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, 151 insertions, 0 deletions

diff --git a/drivers/video/fbdev/q40fb.c b/drivers/video/fbdev/q40fb.c
new file mode 100644
index 000000000..1ff8fa176
--- /dev/null
+++ b/drivers/video/fbdev/q40fb.c
@@ -0,0 +1,151 @@
+/*
+ * linux/drivers/video/q40fb.c -- Q40 frame buffer device
+ *
+ * Copyright (C) 2001
+ *
+ *      Richard Zidlicky <rz@linux-m68k.org>
+ *
+ *  This file is subject to the terms and conditions of the GNU General Public
+ *  License. See the file COPYING in the main directory of this archive for
+ *  more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/mm.h>
+#include <linux/delay.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+
+#include <linux/uaccess.h>
+#include <asm/setup.h>
+#include <asm/q40_master.h>
+#include <linux/fb.h>
+#include <linux/module.h>
+
+#define Q40_PHYS_SCREEN_ADDR 0xFE800000
+
+static struct fb_fix_screeninfo q40fb_fix = {
+	.id		= "Q40",
+	.smem_len	= 1024*1024,
+	.type		= FB_TYPE_PACKED_PIXELS,
+	.visual		= FB_VISUAL_TRUECOLOR,
+	.line_length	= 1024*2,
+	.accel		= FB_ACCEL_NONE,
+};
+
+static const struct fb_var_screeninfo q40fb_var = {
+	.xres		= 1024,
+	.yres		= 512,
+	.xres_virtual	= 1024,
+	.yres_virtual	= 512,
+	.bits_per_pixel	= 16,
+    	.red		= {6, 5, 0},
+	.green		= {11, 5, 0},
+	.blue		= {0, 6, 0},
+	.activate	= FB_ACTIVATE_NOW,
+	.height		= 230,
+	.width		= 300,
+	.vmode		= FB_VMODE_NONINTERLACED,
+};
+
+static int q40fb_setcolreg(unsigned regno, unsigned red, unsigned green,
+			   unsigned blue, unsigned transp,
+			   struct fb_info *info)
+{
+    /*
+     *  Set a single color register. The values supplied have a 16 bit
+     *  magnitude.
+     *  Return != 0 for invalid regno.
+     */
+
+    if (regno > 255)
+	    return 1;
+    red>>=11;
+    green>>=11;
+    blue>>=10;
+
+    if (regno < 16) {
+	((u32 *)info->pseudo_palette)[regno] = ((red & 31) <<6) |
+					       ((green & 31) << 11) |
+					       (blue & 63);
+    }
+    return 0;
+}
+
+static const struct fb_ops q40fb_ops = {
+	.owner		= THIS_MODULE,
+	FB_DEFAULT_IOMEM_OPS,
+	.fb_setcolreg	= q40fb_setcolreg,
+};
+
+static int q40fb_probe(struct platform_device *dev)
+{
+	struct fb_info *info;
+
+	if (!MACH_IS_Q40)
+		return -ENXIO;
+
+	/* mapped in q40/config.c */
+	q40fb_fix.smem_start = Q40_PHYS_SCREEN_ADDR;
+
+	info = framebuffer_alloc(sizeof(u32) * 16, &dev->dev);
+	if (!info)
+		return -ENOMEM;
+
+	info->var = q40fb_var;
+	info->fix = q40fb_fix;
+	info->fbops = &q40fb_ops;
+	info->pseudo_palette = info->par;
+	info->par = NULL;
+	info->screen_base = (char *) q40fb_fix.smem_start;
+
+	if (fb_alloc_cmap(&info->cmap, 256, 0) < 0) {
+		framebuffer_release(info);
+		return -ENOMEM;
+	}
+
+	master_outb(3, DISPLAY_CONTROL_REG);
+
+	if (register_framebuffer(info) < 0) {
+		printk(KERN_ERR "Unable to register Q40 frame buffer\n");
+		fb_dealloc_cmap(&info->cmap);
+		framebuffer_release(info);
+		return -EINVAL;
+	}
+
+	fb_info(info, "Q40 frame buffer alive and kicking !\n");
+	return 0;
+}
+
+static struct platform_driver q40fb_driver = {
+	.probe	= q40fb_probe,
+	.driver	= {
+		.name	= "q40fb",
+	},
+};
+
+static struct platform_device q40fb_device = {
+	.name	= "q40fb",
+};
+
+static int __init q40fb_init(void)
+{
+	int ret = 0;
+
+	if (fb_get_options("q40fb", NULL))
+		return -ENODEV;
+
+	ret = platform_driver_register(&q40fb_driver);
+
+	if (!ret) {
+		ret = platform_device_register(&q40fb_device);
+		if (ret)
+			platform_driver_unregister(&q40fb_driver);
+	}
+	return ret;
+}
+
+module_init(q40fb_init);
+MODULE_LICENSE("GPL");