[PATCH 1/2] spi/rockchip: Round up clock rate divisor to err on the safe side

[PATCH 1/2] spi/rockchip: Round up clock rate divisor to err on the safe side

[Julius Werner]

The Rockchip SPI driver currently calculates its clock rate divisor by
integer dividing the parent rate by the target rate, and then rounding
the result up to the next even number (since the divisor must be
even).

Clock rate divisors should always be rounded up, so that the resulting
frequency is lower or equal to the target. This is correctly done in the
second step here but not in the first, so we still have a risk of
exceeding the desired target frequency (e.g. setting spi-max-frequency
to 40000000 with a parent clock of 99000000 could lead to a divisor of
99000000 / 40000000 == 2 (which is even) that then results in an
effective frequency of 99000000 / 2 == 49500000 (potentially exceeding
the flash chip's specifications).

This patch changes the division to round up to fix this problem.

Signed-off-by: Julius Werner <jwerner-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
---
 drivers/spi/spi-rockchip.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/spi/spi-rockchip.c b/drivers/spi/spi-rockchip.c
index 1a777dc..5e4e52c 100644
--- a/drivers/spi/spi-rockchip.c
+++ b/drivers/spi/spi-rockchip.c
@@ -519,7 +519,7 @@ static void rockchip_spi_config(struct rockchip_spi *rs)
 	}
 
 	/* div doesn't support odd number */
-	div = max_t(u32, rs->max_freq / rs->speed, 1);
+	div = DIV_ROUND_UP(rs->max_freq, rs->speed);
 	div = (div + 1) & 0xfffe;
 
 	writel_relaxed(cr0, rs->regs + ROCKCHIP_SPI_CTRLR0);
-- 
2.1.2

--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 2/2] spi/rockchip: Add device tree property to configure Rx Sample Delay

[Julius Werner]

We have found that we can sometimes see read failures on boards with
high-capacitance SPI lines. It seems that the controller samples the Rx
data line too early, and its register interface has an "Rx Sample Delay"
setting to fine-tune against this issue.

This patch adds a new optional device tree entry that can configure this
delay in terms of nanoseconds. The kernel will calculate the
best-fitting amount of parent clock ticks to program the controller with
based on that.

Signed-off-by: Julius Werner <jwerner-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
---
 .../devicetree/bindings/spi/spi-rockchip.txt        |  4 ++++
 drivers/spi/spi-rockchip.c                          | 21 +++++++++++++++++++++
 2 files changed, 25 insertions(+)

diff --git a/Documentation/devicetree/bindings/spi/spi-rockchip.txt b/Documentation/devicetree/bindings/spi/spi-rockchip.txt
index 467dec4..0c491bd 100644
--- a/Documentation/devicetree/bindings/spi/spi-rockchip.txt
+++ b/Documentation/devicetree/bindings/spi/spi-rockchip.txt
@@ -24,6 +24,9 @@ Optional Properties:
 - dmas: DMA specifiers for tx and rx dma. See the DMA client binding,
 		Documentation/devicetree/bindings/dma/dma.txt
 - dma-names: DMA request names should include "tx" and "rx" if present.
+- rx-sample-delay-ns: nanoseconds to delay after the SCLK edge before sampling
+		Rx data (may need to be fine tuned for high capacitance lines).
+		No delay (0) by default.
 
 
 Example:
@@ -33,6 +36,7 @@ Example:
 		reg = <0xff110000 0x1000>;
 		dmas = <&pdma1 11>, <&pdma1 12>;
 		dma-names = "tx", "rx";
+		rx-sample-delay-ns = <10>;
 		#address-cells = <1>;
 		#size-cells = <0>;
 		interrupts = <GIC_SPI 44 IRQ_TYPE_LEVEL_HIGH>;
diff --git a/drivers/spi/spi-rockchip.c b/drivers/spi/spi-rockchip.c
index 5e4e52c..f89cd5d 100644
--- a/drivers/spi/spi-rockchip.c
+++ b/drivers/spi/spi-rockchip.c
@@ -179,6 +179,7 @@ struct rockchip_spi {
 	u8 tmode;
 	u8 bpw;
 	u8 n_bytes;
+	u8 rsd_nsecs;
 	unsigned len;
 	u32 speed;
 
@@ -493,6 +494,7 @@ static void rockchip_spi_config(struct rockchip_spi *rs)
 {
 	u32 div = 0;
 	u32 dmacr = 0;
+	int rsd = 0;
 
 	u32 cr0 = (CR0_BHT_8BIT << CR0_BHT_OFFSET)
 		| (CR0_SSD_ONE << CR0_SSD_OFFSET);
@@ -522,6 +524,20 @@ static void rockchip_spi_config(struct rockchip_spi *rs)
 	div = DIV_ROUND_UP(rs->max_freq, rs->speed);
 	div = (div + 1) & 0xfffe;
 
