Re: [PATCH 1/2] Input: touchscreen-iproc: Add Broadcom iProc touchscreen driver
flat view
From: Joe Perches <joe@perches.com>
Date: 2014-12-18 02:14:49
Also in:
linux-arm-kernel, linux-devicetree, lkml
On Wed, 2014-12-17 at 17:59 -0800, Jonathan Richardson wrote:
Add initial version of the Broadcom touchscreen driver.
trivia:
quoted hunk ↗ jump to hunk
diff --git a/drivers/input/touchscreen/bcm_iproc_tsc.c b/drivers/input/touchscreen/bcm_iproc_tsc.c
quoted hunk ↗ jump to hunk
+/* Bit values for REGCTL2 */ +#define TS_CONTROLLER_EN_BIT (1 << 16) +#define TS_CONTROLLER_AVGDATA_SHIFT 8 +#define TS_CONTROLLER_AVGDATA_MASK (0x7 << TS_CONTROLLER_AVGDATA_SHIFT) +#define TS_CONTROLLER_PWR_LDO (1<<5) +#define TS_CONTROLLER_PWR_ADC (1<<4) +#define TS_CONTROLLER_PWR_BGP (1<<3) +#define TS_CONTROLLER_PWR_TS (1<<2) +#define TS_WIRE_MODE_BIT (1<<1)
Be nicer to use the same spacing around << or maybe use the BIT macro. []
quoted hunk ↗ jump to hunk
+static int get_tsc_config(struct device_node *np, struct iproc_ts_priv *priv) +{ + int ret; + u32 val; + struct device *dev = &priv->pdev->dev; + + priv->cfg_params = default_config; + + ret = of_property_read_u32(np, "scanning_period", &val); + if (ret >= 0) { + if ((1 <= val) && (val <= 256)) + priv->cfg_params.scanning_period = val; + else { + dev_err(dev, "scanning_period must be [1-256]"); + return -EINVAL; + }
ret is never used so I'd probably remove it
from all these blocks.
It's probably be nicer to invert the logic ald
remove the else.
There's a missing terminating newline too.
Something like:
if (of_property_read_u32(np, "scanning_period", &val) >= 0) {
if (val < 1 || val > 256) {
dev_err(dev, "scanning_period must be [1-256]\n");
return -EINVAL;
}
priv->cfg_params.scanning_period = val;
}
etc...