[PATCHv2 0/4] Clocksource: Flextimer: add LS1 support

[PATCHv2 0/4] Clocksource: Flextimer: add LS1 support

[Xiubo Li]

Change in v2:
- Follows Thomas Gleixner advice.


Xiubo Li (4):
  Clocksource: Flextimer: Use internal clocksource read API.
  Clocksource: Flextimer: Remove the useless code.
  Clocksource: Flextimer: Set cpumask to cpu_possible_mask
  Clocksource: Flextimer: Fix counter clock prescaler calculation.

 drivers/clocksource/fsl_ftm_timer.c | 20 +++++++++++---------
 1 file changed, 11 insertions(+), 9 deletions(-)

-- 
2.1.0.27.g96db324

[PATCHv2 1/4] Clocksource: Flextimer: Use internal clocksource read API.

[Xiubo Li]

Since the Flextimer device will be implemented in BE mode on
LS1 SoC, and in LE mode on Vybrid, LS2 SoCs, so here we need
the endianness judgment before doing the mmio.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
---
 drivers/clocksource/fsl_ftm_timer.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/drivers/clocksource/fsl_ftm_timer.c b/drivers/clocksource/fsl_ftm_timer.c
index 454227d..9c6e935 100644
--- a/drivers/clocksource/fsl_ftm_timer.c
+++ b/drivers/clocksource/fsl_ftm_timer.c
@@ -226,6 +226,11 @@ static int __init ftm_clockevent_init(unsigned long freq, int irq)
 	return 0;
 }
 
+static cycle_t ftm_clocksource_read_up(struct clocksource *c)
+{
+	return ftm_readl(priv->clksrc_base + FTM_CNT) & 0xffff;
+}
+
 static int __init ftm_clocksource_init(unsigned long freq)
 {
 	int err;
@@ -238,7 +243,7 @@ static int __init ftm_clocksource_init(unsigned long freq)
 	sched_clock_register(ftm_read_sched_clock, 16, freq / (1 << priv->ps));
 	err = clocksource_mmio_init(priv->clksrc_base + FTM_CNT, "fsl-ftm",
 				    freq / (1 << priv->ps), 300, 16,
-				    clocksource_mmio_readl_up);
+				    ftm_clocksource_read_up);
 	if (err) {
 		pr_err("ftm: init clock source mmio failed: %d\n", err);
 		return err;
-- 
2.1.0.27.g96db324

Re: [PATCHv2 1/4] Clocksource: Flextimer: Use internal clocksource read API.

[Thomas Gleixner]

On Fri, 5 Sep 2014, Xiubo Li wrote:

> Since the Flextimer device will be implemented in BE mode on
> LS1 SoC, and in LE mode on Vybrid, LS2 SoCs, so here we need
> the endianness judgment before doing the mmio.

WTF? I explained you in great length HOW you should solve that, but
you just ignore it. Instead of trying to understand what I say you
just resend the same nonsense again.

Keep that attitude up, if you want to make sure, that your patches are
ignored in the future.

Thanks,

	tglx

RE: [PATCHv2 1/4] Clocksource: Flextimer: Use internal clocksource read API.

[Li.Xiubo]

Hi Thomas,

Okay.

I will try to fix this.

Thanks,

BRs
Xiubo




> -----Original Message-----
> From: Thomas Gleixner [mailto:tglx@linutronix.de]
> Sent: Friday, September 05, 2014 11:40 PM
> To: Xiubo Li-B47053
> Cc: daniel.lezcano@linaro.org; linux-kernel@vger.kernel.org
> Subject: Re: [PATCHv2 1/4] Clocksource: Flextimer: Use internal clocksource
> read API.
> 
> On Fri, 5 Sep 2014, Xiubo Li wrote:
> 
> > Since the Flextimer device will be implemented in BE mode on
> > LS1 SoC, and in LE mode on Vybrid, LS2 SoCs, so here we need
> > the endianness judgment before doing the mmio.
> 
> WTF? I explained you in great length HOW you should solve that, but
> you just ignore it. Instead of trying to understand what I say you
> just resend the same nonsense again.
> 
> Keep that attitude up, if you want to make sure, that your patches are
> ignored in the future.
> 
> Thanks,
> 
> 	tglx

[PATCHv2 2/4] Clocksource: Flextimer: Remove the useless code.