+	/* Rx sample delay is expressed in parent clock cycles (max 3) */
+	rsd = DIV_ROUND_CLOSEST(rs->rsd_nsecs * (rs->max_freq >> 8),
+				1000000000 >> 8);
+	if (!rsd && rs->rsd_nsecs) {
+		pr_warn_once("rockchip-spi: %u Hz are too slow to express %u ns delay\n",
+			     rs->max_freq, rs->rsd_nsecs);
+	} else if (rsd > 3) {
+		rsd = 3;
+		pr_warn_once("rockchip-spi: %u Hz are too fast to express %u ns delay, clamping at %u ns\n",
+			     rs->max_freq, rs->rsd_nsecs,
+			     rsd * 1000000000U / rs->max_freq);
+	}
+	cr0 |= rsd << CR0_RSD_OFFSET;
+
 	writel_relaxed(cr0, rs->regs + ROCKCHIP_SPI_CTRLR0);
 
 	writel_relaxed(rs->len - 1, rs->regs + ROCKCHIP_SPI_CTRLR1);
@@ -614,6 +630,7 @@ static int rockchip_spi_probe(struct platform_device *pdev)
 	struct rockchip_spi *rs;
 	struct spi_master *master;
 	struct resource *mem;
+	u32 rsd_nsecs;
 
 	master = spi_alloc_master(&pdev->dev, sizeof(struct rockchip_spi));
 	if (!master)
@@ -665,6 +682,10 @@ static int rockchip_spi_probe(struct platform_device *pdev)
 	rs->dev = &pdev->dev;
 	rs->max_freq = clk_get_rate(rs->spiclk);
 
+	if (!of_property_read_u32(pdev->dev.of_node, "rx-sample-delay-ns",
+				  &rsd_nsecs))
+		rs->rsd_nsecs = rsd_nsecs;
+
 	rs->fifo_len = get_fifo_len(rs);
 	if (!rs->fifo_len) {
 		dev_err(&pdev->dev, "Failed to get fifo length\n");
-- 
2.1.2

--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 2/2] spi/rockchip: Add device tree property to configure Rx Sample Delay

[Doug Anderson]

Julius,

On Thu, Mar 26, 2015 at 4:30 PM, Julius Werner <jwerner-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org> wrote:
> We have found that we can sometimes see read failures on boards with
> high-capacitance SPI lines. It seems that the controller samples the Rx
> data line too early, and its register interface has an "Rx Sample Delay"
> setting to fine-tune against this issue.
>
> This patch adds a new optional device tree entry that can configure this
> delay in terms of nanoseconds. The kernel will calculate the
> best-fitting amount of parent clock ticks to program the controller with
> based on that.
>
> Signed-off-by: Julius Werner <jwerner-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
> ---
>  .../devicetree/bindings/spi/spi-rockchip.txt        |  4 ++++
>  drivers/spi/spi-rockchip.c                          | 21 +++++++++++++++++++++
>  2 files changed, 25 insertions(+)

Reviewed-by: Doug Anderson <dianders-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 1/2] spi/rockchip: Round up clock rate divisor to err on the safe side

[Doug Anderson]

Julius,

On Thu, Mar 26, 2015 at 4:30 PM, Julius Werner <jwerner-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org> wrote:
> The Rockchip SPI driver currently calculates its clock rate divisor by
> integer dividing the parent rate by the target rate, and then rounding
> the result up to the next even number (since the divisor must be
> even).
>
> Clock rate divisors should always be rounded up, so that the resulting
> frequency is lower or equal to the target. This is correctly done in the
> second step here but not in the first, so we still have a risk of
> exceeding the desired target frequency (e.g. setting spi-max-frequency
> to 40000000 with a parent clock of 99000000 could lead to a divisor of
> 99000000 / 40000000 == 2 (which is even) that then results in an
> effective frequency of 99000000 / 2 == 49500000 (potentially exceeding
> the flash chip's specifications).
>
> This patch changes the division to round up to fix this problem.
>
> Signed-off-by: Julius Werner <jwerner-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
> ---
>  drivers/spi/spi-rockchip.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)

Reviewed-by: Doug Anderson <dianders-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org>
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 1/2] spi/rockchip: Round up clock rate divisor to err on the safe side

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 308 bytes --]

On Thu, Mar 26, 2015 at 04:30:24PM -0700, Julius Werner wrote:
> The Rockchip SPI driver currently calculates its clock rate divisor by
> integer dividing the parent rate by the target rate, and then rounding
> the result up to the next even number (since the divisor must be
> even).

Applied both, thanks.

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 473 bytes --]