Igt-dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Tvrtko Ursulin <tursulin@ursulin.net>
To: igt-dev@lists.freedesktop.org
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Subject: [igt-dev] [PATCH i-g-t 11/11] trace.pl: Fix request split mode
Date: Tue, 30 Jan 2018 10:11:19 +0000	[thread overview]
Message-ID: <20180130101119.15758-14-tvrtko.ursulin@linux.intel.com> (raw)
In-Reply-To: <20180130101119.15758-1-tvrtko.ursulin@linux.intel.com>

From: Tvrtko Ursulin <tvrtko.ursulin@intel.com>

Request split mode had several bugs, both in the original version and also
after the recent refactorings.

One big one was that it wasn't considering different submit ports as a
reason to split execution, and also that it was too time based instead of
looking at relevant timelines.

In this refactoring we address the former by using the engine timelines
introduced in the previous patch. Secondary port submissions are moved
to follow the preceding submission as a first step in the correction
process.

In the second step, we add context timelines and use then in a similar
fashion to separate start and end time of coalesced requests. For each
coalesced request we know its boundaries by looking at the engine
timeline (via global seqnos), and we know the previous request it should
only start after, by looking at the context timeline.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: John Harrison <John.C.Harrison@intel.com>
---
 scripts/trace.pl | 129 ++++++++++++++++++++++++++++++++++++++++++++-----------
 1 file changed, 103 insertions(+), 26 deletions(-)

diff --git a/scripts/trace.pl b/scripts/trace.pl
index 768bb015da99..653215819f34 100755
--- a/scripts/trace.pl
+++ b/scripts/trace.pl
@@ -572,41 +572,118 @@ sub sortStart {
 	return $val;
 }
 
-my @sorted_keys = sort sortStart keys %db;
-my $re_sort = 0;
+my $re_sort = 1;
+my @sorted_keys;
 
-die "Database changed size?!" unless scalar(@sorted_keys) == $key_count;
+sub maybe_sort_keys
+{
+	if ($re_sort) {
+		@sorted_keys = sort sortStart keys %db;
+		$re_sort = 0;
+		die "Database changed size?!" unless scalar(@sorted_keys) ==
+						     $key_count;
+	}
+}
 
-foreach my $key (@sorted_keys) {
-	my $ring = $db{$key}->{'ring'};
-	my $end = $db{$key}->{'end'};
+maybe_sort_keys();
+
+my %ctx_timelines;
+
+sub sortContext {
+	my $as = $db{$a}->{'seqno'};
+	my $bs = $db{$b}->{'seqno'};
+	my $val;
 
-	# correct duration of merged batches
-	if ($correct_durations and exists $db{$key}->{'no-end'}) {
+	$val = $as <=> $bs;
+
+	die if $val == 0;
+
+	return $val;
+}
+
+sub get_ctx_timeline {
+	my ($ctx, $ring, $key) = @_;
+	my @timeline;
+
+	return $ctx_timelines{$key} if exists $ctx_timelines{$key};
+
+	@timeline = grep { $db{$_}->{'ring'} == $ring and
+			   $db{$_}->{'ctx'} == $ctx } @sorted_keys;
+	# FIXME seqno restart
+	@timeline = sort sortContext @timeline;
+
+	$ctx_timelines{$key} = \@timeline;
+
+	return \@timeline;
+}
+
+# Split out merged batches if requested.
+if ($correct_durations) {
+	my @port1;
+	my @no_end;
+
+	# Shift port1 requests start time to after the previous port0 context on
+	# the same timeline has finished.
+	@port1 = grep { $db{$_}->{'port'} == 1 } @sorted_keys;
+	foreach my $key (@port1) {
+		my $timeline = get_engine_timeline($db{$key}->{'ring'});
+		my $i = List::Util::first { ${$timeline}[$_] eq $key } 0..$#{$timeline};
+		my $prev_key;
+		my $start;
+
+		die if $i == 0;
+
+		$prev_key = ${$timeline}[$i - 1];
+		$start = $db{$prev_key}->{'end'};
+		$db{$key}->{'start'} = $start;
+		die if $start > $db{$key}->{'end'};
+
+		# Timeline has just been modified so we'll need to adjust it.
+		# First delete the current key from the timeline.
+		splice @{$timeline}, $i, 1;
+		# Now find the new correct position for the key and insert it
+		# there.
+		while ($i < $#{$timeline}) {
+			last if $db{${$timeline}[$i]}->{'global'} >= $db{$key}->{'global'};
+			$i = $i + 1;
+		}
+		splice @{$timeline}, $i, 0, $key;
+		$re_sort = 1;
+	}
+
+	maybe_sort_keys();
+
+	# Batch with no-end (no request_out) means it was submitted as part of
+	# colaesced context. This means it's start time should be set to the end
+	# time of a precedeing request on this timeline.
+	@no_end = grep { exists $db{$_}->{'no-end'} } @sorted_keys;
+	foreach my $key (@no_end) {
 		my $ctx = $db{$key}->{'ctx'};
-		my $seqno = $db{$key}->{'seqno'};
-		my $start = $db{$key}->{'start'};
-		my $next_key;
-		my $i = 1;
-
-		do {
-			$next_key = db_key($ring, $ctx, $seqno + $i);
-			$i++;
-		} until (exists $db{$next_key} or $i > $key_count);  # ugly stop hack
-
-		# 20us tolerance
-		if (exists $db{$next_key} and $db{$next_key}->{'start'} < $start + 20) {
-			my $notify = $db{$key}->{'notify'};
+		my $ring = $db{$key}->{'ring'};
+		my $tkey = $ctx . '/' . $ring;
+		my $timeline = get_ctx_timeline($ctx, $ring, $tkey);
+		my $i;
+
+		$i = List::Util::first { ${$timeline}[$_] eq $key } 0..$#{$timeline};
+		if ($i > 0) {
+			my $prev_key = ${$timeline}[$i - 1];
+
+			if ($db{$prev_key}->{'global'} == $db{$key}->{'global'}) {
+				$db{$key}->{'start'} = $db{$prev_key}->{'notify'};
+				$re_sort = 1;
+			}
+		}
+
+		if ($i < $#{$timeline}) {
+			my $next_key = ${$timeline}[$i + 1];
+
+			$db{$next_key}->{'start'} = $db{$key}->{'notify'};
 			$re_sort = 1;
-			$db{$next_key}->{'start'} = $notify;
-			$db{$next_key}->{'start'} = $db{$next_key}->{'end'} if $db{$next_key}->{'start'} > $db{$next_key}->{'end'};
-			die if $db{$next_key}->{'start'} > $db{$next_key}->{'end'};
 		}
-		die if $start > $end;
 	}
 }
 
