[PATCH v6] edac: synps: Added EDAC support for zynq ddr ecc controller
Borislav Petkov
bp at alien8.de
Mon Dec 1 10:26:11 PST 2014
On Mon, Dec 01, 2014 at 09:35:09PM +0530, Punnaiah Choudary Kalluri wrote:
> Added EDAC support for reporting the ecc errors of synopsys ddr controller.
> The ddr ecc controller corrects single bit errors and detects double bit
> errors.
>
> Signed-off-by: Punnaiah Choudary Kalluri <punnaia at xilinx.com>
> ---
> Changes for v6:
> - Fixed indentation and typo errors
> - Corrected the function and structure headers
> Changes for v5:
> - Removed dt binding info as already there is a binding info available
> under memorycontroller. so, updated ecc info there.
> - corrected the ip version and function return types
> Changes for v4:
> - Shorten the macro definitions
> - Corrected the ddr ip version
> - Reverted the file name change
> Changes for v3:
> - Updated maintainer information
> - Driver cleanup as per the review comments
> - Shortened the prefix "sysnopsys" to "synps"
> Changes for v2:
> - Updated the commit header and message
> - Renamed the filenames to synopsys_edac
> - Corrected the compatilble string, commnets
> - Renamed the macros,fucntions and data structures
> ---
> MAINTAINERS | 1 +
> drivers/edac/Kconfig | 7 +
> drivers/edac/Makefile | 1 +
> drivers/edac/synopsys_edac.c | 542 ++++++++++++++++++++++++++++++++++++++++++
> 4 files changed, 551 insertions(+), 0 deletions(-)
> create mode 100644 drivers/edac/synopsys_edac.c
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 0ff630d..7601298 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -1556,6 +1556,7 @@ N: xilinx
> F: drivers/clocksource/cadence_ttc_timer.c
> F: drivers/i2c/busses/i2c-cadence.c
> F: drivers/mmc/host/sdhci-of-arasan.c
> +F: drivers/edac/synopsys_edac.c
>
> ARM SMMU DRIVER
> M: Will Deacon <will.deacon at arm.com>
> diff --git a/drivers/edac/Kconfig b/drivers/edac/Kconfig
> index 7072c28..c6d448e 100644
> --- a/drivers/edac/Kconfig
> +++ b/drivers/edac/Kconfig
> @@ -385,4 +385,11 @@ config EDAC_ALTERA_MC
> preloader must initialize the SDRAM before loading
> the kernel.
>
> +config EDAC_SYNOPSYS
> + tristate "Synopsys DDR Memory Controller"
> + depends on EDAC_MM_EDAC && ARCH_ZYNQ
> + help
> + Support for EDAC on the ECC memory used with the Synopsys DDR
> + memory controller.
Just like the other drivers:
"Support for error detection and correction on the Synopsys DDR memory
controller"
> +
> endif # EDAC
> diff --git a/drivers/edac/Makefile b/drivers/edac/Makefile
> index 359aa49..9b5a095 100644
> --- a/drivers/edac/Makefile
> +++ b/drivers/edac/Makefile
> @@ -67,3 +67,4 @@ obj-$(CONFIG_EDAC_OCTEON_LMC) += octeon_edac-lmc.o
> obj-$(CONFIG_EDAC_OCTEON_PCI) += octeon_edac-pci.o
>
> obj-$(CONFIG_EDAC_ALTERA_MC) += altera_edac.o
> +obj-$(CONFIG_EDAC_SYNOPSYS) += synopsys_edac.o
> diff --git a/drivers/edac/synopsys_edac.c b/drivers/edac/synopsys_edac.c
> new file mode 100644
> index 0000000..2a13a23
> --- /dev/null
> +++ b/drivers/edac/synopsys_edac.c
> @@ -0,0 +1,542 @@
> +/*
> + * Synopsys DDR ECC Driver
> + * This driver is based on ppc4xx_edac.c drivers
> + *
> + * Copyright (C) 2012 - 2014 Xilinx, Inc.
