* Purpose: Generic MCA handling layer
*
* Copyright (C) 2004 FUJITSU LIMITED
- * Copyright (C) Hidetoshi Seto (seto.hidetoshi@jp.fujitsu.com)
+ * Copyright (C) 2004 Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
* Copyright (C) 2005 Silicon Graphics, Inc
* Copyright (C) 2005 Keith Owens <kaos@sgi.com>
+ * Copyright (C) 2006 Russ Anderson <rja@sgi.com>
*/
-#include <linux/config.h>
#include <linux/types.h>
#include <linux/init.h>
#include <linux/sched.h>
#include <linux/interrupt.h>
#include <linux/irq.h>
#include <linux/kallsyms.h>
-#include <linux/smp_lock.h>
#include <linux/bootmem.h>
#include <linux/acpi.h>
#include <linux/timer.h>
ISOLATE_NONE
} isolate_status_t;
+typedef enum {
+ MCA_NOT_RECOVERED = 0,
+ MCA_RECOVERED = 1
+} recovery_status_t;
+
/*
* This pool keeps pointers to the section part of SAL error record
*/
int max_idx; /* Maximum index of section pointer list pool */
} slidx_pool;
+static int
+fatal_mca(const char *fmt, ...)
+{
+ va_list args;
+ char buf[256];
+
+ va_start(args, fmt);
+ vsnprintf(buf, sizeof(buf), fmt, args);
+ va_end(args);
+ ia64_mca_printk(KERN_ALERT "MCA: %s\n", buf);
+
+ return MCA_NOT_RECOVERED;
+}
+
+static int
+mca_recovered(const char *fmt, ...)
+{
+ va_list args;
+ char buf[256];
+
+ va_start(args, fmt);
+ vsnprintf(buf, sizeof(buf), fmt, args);
+ va_end(args);
+ ia64_mca_printk(KERN_INFO "MCA: %s\n", buf);
+
+ return MCA_RECOVERED;
+}
+
/**
* mca_page_isolate - isolate a poisoned page in order not to use it later
* @paddr: poisoned memory location
*/
void
-mca_handler_bh(unsigned long paddr)
+mca_handler_bh(unsigned long paddr, void *iip, unsigned long ipsr)
{
- printk(KERN_ERR
- "OS_MCA: process [pid: %d](%s) encounters MCA (paddr=%lx)\n",
- current->pid, current->comm, paddr);
+ ia64_mlogbuf_dump();
+ printk(KERN_ERR "OS_MCA: process [cpu %d, pid: %d, uid: %d, "
+ "iip: %p, psr: 0x%lx,paddr: 0x%lx](%s) encounters MCA.\n",
+ raw_smp_processor_id(), current->pid, current->uid,
+ iip, ipsr, paddr, current->comm);
spin_lock(&mca_bh_lock);
switch (mca_page_isolate(paddr)) {
return MCA_IS_GLOBAL;
}
+/**
+ * get_target_identifier - Get the valid Cache or Bus check target identifier.
+ * @peidx: pointer of index of processor error section
+ *
+ * Return value:
+ * target address on Success / 0 on Failure
+ */
+static u64
+get_target_identifier(peidx_table_t *peidx)
+{
+ u64 target_address = 0;
+ sal_log_mod_error_info_t *smei;
+ pal_cache_check_info_t *pcci;
+ int i, level = 9;
+
+ /*
+ * Look through the cache checks for a valid target identifier
+ * If more than one valid target identifier, return the one
+ * with the lowest cache level.
+ */
+ for (i = 0; i < peidx_cache_check_num(peidx); i++) {
+ smei = (sal_log_mod_error_info_t *)peidx_cache_check(peidx, i);
+ if (smei->valid.target_identifier && smei->target_identifier) {
+ pcci = (pal_cache_check_info_t *)&(smei->check_info);
+ if (!target_address || (pcci->level < level)) {
+ target_address = smei->target_identifier;
+ level = pcci->level;
+ continue;
+ }
+ }
+ }
+ if (target_address)
+ return target_address;
+
+ /*
+ * Look at the bus check for a valid target identifier
+ */
+ smei = peidx_bus_check(peidx, 0);
+ if (smei && smei->valid.target_identifier)
+ return smei->target_identifier;
+
+ return 0;
+}
+
/**
* recover_from_read_error - Try to recover the errors which type are "read"s.
