Add method/format information to =record-started
Commit Message
On 16-06-06 09:44 AM, Metzger, Markus T wrote:
> I was thinking about Eclipse. It has to handle unknown (e.g. new) formats, anyway.
Right.
Here's a v2:
From ce80ad77f163700e51c1e640ceb933b33ca1d280 Mon Sep 17 00:00:00 2001
From: Simon Marchi <simon.marchi@ericsson.com>
Date: Mon, 30 May 2016 17:29:39 -0400
Subject: [PATCH] Add method/format information to =record-started
Eclipse CDT now supports enabling execution recording using two methods
(full and btrace) and both formats for btrace (bts and pt). In the
event that recording is enabled behind the back of the GUI (by the user
on the command line, or a script), we need to know which method/format
are being used, so it can be correctly reflected in the interface. This
patch adds this information to the =record-started async record.
Before:
=record-started,thread-group="i1"
After:
=record-started,thread-group="i1",method="btrace",format="bts"
=record-started,thread-group="i1",method="btrace",format="pt"
=record-started,thread-group="i1",method="full"
The "format" field is only present when the current method supports
multiple formats (only the btrace method as of now).
gdb/ChangeLog:
* NEWS: Mention the new fields in =record-started.
* common/btrace-common.h (btrace_format_short_string): New function
declaration.
* common/btrace-common.c (btrace_format_short_string): New
function.
* mi/mi-interp.c (mi_record_changed): Output method and format
fields in the =record-started record.
* record-btrace.c (record_btrace_open): Adapt record_changed
notification.
* record-full.c (record_full_open): Likewise.
* record.c (cmd_record_stop): Likewise.
gdb/doc/ChangeLog:
* gdb.texinfo (GDB/MI Async Records): Document method and
format fields in =record-started.
* observer.texi (record_changed): Add method and format
parameters.
gdb/testsuite/ChangeLog:
* gdb.mi/mi-record-changed.exp: Adjust =record-started output
matching.
---
gdb/NEWS | 5 +++++
gdb/common/btrace-common.c | 20 ++++++++++++++++++++
gdb/common/btrace-common.h | 3 +++
gdb/doc/gdb.texinfo | 7 ++++++-
gdb/doc/observer.texi | 7 ++++++-
gdb/mi/mi-interp.c | 25 +++++++++++++++++++++----
gdb/record-btrace.c | 4 +++-
gdb/record-full.c | 2 +-
gdb/record.c | 2 +-
gdb/testsuite/gdb.mi/mi-record-changed.exp | 4 ++--
10 files changed, 68 insertions(+), 11 deletions(-)
Comments
> -----Original Message-----
> From: Simon Marchi [mailto:simon.marchi@ericsson.com]
> Sent: Monday, June 6, 2016 3:49 PM
> To: Metzger, Markus T <markus.t.metzger@intel.com>; gdb-
> patches@sourceware.org
> Subject: Re: [PATCH] Add method/format information to =record-started
> From ce80ad77f163700e51c1e640ceb933b33ca1d280 Mon Sep 17 00:00:00 2001
> From: Simon Marchi <simon.marchi@ericsson.com>
> Date: Mon, 30 May 2016 17:29:39 -0400
> Subject: [PATCH] Add method/format information to =record-started
>
> Eclipse CDT now supports enabling execution recording using two methods
> (full and btrace) and both formats for btrace (bts and pt). In the
> event that recording is enabled behind the back of the GUI (by the user
> on the command line, or a script), we need to know which method/format
> are being used, so it can be correctly reflected in the interface. This
> patch adds this information to the =record-started async record.
>
> Before:
>
> =record-started,thread-group="i1"
>
> After:
>
> =record-started,thread-group="i1",method="btrace",format="bts"
> =record-started,thread-group="i1",method="btrace",format="pt"
> =record-started,thread-group="i1",method="full"
>
> The "format" field is only present when the current method supports
> multiple formats (only the btrace method as of now).
The btrace bits look good to me.
I'm wondering if we want the short or the long form of the format. If we
expect MI front-ends to just print it as-is, the long form may be more useful.
That's also what's printed by "info record". I'm fine either way.
Regards,
Markus.