> + *
> + * This program is free software: you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation, either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + * This file is subject to the terms and conditions of the GNU General Public
> + * License. See the file "COPYING" in the main directory of this archive
> + * for more details
> + */
> +
> +#include <linux/edac.h>
> +#include <linux/module.h>
> +#include <linux/platform_device.h>
> +
> +#include "edac_core.h"
> +
> +/* Number of cs_rows needed per memory controller */
> +#define SYNPS_EDAC_NR_CSROWS 1
> +
> +/* Number of channels per memory controller */
> +#define SYNPS_EDAC_NR_CHANS 1
> +
> +/* Granularity of reported error in bytes */
> +#define SYNPS_EDAC_ERR_GRAIN 1
> +
> +#define SYNPS_EDAC_MSG_SIZE 256
> +
> +#define SYNPS_EDAC_MOD_STRING "synps_edac"
> +#define SYNPS_EDAC_MOD_VER "1"
> +
> +/* Synopsys DDR memory controller registers that are relevant to ECC */
> +#define CTRL_OFST 0x0
> +#define T_ZQ_OFST 0xA4
> +
> +/* ECC control register */
> +#define ECC_CTRL_OFST 0xC4
> +/* ECC log register */
> +#define CE_LOG_OFST 0xC8
> +/* ECC address register */
> +#define CE_ADDR_OFST 0xCC
> +/* ECC data[31:0] register */
> +#define CE_DATA_31_0_OFST 0xD0
> +
> +/* Uncorrectable error info regsisters */
> +#define UE_LOG_OFST 0xDC
> +#define UE_ADDR_OFST 0xE0
> +#define UE_DATA_31_0_OFST 0xE4
> +
> +#define STAT_OFST 0xF0
> +#define SCRUB_OFST 0xF4
> +
> +/* Control regsiter bitfield definitions */
> +#define CTRL_BW_MASK 0xC
> +#define CTRL_BW_SHIFT 2
> +
> +#define DDRCTL_WDTH_16 1
> +#define DDRCTL_WDTH_32 0
> +
> +/* ZQ register bitfield definitions */
> +#define T_ZQ_DDRMODE_MASK 0x2
> +
> +/* ECC control register bitfield definitions */
> +#define ECC_CTRL_CLR_CE_ERR 0x2
> +#define ECC_CTRL_CLR_UE_ERR 0x1
> +
> +/* ECC correctable/uncorrectable error log register definitions */
> +#define CE_LOG_VALID 0x1
> +#define CE_LOG_BITPOS_MASK 0xFE
> +#define CE_LOG_BITPOS_SHIFT 1
> +
> +/* ECC correctable/uncorrectable error address register definitions */
> +#define ADDR_COL_MASK 0xFFF
> +#define ADDR_ROW_MASK 0xFFFF000
> +#define ADDR_ROW_SHIFT 12
> +#define ADDR_BANK_MASK 0x70000000
> +#define ADDR_BANK_SHIFT 28
> +
> +/* ECC statistic regsiter definitions */
> +#define STAT_UECNT_MASK 0xFF
> +#define STAT_CECNT_MASK 0xFF00
> +#define STAT_CECNT_SHIFT 8
> +
> +/* ECC scrub regsiter definitions */
> +#define SCRUB_MODE_MASK 0x7
> +#define SCRUB_MODE_SECDED 0x4
> +
> +/**
> + * struct ecc_error_info - ECC error log information
> + * @row: Row number
> + * @col: Column number
> + * @bank: Bank number
> + * @bitpos: Bit position
> + * @data: Data causing the error
> + */
> +struct ecc_error_info {
> + u32 row;
> + u32 col;
> + u32 bank;
> + u32 bitpos;
> + u32 data;
> +};
> +
> +/**
> + * struct synps_ecc_status - ECC status information to report
> + * @ce_cnt: Correctable error count
> + * @ue_cnt: Uncorrectable error count
> + * @ceinfo: Correctable error log information
> + * @ueinfo: Uncorrectable error log information
> + */
> +struct synps_ecc_status {
> + u32 ce_cnt;
> + u32 ue_cnt;
> + struct ecc_error_info ceinfo;
> + struct ecc_error_info ueinfo;
> +};
> +
> +/**
> + * struct synps_edac_priv - DDR memory controller private instance data
> + * @baseaddr: Base address of the DDR controller
> + * @ce_cnt: Correctable Error count
> + * @ue_cnt: Uncorrectable Error count
> + */
> +struct synps_edac_priv {
> + void __iomem *baseaddr;
> + u32 ce_cnt;
> + u32 ue_cnt;
> +};
> +
> +/**
> + * synps_edac_geterror_info - Get the current ecc error info
> + * @base: Pointer to the base address of the ddr memory controller
> + * @p: Pointer to the synopsys ecc status structure
> + *
> + * This routine determines there is any ecc error or not
> + *
> + * Return: one if there is no error otherwise returns zero
> + */
> +static int synps_edac_geterror_info(void __iomem *base,
> + struct synps_ecc_status *p)
> +{
> + u32 regval;
> + u32 clearval = 0;
Put them on a single line:
u32 regval, clearval = 0;
> +
> + regval = readl(base + STAT_OFST) & (STAT_UECNT_MASK | STAT_CECNT_MASK);
> +
Superfluous newline.
