[PATCH v3 2/2] i2c: designware: add two optional property tx/rx-fifo-size

On 13-06-19 12:22 AM, Wolfram Sang wrote:

On Tue, Jun 11, 2013 at 10:37:10AM +0800, Zhangfei Gao wrote:

quoted

rx-fifo-size and tx-fifo-size will be updated if provided from dts

Signed-off-by: Zhangfei Gao <zhangfei.gao@linaro.org>
CC: Baruch Siach <baruch@tkos.co.il>
---
  .../devicetree/bindings/i2c/i2c-designware.txt     |    7 +++++++
  drivers/i2c/busses/i2c-designware-platdrv.c        |    3 +++
  2 files changed, 10 insertions(+)

diff --git a/Documentation/devicetree/bindings/i2c/i2c-designware.txt b/Documentation/devicetree/bindings/i2c/i2c-designware.txt
index e42a2ee..84717fe 100644
--- a/Documentation/devicetree/bindings/i2c/i2c-designware.txt
+++ b/Documentation/devicetree/bindings/i2c/i2c-designware.txt

@@ -6,6 +6,11 @@ Required properties :
   - reg : Offset and length of the register set for the device
   - interrupts : <IRQ> where IRQ is the interrupt number.

+Optional properties:
+
+ - rx-fifo-size : 1 cell, Rx fifo size
+ - tx-fifo-size : 1 cell, Tx fifo size
+

I'd think the bindings are okay, yet I wonder they are not more used in
practice. So, an ack from a dt-expert would be appreciated.

Thanks Wolfram,
The patches also acked by Baruch, the original writer.

Some platform can not provide fifo info from register 
DW_IC_COMP_PARAM_1, hisilicon read such register is 0.

quoted

+	of_property_read_u32(np, "rx-fifo-size", &dev->rx_fifo_depth);
+	of_property_read_u32(np, "tx-fifo-size", &dev->tx_fifo_depth);

Those can fail.

"rx-fifo-size and tx-fifo-size will be updated if provided from dts"
The thought is if dts have the optional properties, then update.
otherwise just use original value, read from register.

So I thought the read fail is fine?