<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Wed, May 3, 2017 at 12:15 AM, Iago Toral <<a href="mailto:itoral@igalia.com" target="_blank">itoral@igalia.com</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hi Jason Sorry that I missed this. All the changes look good to me and I have sent reviews for the patches that missed them. With this series we pass all the multiview tests.</blockquote><div> </div><div>Thanks! </div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> Iago <div class="HOEnZb"><div class="h5"> On Thu, 2017-04-27 at 09:31 -0700, Jason Ekstrand wrote: > This is mostly a re-send of my earlier patches but there are a few > changes. > I think, at this point, that I'm ready to merge it assuming Iago is > ok with > the changes. The important chages are: > > 1. The lowering pass has been altered to take a view mask instead of > a > subpass. The rest of the anv_pipeline code has also been > modified to > properly incorporate the view mask into the shader key. > 2. There have been two fixes to the code which multiplies the number > of > invocations by the number of views. First was that I > accidentally used > AND instead of ADD. Second was an off-by-one error when adding > ALU > operations to an MI_MATH instruction that made us stomp the > MI_MATH > instruction itself instead of starting on dword 1. > > 3. I've added a trivial implementation of multiDrawIndirect. This > isn't > really required but some of the tests erroneously required it by > mistake. I'm not 100% sure that I like having a software > implementation of multiDrawIndirect but it's no worse (and > probably > slightly better) than an application fall-back would be. > Jason Ekstrand (14): > compiler: Add a system value and varying for ViewIndex > spirv: Bump the SPIR-V header to the latest public version > spirv: Add support for SPV_KHR_multiview > anv/nir: Delete the apply_dynamic_offsets prototype > anv: Add the KHX_multiview boilerplate > anv/pass: Store the per-subpass view mask > anv: Move shader hashing to anv_pipeline > anv/pipeline: Call nir_gather_info later > anv/pipeline: Add a subpass field to anv_pipeline > anv/pipeline: Add shader lowering for multiview > anv/cmd_buffer: Pull indirect draw parameter loading into a helper > anv/cmd_buffer: Emit instanced draws for multiple views > anv: Enable VK_KHX_multiview and SPV_KHR_multiview > anv: Trivially implement multiDrawIndirect > > src/compiler/nir/nir.c | 4 + > src/compiler/nir/nir_intrinsics.h | 1 + > src/compiler/shader_enums.c | 2 + > src/compiler/shader_enums.h | 4 + > src/compiler/spirv/nir_spirv.h | 1 + > src/compiler/spirv/spirv.h | 34 ++++- > src/compiler/spirv/spirv_to_nir.c | 4 + > src/compiler/spirv/vtn_variables.c | 4 + > src/intel/Makefile.sources | 1 + > src/intel/vulkan/anv_device.c | 23 ++- > src/intel/vulkan/anv_entrypoints_gen.py | 1 + > src/intel/vulkan/anv_nir.h | 5 +- > src/intel/vulkan/anv_nir_lower_multiview.c | 234 > +++++++++++++++++++++++++++++ > src/intel/vulkan/anv_pass.c | 19 +++ > src/intel/vulkan/anv_pipeline.c | 83 +++++++--- > src/intel/vulkan/anv_pipeline_cache.c | 27 ---- > src/intel/vulkan/anv_private.h | 15 +- > src/intel/vulkan/genX_cmd_buffer.c | 209 > ++++++++++++++++++++++---- > src/intel/vulkan/genX_pipeline.c | 9 +- > src/mesa/program/prog_print.c | 2 + > 20 files changed, 588 insertions(+), 94 deletions(-) > create mode 100644 src/intel/vulkan/anv_nir_lower_multiview.c > </div></div></blockquote></div> </div></div>