/*
* eeh.c
* Copyright (C) 2001 Dave Engebretsen & Todd Inglett IBM Corporation
- *
+ *
* 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.
- *
+ *
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
-#include <linux/bootmem.h>
#include <linux/init.h>
#include <linux/list.h>
-#include <linux/mm.h>
#include <linux/notifier.h>
#include <linux/pci.h>
#include <linux/proc_fs.h>
#include <linux/rbtree.h>
#include <linux/seq_file.h>
#include <linux/spinlock.h>
+#include <asm/atomic.h>
#include <asm/eeh.h>
#include <asm/io.h>
#include <asm/machdep.h>
#include <asm/rtas.h>
#include <asm/atomic.h>
#include <asm/systemcfg.h>
-#include "pci.h"
+#include <asm/ppc-pci.h>
#undef DEBUG
* were "empty": all reads return 0xff's and all writes are silently
* ignored. EEH slot isolation events can be triggered by parity
* errors on the address or data busses (e.g. during posted writes),
- * which in turn might be caused by dust, vibration, humidity,
- * radioactivity or plain-old failed hardware.
+ * which in turn might be caused by low voltage on the bus, dust,
+ * vibration, humidity, radioactivity or plain-old failed hardware.
*
* Note, however, that one of the leading causes of EEH slot
* freeze events are buggy device drivers, buggy device microcode,
* and sent out for processing.
*/
-/** Bus Unit ID macros; get low and hi 32-bits of the 64-bit BUID */
-#define BUID_HI(buid) ((buid) >> 32)
-#define BUID_LO(buid) ((buid) & 0xffffffff)
-
/* EEH event workqueue setup. */
static DEFINE_SPINLOCK(eeh_eventlist_lock);
LIST_HEAD(eeh_eventlist);
static int eeh_error_buf_size;
/* System monitoring statistics */
+static DEFINE_PER_CPU(unsigned long, no_device);
+static DEFINE_PER_CPU(unsigned long, no_dn);
+static DEFINE_PER_CPU(unsigned long, no_cfg_addr);
+static DEFINE_PER_CPU(unsigned long, ignored_check);
static DEFINE_PER_CPU(unsigned long, total_mmio_ffs);
static DEFINE_PER_CPU(unsigned long, false_positives);
static DEFINE_PER_CPU(unsigned long, ignored_failures);
while (*p) {
parent = *p;
piar = rb_entry(parent, struct pci_io_addr_range, rb_node);
- if (alo < piar->addr_lo) {
+ if (ahi < piar->addr_lo) {
p = &parent->rb_left;
- } else if (ahi > piar->addr_hi) {
+ } else if (alo > piar->addr_hi) {
p = &parent->rb_right;
} else {
if (dev != piar->pcidev ||
piar->pcidev = dev;
piar->flags = flags;
+#ifdef DEBUG
+ printk(KERN_DEBUG "PIAR: insert range=[%lx:%lx] dev=%s\n",
+ alo, ahi, pci_name (dev));
+#endif
+
rb_link_node(&piar->rb_node, parent, p);
rb_insert_color(&piar->rb_node, &pci_io_addr_cache_root.rb_root);
static void __pci_addr_cache_insert_device(struct pci_dev *dev)
{
struct device_node *dn;
+ struct pci_dn *pdn;
int i;
int inserted = 0;
dn = pci_device_to_OF_node(dev);
if (!dn) {
- printk(KERN_WARNING "PCI: no pci dn found for dev=%s\n",
- pci_name(dev));
+ printk(KERN_WARNING "PCI: no pci dn found for dev=%s\n", pci_name(dev));
return;
}
/* Skip any devices for which EEH is not enabled. */
- if (!(dn->eeh_mode & EEH_MODE_SUPPORTED) ||
- dn->eeh_mode & EEH_MODE_NOCHECK) {
+ pdn = PCI_DN(dn);
+ if (!(pdn->eeh_mode & EEH_MODE_SUPPORTED) ||
+ pdn->eeh_mode & EEH_MODE_NOCHECK) {
#ifdef DEBUG
- printk(KERN_INFO "PCI: skip building address cache for=%s\n",
- pci_name(dev));
+ printk(KERN_INFO "PCI: skip building address cache for=%s - %s\n",
+ pci_name(dev), pdn->node->full_name);
#endif
return;
}
* we maintain a cache of devices that can be quickly searched.
