[v3,2/2] Add dprintf and detach test (PR breakpoints/17012)
Commit Message
This adds a test to demonstrate PR 17012, where adding a dprintf in a
linux native process and detaching leaves the trap instruction in the
process.
One issue is that the test does not pass when running with
--target_board=native-extended-gdbserver, partly due to bug 17302 [1].
One quick hack I tried for this was to add a useless "next" between the
call to getpid() and detach, which avoids the bug. There is still one
case where the test fails, and that is with:
- breakpoint always-inserted on
- dprintf-style agent
- disconnected-dprintf on
What happens is that my detach does not actually detach the process,
because some persistent commands (the disconnected dprintf) is present.
However since gdbserver is ran with --once, when gdb disconnects,
gdbserver goes down and takes with it all the processes it spawned and
that are still under its control (which includes my test process).
When the test checks if the test process is still alive, it obvisouly
fails. Investigating about that led me to ask a question on the ML [2]
about the behavior of detach.
Until the remote case is sorted out, would it be ok to disable the
test for all remote targets or KFAIL the tests? I will tackle the
blocking problems in separate patches.
[1] https://sourceware.org/bugzilla/show_bug.cgi?id=17302
[2] https://sourceware.org/ml/gdb/2014-08/msg00115.html
New in v3:
* Follow-up Pedro's review
* Fix wrong include and formatting in test program.
* Remove check to run the test only on Linux.
* Disable test for stubs instead of all remote targets
* Test with all cominations of dprintf-style, breakpoint
always-inserted and disconnected-dprintf.
* Check if "call getpid ()" succeeds.
* Clean restart gdb when re-attaching,
New in v2:
* Verify process existence by trying to attach it with a new gdb.
* Added/updated copyright notices.
gdb/testsuite/ChangeLog:
PR breakpoints/17012
gdb.base/dprintf-detach.c: New file.
gdb.base/dprintf-detach.exp: New file.
---
gdb/testsuite/gdb.base/dprintf-detach.c | 33 ++++++++++++
gdb/testsuite/gdb.base/dprintf-detach.exp | 87 +++++++++++++++++++++++++++++++
2 files changed, 120 insertions(+)
create mode 100644 gdb/testsuite/gdb.base/dprintf-detach.c
create mode 100644 gdb/testsuite/gdb.base/dprintf-detach.exp
Comments
On 08/22/2014 08:05 PM, Simon Marchi wrote:
> Until the remote case is sorted out, would it be ok to disable the
> test for all remote targets or KFAIL the tests? I will tackle the
> blocking problems in separate patches.
Sure.
This is OK otherwise.
Though please push the test and fix as a single commit.
Thanks,
Pedro Alves
On 14-09-04 12:45 PM, Pedro Alves wrote:
> On 08/22/2014 08:05 PM, Simon Marchi wrote:
>
>
>> Until the remote case is sorted out, would it be ok to disable the
>> test for all remote targets or KFAIL the tests? I will tackle the
>> blocking problems in separate patches.
>
> Sure.
I am not sure how to do this though. Is this ok?
# Disable this test when using gdb for now (the inferior crashes when
# detaching due to PR 17302).
set test "re-attach to inferior"
if ![gdb_is_target_remote] {
# Give some time for the ex-inferior to run and hopefully not crash.
sleep 1
gdb_test "attach $inferior_pid" "Attaching to program: $escapedbinfile, process $inferior_pid.*Reading symbols from.*" "$test"
} else {
kfail "breakpoints/17302" "$test"
}
I know that "remote" does not necessarily means gdbserver, but I don't
know how to check for gdbserver specifically. I put the sleep in the condition
so we don't waste time for nothing if we are not going to run the test anyway.
> This is OK otherwise.
>
> Though please push the test and fix as a single commit.
>
> Thanks,
> Pedro Alves
>
On 09/04/2014 07:37 PM, Simon Marchi wrote:
> On 14-09-04 12:45 PM, Pedro Alves wrote:
>> On 08/22/2014 08:05 PM, Simon Marchi wrote:
>>
>>
>>> Until the remote case is sorted out, would it be ok to disable the
>>> test for all remote targets or KFAIL the tests? I will tackle the
>>> blocking problems in separate patches.
