From: Markus Metzger <markus.t.metzger@intel.com>
To: gdb-patches@sourceware.org
Cc: palves@redhat.com, Simon Marchi <simon.marchi@ericsson.com>
Subject: [PATCH v3 2/5] btrace: allow recording to be started (and stopped) for running threads
Date: Mon, 30 Jan 2017 10:06:00 -0000 [thread overview]
Message-ID: <1485770743-6603-3-git-send-email-markus.t.metzger@intel.com> (raw)
In-Reply-To: <1485770743-6603-1-git-send-email-markus.t.metzger@intel.com>
When recording is started for a running thread, GDB was able to start tracing
but then failed to read registers to insert the initial entry for the current
PC. We don't really need that initial entry if we don't know where exactly we
started recording. Skip that step to allow recording to be started while
threads are running.
If we do run into errors, we need to undo the tracing enable to not leak this
thread. The operation did not complete so our caller won't clean up this
thread.
For the BTRACE_FORMAT_PT btrace format, we don't need that initial entry since
it will be recorded in the trace. We can omit the call to btrace_add_pc.
CC: Simon Marchi <simon.marchi@ericsson.com>
2017-01-30 Markus Metzger <markus.t.metzger@intel.com>
gdb/
* btrace.c (btrace_enable): Do not call btrace_add_pc for
BTRACE_FORMAT_PT or if can_access_registers_ptid returns false.
(validate_registers_access_ptid): New.
(btrace_fetch): Call validate_registers_access_ptid.
testsuite/
* gdb.btrace/enable-running.c: New.
* gdb.btrace/enable-running.exp: New.
---
gdb/btrace.c | 46 ++++++++++++++--
gdb/testsuite/gdb.btrace/enable-running.c | 47 ++++++++++++++++
gdb/testsuite/gdb.btrace/enable-running.exp | 85 +++++++++++++++++++++++++++++
3 files changed, 174 insertions(+), 4 deletions(-)
create mode 100644 gdb/testsuite/gdb.btrace/enable-running.c
create mode 100644 gdb/testsuite/gdb.btrace/enable-running.exp
diff --git a/gdb/btrace.c b/gdb/btrace.c
index d266af7..fec2cce 100644
--- a/gdb/btrace.c
+++ b/gdb/btrace.c
@@ -141,6 +141,18 @@ ftrace_debug (const struct btrace_function *bfun, const char *prefix)
prefix, fun, file, level, ibegin, iend);
}
+/* Validate that we can read PTID's registers. */
+
+static void
+validate_registers_access_ptid (ptid_t ptid)
+{
+ struct cleanup *cleanup = save_inferior_ptid ();
+
+ inferior_ptid = ptid;
+ validate_registers_access ();
+ do_cleanups (cleanup);
+}
+
/* Return non-zero if BFUN does not match MFUN and FUN,
return zero otherwise. */
@@ -1472,10 +1484,33 @@ btrace_enable (struct thread_info *tp, const struct btrace_config *conf)
tp->btrace.target = target_enable_btrace (tp->ptid, conf);
- /* Add an entry for the current PC so we start tracing from where we
- enabled it. */
- if (tp->btrace.target != NULL)
- btrace_add_pc (tp);
+ /* We're done if we failed to enable tracing. */
+ if (tp->btrace.target == NULL)
+ return;
+
+ /* We need to undo the enable in case of errors. */
+ TRY
+ {
+ /* Add an entry for the current PC so we start tracing from where we
+ enabled it.
+
+ If we can't access TP's registers, TP is most likely running. In this
+ case, we can't really say where tracing was enabled so it should be
+ safe to simply skip this step.
+
+ This is not relevant for BTRACE_FORMAT_PT since the trace will already
+ start at the PC at which tracing was enabled. */
+ if (conf->format != BTRACE_FORMAT_PT
+ && can_access_registers_ptid (tp->ptid))
+ btrace_add_pc (tp);
+ }
+ CATCH (exception, RETURN_MASK_ALL)
+ {
+ btrace_disable (tp);
+
+ throw_exception (exception);
+ }
+ END_CATCH
}
/* See btrace.h. */
@@ -1709,6 +1744,9 @@ btrace_fetch (struct thread_info *tp)
if (btinfo->replay != NULL)
return;
+ /* Make sure the thread is not running or has already exited. */
+ validate_registers_access_ptid (tp->ptid);
+
btrace_data_init (&btrace);
cleanup = make_cleanup_btrace_data (&btrace);
diff --git a/gdb/testsuite/gdb.btrace/enable-running.c b/gdb/testsuite/gdb.btrace/enable-running.c
new file mode 100644
index 0000000..a2e9c14
--- /dev/null
+++ b/gdb/testsuite/gdb.btrace/enable-running.c
@@ -0,0 +1,47 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+ Copyright 2017 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 <pthread.h>
+
+#define NTHREADS 3
+
+static void *
+test (void *arg)
+{
+ /* Let's hope this is long enough for GDB to enable tracing and check that
+ everything is working as expected. */
+ sleep (10);
+
+ return arg;
+}
+
+int
+main (void)
+{
+ pthread_t th[NTHREADS];
+ int i;
+
+ for (i = 0; i < NTHREADS; ++i)
+ pthread_create (&th[i], NULL, test, NULL);
+
+ test (NULL); /* bp.1 */
+
+ for (i = 0; i < NTHREADS; ++i)
+ pthread_join (th[i], NULL);
+
+ return 0;
+}
diff --git a/gdb/testsuite/gdb.btrace/enable-running.exp b/gdb/testsuite/gdb.btrace/enable-running.exp
new file mode 100644
index 0000000..fe6aa0b
--- /dev/null
+++ b/gdb/testsuite/gdb.btrace/enable-running.exp
@@ -0,0 +1,85 @@
+# This testcase is part of GDB, the GNU debugger.
