linux/arch/powerpc/platforms/pseries/ras.c

373 lines
10 KiB
C
Raw Normal View History

/*
* Copyright (C) 2001 Dave Engebretsen 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
*/
/* Change Activity:
* 2001/09/21 : engebret : Created with minimal EPOW and HW exception support.
* End Change Activity
*/
#include <linux/errno.h>
#include <linux/threads.h>
#include <linux/kernel_stat.h>
#include <linux/signal.h>
#include <linux/sched.h>
#include <linux/ioport.h>
#include <linux/interrupt.h>
#include <linux/timex.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/pci.h>
#include <linux/delay.h>
#include <linux/irq.h>
#include <linux/random.h>
#include <linux/sysrq.h>
#include <linux/bitops.h>
#include <asm/uaccess.h>
#include <asm/system.h>
#include <asm/io.h>
#include <asm/pgtable.h>
#include <asm/irq.h>
#include <asm/cache.h>
#include <asm/prom.h>
#include <asm/ptrace.h>
#include <asm/machdep.h>
#include <asm/rtas.h>
#include <asm/udbg.h>
#include <asm/firmware.h>
#include "pseries.h"
static unsigned char ras_log_buf[RTAS_ERROR_LOG_MAX];
static DEFINE_SPINLOCK(ras_log_buf_lock);
char mce_data_buf[RTAS_ERROR_LOG_MAX];
static int ras_get_sensor_state_token;
static int ras_check_exception_token;
#define EPOW_SENSOR_TOKEN 9
#define EPOW_SENSOR_INDEX 0
#define RAS_VECTOR_OFFSET 0x500
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
static irqreturn_t ras_epow_interrupt(int irq, void *dev_id);
static irqreturn_t ras_error_interrupt(int irq, void *dev_id);
/* #define DEBUG */
2006-07-03 11:36:01 +00:00
static void request_ras_irqs(struct device_node *np,
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
irq_handler_t handler,
const char *name)
{
2006-07-03 11:36:01 +00:00
int i, index, count = 0;
struct of_irq oirq;
const u32 *opicprop;
2006-07-03 11:36:01 +00:00
unsigned int opicplen;
unsigned int virqs[16];
/* Check for obsolete "open-pic-interrupt" property. If present, then
* map those interrupts using the default interrupt host and default
* trigger
*/
opicprop = get_property(np, "open-pic-interrupt", &opicplen);
2006-07-03 11:36:01 +00:00
if (opicprop) {
opicplen /= sizeof(u32);
for (i = 0; i < opicplen; i++) {
if (count > 15)
break;
[PATCH] powerpc: fix trigger handling in the new irq code This patch slightly reworks the new irq code to fix a small design error. I removed the passing of the trigger to the map() calls entirely, it was not a good idea to have one call do two different things. It also fixes a couple of corner cases. Mapping a linux virtual irq to a physical irq now does only that. Setting the trigger is a different action which has a different call. The main changes are: - I no longer call host->ops->map() for an already mapped irq, I just return the virtual number that was already mapped. It was called before to give an opportunity to change the trigger, but that was causing issues as that could happen while the interrupt was in use by a device, and because of the trigger change, map would potentially muck around with things in a racy way. That was causing much burden on a given's controller implementation of map() to get it right. This is much simpler now. map() is only called on the initial mapping of an irq, meaning that you know that this irq is _not_ being used. You can initialize the hardware if you want (though you don't have to). - Controllers that can handle different type of triggers (level/edge/etc...) now implement the standard irq_chip->set_type() call as defined by the generic code. That means that you can use the standard set_irq_type() to configure an irq line manually if you wish or (though I don't like that interface), pass explicit trigger flags to request_irq() as defined by the generic kernel interfaces. Also, using those interfaces guarantees that your controller set_type callback is called with the descriptor lock held, thus providing locking against activity on the same interrupt (including mask/unmask/etc...) automatically. A result is that, for example, MPIC's own map() implementation calls irq_set_type(NONE) to configure the hardware to the default triggers. - To allow the above, the irq_map array entry for the new mapped interrupt is now set before map() callback is called for the controller. - The irq_create_of_mapping() (also used by irq_of_parse_and_map()) function for mapping interrupts from the device-tree now also call the separate set_irq_type(), and only does so if there is a change in the trigger type. - While I was at it, I changed pci_read_irq_line() (which is the helper I would expect most archs to use in their pcibios_fixup() to get the PCI interrupt routing from the device tree) to also handle a fallback when the DT mapping fails consisting of reading the PCI_INTERRUPT_PIN to know wether the device has an interrupt at all, and the the PCI_INTERRUPT_LINE to get an interrupt number from the device. That number is then mapped using the default controller, and the trigger is set to level low. That default behaviour works for several platforms that don't have a proper interrupt tree like Pegasos. If it doesn't work for your platform, then either provide a proper interrupt tree from the firmware so that fallback isn't needed, or don't call pci_read_irq_line() - Add back a bit that got dropped by my main rework patch for properly clearing pending IPIs on pSeries when using a kexec Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-07-10 11:44:42 +00:00
virqs[count] = irq_create_mapping(NULL, *(opicprop++));
2006-07-03 11:36:01 +00:00
if (virqs[count] == NO_IRQ)
printk(KERN_ERR "Unable to allocate interrupt "
"number for %s\n", np->full_name);
else
count++;
}
2006-07-03 11:36:01 +00:00
}
/* Else use normal interrupt tree parsing */
else {
/* First try to do a proper OF tree parsing */
for (index = 0; of_irq_map_one(np, index, &oirq) == 0;
index++) {
if (count > 15)
break;
virqs[count] = irq_create_of_mapping(oirq.controller,
oirq.specifier,
oirq.size);
if (virqs[count] == NO_IRQ)
printk(KERN_ERR "Unable to allocate interrupt "
"number for %s\n", np->full_name);
else
count++;
}
}
/* Now request them */
for (i = 0; i < count; i++) {
if (request_irq(virqs[i], handler, 0, name, NULL)) {
printk(KERN_ERR "Unable to request interrupt %d for "
2006-07-03 11:36:01 +00:00
"%s\n", virqs[i], np->full_name);
return;
}
}
}
/*
* Initialize handlers for the set of interrupts caused by hardware errors
* and power system events.