Intel Deutschland GmbH
Registered Address: Am Campeon 10-12, 85579 Neubiberg, Germany
Tel: +49 89 99 8853-0, www.intel.de
Managing Directors: Christin Eisenschmid, Christian Lamprechter
Chairperson of the Supervisory Board: Nicole Lau
Registered Office: Munich
Commercial Register: Amtsgericht Muenchen HRB 186928
On 16-06-06 10:02 AM, Metzger, Markus T wrote:
> I'm wondering if we want the short or the long form of the format. If we
> expect MI front-ends to just print it as-is, the long form may be more useful.
> That's also what's printed by "info record". I'm fine either way.
The main goal for this is to communicate to Eclipse which method is being used,
so it can update its UI as if the user had done the same action using the UI.
It's really for program/program communication, not program/human. If frontends
wanted a nice human readable string, we could always add it as an extra.
> -----Original Message-----
> From: Simon Marchi [mailto:simon.marchi@ericsson.com]
> Sent: Monday, June 6, 2016 4:28 PM
> To: Metzger, Markus T <markus.t.metzger@intel.com>; gdb-
> patches@sourceware.org
> Subject: Re: [PATCH] Add method/format information to =record-started
>
> On 16-06-06 10:02 AM, Metzger, Markus T wrote:
> > I'm wondering if we want the short or the long form of the format. If we
> > expect MI front-ends to just print it as-is, the long form may be more useful.
> > That's also what's printed by "info record". I'm fine either way.
>
> The main goal for this is to communicate to Eclipse which method is being used,
> so it can update its UI as if the user had done the same action using the UI.
> It's really for program/program communication, not program/human. If
> frontends
> wanted a nice human readable string, we could always add it as an extra.
OK.
Thanks,
Markus.
Intel Deutschland GmbH
Registered Address: Am Campeon 10-12, 85579 Neubiberg, Germany
Tel: +49 89 99 8853-0, www.intel.de
Managing Directors: Christin Eisenschmid, Christian Lamprechter
Chairperson of the Supervisory Board: Nicole Lau
Registered Office: Munich
Commercial Register: Amtsgericht Muenchen HRB 186928
> From: Simon Marchi <simon.marchi@ericsson.com>
> Date: Mon, 6 Jun 2016 09:49:01 -0400
>
> diff --git a/gdb/NEWS b/gdb/NEWS
> index dce79a2..253c8e5 100644
> --- a/gdb/NEWS
> +++ b/gdb/NEWS
> @@ -67,6 +67,11 @@ maint selftest
> including JIT compiling fast tracepoint's conditional expression
> bytecode into native code.
>
> +* MI async record =record-started now includes the method and format used for
> + recording. For example:
> +
> + =record-started,thread-group="i1",method="btrace",format="bts"
> +
> *** Changes in GDB 7.11
This is OK.
> diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
> index 7e89003..f50582a 100644
> --- a/gdb/doc/gdb.texinfo
> +++ b/gdb/doc/gdb.texinfo
> @@ -26423,12 +26423,17 @@ breakpoint commands; @xref{GDB/MI Breakpoint Commands}. The
> Note that if a breakpoint is emitted in the result record of a
> command, then it will not also be emitted in an async record.
>
> -@item =record-started,thread-group="@var{id}"
> +@item =record-started,thread-group="@var{id}",method="@var{method}"[,format="@var{format}"]
> @itemx =record-stopped,thread-group="@var{id}"
> Execution log recording was either started or stopped on an
> inferior. The @var{id} is the @value{GDBN} identifier of the thread
> group corresponding to the affected inferior.
>
> +The @var{method} field indicates the method used to record execution. If the
> +method in use supports multiple recording formats, @var{format} will be present
> +and contain the currently used format. See @xref{Process Record and Replay}
> +for existing method and format values. ^^^
No need for "See" there, as @xref generates that automatically. Or
use "See @ref".
OK for the documentation parts, with the above gotcha fixed.
On 16-06-06 11:13 AM, Eli Zaretskii wrote:
>> From: Simon Marchi <simon.marchi@ericsson.com>
>> Date: Mon, 6 Jun 2016 09:49:01 -0400
>>
>> diff --git a/gdb/NEWS b/gdb/NEWS
>> index dce79a2..253c8e5 100644
>> --- a/gdb/NEWS
>> +++ b/gdb/NEWS
>> @@ -67,6 +67,11 @@ maint selftest
>> including JIT compiling fast tracepoint's conditional expression
>> bytecode into native code.
