Re: i82559 on ebsa patch

[Jonathan Larmour <jifl at eCosCentric dot com>]

Andrew Lunn wrote:
> -// Interrupts are wire-or'd together.
> -#if 1 < CYGNUM_DEVS_ETH_INTEL_I82559_DEV_COUNT
> -#define CYGHWR_DEVS_ETH_INTEL_I82559_DEMUX_ALL
> -#endif // multiple devs, so demux_all needed
> -
> -// define multiple interrupt handling anyway:
> -//? #define CYGHWR_DEVS_ETH_INTRS (SA11X0_GPIO_PIN_10)
> -
> -// This brings in code to ensure missed interrupts are properly
> -// acknowledged so that another interrupt can occur in future.
> -// Only a problem with edge-triggered systems.
> -
> -/* #define CYGHWR_DEVS_ETH_INTEL_I82559_MISSED_INTERRUPT(p_i82559) \
> -     (CYGHWR_DEVS_ETH_INTRS != (CYGHWR_DEVS_ETH_INTRS & *SA11X0_GPIO_PIN_LEVEL))
> -*/
> +// Interrupts are multiplex onto one interrupt pin.
> +#define CYGNUM_DEVS_ETH_INTEL_I82559_SEPARATE_MUX_INTERRUPT \
> +          CYGNUM_HAL_INTERRUPT_PCI_IRQ
Wouldn't it be better to still conditionalise this on 
CYGNUM_DEVS_ETH_INTEL_I82559_DEV_COUNT ? The board may have just one card 
in (for whatever reason) and that's the type of thing that option is for 
(if disabling the individual port drivers in ebsa285_eth_drivers.cdl).

And since I can't work it out, why is DEMUX_ALL no longer needed?

Jifl
--
eCosCentric       http://www.eCosCentric.com/       <info@eCosCentric.com>
--[ "You can complain because roses have thorns, or you ]--
--[  can rejoice because thorns have roses." -Lincoln   ]-- Opinions==mine