7c07012eb1
note_interrupt() increments desc->irq_count for each interrupt even for percpu interrupt handlers, even when they are handled successfully. This causes cacheline bouncing and limits scalability. Instead of incrementing irq_count every time, only start incrementing it after seeing an unhandled irq, which should avoid the cache line bouncing in the common path. This actually should give better consistency in handling misbehaving irqs too, because instead of the first unhandled irq arriving at an arbitrary point in the irq_count cycle, its arrival will begin the irq_count cycle. Cédric reports the result of his IPI throughput test: Millions of IPIs/s ----------- -------------------------------------- upstream upstream patched chips cpus default noirqdebug default (irqdebug) ----------- ----------------------------------------- 1 0-15 4.061 4.153 4.084 0-31 7.937 8.186 8.158 0-47 11.018 11.392 11.233 0-63 11.460 13.907 14.022 2 0-79 8.376 18.105 18.084 0-95 7.338 22.101 22.266 0-111 6.716 25.306 25.473 0-127 6.223 27.814 28.029 Signed-off-by: Nicholas Piggin <npiggin@gmail.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Link: https://lore.kernel.org/r/20210402132037.574661-1-npiggin@gmail.com
471 lines
12 KiB
C
471 lines
12 KiB
C
// SPDX-License-Identifier: GPL-2.0
|
|
/*
|
|
* Copyright (C) 1992, 1998-2004 Linus Torvalds, Ingo Molnar
|
|
*
|
|
* This file contains spurious interrupt handling.
|
|
*/
|
|
|
|
#include <linux/jiffies.h>
|
|
#include <linux/irq.h>
|
|
#include <linux/module.h>
|
|
#include <linux/interrupt.h>
|
|
#include <linux/moduleparam.h>
|
|
#include <linux/timer.h>
|
|
|
|
#include "internals.h"
|
|
|
|
static int irqfixup __read_mostly;
|
|
|
|
#define POLL_SPURIOUS_IRQ_INTERVAL (HZ/10)
|
|
static void poll_spurious_irqs(struct timer_list *unused);
|
|
static DEFINE_TIMER(poll_spurious_irq_timer, poll_spurious_irqs);
|
|
static int irq_poll_cpu;
|
|
static atomic_t irq_poll_active;
|
|
|
|
/*
|
|
* We wait here for a poller to finish.
|
|
*
|
|
* If the poll runs on this CPU, then we yell loudly and return
|
|
* false. That will leave the interrupt line disabled in the worst
|
|
* case, but it should never happen.
|
|
*
|
|
* We wait until the poller is done and then recheck disabled and
|
|
* action (about to be disabled). Only if it's still active, we return
|
|
* true and let the handler run.
|
|
*/
|
|
bool irq_wait_for_poll(struct irq_desc *desc)
|
|
__must_hold(&desc->lock)
|
|
{
|
|
if (WARN_ONCE(irq_poll_cpu == smp_processor_id(),
|
|
"irq poll in progress on cpu %d for irq %d\n",
|
|
smp_processor_id(), desc->irq_data.irq))
|
|
return false;
|
|
|
|
#ifdef CONFIG_SMP
|
|
do {
|
|
raw_spin_unlock(&desc->lock);
|
|
while (irqd_irq_inprogress(&desc->irq_data))
|
|
cpu_relax();
|
|
raw_spin_lock(&desc->lock);
|
|
} while (irqd_irq_inprogress(&desc->irq_data));
|
|
/* Might have been disabled in meantime */
|
|
return !irqd_irq_disabled(&desc->irq_data) && desc->action;
|
|
#else
|
|
return false;
|
|
#endif
|
|
}
|
|
|
|
|
|
/*
|
|
* Recovery handler for misrouted interrupts.