* This routine adds a device to that cache.
*/
-void pci_addr_cache_insert_device(struct pci_dev *dev)
+static void pci_addr_cache_insert_device(struct pci_dev *dev)
{
unsigned long flags;
* the tree multiple times (once per resource).
* But so what; device removal doesn't need to be that fast.
*/
-void pci_addr_cache_remove_device(struct pci_dev *dev)
+static void pci_addr_cache_remove_device(struct pci_dev *dev)
{
unsigned long flags;
{
struct pci_dev *dev = NULL;
+ if (!eeh_subsystem_enabled)
+ return;
+
spin_lock_init(&pci_io_addr_cache_root.piar_lock);
while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
/* --------------------------------------------------------------- */
/* Above lies the PCI Address Cache. Below lies the EEH event infrastructure */
+void eeh_slot_error_detail (struct pci_dn *pdn, int severity)
+{
+ unsigned long flags;
+ int rc;
+
+ /* Log the error with the rtas logger */
+ spin_lock_irqsave(&slot_errbuf_lock, flags);
+ memset(slot_errbuf, 0, eeh_error_buf_size);
+
+ rc = rtas_call(ibm_slot_error_detail,
+ 8, 1, NULL, pdn->eeh_config_addr,
+ BUID_HI(pdn->phb->buid),
+ BUID_LO(pdn->phb->buid), NULL, 0,
+ virt_to_phys(slot_errbuf),
+ eeh_error_buf_size,
+ severity);
+
+ if (rc == 0)
+ log_error(slot_errbuf, ERR_TYPE_RTAS_LOG, 0);
+ spin_unlock_irqrestore(&slot_errbuf_lock, flags);
+}
+
/**
* eeh_register_notifier - Register to find out about EEH events.
* @nb: notifier block to callback on events
* @dn: device node to read
* @rets: array to return results in
*/
-static int read_slot_reset_state(struct device_node *dn, int rets[])
+static int read_slot_reset_state(struct pci_dn *pdn, int rets[])
{
int token, outputs;
outputs = 4;
} else {
token = ibm_read_slot_reset_state;
+ rets[2] = 0; /* fake PE Unavailable info */
outputs = 3;
}
- return rtas_call(token, 3, outputs, rets, dn->eeh_config_addr,
- BUID_HI(dn->phb->buid), BUID_LO(dn->phb->buid));
+ return rtas_call(token, 3, outputs, rets, pdn->eeh_config_addr,
+ BUID_HI(pdn->phb->buid), BUID_LO(pdn->phb->buid));
}
/**
* Since the panic_on_oops sysctl is used to halt the system
* in light of potential corruption, we can use it here.
*/
- if (panic_on_oops)
+ if (panic_on_oops) {
+ struct device_node *dn = pci_device_to_OF_node(dev);
+ eeh_slot_error_detail (PCI_DN(dn), 2 /* Permanent Error */);
panic("EEH: MMIO failure (%d) on device:%s\n", reset_state,
pci_name(dev));
+ }
else {
__get_cpu_var(ignored_failures)++;
printk(KERN_INFO "EEH: Ignored MMIO failure (%d) on device:%s\n",
notifier_call_chain (&eeh_notifier_chain,
EEH_NOTIFY_FREEZE, event);
- __get_cpu_var(slot_resets)++;
-
pci_dev_put(event->dev);
kfree(event);
}
/**
* eeh_token_to_phys - convert EEH address token to phys address
- * @token i/o token, should be address in the form 0xE....
+ * @token i/o token, should be address in the form 0xA....
*/
static inline unsigned long eeh_token_to_phys(unsigned long token)
{
* will query firmware for the EEH status.
*
* Returns 0 if there has not been an EEH error; otherwise returns
- * a non-zero value and queues up a solt isolation event notification.
