[Intel-gfx] [RFC 00/22] Gen7 batch buffer command parser

[Volkin, Bradley D]

On Wed, Feb 05, 2014 at 10:30:00AM -0800, Daniel Vetter wrote:
> On Wed, Feb 5, 2014 at 7:18 PM, Volkin, Bradley D
> <bradley.d.volkin at intel.com> wrote:
> > On Wed, Feb 05, 2014 at 02:28:29AM -0800, Chris Wilson wrote:
> >> On Tue, Nov 26, 2013 at 08:51:17AM -0800, bradley.d.volkin at intel.com wrote:
> >> > From: Brad Volkin <bradley.d.volkin at intel.com>
> >> >
> >> > Certain OpenGL features (e.g. transform feedback, performance monitoring)
> >> > require userspace code to submit batches containing commands such as
> >> > MI_LOAD_REGISTER_IMM to access various registers. Unfortunately, some
> >> > generations of the hardware will noop these commands in "unsecure" batches
> >> > (which includes all userspace batches submitted via i915) even though the
> >> > commands may be safe and represent the intended programming model of the device.
> >> >
> >> > This series introduces a software command parser similar in operation to the
> >> > command parsing done in hardware for unsecure batches. However, the software
> >> > parser allows some operations that would be noop'd by hardware, if the parser
> >> > determines the operation is safe, and submits the batch as "secure" to prevent
> >> > hardware parsing. Currently the series implements this on IVB and HSW.
> >>
> >> Just one more question... Do you have a branch for people to test?
> >
> > Not at the moment. And as mentioned in the v2 cover letter, it's actually not
> > particularly testable (or mergeable for that matter) right now because of a
> > regression in secure dispatch on nightly.
> 
> The command parser itself should still work, even with the regression
> in -nightly. The copying and secure dispatch are obviously fail atm.
> That still leaves regression testing of current userspace and
> micro-optimizing the checker itself as possible things to do. Otoh not
> sure what exactly Chris wanted to test.

To test/merge, we'd have to change the series to take out the part where
patch 02/13 sets I915_DISPATCH_SECURE to avoid a BUG_ON() when i915.enable_cmd_parser=1.
But yes, otherwise the parsing works and I think should be sufficient for
what Chris indicated he wants to test.

- Brad

> -Daniel
> -- 
> Daniel Vetter
> Software Engineer, Intel Corporation
> +41 (0) 79 365 57 48 - http://blog.ffwll.ch