aboutsummaryrefslogtreecommitdiff
path: root/drivers/edac/i82860_edac.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/edac/i82860_edac.c')
-rw-r--r--drivers/edac/i82860_edac.c312
1 files changed, 192 insertions, 120 deletions
diff --git a/drivers/edac/i82860_edac.c b/drivers/edac/i82860_edac.c
index 52596e75f9c..3382f6344e4 100644
--- a/drivers/edac/i82860_edac.c
+++ b/drivers/edac/i82860_edac.c
@@ -9,15 +9,21 @@
* by Thayne Harbaugh of Linux Networx. (http://lnxi.com)
*/
-
-#include <linux/config.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/pci.h>
#include <linux/pci_ids.h>
-#include <linux/slab.h>
-#include "edac_mc.h"
+#include <linux/edac.h>
+#include "edac_core.h"
+
+#define I82860_REVISION " Ver: 2.0.2"
+#define EDAC_MOD_STR "i82860_edac"
+
+#define i82860_printk(level, fmt, arg...) \
+ edac_printk(level, "i82860", fmt, ##arg)
+#define i82860_mc_printk(mci, level, fmt, arg...) \
+ edac_mc_chipset_printk(mci, level, "i82860", fmt, ##arg)
#ifndef PCI_DEVICE_ID_INTEL_82860_0
#define PCI_DEVICE_ID_INTEL_82860_0 0x2531
@@ -48,28 +54,32 @@ struct i82860_error_info {
static const struct i82860_dev_info i82860_devs[] = {
[I82860] = {
- .ctl_name = "i82860"},
+ .ctl_name = "i82860"},
};
-static struct pci_dev *mci_pdev = NULL; /* init dev: in case that AGP code
- has already registered driver */
+static struct pci_dev *mci_pdev; /* init dev: in case that AGP code
+ * has already registered driver
+ */
+static struct edac_pci_ctl_info *i82860_pci;
-static int i82860_registered = 1;
-
-static void i82860_get_error_info (struct mem_ctl_info *mci,
- struct i82860_error_info *info)
+static void i82860_get_error_info(struct mem_ctl_info *mci,
+ struct i82860_error_info *info)
{
+ struct pci_dev *pdev;
+
+ pdev = to_pci_dev(mci->pdev);
+
/*
* 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(mci->pdev, I82860_ERRSTS, &info->errsts);
- pci_read_config_dword(mci->pdev, I82860_EAP, &info->eap);
- pci_read_config_word(mci->pdev, I82860_DERRCTL_STS, &info->derrsyn);
- pci_read_config_word(mci->pdev, I82860_ERRSTS, &info->errsts2);
+ pci_read_config_word(pdev, I82860_ERRSTS, &info->errsts);
+ pci_read_config_dword(pdev, I82860_EAP, &info->eap);
+ pci_read_config_word(pdev, I82860_DERRCTL_STS, &info->derrsyn);
+ pci_read_config_word(pdev, I82860_ERRSTS, &info->errsts2);
- pci_write_bits16(mci->pdev, I82860_ERRSTS, 0x0003, 0x0003);
+ pci_write_bits16(pdev, I82860_ERRSTS, 0x0003, 0x0003);
/*
* If the error is the same for both reads then the first set of reads
@@ -78,16 +88,18 @@ static void i82860_get_error_info (struct mem_ctl_info *mci,
*/
if (!(info->errsts2 & 0x0003))
return;
+
if ((info->errsts ^ info->errsts2) & 0x0003) {
- pci_read_config_dword(mci->pdev, I82860_EAP, &info->eap);
- pci_read_config_word(mci->pdev, I82860_DERRCTL_STS,
- &info->derrsyn);
+ pci_read_config_dword(pdev, I82860_EAP, &info->eap);
+ pci_read_config_word(pdev, I82860_DERRCTL_STS, &info->derrsyn);
}
}
-static int i82860_process_error_info (struct mem_ctl_info *mci,
- struct i82860_error_info *info, int handle_errors)
+static int i82860_process_error_info(struct mem_ctl_info *mci,
+ struct i82860_error_info *info,
+ int handle_errors)
