From: Simon Marchi <simon.marchi@efficios.com>
To: gdb-patches@sourceware.org
Cc: Simon Marchi <simon.marchi@efficios.com>
Subject: [PATCH 01/30] gdb/mi: fix ^running record with multiple MI interpreters
Date: Tue, 2 May 2023 16:49:41 -0400 [thread overview]
Message-ID: <20230502205011.132151-2-simon.marchi@efficios.com> (raw)
In-Reply-To: <20230502205011.132151-1-simon.marchi@efficios.com>
I stumbled on the mi_proceeded and running_result_record_printed
globals, which are shared by all MI interpreter instances (it's unlikely
that people use multiple MI interpreter instances, but it's possible).
After poking at it, I found this bug:
1. Start GDB in MI mode
2. Add a second MI interpreter with the new-ui command
3. Use -exec-run on the second interpreter
This is the output I get on the first interpreter:
=thread-group-added,id="i1"
~"Reading symbols from a.out...\n"
~"New UI allocated\n"
(gdb)
=thread-group-started,id="i1",pid="94718"
=thread-created,id="1",group-id="i1"
^running
*running,thread-id="all"
And this is the output I get on the second intepreter:
=thread-group-added,id="i1"
(gdb)
-exec-run
=thread-group-started,id="i1",pid="94718"
=thread-created,id="1",group-id="i1"
*running,thread-id="all"
The problem here is that the `^running` reply to the -exec-run command
is printed on the wrong UI. It is printed on the first one, it should
be printed on the second (the one on which we sent the -exec-run).
What happens under the hood is that captured_mi_execute_command, while
executing a command for the second intepreter, clears the
running_result_record_printed and mi_proceeded globals.
mi_about_to_proceed then sets mi_proceeded. Then, mi_on_resume_1 gets
called for the first intepreter first. Since the
!running_result_record_printed && mi_proceeded
condition is true, it prints a ^running, and sets
running_result_record_printed. When mi_on_resume_1 gets called for the
second interpreter, running_result_record_printed is already set, so
^running is not printed there.
It took me a while to understand the relationship between these two
variables. I think that in the end, this is what we want to track:
1. When executing an MI command, take note if that command causes a
"proceed". This is done in mi_about_to_proceed.
2. In mi_on_resume_1, if the command indeed caused a "proceed", we want
to output a ^running record. And we want to remember that we did,
because...
3. Back in captured_mi_execute_command, if we did not output a
^running, we want to output a ^done.
Moving those two variables to the mi_interp struture appears to fix it.
Only for the interpreter doing the -exec-run command does the
running_result_record_printed flag get cleared, and therefore only or
that one does the ^running record get printed.
Add a new test for this, that does pretty much what the reproducer above
shows. Without the fix, the test fails because
mi_send_resuming_command_raw never sees the ^running record.
Change-Id: I63ea30e6cb61a8e1dd5ef03377e6003381a9209b
---
gdb/mi/mi-interp.c | 13 ++--
gdb/mi/mi-interp.h | 6 ++
gdb/mi/mi-main.c | 14 ++--
gdb/mi/mi-main.h | 3 -
gdb/testsuite/gdb.mi/run-with-two-mi-uis.c | 7 ++
gdb/testsuite/gdb.mi/run-with-two-mi-uis.exp | 67 ++++++++++++++++++++
gdb/testsuite/lib/mi-support.exp | 26 +++++---
7 files changed, 111 insertions(+), 25 deletions(-)
create mode 100644 gdb/testsuite/gdb.mi/run-with-two-mi-uis.c
create mode 100644 gdb/testsuite/gdb.mi/run-with-two-mi-uis.exp
diff --git a/gdb/mi/mi-interp.c b/gdb/mi/mi-interp.c
index ad33a21374ab..7d8dfd830a4f 100644
--- a/gdb/mi/mi-interp.c
+++ b/gdb/mi/mi-interp.c
@@ -676,7 +676,12 @@ mi_about_to_proceed (void)
return;
}
- mi_proceeded = 1;
+ mi_interp *mi = as_mi_interp (top_level_interpreter ());
+
+ if (mi == nullptr)
+ return;
+
+ mi->mi_proceeded = 1;
}
/* When the element is non-zero, no MI notifications will be emitted in
@@ -960,7 +965,7 @@ mi_on_resume_1 (struct mi_interp *mi,
will make it impossible for frontend to know what's going on.
