2016-01-01 04:33:14 +00:00
|
|
|
# Copyright 2012-2016 Free Software Foundation, Inc.
|
2012-07-31 07:33:16 +00:00
|
|
|
|
|
|
|
# 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/>.
|
|
|
|
|
|
|
|
if [is_remote target] {
|
|
|
|
# The test always runs locally.
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
test suite update - gdb.base/[t-z]
Convert files gdb.base/[t-z]*.exp to use standard_output_file et al.
* gdb.base/testenv.exp, gdb.base/tui-layout.exp,
gdb.base/twice.exp, gdb.base/type-opaque.exp, gdb.base/unload.exp,
gdb.base/unwindonsignal.exp, gdb.base/valgrind-db-attach.exp,
gdb.base/valgrind-infcall.exp, gdb.base/value-double-free.exp,
gdb.base/varargs.exp, gdb.base/watch-cond-infcall.exp,
gdb.base/watch-cond.exp, gdb.base/watch-non-mem.exp,
gdb.base/watch-read.exp, gdb.base/watch-vfork.exp,
gdb.base/watch_thread_num.exp, gdb.base/watchpoint-cond-gone.exp,
gdb.base/watchpoint-delete.exp, gdb.base/watchpoint-hw.exp,
gdb.base/watchpoint-solib.exp, gdb.base/watchpoint.exp,
gdb.base/watchpoints.exp, gdb.base/wchar.exp, gdb.base/whatis.exp:
Use standard_testfile, standard_output_file, prepare_for_testing,
clean_restart.
2013-06-27 19:00:47 +00:00
|
|
|
standard_testfile .c
|
|
|
|
if {[build_executable $testfile.exp $testfile $srcfile {debug}] == -1} {
|
2012-07-31 07:33:16 +00:00
|
|
|
return -1
|
|
|
|
}
|
|
|
|
|
|
|
|
set test "spawn valgrind"
|
|
|
|
set cmd "valgrind --vgdb-error=0 $binfile"
|
2013-06-07 17:31:09 +00:00
|
|
|
set res [remote_spawn host $cmd]
|
2012-07-31 07:33:16 +00:00
|
|
|
if { $res < 0 || $res == "" } {
|
|
|
|
verbose -log "Spawning $cmd failed."
|
|
|
|
unsupported $test
|
|
|
|
return -1
|
|
|
|
}
|
|
|
|
pass $test
|
|
|
|
# Declare GDB now as running.
|
Introduce gdb_interact in testsuite
gdb_interact is a small utility that we have found quite useful to debug
test cases.
Putting gdb_interact in a test suspends it and allows to interact with
gdb to inspect whatever you want. You can then type ">>>" to resume the
test execution. Of course, this is only for gdb devs. It wouldn't make
sense to leave a gdb_interact permanently in a test case.
When starting the interaction with the user, the script prints this
banner:
+------------------------------------------+
| Script interrupted, you can now interact |
| with by gdb. Type >>> to continue. |
+------------------------------------------+
Notes:
* When gdb is launched, the gdb_spawn_id variable (lib/gdb.exp) is
assigned -1. Given the name, I would expect it to contain the gdb
expect spawn id, which is needed for interact. I changed all places
that set gdb_spawn_id to -1 to set it to the actual gdb spawn id
instead.
* When entering the "interact" mode, the last (gdb) prompt is already
eaten by expect, so it doesn't show up on the terminal. Subsequent
prompts do appear though. We tried to print "(gdb)" just before the
interact to replace it. However, it could be misleading if you are
debugging an MI test case, it makes you think that you are typing in a
CLI prompt, when in reality it's MI. In the end I decided that since
the feature is for developers who know what they're doing and that one
is normally consciously using gdb_interact, the script doesn't need
to babysit the user.
* There are probably some quirks depending on where in the script
gdb_interact appears (e.g. it could interfere with following
commands and make them fail), but it works for most cases. Quirks can
always be fixed later.
