From: Doug Evans <dje@google.com>
To: Pedro Alves <palves@redhat.com>
Cc: Jan Kratochvil <jan.kratochvil@redhat.com>,
eliz@gnu.org, gdb-patches@sourceware.org
Subject: Re: [PATCH] Re: Regression for GDB global --statistics
Date: Sat, 12 Jul 2014 00:42:00 -0000 [thread overview]
Message-ID: <21440.24455.769287.540494@ruffy.mtv.corp.google.com> (raw)
In-Reply-To: <53BEC275.8000004@redhat.com>
Pedro Alves writes:
> Hi Doug,
>
> On 07/09/2014 10:10 PM, Doug Evans wrote:
> > Here's a patch.
> >
> > 2014-07-09 Doug Evans <dje@google.com>
> >
> > * maint.c (count_symtabs_and_blocks): Handle NULL
> > current_program_space.
>
> It's not obvious to me how this can be NULL, given we initialize
> it so early. If there's a good reason, could you please add
> a comment here mentioning it?
2014-07-11 Doug Evans <dje@google.com>
* maint.c (count_symtabs_and_blocks): Handle NULL
current_program_space.
(report_command_stats): Check global enabled flag in addition to
recorded enabled flag.
(make_command_stats_cleanup): Handle msg_type == 0, startup.
testsuite/
* gdb.base/maint.exp: Update testing of per-command stats.
diff --git a/gdb/maint.c b/gdb/maint.c
index 5f34af3..c7a937c 100644
--- a/gdb/maint.c
+++ b/gdb/maint.c
@@ -830,13 +830,19 @@ count_symtabs_and_blocks (int *nr_symtabs_ptr, int *nr_primary_symtabs_ptr,
int nr_primary_symtabs = 0;
int nr_blocks = 0;
- ALL_SYMTABS (o, s)
+ /* When collecting statistics during startup, this is called before
+ pretty much anything in gdb has been initialized, and thus
+ current_program_space may be NULL. */
+ if (current_program_space != NULL)
{
- ++nr_symtabs;
- if (s->primary)
+ ALL_SYMTABS (o, s)
{
- ++nr_primary_symtabs;
- nr_blocks += BLOCKVECTOR_NBLOCKS (BLOCKVECTOR (s));
+ ++nr_symtabs;
+ if (s->primary)
+ {
+ ++nr_primary_symtabs;
+ nr_blocks += BLOCKVECTOR_NBLOCKS (BLOCKVECTOR (s));
+ }
}
}
@@ -856,7 +862,7 @@ report_command_stats (void *arg)
struct cmd_stats *start_stats = (struct cmd_stats *) arg;
int msg_type = start_stats->msg_type;
- if (start_stats->time_enabled)
+ if (start_stats->time_enabled && per_command_time)
{
long cmd_time = get_run_time () - start_stats->start_cpu_time;
struct timeval now_wall_time, delta_wall_time, wait_time;
@@ -877,7 +883,7 @@ report_command_stats (void *arg)
(long) delta_wall_time.tv_usec);
}
- if (start_stats->space_enabled)
+ if (start_stats->space_enabled && per_command_space)
{
#ifdef HAVE_SBRK
char *lim = (char *) sbrk (0);
@@ -894,7 +900,7 @@ report_command_stats (void *arg)
#endif
}
- if (start_stats->symtab_enabled)
+ if (start_stats->symtab_enabled && per_command_symtab)
{
int nr_symtabs, nr_primary_symtabs, nr_blocks;
@@ -920,8 +926,14 @@ make_command_stats_cleanup (int msg_type)
{
struct cmd_stats *new_stat;
- /* Early exit if we're not reporting any stats. */
- if (!per_command_time
+ /* Early exit if we're not reporting any stats. It can be expensive to
+ compute the pre-command values so don't collect them at all if we're
+ not reporting stats. Alas this doesn't work in the startup case because
+ we don't know yet whether we will be reporting the stats. For the
