x86/mce: Save and use APEI corrected threshold limit

The MCA threshold limit generally is not something that needs to change during
runtime. It is common for a system administrator to decide on a policy for
their managed systems.

If MCA thresholding is OS-managed, then the threshold limit must be set at
every boot. However, many systems allow the user to set a value in their BIOS.
And this is reported through an APEI HEST entry even if thresholding is not in
FW-First mode.

Use this value, if available, to set the OS-managed threshold limit.  Users
can still override it through sysfs if desired for testing or debug.

APEI is parsed after MCE is initialized. So reset the thresholding blocks
later to pick up the threshold limit.

Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov (AMD) <bp@alien8.de>
Link: https://lore.kernel.org/20251104-wip-mca-updates-v8-0-66c8eacf67b9@amd.com

1 files changed, 16 insertions, 2 deletions

diff --git a/arch/x86/kernel/cpu/mce/amd.c b/arch/x86/kernel/cpu/mce/amd.c
index 940d1a033569..af6e9b5ef288 100644
--- a/arch/x86/kernel/cpu/mce/amd.c
+++ b/arch/x86/kernel/cpu/mce/amd.c
@@ -489,6 +489,18 @@ static void threshold_restart_bank(unsigned int bank, bool intr_en)
 	}
 }
 
+/* Try to use the threshold limit reported through APEI. */
+static u16 get_thr_limit(void)
+{
+	u32 thr_limit = mce_get_apei_thr_limit();
+
+	/* Fallback to old default if APEI limit is not available. */
+	if (!thr_limit)
+		return THRESHOLD_MAX;
+
+	return min(thr_limit, THRESHOLD_MAX);
+}
+
 static void mce_threshold_block_init(struct threshold_block *b, int offset)
 {
 	struct thresh_restart tr = {
@@ -497,7 +509,7 @@ static void mce_threshold_block_init(struct threshold_block *b, int offset)
 		.lvt_off		= offset,
 	};
 
-	b->threshold_limit		= THRESHOLD_MAX;
+	b->threshold_limit		= get_thr_limit();
 	threshold_restart_block(&tr);
 };
 
@@ -1071,7 +1083,7 @@ static int allocate_threshold_blocks(unsigned int cpu, struct threshold_bank *tb
 	b->address		= address;
 	b->interrupt_enable	= 0;
 	b->interrupt_capable	= lvt_interrupt_supported(bank, high);
-	b->threshold_limit	= THRESHOLD_MAX;
+	b->threshold_limit	= get_thr_limit();
 
 	if (b->interrupt_capable) {
 		default_attrs[2] = &interrupt_enable.attr;
@@ -1082,6 +1094,8 @@ static int allocate_threshold_blocks(unsigned int cpu, struct threshold_bank *tb
 
 	list_add(&b->miscj, &tb->miscj);
 
+	mce_threshold_block_init(b, (high & MASK_LVTOFF_HI) >> 20);
+
 	err = kobject_init_and_add(&b->kobj, &threshold_ktype, tb->kobj, get_name(cpu, bank, b));
 	if (err)
 		goto out_free;