ARM: 7205/2: sched_clock: allow sched_clock to be selected at runtime

sched_clock() is yet another blocker on the road to the single
image. This patch implements an idea by Russell King:

http://www.spinics.net/lists/linux-omap/msg49561.html

Instead of asking the platform to implement both sched_clock()
itself and the rollover callback, simply register a read()
function, and let the ARM code care about sched_clock() itself,
the conversion to ns and the rollover. sched_clock() uses
this read() function as an indirection to the platform code.
If the platform doesn't provide a read(), the code falls back
to the jiffy counter (just like the default sched_clock).

This allow some simplifications and possibly some footprint gain
when multiple platforms are compiled in. Among the drawbacks,
the removal of the *_fixed_sched_clock optimization which could
negatively impact some platforms (sa1100, tegra, versatile
and omap).

Tested on 11MPCore, OMAP4 and Tegra.

Cc: Imre Kaloz <kaloz@openwrt.org>
Cc: Eric Miao <eric.y.miao@gmail.com>
Cc: Colin Cross <ccross@android.com>
Cc: Erik Gilling <konkers@android.com>
Cc: Olof Johansson <olof@lixom.net>
Cc: Sascha Hauer <kernel@pengutronix.de>
Cc: Alessandro Rubini <rubini@unipv.it>
Cc: STEricsson <STEricsson_nomadik_linux@list.st.com>
Cc: Lennert Buytenhek <kernel@wantstofly.org>
Cc: Ben Dooks <ben-linux@fluff.org>
Tested-by: Jamie Iles <jamie@jamieiles.com>
Tested-by: Tony Lindgren <tony@atomide.com>
Tested-by: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Acked-by: Nicolas Pitre <nico@linaro.org>
Acked-by: Krzysztof Halasa <khc@pm.waw.pl>
Acked-by: Kukjin Kim <kgene.kim@samsung.com>
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

arch/arm/kernel/sched_clock.c

@@ -14,61 +14,153 @@
 
 #include <asm/sched_clock.h>
 
+struct clock_data {
+	u64 epoch_ns;
+	u32 epoch_cyc;
+	u32 epoch_cyc_copy;
+	u32 mult;
+	u32 shift;
+};
+
 static void sched_clock_poll(unsigned long wrap_ticks);
 static DEFINE_TIMER(sched_clock_timer, sched_clock_poll, 0, 0);
-static void (*sched_clock_update_fn)(void);
+
+static struct clock_data cd = {
+	.mult	= NSEC_PER_SEC / HZ,
+};
+
+static u32 __read_mostly sched_clock_mask = 0xffffffff;
+
+static u32 notrace jiffy_sched_clock_read(void)
+{
+	return (u32)(jiffies - INITIAL_JIFFIES);
+}
+
+static u32 __read_mostly (*read_sched_clock)(void) = jiffy_sched_clock_read;
+
+static inline u64 cyc_to_ns(u64 cyc, u32 mult, u32 shift)
+{
+	return (cyc * mult) >> shift;
+}
+
+static unsigned long long cyc_to_sched_clock(u32 cyc, u32 mask)
+{
+	u64 epoch_ns;
+	u32 epoch_cyc;
+
+	/*
+	 * Load the epoch_cyc and epoch_ns atomically.  We do this by
+	 * ensuring that we always write epoch_cyc, epoch_ns and
+	 * epoch_cyc_copy in strict order, and read them in strict order.
+	 * If epoch_cyc and epoch_cyc_copy are not equal, then we're in
+	 * the middle of an update, and we should repeat the load.
+	 */
+	do {
+		epoch_cyc = cd.epoch_cyc;
+		smp_rmb();
+		epoch_ns = cd.epoch_ns;
+		smp_rmb();
+	} while (epoch_cyc != cd.epoch_cyc_copy);
+
+	return epoch_ns + cyc_to_ns((cyc - epoch_cyc) & mask, cd.mult, cd.shift);
+}
+
+/*
+ * Atomically update the sched_clock epoch.
+ */
+static void notrace update_sched_clock(void)
+{
+	unsigned long flags;
+	u32 cyc;
+	u64 ns;
+
+	cyc = read_sched_clock();
+	ns = cd.epoch_ns +
+		cyc_to_ns((cyc - cd.epoch_cyc) & sched_clock_mask,
+			  cd.mult, cd.shift);
+	/*
+	 * Write epoch_cyc and epoch_ns in a way that the update is
+	 * detectable in cyc_to_fixed_sched_clock().
+	 */
+	raw_local_irq_save(flags);
+	cd.epoch_cyc = cyc;
+	smp_wmb();
+	cd.epoch_ns = ns;
+	smp_wmb();
+	cd.epoch_cyc_copy = cyc;
+	raw_local_irq_restore(flags);
+}
 
