_expect
# command, because which one it ends up in is timing dependent.
send_gdb "continue\n"
# For some reason, i386-*-sysv4 gdb fails to issue the Continuing
# message, but otherwise appears normal (FIXME).
set msg "continue"
gdb_test_multiple "" "$msg" {
-re "^continue\r\nContinuing.\r\n(\r\n|)$" {
pass $msg
}
-re "^continue\r\n\r\n" {
fail "$msg (missing Continuing.)"
}
}
send_inferior "data\n"
# The optional leading \r\n is in case we sent a newline above
# to wake the program, in which case the program now sends it
# back.
# FIXME: The pattern below leads to an expected success on HPUX-11.0
# but the success is spurious. Need to provide the right reg.expr.
# here.
set msg "echo data"
gdb_test_multiple "" $msg {
-i "$inferior_spawn_id" -re "^(\r\n|)data\r\ndata\r\n$" {
pass $msg
}
-i "$gdb_spawn_id" -re "Undefined command.*$gdb_prompt " {
fail $msg
}
}
if { ! [target_info exists gdb,nosignals] } {
# Wait until the program is in the read system call again.
sleep 2
# Stop the program for another test.
set msg "Send Control-C, second time"
send_gdb "\003"
gdb_test_multiple "" "$msg" {
-re "Program received signal SIGINT.*$gdb_prompt $" {
pass "$msg"
}
}
# The "signal" command should deliver the correct signal and
# return to the loop.
set msg "signal SIGINT"
gdb_test_multiple "signal SIGINT" "$msg" {
-re "^signal SIGINT\r\nContinuing with signal SIGINT.\r\n(\r\n|)$" {
pass "$msg"
}
}
# We should be back in the loop.
send_inferior "more data\n"
set msg "echo more data"
gdb_test_multiple "" $msg {
-i "$inferior_spawn_id" -re "^(\r\n|)more data\r\nmore data\r\n$" {
pass $msg
}
}
}
set saw_end_of_file 0
set saw_inferior_exit 0
set msg "send end of file"
send_inferior "\004"
set spawn_list "$inferior_spawn_id"
gdb_test_multiple "" $msg {
-i spawn_list -re "end of file" {
set saw_end_of_file 1
verbose -log "saw \"end of file\""
if {!$saw_inferior_exit} {
# When $inferior_spawn_id != $gdb_spawn_id, such
# as when testing with gdbserver, we may see the
# eof (the process exit, not the string just
# matched) for $inferior_spawn_id before the
# expected gdb output. Clear this so we no longer
# expect anything out of $inferior_spawn_id.
set spawn_list ""
exp_continue
}
}
-i "$gdb_spawn_id" -re "$inferior_exited_re normally.*$gdb_prompt " {
set saw_inferior_exit 1
verbose -log "saw inferior exit"
if {!$saw_end_of_file} {
exp_continue
}
}
}
gdb_assert { $saw_end_of_file && $saw_inferior_exit } $msg
}
}
return 0