{
+ struct dimm_info *dimm;
int row;
if (!(info->errsts2 & 0x0003))
@@ -97,18 +109,25 @@ static int i82860_process_error_info (struct mem_ctl_info *mci,
return 1;
if ((info->errsts ^ info->errsts2) & 0x0003) {
- edac_mc_handle_ce_no_info(mci, "UE overwrote CE");
+ edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci, 1, 0, 0, 0,
+ -1, -1, -1, "UE overwrote CE", "");
info->errsts = info->errsts2;
}
info->eap >>= PAGE_SHIFT;
row = edac_mc_find_csrow_by_page(mci, info->eap);
+ dimm = mci->csrows[row]->channels[0]->dimm;
if (info->errsts & 0x0002)
- edac_mc_handle_ue(mci, info->eap, 0, row, "i82860 UE");
+ edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci, 1,
+ info->eap, 0, 0,
+ dimm->location[0], dimm->location[1], -1,
+ "i82860 UE", "");
else
- edac_mc_handle_ce(mci, info->eap, 0, info->derrsyn, row,
- 0, "i82860 UE");
+ edac_mc_handle_error(HW_EVENT_ERR_UNCORRECTED, mci, 1,
+ info->eap, 0, info->derrsyn,
+ dimm->location[0], dimm->location[1], -1,
+ "i82860 CE", "");
return 1;
}
@@ -117,139 +136,173 @@ static void i82860_check(struct mem_ctl_info *mci)
{
struct i82860_error_info info;
- debugf1("MC%d: " __FILE__ ": %s()\n", mci->mc_idx, __func__);
+ edac_dbg(1, "MC%d\n", mci->mc_idx);
i82860_get_error_info(mci, &info);
i82860_process_error_info(mci, &info, 1);
}
-static int i82860_probe1(struct pci_dev *pdev, int dev_idx)
+static void i82860_init_csrows(struct mem_ctl_info *mci, struct pci_dev *pdev)
{
- int rc = -ENODEV;
- int index;
- struct mem_ctl_info *mci = NULL;
unsigned long last_cumul_size;
+ u16 mchcfg_ddim; /* DRAM Data Integrity Mode 0=none, 2=edac */
+ u16 value;
+ u32 cumul_size;
+ struct csrow_info *csrow;
+ struct dimm_info *dimm;
+ int index;
- u16 mchcfg_ddim; /* DRAM Data Integrity Mode 0=none,2=edac */
-
- /* RDRAM has channels but these don't map onto the abstractions that
- edac uses.
- The device groups from the GRA registers seem to map reasonably
- well onto the notion of a chip select row.
- There are 16 GRA registers and since the name is associated with
- the channel and the GRA registers map to physical devices so we are
- going to make 1 channel for group.
- */
- mci = edac_mc_alloc(0, 16, 1);
- if (!mci)
- return -ENOMEM;
-
- debugf3("MC: " __FILE__ ": %s(): init mci\n", __func__);
-
- mci->pdev = pdev;
- mci->mtype_cap = MEM_FLAG_DDR;
-
-
- mci->edac_ctl_cap = EDAC_FLAG_NONE | EDAC_FLAG_SECDED;
- /* I"m not sure about this but I think that all RDRAM is SECDED */
- mci->edac_cap = EDAC_FLAG_SECDED;
- /* adjust FLAGS */
-
- mci->mod_name = BS_MOD_STR;
- mci->mod_ver = "$Revision: 1.1.2.6 $";
- mci->ctl_name = i82860_devs[dev_idx].ctl_name;
- mci->edac_check = i82860_check;
- mci->ctl_page_to_phys = NULL;
-
- pci_read_config_word(mci->pdev, I82860_MCHCFG, &mchcfg_ddim);
+ pci_read_config_word(pdev, I82860_MCHCFG, &mchcfg_ddim);
mchcfg_ddim = mchcfg_ddim & 0x180;
+ last_cumul_size = 0;
- /*
- * The group row boundary (GRA) reg values are boundary address
+ /* The group row boundary (GRA) reg values are boundary address
* for each DRAM row with a granularity of 16MB. GRA regs are
* cumulative; therefore GRA15 will contain the total memory contained
* in all eight rows.