+ * a non-zero value and queues up a slot isolation event notification.
*
* It is safe to call this routine in an interrupt context.
*/
int ret;
int rets[3];
unsigned long flags;
- int rc, reset_state;
+ int reset_state;
struct eeh_event *event;
+ struct pci_dn *pdn;
__get_cpu_var(total_mmio_ffs)++;
if (!eeh_subsystem_enabled)
return 0;
- if (!dn)
+ if (!dn) {
+ __get_cpu_var(no_dn)++;
return 0;
+ }
+ pdn = PCI_DN(dn);
/* Access to IO BARs might get this far and still not want checking. */
- if (!(dn->eeh_mode & EEH_MODE_SUPPORTED) ||
- dn->eeh_mode & EEH_MODE_NOCHECK) {
+ if (!pdn->eeh_capable || !(pdn->eeh_mode & EEH_MODE_SUPPORTED) ||
+ pdn->eeh_mode & EEH_MODE_NOCHECK) {
+ __get_cpu_var(ignored_check)++;
+#ifdef DEBUG
+ printk ("EEH:ignored check for %s %s\n", pci_name (dev), dn->full_name);
+#endif
return 0;
}
- if (!dn->eeh_config_addr) {
+ if (!pdn->eeh_config_addr) {
+ __get_cpu_var(no_cfg_addr)++;
return 0;
}
* If we already have a pending isolation event for this
* slot, we know it's bad already, we don't need to check...
*/
- if (dn->eeh_mode & EEH_MODE_ISOLATED) {
+ if (pdn->eeh_mode & EEH_MODE_ISOLATED) {
atomic_inc(&eeh_fail_count);
if (atomic_read(&eeh_fail_count) >= EEH_MAX_FAILS) {
/* re-read the slot reset state */
- if (read_slot_reset_state(dn, rets) != 0)
+ if (read_slot_reset_state(pdn, rets) != 0)
rets[0] = -1; /* reset state unknown */
eeh_panic(dev, rets[0]);
}
* function zero of a multi-function device.
* In any case they must share a common PHB.
*/
- ret = read_slot_reset_state(dn, rets);
- if (!(ret == 0 && rets[1] == 1 && (rets[0] == 2 || rets[0] == 4))) {
+ ret = read_slot_reset_state(pdn, rets);
+
+ /* If the call to firmware failed, punt */
+ if (ret != 0) {
+ printk(KERN_WARNING "EEH: read_slot_reset_state() failed; rc=%d dn=%s\n",
+ ret, dn->full_name);
__get_cpu_var(false_positives)++;
return 0;
}
- /* prevent repeated reports of this failure */
- dn->eeh_mode |= EEH_MODE_ISOLATED;
+ /* If EEH is not supported on this device, punt. */
+ if (rets[1] != 1) {
+ printk(KERN_WARNING "EEH: event on unsupported device, rc=%d dn=%s\n",
+ ret, dn->full_name);
+ __get_cpu_var(false_positives)++;
+ return 0;
+ }
- reset_state = rets[0];
+ /* If not the kind of error we know about, punt. */
+ if (rets[0] != 2 && rets[0] != 4 && rets[0] != 5) {
+ __get_cpu_var(false_positives)++;
+ return 0;
+ }
- spin_lock_irqsave(&slot_errbuf_lock, flags);
- memset(slot_errbuf, 0, eeh_error_buf_size);
+ /* Note that config-io to empty slots may fail;
+ * we recognize empty because they don't have children. */
+ if ((rets[0] == 5) && (dn->child == NULL)) {
+ __get_cpu_var(false_positives)++;
+ return 0;
+ }
- rc = rtas_call(ibm_slot_error_detail,
- 8, 1, NULL, dn->eeh_config_addr,
- BUID_HI(dn->phb->buid),
- BUID_LO(dn->phb->buid), NULL, 0,
- virt_to_phys(slot_errbuf),
- eeh_error_buf_size,
- 1 /* Temporary Error */);
+ /* prevent repeated reports of this failure */
+ pdn->eeh_mode |= EEH_MODE_ISOLATED;
+ __get_cpu_var(slot_resets)++;
- if (rc == 0)
- log_error(slot_errbuf, ERR_TYPE_RTAS_LOG, 0);
- spin_unlock_irqrestore(&slot_errbuf_lock, flags);
+ reset_state = rets[0];
+
+ eeh_slot_error_detail (pdn, 1 /* Temporary Error */);
printk(KERN_INFO "EEH: MMIO failure (%d) on device: %s %s\n",
rets[0], dn->name, dn->full_name);
/* Most EEH events are due to device driver bugs. Having
* a stack trace will help the device-driver authors figure
* out what happened. So print that out. */
- dump_stack();
+ if (rets[0] != 5) dump_stack();
schedule_work(&eeh_event_wq);
return 0;
* @token i/o token, should be address in the form 0xA....
