Re: [PATCH 2/11] it8213/piix/slc90e66: remove {it8213,piix,slc90e66}_dma_2_pio()

[Sergei Shtylyov <sshtylyov@ru.mvista.com>]

Hello.

Bartlomiej Zolnierkiewicz wrote:

> Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@gmail.com>

    Sigh... when I was talking about getting rid of that function, I meant 
that we need to also decouple UltraDMA from PIO -- which is what this patch 
misses...
    Actually, I intended to decouple DMA form PIO in this drivers completely 
but that would be a next step...

> Index: b/drivers/ide/pci/it8213.c
> ===================================================================
> --- a/drivers/ide/pci/it8213.c
> +++ b/drivers/ide/pci/it8213.c
> @@ -124,7 +93,7 @@ static void it8213_set_dma_mode(ide_driv
>  	int w_flag		= 0x10 << drive->dn;
>  	int u_speed		= 0;
>  	u16			reg4042, reg4a;
> -	u8			reg48, reg54, reg55;
> +	u8			reg48, reg54, reg55, pio;
>  
>  	pci_read_config_word(dev, maslave, &reg4042);
>  	pci_read_config_byte(dev, 0x48, &reg48);
> @@ -165,7 +134,11 @@ static void it8213_set_dma_mode(ide_driv
>  				pci_write_config_byte(dev, 0x54, reg54 | v_flag);
>  		} else
>  			pci_write_config_byte(dev, 0x54, reg54 & ~v_flag);
> +
> +		pio = 4;

    Why let this stupidity live?

>  	} else {
> +		const u8 mwdma_to_pio[] = { 0, 3, 4 };
> +
>  		if (reg48 & u_flag)
>  			pci_write_config_byte(dev, 0x48, reg48 & ~u_flag);
>  		if (reg4a & a_speed)
> @@ -174,9 +147,14 @@ static void it8213_set_dma_mode(ide_driv
>  			pci_write_config_byte(dev, 0x54, reg54 & ~v_flag);
>  		if (reg55 & w_flag)
>  			pci_write_config_byte(dev, 0x55, (u8) reg55 & ~w_flag);
> +
> +		if (speed >= XFER_MW_DMA_0)
> +			pio = mwdma_to_pio[speed - XFER_MW_DMA_0];
> +		else
> +			pio = 2; /* only SWDMA2 is allowed */

    Well, MWDMA0 is also not allowed, so you could have cut on mwdma_to_pio 
array size.

MBR, Sergei