[linux-dvb] a700 support (was: [patch 5/5] mt312: add attach-time setting to invert lnb-voltage (Matthias Schwarzott))

Eduard Huguet eduardhc at gmail.com
Sun Apr 27 13:42:14 CEST 2008


Thank you very much, Matthias. I was going to try the patch right now,
however I'm finding that it doesn't apply clean to the current HG tree. This
is what I'm getting:

patching file linux/drivers/media/dvb/frontends/Kconfig
Hunk #1 FAILED at 368.
1 out of 1 hunk FAILED -- saving rejects to file
linux/drivers/media/dvb/frontends/Kconfig.rej
patching file linux/drivers/media/dvb/frontends/Makefile
Hunk #1 succeeded at 23 (offset -2 lines).
patching file linux/drivers/media/dvb/frontends/zl10036.c
patching file linux/drivers/media/dvb/frontends/zl10036.h
patching file linux/drivers/media/video/saa7134/Kconfig
patching file linux/drivers/media/video/saa7134/saa7134-cards.c
Hunk #3 succeeded at 5716 (offset 42 lines).
patching file linux/drivers/media/video/saa7134/saa7134-dvb.c

I've tried to manually patch Kconfig by adding the rejected lines, but I
suppose there must something I'm doing wrong: apparently it compiles fine,
but saa7134-dvb is not loaded and the frontend is not being created for the
card (although the card is detected and the video0 device for analog is
there).

I'll apreciate any hints you could give me, I'm impatient to test this.
Plus, I'm seing lately some cross posted messages regarding the chips on
this cards, so maybe there's some hope for it :D

Best regards,
  Eduard




2008/4/27 Matthias Schwarzott <zzam at gentoo.org>:

> On Montag, 14. April 2008, Eduard Huguet wrote:
> > > ---------- Missatge reenviat ----------
> > > From: Matthias Schwarzott <zzam at gentoo.org>
> > > To: linux-dvb at linuxtv.org
> > > Date: Sat, 12 Apr 2008 17:04:50 +0200
> > > Subject: [linux-dvb] [patch 5/5] mt312: add attach-time setting to
> invert
> > > lnb-voltage
> > > Add a setting to config struct for inversion of lnb-voltage.
> > > Needed for support of Avermedia A700 cards.
> > >
> > > Signed-off-by: Matthias Schwarzott <zzam at gentoo.org>
> > > Index: v4l-dvb/linux/drivers/media/dvb/frontends/mt312.c
> > > ===================================================================
> > > --- v4l-dvb.orig/linux/drivers/media/dvb/frontends/mt312.c
> > > +++ v4l-dvb/linux/drivers/media/dvb/frontends/mt312.c
> > > @@ -422,11 +422,16 @@ static int mt312_set_voltage(struct dvb_
> > >   {
> > >         struct mt312_state *state = fe->demodulator_priv;
> > >         const u8 volt_tab[3] = { 0x00, 0x40, 0x00 };
> > > +       u8 val;
> > >
> > >         if (v > SEC_VOLTAGE_OFF)
> > >                 return -EINVAL;
> > >
> > > -       return mt312_writereg(state, DISEQC_MODE, volt_tab[v]);
> > > +       val = volt_tab[v];
> > > +       if (state->config->voltage_inverted)
> > > +               val ^= 0x40;
> > > +
> > > +       return mt312_writereg(state, DISEQC_MODE, val);
> > >   }
> > >
> > >   static int mt312_read_status(struct dvb_frontend *fe, fe_status_t
> *s)
> > > Index: v4l-dvb/linux/drivers/media/dvb/frontends/mt312.h
> > > ===================================================================
> > > --- v4l-dvb.orig/linux/drivers/media/dvb/frontends/mt312.h
> > > +++ v4l-dvb/linux/drivers/media/dvb/frontends/mt312.h
> > > @@ -31,6 +31,9 @@
> > >   struct mt312_config {
> > >         /* the demodulator's i2c address */
> > >         u8 demod_address;
> > > +
> > > +       /* inverted voltage setting */
> > > +       int voltage_inverted:1;
> > >   };
> > >
> > >   #if defined(CONFIG_DVB_MT312) || (defined(CONFIG_DVB_MT312_MODULE)
> &&
> > > defined(MODULE))
> > > --
> >
> > Thanks for the patches. ¿Is your lastest unified diff on your page
> > (a700_full_20080412.diff) equivalent to these patches or must they be
> > applied separately?
> Another 2 weeks later. These patches already have been applied at hg
> level. So
> now the new remaining full patch (a700_full_20080427.diff) is to be
> applied
> on top.
>
> It contains a zl10036 driver and the changes to the glue code in
> saa7134-dvb.
> zl10036 works most of the time here, but I am not fully happy with its
> current
> design.
>
> Open issues:
> * Should set_params routine be kept spiltted as it is?
> * Is bandwidth handling sane?
> Now calc the needed bw by symbolrate and add 3MHz.
> Datasheet suggests to start at max setting and decrease as possible after
> the
> lock is established (using freq. offset info from demod).
> * The used gain values are found by try and error and do not work for all
> transponders here.
> The most difficult transponder here was RTL - and that now works for me
> with
> exactly these gain settings (rfg=0, ba=1, bg=1).
>
> >
> > I'll try to some tests tonight, if you have made some progress. By the
> way,
> > ¿could you tell me if it's better to use use_frontend=0 or 1 for
> > saa7134-dvb module? I think that this changes the driver used for
> frontend,
> > but I'm not sure.
>
> use_frontend is intended to be used for choosing between dvb-s and dvb-t
> or
> similar if the frontend offers both exclusively. I abused this parameter
> to
> choose between the two driver for zl10313 demod. This is no longer used,
> so
> forget about this parameter.
>
> Regards
> Matthias
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.linuxtv.org/pipermail/linux-dvb/attachments/20080427/1d9d0d49/attachment.htm 


More information about the linux-dvb mailing list