The following VulkanCTS 1.1.4 tests will not complete on Braswell hardware: dEQP-VK.spirv_assembly.instruction.graphics.spirv_ids_abuse.lots_ids_vert dEQP-VK.spirv_assembly.instruction.graphics.spirv_ids_abuse.lots_ids_tessc dEQP-VK.spirv_assembly.instruction.graphics.spirv_ids_abuse.lots_ids_frag dEQP-VK.spirv_assembly.instruction.graphics.spirv_ids_abuse.lots_ids_geom
The tests are crashing due to allocation failure after running for some time which is caused by enormous size of the resulting native code of the shaders. Initially it was bisected to: 8fdee457a4cb20f4587b5ec817aa1f9325bd5f1c is the first bad commit commit 8fdee457a4cb20f4587b5ec817aa1f9325bd5f1c Author: Jason Ekstrand <jason.ekstrand@intel.com> Date: Thu Jan 10 13:39:05 2019 -0600 anv/pipeline: Move lower_explicit_io much later Now that nir_opt_copy_prop_vars can properly handle array derefs on vectors, it's safe to move UBO and SSBO lowering to late in the pipeline. This should allow NIR to actually start optimizing SSBO access. Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com> However this commit just triggered the underlying issue - force loop unrolling without any check for resulting unrolled size. I've sent a merge request to fix this: https://gitlab.freedesktop.org/mesa/mesa/merge_requests/1365
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/mesa/mesa/issues/1818.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.