*/
static int __init init_ras_IRQ(void)
{
struct device_node *np;
ras_get_sensor_state_token = rtas_token("get-sensor-state");
ras_check_exception_token = rtas_token("check-exception");
/* Internal Errors */
np = of_find_node_by_path("/event-sources/internal-errors");
if (np != NULL) {
2006-07-03 11:36:01 +00:00
request_ras_irqs(np, ras_error_interrupt, "RAS_ERROR");
of_node_put(np);
}
/* EPOW Events */
np = of_find_node_by_path("/event-sources/epow-events");
if (np != NULL) {
2006-07-03 11:36:01 +00:00
request_ras_irqs(np, ras_epow_interrupt, "RAS_EPOW");
of_node_put(np);
}
return 0;
}
__initcall(init_ras_IRQ);
/*
* Handle power subsystem events (EPOW).
*
* Presently we just log the event has occurred. This should be fixed
* to examine the type of power failure and take appropriate action where
* the time horizon permits something useful to be done.
*/
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
static irqreturn_t ras_epow_interrupt(int irq, void *dev_id)
{
int status = 0xdeadbeef;
int state = 0;
int critical;
status = rtas_call(ras_get_sensor_state_token, 2, 2, &state,
EPOW_SENSOR_TOKEN, EPOW_SENSOR_INDEX);
if (state > 3)
critical = 1; /* Time Critical */
else
critical = 0;
spin_lock(&ras_log_buf_lock);
status = rtas_call(ras_check_exception_token, 6, 1, NULL,
RAS_VECTOR_OFFSET,
2006-07-03 11:36:01 +00:00
irq_map[irq].hwirq,
RTAS_EPOW_WARNING | RTAS_POWERMGM_EVENTS,
critical, __pa(&ras_log_buf),
rtas_get_error_log_max());
udbg_printf("EPOW <0x%lx 0x%x 0x%x>\n",
*((unsigned long *)&ras_log_buf), status, state);
printk(KERN_WARNING "EPOW <0x%lx 0x%x 0x%x>\n",
*((unsigned long *)&ras_log_buf), status, state);
/* format and print the extended information */
log_error(ras_log_buf, ERR_TYPE_RTAS_LOG, 0);
spin_unlock(&ras_log_buf_lock);
return IRQ_HANDLED;
}
/*
* Handle hardware error interrupts.
*
* RTAS check-exception is called to collect data on the exception. If
* the error is deemed recoverable, we log a warning and return.
* For nonrecoverable errors, an error is logged and we stop all processing
* as quickly as possible in order to prevent propagation of the failure.
*/
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 13:55:46 +00:00
static irqreturn_t ras_error_interrupt(int irq, void *dev_id)
{
struct rtas_error_log *rtas_elog;
int status = 0xdeadbeef;
int fatal;
spin_lock(&ras_log_buf_lock);
status = rtas_call(ras_check_exception_token, 6, 1, NULL,
RAS_VECTOR_OFFSET,
2006-07-03 11:36:01 +00:00
irq_map[irq].hwirq,
RTAS_INTERNAL_ERROR, 1 /*Time Critical */,
__pa(&ras_log_buf),
rtas_get_error_log_max());
rtas_elog = (struct rtas_error_log *)ras_log_buf;
if ((status == 0) && (rtas_elog->severity >= RTAS_SEVERITY_ERROR_SYNC))
fatal = 1;
else
fatal = 0;
/* format and print the extended information */
log_error(ras_log_buf, ERR_TYPE_RTAS_LOG, fatal);
if (fatal) {
udbg_printf("Fatal HW Error <0x%lx 0x%x>\n",
*((unsigned long *)&ras_log_buf), status);
printk(KERN_EMERG "Error: Fatal hardware error <0x%lx 0x%x>\n",
*((unsigned long *)&ras_log_buf), status);
#ifndef DEBUG
/* Don't actually power off when debugging so we can test
* without actually failing while injecting errors.