>>
>> +* MI async record =record-started now includes the method and format used for
>> + recording. For example:
>> +
>> + =record-started,thread-group="i1",method="btrace",format="bts"
>> +
>> *** Changes in GDB 7.11
>
> This is OK.
>
>> diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
>> index 7e89003..f50582a 100644
>> --- a/gdb/doc/gdb.texinfo
>> +++ b/gdb/doc/gdb.texinfo
>> @@ -26423,12 +26423,17 @@ breakpoint commands; @xref{GDB/MI Breakpoint Commands}. The
>> Note that if a breakpoint is emitted in the result record of a
>> command, then it will not also be emitted in an async record.
>>
>> -@item =record-started,thread-group="@var{id}"
>> +@item =record-started,thread-group="@var{id}",method="@var{method}"[,format="@var{format}"]
>> @itemx =record-stopped,thread-group="@var{id}"
>> Execution log recording was either started or stopped on an
>> inferior. The @var{id} is the @value{GDBN} identifier of the thread
>> group corresponding to the affected inferior.
>>
>> +The @var{method} field indicates the method used to record execution. If the
>> +method in use supports multiple recording formats, @var{format} will be present
>> +and contain the currently used format. See @xref{Process Record and Replay}
>> +for existing method and format values. ^^^
>
> No need for "See" there, as @xref generates that automatically. Or
> use "See @ref".
>
> OK for the documentation parts, with the above gotcha fixed.
Oh it slipped passed my eyes when reading, but you're right.
Thanks, this is now pushed.
@@ -67,6 +67,11 @@ maint selftest
including JIT compiling fast tracepoint's conditional expression
bytecode into native code.
+* MI async record =record-started now includes the method and format used for
+ recording. For example:
+
+ =record-started,thread-group="i1",method="btrace",format="bts"
+
*** Changes in GDB 7.11
* GDB now supports debugging kernel-based threads on FreeBSD.
@@ -43,6 +43,26 @@ btrace_format_string (enum btrace_format format)
/* See btrace-common.h. */
+const char *
+btrace_format_short_string (enum btrace_format format)
+{
+ switch (format)
+ {
+ case BTRACE_FORMAT_NONE:
+ return "unknown";
+
+ case BTRACE_FORMAT_BTS:
+ return "bts";
+
+ case BTRACE_FORMAT_PT:
+ return "pt";
+ }
+
+ internal_error (__FILE__, __LINE__, _("Unknown branch trace format"));
+}
+
+/* See btrace-common.h. */
+
void
btrace_data_init (struct btrace_data *data)
{
@@ -214,6 +214,9 @@ enum btrace_error
/* Return a string representation of FORMAT. */
extern const char *btrace_format_string (enum btrace_format format);
+/* Return an abbreviation string representation of FORMAT. */
+extern const char *btrace_format_short_string (enum btrace_format format);
+
/* Initialize DATA. */
extern void btrace_data_init (struct btrace_data *data);
@@ -26423,12 +26423,17 @@ breakpoint commands; @xref{GDB/MI Breakpoint Commands}. The
Note that if a breakpoint is emitted in the result record of a
command, then it will not also be emitted in an async record.
-@item =record-started,thread-group="@var{id}"
+@item =record-started,thread-group="@var{id}",method="@var{method}"[,format="@var{format}"]
@itemx =record-stopped,thread-group="@var{id}"
Execution log recording was either started or stopped on an
inferior. The @var{id} is the @value{GDBN} identifier of the thread
group corresponding to the affected inferior.
+The @var{method} field indicates the method used to record execution. If the
+method in use supports multiple recording formats, @var{format} will be present
+and contain the currently used format. See @xref{Process Record and Replay}
+for existing method and format values.
+
@item =cmd-param-changed,param=@var{param},value=@var{value}
Reports that a parameter of the command @code{set @var{param}} is
changed to @var{value}. In the multi-word @code{set} command,
@@ -141,11 +141,16 @@ at the entry-point instruction. For @samp{attach} and @samp{core},
inferior, and before any information on the inferior has been printed.
@end deftypefun
-@deftypefun void record_changed (struct inferior *@var{inferior}, int @var{started})
+@deftypefun void record_changed (struct inferior *@var{inferior}, int @var{started}, const char *@var{method}, const char *@var{format})
The status of process record for inferior @var{inferior} in
@value{GDBN} has changed. The process record is started if
@var{started} is true, and the process record is stopped if
@var{started} is false.
