From: Hui Zhu <teawater@gmail.com>
To: Pedro Alves <palves@redhat.com>
Cc: Tom Tromey <tromey@redhat.com>,
gdb-patches@sourceware.org, Keith Seitz <keiths@redhat.com>,
Yao Qi <yao@codesourcery.com>
Subject: Re: [RFC] PR 15075 dprintf interferes with "next"
Date: Fri, 07 Jun 2013 03:16:00 -0000 [thread overview]
Message-ID: <CANFwon2oFaTc0ey4B7z=OXcne42SRwp_mozfOYKU4nx-K+80nA@mail.gmail.com> (raw)
In-Reply-To: <51ACD6ED.7040604@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 2216 bytes --]
Hi Pedro,
Thanks for your review.
On Tue, Jun 4, 2013 at 1:48 AM, Pedro Alves <palves@redhat.com> wrote:
> On 06/03/2013 05:06 AM, Hui Zhu wrote:
>> Ping http://sourceware.org/ml/gdb-patches/2013-05/msg00958.html
>>
>
> As this exposes the non-stop racy failure, we should fix it that one
> first. Failing that, we should kfail or skip the test with
> remote targets. Let's consider the latter option later if we don't
> manage to address the race timely. As I said on:
>
> http://sourceware.org/ml/gdb-patches/2013-05/msg01111.html
>
> I'm investigating this. I have a prototype patch, but I need
> a bit more to handle some details, like what to do with
> signal catchpoints when we find threads had been stopped with
> a signal (I'm currently thinking of skipping the catchpoints).
> I'm composing a test to exercise/expose this kind of stuff,
> for a better RFC.
I add a check in the begin of dprintf-non-stop.exp. Then if the
target in remote, this test will not support.
Please help me review it.
Thanks,
Hui
2013-06-07 Yao Qi <yao@codesourcery.com>
Hui Zhu <hui@codesourcery.com>
Pedro Alves <palves@redhat.com>
PR breakpoints/15075
PR breakpoints/15434
* breakpoint.c (bpstat_stop_status): Call
b->ops->after_condition_true.
(update_dprintf_command_list): Don't append "continue" command
to the command list of dprintf breakpoint.
(base_breakpoint_after_condition_true): New function.
(base_breakpoint_ops): Add base_breakpoint_after_condition_true.
(dprintf_create_breakpoints_sal,
dprintf_after_condition_true): New functions.
(initialize_breakpoint_ops): Set dprintf_create_breakpoints_sal
and dprintf_after_condition_true.
* breakpoint.h (breakpoint_ops): Add after_condition_true.
2013-06-07 Yao Qi <yao@codesourcery.com>
Hui Zhu <hui@codesourcery.com>
PR breakpoints/15075
PR breakpoints/15434
* gdb.base/dprintf-next.c: New file.
* gdb.base/dprintf-next.exp: New file.
* gdb.base/dprintf-non-stop.c: New file.
* gdb.base/dprintf-non-stop.exp: New file.
* gdb.base/dprintf.exp: Don't check "continue" in the output
of "info breakpoints".
* gdb.mi/mi-breakpoint-changed.exp (test_insert_delete_modify):
Don't check "continue" in script field.