The idea and original implementation was contributed by Anders
Granlund, a colleague of mine. Thanks to him.
gdb/testsuite/ChangeLog:
* gdb.base/statistics.exp: Assign spawn id to gdb_spawn_id.
* gdb.base/valgrind-db-attach.exp: Same.
* gdb.base/valgrind-infcall.exp: Same.
* lib/mi-support.exp (default_mi_gdb_start): Same.
* lib/prompt.exp (default_prompt_gdb_start): Same.
* lib/gdb.exp (default_gdb_spawn): Same.
(gdb_interact): New.
2015-01-22 19:33:04 +00:00
|
|
|
set gdb_spawn_id $res
|
2012-07-31 07:33:16 +00:00
|
|
|
|
|
|
|
# GDB started by vgdb stops already after the startup is executed, like with
|
|
|
|
# non-extended gdbserver. It is also not correct to run/attach the inferior.
|
|
|
|
set use_gdb_stub 1
|
|
|
|
|
|
|
|
set test "valgrind started"
|
|
|
|
# The trailing '.' differs for different memcheck versions.
|
|
|
|
gdb_test_multiple "" $test {
|
|
|
|
-re "Memcheck, a memory error detector\\.?\r\n" {
|
|
|
|
pass $test
|
|
|
|
}
|
|
|
|
-re "valgrind: failed to start tool 'memcheck' for platform '.*': No such file or directory" {
|
|
|
|
unsupported $test
|
|
|
|
return -1
|
|
|
|
}
|
|
|
|
-re "valgrind: wrong ELF executable class" {
|
|
|
|
unsupported $test
|
|
|
|
return -1
|
|
|
|
}
|
|
|
|
-re "command not found" {
|
|
|
|
# The spawn succeeded, but then valgrind was not found - e.g. if
|
|
|
|
# we spawned SSH to a remote system.
|
|
|
|
unsupported $test
|
|
|
|
return -1
|
|
|
|
}
|
2012-08-01 09:02:50 +00:00
|
|
|
-re "valgrind: Bad option.*--vgdb-error=0" {
|
2012-07-31 07:33:16 +00:00
|
|
|
# valgrind is not >= 3.7.0.
|
|
|
|
unsupported $test
|
|
|
|
return -1
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
set test "vgdb prompt"
|
|
|
|
# The trailing '.' differs for different memcheck versions.
|
|
|
|
gdb_test_multiple "" $test {
|
|
|
|
-re " (target remote | \[^\r\n\]*/vgdb \[^\r\n\]*)\r\n" {
|
|
|
|
set vgdbcmd $expect_out(1,string)
|
|
|
|
pass $test
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
# Do not kill valgrind.
|
testsuite: tcl exec& -> 'kill -9 $pid' is racy (attach-many-short-lived-thread.exp races and others)
The buildbots show that attach-many-short-lived-thread.exp is racy.
But after staring at debug logs and playing with SystemTap scripts for
a (long) while, I figured out that neither GDB, nor the kernel nor the
test's program itself are at fault.
The problem is simply that the testsuite machinery is currently
subject to PID-reuse races. The attach-many-short-lived-threads.c
test program just happens to be much more susceptible to trigger this
race because threads and processes share the same number space on
Linux, and the test spawns many many short lived threads in
succession, thus enlarging the race window a lot.
Part of the problem is that several tests spawn processes with "exec&"
(in order to test the "attach" command) , and then at the end of the
test, to make sure things are cleaned up, issue a 'remote_spawn "kill
-p $testpid"'. Since with tcl's "exec&", tcl itself is responsible
for reaping the process's exit status, when we go kill the process,
testpid may have already exited _and_ its status may have (and often
has) been reaped already. Thus it can happen that another process
meanwhile reuses $testpid, and that "kill" command kills the wrong
process... Frequently, that happens to be
attach-many-short-lived-thread, but this explains other test's races
as well.
