[linux-dvb] [emard@softhome.net: Re: dvb-kernel stv0299 fixed for budgets including satelco]

[Emard <emard@softhome.net>]

----- Forwarded message from Emard <emard@softhome.net> -----

From: Emard <emard@softhome.net>
To: Holger Waechtler <holger@convergence.de>
Subject: Re: [linux-dvb] dvb-kernel stv0299 fixed for budgets including satelco
Reply-To: emard@softhome.net

On Fri, Jan 24, 2003 at 01:10:24PM +0100, Holger Waechtler wrote:
> diff -pur 
> orig/dvb-kernel/linux/drivers/media/dvb/ttpci-budget/budget-core.c dvb
> -kernel/linux/drivers/media/dvb/ttpci-budget/budget-core.c
> --- orig/dvb-kernel/linux/drivers/media/dvb/ttpci-budget/budget-core.c 
> Thu Jan
> 23 20:00:03 2003
> +++ dvb-kernel/linux/drivers/media/dvb/ttpci-budget/budget-core.c 
> Fri Jan
> 24 00:38:04 2003
> @@ -406,6 +406,8 @@ int budget_attach (struct saa7146_dev* d
> 
>         dvb_register_adapter(&budget->dvb_adapter, budget->card->name);
> 
> +       saa7146_setgpio(dev, 2, SAA7146_GPIO_OUTLO); /* frontend power 
> off */
> +
> 
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> shouldn't this be handled by the FE_INIT/FE_SLEEP ioctl catched in 
> budget_diseqc_ioctl()??
> 
> Holger
> 

I have looked at the code and it seems not so simple. 

You see, the problem exists on satelco, i.e. a budget card 
with diseqc generated by the hardware, so budget_diseqc_ioctl() 
will never be executed because there's some 

if(card type == BUDGET_TT) then register budget_diseqc_ioctl().

and satelco card type != BUDGET_TT but something like 
BUDGET_TT_CI_HW_DISEQC

Besides, from my previous remark, i2c initialization may require
frontend powered down to successfully perform, so even if we move
the power commands to budget_diseqc_ioctl(), we might have to 
call it back from budget_attach().

So I can only stay with the above dmafe.patch as a solution 
that's simple and clear.

Emard

----- End forwarded message -----


-- 
Info:
To unsubscribe send a mail to listar@linuxtv.org with "unsubscribe linux-dvb" as subject.