On 12/13/2013 08:55 AM, Daniel Lezcano wrote:

On 12/12/2013 06:36 PM, ivan.khoronzhuk wrote:

quoted

On 12/12/2013 05:51 PM, Daniel Lezcano wrote:

quoted

On 12/11/2013 07:00 PM, Ivan Khoronzhuk wrote:

quoted

Add broadcast clock-event device for the Keystone arch.

The timer can be configured as a general-purpose 64-bit timer,
dual general-purpose 32-bit timers. When configured as dual 32-bit
timers, each half can operate in conjunction (chain mode) or
independently (unchained mode) of each other.

Signed-off-by: Ivan Khoronzhuk <redacted>
---
   drivers/clocksource/Makefile         |    1 +
   drivers/clocksource/timer-keystone.c |  223
++++++++++++++++++++++++++++++++++
   2 files changed, 224 insertions(+)
   create mode 100644 drivers/clocksource/timer-keystone.c

diff --git a/drivers/clocksource/Makefile 

b/drivers/clocksource/Makefile
index 33621ef..2acf3fc 100644

--- a/drivers/clocksource/Makefile
+++ b/drivers/clocksource/Makefile

@@ -36,3 +36,4 @@ obj-$(CONFIG_ARM_ARCH_TIMER)        += 

arm_arch_timer.o
   obj-$(CONFIG_ARM_GLOBAL_TIMER)        += arm_global_timer.o
   obj-$(CONFIG_CLKSRC_METAG_GENERIC)    += metag_generic.o
   obj-$(CONFIG_ARCH_HAS_TICK_BROADCAST)    += dummy_timer.o
+obj-$(CONFIG_ARCH_KEYSTONE)        += timer-keystone.o

diff --git a/drivers/clocksource/timer-keystone.c

b/drivers/clocksource/timer-keystone.c
new file mode 100644
index 0000000..4a8083a

--- /dev/null
+++ b/drivers/clocksource/timer-keystone.c

@@ -0,0 +1,223 @@
+/*
+ * Keystone broadcast clock-event
+ *
+ * Copyright 2013 Texas Instruments, Inc.
+ *
+ * Author: Ivan Khoronzhuk <ivan.khoronzhuk@ti.com>
+ *
+ * This program is free software; you can redistribute it and/or 

modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ */
+
+#include <linux/clk.h>
+#include <linux/clockchips.h>
+#include <linux/clocksource.h>
+#include <linux/interrupt.h>
+#include <linux/of_address.h>
+#include <linux/of_irq.h>
+
+#define TIMER_NAME            "timer64-event"
+
+/* Timer register offsets */
+#define TIM12                0x10
+#define TIM34                0x14
+#define PRD12                0x18
+#define PRD34                0x1c
+#define TCR                0x20
+#define TGCR                0x24
+#define INTCTLSTAT            0x44
+
+/* Timer register bitfields */
+#define TCR_ENAMODE_MASK        0xC0
+#define TCR_ENAMODE_ONESHOT_MASK    0x40
+#define TCR_ENAMODE_PERIODIC_MASK    0x80
+
+#define TGCR_TIM_UNRESET_MASK        0x03
+#define INTCTLSTAT_ENINT_MASK        0x01
+
+/**
+ * struct keystone_timer: holds timer's data
+ * @base: timer memory base address
+ * @clk_rate: source clock rate
+ * @irqacrion: interrupt action descriptor
+ * @event_dev: event device based on timer
+ */
+static struct keystone_timer {
+    void __iomem *base;
+    unsigned long clk_rate;
+    struct irqaction irqaction;
+    struct clock_event_device event_dev;
+} timer;
+
+static inline u32 keystone_timer_readl(unsigned long rg)
+{
+    return readl(timer.base + rg);
+}
+
+static inline void keystone_timer_writel(u32 val, unsigned long rg)
+{
+    writel(val, timer.base + rg);
+}
+
+/**
+ * keystone_timer_config: configures timer to work in
oneshot/periodic modes, in
+ * other cases disables timer.
+ * @ mode: mode to configure
+ * @ period: cycles number to configure for
+ */
+static int keystone_timer_config(u64 period, enum clock_event_mode 
mode)
+{
+    u32 tcr;
+
+    tcr = keystone_timer_readl(TCR);
+
+    /* disable timer */
+    tcr &= ~(TCR_ENAMODE_MASK);
+    keystone_timer_writel(tcr, TCR);