In the attach-many-short-lived-threads test, it sometimes manifests
like this:
(gdb) file /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads...done.
(gdb) Loaded /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads into /home/pedro/gdb/mygit/build/gdb/testsuite/../../gdb/gdb
attach 5940
Attaching to program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads, process 5940
warning: process 5940 is a zombie - the process has already terminated
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ptrace: Operation not permitted.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: attach
info threads
No threads.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: no new threads
set breakpoint always-inserted on
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: set breakpoint always-inserted on
Other times the process dies while the test is ongoing (the process is
ptrace-stopped):
(gdb) print again = 1
Cannot access memory at address 0x6020cc
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: reset timer in the inferior
(Recall that on Linux, SIGKILL is not interceptable)
And other times it dies just while we're detaching:
$4 = 319
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 2: print seconds_left
detach
Can't detach Thread 0x7fb13b7de700 (LWP 1842): No such process
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: detach
GDB mishandles the latter (it should ignore ESRCH while detaching just
like when continuing), but that's another story.
The fix here is to change spawn_wait_for_attach to use Expect's
'spawn' command instead of Tcl's 'exec&' to spawn programs, because
with spawn we control when to wait for/reap the process. That allows
killing the process by PID without being subject to pid-reuse races,
because even if the process is already dead, the kernel won't reuse
the process's PID until the zombie is reaped.
The other part of the problem lies in DejaGnu itself, unfortunately.
I have occasionally seen tests (attach-many-short-lived-threads
included, but not only that one) die with a random inexplicable
SIGTERM too, and that too is caused by the same reason, except that in
that case, the rogue SIGTERM is sent from this bit in DejaGnu's remote.exp:
exec sh -c "exec > /dev/null 2>&1 && (kill -2 $pgid || kill -2 $pid) && sleep 5 && (kill $pgid || kill $pid) && sleep 5 && (kill -9 $pgid || kill -9 $pid) &"
...
catch "wait -i $shell_id"
Even if the program exits promptly, that whole cascade of kills
carries on in the background, thus potentially killing the poor
process that manages to reuse $pid...
I sent a fix for that to the DejaGnu list:
http://lists.gnu.org/archive/html/dejagnu/2015-07/msg00000.html
With both patches in place, I haven't seen
attach-many-short-lived-threads.exp fail again.
Tested on x86_64 Fedora 20, native, gdbserver and extended-gdbserver.
gdb/testsuite/ChangeLog:
2015-07-31 Pedro Alves <palves@redhat.com>
* gdb.base/attach-pie-misread.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* gdb.base/attach-pie-noexec.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/attach-twice.exp: Likewise.
* gdb.base/attach.exp: Likewise.
(do_command_attach_tests): Use gdb_spawn_with_cmdline_opts and
gdb_test_multiple.
* gdb.base/solib-overlap.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/valgrind-infcall.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
* gdb.threads/attach-into-signal.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.threads/attach-many-short-lived-threads.exp: Adjust to
spawn_wait_for_attach returning a spawn id instead of a pid. Use
spawn_id_get_pid and kill_wait_spawned_process.
* gdb.threads/attach-stopped.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/break-interp.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* lib/gdb.exp (can_spawn_for_attach): Adjust comment.
(kill_wait_spawned_process, spawn_id_get_pid): New procedures.
(spawn_wait_for_attach): Use spawn instead of exec to spawn
processes. Don't map cygwin/windows pids here. Now returns a
spawn id list.
2015-07-31 19:06:24 +00:00
|
|
|
set valgrind_spawn_id [board_info host fileid]
|
2012-07-31 07:33:16 +00:00
|
|
|
unset gdb_spawn_id
|
|
|
|
set board [host_info name]
|
|
|
|
unset_board_info fileid
|
|
|
|
|
test suite update - gdb.base/[t-z]
Convert files gdb.base/[t-z]*.exp to use standard_output_file et al.
