diff options
Diffstat (limited to 'drivers/edac')
-rw-r--r-- | drivers/edac/Kconfig | 15 | ||||
-rw-r--r-- | drivers/edac/Makefile | 2 | ||||
-rw-r--r-- | drivers/edac/amd64_edac.c | 175 | ||||
-rw-r--r-- | drivers/edac/cpc925_edac.c | 6 | ||||
-rw-r--r-- | drivers/edac/edac_core.h | 2 | ||||
-rw-r--r-- | drivers/edac/edac_device.c | 5 | ||||
-rw-r--r-- | drivers/edac/edac_mc.c | 4 | ||||
-rw-r--r-- | drivers/edac/edac_mce_amd.c | 2 | ||||
-rw-r--r-- | drivers/edac/edac_pci.c | 4 | ||||
-rw-r--r-- | drivers/edac/i3200_edac.c | 527 | ||||
-rw-r--r-- | drivers/edac/mpc85xx_edac.c | 30 | ||||
-rw-r--r-- | drivers/edac/mv64x60_edac.c | 22 |
12 files changed, 654 insertions, 140 deletions
diff --git a/drivers/edac/Kconfig b/drivers/edac/Kconfig index 4339b1a..02127e5 100644 --- a/drivers/edac/Kconfig +++ b/drivers/edac/Kconfig @@ -59,7 +59,7 @@ config EDAC_MM_EDAC config EDAC_AMD64 tristate "AMD64 (Opteron, Athlon64) K8, F10h, F11h" - depends on EDAC_MM_EDAC && K8_NB && X86_64 && PCI + depends on EDAC_MM_EDAC && K8_NB && X86_64 && PCI && CPU_SUP_AMD help Support for error detection and correction on the AMD 64 Families of Memory Controllers (K8, F10h and F11h) @@ -133,6 +133,13 @@ config EDAC_I3000 Support for error detection and correction on the Intel 3000 and 3010 server chipsets. +config EDAC_I3200 + tristate "Intel 3200" + depends on EDAC_MM_EDAC && PCI && X86 && EXPERIMENTAL + help + Support for error detection and correction on the Intel + 3200 and 3210 server chipsets. + config EDAC_X38 tristate "Intel X38" depends on EDAC_MM_EDAC && PCI && X86 @@ -176,11 +183,11 @@ config EDAC_I5100 San Clemente MCH. config EDAC_MPC85XX - tristate "Freescale MPC85xx" - depends on EDAC_MM_EDAC && FSL_SOC && MPC85xx + tristate "Freescale MPC83xx / MPC85xx" + depends on EDAC_MM_EDAC && FSL_SOC && (PPC_83xx || MPC85xx) help Support for error detection and correction on the Freescale - MPC8560, MPC8540, MPC8548 + MPC8349, MPC8560, MPC8540, MPC8548 config EDAC_MV64X60 tristate "Marvell MV64x60" diff --git a/drivers/edac/Makefile b/drivers/edac/Makefile index cfa033c..7a473bb 100644 --- a/drivers/edac/Makefile +++ b/drivers/edac/Makefile @@ -32,6 +32,7 @@ obj-$(CONFIG_EDAC_I82443BXGX) += i82443bxgx_edac.o obj-$(CONFIG_EDAC_I82875P) += i82875p_edac.o obj-$(CONFIG_EDAC_I82975X) += i82975x_edac.o obj-$(CONFIG_EDAC_I3000) += i3000_edac.o +obj-$(CONFIG_EDAC_I3200) += i3200_edac.o obj-$(CONFIG_EDAC_X38) += x38_edac.o obj-$(CONFIG_EDAC_I82860) += i82860_edac.o obj-$(CONFIG_EDAC_R82600) += r82600_edac.o @@ -49,3 +50,4 @@ obj-$(CONFIG_EDAC_CELL) += cell_edac.o obj-$(CONFIG_EDAC_PPC4XX) += ppc4xx_edac.o obj-$(CONFIG_EDAC_AMD8111) += amd8111_edac.o obj-$(CONFIG_EDAC_AMD8131) += amd8131_edac.o + diff --git a/drivers/edac/amd64_edac.c b/drivers/edac/amd64_edac.c index 173dc4a..4e551e6 100644 --- a/drivers/edac/amd64_edac.c +++ b/drivers/edac/amd64_edac.c @@ -1255,7 +1255,9 @@ static int k8_dbam_map_to_pages(struct amd64_pvt *pvt, int dram_map) */ static int f10_early_channel_count(struct amd64_pvt *pvt) { + int dbams[] = { DBAM0, DBAM1 }; int err = 0, channels = 0; + int i, j; u32 dbam; err = pci_read_config_dword(pvt->dram_f2_ctl, F10_DCLR_0, &pvt->dclr0); @@ -1288,46 +1290,19 @@ static int f10_early_channel_count(struct amd64_pvt *pvt) * is more than just one DIMM present in unganged mode. Need to check * both controllers since DIMMs can be placed in either one. */ - channels = 0; - err = pci_read_config_dword(pvt->dram_f2_ctl, DBAM0, &dbam); - if (err) - goto err_reg; - - if (DBAM_DIMM(0, dbam) > 0) - channels++; - if (DBAM_DIMM(1, dbam) > 0) - channels++; - if (DBAM_DIMM(2, dbam) > 0) - channels++; - if (DBAM_DIMM(3, dbam) > 0) - channels++; - - /* If more than 2 DIMMs are present, then we have 2 channels */ - if (channels > 2) - channels = 2; - else if (channels == 0) { - /* No DIMMs on DCT0, so look at DCT1 */ - err = pci_read_config_dword(pvt->dram_f2_ctl, DBAM1, &dbam); + for (i = 0; i < ARRAY_SIZE(dbams); i++) { + err = pci_read_config_dword(pvt->dram_f2_ctl, dbams[i], &dbam); if (err) goto err_reg; - if (DBAM_DIMM(0, dbam) > 0) - channels++; - if (DBAM_DIMM(1, dbam) > 0) - channels++; - if (DBAM_DIMM(2, dbam) > 0) - channels++; - if (DBAM_DIMM(3, dbam) > 0) - channels++; - - if (channels > 2) - channels = 2; + for (j = 0; j < 4; j++) { + if (DBAM_DIMM(j, dbam) > 0) { + channels++; + break; + } + } } - /* If we found ALL 0 values, then assume just ONE DIMM-ONE Channel */ - if (channels == 0) - channels = 1; - debugf0("MCT channel count: %d\n", channels); return channels; @@ -2766,30 +2741,53 @@ static void amd64_restore_ecc_error_reporting(struct amd64_pvt *pvt) wrmsr_on_cpus(cpumask, K8_MSR_MCGCTL, msrs); } -static void check_mcg_ctl(void *ret) +/* get all cores on this DCT */ +static void get_cpus_on_this_dct_cpumask(cpumask_t *mask, int nid) { - u64 msr_val = 0; - u8 nbe; - - rdmsrl(MSR_IA32_MCG_CTL, msr_val); - nbe = msr_val & K8_MSR_MCGCTL_NBE; - - debugf0("core: %u, MCG_CTL: 0x%llx, NB MSR is %s\n", - raw_smp_processor_id(), msr_val, - (nbe ? "enabled" : "disabled")); + int cpu; - if (!nbe) - *(int *)ret = 0; + for_each_online_cpu(cpu) + if (amd_get_nb_id(cpu) == nid) + cpumask_set_cpu(cpu, mask); } /* check MCG_CTL on all the cpus on this node */ -static int amd64_mcg_ctl_enabled_on_cpus(const cpumask_t *mask) +static bool amd64_nb_mce_bank_enabled_on_node(int nid) { - int ret = 1; - preempt_disable(); - smp_call_function_many(mask, check_mcg_ctl, &ret, 1); - preempt_enable(); + cpumask_t mask; + struct msr *msrs; + int cpu, nbe, idx = 0; + bool ret = false; + cpumask_clear(&mask); + + get_cpus_on_this_dct_cpumask(&mask, nid); + + msrs = kzalloc(sizeof(struct msr) * cpumask_weight(&mask), GFP_KERNEL); + if (!msrs) { + amd64_printk(KERN_WARNING, "%s: error allocating msrs\n", + __func__); + return false; + } + + rdmsr_on_cpus(&mask, MSR_IA32_MCG_CTL, msrs); + + for_each_cpu(cpu, &mask) { + nbe = msrs[idx].l & K8_MSR_MCGCTL_NBE; + + debugf0("core: %u, MCG_CTL: 0x%llx, NB MSR is %s\n", + cpu, msrs[idx].q, + (nbe ? "enabled" : "disabled")); + + if (!nbe) + goto out; + + idx++; + } + ret = true; + +out: + kfree(msrs); return ret; } @@ -2799,71 +2797,46 @@ static int amd64_mcg_ctl_enabled_on_cpus(const cpumask_t *mask) * the memory system completely. A command line option allows to force-enable * hardware ECC later in amd64_enable_ecc_error_reporting(). */ +static const char *ecc_warning = + "WARNING: ECC is disabled by BIOS. Module will NOT be loaded.\n" + " Either Enable ECC in the BIOS, or set 'ecc_enable_override'.\n" + " Also, use of the override can cause unknown side effects.\n"; + static int amd64_check_ecc_enabled(struct amd64_pvt *pvt) { u32 value; - int err = 0, ret = 0; + int err = 0; u8 ecc_enabled = 0; + bool nb_mce_en = false; err = pci_read_config_dword(pvt->misc_f3_ctl, K8_NBCFG, &value); if (err) debugf0("Reading K8_NBCTL failed\n"); ecc_enabled = !!(value & K8_NBCFG_ECC_ENABLE); + if (!ecc_enabled) + amd64_printk(KERN_WARNING, "This node reports that Memory ECC " + "is currently disabled, set F3x%x[22] (%s).\n", + K8_NBCFG, pci_name(pvt->misc_f3_ctl)); + else + amd64_printk(KERN_INFO, "ECC is enabled by BIOS.\n"); - ret = amd64_mcg_ctl_enabled_on_cpus(cpumask_of_node(pvt->mc_node_id)); - - debugf0("K8_NBCFG=0x%x, DRAM ECC is %s\n", value, - (value & K8_NBCFG_ECC_ENABLE ? "enabled" : "disabled")); - - if (!ecc_enabled || !ret) { - if (!ecc_enabled) { - amd64_printk(KERN_WARNING, "This node reports that " - "Memory ECC is currently " - "disabled.\n"); + nb_mce_en = amd64_nb_mce_bank_enabled_on_node(pvt->mc_node_id); + if (!nb_mce_en) + amd64_printk(KERN_WARNING, "NB MCE bank disabled, set MSR " + "0x%08x[4] on node %d to enable.\n", + MSR_IA32_MCG_CTL, pvt->mc_node_id); - amd64_printk(KERN_WARNING, "bit 0x%lx in register " - "F3x%x of the MISC_CONTROL device (%s) " - "should be enabled\n", K8_NBCFG_ECC_ENABLE, - K8_NBCFG, pci_name(pvt->misc_f3_ctl)); - } - if (!ret) { - amd64_printk(KERN_WARNING, "bit 0x%016lx in MSR 0x%08x " - "of node %d should be enabled\n", - K8_MSR_MCGCTL_NBE, MSR_IA32_MCG_CTL, - pvt->mc_node_id); - } + if (!ecc_enabled || !nb_mce_en) { if (!ecc_enable_override) { - amd64_printk(KERN_WARNING, "WARNING: ECC is NOT " - "currently enabled by the BIOS. Module " - "will NOT be loaded.