x86/amd_node: Fix PCI device reference counting in amd_smn_init()grafted

The local "root" pointer is a temporary variable used during the device
search. Therefore, refcount related to the search iterators should be cleaned
up after the search is complete.

Use the __free() cleanup macro to ensure the refcount is decremented when the
temporary pointer goes out of scope.

Additionally, increment the refcount when caching a root pointer. This ensures
the in-use refcount is separate from the temporary search refcounting.

Finally, drop the redundant "root = NULL" before the second search loop. The
pci_get_class() iterator always decrements the refcount of its "from"
argument, so the first loop can only fall through with "root" already NULL.

Fixes: 0a4b61d9c2e4 ("x86/amd_node: Fix AMD root device caching")
Closes: https://sashiko.dev/#/patchset/20260806160159.230453-1-jason.andryuk%40amd.com
Reported-by: Sashiko <sashiko-bot@kernel.org>
Assisted-by: LLM
Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Reviewed-by: Mario Limonciello (AMD) <superm1@kernel.org>
Cc: <stable@kernel.org>
Link: https://patch.msgid.link/20260903154325.74343-1-yazen.ghannam@amd.com

1 files changed, 56 insertions, 0 deletions

diff --git a/arch/alpha/lib/udelay.c b/arch/alpha/lib/udelay.c
new file mode 100644
index 000000000..873648263
--- /dev/null
+++ b/arch/alpha/lib/udelay.c
@@ -0,0 +1,56 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 1993, 2000 Linus Torvalds
+ *
+ * Delay routines, using a pre-computed "loops_per_jiffy" value.
+ */
+
+#include <linux/module.h>
+#include <linux/sched.h> /* for udelay's use of smp_processor_id */
+#include <asm/param.h>
+#include <asm/smp.h>
+#include <linux/delay.h>
+
+/*
+ * Use only for very small delays (< 1 msec). 
+ *
+ * The active part of our cycle counter is only 32-bits wide, and
+ * we're treating the difference between two marks as signed.  On
+ * a 1GHz box, that's about 2 seconds.
+ */
+
+void
+__delay(int loops)
+{
+	int tmp;
+	__asm__ __volatile__(
+		"	rpcc %0\n"
+		"	addl %1,%0,%1\n"
+		"1:	rpcc %0\n"
+		"	subl %1,%0,%0\n"
+		"	bgt %0,1b"
+		: "=&r" (tmp), "=r" (loops) : "1"(loops));
+}
+EXPORT_SYMBOL(__delay);
+
+#ifdef CONFIG_SMP
+#define LPJ	 cpu_data[smp_processor_id()].loops_per_jiffy
+#else
+#define LPJ	 loops_per_jiffy
+#endif
+
+void
+udelay(unsigned long usecs)
+{
+	usecs *= (((unsigned long)HZ << 32) / 1000000) * LPJ;
+	__delay((long)usecs >> 32);
+}
+EXPORT_SYMBOL(udelay);
+
+void
+ndelay(unsigned long nsecs)
+{
+	nsecs *= (((unsigned long)HZ << 32) / 1000000000) * LPJ;
+	__delay((long)nsecs >> 32);
+}
+EXPORT_SYMBOL(ndelay);