* gdb.base/testenv.exp, gdb.base/tui-layout.exp,
gdb.base/twice.exp, gdb.base/type-opaque.exp, gdb.base/unload.exp,
gdb.base/unwindonsignal.exp, gdb.base/valgrind-db-attach.exp,
gdb.base/valgrind-infcall.exp, gdb.base/value-double-free.exp,
gdb.base/varargs.exp, gdb.base/watch-cond-infcall.exp,
gdb.base/watch-cond.exp, gdb.base/watch-non-mem.exp,
gdb.base/watch-read.exp, gdb.base/watch-vfork.exp,
gdb.base/watch_thread_num.exp, gdb.base/watchpoint-cond-gone.exp,
gdb.base/watchpoint-delete.exp, gdb.base/watchpoint-hw.exp,
gdb.base/watchpoint-solib.exp, gdb.base/watchpoint.exp,
gdb.base/watchpoints.exp, gdb.base/wchar.exp, gdb.base/whatis.exp:
Use standard_testfile, standard_output_file, prepare_for_testing,
clean_restart.
2013-06-27 19:00:47 +00:00
|
|
|
clean_restart $testfile
|
2012-07-31 07:33:16 +00:00
|
|
|
|
2015-08-06 22:38:51 +00:00
|
|
|
# Make sure we're disconnected, in case we're testing with the
|
|
|
|
# native-extended-gdbserver board, where gdb_start/gdb_load spawn
|
|
|
|
# gdbserver and connect to it.
|
|
|
|
gdb_test "disconnect" ".*"
|
|
|
|
|
2012-08-03 16:03:07 +00:00
|
|
|
gdb_test "$vgdbcmd" " in \\.?_start .*" "target remote for vgdb"
|
2012-07-31 07:33:16 +00:00
|
|
|
|
|
|
|
gdb_test "monitor v.set gdb_output" "valgrind output will go to gdb.*"
|
|
|
|
|
|
|
|
set continue_count 1
|
2012-08-07 18:04:12 +00:00
|
|
|
set loop 1
|
|
|
|
while {$loop && $continue_count < 100} {
|
2012-07-31 07:33:16 +00:00
|
|
|
set test "continue #$continue_count"
|
|
|
|
gdb_test_multiple "continue" "" {
|
|
|
|
-re "Invalid free\\(\\).*: main .*\r\n$gdb_prompt $" {
|
|
|
|
pass $test
|
2012-08-07 18:04:12 +00:00
|
|
|
set loop 0
|
2012-07-31 07:33:16 +00:00
|
|
|
}
|
2012-08-03 16:03:07 +00:00
|
|
|
-re "Remote connection closed.*\r\n$gdb_prompt $" {
|
|
|
|
fail "$test (remote connection closed)"
|
2013-03-17 20:37:32 +00:00
|
|
|
# Only if valgrind got stuck.
|
testsuite: tcl exec& -> 'kill -9 $pid' is racy (attach-many-short-lived-thread.exp races and others)
The buildbots show that attach-many-short-lived-thread.exp is racy.
But after staring at debug logs and playing with SystemTap scripts for
a (long) while, I figured out that neither GDB, nor the kernel nor the
test's program itself are at fault.
The problem is simply that the testsuite machinery is currently
subject to PID-reuse races. The attach-many-short-lived-threads.c
test program just happens to be much more susceptible to trigger this
race because threads and processes share the same number space on
Linux, and the test spawns many many short lived threads in
succession, thus enlarging the race window a lot.
Part of the problem is that several tests spawn processes with "exec&"
(in order to test the "attach" command) , and then at the end of the
test, to make sure things are cleaned up, issue a 'remote_spawn "kill
-p $testpid"'. Since with tcl's "exec&", tcl itself is responsible
for reaping the process's exit status, when we go kill the process,
testpid may have already exited _and_ its status may have (and often
has) been reaped already. Thus it can happen that another process
meanwhile reuses $testpid, and that "kill" command kills the wrong
process... Frequently, that happens to be
attach-many-short-lived-thread, but this explains other test's races
as well.