\n" - " Either Enable ECC in the BIOS, " - "or use the 'ecc_enable_override' " - "parameter.\n" - " Might be a BIOS bug, if BIOS says " - "ECC is enabled\n" - " Use of the override can cause " - "unknown side effects.\n"); - ret = -ENODEV; - } else - /* - * enable further driver loading if ECC enable is - * overridden. - */ - ret = 0; - } else { - amd64_printk(KERN_INFO, - "ECC is enabled by BIOS, Proceeding " - "with EDAC module initialization\n"); - - /* Signal good ECC status */ - ret = 0; - + amd64_printk(KERN_WARNING, "%s", ecc_warning); + return -ENODEV; + } + } else /* CLEAR the override, since BIOS controlled it */ ecc_enable_override = 0; - } - return ret; + return 0; } struct mcidev_sysfs_attribute sysfs_attrs[ARRAY_SIZE(amd64_dbg_attrs) + diff --git a/drivers/edac/cpc925_edac.c b/drivers/edac/cpc925_edac.c index 8c54196..3d50274 100644 --- a/drivers/edac/cpc925_edac.c +++ b/drivers/edac/cpc925_edac.c @@ -885,14 +885,14 @@ static int __devinit cpc925_probe(struct platform_device *pdev) if (!devm_request_mem_region(&pdev->dev, r->start, - r->end - r->start + 1, + resource_size(r), pdev->name)) { cpc925_printk(KERN_ERR, "Unable to request mem region\n"); res = -EBUSY; goto err1; } - vbase = devm_ioremap(&pdev->dev, r->start, r->end - r->start + 1); + vbase = devm_ioremap(&pdev->dev, r->start, resource_size(r)); if (!vbase) { cpc925_printk(KERN_ERR, "Unable to ioremap device\n"); res = -ENOMEM; @@ -953,7 +953,7 @@ err3: cpc925_mc_exit(mci); edac_mc_free(mci); err2: - devm_release_mem_region(&pdev->dev, r->start, r->end-r->start+1); + devm_release_mem_region(&pdev->dev, r->start, resource_size(r)); err1: devres_release_group(&pdev->dev, cpc925_probe); out: diff --git a/drivers/edac/edac_core.h b/drivers/edac/edac_core.h index 871c13b..12f355c 100644 --- a/drivers/edac/edac_core.h +++ b/drivers/edac/edac_core.h @@ -286,7 +286,7 @@ enum scrub_type { * is irrespective of the memory devices being mounted * on both sides of the memory stick. * - * Socket set: All of the memory sticks that are required for for + * Socket set: All of the memory sticks that are required for * a single memory access or all of the memory sticks * spanned by a chip-select row. A single socket set * has two chip-select rows and if double-sided sticks diff --git a/drivers/edac/edac_device.c b/drivers/edac/edac_device.c index b02a6a6..d5e13c9 100644 --- a/drivers/edac/edac_device.c +++ b/drivers/edac/edac_device.c @@ -356,7 +356,6 @@ static void complete_edac_device_list_del(struct rcu_head *head) edac_dev = container_of(head, struct edac_device_ctl_info, rcu); INIT_LIST_HEAD(&edac_dev->link); - complete(&edac_dev->removal_complete); } /* @@ -369,10 +368,8 @@ static void del_edac_device_from_global_list(struct edac_device_ctl_info *edac_device) { list_del_rcu(&edac_device->link); - - init_completion(&edac_device->removal_complete); call_rcu(&edac_device->rcu, complete_edac_device_list_del); - wait_for_completion(&edac_device->removal_complete); + rcu_barrier(); } /* diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c index 335b7eb..b629c41 100644 --- a/drivers/edac/edac_mc.c +++ b/drivers/edac/edac_mc.c @@ -418,16 +418,14 @@ static void complete_mc_list_del(struct rcu_head *head) mci = container_of(head, struct mem_ctl_info, rcu); INIT_LIST_HEAD(&mci->link); - complete(&mci->complete); } static void del_mc_from_global_list(struct mem_ctl_info *mci) { atomic_dec(&edac_handlers); list_del_rcu(&mci->link); - init_completion(&mci->complete); call_rcu(&mci->rcu, complete_mc_list_del); - wait_for_completion(&mci->complete); + rcu_barrier(); } /** diff --git a/drivers/edac/edac_mce_amd.c b/drivers/edac/edac_mce_amd.c index c8ca713..0c21c37 100644 --- a/drivers/edac/edac_mce_amd.c +++ b/drivers/edac/edac_mce_amd.c @@ -405,7 +405,7 @@ void decode_mce(struct mce *m) regs.nbsh = (u32)(m->status >> 32); regs.nbeal = (u32) m->addr; regs.nbeah = (u32)(m->addr >> 32); - node = per_cpu(cpu_llc_id, m->extcpu); + node = amd_get_nb_id(m->extcpu); amd_decode_nb_mce(node, ®s, 1); break; diff --git a/drivers/edac/edac_pci.c b/drivers/edac/edac_pci.c index 30b585b..efb5d56 100644 --- a/drivers/edac/edac_pci.c +++ b/drivers/edac/edac_pci.c @@ -174,7 +174,6 @@ static void complete_edac_pci_list_del(struct rcu_head *head) pci = container_of(head, struct edac_pci_ctl_info, rcu); INIT_LIST_HEAD(&pci->link); - complete(&pci->complete); } /* @@ -185,9 +184,8 @@ static void complete_edac_pci_list_del(struct rcu_head *head) static void del_edac_pci_from_global_list(struct edac_pci_ctl_info *pci) { list_del_rcu(&pci->link); - init_completion(&pci->complete); call_rcu(&pci->rcu, complete_edac_pci_list_del); - wait_for_completion(&pci->complete); + rcu_barrier(); } #if 0 diff --git a/drivers/edac/i3200_edac.c b/drivers/edac/i3200_edac.c new file mode 100644 index 0000000..fde4db9 --- /dev/null +++ b/drivers/edac/i3200_edac.c @@ -0,0 +1,527 @@ +/* + * Intel 3200/3210 Memory Controller kernel module + * Copyright (C) 2008-2009 Akamai Technologies, Inc. + * Portions by Hitoshi Mitake <h.mitake@gmail.com>. + * + * This file may be distributed under the terms of the + * GNU General Public License. + */ + +#include <linux/module.h> +#include <linux/init.h> +#include <linux/pci.h> +#include <linux/pci_ids.h> +#include <linux/slab.h> +#include <linux/edac.h> +#include <linux/io.h> +#include "edac_core.h" + +#define I3200_REVISION "1.1" + +#define EDAC_MOD_STR "i3200_edac" + +#define PCI_DEVICE_ID_INTEL_3200_HB 0x29f0 + +#define I3200_RANKS 8 +#define I3200_RANKS_PER_CHANNEL 4 +#define I3200_CHANNELS 2 + +/* Intel 3200 register addresses - device 0 function 0 - DRAM Controller */ + +#define I3200_MCHBAR_LOW 0x48 /* MCH Memory Mapped Register BAR */ +#define I3200_MCHBAR_HIGH 0x4c +#define I3200_MCHBAR_MASK 0xfffffc000ULL /* bits 35:14 */ +#define I3200_MMR_WINDOW_SIZE 16384 + +#define I3200_TOM 0xa0 /* Top of Memory (16b) + * + * 15:10 reserved + * 9:0 total populated physical memory + */ +#define I3200_TOM_MASK 0x3ff /* bits 9:0 */ +#define I3200_TOM_SHIFT 26 /* 64MiB grain */ + +#define I3200_ERRSTS 0xc8 /* Error Status Register (16b) + * + * 15 reserved + * 14 Isochronous TBWRR Run Behind FIFO Full + * (ITCV) + * 13 Isochronous TBWRR Run Behind FIFO Put + * (ITSTV) + * 12 reserved + * 11 MCH Thermal Sensor Event + * for SMI/SCI/SERR (GTSE) + * 10 reserved + * 9 LOCK to non-DRAM Memory Flag (LCKF) + * 8 reserved + * 7 DRAM Throttle Flag (DTF) + * 6:2 reserved + * 1 Multi-bit DRAM ECC Error Flag (DMERR) + * 0 Single-bit DRAM ECC Error Flag (DSERR) + */ +#define I3200_ERRSTS_UE 0x0002 +#define I3200_ERRSTS_CE 0x0001 +#define I3200_ERRSTS_BITS (I3200_ERRSTS_UE | I3200_ERRSTS_CE) + + +/* Intel MMIO register space - device 0 function 0 - MMR space */ + +#define I3200_C0DRB 0x200 /* Channel 0 DRAM Rank Boundary (16b x 4) + * + * 15:10 reserved + * 9:0 Channel 0 DRAM Rank Boundary Address + */ +#define I3200_C1DRB 0x600 /* Channel 1 DRAM Rank Boundary (16b x 4) */ +#define I3200_DRB_MASK 0x3ff /* bits 9:0 */ +#define I3200_DRB_SHIFT 26 /* 64MiB grain */ + +#define I3200_C0ECCERRLOG 0x280 /* Channel 0 ECC Error Log (64b) + * + * 63:48 Error Column Address (ERRCOL) + * 47:32 Error Row Address (ERRROW) + * 31:29 Error Bank Address (ERRBANK) + * 28:27 Error Rank Address (ERRRANK) + * 26:24 reserved + * 23:16 Error Syndrome (ERRSYND) + * 15: 2 reserved + * 1 Multiple Bit Error Status (MERRSTS) + * 0 Correctable Error Status (CERRSTS) + */ +#define I3200_C1ECCERRLOG 0x680 /* Chan 1 ECC Error Log (64b) */ +#define I3200_ECCERRLOG_CE 0x1 +#define I3200_ECCERRLOG_UE 0x2 +#define I3200_ECCERRLOG_RANK_BITS 0x18000000 +#define I3200_ECCERRLOG_RANK_SHIFT 27 +#define I3200_ECCERRLOG_SYNDROME_BITS 0xff0000 +#define I3200_ECCERRLOG_SYNDROME_SHIFT 16 +#define I3200_CAPID0 0xe0 /* P.95 of spec for details */ + +struct i3200_priv { + void __iomem *window; +}; + +static int nr_channels; + +static int how_many_channels(struct pci_dev *pdev) +{ + unsigned char capid0_8b; /* 8th byte of CAPID0 */ + + pci_read_config_byte(pdev, I3200_CAPID0 + 8, &capid0_8b); + if (capid0_8b & 0x20) { /* check DCD: Dual Channel Disable */ + debugf0("In single channel mode.