linux-kernel - [PATCH RESEND v3 11/18] EDAC/synopsys: Read full data pattern on errors

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Fri, 30 Sep 2022 02:35:23 +0300
From:   Serge Semin <Sergey.Semin@...kalelectronics.ru>
To:     Michal Simek <michal.simek@...inx.com>,
        Borislav Petkov <bp@...en8.de>,
        Mauro Carvalho Chehab <mchehab@...nel.org>,
        Tony Luck <tony.luck@...el.com>,
        James Morse <james.morse@....com>,
        Robert Richter <rric@...nel.org>
CC:     Serge Semin <Sergey.Semin@...kalelectronics.ru>,
        Serge Semin <fancer.lancer@...il.com>,
        Alexey Malahov <Alexey.Malahov@...kalelectronics.ru>,
        Michail Ivanov <Michail.Ivanov@...kalelectronics.ru>,
        Pavel Parkhomenko <Pavel.Parkhomenko@...kalelectronics.ru>,
        Punnaiah Choudary Kalluri 
        <punnaiah.choudary.kalluri@...inx.com>,
        Manish Narani <manish.narani@...inx.com>,
        Dinh Nguyen <dinguyen@...nel.org>,
        <linux-arm-kernel@...ts.infradead.org>,
        <linux-edac@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: [PATCH RESEND v3 11/18] EDAC/synopsys: Read full data pattern on errors

DW uMCTL2 DDRC calculates ECC for the DQ-bus word. If non-Full bus width
mode is activated the leftover DQ-bits will be padded with zeros, but the
ECC syndrome is calculated for the whole width anyway [1]. For some reason
the DW uMCTL2 DDRC driver currently doesn't read the whole SDRAM word in
case of the ECC errors even though the 64-bit DQ-bus has been supported
for a long time. Let's fix that by extending the data field of the ECC
error info structure and reading the upper 32-bits part of the data
pattern if an ECC error happens and the DDR controller has been configured
with the 64-bits DQ bus. As before the data will be printed as a part of
the custom error message passed to the edac_mc_handle_error() method.

[1] DesignWare® Cores Enhanced Universal DDR Memory Controller (uMCTL2)
    Databook, Version 3.91a, October 2020, p.424-425

Signed-off-by: Serge Semin <Sergey.Semin@...kalelectronics.ru>
---
 drivers/edac/synopsys_edac.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/edac/synopsys_edac.c b/drivers/edac/synopsys_edac.c
index 1cfbc5a2a564..00417f368893 100644
--- a/drivers/edac/synopsys_edac.c
+++ b/drivers/edac/synopsys_edac.c
@@ -309,7 +309,7 @@ struct snps_ecc_error_info {
 	u32 bank;
 	u32 bankgrp;
 	u32 bitpos;
-	u32 data;
+	u64 data;
 };
 
 /**
@@ -418,6 +418,8 @@ static int snps_get_error_info(struct snps_edac_priv *priv)
 	p->ceinfo.col = FIELD_GET(ECC_CEADDR1_COL_MASK, regval);
 
 	p->ceinfo.data = readl(base + ECC_CSYND0_OFST);
+	if (priv->info.dq_width == SNPS_DQ_64)
+		p->ceinfo.data |= (u64)readl(base + ECC_CSYND1_OFST) << 32;
 
 	edac_dbg(2, "ECCCSYN0: 0x%08X ECCCSYN1: 0x%08X ECCCSYN2: 0x%08X\n",
 		 readl(base + ECC_CSYND0_OFST), readl(base + ECC_CSYND1_OFST),
@@ -436,6 +438,8 @@ static int snps_get_error_info(struct snps_edac_priv *priv)
 	p->ueinfo.col = FIELD_GET(ECC_CEADDR1_COL_MASK, regval);
 
 	p->ueinfo.data = readl(base + ECC_UESYND0_OFST);
+	if (priv->info.dq_width == SNPS_DQ_64)
+		p->ueinfo.data |= (u64)readl(base + ECC_UESYND1_OFST) << 32;
 
 out:
 	spin_lock_irqsave(&priv->lock, flags);
@@ -466,7 +470,7 @@ static void snps_handle_error(struct mem_ctl_info *mci, struct snps_ecc_status *
 		pinf = &p->ceinfo;
 
 		snprintf(priv->message, SNPS_EDAC_MSG_SIZE,
-			 "Row %d Col %d Bank %d Bank Group %d Bit %d Data 0x%08x",
+			 "Row %d Col %d Bank %d Bank Group %d Bit %d Data 0x%08llx",
 			 pinf->row, pinf->col, pinf->bank, pinf->bankgrp,
 			 pinf->bitpos, pinf->data);
 
@@ -479,8 +483,9 @@ static void snps_handle_error(struct mem_ctl_info *mci, struct snps_ecc_status *
 		pinf = &p->ueinfo;
 
 		snprintf(priv->message, SNPS_EDAC_MSG_SIZE,
-			 "Row %d Col %d Bank %d Bank Group %d",
-			 pinf->row, pinf->col, pinf->bank, pinf->bankgrp);
+			 "Row %d Col %d Bank %d Bank Group %d Data 0x%08llx",
+			 pinf->row, pinf->col, pinf->bank, pinf->bankgrp,
+			 pinf->data);
 
 		edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci,
 				     p->ue_cnt, 0, 0, 0, 0, 0, -1,
-- 
2.37.3