[PATCH 4/4] Modifications to the driver mb86a20s‏‏

[PATCH 4/4] Modifications to the driver mb86a20s‏‏

[Manoel PN]

[-- Attachment #1: Type: text/plain, Size: 266 bytes --]



This patch implement changes to the function mb86a20s_read_signal_strength.

The original function, binary search, does not work with device dtb08.

I would like to know if this function works.


Signed-off-by: Manoel Pinheiro <pinusdtv@hotmail.com>


 		 	   		  

[-- Attachment #2: signal_strength.patch --]
[-- Type: application/octet-stream, Size: 1210 bytes --]

diff --git a/drivers/media/dvb/frontends/mb86a20s.c b/drivers/media/dvb/frontends/mb86a20s.c
index 0f867a5..8f39da5 100644
--- a/drivers/media/dvb/frontends/mb86a20s.c
+++ b/drivers/media/dvb/frontends/mb86a20s.c
@@ -414,6 +414,12 @@ err:
 static int mb86a20s_read_signal_strength(struct dvb_frontend *fe, u16 *strength)
 {
 	struct mb86a20s_state *state = fe->demodulator_priv;
+	int i, val, val2;
+
+#if 0
+	/*
+	 * Binary search don't work with DTB08
+	 */
 	unsigned rf_max, rf_min, rf;
 	u8	 val;
 
@@ -447,6 +453,30 @@ static int mb86a20s_read_signal_strength(struct dvb_frontend *fe, u16 *strength)
 
 	if (fe->ops.i2c_gate_ctrl)
 		fe->ops.i2c_gate_ctrl(fe, 1);
+#endif
+
+	dprintk("\n");
+
+	*strength = 0;
+
+	for (i = 0; i < 10; i++)
+	{
+		val = mb86a20s_readreg(state, 0x0a);
+		if (val >= 2) {
+			mb86a20s_writereg(state, 0x04, 0x25);
+			val2 = mb86a20s_readreg(state, 0x05);
+			mb86a20s_writereg(state, 0x04, 0x26);
+			val = mb86a20s_readreg(state, 0x05);
+			if (val >= 0 && val2 >= 0) {
+				val = (val2 << 8) | val;
+				val2 = (val * 0x100100) >> 16;
+				dprintk("signal strength = %i\n", val2);
+				*strength = (u16)val2;
+				return 0;
+			}
+		}
+		msleep(10);
+	}
 
 	return 0;
 }

Re: [PATCH 4/4] Modifications to the driver mb86a20s‏‏

[Mauro Carvalho Chehab]

Em 13-05-2011 04:13, Manoel PN escreveu:
> 
> 
> This patch implement changes to the function mb86a20s_read_signal_strength.
> 
> The original function, binary search, does not work with device dtb08.

Hmm... that's interesting. Maybe due to a different mb86a20s chip stepping?

> 
> I would like to know if this function works.

I'll test it when I return back to Brazil.

Please send this patch in separate, as a RFC, and keep both methods enabled,
printing debug messages. This way, I can test it with the two devices I have,
comparing the SNR relations provided by both ways.

Thanks,
Mauro.