\n"); + return 1; + } else { + debugf0("In dual channel mode.\n"); + return 2; + } +} + +static unsigned long eccerrlog_syndrome(u64 log) +{ + return (log & I3200_ECCERRLOG_SYNDROME_BITS) >> + I3200_ECCERRLOG_SYNDROME_SHIFT; +} + +static int eccerrlog_row(int channel, u64 log) +{ + u64 rank = ((log & I3200_ECCERRLOG_RANK_BITS) >> + I3200_ECCERRLOG_RANK_SHIFT); + return rank | (channel * I3200_RANKS_PER_CHANNEL); +} + +enum i3200_chips { + I3200 = 0, +}; + +struct i3200_dev_info { + const char *ctl_name; +}; + +struct i3200_error_info { + u16 errsts; + u16 errsts2; + u64 eccerrlog[I3200_CHANNELS]; +}; + +static const struct i3200_dev_info i3200_devs[] = { + [I3200] = { + .ctl_name = "i3200" + }, +}; + +static struct pci_dev *mci_pdev; +static int i3200_registered = 1; + + +static void i3200_clear_error_info(struct mem_ctl_info *mci) +{ + struct pci_dev *pdev; + + pdev = to_pci_dev(mci->dev); + + /* + * Clear any error bits. + * (Yes, we really clear bits by writing 1 to them.) + */ + pci_write_bits16(pdev, I3200_ERRSTS, I3200_ERRSTS_BITS, + I3200_ERRSTS_BITS); +} + +static void i3200_get_and_clear_error_info(struct mem_ctl_info *mci, + struct i3200_error_info *info) +{ + struct pci_dev *pdev; + struct i3200_priv *priv = mci->pvt_info; + void __iomem *window = priv->window; + + pdev = to_pci_dev(mci->dev); + + /* + * This is a mess because there is no atomic way to read all the + * registers at once and the registers can transition from CE being + * overwritten by UE. + */ + pci_read_config_word(pdev, I3200_ERRSTS, &info->errsts); + if (!(info->errsts & I3200_ERRSTS_BITS)) + return; + + info->eccerrlog[0] = readq(window + I3200_C0ECCERRLOG); + if (nr_channels == 2) + info->eccerrlog[1] = readq(window + I3200_C1ECCERRLOG); + + pci_read_config_word(pdev, I3200_ERRSTS, &info->errsts2); + + /* + * If the error is the same for both reads then the first set + * of reads is valid. If there is a change then there is a CE + * with no info and the second set of reads is valid and + * should be UE info. + */ + if ((info->errsts ^ info->errsts2) & I3200_ERRSTS_BITS) { + info->eccerrlog[0] = readq(window + I3200_C0ECCERRLOG); + if (nr_channels == 2) + info->eccerrlog[1] = readq(window + I3200_C1ECCERRLOG); + } + + i3200_clear_error_info(mci); +} + +static void i3200_process_error_info(struct mem_ctl_info *mci, + struct i3200_error_info *info) +{ + int channel; + u64 log; + + if (!(info->errsts & I3200_ERRSTS_BITS)) + return; + + if ((info->errsts ^ info->errsts2) & I3200_ERRSTS_BITS) { + edac_mc_handle_ce_no_info(mci, "UE overwrote CE"); + info->errsts = info->errsts2; + } + + for (channel = 0; channel < nr_channels; channel++) { + log = info->eccerrlog[channel]; + if (log & I3200_ECCERRLOG_UE) { + edac_mc_handle_ue(mci, 0, 0, + eccerrlog_row(channel, log), + "i3200 UE"); + } else if (log & I3200_ECCERRLOG_CE) { + edac_mc_handle_ce(mci, 0, 0, + eccerrlog_syndrome(log), + eccerrlog_row(channel, log), 0, + "i3200 CE"); + } + } +} + +static void i3200_check(struct mem_ctl_info *mci) +{ + struct i3200_error_info info; + + debugf1("MC%d: %s()\n", mci->mc_idx, __func__); + i3200_get_and_clear_error_info(mci, &info); + i3200_process_error_info(mci, &info); +} + + +void __iomem *i3200_map_mchbar(struct pci_dev *pdev) +{ + union { + u64 mchbar; + struct { + u32 mchbar_low; + u32 mchbar_high; + }; + } u; + void __iomem *window; + + pci_read_config_dword(pdev, I3200_MCHBAR_LOW, &u.mchbar_low); + pci_read_config_dword(pdev, I3200_MCHBAR_HIGH, &u.mchbar_high); + u.mchbar &= I3200_MCHBAR_MASK; + + if (u.mchbar != (resource_size_t)u.mchbar) { + printk(KERN_ERR + "i3200: mmio space beyond accessible range (0x%llx)\n", + (unsigned long long)u.mchbar); + return NULL; + } + + window = ioremap_nocache(u.mchbar, I3200_MMR_WINDOW_SIZE); + if (!window) + printk(KERN_ERR "i3200: cannot map mmio space at 0x%llx\n", + (unsigned long long)u.mchbar); + + return window; +} + + +static void i3200_get_drbs(void __iomem *window, + u16 drbs[I3200_CHANNELS][I3200_RANKS_PER_CHANNEL]) +{ + int i; + + for (i = 0; i < I3200_RANKS_PER_CHANNEL; i++) { + drbs[0][i] = readw(window + I3200_C0DRB + 2*i) & I3200_DRB_MASK; + drbs[1][i] = readw(window + I3200_C1DRB + 2*i) & I3200_DRB_MASK; + } +} + +static bool i3200_is_stacked(struct pci_dev *pdev, + u16 drbs[I3200_CHANNELS][I3200_RANKS_PER_CHANNEL]) +{ + u16 tom; + + pci_read_config_word(pdev, I3200_TOM, &tom); + tom &= I3200_TOM_MASK; + + return drbs[I3200_CHANNELS - 1][I3200_RANKS_PER_CHANNEL - 1] == tom; +} + +static unsigned long drb_to_nr_pages( + u16 drbs[I3200_CHANNELS][I3200_RANKS_PER_CHANNEL], bool stacked, + int channel, int rank) +{ + int n; + + n = drbs[channel][rank]; + if (rank > 0) + n -= drbs[channel][rank - 1]; + if (stacked && (channel == 1) && + drbs[channel][rank] == drbs[channel][I3200_RANKS_PER_CHANNEL - 1]) + n -= drbs[0][I3200_RANKS_PER_CHANNEL - 1]; + + n <<= (I3200_DRB_SHIFT - PAGE_SHIFT); + return n; +} + +static int i3200_probe1(struct pci_dev *pdev, int dev_idx) +{ + int rc; + int i; + struct mem_ctl_info *mci = NULL; + unsigned long last_page; + u16 drbs[I3200_CHANNELS][I3200_RANKS_PER_CHANNEL]; + bool stacked; + void __iomem *window; + struct i3200_priv *priv; + + debugf0("MC: %s()\n", __func__); + + window = i3200_map_mchbar(pdev); + if (!window) + return -ENODEV; + + i3200_get_drbs(window, drbs); + nr_channels = how_many_channels(pdev); + + mci = edac_mc_alloc(sizeof(struct i3200_priv), I3200_RANKS, + nr_channels, 0); + if (!mci) + return -ENOMEM; + + debugf3("MC: %s(): init mci\n", __func__); + + mci->dev = &pdev->dev; + mci->mtype_cap = MEM_FLAG_DDR2; + + mci->edac_ctl_cap = EDAC_FLAG_SECDED; + mci->edac_cap = EDAC_FLAG_SECDED; + + mci->mod_name = EDAC_MOD_STR; + mci->mod_ver = I3200_REVISION; + mci->ctl_name = i3200_devs[dev_idx].ctl_name; + mci->dev_name = pci_name(pdev); + mci->edac_check = i3200_check; + mci->ctl_page_to_phys = NULL; + priv = mci->pvt_info; + priv->window = window; + + stacked = i3200_is_stacked(pdev, drbs); + + /* + * The dram rank boundary (DRB) reg values are boundary addresses + * for each DRAM rank with a granularity of 64MB. DRB regs are + * cumulative; the last one will contain the total memory + * contained in all ranks. + */ + last_page = -1UL; + for (i = 0; i < mci->nr_csrows; i++) { + unsigned long nr_pages; + struct csrow_info *csrow = &mci->csrows[i]; + + nr_pages = drb_to_nr_pages(drbs, stacked, + i / I3200_RANKS_PER_CHANNEL, + i % I3200_RANKS_PER_CHANNEL); + + if (nr_pages == 0) { + csrow->mtype = MEM_EMPTY; + continue; + } + + csrow->first_page = last_page + 1; + last_page += nr_pages; + csrow->last_page = last_page; + csrow->nr_pages = nr_pages; + + csrow->grain = nr_pages << PAGE_SHIFT; + csrow->mtype = MEM_DDR2; + csrow->dtype = DEV_UNKNOWN; + csrow->edac_mode = EDAC_UNKNOWN; + } + + i3200_clear_error_info(mci); + + rc = -ENODEV; + if (edac_mc_add_mc(mci)) { + debugf3("MC: %s(): failed edac_mc_add_mc()\n", __func__); + goto fail; + } + + /* get this far and it's successful */ + debugf3("MC: %s(): success\n", __func__); + return 0; + +fail: + iounmap(window); + if (mci) + edac_mc_free(mci); + + return rc; +} + +static int __devinit i3200_init_one(struct pci_dev *pdev, + const struct pci_device_id *ent) +{ + int rc; + + debugf0("MC: %s()\n", __func__); + + if (pci_enable_device(pdev) < 0) + return -EIO; + + rc = i3200_probe1(pdev, ent->driver_data); + if (!mci_pdev) + mci_pdev = pci_dev_get(pdev); + + return rc; +} + +static void __devexit i3200_remove_one(struct pci_dev *pdev) +{ + struct mem_ctl_info *mci; + struct i3200_priv *priv; + + debugf0("%s()\n", __func__); + + mci = edac_mc_del_mc(&pdev->dev); + if (!mci) + return; + + priv = mci->pvt_info; + iounmap(priv->window); + + edac_mc_free(mci); +} + +static const struct pci_device_id i3200_pci_tbl[] __devinitdata = { + { + PCI_VEND_DEV(INTEL, 3200_HB), PCI_ANY_ID, PCI_ANY_ID, 0, 0, + I3200}, + { + 0, + } /* 0 terminated list. */ +}; + +MODULE_DEVICE_TABLE(pci, i3200_pci_tbl); + +static struct pci_driver i3200_driver = { + .name = EDAC_MOD_STR, + .probe = i3200_init_one, + .remove = __devexit_p(i3200_remove_one), + .id_table = i3200_pci_tbl, +}; + +static int __init i3200_init(void) +{ + int pci_rc; + + debugf3("MC: %s()\n", __func__); + + /* Ensure that the OPSTATE is set correctly for POLL or NMI */ + opstate_init(); + + pci_rc = pci_register_driver(&i3200_driver); + if (pci_rc < 0) + goto fail0; + + if (!mci_pdev) { + i3200_registered = 0; + mci_pdev = pci_get_device(PCI_VENDOR_ID_INTEL, + PCI_DEVICE_ID_INTEL_3200_HB, NULL); + if (!mci_pdev) { + debugf0("i3200 pci_get_device fail\n"); + pci_rc = -ENODEV; + goto fail1; + } + + pci_rc = i3200_init_one(mci_pdev, i3200_pci_tbl); + if (pci_rc < 0) { + debugf0("i3200 init fail\n"); + pci_rc = -ENODEV; + goto fail1; + } + } + + return 0; + +fail1: + pci_unregister_driver(&i3200_driver); + +fail0: + if (mci_pdev) + pci_dev_put(mci_pdev); + + return pci_rc; +} + +static void __exit i3200_exit(void) +{ + debugf3("MC: %s()\n", __func__); + + pci_unregister_driver(&i3200_driver); + if (!i3200_registered) { + i3200_remove_one(mci_pdev); + pci_dev_put(mci_pdev); + } +} + +module_init(i3200_init); +module_exit(i3200_exit); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Akamai Technologies, Inc."); +MODULE_DESCRIPTION("MC support for Intel 3200 memory hub controllers"); + +module_param(edac_op_state, int, 0444); +MODULE_PARM_DESC(edac_op_state, "EDAC Error Reporting state: 0=Poll,1=NMI"); diff --git a/drivers/edac/mpc85xx_edac.c b/drivers/edac/mpc85xx_edac.c index 3f2ccfc..157f650 100644 --- a/drivers/edac/mpc85xx_edac.c +++ b/drivers/edac/mpc85xx_edac.c @@ -41,7 +41,9 @@ static u32 orig_pci_err_en; #endif static u32 orig_l2_err_disable; +#ifdef CONFIG_MPC85xx static u32 orig_hid1[2]; +#endif /************************ MC SYSFS parts ***********************************/ @@ -646,6 +648,7 @@ static struct of_device_id mpc85xx_l2_err_of_match[] = { { .compatible = "fsl,mpc8560-l2-cache-controller", }, { .compatible = "fsl,mpc8568-l2-cache-controller", }, { .compatible = "fsl,mpc8572-l2-cache-controller", }, + { .compatible = "fsl,p2020-l2-cache-controller", }, {}, }; @@ -788,19 +791,20 @@ static void __devinit mpc85xx_init_csrows(struct mem_ctl_info *mci) csrow = &mci->csrows[index]; cs_bnds = in_be32(pdata->mc_vbase + MPC85XX_MC_CS_BNDS_0 + (index * MPC85XX_MC_CS_BNDS_OFS)); - start = (cs_bnds & 0xfff0000) << 4; - end = ((cs_bnds & 0xfff) << 20); - if (start) - start |= 0xfffff; - if (end) - end |= 0xfffff; + + start = (cs_bnds & 0xffff0000) >> 16; + end = (cs_bnds & 0x0000ffff); if (start == end) continue; /* not populated */ + start <<= (24 - PAGE_SHIFT); + end <<= (24 - PAGE_SHIFT); + end |= (1 << (24 - PAGE_SHIFT)) - 1; + csrow->first_page = start >> PAGE_SHIFT; csrow->last_page = end >> PAGE_SHIFT; - csrow->nr_pages = csrow->last_page + 1 - csrow->first_page; + csrow->nr_pages = end + 1 - start; csrow->grain = 8; csrow->mtype = mtype; csrow->dtype = DEV_UNKNOWN; @@ -984,6 +988,8 @@ static struct of_device_id mpc85xx_mc_err_of_match[] = { { .compatible = "fsl,mpc8560-memory-controller", }, { .compatible = "fsl,mpc8568-memory-controller", }, { .compatible = "fsl,mpc8572-memory-controller", }, + { .compatible = "fsl,mpc8349-memory-controller", }, + { .compatible = "fsl,p2020-memory-controller", }, {}, }; @@ -999,13 +1005,13 @@ static struct of_platform_driver mpc85xx_mc_err_driver = { }, }; - +#ifdef CONFIG_MPC85xx static void __init mpc85xx_mc_clear_rfxe(void *data) { orig_hid1[smp_processor_id()] = mfspr(SPRN_HID1); mtspr(SPRN_HID1, (orig_hid1[smp_processor_id()] & ~0x20000)); } - +#endif static int __init