> + if (!regval)
> + return 1;
> +
> + memset(p, 0, sizeof(*p));
See the comment below about when to memset this.
> +
> + p->ce_cnt = (regval & STAT_CECNT_MASK) >> STAT_CECNT_SHIFT;
> + p->ue_cnt = regval & STAT_UECNT_MASK;
This looks clumsy. Just do this instead:
int ret = 0;
regval = readl(base + STAT_OFST);
p->ce_cnt = (regval & STAT_CECNT_MASK) >> STAT_CECNT_SHIFT;
p->ue_cnt = regval & STAT_UECNT_MASK;
if (pe->ce_cnt ... ) {
__assign_error_info(CE, ...);
ret = 1;
}
if (pe->ue_cnt ...) {
__assign_error_info(UE, ...);
ret = 1;
}
...
return ret;
and __assign_error_info() gets the needed register values and does only
the assignments according to the first argument which tells it what to
assign - a CE or an UE.
> +
> + regval = readl(base + CE_LOG_OFST);
> + if (!(p->ce_cnt && (regval & CE_LOG_VALID)))
> + goto ue_err;
> + p->ceinfo.bitpos = (regval & CE_LOG_BITPOS_MASK) >> CE_LOG_BITPOS_SHIFT;
> + regval = readl(base + CE_ADDR_OFST);
> + p->ceinfo.row = (regval & ADDR_ROW_MASK) >> ADDR_ROW_SHIFT;
> + p->ceinfo.col = regval & ADDR_COL_MASK;
> + p->ceinfo.bank = (regval & ADDR_BANK_MASK) >> ADDR_BANK_SHIFT;
> + p->ceinfo.data = readl(base + CE_DATA_31_0_OFST);
Right so this [*] ...
> + edac_dbg(3, "ce bitposition: %d data: %d\n", p->ceinfo.bitpos,
> + p->ceinfo.data);
> + clearval = ECC_CTRL_CLR_CE_ERR;
> +
> +ue_err:
> + regval = readl(base + UE_LOG_OFST);
> + if (!(p->ue_cnt && (regval & CE_LOG_VALID)))
> + goto out;
This is funny: you do test CE_LOG_VALID here for UE and above for CE
errors. If it is the same bit, just call it ERR_VALID so that it is
clear.
> +
> + regval = readl(base + UE_ADDR_OFST);
> + p->ueinfo.row = (regval & ADDR_ROW_MASK) >> ADDR_ROW_SHIFT;
> + p->ueinfo.col = regval & ADDR_COL_MASK;
> + p->ueinfo.bank = (regval & ADDR_BANK_MASK) >> ADDR_BANK_SHIFT;
> + p->ueinfo.data = readl(base + UE_DATA_31_0_OFST);
... [*] and this looks pretty similar and could be carved out in a
__assign_error_info() function.
Also, you can slim down struct synps_ecc_status by having a single
struct ecc_error_info embedded in there and a flag which says what type
of error this struct is carrying.