+#
+# Copyright 2017 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 { [skip_btrace_tests] } { return -1 }
+
+standard_testfile
+if {[gdb_compile_pthreads "$srcdir/$subdir/$srcfile" "$binfile" executable {debug}] != "" } {
+ return -1
+}
+
+# We need to enable non-stop mode for the remote case.
+save_vars { GDBFLAGS } {
+ append GDBFLAGS " -ex \"set non-stop on\""
+ clean_restart $testfile
+}
+
+if ![runto_main] {
+ return -1
+}
+
+set bp_1 [gdb_get_line_number "bp.1" $srcfile]
+
+gdb_breakpoint $bp_1
+gdb_continue_to_breakpoint "cont to $bp_1" ".*$bp_1\r\n.*"
+gdb_test "cont&" "Continuing\."
+
+# All threads are running. Let's start recording.
+gdb_test_no_output "record btrace"
+
+proc check_tracing_enabled { thread } {
+ global gdb_prompt
+
+ with_test_prefix "thread $thread" {
+ gdb_test "thread $thread" "(running).*" "is running"
+
+ # We can't read the trace while the thread is running.
+ gdb_test "info record" "Selected thread is running\." \
+ "info record while running"
+
+ # Stop the thread before reading the trace.
+ gdb_test_multiple "interrupt" "interrupt" {
+ -re "interrupt\r\n$gdb_prompt " {
+ pass "interrupt"
+ }
+ }
+ # Wait until the thread actually stopped.
+ gdb_test_multiple "" "stopped" {
+ -re "Thread $thread.*stopped\." {
+ pass "stopped"
+ }
+ }
+ # We will consume the thread's current location as part of the
+ # "info record" output.
+ gdb_test "info record" [multi_line \
+ "Active record target: record-btrace" \
+ "Recording format: .*" \
+ "Recorded \[0-9\]+ instructions \[^\\\r\\\n\]*" \
+ ]
+
+ # Continue the thread again.
+ gdb_test "cont&" "Continuing\."
+ }
+}
+
+# Check that recording was started on each thread.
+foreach thread {1 2 3 4} {
+ check_tracing_enabled $thread
+}
+
+# Stop recording while all threads are running.
+gdb_test "record stop" "Process record is stopped \[^\\\r\\\n\]*"
--
1.8.3.1
next prev parent reply other threads:[~2017-01-30 10:06 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-30 10:05 [PATCH v3 0/5] thread, btrace: allow "record btrace" " Markus Metzger
2017-01-30 10:05 ` [PATCH v3 4/5] btrace, testsuite: fix extended-remote non-stop test Markus Metzger
2017-01-31 13:00 ` Pedro Alves
2017-01-30 10:06 ` [PATCH v3 3/5] btrace: add unsupported/untested messages when skipping tests Markus Metzger
2017-01-31 13:00 ` Pedro Alves
2017-01-30 10:06 ` Markus Metzger [this message]
2017-01-31 12:59 ` [PATCH v3 2/5] btrace: allow recording to be started (and stopped) for running threads Pedro Alves
2017-01-31 16:06 ` Metzger, Markus T
2017-01-31 16:36 ` Pedro Alves
2017-02-01 8:53 ` Metzger, Markus T
2017-02-01 9:08 ` Metzger, Markus T
2017-02-01 10:40 ` Pedro Alves
2017-01-30 10:06 ` [PATCH v3 1/5] thread: add can_access_registers_ptid Markus Metzger
2017-01-30 10:06 ` [PATCH v3 5/5] btrace, testsuite: fix extended-remote fail Markus Metzger
2017-01-31 13:00 ` Pedro Alves
2017-01-31 16:06 ` Metzger, Markus T
2017-01-31 16:42 ` Pedro Alves
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1485770743-6603-3-git-send-email-markus.t.metzger@intel.com \
--to=markus.t.metzger@intel.com \
--cc=gdb-patches@sourceware.org \
--cc=palves@redhat.com \
--cc=simon.marchi@ericsson.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox