git.ipfire.org Git - thirdparty/kernel/stable-queue.git/blob

   1 From a57c1dcb93e43357ed3f666e5a2b5d5071dd3930 Mon Sep 17 00:00:00 2001
   2 From: Seth Barry <seth@cyberseth.com>
   3 Date: Sun, 27 Sep 2009 16:42:29 -0300
   4 Subject: V4L/DVB (13109): tda18271: fix signedness issue in tda18271_rf_tracking_filters_init
   5
   6 From: Seth Barry <seth@cyberseth.com>
   7
   8 commit a57c1dcb93e43357ed3f666e5a2b5d5071dd3930 upstream.
   9
  10 While having tda18271 module set with debug=17 (cal & info prints) and
  11 cal=0 (delay calibration process until first use) - I discovered that
  12 during the calibration process, if the frequency test for 69750000
  13 returned a bcal of 0 (see tda18721-fe.c in tda18271_powerscan func) that
  14 the tuner wouldn't be able to pickup any of the frequencies in the range
  15 (all the other frequencies bands returned bcal=1).  I spent some time
  16 going over the code and the NXP's tda18271 spec (ver.4 of it i think) and
  17 adding a lot of debug prints and walking/stepping through the calibration
  18 process.  I found that when the powerscan fails to find a frequency, the
  19 rf calibration is not run and the default value is supposed to be used in
  20 its place (pulled from the RF_CAL_map table) - but something was getting
  21 goofed up there.
  22
  23 Now, my c coding skills are very rusty, but i think root of the problem is
  24 a signedness issue with the math operation for calculating the rf_a1 and
  25 rf_a2 values in tda18271_rf_tracking_filters_init func, which results in
  26 values like 20648 for rf_a1 (when it should probably have a value like 0,
  27 or so slightly negative that it should be zero - this bad value for rf_a1
  28 would in turn makes the approx calc within
  29 tda18271c2_rf_tracking_filters_correction go out of whack).  The simplest
  30 solution i found was to explicitly convert the signedness of the
  31 denominator to avoid the implicit conversion.  The values placed into the
  32 u32 rf_freq array should never exceed about 900mhz, so i think the s32 max
  33 value shouldn't be an issue in this case.
  34
  35 I've tested it out a little, and even when i get a bcal=0 with the
  36 modified code, the default calibration value gets used, rf_a1 is zero, and
  37 the tuner seems to lock on the stream and mythtv seems to play it fine.
  38
  39 Signed-off-by: Seth Barry <seth@cyberseth.com>
  40 Signed-off-by: Michael Krufky <mkrufky@kernellabs.com>
  41 Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
  42 Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
  43
  44 ---
  45  drivers/media/common/tuners/tda18271-fe.c |    4 ++--
  46  1 file changed, 2 insertions(+), 2 deletions(-)
  47
  48 --- a/drivers/media/common/tuners/tda18271-fe.c
  49 +++ b/drivers/media/common/tuners/tda18271-fe.c
  50 @@ -595,13 +595,13 @@ static int tda18271_rf_tracking_filters_
  51                 case RF2:
  52                         map[i].rf_a1 = (prog_cal[RF2] - prog_tab[RF2] -
  53                                         prog_cal[RF1] + prog_tab[RF1]) /
  54 -                               ((rf_freq[RF2] - rf_freq[RF1]) / 1000);
  55 +                               (s32)((rf_freq[RF2] - rf_freq[RF1]) / 1000);
  56                         map[i].rf2   = rf_freq[RF2] / 1000;
  57                         break;
  58                 case RF3:
  59                         map[i].rf_a2 = (prog_cal[RF3] - prog_tab[RF3] -
  60                                         prog_cal[RF2] + prog_tab[RF2]) /
  61 -                               ((rf_freq[RF3] - rf_freq[RF2]) / 1000);
  62 +                               (s32)((rf_freq[RF3] - rf_freq[RF2]) / 1000);
  63                         map[i].rf_b2 = prog_cal[RF2] - prog_tab[RF2];
  64                         map[i].rf3   = rf_freq[RF3] / 1000;
  65                         break;