 static void sched_clock_poll(unsigned long wrap_ticks)
 {
 	mod_timer(&sched_clock_timer, round_jiffies(jiffies + wrap_ticks));
-	sched_clock_update_fn();
+	update_sched_clock();
 }
 
-void __init init_sched_clock(struct clock_data *cd, void (*update)(void),
-	unsigned int clock_bits, unsigned long rate)
+void __init setup_sched_clock(u32 (*read)(void), int bits, unsigned long rate)
 {
 	unsigned long r, w;
 	u64 res, wrap;
 	char r_unit;
 
-	sched_clock_update_fn = update;
+	BUG_ON(bits > 32);
+	WARN_ON(!irqs_disabled());
+	WARN_ON(read_sched_clock != jiffy_sched_clock_read);
+	read_sched_clock = read;
+	sched_clock_mask = (1 << bits) - 1;
 
 	/* calculate the mult/shift to convert counter ticks to ns. */
-	clocks_calc_mult_shift(&cd->mult, &cd->shift, rate, NSEC_PER_SEC, 0);
+	clocks_calc_mult_shift(&cd.mult, &cd.shift, rate, NSEC_PER_SEC, 0);
 
 	r = rate;
 	if (r >= 4000000) {
 		r /= 1000000;
 		r_unit = 'M';
-	} else {
+	} else if (r >= 1000) {
 		r /= 1000;
 		r_unit = 'k';
-	}
+	} else
+		r_unit = ' ';
 
 	/* calculate how many ns until we wrap */
-	wrap = cyc_to_ns((1ULL << clock_bits) - 1, cd->mult, cd->shift);
+	wrap = cyc_to_ns((1ULL << bits) - 1, cd.mult, cd.shift);
 	do_div(wrap, NSEC_PER_MSEC);
 	w = wrap;
 
 	/* calculate the ns resolution of this counter */
-	res = cyc_to_ns(1ULL, cd->mult, cd->shift);
+	res = cyc_to_ns(1ULL, cd.mult, cd.shift);
 	pr_info("sched_clock: %u bits at %lu%cHz, resolution %lluns, wraps every %lums\n",
-		clock_bits, r, r_unit, res, w);
+		bits, r, r_unit, res, w);
 
 	/*
 	 * Start the timer to keep sched_clock() properly updated and
 	 * sets the initial epoch.
 	 */
 	sched_clock_timer.data = msecs_to_jiffies(w - (w / 10));
-	update();
+	update_sched_clock();
 
 	/*
 	 * Ensure that sched_clock() starts off at 0ns
 	 */
-	cd->epoch_ns = 0;
+	cd.epoch_ns = 0;
+
+	pr_debug("Registered %pF as sched_clock source\n", read);
+}
+
+unsigned long long notrace sched_clock(void)
+{
+	u32 cyc = read_sched_clock();
+	return cyc_to_sched_clock(cyc, sched_clock_mask);
 }
 
 void __init sched_clock_postinit(void)
 {
+	/*
+	 * If no sched_clock function has been provided at that point,
+	 * make it the final one one.
+	 */
+	if (read_sched_clock == jiffy_sched_clock_read)
+		setup_sched_clock(jiffy_sched_clock_read, 32, HZ);
+
 	sched_clock_poll(sched_clock_timer.data);
 }