Re: __i915_spin_request() sucks

[Mike Galbraith <umgwanakikbuti@gmail.com>]

On Fri, 2015-11-13 at 08:36 -0700, Jens Axboe wrote:
> Previous patch was obvious pre-coffee crap, this should work for using
> ktime to spin max 1usec.

1us seems a tad low.  I doubt the little wooden gears and pulleys of my
core2 Toshiba Satellite lappy can get one loop ground out in a usec :)

> 
> diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
> index 5cf4a1998273..21192e55c33c 100644
> --- a/drivers/gpu/drm/i915/i915_gem.c
> +++ b/drivers/gpu/drm/i915/i915_gem.c
> @@ -1148,17 +1148,18 @@ static bool missed_irq(struct drm_i915_private *dev_priv,
>  
>  static int __i915_spin_request(struct drm_i915_gem_request *req)
>  {
> -	unsigned long timeout;
> +	ktime_t timeout;
>  
>  	if (i915_gem_request_get_ring(req)->irq_refcount)
>  		return -EBUSY;
>  
> -	timeout = jiffies + 1;
> +	timeout = ktime_get();
> +	ktime_add_us(timeout, 1);
>  	while (!need_resched()) {
>  		if (i915_gem_request_completed(req, true))
>  			return 0;
>  
> -		if (time_after_eq(jiffies, timeout))
> +		if (ktime_after(ktime_get(), timeout))
>  			break;
>  
>  		cpu_relax_lowlatency();
>