Quoting Tvrtko Ursulin (2019-09-27 09:47:59) > > On 25/09/2019 11:01, Chris Wilson wrote: > > -int __i915_vma_do_pin(struct i915_vma *vma, > > - u64 size, u64 alignment, u64 flags) > > +static bool try_qad_pin(struct i915_vma *vma, unsigned int flags) > > What does the QAD stand for? quick and dirty [snip] > High level concept looks okay to me, and the low level details as well. > But it is so huge that I cannot guarantee I haven't missed something in > the middle. So for any hypothetical missed thing in that last third, I > put my faith that before release we catch it and fix it. It's such a huge seismic shift, that we will have to make again before too long, that some little detail will have slipped through. All we can do is to hope to capture anything that escapes the many machine months of testing (but was it testing the right thing???) into a regression test to make sure the next wave of changes is better. Plus we can look forward to all the opportunities to optimise the code all over again. :| -Chris _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
