From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mga02.intel.com (mga02.intel.com [134.134.136.20]) by sourceware.org (Postfix) with ESMTPS id 653F5385800F for ; Tue, 16 Mar 2021 09:35:04 +0000 (GMT) DMARC-Filter: OpenDMARC Filter v1.3.2 sourceware.org 653F5385800F IronPort-SDR: kwV8/tQZzeqySlTuXT3lliU40y76+X5CIAeQYjh1qsy+4Xb7mmiKbwD3xjLYT3adEqEw1gi+Pi NzQcqC3C7V7g== X-IronPort-AV: E=McAfee;i="6000,8403,9924"; a="176360916" X-IronPort-AV: E=Sophos;i="5.81,251,1610438400"; d="scan'208";a="176360916" Received: from fmsmga005.fm.intel.com ([10.253.24.32]) by orsmga101.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 16 Mar 2021 02:35:02 -0700 IronPort-SDR: 4k6qaieUoEkWTug5Ea5h4ll/0s2bMGJ0cH+XHBnJF5Y3mOQbTTmNTk0ioNJVQPOzHIF5EW9gfd MsVvCno2XK9g== X-IronPort-AV: E=Sophos;i="5.81,251,1610438400"; d="scan'208";a="605193540" Received: from labpc2407.iul.intel.com (HELO localhost) ([172.28.50.61]) by fmsmga005-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 16 Mar 2021 02:35:01 -0700 From: Markus Metzger To: gdb-patches@sourceware.org Subject: [PATCH 1/4] gdb, infrun, btrace: fix reverse/replay stepping at end of execution history Date: Tue, 16 Mar 2021 10:34:58 +0100 Message-Id: <20210316093501.936148-2-markus.t.metzger@intel.com> X-Mailer: git-send-email 2.29.2 In-Reply-To: <20210316093501.936148-1-markus.t.metzger@intel.com> References: <20210316093501.936148-1-markus.t.metzger@intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit X-Spam-Status: No, score=-9.5 required=5.0 tests=BAYES_00, GIT_PATCH_0, KAM_DMARC_STATUS, KAM_SHORT, RCVD_IN_MSPIKE_H3, RCVD_IN_MSPIKE_WL, SPF_HELO_NONE, SPF_PASS, TXREP autolearn=ham autolearn_force=no version=3.4.2 X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on server2.sourceware.org X-BeenThere: gdb-patches@sourceware.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Gdb-patches mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 16 Mar 2021 09:35:06 -0000 When trying to step over a breakopint at the end of the trace, the step-over will fail with no-history. This does not clear step_over_info so a subsequent resume will cause GDB to not resume the thread and expect a SIGTRAP to complete the step-over. This will never come causing GDB to hang in the wait-for-event poll. That step-over failed after actually completing the step. This is wrong. Fix it by moving the end of execution history check to before we are stepping. This exposes another issue, however. When completing a step-over at the end of the execution history, we implicitly stop replaying that thread. A continue command would resume after the step-over and, since we're no longer replaying, would continue recording. Fix that by recording the replay state in the thread's control state and failing with no-history in keep_going if we're switching from replay to recording. gdb/ChangeLog: 2021-01-14 Markus Metzger * gdbthread.h (struct thread_control_state) : New. * infrun.c (clear_proceed_status_thread): Set thread_control_state.is_replaying. (keep_going_pass_signal): Check thread_control_state.is_replaying. * record-btrace.c (record_btrace_single_step_forward): Move end of execution history check. gdb/testsuite/ChangeLog: 2021-01-13 Markus Metzger * gdb.btrace/cont-hang.exp: New file. * gdb.btrace/step-hang.exp: New file. * gdb.btrace/stepn.exp: New file. --- gdb/gdbthread.h | 3 ++ gdb/infrun.c | 22 ++++++++++ gdb/record-btrace.c | 19 ++++----- gdb/testsuite/gdb.btrace/cont-hang.exp | 47 +++++++++++++++++++++ gdb/testsuite/gdb.btrace/step-hang.exp | 46 +++++++++++++++++++++ gdb/testsuite/gdb.btrace/stepn.exp | 56 ++++++++++++++++++++++++++ 6 files changed, 184 insertions(+), 9 deletions(-) create mode 100644 