Or is it technically possible to have both a CE and an UE coming out of
the hardware so that you want to carry two errors at the same time?
In any case, you can split this synps_edac_geterror_info() function into
a helper which assigns to the different p-> fields and you hand-in only
the regvals you've read before, i.e. the __assign_error_info() example
above.
> + clearval |= ECC_CTRL_CLR_UE_ERR;
> +
> +out:
> + writel(clearval, base + ECC_CTRL_OFST);
> + writel(0x0, base + ECC_CTRL_OFST);
> +
> + return 0;
> +}
> +
> +/**
> + * synps_edac_handle_error - Handle controller error types CE and UE
> + * @mci: Pointer to the edac memory controller instance
> + * @p: Pointer to the synopsys ecc status structure
> + *
> + * This routine handles the controller ECC correctable and un correctable
> + * error.
> + */
> +static void synps_edac_handle_error(struct mem_ctl_info *mci,
> + struct synps_ecc_status *p)
> +{
> + char message[SYNPS_EDAC_MSG_SIZE];
You could preallocate this on driver init so you don't do relatively big
stack allocations on the error reporting path and remain lean.
> + struct ecc_error_info *pinf;
> +
> + if (p->ce_cnt) {
> + pinf = &p->ceinfo;
> + snprintf(message, SYNPS_EDAC_MSG_SIZE,
> + "DDR ECC error type :%s Row %d Bank %d Col %d ",
> + "CE", pinf->row, pinf->bank, pinf->col);
> + edac_mc_handle_error(HW_EVENT_ERR_CORRECTED, mci,
> + p->ce_cnt, 0, 0, 0, 0, 0, -1,
> + message, "");
> + }
> +
> + if (p->ue_cnt) {
> + pinf = &p->ueinfo;
> + snprintf(message, SYNPS_EDAC_MSG_SIZE,
> + "DDR ECC error type :%s Row %d Bank %d Col %d ",
> + "UE", pinf->row, pinf->bank, pinf->col);
> + edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci,
> + p->ue_cnt, 0, 0, 0, 0, 0, -1,
> + message, "");
> + }
If you memset(p, 0,...) here, after consumption, you don't need to do it
anywhere else and be sure that *p would be always clean and ready for
the next error.
> +}
> +
> +/**
> + * synps_edac_check - Check controller for ECC errors
> + * @mci: Pointer to the edac memory controller instance
> + *
> + * This routine is used to check and post ECC errors and is called by
All those comments start with the boilerplate "This rounine is
blabla...". Just write only the "blabla... " :) In this particular case:
"Check and post ECC errors. Called by the polling thread."
Much faster to read :)
> + * the EDAC polling thread
> + */
> +static void synps_edac_check(struct mem_ctl_info *mci)
> +{
> + struct synps_edac_priv *priv = mci->pvt_info;
> + struct synps_ecc_status stat;
> + int status;
> +
> + status = synps_edac_geterror_info(priv->baseaddr, &stat);
> + if (status)
> + return;
> +
> + priv->ce_cnt += stat.ce_cnt;
> + priv->ue_cnt += stat.ue_cnt;
> + synps_edac_handle_error(mci, &stat);
> +
> + edac_dbg(3, "Total error count ce %d ue %d\n",
> + priv->ce_cnt, priv->ue_cnt);
> +}
> +
> +/**
> + * synps_edac_get_dtype - Return the controller memory width
> + * @base: Pointer to the ddr memory contoller base address
> + *
> + * This routine returns the EDAC device type width appropriate for the
> + * current controller configuration.
> + *
> + * Return: a device type width enumeration.
> + */
> +static enum dev_type synps_edac_get_dtype(const void __iomem *base)
> +{
> + enum dev_type dt;
> + u32 width;
> +
> + width = readl(base + CTRL_OFST);
> + width = (width & CTRL_BW_MASK) >> CTRL_BW_SHIFT;
> +
> + switch (width) {
> + case DDRCTL_WDTH_16:
> + dt = DEV_X2;
> + break;
> + case DDRCTL_WDTH_32:
> + dt = DEV_X4;
> + break;
> + default:
> + dt = DEV_UNKNOWN;
> + }
> +
> + return dt;
> +}
> +
> +/**
> + * synps_edac_get_eccstate - Return the controller ecc enable/disable status
> + * @base: Pointer to the ddr memory contoller base address
> + *
> + * This routine returns the ECC enable/disable status for the controller
> + *
> + * Return: a ecc status boolean i.e true/false - enabled/disabled.
