@@ -9,6 +9,7 @@
#include <linux/edac.h>
#include <linux/module.h>
#include <linux/platform_device.h>
+#include <linux/spinlock.h>
#include <linux/interrupt.h>
#include <linux/of.h>
#include <linux/of_device.h>
@@ -300,6 +301,7 @@ struct synps_ecc_status {
/**
* struct synps_edac_priv - DDR memory controller private instance data.
* @baseaddr: Base address of the DDR controller.
+ * @reglock: Concurrent CSRs access lock.
* @message: Buffer for framing the event specific info.
* @stat: ECC status information.
* @p_data: Platform data.
@@ -314,6 +316,7 @@ struct synps_ecc_status {
*/
struct synps_edac_priv {
void __iomem *baseaddr;
+ spinlock_t reglock;
char message[SYNPS_EDAC_MSG_SIZE];
struct synps_ecc_status stat;
const struct synps_platform_data *p_data;
@@ -409,7 +412,8 @@ static int zynq_get_error_info(struct synps_edac_priv *priv)
static int zynqmp_get_error_info(struct synps_edac_priv *priv)
{
struct synps_ecc_status *p;
- u32 regval, clearval = 0;
+ u32 regval, clearval;
+ unsigned long flags;
void __iomem *base;
base = priv->baseaddr;
@@ -453,10 +457,14 @@ static int zynqmp_get_error_info(struct synps_edac_priv *priv)
p->ueinfo.blknr = (regval & ECC_CEADDR1_BLKNR_MASK);
p->ueinfo.data = readl(base + ECC_UESYND0_OFST);
out:
- clearval = ECC_CTRL_CLR_CE_ERR | ECC_CTRL_CLR_CE_ERRCNT;
- clearval |= ECC_CTRL_CLR_UE_ERR | ECC_CTRL_CLR_UE_ERRCNT;
+ spin_lock_irqsave(&priv->reglock, flags);
+
+ clearval = readl(base + ECC_CLR_OFST) |
+ ECC_CTRL_CLR_CE_ERR | ECC_CTRL_CLR_CE_ERRCNT |
+ ECC_CTRL_CLR_UE_ERR | ECC_CTRL_CLR_UE_ERRCNT;
writel(clearval, base + ECC_CLR_OFST);
- writel(0x0, base + ECC_CLR_OFST);
+
+ spin_unlock_irqrestore(&priv->reglock, flags);
return 0;
}
@@ -516,24 +524,41 @@ static void handle_error(struct mem_ctl_info *mci, struct synps_ecc_status *p)
static void enable_intr(struct synps_edac_priv *priv)
{
+ unsigned long flags;
+
/* Enable UE/CE Interrupts */
- if (priv->p_data->quirks & DDR_ECC_INTR_SELF_CLEAR)
- writel(DDR_UE_MASK | DDR_CE_MASK,
- priv->baseaddr + ECC_CLR_OFST);
- else
+ if (!(priv->p_data->quirks & DDR_ECC_INTR_SELF_CLEAR)) {
writel(DDR_QOSUE_MASK | DDR_QOSCE_MASK,
priv->baseaddr + DDR_QOS_IRQ_EN_OFST);
+ return;
+ }
+
+ spin_lock_irqsave(&priv->reglock, flags);
+
+ writel(DDR_UE_MASK | DDR_CE_MASK,
+ priv->baseaddr + ECC_CLR_OFST);
+
+ spin_unlock_irqrestore(&priv->reglock, flags);
}
static void disable_intr(struct synps_edac_priv *priv)
{
+ unsigned long flags;
+
/* Disable UE/CE Interrupts */
- if (priv->p_data->quirks & DDR_ECC_INTR_SELF_CLEAR)
- writel(0x0, priv->baseaddr + ECC_CLR_OFST);
- else
+ if (!(priv->p_data->quirks & DDR_ECC_INTR_SELF_CLEAR)) {
writel(DDR_QOSUE_MASK | DDR_QOSCE_MASK,
priv->baseaddr + DDR_QOS_IRQ_DB_OFST);
+
+ return;
+ }
+
+ spin_lock_irqsave(&priv->reglock, flags);
+
+ writel(0, priv->baseaddr + ECC_CLR_OFST);
+
+ spin_unlock_irqrestore(&priv->reglock, flags);
}
/**
@@ -577,8 +602,6 @@ static irqreturn_t intr_handler(int irq, void *dev_id)
/* v3.0 of the controller does not have this register */
if (!(priv->p_data->quirks & DDR_ECC_INTR_SELF_CLEAR))
writel(regval, priv->baseaddr + DDR_QOS_IRQ_STAT_OFST);
- else
- enable_intr(priv);
return IRQ_HANDLED;
}
@@ -1360,6 +1383,7 @@ static int mc_probe(struct platform_device *pdev)
priv = mci->pvt_info;
priv->baseaddr = baseaddr;
priv->p_data = p_data;
+ spin_lock_init(&priv->reglock);
mc_init(mci, pdev);
The race condition around the ECCCLR register access happens in the IRQ disable method called in the device remove() procedure and in the ECC IRQ handler: 1. Enable IRQ: a. ECCCLR = EN_CE | EN_UE 2. Disable IRQ: a. ECCCLR = 0 3. IRQ handler: a. ECCCLR = CLR_CE | CLR_CE_CNT | CLR_CE | CLR_CE_CNT b. ECCCLR = 0 c. ECCCLR = EN_CE | EN_UE So if the IRQ disabling procedure is called concurrently with the IRQ handler method the IRQ might be actually left enabled due to the statement 3c. The root cause of the problem is that ECCCLR register (which since v3.10a has been called as ECCCTL) has intermixed ECC status data clear flags and the IRQ enable/disable flags. Thus the IRQ disabling (clear EN flags) and handling (write 1 to clear ECC status data) procedures must be serialised around the ECCCTL register modification to prevent the race. So fix the problem described above by adding the spin-lock around the ECCCLR modifications and preventing the IRQ-handler from modifying the IRQs enable flags (there is no point in disabling the IRQ and then re-enabling it again within a single IRQ handler call, see the statements 3a/3b and 3c above). Fixes: f7824ded4149 ("EDAC/synopsys: Add support for version 3 of the Synopsys EDAC DDR") Signed-off-by: Serge Semin <fancer.lancer@gmail.com> --- Cc: Sherry Sun <sherry.sun@nxp.com> Changelog v4: - This is a new patch detached from [PATCH v3 01/17] EDAC/synopsys: Fix native uMCTL2 IRQs handling procedure - Rename lock to reglock (Borislav) --- drivers/edac/synopsys_edac.c | 50 ++++++++++++++++++++++++++---------- 1 file changed, 37 insertions(+), 13 deletions(-)