In future (MI3), we'll be outputting "^done" here. */
- if (!running_result_record_printed && mi_proceeded)
+ if (!mi->running_result_record_printed && mi->mi_proceeded)
{
gdb_printf (mi->raw_stdout, "%s^running\n",
current_token ? current_token : "");
@@ -976,9 +981,9 @@ mi_on_resume_1 (struct mi_interp *mi,
for (thread_info *tp : all_non_exited_threads (targ, ptid))
mi_output_running (tp);
- if (!running_result_record_printed && mi_proceeded)
+ if (!mi->running_result_record_printed && mi->mi_proceeded)
{
- running_result_record_printed = 1;
+ mi->running_result_record_printed = 1;
/* This is what gdb used to do historically -- printing prompt
even if it cannot actually accept any input. This will be
surely removed for MI3, and may be removed even earlier. */
diff --git a/gdb/mi/mi-interp.h b/gdb/mi/mi-interp.h
index e07be12f87ac..eb81cbe6cada 100644
--- a/gdb/mi/mi-interp.h
+++ b/gdb/mi/mi-interp.h
@@ -64,6 +64,12 @@ class mi_interp final : public interp
/* MI's CLI builder (wraps OUT). */
struct ui_out *cli_uiout;
+
+ int running_result_record_printed = 1;
+
+ /* Flag indicating that the target has proceeded since the last
+ command was issued. */
+ int mi_proceeded;
};
/* Output the shared object attributes to UIOUT. */
diff --git a/gdb/mi/mi-main.c b/gdb/mi/mi-main.c
index 35c74c407ee8..ab9184bc3585 100644
--- a/gdb/mi/mi-main.c
+++ b/gdb/mi/mi-main.c
@@ -84,12 +84,6 @@ char *current_token;
command including all option, and make it possible. */
static struct mi_parse *current_context;
-int running_result_record_printed = 1;
-
-/* Flag indicating that the target has proceeded since the last
- command was issued. */
-int mi_proceeded;
-
static void mi_cmd_execute (struct mi_parse *parse);
static void mi_execute_async_cli_command (const char *cli_command,
@@ -1803,8 +1797,8 @@ captured_mi_execute_command (struct ui_out *uiout, struct mi_parse *context)
scoped_restore save_token = make_scoped_restore (¤t_token,
context->token);
- running_result_record_printed = 0;
- mi_proceeded = 0;
+ mi->running_result_record_printed = 0;
+ mi->mi_proceeded = 0;
switch (context->op)
{
case MI_COMMAND:
@@ -1822,7 +1816,7 @@ captured_mi_execute_command (struct ui_out *uiout, struct mi_parse *context)
to directly use the mi_interp's uiout, since the command
could have reset the interpreter, in which case the current
uiout will most likely crash in the mi_out_* routines. */
- if (!running_result_record_printed)
+ if (!mi->running_result_record_printed)
{
gdb_puts (context->token, mi->raw_stdout);
/* There's no particularly good reason why target-connect results
@@ -1861,7 +1855,7 @@ captured_mi_execute_command (struct ui_out *uiout, struct mi_parse *context)
|| current_interp_named_p (INTERP_MI3)
|| current_interp_named_p (INTERP_MI4))
{
- if (!running_result_record_printed)
+ if (!mi->running_result_record_printed)
{
gdb_puts (context->token, mi->raw_stdout);
gdb_puts ("^done", mi->raw_stdout);
diff --git a/gdb/mi/mi-main.h b/gdb/mi/mi-main.h
index ff1d4ed84bc8..b9fe295d3306 100644
--- a/gdb/mi/mi-main.h
+++ b/gdb/mi/mi-main.h
@@ -36,9 +36,6 @@ extern int mi_async_p (void);
extern char *current_token;
-extern int running_result_record_printed;
-extern int mi_proceeded;
-
struct mi_suppress_notification
{
/* Breakpoint notification suppressed? */
diff --git a/gdb/testsuite/gdb.mi/run-with-two-mi-uis.c b/gdb/testsuite/gdb.mi/run-with-two-mi-uis.c
new file mode 100644
index 000000000000..1e2428e0d544
--- /dev/null
+++ b/gdb/testsuite/gdb.mi/run-with-two-mi-uis.c
@@ -0,0 +1,7 @@
+#include <unistd.h>
+
+int
+main (void)
+{
+ sleep (1234);
+}
diff --git a/gdb/testsuite/gdb.mi/run-with-two-mi-uis.exp b/gdb/testsuite/gdb.mi/run-with-two-mi-uis.exp
new file mode 100644
index 000000000000..9a049c98ea7a
--- /dev/null
+++ b/gdb/testsuite/gdb.mi/run-with-two-mi-uis.exp
@@ -0,0 +1,67 @@
+# Copyright 2023 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/>.