> + */
> +static bool synps_edac_get_eccstate(void __iomem *base)
> +{
> + enum dev_type dt;
> + u32 ecctype;
> + bool state = false;
> +
> + dt = synps_edac_get_dtype(base);
> + if (dt == DEV_UNKNOWN)
> + return state;
> +
> + ecctype = readl(base + SCRUB_OFST) & SCRUB_MODE_MASK;
> +
> + if ((ecctype == SCRUB_MODE_SECDED) && (dt == DEV_X2)) {
> + state = true;
> + writel(0x0, base + ECC_CTRL_OFST);
Out of curiosity, why is that register write needed here? Maybe
forgotten? It looks unbalanced...
> + }
> +
> + return state;
> +}
> +
> +/**
> + * synps_edac_get_memsize - reads the size of the attached memory device
> + *
> + * Return: the memory size in bytes
> + */
> +static u32 synps_edac_get_memsize(void)
> +{
> + struct sysinfo inf;
> +
> + si_meminfo(&inf);
> +
> + return inf.totalram * inf.mem_unit;
> +}
> +
> +/**
> + * synps_edac_get_mtype - Returns controller memory type
> + * @base: pointer to the synopsys ecc status structure
> + *
> + * This routine returns the EDAC memory type appropriate for the
> + * current controller configuration.
> + *
> + * Return: a memory type enumeration.
> + */
> +static enum mem_type synps_edac_get_mtype(const void __iomem *base)
> +{
> + enum mem_type mt;
> + u32 memtype;
> +
> + memtype = readl(base + T_ZQ_OFST);
> +
> + if (memtype & T_ZQ_DDRMODE_MASK)
> + mt = MEM_DDR3;
> + else
> + mt = MEM_DDR2;
> +
> + return mt;
> +}
> +
> +/**
> + * synps_edac_init_csrows - Initialize the cs row data
> + * @mci: Pointer to the edac memory controller instance
> + *
> + * This routine initializes the chip select rows associated
> + * with the EDAC memory controller instance
> + *
> + * Return: Unconditionally 0.
> + */
> +static int synps_edac_init_csrows(struct mem_ctl_info *mci)
> +{
> + struct csrow_info *csi;
> + struct dimm_info *dimm;
> + struct synps_edac_priv *priv = mci->pvt_info;
> + u32 size;
> + int row, j;
> +
> + for (row = 0; row < mci->nr_csrows; row++) {
> + csi = mci->csrows[row];
> + size = synps_edac_get_memsize();
> +
> + for (j = 0; j < csi->nr_channels; j++) {
> + dimm = csi->channels[j]->dimm;
> + dimm->edac_mode = EDAC_FLAG_SECDED;
> + dimm->mtype = synps_edac_get_mtype(priv->baseaddr);
> + dimm->nr_pages =
> + (size >> PAGE_SHIFT) / csi->nr_channels;
No need for the linebreak here, even if it sticks out of 80 cols. The
break is uglier.
You could even align on the "=" signs for even better readability.
> + dimm->grain = SYNPS_EDAC_ERR_GRAIN;
> + dimm->dtype = synps_edac_get_dtype(priv->baseaddr);
> + }
> + }
> +
> + return 0;
> +}
> +
> +/**
> + * synps_edac_mc_init - Initialize driver instance
> + * @mci: Pointer to the edac memory controller instance
> + * @pdev: Pointer to the platform_device struct
> + *
> + * This routine performs initialization of the EDAC memory controller
> + * instance and related driver-private data associated with the
> + * memory controller the instance is bound to.
> + *
> + * Return: 0 if OK; otherwise, < 0 on error.
Where do we even return < 0 since synps_edac_init_csrows() returns
"Unconditionally 0." ?
