[media] lmedm04: fix the range for relative measurements
authorMauro Carvalho Chehab <mchehab@osg.samsung.com>
Tue, 23 Jun 2015 15:29:46 +0000 (12:29 -0300)
committerMauro Carvalho Chehab <mchehab@osg.samsung.com>
Wed, 24 Jun 2015 11:38:30 +0000 (08:38 -0300)
Relative measurements are typically between 0 and 0xffff. However,
for some tuners (TUNER_S7395 and TUNER_S0194), the range were from
0 to 0xff00, with means that 100% is never archived.
Also, TUNER_RS2000 uses a more complex math.

So, create a macro that does the conversion using bit operations
and use it for all conversions.

The code is also easier to read with is a bonus.

While here, remove a bogus comment.

Signed-off-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
drivers/media/usb/dvb-usb-v2/lmedm04.c

index fcef2a3..4cc55b3 100644 (file)
@@ -257,6 +257,9 @@ static int lme2510_enable_pid(struct dvb_usb_device *d, u8 index, u16 pid_out)
        return ret;
 }
 
+/* Convert range from 0x00-0xff to 0x0000-0xffff */
+#define reg_to_16bits(x)       ((x) | ((x) << 8))
+
 static void lme2510_update_stats(struct dvb_usb_adapter *adap)
 {
        struct lme2510_state *st = adap_to_priv(adap);
@@ -288,23 +291,17 @@ static void lme2510_update_stats(struct dvb_usb_adapter *adap)
 
        switch (st->tuner_config) {
        case TUNER_LG:
-               s_tmp = 0xff - st->signal_level;
-               s_tmp |= s_tmp << 8;
-
-               c_tmp = 0xff - st->signal_sn;
-               c_tmp |= c_tmp << 8;
+               s_tmp = reg_to_16bits(0xff - st->signal_level);
+               c_tmp = reg_to_16bits(0xff - st->signal_sn);
                break;
-       /* fall through */
        case TUNER_S7395:
        case TUNER_S0194:
                s_tmp = 0xffff - (((st->signal_level * 2) << 8) * 5 / 4);
-
-               c_tmp = ((0xff - st->signal_sn - 0xa1) * 3) << 8;
+               c_tmp = reg_to_16bits((0xff - st->signal_sn - 0xa1) * 3);
                break;
        case TUNER_RS2000:
-               s_tmp = st->signal_level * 0xffff / 0xff;
-
-               c_tmp = st->signal_sn * 0xffff / 0x7f;
+               s_tmp = reg_to_16bits(st->signal_level);
+               c_tmp = reg_to_16bits(st->signal_sn);
        }
 
        c->strength.len = 1;