Re: [PATCH] Re: 2.6.17-rc5-git1: regression: resume from suspend(RAM) fails: libata issue

[Mark Lord <liml@rtr.ca>]

Jens Axboe wrote:
> +               u8 status1, status2, status3;
> +               status1 = ata_chk_status(ap);
> +               ata_busy_wait(ap, ATA_BUSY, 200000);
> +               status2 = ata_chk_status(ap);
> +               status3 = ata_wait_idle(ap);
> +               printk("s 1/2/3 %x/%x/%x\n", status1, status2, status3);
> 
> with the rest being as posted before. It gives me:
> 
> s 1/2/3 80/50/50 on resume.
...

> With that hunk restored to normal, no change. Resume still works, status
> is still 80/50/50.

So, just the "status1 = ata_chk_status(ap);" seems to be enough
to make it work for you, in place of the earlier 500msec delay we started with?

Curious.

Does it still work when the "status1 = ata_chk_status(ap);" is replaced with udelay(1) ??