> + */
> +static int synps_edac_mc_init(struct mem_ctl_info *mci,
> + struct platform_device *pdev)
> +{
> + int status;
> + struct synps_edac_priv *priv;
> +
> + mci->pdev = &pdev->dev;
> + priv = mci->pvt_info;
> + platform_set_drvdata(pdev, mci);
> +
> + /* Initialize controller capabilities and configuration */
> + mci->mtype_cap = MEM_FLAG_DDR3 | MEM_FLAG_DDR2;
> + mci->edac_ctl_cap = EDAC_FLAG_NONE | EDAC_FLAG_SECDED;
> + mci->scrub_cap = SCRUB_HW_SRC;
> + mci->scrub_mode = SCRUB_NONE;
> +
> + mci->edac_cap = EDAC_FLAG_SECDED;
> + mci->ctl_name = "synps_ddr_controller";
> + mci->dev_name = SYNPS_EDAC_MOD_STRING;
> + mci->mod_name = SYNPS_EDAC_MOD_VER;
> + mci->mod_ver = "1";
> +
> + edac_op_state = EDAC_OPSTATE_POLL;
> + mci->edac_check = synps_edac_check;
> + mci->ctl_page_to_phys = NULL;
> +
> + /*
> + * Initialize the MC control structure 'csrows' table
> + * with the mapping and control information.
> + */
That is too much and too obvious a comment.
> + status = synps_edac_init_csrows(mci);
> +
> + return status;
> +}
> +
> +/**
> + * synps_edac_mc_probe - Check controller and bind driver
> + * @pdev: Pointer to the platform_device struct
> + *
> + * This routine probes a specific controller
> + * instance for binding with the driver.
Ditto.
> + *
> + * Return: 0 if the controller instance was successfully bound to the
> + * driver; otherwise, < 0 on error.
> + */
> +static int synps_edac_mc_probe(struct platform_device *pdev)
> +{
> + struct mem_ctl_info *mci;
> + struct edac_mc_layer layers[2];
> + struct synps_edac_priv *priv;
> + int rc;
> + struct resource *res;
> + void __iomem *baseaddr;
> +
> + res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> + baseaddr = devm_ioremap_resource(&pdev->dev, res);
> + if (IS_ERR(baseaddr))
> + return PTR_ERR(baseaddr);
> +
> + if (synps_edac_get_eccstate(baseaddr) == false) {
if (!synps_edac_get_eccstate(...)) {
> + edac_printk(KERN_INFO, EDAC_MC, "ECC not enabled\n");
> + return -ENXIO;
> + }
> +
> + layers[0].type = EDAC_MC_LAYER_CHIP_SELECT;
> + layers[0].size = SYNPS_EDAC_NR_CSROWS;
> + layers[0].is_virt_csrow = true;
> + layers[1].type = EDAC_MC_LAYER_CHANNEL;
> + layers[1].size = SYNPS_EDAC_NR_CHANS;
> + layers[1].is_virt_csrow = false;
> +
> + mci = edac_mc_alloc(0, ARRAY_SIZE(layers), layers,
> + sizeof(struct synps_edac_priv));
> + if (!mci) {
> + edac_printk(KERN_ERR, EDAC_MC,
> + "Failed memory allocation for mc instance\n");
> + return -ENOMEM;
> + }
> +
> + priv = mci->pvt_info;
> + priv->baseaddr = baseaddr;
> + rc = synps_edac_mc_init(mci, pdev);
> + if (rc) {
> + edac_printk(KERN_ERR, EDAC_MC,
> + "Failed to initialize instance\n");
> + goto free_edac_mc;
> + }
> +
> + rc = edac_mc_add_mc(mci);
> + if (rc) {
> + edac_printk(KERN_ERR, EDAC_MC,
> + "Failed to register with EDAC core\n");
> + goto del_edac_mc;
This should be free_edac_mc as you do edac_mc_del_mc() only if
edac_mc_add_mc() has succeeded.
--
Regards/Gruss,
Boris.
Sent from a fat crate under my desk. Formatting is fine.
--
More information about the linux-arm-kernel
mailing list