* @slidx: pointer of index of SAL error record
peidx_table_t *peidx, pal_bus_check_info_t *pbci,
struct ia64_sal_os_state *sos)
{
- sal_log_mod_error_info_t *smei;
+ u64 target_identifier;
pal_min_state_area_t *pmsa;
struct ia64_psr *psr1, *psr2;
ia64_fptr_t *mca_hdlr_bh = (ia64_fptr_t*)mca_handler_bhhook;
/* Is target address valid? */
- if (!pbci->tv)
- return 0;
+ target_identifier = get_target_identifier(peidx);
+ if (!target_identifier)
+ return fatal_mca("target address not valid");
/*
* cpu read or memory-mapped io read
/* Is minstate valid? */
if (!peidx_bottom(peidx) || !(peidx_bottom(peidx)->valid.minstate))
- return 0;
+ return fatal_mca("minstate not valid");
psr1 =(struct ia64_psr *)&(peidx_minstate_area(peidx)->pmsa_ipsr);
+ psr2 =(struct ia64_psr *)&(peidx_minstate_area(peidx)->pmsa_xpsr);
/*
* Check the privilege level of interrupted context.
* If it is user-mode, then terminate affected process.
*/
- if (psr1->cpl != 0) {
- smei = peidx_bus_check(peidx, 0);
- if (smei->valid.target_identifier) {
- /*
- * setup for resume to bottom half of MCA,
- * "mca_handler_bhhook"
- */
- pmsa = sos->pal_min_state;
- /* pass to bhhook as 1st argument (gr8) */
- pmsa->pmsa_gr[8-1] = smei->target_identifier;
- /* set interrupted return address (but no use) */
- pmsa->pmsa_br0 = pmsa->pmsa_iip;
- /* change resume address to bottom half */
- pmsa->pmsa_iip = mca_hdlr_bh->fp;
- pmsa->pmsa_gr[1-1] = mca_hdlr_bh->gp;
- /* set cpl with kernel mode */
- psr2 = (struct ia64_psr *)&pmsa->pmsa_ipsr;
- psr2->cpl = 0;
- psr2->ri = 0;
- psr2->i = 0;
-
- return 1;
- }
+ pmsa = sos->pal_min_state;
+ if (psr1->cpl != 0 ||
+ ((psr2->cpl != 0) && mca_recover_range(pmsa->pmsa_iip))) {
+ /*
+ * setup for resume to bottom half of MCA,
+ * "mca_handler_bhhook"
+ */
+ /* pass to bhhook as argument (gr8, ...) */
+ pmsa->pmsa_gr[8-1] = target_identifier;
+ pmsa->pmsa_gr[9-1] = pmsa->pmsa_iip;
+ pmsa->pmsa_gr[10-1] = pmsa->pmsa_ipsr;
+ /* set interrupted return address (but no use) */
+ pmsa->pmsa_br0 = pmsa->pmsa_iip;
+ /* change resume address to bottom half */
+ pmsa->pmsa_iip = mca_hdlr_bh->fp;
+ pmsa->pmsa_gr[1-1] = mca_hdlr_bh->gp;
+ /* set cpl with kernel mode */
+ psr2 = (struct ia64_psr *)&pmsa->pmsa_ipsr;
+ psr2->cpl = 0;
+ psr2->ri = 0;
+ psr2->bn = 1;
+ psr2->i = 0;
+
+ return mca_recovered("user memory corruption. "
+ "kill affected process - recovered.");
}
- return 0;
+ return fatal_mca("kernel context not recovered, iip 0x%lx\n",
+ pmsa->pmsa_iip);
}
/**
default:
break;
}
+ } else if (psp->cc && !psp->bc) { /* Cache error */
+ status = recover_from_read_error(slidx, peidx, pbci, sos);
}
return status;
}
+/*
+ * recover_from_tlb_check
+ * @peidx: pointer of index of processor error section
+ *
+ * Return value:
+ * 1 on Success / 0 on Failure
+ */
+static int
+recover_from_tlb_check(peidx_table_t *peidx)
+{
+ sal_log_mod_error_info_t *smei;
+ pal_tlb_check_info_t *ptci;
+
+ smei = (sal_log_mod_error_info_t *)peidx_tlb_check(peidx, 0);
+ ptci = (pal_tlb_check_info_t *)&(smei->check_info);
+
+ /*
+ * Look for signature of a duplicate TLB DTC entry, which is
+ * a SW bug and always fatal.