Please move this write after the switch (cf. linked comment below)

See below

quoted

quoted

+    /* set enable mode or leave disabled timer */
+    switch (mode) {
+    case CLOCK_EVT_MODE_ONESHOT:
+        tcr |= TCR_ENAMODE_ONESHOT_MASK;
+        break;
+    case CLOCK_EVT_MODE_PERIODIC:
+        tcr |= TCR_ENAMODE_PERIODIC_MASK;
+        break;
+    default:
+        return 0;

Shouldn't you return an error ?

It returns value only for convenient usage in keystone_set_next_event(),
see [1].

In general, this function is used to configure the timer.
The procedure is that we need to stop the timer then set it.
If timer needs to be disabled (shutdown mode), function disables it 
and returns.

quoted

quoted

+    }
+
+    /* reset counter to zero, set new period */
+    keystone_timer_writel(0, TIM12);
+    keystone_timer_writel(0, TIM34);
+    keystone_timer_writel(period & 0xffffffff, PRD12);
+    keystone_timer_writel(period >> 32, PRD34);
+
+    /* enable timer */
+    keystone_timer_writel(tcr, TCR);
+    return 0;
+}
+
+static irqreturn_t keystone_timer_interrupt(int irq, void *dev_id)
+{
+    struct clock_event_device *evt = &timer.event_dev;
+
+    evt->event_handler(evt);
+    return IRQ_HANDLED;
+}
+
+static int keystone_set_next_event(unsigned long cycles,
+                  struct clock_event_device *evt)
+{
+    return keystone_timer_config(cycles, evt->mode);

[1]

quoted

quoted

+}
+
+static void keystone_set_mode(enum clock_event_mode mode,
+                 struct clock_event_device *evt)
+{
+    u64 period;
+
+    switch (mode) {
+    case CLOCK_EVT_MODE_PERIODIC:
+        period = timer.clk_rate / (HZ);
+        keystone_timer_config(period, CLOCK_EVT_MODE_PERIODIC);
+        break;
+    case CLOCK_EVT_MODE_UNUSED:
+    case CLOCK_EVT_MODE_SHUTDOWN:
+    case CLOCK_EVT_MODE_ONESHOT:
+        keystone_timer_config(0, CLOCK_EVT_MODE_SHUTDOWN);

I don't really this code assuming the side effect of specifying a
unsupported mode will disable the timer.

In case of an unsupported mode the function does nothing.
Is it not correct?

It disables the timer only in 3 cases.
CLOCK_EVT_MODE_UNUSED
CLOCK_EVT_MODE_SHUTDOWN
CLOCK_EVT_MODE_ONESHOT

For CLOCK_EVT_MODE_ONESHOT,
It is supposed that then keystone_set_next_even() will re-enable the 
timer.

The code looks like is doing what is expected. It is not about a bug but 
a coding style issue. I am not in favor of passing the 
CLOCK_EVT_MODE_SHUTDOWN which default in keystone_timer_config on 
returning no error. The caller assumes the function will exit without 
re-enabling the timer. It is correct from an implementation POV but IMO 
it is a bit subtle and prone to errors for the future modifications.

If keystone_timer_config(period, __A_VALUE_FROM_SPACE__) is called, it 
will return no error and disables the timer.

IMO, for the sake of clarity, defining the different modes in the switch 
and returning an error on default would be more convenient.

For this reason, you shouldn't disable the timer before checking the 
correctness of the mode, thus the comment about moving the timer 
disabling after the switch.

Ok,
I'll add separate function "keystone_timer_disable()" for using in keystone_set_mode().
The keystone_timer_config() will not be used for disabling the timer any more.
In case of an unsupported mode the keystone_timer_config() will return -1.


-- 
Regards,
Ivan Khoronzhuk