|
|
*/
|
|
static int try_one_irq(struct irq_desc *desc, bool force)
|
|
{
|
|
irqreturn_t ret = IRQ_NONE;
|
|
struct irqaction *action;
|
|
|
|
raw_spin_lock(&desc->lock);
|
|
|
|
/*
|
|
* PER_CPU, nested thread interrupts and interrupts explicitly
|
|
* marked polled are excluded from polling.
|
|
*/
|
|
if (irq_settings_is_per_cpu(desc) ||
|
|
irq_settings_is_nested_thread(desc) ||
|
|
irq_settings_is_polled(desc))
|
|
goto out;
|
|
|
|
/*
|
|
* Do not poll disabled interrupts unless the spurious
|
|
* disabled poller asks explicitly.
|
|
*/
|
|
if (irqd_irq_disabled(&desc->irq_data) && !force)
|
|
goto out;
|
|
|
|
/*
|
|
* All handlers must agree on IRQF_SHARED, so we test just the
|
|
* first.
|
|
*/
|
|
action = desc->action;
|
|
if (!action || !(action->flags & IRQF_SHARED) ||
|
|
(action->flags & __IRQF_TIMER))
|
|
goto out;
|
|
|
|
/* Already running on another processor */
|
|
if (irqd_irq_inprogress(&desc->irq_data)) {
|
|
/*
|
|
* Already running: If it is shared get the other
|
|
* CPU to go looking for our mystery interrupt too
|
|
*/
|
|
desc->istate |= IRQS_PENDING;
|
|
goto out;
|
|
}
|
|
|
|
/* Mark it poll in progress */
|
|
desc->istate |= IRQS_POLL_INPROGRESS;
|
|
do {
|
|
if (handle_irq_event(desc) == IRQ_HANDLED)
|
|
ret = IRQ_HANDLED;
|
|
/* Make sure that there is still a valid action */
|
|
action = desc->action;
|
|
} while ((desc->istate & IRQS_PENDING) && action);
|
|
desc->istate &= ~IRQS_POLL_INPROGRESS;
|
|
out:
|
|
raw_spin_unlock(&desc->lock);
|
|
return ret == IRQ_HANDLED;
|
|
}
|
|
|
|
static int misrouted_irq(int irq)
|
|
{
|
|
struct irq_desc *desc;
|
|
int i, ok = 0;
|
|
|
|
if (atomic_inc_return(&irq_poll_active) != 1)
|
|
goto out;
|
|
|
|
irq_poll_cpu = smp_processor_id();
|
|
|
|
for_each_irq_desc(i, desc) {
|
|
if (!i)
|
|
continue;
|
|
|
|
if (i == irq) /* Already tried */
|
|
continue;
|
|
|
|
if (try_one_irq(desc, false))
|
|
ok = 1;
|
|
}
|
|
out:
|
|
atomic_dec(&irq_poll_active);
|
|
/* So the caller can adjust the irq error counts */
|
|
return ok;
|
|
}
|
|
|
|
static void poll_spurious_irqs(struct timer_list *unused)
|
|
{
|
|
struct irq_desc *desc;
|
|
int i;
|
|
|
|
if (atomic_inc_return(&irq_poll_active) != 1)
|
|
goto out;
|
|
irq_poll_cpu = smp_processor_id();
|
|
|
|
for_each_irq_desc(i, desc) {
|
|
unsigned int state;
|
|
|
|
if (!i)
|
|
continue;
|
|
|
|
/* Racy but it doesn't matter */
|
|
state = desc->istate;
|
|
barrier();
|
|
if (!(state & IRQS_SPURIOUS_DISABLED))
|
|
continue;
|
|
|
|
local_irq_disable();
|
|
try_one_irq(desc, true);
|
|
local_irq_enable();
|
|
}
|
|
out:
|
|
atomic_dec(&irq_poll_active);
|
|
mod_timer(&poll_spurious_irq_timer,
|
|
jiffies + POLL_SPURIOUS_IRQ_INTERVAL);
|
|
}
|
|
|
|
static inline int bad_action_ret(irqreturn_t action_ret)
|
|
{
|
|
unsigned int r = action_ret;
|
|
|
|
if (likely(r <= (IRQ_HANDLED | IRQ_WAKE_THREAD)))
|
|
return 0;
|
|
return 1;
|
|
}
|
|
|
|
/*
|
|
* If 99,900 of the previous 100,000 interrupts have not been handled
|
|
* then assume that the IRQ is stuck in some manner. Drop a diagnostic
|
|
* and try to turn the IRQ off.