* Error data will not be logged to syslog.
*/
ppc_md.power_off();
#endif
} else {
udbg_printf("Recoverable HW Error <0x%lx 0x%x>\n",
*((unsigned long *)&ras_log_buf), status);
printk(KERN_WARNING
"Warning: Recoverable hardware error <0x%lx 0x%x>\n",
*((unsigned long *)&ras_log_buf), status);
}
spin_unlock(&ras_log_buf_lock);
return IRQ_HANDLED;
}
/* Get the error information for errors coming through the
* FWNMI vectors. The pt_regs' r3 will be updated to reflect
* the actual r3 if possible, and a ptr to the error log entry
* will be returned if found.
*
* The mce_data_buf does not have any locks or protection around it,
* if a second machine check comes in, or a system reset is done
* before we have logged the error, then we will get corruption in the
* error log. This is preferable over holding off on calling
* ibm,nmi-interlock which would result in us checkstopping if a
* second machine check did come in.
*/
static struct rtas_error_log *fwnmi_get_errinfo(struct pt_regs *regs)
{
unsigned long errdata = regs->gpr[3];
struct rtas_error_log *errhdr = NULL;
unsigned long *savep;
if ((errdata >= 0x7000 && errdata < 0x7fff0) ||
(errdata >= rtas.base && errdata < rtas.base + rtas.size - 16)) {
savep = __va(errdata);
regs->gpr[3] = savep[0]; /* restore original r3 */
memset(mce_data_buf, 0, RTAS_ERROR_LOG_MAX);
memcpy(mce_data_buf, (char *)(savep + 1), RTAS_ERROR_LOG_MAX);
errhdr = (struct rtas_error_log *)mce_data_buf;
} else {
printk("FWNMI: corrupt r3\n");
}
return errhdr;
}
/* Call this when done with the data returned by FWNMI_get_errinfo.
* It will release the saved data area for other CPUs in the
* partition to receive FWNMI errors.
*/
static void fwnmi_release_errinfo(void)
{
int ret = rtas_call(rtas_token("ibm,nmi-interlock"), 0, 1, NULL);
if (ret != 0)
printk("FWNMI: nmi-interlock failed: %d\n", ret);
}
int pSeries_system_reset_exception(struct pt_regs *regs)
{
if (fwnmi_active) {
struct rtas_error_log *errhdr = fwnmi_get_errinfo(regs);
if (errhdr) {
/* XXX Should look at FWNMI information */
}
fwnmi_release_errinfo();
}
return 0; /* need to perform reset */
}
/*
* See if we can recover from a machine check exception.
* This is only called on power4 (or above) and only via
* the Firmware Non-Maskable Interrupts (fwnmi) handler
* which provides the error analysis for us.
*
* Return 1 if corrected (or delivered a signal).
* Return 0 if there is nothing we can do.
*/
static int recover_mce(struct pt_regs *regs, struct rtas_error_log * err)
{
int nonfatal = 0;
if (err->disposition == RTAS_DISP_FULLY_RECOVERED) {
/* Platform corrected itself */
nonfatal = 1;
} else if ((regs->msr & MSR_RI) &&
user_mode(regs) &&
err->severity == RTAS_SEVERITY_ERROR_SYNC &&
err->disposition == RTAS_DISP_NOT_RECOVERED &&
err->target == RTAS_TARGET_MEMORY &&
err->type == RTAS_TYPE_ECC_UNCORR &&
!(current->pid == 0 || is_init(current))) {
/* Kill off a user process with an ECC error */
printk(KERN_ERR "MCE: uncorrectable ecc error for pid %d\n",
current->pid);
/* XXX something better for ECC error? */
_exception(SIGBUS, regs, BUS_ADRERR, regs->nip);
nonfatal = 1;
}
log_error((char *)err, ERR_TYPE_RTAS_LOG, !nonfatal);
return nonfatal;
}
/*
* Handle a machine check.
*
* Note that on Power 4 and beyond Firmware Non-Maskable Interrupts (fwnmi)
* should be present. If so the handler which called us tells us if the
* error was recovered (never true if RI=0).
*
* On hardware prior to Power 4 these exceptions were asynchronous which
* means we can't tell exactly where it occurred and so we can't recover.
*/
int pSeries_machine_check_exception(struct pt_regs *regs)
{
struct rtas_error_log *errp;
if (fwnmi_active) {
errp = fwnmi_get_errinfo(regs);
fwnmi_release_errinfo();
if (errp && recover_mce(regs, errp))
return 1;
}
return 0;
}