+ */
+ if (ptci->op == PAL_TLB_CHECK_OP_PURGE
+ && !(ptci->itr || ptci->dtc || ptci->itc))
+ return fatal_mca("Duplicate TLB entry");
+
+ return mca_recovered("TLB check recovered");
+}
+
/**
* recover_from_processor_error
* @platform: whether there are some platform error section or not
* Return value:
* 1 on Success / 0 on Failure
*/
-/*
- * Later we try to recover when below all conditions are satisfied.
- * 1. Only one processor error section is exist.
- * 2. BUS_CHECK is exist and the others are not exist.(Except TLB_CHECK)
- * 3. The entry of BUS_CHECK_INFO is 1.
- * 4. "External bus error" flag is set and the others are not set.
- */
static int
recover_from_processor_error(int platform, slidx_table_t *slidx,
* The machine check is corrected.
*/
if (psp->cm == 1)
- return 1;
+ return mca_recovered("machine check is already corrected.");
/*
* The error was not contained. Software must be reset.
*/
if (psp->us || psp->ci == 0)
- return 0;
+ return fatal_mca("error not contained");
+
+ /*
+ * Look for recoverable TLB check
+ */
+ if (psp->tc && !(psp->cc || psp->bc || psp->rc || psp->uc))
+ return recover_from_tlb_check(peidx);
/*
* The cache check and bus check bits have four possible states
* cc bc
- * 0 0 Weird record, not recovered
- * 1 0 Cache error, not recovered
- * 0 1 I/O error, attempt recovery
* 1 1 Memory error, attempt recovery
+ * 1 0 Cache error, attempt recovery
+ * 0 1 I/O error, attempt recovery
+ * 0 0 Other error type, not recovered
*/
- if (psp->bc == 0 || pbci == NULL)
- return 0;
+ if (psp->cc == 0 && (psp->bc == 0 || pbci == NULL))
+ return fatal_mca("No cache or bus check");
/*
- * Sorry, we cannot handle so many.
+ * Cannot handle more than one bus check.
*/
if (peidx_bus_check_num(peidx) > 1)
- return 0;
- /*
- * Well, here is only one bus error.
- */
- if (pbci->ib || pbci->cc)
- return 0;
+ return fatal_mca("Too many bus checks");
+
+ if (pbci->ib)
+ return fatal_mca("Internal Bus error");
if (pbci->eb && pbci->bsi > 0)
- return 0;
+ return fatal_mca("External bus check fatal status");
/*
- * This is a local MCA and estimated as recoverble external bus error.
- * (e.g. a load from poisoned memory)
- * This means "there are some platform errors".
+ * This is a local MCA and estimated as a recoverable error.
*/
if (platform)
return recover_from_platform_error(slidx, peidx, pbci, sos);
+
/*
* On account of strange SAL error record, we cannot recover.
*/
- return 0;
+ return fatal_mca("Strange SAL record");
}
/**
/* Now, OS can recover when there is one processor error section */
if (n_proc_err > 1)
- return 0;
- else if (n_proc_err == 0) {
- /* Weird SAL record ... We need not to recover */
-
- return 1;
- }
+ return fatal_mca("Too Many Errors");
+ else if (n_proc_err == 0)
+ /* Weird SAL record ... We can't do anything */
+ return fatal_mca("Weird SAL record");
/* Make index of processor error section */
mca_make_peidx((sal_log_processor_info_t*)
/* Check whether MCA is global or not */
if (is_mca_global(&peidx, &pbci, sos))
- return 0;
+ return fatal_mca("global MCA");
/* Try to recover a processor error */
return recover_from_processor_error(platform_err, &slidx, &peidx,