+
+# Test doing an -exec-run while there are two MI UIs.
+
+load_lib mi-support.exp
+
+standard_testfile
+
+if {[build_executable $testfile.exp $testfile ${srcfile} "debug"] == -1} {
+ untested "failed to compile"
+ return
+}
+
+# Run one configuration of the test.
+#
+# UI_TO_RUN is the UI that should issue the run command.
+
+proc do_test { ui_to_run } {
+ if {[mi_clean_restart $::binfile "separate-mi-tty"] != 0} {
+ fail "could not start gdb"
+ return
+ }
+
+ with_spawn_id $::gdb_main_spawn_id {
+ lassign [create_mi_ui] second_mi_spawn_id second_mi_tty_name
+ }
+
+ with_spawn_id $second_mi_spawn_id {
+ gdb_expect {
+ -re "=thread-group-added,id=\"i1\"\r\n$::mi_gdb_prompt$" {
+ pass "consume"
+ }
+ }
+ }
+
+ if { $ui_to_run == "first" } {
+ set spawn_id_to_run $::mi_spawn_id
+ } elseif { $ui_to_run == "second" } {
+ set spawn_id_to_run $second_mi_spawn_id
+ } else {
+ error "invalid ui_to_run value"
+ }
+
+ with_spawn_id $spawn_id_to_run {
+ # mi_runto_main implicitly verifies that the UI doing the -exec-run gets
+ # the expected ^running record.
+ mi_runto_main
+ }
+}
+
+foreach_with_prefix ui_to_run {first second} {
+ do_test $ui_to_run
+}
+
diff --git a/gdb/testsuite/lib/mi-support.exp b/gdb/testsuite/lib/mi-support.exp
index 0d830d8e4ae5..2ff4ab93ea82 100644
--- a/gdb/testsuite/lib/mi-support.exp
+++ b/gdb/testsuite/lib/mi-support.exp
@@ -131,6 +131,21 @@ proc mi_create_inferior_pty {} {
}
}
+# Create a new pty, and reate a new MI UI (using the new-ui command) on it.
+#
+# Return a list with the spawn id for that pty and the pty file name.
+
+proc create_mi_ui {} {
+ spawn -pty
+ set tty_name $spawn_out(slave,name)
+ gdb_test_multiple "new-ui mi $tty_name" "new-ui" {
+ -re "New UI allocated\r\n$::gdb_prompt $" {
+ }
+ }
+
+ return [list $spawn_id $tty_name]
+}
+
#
# Like default_mi_gdb_start below, but the MI is created as a separate
# ui in a new tty. The global MI_SPAWN_ID is updated to point at the
@@ -154,13 +169,7 @@ proc mi_gdb_start_separate_mi_tty { { flags {} } } {
gdb_start
# Create the new PTY for the MI UI.
- spawn -pty
- set mi_spawn_id $spawn_id
- set mi_tty_name $spawn_out(slave,name)
- gdb_test_multiple "new-ui mi $mi_tty_name" "new-ui" {
- -re "New UI allocated\r\n$gdb_prompt $" {
- }
- }
+ lassign [create_mi_ui] mi_spawn_id mi_tty_name
# Switch to the MI channel.
set gdb_main_spawn_id $gdb_spawn_id
@@ -822,7 +831,7 @@ proc mi_gdb_test { args } {
fail "$errmsg"
return -1
}
- -re ".*$mi_gdb_prompt\[ \]*$" {
+ -re "(.*$mi_gdb_prompt\[ \]*)$" {
if {![string match "" $message]} {
fail "$message (unexpected output)"