|
|
*
|
|
* (The other 100-of-100,000 interrupts may have been a correctly
|
|
* functioning device sharing an IRQ with the failing one)
|
|
*/
|
|
static void __report_bad_irq(struct irq_desc *desc, irqreturn_t action_ret)
|
|
{
|
|
unsigned int irq = irq_desc_get_irq(desc);
|
|
struct irqaction *action;
|
|
unsigned long flags;
|
|
|
|
if (bad_action_ret(action_ret)) {
|
|
printk(KERN_ERR "irq event %d: bogus return value %x\n",
|
|
irq, action_ret);
|
|
} else {
|
|
printk(KERN_ERR "irq %d: nobody cared (try booting with "
|
|
"the \"irqpoll\" option)\n", irq);
|
|
}
|
|
dump_stack();
|
|
printk(KERN_ERR "handlers:\n");
|
|
|
|
/*
|
|
* We need to take desc->lock here. note_interrupt() is called
|
|
* w/o desc->lock held, but IRQ_PROGRESS set. We might race
|
|
* with something else removing an action. It's ok to take
|
|
* desc->lock here. See synchronize_irq().
|
|
*/
|
|
raw_spin_lock_irqsave(&desc->lock, flags);
|
|
for_each_action_of_desc(desc, action) {
|
|
printk(KERN_ERR "[<%p>] %ps", action->handler, action->handler);
|
|
if (action->thread_fn)
|
|
printk(KERN_CONT " threaded [<%p>] %ps",
|
|
action->thread_fn, action->thread_fn);
|
|
printk(KERN_CONT "\n");
|
|
}
|
|
raw_spin_unlock_irqrestore(&desc->lock, flags);
|
|
}
|
|
|
|
static void report_bad_irq(struct irq_desc *desc, irqreturn_t action_ret)
|
|
{
|
|
static int count = 100;
|
|
|
|
if (count > 0) {
|
|
count--;
|
|
__report_bad_irq(desc, action_ret);
|
|
}
|
|
}
|
|
|
|
static inline int
|
|
try_misrouted_irq(unsigned int irq, struct irq_desc *desc,
|
|
irqreturn_t action_ret)
|
|
{
|
|
struct irqaction *action;
|
|
|
|
if (!irqfixup)
|
|
return 0;
|
|
|
|
/* We didn't actually handle the IRQ - see if it was misrouted? */
|
|
if (action_ret == IRQ_NONE)
|
|
return 1;
|
|
|
|
/*
|
|
* But for 'irqfixup == 2' we also do it for handled interrupts if
|
|
* they are marked as IRQF_IRQPOLL (or for irq zero, which is the
|
|
* traditional PC timer interrupt.. Legacy)
|
|
*/
|
|
if (irqfixup < 2)
|
|
return 0;
|
|
|
|
if (!irq)
|
|
return 1;
|
|
|
|
/*
|
|
* Since we don't get the descriptor lock, "action" can
|
|
* change under us. We don't really care, but we don't
|
|
* want to follow a NULL pointer. So tell the compiler to
|
|
* just load it once by using a barrier.