In the attach-many-short-lived-threads test, it sometimes manifests
like this:
(gdb) file /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads...done.
(gdb) Loaded /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads into /home/pedro/gdb/mygit/build/gdb/testsuite/../../gdb/gdb
attach 5940
Attaching to program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads, process 5940
warning: process 5940 is a zombie - the process has already terminated
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ptrace: Operation not permitted.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: attach
info threads
No threads.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: no new threads
set breakpoint always-inserted on
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: set breakpoint always-inserted on
Other times the process dies while the test is ongoing (the process is
ptrace-stopped):
(gdb) print again = 1
Cannot access memory at address 0x6020cc
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: reset timer in the inferior
(Recall that on Linux, SIGKILL is not interceptable)
And other times it dies just while we're detaching:
$4 = 319
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 2: print seconds_left
detach
Can't detach Thread 0x7fb13b7de700 (LWP 1842): No such process
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: detach
GDB mishandles the latter (it should ignore ESRCH while detaching just
like when continuing), but that's another story.
The fix here is to change spawn_wait_for_attach to use Expect's
'spawn' command instead of Tcl's 'exec&' to spawn programs, because
with spawn we control when to wait for/reap the process. That allows
killing the process by PID without being subject to pid-reuse races,
because even if the process is already dead, the kernel won't reuse
the process's PID until the zombie is reaped.
The other part of the problem lies in DejaGnu itself, unfortunately.
I have occasionally seen tests (attach-many-short-lived-threads
included, but not only that one) die with a random inexplicable
SIGTERM too, and that too is caused by the same reason, except that in
that case, the rogue SIGTERM is sent from this bit in DejaGnu's remote.exp:
exec sh -c "exec > /dev/null 2>&1 && (kill -2 $pgid || kill -2 $pid) && sleep 5 && (kill $pgid || kill $pid) && sleep 5 && (kill -9 $pgid || kill -9 $pid) &"
...
catch "wait -i $shell_id"
Even if the program exits promptly, that whole cascade of kills
carries on in the background, thus potentially killing the poor
process that manages to reuse $pid...
I sent a fix for that to the DejaGnu list:
http://lists.gnu.org/archive/html/dejagnu/2015-07/msg00000.html
With both patches in place, I haven't seen
attach-many-short-lived-threads.exp fail again.
Tested on x86_64 Fedora 20, native, gdbserver and extended-gdbserver.
gdb/testsuite/ChangeLog:
2015-07-31 Pedro Alves <palves@redhat.com>
* gdb.base/attach-pie-misread.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* gdb.base/attach-pie-noexec.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/attach-twice.exp: Likewise.
* gdb.base/attach.exp: Likewise.
(do_command_attach_tests): Use gdb_spawn_with_cmdline_opts and
gdb_test_multiple.
* gdb.base/solib-overlap.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/valgrind-infcall.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
* gdb.threads/attach-into-signal.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.threads/attach-many-short-lived-threads.exp: Adjust to
spawn_wait_for_attach returning a spawn id instead of a pid. Use
spawn_id_get_pid and kill_wait_spawned_process.
* gdb.threads/attach-stopped.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/break-interp.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* lib/gdb.exp (can_spawn_for_attach): Adjust comment.
(kill_wait_spawned_process, spawn_id_get_pid): New procedures.
(spawn_wait_for_attach): Use spawn instead of exec to spawn
processes. Don't map cygwin/windows pids here. Now returns a
spawn id list.
2015-07-31 19:06:24 +00:00
|
|
|
kill_wait_spawned_process $valgrind_spawn_id
|
2012-08-03 16:03:07 +00:00
|
|
|
return -1
|
|
|
|
}
|
2012-08-07 18:04:12 +00:00
|
|
|
-re "The program is not being run\\.\r\n$gdb_prompt $" {
|
|
|
|
fail "$test (valgrind vgdb has terminated)"
|
2013-03-17 20:37:32 +00:00
|
|
|
# Only if valgrind got stuck.
|
testsuite: tcl exec& -> 'kill -9 $pid' is racy (attach-many-short-lived-thread.exp races and others)
The buildbots show that attach-many-short-lived-thread.exp is racy.