* @val value, should be all 1's (XXX why do we need this arg??)
*
- * Check for an eeh failure at the given token address.
* Check for an EEH failure at the given token address. Call this
* routine if the result of a read was all 0xff's and you want to
* find out if this is due to an EEH slot freeze event. This routine
/* Finding the phys addr + pci device; this is pretty quick. */
addr = eeh_token_to_phys((unsigned long __force) token);
dev = pci_get_device_by_addr(addr);
- if (!dev)
+ if (!dev) {
+ __get_cpu_var(no_device)++;
return val;
+ }
dn = pci_device_to_OF_node(dev);
eeh_dn_check_failure (dn, dev);
u32 *device_id = (u32 *)get_property(dn, "device-id", NULL);
u32 *regs;
int enable;
+ struct pci_dn *pdn = PCI_DN(dn);
- dn->eeh_mode = 0;
+ pdn->eeh_mode = 0;
if (status && strcmp(status, "ok") != 0)
return NULL; /* ignore devices with bad status */
/* There is nothing to check on PCI to ISA bridges */
if (dn->type && !strcmp(dn->type, "isa")) {
- dn->eeh_mode |= EEH_MODE_NOCHECK;
+ pdn->eeh_mode |= EEH_MODE_NOCHECK;
return NULL;
}
enable = 0;
if (!enable)
- dn->eeh_mode |= EEH_MODE_NOCHECK;
+ pdn->eeh_mode |= EEH_MODE_NOCHECK;
/* Ok... see if this device supports EEH. Some do, some don't,
* and the only way to find out is to check each and every one. */
EEH_ENABLE);
if (ret == 0) {
eeh_subsystem_enabled = 1;
- dn->eeh_mode |= EEH_MODE_SUPPORTED;
- dn->eeh_config_addr = regs[0];
+ pdn->eeh_mode |= EEH_MODE_SUPPORTED;
+ pdn->eeh_config_addr = regs[0];
#ifdef DEBUG
printk(KERN_DEBUG "EEH: %s: eeh enabled\n", dn->full_name);
#endif
/* This device doesn't support EEH, but it may have an
* EEH parent, in which case we mark it as supported. */
- if (dn->parent && (dn->parent->eeh_mode & EEH_MODE_SUPPORTED)) {
+ if (dn->parent && PCI_DN(dn->parent)
+ && (PCI_DN(dn->parent)->eeh_mode & EEH_MODE_SUPPORTED)) {
/* Parent supports EEH. */
- dn->eeh_mode |= EEH_MODE_SUPPORTED;
- dn->eeh_config_addr = dn->parent->eeh_config_addr;
+ pdn->eeh_mode |= EEH_MODE_SUPPORTED;
+ pdn->eeh_config_addr = PCI_DN(dn->parent)->eeh_config_addr;
return NULL;
}
}
dn->full_name);
}
- return NULL;
+ return NULL;
}
/*
struct device_node *phb, *np;
struct eeh_early_enable_info info;
+ spin_lock_init(&slot_errbuf_lock);
+
np = of_find_node_by_path("/rtas");
if (np == NULL)
return;
unsigned long buid;
buid = get_phb_buid(phb);
- if (buid == 0)
+ if (buid == 0 || PCI_DN(phb) == NULL)
continue;
info.buid_lo = BUID_LO(buid);
struct pci_controller *phb;
struct eeh_early_enable_info info;
- if (!dn)
+ if (!dn || !PCI_DN(dn))
return;
- phb = dn->phb;
+ phb = PCI_DN(dn)->phb;