-@sorted_keys = sort sortStart keys %db if $re_sort;
+maybe_sort_keys();
 
 # GPU time accounting
 my (%running, %runnable, %queued, %batch_avg, %batch_total_avg, %batch_count);
-- 
2.14.1

_______________________________________________
igt-dev mailing list
igt-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/igt-dev

  parent reply	other threads:[~2018-01-30 10:11 UTC|newest]

Thread overview: 21+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-01-30 10:11 [igt-dev] [PATCH i-g-t 00/11] trace.pl fixes Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 01/11] scripts/trace.pl: More hash key optimisations Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 02/11] scripts/trace.pl: Sort order Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 03/11] scripts/trace.pl: Calculate stats only after all munging Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 04/11] scripts/trace.pl: Simplify 'end' & 'notify' generation Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 05/11] trace.pl: Use context hw_id as context id Tvrtko Ursulin
2018-01-30 12:29   ` Lionel Landwerlin
2018-02-06 18:37     ` Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 06/11] trace.pl: Move sortQueue near its user Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 06/11] trace.pl: Move sortQueue near to the user Tvrtko Ursulin
2018-01-30 10:34   ` Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 07/11] trace.pl: Move min/max timestamp lookup to last loop Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 08/11] trace.pl: Fix engine busy accounting in split mode Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 08/11] trace.pl: Split " Tvrtko Ursulin
2018-01-30 10:34   ` Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 09/11] trace.pl: Add support for colouring context execution Tvrtko Ursulin
2018-01-30 10:11 ` [igt-dev] [PATCH i-g-t 10/11] trace.pl: Fix incomplete request handling Tvrtko Ursulin
2018-01-30 10:11 ` Tvrtko Ursulin [this message]
2018-01-30 15:05 ` [igt-dev] ✓ Fi.CI.BAT: success for trace.pl fixes Patchwork
  -- strict thread matches above, loose matches on Subject: below --
2018-03-06 12:43 [Intel-gfx] [PATCH i-g-t v2 00/11] " Tvrtko Ursulin
2018-03-06 12:43 ` [igt-dev] [PATCH i-g-t 11/11] trace.pl: Fix request split mode Tvrtko Ursulin
2018-07-09 13:19 [Intel-gfx] [PATCH i-g-t 00/11] trace.pl fixes and improvements Tvrtko Ursulin
2018-07-09 13:20 ` [igt-dev] [PATCH i-g-t 11/11] trace.pl: Fix request split mode Tvrtko Ursulin

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=20180130101119.15758-14-tvrtko.ursulin@linux.intel.com \
    --to=tursulin@ursulin.net \
    --cc=igt-dev@lists.freedesktop.org \
    --cc=tvrtko.ursulin@intel.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