[-- Attachment #2: dprintf-continue.txt --]
[-- Type: text/plain, Size: 5391 bytes --]
--- a/gdb/breakpoint.c
+++ b/gdb/breakpoint.c
@@ -5299,6 +5299,8 @@ bpstat_stop_status (struct address_space
if (command_line_is_silent (bs->commands
? bs->commands->commands : NULL))
bs->print = 0;
+
+ b->ops->after_condition_true (bs);
}
}
@@ -8934,25 +8936,16 @@ update_dprintf_command_list (struct brea
_("Invalid dprintf style."));
gdb_assert (printf_line != NULL);
- /* Manufacture a printf/continue sequence. */
+ /* Manufacture a printf sequence. */
{
- struct command_line *printf_cmd_line, *cont_cmd_line = NULL;
-
- if (strcmp (dprintf_style, dprintf_style_agent) != 0)
- {
- cont_cmd_line = xmalloc (sizeof (struct command_line));
- cont_cmd_line->control_type = simple_control;
- cont_cmd_line->body_count = 0;
- cont_cmd_line->body_list = NULL;
- cont_cmd_line->next = NULL;
- cont_cmd_line->line = xstrdup ("continue");
- }
+ struct command_line *printf_cmd_line
+ = xmalloc (sizeof (struct command_line));
printf_cmd_line = xmalloc (sizeof (struct command_line));
printf_cmd_line->control_type = simple_control;
printf_cmd_line->body_count = 0;
printf_cmd_line->body_list = NULL;
- printf_cmd_line->next = cont_cmd_line;
+ printf_cmd_line->next = NULL;
printf_cmd_line->line = printf_line;
breakpoint_set_commands (b, printf_cmd_line);
@@ -12738,6 +12731,14 @@ base_breakpoint_explains_signal (struct
return BPSTAT_SIGNAL_HIDE;
}
+/* The default "after_condition_true" method. */
+
+static void
+base_breakpoint_after_condition_true (struct bpstats *bs)
+{
+ /* Nothing to do. */
+}
+
struct breakpoint_ops base_breakpoint_ops =
{
base_breakpoint_dtor,
@@ -12757,7 +12758,8 @@ struct breakpoint_ops base_breakpoint_op
base_breakpoint_create_sals_from_address,
base_breakpoint_create_breakpoints_sal,
base_breakpoint_decode_linespec,
- base_breakpoint_explains_signal
+ base_breakpoint_explains_signal,
+ base_breakpoint_after_condition_true,
};
/* Default breakpoint_ops methods. */
@@ -13351,6 +13353,76 @@ dprintf_print_recreate (struct breakpoin
print_recreate_thread (tp, fp);
}
+/* Implement the "create_breakpoints_sal" breakpoint_ops method for
+ dprintf. */
+
+static void
+dprintf_create_breakpoints_sal (struct gdbarch *gdbarch,
+ struct linespec_result *canonical,
+ struct linespec_sals *lsal,
+ char *cond_string,
+ char *extra_string,
+ enum bptype type_wanted,
+ enum bpdisp disposition,
+ int thread,
+ int task, int ignore_count,
+ const struct breakpoint_ops *ops,
+ int from_tty, int enabled,
+ int internal, unsigned flags)
+{
+ struct breakpoint *b;
+
+ create_breakpoints_sal_default (gdbarch, canonical, lsal,
+ cond_string, extra_string,
+ type_wanted,
+ disposition, thread, task,
+ ignore_count, ops, from_tty,
+ enabled, internal, flags);
+
+ b = get_breakpoint (breakpoint_count);
+ gdb_assert (b != NULL);
+
+ breakpoint_set_silent (b, 0);
+}
+
+/* Implement the "after_condition_true" breakpoint_ops method for
+ dprintf.
+
+ dprintf's are implemented with regular commands in their command
+ list, but we run the commands here instead of before presenting the
+ stop to the user, as dprintf's don't actually cause a stop. This
+ also makes it so that the commands of multiple dprintfs at the same
+ address are all handled. */
+
+static void
+dprintf_after_condition_true (struct bpstats *bs)
+{
+ struct cleanup *old_chain;
+ struct bpstats tmp_bs = { NULL };
+ struct bpstats *tmp_bs_p = &tmp_bs;
+
+ /* dprintf's never cause a stop. This wasn't set in the
+ check_status hook instead because that would make the dprintf's