}
@@ -1082,6 +1091,7 @@ proc mi_runto_helper {func run_or_continue args} {
# file.", etc. to the CLI stream.
set extra_output "&\"\[^\r\n\]+\"\r\n"
}
+
mi_gdb_test "200-break-insert [join $extra_opts " "] -t $func" "${extra_output}200\\^done,$bp" \
"breakpoint at $func"
--
2.40.1
next prev parent reply other threads:[~2023-05-02 20:50 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-02 20:49 [PATCH 00/30] Switch interpreters to use virtual methods Simon Marchi
2023-05-02 20:49 ` Simon Marchi [this message]
2023-05-03 14:42 ` [PATCH 01/30] gdb/mi: fix ^running record with multiple MI interpreters Alexandra Petlanova Hajkova
2023-05-03 15:02 ` Simon Marchi
2023-05-29 14:53 ` Simon Marchi
2023-05-02 20:49 ` [PATCH 02/30] gdb/mi: make current_token a field of mi_interp Simon Marchi
2023-05-04 12:31 ` Alexandra Petlanova Hajkova
2023-05-02 20:49 ` [PATCH 03/30] gdb: add interp::on_signal_received method Simon Marchi
2023-05-04 14:38 ` Alexandra Petlanova Hajkova
2023-05-02 20:49 ` [PATCH 04/30] gdb: add interp::on_normal_stop method Simon Marchi
2023-05-02 20:49 ` [PATCH 05/30] gdb: add interp::on_signal_exited method Simon Marchi
2023-05-02 20:49 ` [PATCH 06/30] gdb: add interp::on_exited method Simon Marchi
2023-05-02 20:49 ` [PATCH 07/30] gdb: add interp::on_no_history method Simon Marchi
2023-05-02 20:49 ` [PATCH 08/30] gdb: add interp::on_sync_execution_done method Simon Marchi
2023-05-02 20:49 ` [PATCH 09/30] gdb: add interp::on_command_error method Simon Marchi
2023-05-02 20:49 ` [PATCH 10/30] gdb: add interp::on_user_selected_context_changed method Simon Marchi
2023-05-02 20:49 ` [PATCH 11/30] gdb: add interp::on_new_thread method Simon Marchi
2023-05-02 20:49 ` [PATCH 12/30] gdb: add interp::on_thread_exited method Simon Marchi
2023-05-02 20:49 ` [PATCH 13/30] gdb: add interp::on_inferior_added method Simon Marchi
2023-05-02 20:49 ` [PATCH 14/30] gdb: add interp::on_inferior_appeared method Simon Marchi
2023-05-02 20:49 ` [PATCH 15/30] gdb: add interp::on_inferior_disappeared method Simon Marchi
2023-05-02 20:49 ` [PATCH 16/30] gdb: add interp::on_inferior_removed method Simon Marchi
2023-05-02 20:49 ` [PATCH 17/30] gdb: add interp::on_record_changed method Simon Marchi
2023-05-02 20:49 ` [PATCH 18/30] gdb: add interp::on_target_resumed method Simon Marchi
2023-05-02 20:49 ` [PATCH 19/30] gdb: add interp::on_solib_loaded method Simon Marchi
2023-05-02 20:50 ` [PATCH 20/30] gdb: add interp::on_solib_unloaded method Simon Marchi
2023-05-02 20:50 ` [PATCH 21/30] gdb: add interp::on_about_to_proceed method Simon Marchi
2023-05-02 20:50 ` [PATCH 22/30] gdb: add interp::on_traceframe_changed method Simon Marchi
2023-05-02 20:50 ` [PATCH 23/30] gdb: add interp::on_tsv_created method Simon Marchi
2023-05-02 20:50 ` [PATCH 24/30] gdb: add interp::on_tsv_deleted method Simon Marchi
2023-05-02 20:50 ` [PATCH 25/30] gdb: add interp::on_tsv_modified method Simon Marchi
2023-05-02 20:50 ` [PATCH 26/30] gdb: add interp::on_breakpoint_created method Simon Marchi
2023-05-02 20:50 ` [PATCH 27/30] gdb: add interp::on_breakpoint_deleted method Simon Marchi
2023-05-02 20:50 ` [PATCH 28/30] gdb: add interp::on_breakpoint_modified method Simon Marchi
2023-05-02 20:50 ` [PATCH 29/30] gdb: add interp::on_param_changed method Simon Marchi
2023-05-02 20:50 ` [PATCH 30/30] gdb: add interp::on_memory_changed method Simon Marchi
2023-05-02 20:50 ` [PATCH 00/30] Make interpreters use virtual methods (instead of observers) Simon Marchi
2023-05-23 12:43 ` [PATCH 00/30] Switch interpreters to use virtual methods Simon Marchi
2023-05-30 19:09 ` Simon Marchi
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=20230502205011.132151-2-simon.marchi@efficios.com \
--to=simon.marchi@efficios.com \
--cc=gdb-patches@sourceware.org \
/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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).