> Hello
Hello Remi!
Thank you for looking at the Avercard with the CX23885 driver.
>
> Antti, redirected me toward you,
What exactly is your question?
> And
>
> If it's at the least at the same "stage" as the HC81 , I think it deserve's to
> be listed in "cards".h
Possibly. Tell us again 1) what you have working reliably 2) what
isn't working but the driver is exposing and 3) what the driver is
exposing but has not been tested.
> So people will know right away, that this card has been identified by the V4L
> community, and dont have
Anyone working on that card would arrive here on the mailing list, I
don't think that's going to be an issue. I suggest you focus on
getting feature of the card working reliably, produce a patchset and
I'm sure it will get reviewed, refined then eventually merged.
Other comments below.
> > + [CX23885_BOARD_AVERMEDIA_A306] = {
> > + .name = "AVerTV Hybrid Minicard PCIe A306",
> > + .tuner_type = TUNER_XC2028,
> > + .tuner_addr = 0x61, /* 0xc2 >> 1 */
> > + .tuner_bus = 1,
> > + .porta = CX23885_ANALOG_VIDEO,
> > + .portb = CX23885_MPEG_ENCODER,
> > + .input = {{
> > + .type = CX23885_VMUX_TELEVISION,
> > + .vmux = CX25840_VIN2_CH1 |
> > + CX25840_VIN5_CH2 |
> > + CX25840_NONE0_CH3 |
> > + CX25840_NONE1_CH3,
> > + .amux = CX25840_AUDIO8,
> > + }, {
> > + .type = CX23885_VMUX_SVIDEO,
> > + .vmux = CX25840_VIN8_CH1 |
> > + CX25840_NONE_CH2 |
> > + CX25840_VIN7_CH3 |
> > + CX25840_SVIDEO_ON,
> > + .amux = CX25840_AUDIO6,
> > + }, {
> > + .type = CX23885_VMUX_COMPONENT,
> > + .vmux = CX25840_VIN1_CH1 |
> > + CX25840_NONE_CH2 |
> > + CX25840_NONE0_CH3 |
> > + CX25840_NONE1_CH3,
> > + .amux = CX25840_AUDIO6,
> > + }},
> > +
> > + }
Does the card have a MPEG2 hardware compressor and is it functioning
properly? (/dev/video1)
Are both svideo and component inputs working correctly in tvtime?
What about audio inputs? Does the card have any audio inputs and is
the driver acting and exposing those features correctly?
If not then please remove any of these sections.
> > + case CX23885_BOARD_AVERMEDIA_A306:
> > + cx_clear(MC417_CTL, 1);
> > + /* GPIO-0,1,2 setup direction as output */
> > + cx_set(GP0_IO, 0x00070000);
> > + mdelay(10);
> > + /* AF9013 demod reset */
> > + cx_set(GP0_IO, 0x00010001);
> > + mdelay(10);
> > + cx_clear(GP0_IO, 0x00010001);
> > + mdelay(10);
> > + cx_set(GP0_IO, 0x00010001);
> > + mdelay(10);
> > + /* demod tune? */
> > + cx_clear(GP0_IO, 0x00030003);
> > + mdelay(10);
> > + cx_set(GP0_IO, 0x00020002);
> > + mdelay(10);
> > + cx_set(GP0_IO, 0x00010001);
> > + mdelay(10);
> > + cx_clear(GP0_IO, 0x00020002);
> > + /* XC3028L tuner reset */
> > + cx_set(GP0_IO, 0x00040004);
> > + cx_clear(GP0_IO, 0x00040004);
> > + cx_set(GP0_IO, 0x00040004);
> > + mdelay(60);
> > + break;
You're setting and clearing the GPIO direction enable registers (upper
16 bits), this isn't a good idea.
If you want to drive a GPIO in a specific direction (lower 8 bits),
perhaps for a reset, instead do this:
/* AF9013 demod reset */
cx_set(GP0_IO, 0x00010001); /* Establish the direction of the GPIO and
it's signal level)
mdelay(10);
cx_clear(GP0_IO, 0x00000001); /* change the signal level, drive to low */
mdelay(10);
cx_set(GP0_IO, 0x00000001); /* back to high */
mdelay(10);
Repeat this example for other other 'demod tune' and 3028 resets you
are doing, don't toggle the upper 16bits, else you leave the GPIO
floating, a bad idea.
- Steve
--
Steven Toth - Kernel Labs
http://www.kernellabs.com
--
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