+
+When @var{started} is true, @var{method} indicates the short name of the method
+used for recording. If the method supports multiple formats, @var{format}
+indicates which one is being used, otherwise it is NULL. When @var{started} is
+false, they are both NULL.
@end deftypefun
@deftypefun void solib_loaded (struct so_list *@var{solib})
@@ -65,7 +65,8 @@ static void mi_on_no_history (void);
static void mi_new_thread (struct thread_info *t);
static void mi_thread_exit (struct thread_info *t, int silent);
-static void mi_record_changed (struct inferior*, int);
+static void mi_record_changed (struct inferior*, int, const char *,
+ const char *);
static void mi_inferior_added (struct inferior *inf);
static void mi_inferior_appeared (struct inferior *inf);
static void mi_inferior_exit (struct inferior *inf);
@@ -399,7 +400,8 @@ mi_thread_exit (struct thread_info *t, int silent)
/* Emit notification on changing the state of record. */
static void
-mi_record_changed (struct inferior *inferior, int started)
+mi_record_changed (struct inferior *inferior, int started, const char *method,
+ const char *format)
{
struct mi_interp *mi = (struct mi_interp *) top_level_interpreter_data ();
struct cleanup *old_chain;
@@ -407,8 +409,23 @@ mi_record_changed (struct inferior *inferior, int started)
old_chain = make_cleanup_restore_target_terminal ();
target_terminal_ours_for_output ();
- fprintf_unfiltered (mi->event_channel, "record-%s,thread-group=\"i%d\"",
- started ? "started" : "stopped", inferior->num);
+ if (started)
+ {
+ if (format != NULL)
+ fprintf_unfiltered (
+ mi->event_channel,
+ "record-started,thread-group=\"i%d\",method=\"%s\",format=\"%s\"",
+ inferior->num, method, format);
+ else
+ fprintf_unfiltered (
+ mi->event_channel,
+ "record-started,thread-group=\"i%d\",method=\"%s\"",
+ inferior->num, method);
+ }
+ else
+ fprintf_unfiltered (mi->event_channel,
+ "record-stopped,thread-group=\"i%d\"", inferior->num);
+
gdb_flush (mi->event_channel);
@@ -206,6 +206,7 @@ record_btrace_open (const char *args, int from_tty)
{
struct cleanup *disable_chain;
struct thread_info *tp;
+ const char *format;
DEBUG ("open");
@@ -234,7 +235,8 @@ record_btrace_open (const char *args, int from_tty)
NULL);
record_btrace_generating_corefile = 0;
- observer_notify_record_changed (current_inferior (), 1);
+ format = btrace_format_short_string (record_btrace_conf.format);
+ observer_notify_record_changed (current_inferior (), 1, "btrace", format);
discard_cleanups (disable_chain);
}
@@ -869,7 +869,7 @@ record_full_open (const char *name, int from_tty)
record_full_init_record_breakpoints ();
- observer_notify_record_changed (current_inferior (), 1);
+ observer_notify_record_changed (current_inferior (), 1, "full", NULL);
}
/* "to_close" target method. Close the process record target. */
@@ -268,7 +268,7 @@ cmd_record_stop (char *args, int from_tty)
printf_unfiltered (_("Process record is stopped and all execution "
"logs are deleted.\n"));
- observer_notify_record_changed (current_inferior (), 0);
+ observer_notify_record_changed (current_inferior (), 0, NULL, NULL);
}
/* The "set record" command. */
@@ -31,14 +31,14 @@ if [mi_gdb_start] {
}
mi_run_to_main
-mi_gdb_test "record" ".*=record-started,thread-group=\"i${decimal}\".*\\^done" \
+mi_gdb_test "record" ".*=record-started,thread-group=\"i${decimal}\",method=\"full\".*\\^done" \
"record"
mi_gdb_test "record stop" \
".*=record-stopped,thread-group=\"i${decimal}\".*\\^done" \
"record end"
mi_gdb_test "target record" \
- ".*=record-started,thread-group=\"i${decimal}\".*\\^done" \
+ ".*=record-started,thread-group=\"i${decimal}\",method=\"full\".*\\^done" \
"target record"
return 0