*/
- for (last_cumul_size = index = 0; index < mci->nr_csrows; index++) {
- u16 value;
- u32 cumul_size;
- struct csrow_info *csrow = &mci->csrows[index];
-
- pci_read_config_word(mci->pdev, I82860_GBA + index * 2,
- &value);
+ for (index = 0; index < mci->nr_csrows; index++) {
+ csrow = mci->csrows[index];
+ dimm = csrow->channels[0]->dimm;
+ pci_read_config_word(pdev, I82860_GBA + index * 2, &value);
cumul_size = (value & I82860_GBA_MASK) <<
- (I82860_GBA_SHIFT - PAGE_SHIFT);
- debugf3("MC: " __FILE__ ": %s(): (%d) cumul_size 0x%x\n",
- __func__, index, cumul_size);
+ (I82860_GBA_SHIFT - PAGE_SHIFT);
+ edac_dbg(3, "(%d) cumul_size 0x%x\n", index, cumul_size);
+
if (cumul_size == last_cumul_size)
continue; /* not populated */
csrow->first_page = last_cumul_size;
csrow->last_page = cumul_size - 1;
- csrow->nr_pages = cumul_size - last_cumul_size;
+ dimm->nr_pages = cumul_size - last_cumul_size;
last_cumul_size = cumul_size;
- csrow->grain = 1 << 12; /* I82860_EAP has 4KiB reolution */
- csrow->mtype = MEM_RMBS;
- csrow->dtype = DEV_UNKNOWN;
- csrow->edac_mode = mchcfg_ddim ? EDAC_SECDED : EDAC_NONE;
+ dimm->grain = 1 << 12; /* I82860_EAP has 4KiB reolution */
+ dimm->mtype = MEM_RMBS;
+ dimm->dtype = DEV_UNKNOWN;
+ dimm->edac_mode = mchcfg_ddim ? EDAC_SECDED : EDAC_NONE;
}
+}
+
+static int i82860_probe1(struct pci_dev *pdev, int dev_idx)
+{
+ struct mem_ctl_info *mci;
+ struct edac_mc_layer layers[2];
+ struct i82860_error_info discard;
+
+ /*
+ * RDRAM has channels but these don't map onto the csrow abstraction.
+ * According with the datasheet, there are 2 Rambus channels, supporting
+ * up to 16 direct RDRAM devices.
+ * The device groups from the GRA registers seem to map reasonably
+ * well onto the notion of a chip select row.
+ * There are 16 GRA registers and since the name is associated with
+ * the channel and the GRA registers map to physical devices so we are
+ * going to make 1 channel for group.
+ */
+ layers[0].type = EDAC_MC_LAYER_CHANNEL;
+ layers[0].size = 2;
+ layers[0].is_virt_csrow = true;
+ layers[1].type = EDAC_MC_LAYER_SLOT;
+ layers[1].size = 8;
+ layers[1].is_virt_csrow = true;
+ mci = edac_mc_alloc(0, ARRAY_SIZE(layers), layers, 0);
+ if (!mci)
+ return -ENOMEM;
- /* clear counters */
- pci_write_bits16(mci->pdev, I82860_ERRSTS, 0x0003, 0x0003);
+ edac_dbg(3, "init mci\n");
+ mci->pdev = &pdev->dev;
+ mci->mtype_cap = MEM_FLAG_DDR;
+ mci->edac_ctl_cap = EDAC_FLAG_NONE | EDAC_FLAG_SECDED;
+ /* I"m not sure about this but I think that all RDRAM is SECDED */
+ mci->edac_cap = EDAC_FLAG_SECDED;
+ mci->mod_name = EDAC_MOD_STR;
+ mci->mod_ver = I82860_REVISION;
+ mci->ctl_name = i82860_devs[dev_idx].ctl_name;
+ mci->dev_name = pci_name(pdev);
+ mci->edac_check = i82860_check;
+ mci->ctl_page_to_phys = NULL;
+ i82860_init_csrows(mci, pdev);
+ i82860_get_error_info(mci, &discard); /* clear counters */
+ /* Here we assume that we will never see multiple instances of this
+ * type of memory controller. The ID is therefore hardcoded to 0.
+ */
if (edac_mc_add_mc(mci)) {
- debugf3("MC: " __FILE__
- ": %s(): failed edac_mc_add_mc()\n",
+ edac_dbg(3, "failed edac_mc_add_mc()\n");
+ goto fail;
+ }
+
+ /* allocating generic PCI control info */
+ i82860_pci = edac_pci_create_generic_ctl(&pdev->dev, EDAC_MOD_STR);
+ if (!i82860_pci) {
+ printk(KERN_WARNING
+ "%s(): Unable to create PCI control\n",
+ __func__);
+ printk(KERN_WARNING
+ "%s(): PCI error report via EDAC not setup\n",
__func__);
- edac_mc_free(mci);
- } else {
- /* get this far and it's successful */
- debugf3("MC: " __FILE__ ": %s(): success\n", __func__);
- rc = 0;
}
- return rc;
