On Mon, Nov 4, 2013 at 6:22 PM, Mateo Lozano, Oscar <oscar.mateo@xxxxxxxxx> wrote: > Well, I left the return value so that the user could decide whether or not to assert (e.g. gem_reloc_vs_gpu probably wants to assert, but if we end up adding a call to igt_drop_caches_set() inside gem_quiescent_gpu() and we assert, then we are virtually making *all* tests depend on i915_gem_drop_caches being available...). You know better than anyone if what I just said makes sense, so I´ll leave the ultimate decision to you :) Hm, I've forgotten that there are kernels without the drop caches interface. I've just checked and the first version with it is 3.9. That's old enough for me to no longer care, at least wrt testing stuff with i-g-t. If anyone wants to regression test such old kernels they can pick an older version of igt. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/intel-gfx
