davinci: Add compare register support to timer code

The Timer64p timer has 8 compare registers that can
be used to generate interrupts when the timer value
matches the compare reg's value.  They do not disturb
the timer itself.  This can be useful when there is
only one timer available for both clock events and
clocksource.

When enabled, the clocksource remains a continuous
32-bit counter but the clock event will no longer
support periodic interrupts.  Instead only oneshot
timers will be supported and implemented by setting
the compare register to the current timer value plus
the period that the clock event subsystem is requesting.

Compare registers support is enabled automatically
when the following conditions are met:
1) The same timer is being used for clock events
   and clocksource.
2) The timer is the bottom half (32 bits) of the
   64-bit timer (hardware limitation).
3) The the compare register offset and irq are
   not zero.

Since the timer is always running, there is a hardware
race in timer32_config() between reading the current
timer value, and adding the period to the current
timer value and writing the compare register.
Testing on a da830 evm board with the timer clocked
at 24 MHz and the processor clocked at 300 MHz,
showed the number of counter ticks to do this ranged
from 20-53 (~1-2.2 usecs) but usually around 41 ticks.
This includes some artifacts from collecting the
information.  So, the minimum period should be
at least 5 usecs to be safe.

There is also an non-critical lower limit that
the period should be since there is no point in
setting an event that is much shorter than the
time it takes to set the event, and get & handle
the timer interrupt for that event.  There can
also be all sorts of delays from activities
occuring elsewhere in the system (including
hardware activitis like cache & TLB management).
These are virtually impossible to quantify so a
minimum period of 50 usecs was chosen.  That will
certianly be enough to avoid the actual hardware
race but hopefully not large enough to cause
unreasonably course-grained timers.
Signed-off-by: Mark A. Greer <mgreer@mvista.com>
Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>

arch/arm/mach-davinci/include/mach/common.h

@@ -30,6 +30,8 @@ struct davinci_timer_instance {
 	void __iomem	*base;
 	u32		bottom_irq;
 	u32		top_irq;
+	unsigned long	cmp_off;
+	unsigned int	cmp_irq;
 };
 
 struct davinci_timer_info {

arch/arm/mach-davinci/time.c

@@ -55,6 +55,16 @@ enum {
 #define TGCR			0x24
 #define WDTCR			0x28
 
+/* Offsets of the 8 compare registers */
+#define	CMP12_0			0x60
+#define	CMP12_1			0x64
+#define	CMP12_2			0x68
+#define	CMP12_3			0x6c
+#define	CMP12_4			0x70
+#define	CMP12_5			0x74
+#define	CMP12_6			0x78
+#define	CMP12_7			0x7c
+
 /* Timer register bitfields */
 #define TCR_ENAMODE_DISABLE          0x0
 #define TCR_ENAMODE_ONESHOT          0x1
@@ -85,6 +95,7 @@ struct timer_s {
 	unsigned int id;
 	unsigned long period;
 	unsigned long opts;
+	unsigned long flags;
 	void __iomem *base;
 	unsigned long tim_off;
 	unsigned long prd_off;
@@ -94,9 +105,13 @@ struct timer_s {
 static struct timer_s timers[];
 
 /* values for 'opts' field of struct timer_s */
-#define TIMER_OPTS_DISABLED   0x00
-#define TIMER_OPTS_ONESHOT    0x01
-#define TIMER_OPTS_PERIODIC   0x02
+#define TIMER_OPTS_DISABLED		0x01
+#define TIMER_OPTS_ONESHOT		0x02
+#define TIMER_OPTS_PERIODIC		0x04
+#define TIMER_OPTS_STATE_MASK		0x07
+
+#define TIMER_OPTS_USE_COMPARE		0x80000000
+#define USING_COMPARE(t)		((t)->opts & TIMER_OPTS_USE_COMPARE)
 
 static char *id_to_name[] = {
 	[T0_BOT]	= "timer0_0",
@@ -107,7 +122,24 @@ static char *id_to_name[] = {
 
 static int timer32_config(struct timer_s *t)
 {
-	u32 tcr = __raw_readl(t->base + TCR);
+	u32 tcr;
+	struct davinci_soc_info *soc_info = davinci_get_soc_info();
+
+	if (USING_COMPARE(t)) {
+		struct davinci_timer_instance *dtip =
+				soc_info->timer_info->timers;
+		int event_timer = ID_TO_TIMER(timers[TID_CLOCKEVENT].id);
+
+		/*
+		 * Next interrupt should be the current time reg value plus
+		 * the new period (using 32-bit unsigned addition/wrapping
+		 * to 0 on overflow).  This assumes that the clocksource
+		 * is setup to count to 2^32-1 before wrapping around to 0.
+		 */
+		__raw_writel(__raw_readl(t->base + t->tim_off) + t->period,
+			t->base + dtip[event_timer].cmp_off);
+	} else {
+		tcr = __raw_readl(t->base + TCR);
 
 		/* disable timer */
 		tcr &= ~(TCR_ENAMODE_MASK << t->enamode_shift);
@@ -118,13 +150,13 @@ static int timer32_config(struct timer_s *t)
 		__raw_writel(t->period, t->base + t->prd_off);
 
