On 10/02/2014 08:21 AM, Matthias Schwarzott wrote:
This is used later for is_tuner function that switches i2c behaviour for
some tuners.
V2: Add comments about possible improvements for port_3 switch function.
Signed-off-by: Matthias Schwarzott <zzam@xxxxxxxxxx>
---
drivers/media/usb/cx231xx/cx231xx-avcore.c | 10 ++++++++++
drivers/media/usb/cx231xx/cx231xx.h | 1 +
2 files changed, 11 insertions(+)
diff --git a/drivers/media/usb/cx231xx/cx231xx-avcore.c b/drivers/media/usb/cx231xx/cx231xx-avcore.c
index 40a6987..148b5fa 100644
--- a/drivers/media/usb/cx231xx/cx231xx-avcore.c
+++ b/drivers/media/usb/cx231xx/cx231xx-avcore.c
@@ -1272,6 +1272,12 @@ int cx231xx_enable_i2c_port_3(struct cx231xx *dev, bool is_port_3)
if (dev->board.dont_use_port_3)
is_port_3 = false;
+
+ /* Should this code check dev->port_3_switch_enabled first */
+ /* to skip unnecessary reading of the register? */
+ /* If yes, the flag dev->port_3_switch_enabled must be initialized */
+ /* correctly. */
That multiline comment is violates Linux CodingStyle. See
Documentation/CodingStyle
+
status = cx231xx_read_ctrl_reg(dev, VRT_GET_REGISTER,
PWR_CTL_EN, value, 4);
if (status < 0)
@@ -1294,6 +1300,10 @@ int cx231xx_enable_i2c_port_3(struct cx231xx *dev, bool is_port_3)
status = cx231xx_write_ctrl_reg(dev, VRT_SET_REGISTER,
PWR_CTL_EN, value, 4);
+ /* remember status of the switch for usage in is_tuner */
+ if (status >= 0)
+ dev->port_3_switch_enabled = is_port_3;
+
return status;
}
diff --git a/drivers/media/usb/cx231xx/cx231xx.h b/drivers/media/usb/cx231xx/cx231xx.h
index f03338b..8a3c97b 100644
--- a/drivers/media/usb/cx231xx/cx231xx.h
+++ b/drivers/media/usb/cx231xx/cx231xx.h
@@ -629,6 +629,7 @@ struct cx231xx {
/* I2C adapters: Master 1 & 2 (External) & Master 3 (Internal only) */
struct cx231xx_i2c i2c_bus[3];
unsigned int xc_fw_load_done:1;
+ unsigned int port_3_switch_enabled:1;
/* locks */
struct mutex gpio_i2c_lock;
struct mutex i2c_lock;
I trust functionality is correct. However, I expected to see mux
switching logic inside mux select() / deselect(), maybe with caching mux
switch position in order to avoid I/O needed for checking mux switch
position.
But as I don't know that driver internals very well, I am not adding
reviewed by tag, which does not mean that is wrong. It is simply because
I don't know.
regards
Antti
--
http://palosaari.fi/
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
