[Intel-gfx] [PATCH 11/19] drm/i915: Do a nonblocking wait first in pread/pwrite

[Chris Wilson]

On Fri, Aug 05, 2016 at 10:08:30AM +0300, Joonas Lahtinen wrote:
> On to, 2016-08-04 at 20:52 +0100, Chris Wilson wrote:
> > If we try and read or write to an active request, we first must wait
> > upon the GPU completing that request. Let's do that without holding the
> > mutex (and so allow someone else to access the GPU whilst we wait). Upn
> 
> STILL TYPO                                                    Upon ---^
> 
> >  	/* Bounds check destination. */
> >  	if (args->offset > obj->base.size ||
> >  	    args->size > obj->base.size - args->offset) {
> >  		ret = -EINVAL;
> > -		goto out;
> > +		goto err;
> >  	}
> >  
> > -	trace_i915_gem_object_pwrite(obj, args->offset, args->size);
> > +	ret = __unsafe_wait_rendering(obj, to_rps_client(file), false);
> > +	if (ret)
> > +		goto err;
> >  
> > +	intel_runtime_pm_get(dev_priv);
> > +
> > +	ret = i915_mutex_lock_interruptible(dev);
> > +	if (ret)
> > +		goto err_rpm;
> > +
> > +	trace_i915_gem_object_pwrite(obj, args->offset, args->size);
> 
> This trace is still moved, maybe add your reasoning to commit message.

I think the trace is not good enough. We know we entered the ioctl, we
can trace that itself. So the question is what information is valuable
at this point, and that would be which path we take (gtt, shmem, phys).

Anyway, I'll bump it back to the start and we can add a review of
tracepoints to the wishlist.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre