Merge: skx_common: use driver decoder when possible

MR: https://gitlab.com/redhat/centos-stream/src/kernel/centos-stream-9/-/merge_requests/1466

Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2124913
Tested: by me, sanity

This patchset backports support for decoding errors in the driver when
possible instead of using BIOS, which improves the performance. As
production systems might see storms of corrected errors in thousands per
second, this can be very useful. Code will fall back to BIOS when driver
decoding is not possible.

Signed-off-by: Aristeu Rozanski <arozansk@redhat.com>

Approved-by: David Arcari <darcari@redhat.com>
Approved-by: Prarit Bhargava <prarit@redhat.com>
Approved-by: Rafael Aquini <aquini@redhat.com>
Approved-by: Herton R. Krzesinski <herton@redhat.com>

Signed-off-by: Frantisek Hrbata <fhrbata@redhat.com>

Author: Frantisek Hrbata

drivers/edac/skx_base.c

@@ -714,8 +714,13 @@ static int __init skx_init(void)
 
 	skx_set_decode(skx_decode, skx_show_retry_rd_err_log);
 
-	if (nvdimm_count && skx_adxl_get() == -ENODEV)
-		skx_printk(KERN_NOTICE, "Only decoding DDR4 address!\n");
+	if (nvdimm_count && skx_adxl_get() != -ENODEV) {
+		skx_set_decode(NULL, skx_show_retry_rd_err_log);
+	} else {
+		if (nvdimm_count)
+			skx_printk(KERN_NOTICE, "Only decoding DDR4 address!\n");
+		skx_set_decode(skx_decode, skx_show_retry_rd_err_log);
+	}
 
 	/* Ensure that the OPSTATE is set correctly for POLL or NMI */
 	opstate_init();

drivers/edac/skx_common.c

@@ -40,7 +40,7 @@ static char *adxl_msg;
 static unsigned long adxl_nm_bitmap;
 
 static char skx_msg[MSG_SIZE];
-static skx_decode_f skx_decode;
+static skx_decode_f driver_decode;
 static skx_show_retry_log_f skx_show_retry_rd_err_log;
 static u64 skx_tolm, skx_tohm;
 static LIST_HEAD(dev_edac_list);
@@ -173,6 +173,8 @@ static bool skx_adxl_decode(struct decoded_addr *res, bool error_in_1st_level_me
 			break;
 	}
 
+	res->decoded_by_adxl = true;
+
 	return true;
 }
 
@@ -183,7 +185,7 @@ void skx_set_mem_cfg(bool mem_cfg_2lm)
 
 void skx_set_decode(skx_decode_f decode, skx_show_retry_log_f show_retry_log)
 {
-	skx_decode = decode;
+	driver_decode = decode;
 	skx_show_retry_rd_err_log = show_retry_log;
 }
 
@@ -588,19 +590,19 @@ static void skx_mce_output_error(struct mem_ctl_info *mci,
 			break;
 		}
 	}
-	if (adxl_component_count) {
+	if (res->decoded_by_adxl) {
 		len = snprintf(skx_msg, MSG_SIZE, "%s%s err_code:0x%04x:0x%04x %s",
 			 overflow ? " OVERFLOW" : "",
 			 (uncorrected_error && recoverable) ? " recoverable" : "",
 			 mscod, errcode, adxl_msg);
 	} else {
 		len = snprintf(skx_msg, MSG_SIZE,
-			 "%s%s err_code:0x%04x:0x%04x socket:%d imc:%d rank:%d bg:%d ba:%d row:0x%x col:0x%x",
+			 "%s%s err_code:0x%04x:0x%04x ProcessorSocketId:0x%x MemoryControllerId:0x%x PhysicalRankId:0x%x Row:0x%x Column:0x%x Bank:0x%x BankGroup:0x%x",
 			 overflow ? " OVERFLOW" : "",
 			 (uncorrected_error && recoverable) ? " recoverable" : "",
 			 mscod, errcode,
 			 res->socket, res->imc, res->rank,
-			 res->bank_group, res->bank_address, res->row, res->column);
+			 res->row, res->column, res->bank_address, res->bank_group);
 	}
 
 	if (skx_show_retry_rd_err_log)
@@ -648,11 +650,11 @@ int skx_mce_check_error(struct notifier_block *nb, unsigned long val,
 	memset(&res, 0, sizeof(res));
 	res.addr = mce->addr;
 
-	if (adxl_component_count) {
-		if (!skx_adxl_decode(&res, skx_error_in_1st_level_mem(mce)))
+	/* Try driver decoder first */
+	if (!(driver_decode && driver_decode(&res))) {
+		/* Then try firmware decoder (ACPI DSM methods) */
+		if (!(adxl_component_count && skx_adxl_decode(&res, skx_error_in_1st_level_mem(mce))))
 			return NOTIFY_DONE;
-	} else if (!skx_decode || !skx_decode(&res)) {
-		return NOTIFY_DONE;
 	}
 
 	mci = res.dev->imc[res.imc].mci;

drivers/edac/skx_common.h

@@ -136,6 +136,7 @@ struct decoded_addr {
 	int	column;
 	int	bank_address;
 	int	bank_group;
+	bool	decoded_by_adxl;
 };
 
 struct res_config {