gdb/testsuite/gdb.btrace/cont-hang.exp create mode 100644 gdb/testsuite/gdb.btrace/step-hang.exp create mode 100644 gdb/testsuite/gdb.btrace/stepn.exp diff --git a/gdb/gdbthread.h b/gdb/gdbthread.h index eef37f79e6a..563b7bd8954 100644 --- a/gdb/gdbthread.h +++ b/gdb/gdbthread.h @@ -159,6 +159,9 @@ struct thread_control_state command. This is used to decide whether "set scheduler-locking step" behaves like "on" or "off". */ int stepping_command = 0; + + /* Whether the thread was replaying when the command was issued. */ + bool is_replaying = false; }; /* Inferior thread specific part of `struct infcall_suspend_state'. */ diff --git a/gdb/infrun.c b/gdb/infrun.c index a271220b261..591fda93d21 100644 --- a/gdb/infrun.c +++ b/gdb/infrun.c @@ -2658,6 +2658,8 @@ clear_proceed_status_thread (struct thread_info *tp) /* Discard any remaining commands or status from previous stop. */ bpstat_clear (&tp->control.stop_bpstat); + + tp->control.is_replaying = target_record_is_replaying (tp->ptid); } void @@ -7813,6 +7815,26 @@ keep_going_pass_signal (struct execution_control_state *ecs) gdb_assert (ecs->event_thread->ptid == inferior_ptid); gdb_assert (!ecs->event_thread->resumed); + /* When a thread reaches the end of its execution history, it automatically + stops replaying. This is so the user doesn't need to explicitly stop it + with a separate command. + + We do not want a single command (e.g. continue) to transition from + replaying to recording, though, e.g. when starting from a breakpoint we + needed to step over at the end of the trace. When we reach the end of the + execution history during stepping, stop with no-history. + + The other direction is fine. When we're at the end of the execution + history, we may reverse-continue to start replaying. */ + if (ecs->event_thread->control.is_replaying + && !target_record_is_replaying (ecs->event_thread->ptid)) + { + gdb::observers::no_history.notify (); + stop_waiting (ecs); + normal_stop (); + return; + } + /* Save the pc before execution, to compare with pc after stop. */ ecs->event_thread->prev_pc = regcache_read_pc_protected (get_thread_regcache (ecs->event_thread)); diff --git a/gdb/record-btrace.c b/gdb/record-btrace.c index d9cc7a3b6d8..c36f3350b29 100644 --- a/gdb/record-btrace.c +++ b/gdb/record-btrace.c @@ -2350,6 +2350,16 @@ record_btrace_single_step_forward (struct thread_info *tp) if (replay == NULL) return btrace_step_no_history (); + /* The execution trace contains (and ends with) the current instruction. + This instruction has not been executed, yet, so the trace really ends + one instruction earlier. + + We'd fail later on in btrace_insn_next () but we must not trigger + breakpoints as we're not really able to step. */ + btrace_insn_end (&end, btinfo); + if (btrace_insn_cmp (replay, &end) == 0) + return btrace_step_no_history (); + /* Check if we're stepping a breakpoint. */ if (record_btrace_replay_at_breakpoint (tp)) return btrace_step_stopped (); @@ -2372,15 +2382,6 @@ record_btrace_single_step_forward (struct thread_info *tp) } while (btrace_insn_get (replay) == NULL); - /* Determine the end of the instruction trace. */ - btrace_insn_end (&end, btinfo); - - /* The execution trace contains (and ends with) the current instruction. - This instruction has not been executed, yet, so the trace really ends - one instruction earlier. */ - if (btrace_insn_cmp (replay, &end) == 0) - return btrace_step_no_history (); - return btrace_step_spurious (); } diff --git a/gdb/testsuite/gdb.btrace/cont-hang.exp b/gdb/testsuite/gdb.btrace/cont-hang.exp new file mode 100644 index 00000000000..59d0e17599e --- /dev/null +++ b/gdb/testsuite/gdb.btrace/cont-hang.exp @@ -0,0 +1,47 @@ +# This testcase is part of GDB, the