sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling buggrafted

alb_break_llc() decides whether to break LLC preference during active
load balance. It does so by testing that every runnable fair task on the
source rq prefers its LLC:

	env->src_rq->nr_pref_llc_running == env->src_rq->cfs.h_nr_runnable

But the two counters cover different sets. nr_pref_llc_running is updated
in account_llc_enqueue()/account_llc_dequeue(), next to cfs_rq->nr_queued,
so it follows queued tasks. h_nr_runnable is updated in set_delayed()/
clear_delayed() and drops delay-dequeued tasks.

So under DELAY_DEQUEUE, a preferring task that goes to sleep stays counted
in nr_pref_llc_running while h_nr_runnable falls. The equality then breaks,
alb_break_llc() returns false, and active balance is free to pull a task
off its preferred LLC. Active balance only moves runnable tasks, and this
is the only LLC check it consults: once the stopper runs, LBF_ACTIVE_LB
skips the per-task test in can_migrate_task(). The runnable set is the one
we want.

Fix it on the counter side. A task should be counted in
nr_pref_llc_running exactly while it is both queued on its preferred LLC
(pref_llc_queued) and runnable (!sched_delayed). Define that membership
once in task_pref_llc_runnable(), and adjust the counter only through
pref_llc_running_inc()/pref_llc_running_dec() from the four sites that
change either input: account_llc_enqueue(), account_llc_dequeue(),
set_delayed() and clear_delayed(). Gating every update on the same
predicate keeps the delay, wake and dequeue paths from double-counting
or underflowing; see the comments at those sites for the ordering.

nr_llc_running and sd->llc_counts are not touched and stay on queued
semantics.

Fixes: 714059f79ff0 ("sched/cache: Handle moving single tasks to/from their preferred LLC")
Closes: https://lore.kernel.org/lkml/20260827135000.735138-1-zhanxusheng@xiaomi.com/
Reported-by: Zhan Xusheng <zhanxusheng@xiaomi.com>
Suggested-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Kayra Cizmeci <kayracizmeci@gmail.com>
Cc: <stable@kernel.org> # v7.2.x
Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com

1 files changed, 98 insertions, 0 deletions

diff --git a/scripts/checkincludes.pl b/scripts/checkincludes.pl
new file mode 100755
index 000000000..b514a956d
--- /dev/null
+++ b/scripts/checkincludes.pl
@@ -0,0 +1,98 @@
+#!/usr/bin/env perl
+# SPDX-License-Identifier: GPL-2.0
+#
+# checkincludes: find/remove files included more than once
+#
+# Copyright abandoned, 2000, Niels Kristian Bech Jensen <nkbj@image.dk>.
+# Copyright 2009 Luis R. Rodriguez <mcgrof@gmail.com>
+#
+# This script checks for duplicate includes. It also has support
+# to remove them in place. Note that this will not take into
+# consideration macros so you should run this only if you know
+# you do have real dups and do not have them under #ifdef's. You
+# could also just review the results.
+
+use strict;
+
+sub usage {
+	print "Usage: checkincludes.pl [-r]\n";
+	print "By default we just warn of duplicates\n";
+	print "To remove duplicated includes in place use -r\n";
+	exit 1;
+}
+
+my $remove = 0;
+
+if ($#ARGV < 0) {
+	usage();
+}
+
+if ($#ARGV >= 1) {
+	if ($ARGV[0] =~ /^-/) {
+		if ($ARGV[0] eq "-r") {
+			$remove = 1;
+			shift;
+		} else {
+			usage();
+		}
+	}
+}
+
+my $dup_counter = 0;
+
+foreach my $file (@ARGV) {
+	open(my $f, '<', $file)
+	    or die "Cannot open $file: $!.\n";
+
+	my %includedfiles = ();
+	my @file_lines = ();
+
+	while (<$f>) {
+		if (m/^\s*#\s*include\s*[<"](\S*)[>"]/o) {
+			++$includedfiles{$1};
+		}
+		push(@file_lines, $_);
+	}
+
+	close($f);
+
+	if (!$remove) {
+		foreach my $filename (keys %includedfiles) {
+			if ($includedfiles{$filename} > 1) {
+				print "$file: $filename is included more than once.\n";
+				++$dup_counter;
+			}
+		}
+		next;
+	}
+
+	open($f, '>', $file)
+	    or die("Cannot write to $file: $!");
+
+	my $dups = 0;
+	foreach (@file_lines) {
+		if (m/^\s*#\s*include\s*[<"](\S*)[>"]/o) {
+			foreach my $filename (keys %includedfiles) {
+				if ($1 eq $filename) {
+					if ($includedfiles{$filename} > 1) {
+						$includedfiles{$filename}--;
+						$dups++;
+						++$dup_counter;
+					} else {
+						print {$f} $_;
+					}
+				}
+			}
+		} else {
+			print {$f} $_;
+		}
+	}
+	if ($dups > 0) {
+		print "$file: removed $dups duplicate includes\n";
+	}
+	close($f);
+}
+
+if ($dup_counter == 0) {
+	print "No duplicate includes found.\n";
+}