if (NULL == phb || 0 == phb->buid) {
- printk(KERN_WARNING "EEH: Expected buid but found none\n");
+ printk(KERN_WARNING "EEH: Expected buid but found none for %s\n",
+ dn->full_name);
+ dump_stack();
return;
}
info.buid_lo = BUID_LO(phb->buid);
early_enable_eeh(dn, &info);
}
-EXPORT_SYMBOL(eeh_add_device_early);
+EXPORT_SYMBOL_GPL(eeh_add_device_early);
/**
* eeh_add_device_late - perform EEH initialization for the indicated pci device
*/
void eeh_add_device_late(struct pci_dev *dev)
{
+ struct device_node *dn;
+
if (!dev || !eeh_subsystem_enabled)
return;
printk(KERN_DEBUG "EEH: adding device %s\n", pci_name(dev));
#endif
+ pci_dev_get (dev);
+ dn = pci_device_to_OF_node(dev);
+ PCI_DN(dn)->pcidev = dev;
+
pci_addr_cache_insert_device (dev);
}
-EXPORT_SYMBOL(eeh_add_device_late);
+EXPORT_SYMBOL_GPL(eeh_add_device_late);
/**
* eeh_remove_device - undo EEH setup for the indicated pci device
*/
void eeh_remove_device(struct pci_dev *dev)
{
+ struct device_node *dn;
if (!dev || !eeh_subsystem_enabled)
return;
printk(KERN_DEBUG "EEH: remove device %s\n", pci_name(dev));
#endif
pci_addr_cache_remove_device(dev);
+
+ dn = pci_device_to_OF_node(dev);
+ PCI_DN(dn)->pcidev = NULL;
+ pci_dev_put (dev);
}
-EXPORT_SYMBOL(eeh_remove_device);
+EXPORT_SYMBOL_GPL(eeh_remove_device);
static int proc_eeh_show(struct seq_file *m, void *v)
{
unsigned int cpu;
unsigned long ffs = 0, positives = 0, failures = 0;
unsigned long resets = 0;
+ unsigned long no_dev = 0, no_dn = 0, no_cfg = 0, no_check = 0;
for_each_cpu(cpu) {
ffs += per_cpu(total_mmio_ffs, cpu);
positives += per_cpu(false_positives, cpu);
failures += per_cpu(ignored_failures, cpu);
resets += per_cpu(slot_resets, cpu);
+ no_dev += per_cpu(no_device, cpu);
+ no_dn += per_cpu(no_dn, cpu);
+ no_cfg += per_cpu(no_cfg_addr, cpu);
+ no_check += per_cpu(ignored_check, cpu);
}
if (0 == eeh_subsystem_enabled) {
seq_printf(m, "eeh_total_mmio_ffs=%ld\n", ffs);
} else {
seq_printf(m, "EEH Subsystem is enabled\n");
- seq_printf(m, "eeh_total_mmio_ffs=%ld\n"
- "eeh_false_positives=%ld\n"
- "eeh_ignored_failures=%ld\n"
- "eeh_slot_resets=%ld\n"
- "eeh_fail_count=%d\n",
- ffs, positives, failures, resets,
- eeh_fail_count.counter);
+ seq_printf(m,
+ "no device=%ld\n"
+ "no device node=%ld\n"
+ "no config address=%ld\n"
+ "check not wanted=%ld\n"
+ "eeh_total_mmio_ffs=%ld\n"
+ "eeh_false_positives=%ld\n"
+ "eeh_ignored_failures=%ld\n"
+ "eeh_slot_resets=%ld\n",
+ no_dev, no_dn, no_cfg, no_check,
+ ffs, positives, failures, resets);
}
return 0;
projects / mirror_ubuntu-bionic-kernel.git / blobdiff