+ condition not be evaluated. */
+ bs->stop = 0;
+
+ /* Run the command list here. Take ownership of it instead of
+ copying. We never want these commands to run later in
+ bpstat_do_actions, if a breakpoint that causes a stop happens to
+ be set at same address as this dprintf, or even if running the
+ commands here throws. */
+ tmp_bs.commands = bs->commands;
+ bs->commands = NULL;
+ old_chain = make_cleanup_decref_counted_command_line (&tmp_bs.commands);
+
+ bpstat_do_actions_1 (&tmp_bs_p);
+
+ /* 'tmp_bs.commands' will usually be NULL by now, but
+ bpstat_do_actions_1 may return early without processing the whole
+ list. */
+ do_cleanups (old_chain);
+}
+
/* The breakpoint_ops structure to be used on static tracepoints with
markers (`-m'). */
@@ -15847,6 +15919,8 @@ initialize_breakpoint_ops (void)
ops->print_it = bkpt_print_it;
ops->print_mention = bkpt_print_mention;
ops->print_recreate = dprintf_print_recreate;
+ ops->create_breakpoints_sal = dprintf_create_breakpoints_sal;
+ ops->after_condition_true = dprintf_after_condition_true;
}
/* Chain containing all defined "enable breakpoint" subcommands. */
--- a/gdb/breakpoint.h
+++ b/gdb/breakpoint.h
@@ -614,6 +614,10 @@ struct breakpoint_ops
'catch signal' interact properly with 'handle'; see
bpstat_explains_signal. */
enum bpstat_signal_value (*explains_signal) (struct breakpoint *);
+
+ /* Called after evaluating the breakpoint's condition,
+ and only if it evaluated true. */
+ void (*after_condition_true) (struct bpstats *bs);
};
/* Helper for breakpoint_ops->print_recreate implementations. Prints
[-- Attachment #3: dprintf-continue-test.txt --]
[-- Type: text/plain, Size: 6024 bytes --]
--- /dev/null
+++ b/gdb/testsuite/gdb.base/dprintf-next.c
@@ -0,0 +1,26 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+ Copyright (C) 2013 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/>. */
+
+int
+main (void)
+{
+ int x = 5;
+
+ ++x; /* Next without dprintf. */
+ ++x; /* Set dprintf here. */
+ return x - 7;
+}
--- /dev/null
+++ b/gdb/testsuite/gdb.base/dprintf-next.exp
@@ -0,0 +1,36 @@
+# Copyright 2013 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/>.
+
+standard_testfile
+
+set executable $testfile
+set expfile $testfile.exp
+
+set dp_location [gdb_get_line_number "Set dprintf here"]
+
+if [prepare_for_testing "failed to prepare for dprintf with next" \
+ ${testfile} ${srcfile} {debug}] {
+ return -1
+}
+
+if ![runto_main] {
+ fail "Can't run to main"
+ return -1
+}
+
+gdb_test "dprintf $dp_location, \"%d\\n\", x" \
+ "Dprintf .*"
+
+gdb_test "next" "\\+\\+x\;.*\/\* Next without dprintf.*" "next 1"
+gdb_test "next" "\\+\\+x\;.*\/\* Set dprintf here.*" "next 2"
--- /dev/null
+++ b/gdb/testsuite/gdb.base/dprintf-non-stop.c
@@ -0,0 +1,30 @@
+/* This testcase is part of GDB, the GNU debugger.
+
+ Copyright (C) 2013 Free Software Foundation, Inc.
+ Contributed by Hui Zhu <hui@codesourcery.com>
+
+ 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/>. */
+
+void
+foo ()
+{
+}
+
+int
+main ()
+{
+ foo ();
+ sleep (3);
+ return 0;
+}
--- /dev/null
+++ b/gdb/testsuite/gdb.base/dprintf-non-stop.exp
@@ -0,0 +1,54 @@
+# Copyright (C) 2013 Free Software Foundation, Inc.
+# Contributed by Hui Zhu <hui@codesourcery.com>
+
+# 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] then {
+ unsupported "Dprintf with non-stop is not supported."
+ return 0
+}
+
+standard_testfile
+
+if [prepare_for_testing "failed to prepare for dprintf with non-stop" \
+ ${testfile} ${srcfile} {debug}] {
+ return -1
+}
+
+gdb_test_no_output "set target-async on"