[Xiubo Li]

The clock envnt counter will be enabled in proper time and proper
place when needed.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
---
 drivers/clocksource/fsl_ftm_timer.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/drivers/clocksource/fsl_ftm_timer.c b/drivers/clocksource/fsl_ftm_timer.c
index 9c6e935..ccb8c32 100644
--- a/drivers/clocksource/fsl_ftm_timer.c
+++ b/drivers/clocksource/fsl_ftm_timer.c
@@ -221,8 +221,6 @@ static int __init ftm_clockevent_init(unsigned long freq, int irq)
 					freq / (1 << priv->ps),
 					1, 0xffff);
 
-	ftm_counter_enable(priv->clkevt_base);
-
 	return 0;
 }
 
-- 
2.1.0.27.g96db324

[PATCHv2 3/4] Clocksource: Flextimer: Set cpumask to cpu_possible_mask

[Xiubo Li]

The Flextimer is not tied to CPU0, make it usable on any CPU.
For Vybrid there is only one CPU, while for LS1+ there are more
than one.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
---
 drivers/clocksource/fsl_ftm_timer.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/clocksource/fsl_ftm_timer.c b/drivers/clocksource/fsl_ftm_timer.c
index ccb8c32..c0ce603 100644
--- a/drivers/clocksource/fsl_ftm_timer.c
+++ b/drivers/clocksource/fsl_ftm_timer.c
@@ -214,7 +214,7 @@ static int __init ftm_clockevent_init(unsigned long freq, int irq)
 		return err;
 	}
 
-	ftm_clockevent.cpumask = cpumask_of(0);
+	ftm_clockevent.cpumask = cpu_all_mask;
 	ftm_clockevent.irq = irq;
 
 	clockevents_config_and_register(&ftm_clockevent,
-- 
2.1.0.27.g96db324

[PATCHv2 4/4] Clocksource: Flextimer: Fix counter clock prescaler calculation.

[Xiubo Li]

The old code will use the uncorrect clock prescaler after the
loop and the 'ps' must minus one after the loop.

This patch will use new style of this loop and will fix this bug
at the same time.

Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
---
 drivers/clocksource/fsl_ftm_timer.c | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/drivers/clocksource/fsl_ftm_timer.c b/drivers/clocksource/fsl_ftm_timer.c
index c0ce603..2b0983d 100644
--- a/drivers/clocksource/fsl_ftm_timer.c
+++ b/drivers/clocksource/fsl_ftm_timer.c
@@ -300,16 +300,15 @@ static unsigned long __init ftm_clk_init(struct device_node *np)
 
 static int __init ftm_calc_closest_round_cyc(unsigned long freq)
 {
-	priv->ps = 0;
+	int div;
 
 	/* The counter register is only using the lower 16 bits, and
 	 * if the 'freq' value is to big here, then the periodic_cyc
 	 * may exceed 0xFFFF.
 	 */
-	do {
-		priv->periodic_cyc = DIV_ROUND_CLOSEST(freq,
-						HZ * (1 << priv->ps++));
-	} while (priv->periodic_cyc > 0xFFFF);
+	for (priv->ps = 0, priv->periodic_cyc = ~0UL, div = HZ;
+	     priv->periodic_cyc > 0xffff; priv->ps++, div *= 2)
+		priv->periodic_cyc = DIV_ROUND_CLOSEST(freq, div);
 
 	if (priv->ps > FTM_PS_MAX) {
 		pr_err("ftm: the prescaler is %lu > %d\n",
-- 
2.1.0.27.g96db324