But after staring at debug logs and playing with SystemTap scripts for
a (long) while, I figured out that neither GDB, nor the kernel nor the
test's program itself are at fault.
The problem is simply that the testsuite machinery is currently
subject to PID-reuse races. The attach-many-short-lived-threads.c
test program just happens to be much more susceptible to trigger this
race because threads and processes share the same number space on
Linux, and the test spawns many many short lived threads in
succession, thus enlarging the race window a lot.
Part of the problem is that several tests spawn processes with "exec&"
(in order to test the "attach" command) , and then at the end of the
test, to make sure things are cleaned up, issue a 'remote_spawn "kill
-p $testpid"'. Since with tcl's "exec&", tcl itself is responsible
for reaping the process's exit status, when we go kill the process,
testpid may have already exited _and_ its status may have (and often
has) been reaped already. Thus it can happen that another process
meanwhile reuses $testpid, and that "kill" command kills the wrong
process... Frequently, that happens to be
attach-many-short-lived-thread, but this explains other test's races
as well.
In the attach-many-short-lived-threads test, it sometimes manifests
like this:
(gdb) file /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads...done.
(gdb) Loaded /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads into /home/pedro/gdb/mygit/build/gdb/testsuite/../../gdb/gdb
attach 5940
Attaching to program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads, process 5940
warning: process 5940 is a zombie - the process has already terminated
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ptrace: Operation not permitted.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: attach
info threads
No threads.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: no new threads
set breakpoint always-inserted on
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: set breakpoint always-inserted on
Other times the process dies while the test is ongoing (the process is
ptrace-stopped):
(gdb) print again = 1
Cannot access memory at address 0x6020cc
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: reset timer in the inferior
(Recall that on Linux, SIGKILL is not interceptable)
And other times it dies just while we're detaching:
$4 = 319
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 2: print seconds_left
detach
Can't detach Thread 0x7fb13b7de700 (LWP 1842): No such process
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: detach
GDB mishandles the latter (it should ignore ESRCH while detaching just
like when continuing), but that's another story.
The fix here is to change spawn_wait_for_attach to use Expect's
'spawn' command instead of Tcl's 'exec&' to spawn programs, because
with spawn we control when to wait for/reap the process. That allows
killing the process by PID without being subject to pid-reuse races,
because even if the process is already dead, the kernel won't reuse
the process's PID until the zombie is reaped.
The other part of the problem lies in DejaGnu itself, unfortunately.
I have occasionally seen tests (attach-many-short-lived-threads
included, but not only that one) die with a random inexplicable
SIGTERM too, and that too is caused by the same reason, except that in
that case, the rogue SIGTERM is sent from this bit in DejaGnu's remote.exp:
exec sh -c "exec > /dev/null 2>&1 && (kill -2 $pgid || kill -2 $pid) && sleep 5 && (kill $pgid || kill $pid) && sleep 5 && (kill -9 $pgid || kill -9 $pid) &"
...
catch "wait -i $shell_id"
Even if the program exits promptly, that whole cascade of kills
carries on in the background, thus potentially killing the poor
process that manages to reuse $pid...
I sent a fix for that to the DejaGnu list:
http://lists.gnu.org/archive/html/dejagnu/2015-07/msg00000.html
With both patches in place, I haven't seen
attach-many-short-lived-threads.exp fail again.
Tested on x86_64 Fedora 20, native, gdbserver and extended-gdbserver.
gdb/testsuite/ChangeLog:
2015-07-31 Pedro Alves <palves@redhat.com>
* gdb.base/attach-pie-misread.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* gdb.base/attach-pie-noexec.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/attach-twice.exp: Likewise.