GNU debugger. +# +# Copyright 2021 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 . + +# Test that we do not hang when trying to continue over a breakpoint at +# the end of the trace. + +if { [skip_btrace_tests] } { + unsupported "target does not support record-btrace" + return -1 +} + +standard_testfile record_goto.c +if [prepare_for_testing "failed to prepare" $testfile $srcfile] { + return -1 +} + +if ![runto_main] { + untested "failed to run to main" + return -1 +} + +# Trace the call to the test function. +gdb_test_no_output "record btrace" +gdb_test "next" "main\.3.*" + +# We need to be replaying, otherwise, we'd just continue recording. +gdb_test "reverse-stepi" +gdb_test "break" + +# Continuing will step over the breakpoint and then run into the end of +# the execution history. This ends replay, so we can continue recording. +gdb_test "continue" "No more reverse-execution history.*" +gdb_continue_to_end diff --git a/gdb/testsuite/gdb.btrace/step-hang.exp b/gdb/testsuite/gdb.btrace/step-hang.exp new file mode 100644 index 00000000000..e175ff5c35d --- /dev/null +++ b/gdb/testsuite/gdb.btrace/step-hang.exp @@ -0,0 +1,46 @@ +# This testcase is part of GDB, the GNU debugger. +# +# Copyright 2021 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 . + +# Test that we do not hang when trying to step over a breakpoint at the +# end of the trace. + +if { [skip_btrace_tests] } { + unsupported "target does not support record-btrace" + return -1 +} + +standard_testfile record_goto.c +if [prepare_for_testing "failed to prepare" $testfile $srcfile] { + return -1 +} + +if ![runto_main] { + untested "failed to run to main" + return -1 +} + +# Trace the call to the test function. +gdb_test_no_output "record btrace" +gdb_test "next" "main\.3.*" + +# We need to be replaying, otherwise, we'd just continue recording. +gdb_test "reverse-stepi" +gdb_test "break" + +# Stepping over the breakpoint ends replaying and we can continue recording. +gdb_test "step" "main\.3.*" +gdb_continue_to_end diff --git a/gdb/testsuite/gdb.btrace/stepn.exp b/gdb/testsuite/gdb.btrace/stepn.exp new file mode 100644 index 00000000000..686ca8ceb5e --- /dev/null +++ b/gdb/testsuite/gdb.btrace/stepn.exp @@ -0,0 +1,56 @@ +# This testcase is part of GDB, the GNU debugger. +# +# Copyright 2021 Free Software Foundation, Inc. +# +# Contributed by Intel Corp. +# +# 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 . + +# Test that step n does not start recording when issued while replaying. + +if { [skip_btrace_tests] } { + unsupported "target does not support record-btrace" + return -1 +} + +standard_testfile record_goto.c +if [prepare_for_testing "failed to prepare" $testfile $srcfile] { + return -1 +} + +if ![runto_main] { + untested "failed to run to main" + return -1 +} + +# Trace the call to the test function. +gdb_test_no_output "record btrace" +gdb_test "next" "main\.3.*" + +# Stepping should bring us to the end of the execution history, but should +# not resume recording. +with_test_prefix "stepi" { + gdb_test "reverse-stepi" + gdb_test "stepi 5" "No more reverse-execution history.*main\.3.*" +} + +with_test_prefix "step" { + gdb_test "reverse-step" + gdb_test "step 5" "No more reverse-execution history.*main\.3.*" +} + +with_test_prefix "next" { + gdb_test "reverse-step" + gdb_test "next 5" "No more reverse-execution history.*main\.3.*" +} -- 2.29.2 Intel Deutschland GmbH Registered Address: Am Campeon 10, 85579 Neubiberg, Germany Tel: +49 89 99 8853-0, www.intel.de Managing Directors: Christin Eisenschmid, Sharon Heck, Tiffany Doon Silva Chairperson of the Supervisory Board: Nicole Lau Registered Office: Munich Commercial Register: Amtsgericht Muenchen HRB 186928