Re: [PATCH 5/5] mvsas : redesign the mvsas driver architecture.

[Reinhard Nissl <rnissl@gmx.de>]

Hi,

Ke Wei schrieb:

> +		mvs_64xx_detect_porttype(mvi, i);
> +	}
> +	if (mvi->flags & MVF_FLAG_SOC) {
> +		/* set select registers */
> +		writel(0x0E008000, regs + 0x000);
> +		writel(0x59000008, regs + 0x004);
> +		writel(0x20, regs + 0x008);
> +		writel(0x20, regs + 0x00c);
> +		writel(0x20, regs + 0x010);
> +		writel(0x20, regs + 0x014);
> +		writel(0x20, regs + 0x018);
> +		writel(0x20, regs + 0x01c);
> +	}
> +	for (i = 0; i < mvi->chip->n_phy; i++) {
> +		/* clear phy int status */
> +		tmp = mvs_read_port_irq_stat(mvi, i);
> +		tmp &= ~PHYEV_SIG_FIS;
> +		mvs_write_port_irq_stat(mvi, i, tmp);
> +
> +		/* set phy int mask */
> +		tmp = PHYEV_RDY_CH | PHYEV_BROAD_CH | PHYEV_UNASSOC_FIS |
> +			PHYEV_ID_DONE | PHYEV_DCDR_ERR | PHYEV_CRC_ERR |
> +			PHYEV_DEC_ERR;
> +		mvs_write_port_irq_mask(mvi, i, tmp);
> +
> +		msleep(100);
> +		mvs_update_phyinfo(mvi, i, 1);

I assume that this code doesn't work with my Seagate 10k6 drive
ST3450856SS 0003. Looks like it still just waits for 100 ms which
seems to be not sufficient for this drive (see the thread on this
list regarding this drive).

Is there no way to determine that port detection has finished?

Bye.
-- 
Dipl.-Inform. (FH) Reinhard Nissl
mailto:rnissl@gmx.de