* gdb.base/attach.exp: Likewise.
(do_command_attach_tests): Use gdb_spawn_with_cmdline_opts and
gdb_test_multiple.
* gdb.base/solib-overlap.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/valgrind-infcall.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
* gdb.threads/attach-into-signal.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.threads/attach-many-short-lived-threads.exp: Adjust to
spawn_wait_for_attach returning a spawn id instead of a pid. Use
spawn_id_get_pid and kill_wait_spawned_process.
* gdb.threads/attach-stopped.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/break-interp.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* lib/gdb.exp (can_spawn_for_attach): Adjust comment.
(kill_wait_spawned_process, spawn_id_get_pid): New procedures.
(spawn_wait_for_attach): Use spawn instead of exec to spawn
processes. Don't map cygwin/windows pids here. Now returns a
spawn id list.
2015-07-31 19:06:24 +00:00
|
|
|
kill_wait_spawned_process $valgrind_spawn_id
|
2012-08-07 18:04:12 +00:00
|
|
|
return -1
|
|
|
|
}
|
2012-07-31 07:33:16 +00:00
|
|
|
-re "\r\n$gdb_prompt $" {
|
|
|
|
pass "$test (false warning)"
|
|
|
|
}
|
|
|
|
}
|
|
|
|
set continue_count [expr $continue_count + 1]
|
|
|
|
}
|
|
|
|
|
|
|
|
set test "p gdb_test_infcall ()"
|
|
|
|
gdb_test_multiple $test $test {
|
|
|
|
-re "unhandled instruction bytes.*\r\n$gdb_prompt $" {
|
|
|
|
fail $test
|
|
|
|
}
|
|
|
|
-re "Continuing \\.\\.\\..*\r\n\\\$1 = 2\r\n$gdb_prompt $" {
|
|
|
|
pass $test
|
|
|
|
}
|
|
|
|
}
|
2013-02-27 18:46:52 +00:00
|
|
|
|
|
|
|
# Only if valgrind got stuck.
|
testsuite: tcl exec& -> 'kill -9 $pid' is racy (attach-many-short-lived-thread.exp races and others)
The buildbots show that attach-many-short-lived-thread.exp is racy.
But after staring at debug logs and playing with SystemTap scripts for
a (long) while, I figured out that neither GDB, nor the kernel nor the
test's program itself are at fault.
The problem is simply that the testsuite machinery is currently
subject to PID-reuse races. The attach-many-short-lived-threads.c
test program just happens to be much more susceptible to trigger this
race because threads and processes share the same number space on
Linux, and the test spawns many many short lived threads in
succession, thus enlarging the race window a lot.
Part of the problem is that several tests spawn processes with "exec&"
(in order to test the "attach" command) , and then at the end of the
test, to make sure things are cleaned up, issue a 'remote_spawn "kill
-p $testpid"'. Since with tcl's "exec&", tcl itself is responsible
for reaping the process's exit status, when we go kill the process,
testpid may have already exited _and_ its status may have (and often
has) been reaped already. Thus it can happen that another process
meanwhile reuses $testpid, and that "kill" command kills the wrong
process... Frequently, that happens to be
attach-many-short-lived-thread, but this explains other test's races
as well.
In the attach-many-short-lived-threads test, it sometimes manifests
like this:
(gdb) file /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads
Reading symbols from /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads...done.
(gdb) Loaded /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads into /home/pedro/gdb/mygit/build/gdb/testsuite/../../gdb/gdb
attach 5940
Attaching to program: /home/pedro/gdb/mygit/build/gdb/testsuite/gdb.threads/attach-many-short-lived-threads, process 5940
warning: process 5940 is a zombie - the process has already terminated
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
ptrace: Operation not permitted.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: attach
info threads
No threads.