|
|
*/
|
|
action = desc->action;
|
|
barrier();
|
|
return action && (action->flags & IRQF_IRQPOLL);
|
|
}
|
|
|
|
#define SPURIOUS_DEFERRED 0x80000000
|
|
|
|
void note_interrupt(struct irq_desc *desc, irqreturn_t action_ret)
|
|
{
|
|
unsigned int irq;
|
|
|
|
if (desc->istate & IRQS_POLL_INPROGRESS ||
|
|
irq_settings_is_polled(desc))
|
|
return;
|
|
|
|
if (bad_action_ret(action_ret)) {
|
|
report_bad_irq(desc, action_ret);
|
|
return;
|
|
}
|
|
|
|
/*
|
|
* We cannot call note_interrupt from the threaded handler
|
|
* because we need to look at the compound of all handlers
|
|
* (primary and threaded). Aside of that in the threaded
|
|
* shared case we have no serialization against an incoming
|
|
* hardware interrupt while we are dealing with a threaded
|
|
* result.
|
|
*
|
|
* So in case a thread is woken, we just note the fact and
|
|
* defer the analysis to the next hardware interrupt.
|
|
*
|
|
* The threaded handlers store whether they successfully
|
|
* handled an interrupt and we check whether that number
|
|
* changed versus the last invocation.
|
|
*
|
|
* We could handle all interrupts with the delayed by one
|
|
* mechanism, but for the non forced threaded case we'd just
|
|
* add pointless overhead to the straight hardirq interrupts
|
|
* for the sake of a few lines less code.
|
|
*/
|
|
if (action_ret & IRQ_WAKE_THREAD) {
|
|
/*
|
|
* There is a thread woken. Check whether one of the
|
|
* shared primary handlers returned IRQ_HANDLED. If
|
|
* not we defer the spurious detection to the next
|
|
* interrupt.
|
|
*/
|
|
if (action_ret == IRQ_WAKE_THREAD) {
|
|
int handled;
|
|
/*
|
|
* We use bit 31 of thread_handled_last to
|
|
* denote the deferred spurious detection
|
|
* active. No locking necessary as
|
|
* thread_handled_last is only accessed here
|
|
* and we have the guarantee that hard
|
|
* interrupts are not reentrant.
|
|
*/
|
|
if (!(desc->threads_handled_last & SPURIOUS_DEFERRED)) {
|
|
desc->threads_handled_last |= SPURIOUS_DEFERRED;
|
|
return;
|
|
}
|
|
/*
|
|
* Check whether one of the threaded handlers
|
|
* returned IRQ_HANDLED since the last
|
|
* interrupt happened.
|
|
*
|
|
* For simplicity we just set bit 31, as it is
|
|
* set in threads_handled_last as well. So we
|
|
* avoid extra masking. And we really do not
|
|
* care about the high bits of the handled
|
|
* count. We just care about the count being
|
|
* different than the one we saw before.
|
|
*/
|
|
handled = atomic_read(&desc->threads_handled);
|
|
handled |= SPURIOUS_DEFERRED;
|
|
if (handled != desc->threads_handled_last) {
|
|
action_ret = IRQ_HANDLED;
|
|
/*
|
|
* Note: We keep the SPURIOUS_DEFERRED
|
|
* bit set. We are handling the
|
|
* previous invocation right now.
|
|
* Keep it for the current one, so the
|
|
* next hardware interrupt will
|
|
* account for it.
|
|
*/
|
|
desc->threads_handled_last = handled;
|
|
} else {
|
|
/*
|
|
* None of the threaded handlers felt
|
|
* responsible for the last interrupt
|
|
*
|
|
* We keep the SPURIOUS_DEFERRED bit
|
|
* set in threads_handled_last as we
|
|
* need to account for the current
|
|
* interrupt as well.
|
|
*/
|
|
action_ret = IRQ_NONE;
|
|
}
|
|
} else {
|
|
/*
|
|
* One of the primary handlers returned
|
|
* IRQ_HANDLED. So we don't care about the
|
|
* threaded handlers on the same line. Clear
|
|
* the deferred detection bit.