+
+ /* get this far and it's successful */
+ edac_dbg(3, "success\n");
+
+ return 0;
+
+fail:
+ edac_mc_free(mci);
+ return -ENODEV;
}
/* returns count (>= 0), or negative on error */
-static int __devinit i82860_init_one(struct pci_dev *pdev,
- const struct pci_device_id *ent)
+static int i82860_init_one(struct pci_dev *pdev,
+ const struct pci_device_id *ent)
{
int rc;
- debugf0("MC: " __FILE__ ": %s()\n", __func__);
+ edac_dbg(0, "\n");
+ i82860_printk(KERN_INFO, "i82860 init one\n");
- printk(KERN_INFO "i82860 init one\n");
- if(pci_enable_device(pdev) < 0)
+ if (pci_enable_device(pdev) < 0)
return -EIO;
+
rc = i82860_probe1(pdev, ent->driver_data);
- if(rc == 0)
+
+ if (rc == 0)
mci_pdev = pci_dev_get(pdev);
+
return rc;
}
-static void __devexit i82860_remove_one(struct pci_dev *pdev)
+static void i82860_remove_one(struct pci_dev *pdev)
{
struct mem_ctl_info *mci;
- debugf0(__FILE__ ": %s()\n", __func__);
+ edac_dbg(0, "\n");
+
+ if (i82860_pci)
+ edac_pci_release_generic_ctl(i82860_pci);
+
+ if ((mci = edac_mc_del_mc(&pdev->dev)) == NULL)
+ return;
- mci = edac_mc_find_mci_by_pdev(pdev);
- if ((mci != NULL) && (edac_mc_del_mc(mci) == 0))
- edac_mc_free(mci);
+ edac_mc_free(mci);
}
-static const struct pci_device_id i82860_pci_tbl[] __devinitdata = {
- {PCI_VEND_DEV(INTEL, 82860_0), PCI_ANY_ID, PCI_ANY_ID, 0, 0,
+static const struct pci_device_id i82860_pci_tbl[] = {
+ {
+ PCI_VEND_DEV(INTEL, 82860_0), PCI_ANY_ID, PCI_ANY_ID, 0, 0,
I82860},
- {0,} /* 0 terminated list. */
+ {
+ 0,
+ } /* 0 terminated list. */
};
MODULE_DEVICE_TABLE(pci, i82860_pci_tbl);
static struct pci_driver i82860_driver = {
- .name = BS_MOD_STR,
+ .name = EDAC_MOD_STR,
.probe = i82860_init_one,
- .remove = __devexit_p(i82860_remove_one),
+ .remove = i82860_remove_one,
.id_table = i82860_pci_tbl,
};
@@ -257,43 +310,62 @@ static int __init i82860_init(void)
{
int pci_rc;
- debugf3("MC: " __FILE__ ": %s()\n", __func__);
+ edac_dbg(3, "\n");
+
+ /* Ensure that the OPSTATE is set correctly for POLL or NMI */
+ opstate_init();
+
if ((pci_rc = pci_register_driver(&i82860_driver)) < 0)
- return pci_rc;
+ goto fail0;
if (!mci_pdev) {
- i82860_registered = 0;
mci_pdev = pci_get_device(PCI_VENDOR_ID_INTEL,
- PCI_DEVICE_ID_INTEL_82860_0, NULL);
+ PCI_DEVICE_ID_INTEL_82860_0, NULL);
+
if (mci_pdev == NULL) {
- debugf0("860 pci_get_device fail\n");
- return -ENODEV;
+ edac_dbg(0, "860 pci_get_device fail\n");
+ pci_rc = -ENODEV;
+ goto fail1;
}
+
pci_rc = i82860_init_one(mci_pdev, i82860_pci_tbl);
+
if (pci_rc < 0) {
- debugf0("860 init fail\n");
- pci_dev_put(mci_pdev);
- return -ENODEV;
+ edac_dbg(0, "860 init fail\n");
+ pci_rc = -ENODEV;
+ goto fail1;
}
}
+
return 0;
+
+fail1:
+ pci_unregister_driver(&i82860_driver);
+
+fail0:
+ if (mci_pdev != NULL)
+ pci_dev_put(mci_pdev);
+
+ return pci_rc;
}
static void __exit i82860_exit(void)
{
- debugf3("MC: " __FILE__ ": %s()\n", __func__);
+ edac_dbg(3, "\n");
pci_unregister_driver(&i82860_driver);
- if (!i82860_registered) {
- i82860_remove_one(mci_pdev);
+
+ if (mci_pdev != NULL)
pci_dev_put(mci_pdev);
- }
}
module_init(i82860_init);
module_exit(i82860_exit);
MODULE_LICENSE("GPL");
-MODULE_AUTHOR
- ("Red Hat Inc. (http://www.redhat.com.com) Ben Woodard <woodard@redhat.com>");
+MODULE_AUTHOR("Red Hat Inc. (http://www.redhat.com) "
+ "Ben Woodard <woodard@redhat.com>");
MODULE_DESCRIPTION("ECC support for Intel 82860 memory hub controllers");
+
+module_param(edac_op_state, int, 0444);
+MODULE_PARM_DESC(edac_op_state, "EDAC Error Reporting state: 0=Poll,1=NMI");
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-12-14 09:19:20 +0000