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, 95 insertions, 0 deletions

diff --git a/arch/arm/lib/putuser.S b/arch/arm/lib/putuser.S
new file mode 100644
index 000000000..bdd8836dc
--- /dev/null
+++ b/arch/arm/lib/putuser.S
@@ -0,0 +1,95 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ *  linux/arch/arm/lib/putuser.S
+ *
+ *  Copyright (C) 2001 Russell King
+ *
+ *  Idea from x86 version, (C) Copyright 1998 Linus Torvalds
+ *
+ * These functions have a non-standard call interface to make
+ * them more efficient, especially as they return an error
+ * value in addition to the "real" return value.
+ *
+ * __put_user_X
+ *
+ * Inputs:	r0 contains the address
+ *		r1 contains the address limit, which must be preserved
+ *		r2, r3 contains the value
+ * Outputs:	r0 is the error code
+ *		lr corrupted
+ *
+ * No other registers must be altered.  (see <asm/uaccess.h>
+ * for specific ASM register usage).
+ *
+ * Note that ADDR_LIMIT is either 0 or 0xc0000000
+ * Note also that it is intended that __put_user_bad is not global.
+ */
+#include <linux/linkage.h>
+#include <asm/assembler.h>
+#include <asm/errno.h>
+#include <asm/domain.h>
+
+ENTRY(__put_user_1)
+	check_uaccess r0, 1, r1, ip, __put_user_bad
+1: TUSER(strb)	r2, [r0]
+	mov	r0, #0
+	ret	lr
+ENDPROC(__put_user_1)
+
+ENTRY(__put_user_2)
+	check_uaccess r0, 2, r1, ip, __put_user_bad
+#if __LINUX_ARM_ARCH__ >= 6
+
+2: TUSER(strh)	r2, [r0]
+
+#else
+
+	mov	ip, r2, lsr #8
+#ifndef __ARMEB__
+2: TUSER(strb)	r2, [r0], #1
+3: TUSER(strb)	ip, [r0]
+#else
+2: TUSER(strb)	ip, [r0], #1
+3: TUSER(strb)	r2, [r0]
+#endif
+
+#endif /* __LINUX_ARM_ARCH__ >= 6 */
+	mov	r0, #0
+	ret	lr
+ENDPROC(__put_user_2)
+
+ENTRY(__put_user_4)
+	check_uaccess r0, 4, r1, ip, __put_user_bad
+4: TUSER(str)	r2, [r0]
+	mov	r0, #0
+	ret	lr
+ENDPROC(__put_user_4)
+
+ENTRY(__put_user_8)
+	check_uaccess r0, 8, r1, ip, __put_user_bad
+#ifdef CONFIG_THUMB2_KERNEL
+5: TUSER(str)	r2, [r0]
+6: TUSER(str)	r3, [r0, #4]
+#else
+5: TUSER(str)	r2, [r0], #4
+6: TUSER(str)	r3, [r0]
+#endif
+	mov	r0, #0
+	ret	lr
+ENDPROC(__put_user_8)
+
+__put_user_bad:
+	mov	r0, #-EFAULT
+	ret	lr
+ENDPROC(__put_user_bad)
+
+.pushsection __ex_table, "a"
+	.long	1b, __put_user_bad
+	.long	2b, __put_user_bad
+#if __LINUX_ARM_ARCH__ < 6
+	.long	3b, __put_user_bad
+#endif
+	.long	4b, __put_user_bad
+	.long	5b, __put_user_bad
+	.long	6b, __put_user_bad
+.popsection