1fa405552e
Calculate the MIPS clockevent device's min_delta_ns dynamically based on the time it takes to perform the mips_next_event() sequence. Virtualisation in particular makes the current fixed min_delta of 0x300 inappropriate under some circumstances, as the CP0_Count and CP0_Compare registers may be being emulated by the hypervisor, and the frequency may not correspond directly to the CPU frequency. We actually use twice the median of multiple 75th percentiles of multiple measurements of how long the mips_next_event() sequence takes, in order to fairly efficiently eliminate outliers due to unexpected hypervisor latency (which would need handling with retries when it occurs during normal operation anyway). Signed-off-by: James Hogan <james.hogan@imgtec.com> Cc: Daniel Lezcano <daniel.lezcano@linaro.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/13176/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
304 lines
7.7 KiB
C
304 lines
7.7 KiB
C
/*
|
|
* This file is subject to the terms and conditions of the GNU General Public
|
|
* License. See the file "COPYING" in the main directory of this archive
|
|
* for more details.
|
|
*
|
|
* Copyright (C) 2007 MIPS Technologies, Inc.
|
|
* Copyright (C) 2007 Ralf Baechle <ralf@linux-mips.org>
|
|
*/
|
|
#include <linux/clockchips.h>
|
|
#include <linux/interrupt.h>
|
|
#include <linux/percpu.h>
|
|
#include <linux/smp.h>
|
|
#include <linux/irq.h>
|
|
|
|
#include <asm/time.h>
|
|
#include <asm/cevt-r4k.h>
|
|
|
|
static int mips_next_event(unsigned long delta,
|
|
struct clock_event_device *evt)
|
|
{
|
|
unsigned int cnt;
|
|
int res;
|
|
|
|
cnt = read_c0_count();
|
|
cnt += delta;
|
|
write_c0_compare(cnt);
|
|
res = ((int)(read_c0_count() - cnt) >= 0) ? -ETIME : 0;
|
|
return res;
|
|
}
|
|
|
|
/**
|
|
* calculate_min_delta() - Calculate a good minimum delta for mips_next_event().
|
|
*
|
|
* Running under virtualisation can introduce overhead into mips_next_event() in
|
|
* the form of hypervisor emulation of CP0_Count/CP0_Compare registers,
|
|
* potentially with an unnatural frequency, which makes a fixed min_delta_ns
|
|
* value inappropriate as it may be too small.
|
|
*
|
|
* It can also introduce occasional latency from the guest being descheduled.
|
|
*
|
|
* This function calculates a good minimum delta based roughly on the 75th
|
|
* percentile of the time taken to do the mips_next_event() sequence, in order
|
|
* to handle potentially higher overhead while also eliminating outliers due to
|
|
* unpredictable hypervisor latency (which can be handled by retries).
|
|
*
|
|
* Return: An appropriate minimum delta for the clock event device.
|
|
*/
|
|
static unsigned int calculate_min_delta(void)
|
|
{
|
|
unsigned int cnt, i, j, k, l;
|
|
unsigned int buf1[4], buf2[3];
|
|
unsigned int min_delta;
|
|
|
|
/*
|
|
* Calculate the median of 5 75th percentiles of 5 samples of how long
|
|
* it takes to set CP0_Compare = CP0_Count + delta.
|
|
*/
|
|
for (i = 0; i < 5; ++i) {
|
|
for (j = 0; j < 5; ++j) {
|
|
/*
|
|
* This is like the code in mips_next_event(), and
|
|
* directly measures the borderline "safe" delta.