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: no new threads
set breakpoint always-inserted on
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 1: set breakpoint always-inserted on
Other times the process dies while the test is ongoing (the process is
ptrace-stopped):
(gdb) print again = 1
Cannot access memory at address 0x6020cc
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: reset timer in the inferior
(Recall that on Linux, SIGKILL is not interceptable)
And other times it dies just while we're detaching:
$4 = 319
(gdb) PASS: gdb.threads/attach-many-short-lived-threads.exp: iter 2: print seconds_left
detach
Can't detach Thread 0x7fb13b7de700 (LWP 1842): No such process
(gdb) FAIL: gdb.threads/attach-many-short-lived-threads.exp: iter 2: detach
GDB mishandles the latter (it should ignore ESRCH while detaching just
like when continuing), but that's another story.
The fix here is to change spawn_wait_for_attach to use Expect's
'spawn' command instead of Tcl's 'exec&' to spawn programs, because
with spawn we control when to wait for/reap the process. That allows
killing the process by PID without being subject to pid-reuse races,
because even if the process is already dead, the kernel won't reuse
the process's PID until the zombie is reaped.
The other part of the problem lies in DejaGnu itself, unfortunately.
I have occasionally seen tests (attach-many-short-lived-threads
included, but not only that one) die with a random inexplicable
SIGTERM too, and that too is caused by the same reason, except that in
that case, the rogue SIGTERM is sent from this bit in DejaGnu's remote.exp:
exec sh -c "exec > /dev/null 2>&1 && (kill -2 $pgid || kill -2 $pid) && sleep 5 && (kill $pgid || kill $pid) && sleep 5 && (kill -9 $pgid || kill -9 $pid) &"
...
catch "wait -i $shell_id"
Even if the program exits promptly, that whole cascade of kills
carries on in the background, thus potentially killing the poor
process that manages to reuse $pid...
I sent a fix for that to the DejaGnu list:
http://lists.gnu.org/archive/html/dejagnu/2015-07/msg00000.html
With both patches in place, I haven't seen
attach-many-short-lived-threads.exp fail again.
Tested on x86_64 Fedora 20, native, gdbserver and extended-gdbserver.
gdb/testsuite/ChangeLog:
2015-07-31 Pedro Alves <palves@redhat.com>
* gdb.base/attach-pie-misread.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* gdb.base/attach-pie-noexec.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/attach-twice.exp: Likewise.
* gdb.base/attach.exp: Likewise.
(do_command_attach_tests): Use gdb_spawn_with_cmdline_opts and
gdb_test_multiple.
* gdb.base/solib-overlap.exp: Adjust to spawn_wait_for_attach
returning a spawn id instead of a pid. Use spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/valgrind-infcall.exp: Likewise.
* gdb.multi/multi-attach.exp: Likewise.
* gdb.python/py-prompt.exp: Likewise.
* gdb.python/py-sync-interp.exp: Likewise.
* gdb.server/ext-attach.exp: Likewise.
* gdb.threads/attach-into-signal.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.threads/attach-many-short-lived-threads.exp: Adjust to
spawn_wait_for_attach returning a spawn id instead of a pid. Use
spawn_id_get_pid and kill_wait_spawned_process.
* gdb.threads/attach-stopped.exp (corefunc): Use
spawn_wait_for_attach, spawn_id_get_pid and
kill_wait_spawned_process.
* gdb.base/break-interp.exp: Rename $res to $test_spawn_id.
Use spawn_id_get_pid. Wait for spawn id after eof. Use
kill_wait_spawned_process instead of explicit "kill -9".
* lib/gdb.exp (can_spawn_for_attach): Adjust comment.
(kill_wait_spawned_process, spawn_id_get_pid): New procedures.
(spawn_wait_for_attach): Use spawn instead of exec to spawn
processes. Don't map cygwin/windows pids here. Now returns a
spawn id list.
2015-07-31 19:06:24 +00:00
|
|
|
kill_wait_spawned_process $valgrind_spawn_id
|