+gdb_test_no_output "set non-stop on"
+
+if ![runto main] {
+ fail "Can't run to main"
+ return -1
+}
+
+gdb_test "dprintf foo,\"At foo entry\\n\"" "Dprintf .*"
+
+send_gdb "continue &\n"
+exec sleep 1
+
+set test "interrupt"
+gdb_test_multiple $test $test {
+ -re "interrupt\r\n$gdb_prompt " {
+ pass $test
+ }
+}
+
+set test "inferior stopped"
+gdb_test_multiple "" $test {
+ -re "\r\n\\\[.* \[0-9\]+\\\] #1 stopped\\\.\r\n" {
+ pass $test
+ }
+}
--- a/gdb/testsuite/gdb.base/dprintf.exp
+++ b/gdb/testsuite/gdb.base/dprintf.exp
@@ -50,10 +50,8 @@ gdb_test_sequence "info breakpoints" "dp
"\[\r\n\]2 breakpoint"
"\[\r\n\]3 dprintf"
"\[\r\n\] printf \"At foo entry\\\\n\""
- "\[\r\n\] continue"
"\[\r\n\]4 dprintf"
"\[\r\n\] printf \"arg=%d, g=%d\\\\n\", arg, g"
- "\[\r\n\] continue"
}
gdb_test "break $bp_location1" \
@@ -135,4 +133,3 @@ if $target_can_dprintf {
gdb_test "set dprintf-style foobar" "Undefined item: \"foobar\"." \
"Set dprintf style to an unrecognized type"
-
--- a/gdb/testsuite/gdb.mi/mi-breakpoint-changed.exp
+++ b/gdb/testsuite/gdb.mi/mi-breakpoint-changed.exp
@@ -96,7 +96,7 @@ proc test_insert_delete_modify { } {
$test
set test "dprintf marker, \"arg\" \""
mi_gdb_test $test \
- {.*=breakpoint-created,bkpt=\{number="6",type="dprintf".*,script=\{\"printf \\\\\"arg\\\\\" \\\\\"\",\"continue\"\}.*\}\r\n\^done} \
+ {.*=breakpoint-created,bkpt=\{number="6",type="dprintf".*,script=\{\"printf \\\\\"arg\\\\\" \\\\\"\"\}.*\}\r\n\^done} \
$test
# 2. when modifying condition
next prev parent reply other threads:[~2013-06-07 3:16 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-02-18 13:09 Yao Qi
2013-02-18 21:36 ` Joel Brobecker
2013-02-21 16:36 ` Tom Tromey
2013-04-24 1:24 ` Hui Zhu
2013-04-24 6:06 ` Keith Seitz
2013-04-24 13:30 ` Hui Zhu
2013-04-24 14:03 ` Yao Qi
2013-04-24 14:09 ` Hui Zhu
2013-05-16 7:29 ` Hui Zhu
2013-05-17 21:01 ` Pedro Alves
2013-05-22 10:22 ` Hui Zhu
2013-05-22 12:46 ` Pedro Alves
2013-05-28 0:02 ` Hui Zhu
2013-05-28 3:36 ` Yao Qi
2013-05-29 10:08 ` Hui Zhu
2013-06-03 4:07 ` Hui Zhu
2013-06-03 17:48 ` Pedro Alves
2013-06-07 3:16 ` Hui Zhu [this message]
2013-06-17 7:36 ` Hui Zhu
2013-06-18 18:16 ` Pedro Alves
2013-06-24 8:36 ` Hui Zhu
2013-06-24 22:06 ` Pedro Alves
2013-06-25 9:14 ` Hui Zhu
2013-06-25 11:47 ` Pedro Alves
2013-06-25 13:02 ` Hui Zhu
2013-06-25 14:06 ` Pedro Alves
2013-06-26 2:54 ` Hui Zhu
2013-05-22 14:04 ` Tom Tromey
2013-02-22 17:39 ` DPrintf feedback (was: [RFC] PR 15075 dprintf interferes with "next") Marc Khouzam
2013-02-22 19:32 ` DPrintf feedback Tom Tromey
2013-02-22 20:37 ` Marc Khouzam
2013-02-26 21:12 ` Marc Khouzam
2013-02-28 15:32 ` [PATCH 1/4] Fix dprintf bugs Yao Qi
2013-02-28 13:17 ` [PATCH 3/4] Test dprintf breakpoint works correctly with other breakpoints Yao Qi
2013-02-28 13:17 ` [PATCH 2/4] Test case of conditional dprintf Yao Qi
2013-02-28 14:48 ` [PATCH 4/4] Test case on setting dprintf commands Yao Qi
2013-02-28 16:30 ` [PATCH 1/4] Fix dprintf bugs Eli Zaretskii
2013-03-07 7:45 ` Yao Qi
2013-03-03 2:21 ` Marc Khouzam
2013-03-07 6:47 ` Yao Qi
2013-03-07 14:06 ` Marc Khouzam
2013-03-07 14:36 ` Yao Qi
2013-03-07 14:49 ` Marc Khouzam
2013-03-07 14:59 ` Yao Qi
2013-03-08 15:49 ` Marc Khouzam
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='CANFwon2oFaTc0ey4B7z=OXcne42SRwp_mozfOYKU4nx-K+80nA@mail.gmail.com' \
--to=teawater@gmail.com \
--cc=gdb-patches@sourceware.org \
--cc=keiths@redhat.com \
--cc=palves@redhat.com \
--cc=tromey@redhat.com \
--cc=yao@codesourcery.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