|
|
*/
|
|
cnt = read_c0_count();
|
|
write_c0_compare(cnt);
|
|
cnt = read_c0_count() - cnt;
|
|
|
|
/* Sorted insert into buf1 */
|
|
for (k = 0; k < j; ++k) {
|
|
if (cnt < buf1[k]) {
|
|
l = min_t(unsigned int,
|
|
j, ARRAY_SIZE(buf1) - 1);
|
|
for (; l > k; --l)
|
|
buf1[l] = buf1[l - 1];
|
|
break;
|
|
}
|
|
}
|
|
if (k < ARRAY_SIZE(buf1))
|
|
buf1[k] = cnt;
|
|
}
|
|
|
|
/* Sorted insert of 75th percentile into buf2 */
|
|
for (k = 0; k < i; ++k) {
|
|
if (buf1[ARRAY_SIZE(buf1) - 1] < buf2[k]) {
|
|
l = min_t(unsigned int,
|
|
i, ARRAY_SIZE(buf2) - 1);
|
|
for (; l > k; --l)
|
|
buf2[l] = buf2[l - 1];
|
|
break;
|
|
}
|
|
}
|
|
if (k < ARRAY_SIZE(buf2))
|
|
buf2[k] = buf1[ARRAY_SIZE(buf1) - 1];
|
|
}
|
|
|
|
/* Use 2 * median of 75th percentiles */
|
|
min_delta = buf2[ARRAY_SIZE(buf2) - 1] * 2;
|
|
|
|
/* Don't go too low */
|
|
if (min_delta < 0x300)
|
|
min_delta = 0x300;
|
|
|
|
pr_debug("%s: median 75th percentile=%#x, min_delta=%#x\n",
|
|
__func__, buf2[ARRAY_SIZE(buf2) - 1], min_delta);
|
|
return min_delta;
|
|
}
|
|
|
|
DEFINE_PER_CPU(struct clock_event_device, mips_clockevent_device);
|
|
int cp0_timer_irq_installed;
|
|
|
|
/*
|
|
* Possibly handle a performance counter interrupt.
|
|
* Return true if the timer interrupt should not be checked
|
|
*/
|
|
static inline int handle_perf_irq(int r2)
|
|
{
|
|
/*
|
|
* The performance counter overflow interrupt may be shared with the
|
|
* timer interrupt (cp0_perfcount_irq < 0). If it is and a
|
|
* performance counter has overflowed (perf_irq() == IRQ_HANDLED)
|
|
* and we can't reliably determine if a counter interrupt has also
|
|
* happened (!r2) then don't check for a timer interrupt.
|
|
*/
|
|
return (cp0_perfcount_irq < 0) &&
|
|
perf_irq() == IRQ_HANDLED &&
|
|
!r2;
|
|
}
|
|
|
|
irqreturn_t c0_compare_interrupt(int irq, void *dev_id)
|
|
{
|
|
const int r2 = cpu_has_mips_r2_r6;
|
|
struct clock_event_device *cd;
|
|
int cpu = smp_processor_id();
|
|
|
|
/*
|
|
* Suckage alert:
|
|
* Before R2 of the architecture there was no way to see if a
|
|
* performance counter interrupt was pending, so we have to run
|
|
* the performance counter interrupt handler anyway.
|
|
*/
|
|
if (handle_perf_irq(r2))
|
|
return IRQ_HANDLED;
|
|
|
|
/*
|
|
* The same applies to performance counter interrupts. But with the
|
|
* above we now know that the reason we got here must be a timer
|
|
* interrupt. Being the paranoiacs we are we check anyway.
|
|
*/
|
|
if (!r2 || (read_c0_cause() & CAUSEF_TI)) {
|
|
/* Clear Count/Compare Interrupt */
|
|
write_c0_compare(read_c0_compare());
|
|
cd = &per_cpu(mips_clockevent_device, cpu);
|
|
cd->event_handler(cd);
|
|
|
|
return IRQ_HANDLED;
|
|
}
|
|
|
|
return IRQ_NONE;
|
|
}
|
|
|
|
struct irqaction c0_compare_irqaction = {
|
|
.handler = c0_compare_interrupt,
|
|
/*
|
|
* IRQF_SHARED: The timer interrupt may be shared with other interrupts
|
|
* such as perf counter and FDC interrupts.
|
|
*/
|
|
.flags = IRQF_PERCPU | IRQF_TIMER | IRQF_SHARED,
|
|
.name = "timer",
|
|
};
|
|
|
|
|
|
void mips_event_handler(struct clock_event_device *dev)
|
|
{
|
|
}
|
|
|
|
/*
|
|
* FIXME: This doesn't hold for the relocated E9000 compare interrupt.
|
|
*/
|
|
static int c0_compare_int_pending(void)
|
|
{
|
|
/* When cpu_has_mips_r2, this checks Cause.TI instead of Cause.IP7 */
|
|
return (read_c0_cause() >> cp0_compare_irq_shift) & (1ul << CAUSEB_IP);
|
|
}
|
|
|
|
/*
|
|
* Compare interrupt can be routed and latched outside the core,
|
|
* so wait up to worst case number of cycle counter ticks for timer interrupt
|
|
* changes to propagate to the cause register.