mpc85xx_mc_init(void) { @@ -1038,26 +1044,32 @@ static int __init mpc85xx_mc_init(void) printk(KERN_WARNING EDAC_MOD_STR "PCI fails to register\n"); #endif +#ifdef CONFIG_MPC85xx /* * need to clear HID1[RFXE] to disable machine check int * so we can catch it */ if (edac_op_state == EDAC_OPSTATE_INT) on_each_cpu(mpc85xx_mc_clear_rfxe, NULL, 0); +#endif return 0; } module_init(mpc85xx_mc_init); +#ifdef CONFIG_MPC85xx static void __exit mpc85xx_mc_restore_hid1(void *data) { mtspr(SPRN_HID1, orig_hid1[smp_processor_id()]); } +#endif static void __exit mpc85xx_mc_exit(void) { +#ifdef CONFIG_MPC85xx on_each_cpu(mpc85xx_mc_restore_hid1, NULL, 0); +#endif #ifdef CONFIG_PCI of_unregister_platform_driver(&mpc85xx_pci_err_driver); #endif diff --git a/drivers/edac/mv64x60_edac.c b/drivers/edac/mv64x60_edac.c index 5131aaa..a6b9fec 100644 --- a/drivers/edac/mv64x60_edac.c +++ b/drivers/edac/mv64x60_edac.c @@ -90,7 +90,7 @@ static int __init mv64x60_pci_fixup(struct platform_device *pdev) return -ENOENT; } - pci_serr = ioremap(r->start, r->end - r->start + 1); + pci_serr = ioremap(r->start, resource_size(r)); if (!pci_serr) return -ENOMEM; @@ -140,7 +140,7 @@ static int __devinit mv64x60_pci_err_probe(struct platform_device *pdev) if (!devm_request_mem_region(&pdev->dev, r->start, - r->end - r->start + 1, + resource_size(r), pdata->name)) { printk(KERN_ERR "%s: Error while requesting mem region\n", __func__); @@ -150,7 +150,7 @@ static int __devinit mv64x60_pci_err_probe(struct platform_device *pdev) pdata->pci_vbase = devm_ioremap(&pdev->dev, r->start, - r->end - r->start + 1); + resource_size(r)); if (!pdata->pci_vbase) { printk(KERN_ERR "%s: Unable to setup PCI err regs\n", __func__); res = -ENOMEM; @@ -306,7 +306,7 @@ static int __devinit mv64x60_sram_err_probe(struct platform_device *pdev) if (!devm_request_mem_region(&pdev->dev, r->start, - r->end - r->start + 1, + resource_size(r), pdata->name)) { printk(KERN_ERR "%s: Error while request mem region\n", __func__); @@ -316,7 +316,7 @@ static int __devinit mv64x60_sram_err_probe(struct platform_device *pdev) pdata->sram_vbase = devm_ioremap(&pdev->dev, r->start, - r->end - r->start + 1); + resource_size(r)); if (!pdata->sram_vbase) { printk(KERN_ERR "%s: Unable to setup SRAM err regs\n", __func__); @@ -474,7 +474,7 @@ static int __devinit mv64x60_cpu_err_probe(struct platform_device *pdev) if (!devm_request_mem_region(&pdev->dev, r->start, - r->end - r->start + 1, + resource_size(r), pdata->name)) { printk(KERN_ERR "%s: Error while requesting mem region\n", __func__); @@ -484,7 +484,7 @@ static int __devinit mv64x60_cpu_err_probe(struct platform_device *pdev) pdata->cpu_vbase[0] = devm_ioremap(&pdev->dev, r->start, - r->end - r->start + 1); + resource_size(r)); if (!pdata->cpu_vbase[0]) { printk(KERN_ERR "%s: Unable to setup CPU err regs\n", __func__); res = -ENOMEM; @@ -501,7 +501,7 @@ static int __devinit mv64x60_cpu_err_probe(struct platform_device *pdev) if (!devm_request_mem_region(&pdev->dev, r->start, - r->end - r->start + 1, + resource_size(r), pdata->name)) { printk(KERN_ERR "%s: Error while requesting mem region\n", __func__); @@ -511,7 +511,7 @@ static int __devinit mv64x60_cpu_err_probe(struct platform_device *pdev) pdata->cpu_vbase[1] = devm_ioremap(&pdev->dev, r->start, - r->end - r->start + 1); + resource_size(r)); if (!pdata->cpu_vbase[1]) { printk(KERN_ERR "%s: Unable to setup CPU err regs\n", __func__); res = -ENOMEM; @@ -726,7 +726,7 @@ static int __devinit mv64x60_mc_err_probe(struct platform_device *pdev) if (!devm_request_mem_region(&pdev->dev, r->start, - r->end - r->start + 1, + resource_size(r), pdata->name)) { printk(KERN_ERR "%s: Error while requesting mem region\n", __func__); @@ -736,7 +736,7 @@ static int __devinit mv64x60_mc_err_probe(struct platform_device *pdev) pdata->mc_vbase = devm_ioremap(&pdev->dev, r->start, - r->end - r->start + 1); + resource_size(r)); if (!pdata->mc_vbase) { printk(KERN_ERR "%s: Unable to setup MC err regs\n", __func__); res = -ENOMEM; |