>>
>> Sure.
>
> I am not sure how to do this though. Is this ok?
Looks like it. Alternatively, you can use setup_kfail
and always do the re-attach.
>
> # Disable this test when using gdb for now (the inferior crashes when
> # detaching due to PR 17302).
> set test "re-attach to inferior"
> if ![gdb_is_target_remote] {
> # Give some time for the ex-inferior to run and hopefully not crash.
> sleep 1
>
> gdb_test "attach $inferior_pid" "Attaching to program: $escapedbinfile, process $inferior_pid.*Reading symbols from.*" "$test"
> } else {
> kfail "breakpoints/17302" "$test"
"server/17302"
> }
>
> I know that "remote" does not necessarily means gdbserver, but I don't
> know how to check for gdbserver specifically.
There's a check in qtro.exp you could reuse. See "probe for GDBserver".
Thanks,
Pedro Alves
new file mode 100644
@@ -0,0 +1,33 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+ Copyright 2014 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/>. */
+
+#include <unistd.h>
+
+static void
+function (void)
+{
+ sleep (1);
+}
+
+int
+main (void)
+{
+ int i;
+
+ for (i = 0; i < 30; i++)
+ function ();
+}
new file mode 100644
@@ -0,0 +1,87 @@
+# Copyright 2014 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 test checks that inserting a dprintf and detaching does not crash
+# the program.
+#
+# Related bug: https://sourceware.org/bugzilla/show_bug.cgi?id=17012
+
+# The test relies on "detach/attach".
+if { [target_info exists use_gdb_stub] } then {
+ return 0
+}
+
+standard_testfile
+set escapedbinfile [string_to_regexp ${binfile}]
+
+if [prepare_for_testing "failed to prepare for dprintf-detach" \
+ ${testfile} ${srcfile} {debug}] {
+ return -1
+}
+
+proc dprintf_detach_test { breakpoint_always_inserted dprintf_style disconnected_dprintf } {
+ set test_prefix "bai=${breakpoint_always_inserted} ds=${dprintf_style} dd=${disconnected_dprintf}"
+ global binfile decimal gdb_prompt escapedbinfile
+
+ with_test_prefix "$test_prefix" {
+ # Start with a clean gdb
+ clean_restart ${binfile}
+
+ gdb_test_no_output "set breakpoint always-inserted ${breakpoint_always_inserted}"
+ gdb_test_no_output "set dprintf-style ${dprintf_style}"
+ gdb_test_no_output "set disconnected-dprintf ${disconnected_dprintf}"
+
+ if ![runto_main] {
+ fail "Can't run to main"
+ return -1
+ }
+
+ # Get PID of test program.
+ set inferior_pid -1
+ set test "get inferior process ID"
+ gdb_test_multiple "call getpid ()" $test {
+ -re ".* = ($decimal).*$gdb_prompt $" {
+ set inferior_pid $expect_out(1,string)
+ pass $test
+ }
+ }
+
+ if {$inferior_pid == -1} {
+ return
+ }
+
+ # Add a dprintf and detach.
+ gdb_test "dprintf function, \"hello\"" "Dprintf .*" "dprintf insertion"
+ gdb_test "detach" "Detaching from program: .*$escapedbinfile, .*" "detach program"
+
+ gdb_exit
+
+ # Give some time for the ex-inferior to run and hopefully not crash.
+ sleep 1
+
+ # Check that the process still exists by attaching a new gdb to it.
+ clean_restart ${binfile}
+
+ gdb_test "attach $inferior_pid" "Attaching to program: $escapedbinfile, process $inferior_pid.*Reading symbols from.*" "re-attach to inferior"
+ }
+}
+
+foreach breakpoint_always_inserted { "on" "off" } {
+ foreach dprintf_style { "gdb" "call" "agent" } {
+ foreach disconnected_dprintf { "on" "off" } {
+ dprintf_detach_test $breakpoint_always_inserted $dprintf_style $disconnected_dprintf
+ }
+ }
+}