|
|
*/
|
|
#define COMPARE_INT_SEEN_TICKS 50
|
|
|
|
int c0_compare_int_usable(void)
|
|
{
|
|
unsigned int delta;
|
|
unsigned int cnt;
|
|
|
|
#ifdef CONFIG_KVM_GUEST
|
|
return 1;
|
|
#endif
|
|
|
|
/*
|
|
* IP7 already pending? Try to clear it by acking the timer.
|
|
*/
|
|
if (c0_compare_int_pending()) {
|
|
cnt = read_c0_count();
|
|
write_c0_compare(cnt);
|
|
back_to_back_c0_hazard();
|
|
while (read_c0_count() < (cnt + COMPARE_INT_SEEN_TICKS))
|
|
if (!c0_compare_int_pending())
|
|
break;
|
|
if (c0_compare_int_pending())
|
|
return 0;
|
|
}
|
|
|
|
for (delta = 0x10; delta <= 0x400000; delta <<= 1) {
|
|
cnt = read_c0_count();
|
|
cnt += delta;
|
|
write_c0_compare(cnt);
|
|
back_to_back_c0_hazard();
|
|
if ((int)(read_c0_count() - cnt) < 0)
|
|
break;
|
|
/* increase delta if the timer was already expired */
|
|
}
|
|
|
|
while ((int)(read_c0_count() - cnt) <= 0)
|
|
; /* Wait for expiry */
|
|
|
|
while (read_c0_count() < (cnt + COMPARE_INT_SEEN_TICKS))
|
|
if (c0_compare_int_pending())
|
|
break;
|
|
if (!c0_compare_int_pending())
|
|
return 0;
|
|
cnt = read_c0_count();
|
|
write_c0_compare(cnt);
|
|
back_to_back_c0_hazard();
|
|
while (read_c0_count() < (cnt + COMPARE_INT_SEEN_TICKS))
|
|
if (!c0_compare_int_pending())
|
|
break;
|
|
if (c0_compare_int_pending())
|
|
return 0;
|
|
|
|
/*
|
|
* Feels like a real count / compare timer.
|
|
*/
|
|
return 1;
|
|
}
|
|
|
|
unsigned int __weak get_c0_compare_int(void)
|
|
{
|
|
return MIPS_CPU_IRQ_BASE + cp0_compare_irq;
|
|
}
|
|
|
|
int r4k_clockevent_init(void)
|
|
{
|
|
unsigned int cpu = smp_processor_id();
|
|
struct clock_event_device *cd;
|
|
unsigned int irq, min_delta;
|
|
|
|
if (!cpu_has_counter || !mips_hpt_frequency)
|
|
return -ENXIO;
|
|
|
|
if (!c0_compare_int_usable())
|
|
return -ENXIO;
|
|
|
|
/*
|
|
* With vectored interrupts things are getting platform specific.
|
|
* get_c0_compare_int is a hook to allow a platform to return the
|
|
* interrupt number of its liking.
|
|
*/
|
|
irq = get_c0_compare_int();
|
|
|
|
cd = &per_cpu(mips_clockevent_device, cpu);
|
|
|
|
cd->name = "MIPS";
|
|
cd->features = CLOCK_EVT_FEAT_ONESHOT |
|
|
CLOCK_EVT_FEAT_C3STOP |
|
|
CLOCK_EVT_FEAT_PERCPU;
|
|
|
|
clockevent_set_clock(cd, mips_hpt_frequency);
|
|
|
|
/* Calculate the min / max delta */
|
|
cd->max_delta_ns = clockevent_delta2ns(0x7fffffff, cd);
|
|
min_delta = calculate_min_delta();
|
|
cd->min_delta_ns = clockevent_delta2ns(min_delta, cd);
|
|
|
|
cd->rating = 300;
|
|
cd->irq = irq;
|
|
cd->cpumask = cpumask_of(cpu);
|
|
cd->set_next_event = mips_next_event;
|
|
cd->event_handler = mips_event_handler;
|
|
|
|
clockevents_register_device(cd);
|
|
|
|
if (cp0_timer_irq_installed)
|
|
return 0;
|
|
|
|
cp0_timer_irq_installed = 1;
|
|
|
|
setup_irq(irq, &c0_compare_irqaction);
|
|
|
|
return 0;
|
|
}
|
|
|