+ startup case collect the data anyway (it should be cheap at this point),
+ and leave it to the reporter to decide whether to print them. */
+ if (msg_type != 0
+ && !per_command_time
&& !per_command_space
&& !per_command_symtab)
return make_cleanup (null_cleanup, 0);
@@ -930,7 +942,7 @@ make_command_stats_cleanup (int msg_type)
new_stat->msg_type = msg_type;
- if (per_command_space)
+ if (msg_type == 0 || per_command_space)
{
#ifdef HAVE_SBRK
char *lim = (char *) sbrk (0);
@@ -939,14 +951,14 @@ make_command_stats_cleanup (int msg_type)
#endif
}
- if (per_command_time)
+ if (msg_type == 0 || per_command_time)
{
new_stat->start_cpu_time = get_run_time ();
gettimeofday (&new_stat->start_wall_time, NULL);
new_stat->time_enabled = 1;
}
- if (per_command_symtab)
+ if (msg_type == 0 || per_command_symtab)
{
int nr_symtabs, nr_primary_symtabs, nr_blocks;
diff --git a/gdb/testsuite/gdb.base/maint.exp b/gdb/testsuite/gdb.base/maint.exp
index 64753b7..21d0a31 100644
--- a/gdb/testsuite/gdb.base/maint.exp
+++ b/gdb/testsuite/gdb.base/maint.exp
@@ -130,10 +130,11 @@ gdb_expect {
# tests here!!
gdb_test_no_output "maint check-symtabs"
+# Test per-command stats.
gdb_test_no_output "maint set per-command on"
-
-gdb_test "maint set per-command off" \
+gdb_test "pwd" \
"Command execution time: \[0-9.\]+ \\(cpu\\), \[0-9.\]+ \\(wall\\)\[\r\n\]+Space used: $decimal \\(\\+$decimal for this command\\)\[\r\n\]+#symtabs: $decimal \\(\\+$decimal\\), #primary symtabs: $decimal \\(\\+$decimal\\), #blocks: $decimal \\(\\+$decimal\\)"
+gdb_test_no_output "maint set per-command off"
gdb_test "maint demangle" \
"\"maintenance demangle\" takes an argument to demangle\\."
next prev parent reply other threads:[~2014-07-11 22:05 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-03-15 23:44 [doc RFA] New commands: mt set per-command on|off Doug Evans
2013-03-17 19:04 ` Eli Zaretskii
[not found] ` <20807.35450.682935.524373@ruffy2.mtv.corp.google.com>
2013-03-20 17:24 ` Eli Zaretskii
2013-03-21 17:54 ` Doug Evans
2013-03-28 23:27 ` doc/ build regression with texinfo-5.1 [Re: [doc RFA] New commands: mt set per-command on|off] Jan Kratochvil
2013-03-29 0:27 ` Eli Zaretskii
2013-03-29 16:50 ` Jan Kratochvil
2013-03-29 16:57 ` Doug Evans
2013-03-29 16:58 ` Jan Kratochvil
2013-03-29 21:50 ` Eli Zaretskii
2014-06-30 17:13 ` Regression for GDB global --statistics " Jan Kratochvil
2014-07-09 21:11 ` [PATCH] Re: Regression for GDB global --statistics Doug Evans
2014-07-10 16:42 ` Pedro Alves
2014-07-12 0:42 ` Doug Evans [this message]
2014-07-15 9:43 ` Pedro Alves
2014-07-17 12:02 ` Doug Evans
2014-07-11 21:33 ` Jan Kratochvil
2014-07-11 22:05 ` Doug Evans
2014-07-12 20:25 ` [testsuite patch] " Jan Kratochvil
[not found] ` <CADPb22SxWYQ-6tK85p2koVQjxrFF0o3OtmEd79R0q-_xWcJ+Mg@mail.gmail.com>
2014-07-17 12:23 ` [testsuite commit] " Jan Kratochvil
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=21440.24455.769287.540494@ruffy.mtv.corp.google.com \
--to=dje@google.com \
--cc=eliz@gnu.org \
--cc=gdb-patches@sourceware.org \
--cc=jan.kratochvil@redhat.com \
--cc=palves@redhat.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).