 		/* Set enable mode */
-	if (t->opts & TIMER_OPTS_ONESHOT) {
+		if (t->opts & TIMER_OPTS_ONESHOT)
 			tcr |= TCR_ENAMODE_ONESHOT << t->enamode_shift;
-	} else if (t->opts & TIMER_OPTS_PERIODIC) {
+		else if (t->opts & TIMER_OPTS_PERIODIC)
 			tcr |= TCR_ENAMODE_PERIODIC << t->enamode_shift;
-	}
 
 		__raw_writel(tcr, t->base + TCR);
+	}
 	return 0;
 }
 
@@ -222,8 +254,11 @@ static void __init timer_init(void)
 		/* Register interrupt */
 		t->irqaction.name = t->name;
 		t->irqaction.dev_id = (void *)t;
-		if (t->irqaction.handler != NULL)
+
+		if (t->irqaction.handler != NULL) {
+			irq = USING_COMPARE(t) ? dtip[i].cmp_irq : irq;
 			setup_irq(irq, &t->irqaction);
+		}
 
 		timer32_config(&timers[i]);
 	}
@@ -268,15 +303,18 @@ static void davinci_set_mode(enum clock_event_mode mode,
 	switch (mode) {
 	case CLOCK_EVT_MODE_PERIODIC:
 		t->period = davinci_clock_tick_rate / (HZ);
-		t->opts = TIMER_OPTS_PERIODIC;
+		t->opts &= ~TIMER_OPTS_STATE_MASK;
+		t->opts |= TIMER_OPTS_PERIODIC;
 		timer32_config(t);
 		break;
 	case CLOCK_EVT_MODE_ONESHOT:
-		t->opts = TIMER_OPTS_ONESHOT;
+		t->opts &= ~TIMER_OPTS_STATE_MASK;
+		t->opts |= TIMER_OPTS_ONESHOT;
 		break;
 	case CLOCK_EVT_MODE_UNUSED:
 	case CLOCK_EVT_MODE_SHUTDOWN:
-		t->opts = TIMER_OPTS_DISABLED;
+		t->opts &= ~TIMER_OPTS_STATE_MASK;
+		t->opts |= TIMER_OPTS_DISABLED;
 		break;
 	case CLOCK_EVT_MODE_RESUME:
 		break;
@@ -295,12 +333,40 @@ static void __init davinci_timer_init(void)
 {
 	struct clk *timer_clk;
 	struct davinci_soc_info *soc_info = &davinci_soc_info;
-
+	unsigned int clockevent_id;
+	unsigned int clocksource_id;
 	static char err[] __initdata = KERN_ERR
 		"%s: can't register clocksource!\n";
 
-	timers[TID_CLOCKEVENT].id = soc_info->timer_info->clockevent_id;
-	timers[TID_CLOCKSOURCE].id = soc_info->timer_info->clocksource_id;
+	clockevent_id = soc_info->timer_info->clockevent_id;
+	clocksource_id = soc_info->timer_info->clocksource_id;
+
+	timers[TID_CLOCKEVENT].id = clockevent_id;
+	timers[TID_CLOCKSOURCE].id = clocksource_id;
+
+	/*
+	 * If using same timer for both clock events & clocksource,
+	 * a compare register must be used to generate an event interrupt.
+	 * This is equivalent to a oneshot timer only (not periodic).
+	 */
+	if (clockevent_id == clocksource_id) {
+		struct davinci_timer_instance *dtip =
+				soc_info->timer_info->timers;
+		int event_timer = ID_TO_TIMER(clockevent_id);
+
+		/* Only bottom timers can use compare regs */
+		if (IS_TIMER_TOP(clockevent_id))
+			pr_warning("davinci_timer_init: Invalid use"
+				" of system timers.  Results unpredictable.\n");
+		else if ((dtip[event_timer].cmp_off == 0)
+				|| (dtip[event_timer].cmp_irq == 0))
+			pr_warning("davinci_timer_init:  Invalid timer instance"
+				" setup.  Results unpredictable.\n");
+		else {
+			timers[TID_CLOCKEVENT].opts |= TIMER_OPTS_USE_COMPARE;
+			clockevent_davinci.features = CLOCK_EVT_FEAT_ONESHOT;
+		}
+	}
 
 	/* init timer hw */
 	timer_init();
@@ -312,7 +378,7 @@ static void __init davinci_timer_init(void)
 	davinci_clock_tick_rate = clk_get_rate(timer_clk);
 
 	/* setup clocksource */
-	clocksource_davinci.name = id_to_name[timers[TID_CLOCKSOURCE].id];
+	clocksource_davinci.name = id_to_name[clocksource_id];
 	clocksource_davinci.mult =
 		clocksource_khz2mult(davinci_clock_tick_rate/1000,
 				     clocksource_davinci.shift);
@@ -325,8 +391,7 @@ static void __init davinci_timer_init(void)
 					 clockevent_davinci.shift);
 	clockevent_davinci.max_delta_ns =
 		clockevent_delta2ns(0xfffffffe, &clockevent_davinci);
-	clockevent_davinci.min_delta_ns =
-		clockevent_delta2ns(1, &clockevent_davinci);
+	clockevent_davinci.min_delta_ns = 50000; /* 50 usec */
 
 	clockevent_davinci.cpumask = cpumask_of(0);
 	clockevents_register_device(&clockevent_davinci);