|
|
*
|
|
* In theory we could/should check whether the
|
|
* deferred bit is set and take the result of
|
|
* the previous run into account here as
|
|
* well. But it's really not worth the
|
|
* trouble. If every other interrupt is
|
|
* handled we never trigger the spurious
|
|
* detector. And if this is just the one out
|
|
* of 100k unhandled ones which is handled
|
|
* then we merily delay the spurious detection
|
|
* by one hard interrupt. Not a real problem.
|
|
*/
|
|
desc->threads_handled_last &= ~SPURIOUS_DEFERRED;
|
|
}
|
|
}
|
|
|
|
if (unlikely(action_ret == IRQ_NONE)) {
|
|
/*
|
|
* If we are seeing only the odd spurious IRQ caused by
|
|
* bus asynchronicity then don't eventually trigger an error,
|
|
* otherwise the counter becomes a doomsday timer for otherwise
|
|
* working systems
|
|
*/
|
|
if (time_after(jiffies, desc->last_unhandled + HZ/10))
|
|
desc->irqs_unhandled = 1;
|
|
else
|
|
desc->irqs_unhandled++;
|
|
desc->last_unhandled = jiffies;
|
|
}
|
|
|
|
irq = irq_desc_get_irq(desc);
|
|
if (unlikely(try_misrouted_irq(irq, desc, action_ret))) {
|
|
int ok = misrouted_irq(irq);
|
|
if (action_ret == IRQ_NONE)
|
|
desc->irqs_unhandled -= ok;
|
|
}
|
|
|
|
if (likely(!desc->irqs_unhandled))
|
|
return;
|
|
|
|
/* Now getting into unhandled irq detection */
|
|
desc->irq_count++;
|
|
if (likely(desc->irq_count < 100000))
|
|
return;
|
|
|
|
desc->irq_count = 0;
|
|
if (unlikely(desc->irqs_unhandled > 99900)) {
|
|
/*
|
|
* The interrupt is stuck
|
|
*/
|
|
__report_bad_irq(desc, action_ret);
|
|
/*
|
|
* Now kill the IRQ
|
|
*/
|
|
printk(KERN_EMERG "Disabling IRQ #%d\n", irq);
|
|
desc->istate |= IRQS_SPURIOUS_DISABLED;
|
|
desc->depth++;
|
|
irq_disable(desc);
|
|
|
|
mod_timer(&poll_spurious_irq_timer,
|
|
jiffies + POLL_SPURIOUS_IRQ_INTERVAL);
|
|
}
|
|
desc->irqs_unhandled = 0;
|
|
}
|
|
|
|
bool noirqdebug __read_mostly;
|
|
|
|
int noirqdebug_setup(char *str)
|
|
{
|
|
noirqdebug = 1;
|
|
printk(KERN_INFO "IRQ lockup detection disabled\n");
|
|
|
|
return 1;
|
|
}
|
|
|
|
__setup("noirqdebug", noirqdebug_setup);
|
|
module_param(noirqdebug, bool, 0644);
|
|
MODULE_PARM_DESC(noirqdebug, "Disable irq lockup detection when true");
|
|
|
|
static int __init irqfixup_setup(char *str)
|
|
{
|
|
irqfixup = 1;
|
|
printk(KERN_WARNING "Misrouted IRQ fixup support enabled.\n");
|
|
printk(KERN_WARNING "This may impact system performance.\n");
|
|
|
|
return 1;
|
|
}
|
|
|
|
__setup("irqfixup", irqfixup_setup);
|
|
module_param(irqfixup, int, 0644);
|
|
|
|
static int __init irqpoll_setup(char *str)
|
|
{
|
|
irqfixup = 2;
|
|
printk(KERN_WARNING "Misrouted IRQ fixup and polling support "
|
|
"enabled\n");
|
|
printk(KERN_WARNING "This may significantly impact system "
|
|
"performance\n");
|
|
return 1;
|
|
}
|
|
|
|
__setup("irqpoll", irqpoll_setup);
|