| Commit message (Expand) | Author | Age | Files | Lines |
... | |
* | aco, radv: Rename aco_*_key to aco_*_info. | Timur Kristóf | 2023-03-08 | 1 | -41/+37 |
* | aco, radv: Add load_grid_size_from_user_sgpr to aco options. | Qiang Yu | 2023-03-08 | 1 | -1/+1 |
* | aco: Generalize vs_inputs to args_pending_vmem. | Timur Kristóf | 2023-03-08 | 1 | -15/+8 |
* | amd: add nir_intrinsic_xfb_counter_sub_amd and fix overflowed streamout offsets | Marek Olšák | 2023-03-07 | 1 | -0/+3 |
* | aco: remove VOP[123C]P? structs | Georg Lehmann | 2023-03-07 | 1 | -33/+35 |
* | aco: implement nir_op_unpack_32_4x8 | Marek Olšák | 2023-03-03 | 1 | -1/+4 |
* | aco: only ls and ps use store output now | Qiang Yu | 2023-02-28 | 1 | -4/+7 |
* | aco: use and swizzle mask in dpp quad perm | Georg Lehmann | 2023-02-27 | 1 | -2/+2 |
* | aco/gfx11: use dpp_row_xmask and dpp_row_share | Georg Lehmann | 2023-02-27 | 1 | -0/+6 |
* | aco: use v_permlane(x)16_b32 for masked swizzle | Georg Lehmann | 2023-02-27 | 1 | -1/+15 |
* | amd,nir: remove byte_permute_amd intrinsic | Georg Lehmann | 2023-02-22 | 1 | -9/+0 |
* | nir: add assertions that loops don't have a Continue Construct | Daniel Schürmann | 2023-02-21 | 1 | -0/+1 |
* | radv: unconditionally enable scratch for RT shaders | Daniel Schürmann | 2023-02-16 | 1 | -0/+1 |
* | aco: Don't set scalar offset on buffer load instructions when it's zero. | Timur Kristóf | 2023-02-16 | 1 | -1/+3 |
* | aco: Get rid of redundant load_vmem_mubuf function. | Timur Kristóf | 2023-02-16 | 1 | -26/+13 |
* | ac: Port ACO's get_fetch_format to ac_get_safe_fetch_size. | Timur Kristóf | 2023-02-16 | 1 | -54/+4 |
* | aco: don't modify exec in p_interp_gfx11 | Rhys Perry | 2023-02-08 | 1 | -6/+5 |
* | ac: move ring_offsets to ac_shader_args | Rhys Perry | 2023-02-06 | 1 | -1/+1 |
* | nir,ac/llvm,aco,radv,radeonsi: remove nir_export_vertex_amd | Qiang Yu | 2023-02-03 | 1 | -138/+0 |
* | nir,ac/llvm,aco: remove nir_export_primitive_amd | Qiang Yu | 2023-02-03 | 1 | -57/+0 |
* | amd,radeonsi: implement nir_load_force_vrs_rates_amd in driver abi | Qiang Yu | 2023-02-03 | 1 | -5/+0 |
* | aco: implement nir_export_amd | Qiang Yu | 2023-02-03 | 1 | -0/+36 |
* | aco: set has_color_exports with GPL | Rhys Perry | 2023-01-27 | 1 | -0/+3 |
* | radv, aco, ac: Implement pack_half_2x16_rtz_split. | Timur Kristóf | 2023-01-26 | 1 | -0/+1 |
* | aco: Remove dynamic VS input loads. | Timur Kristóf | 2023-01-26 | 1 | -37/+1 |
* | aco: add support for fp32 addition atomics | Rhys Perry | 2023-01-17 | 1 | -0/+16 |
* | aco: restore semantic_can_reorder for GS output stores | Rhys Perry | 2023-01-16 | 1 | -1/+5 |
* | radv,aco: use ac_nir_lower_legacy_gs | Rhys Perry | 2023-01-16 | 1 | -102/+2 |
* | aco: Pass correct number of coords to Vega 1D LOD instruction. | Bas Nieuwenhuizen | 2023-01-13 | 1 | -1/+1 |
* | radv,aco: fix enable_mrt_output_nan_fixup for RAGE2 again | Samuel Pitoiset | 2023-01-11 | 1 | -1/+1 |
* | aco: Use NSA on GFX11 with more than 5 vaddr registers. | Georg Lehmann | 2023-01-11 | 1 | -38/+48 |
* | aco: allow Builder::Result to be dereferenced | Rhys Perry | 2023-01-10 | 1 | -11/+9 |
* | aco/gfx11: implement load_input_vertex | Rhys Perry | 2022-12-16 | 1 | -19/+5 |
* | aco/gfx11: export mrtz in discard early exit for non-color shaders | Rhys Perry | 2022-12-16 | 1 | -0/+6 |
* | aco: Emulate Wave64 bpermute on GFX11. | Timur Kristóf | 2022-12-14 | 1 | -1/+3 |
* | aco: Stylistic changes to emit_gfx10_wave64_bpermute. | Timur Kristóf | 2022-12-14 | 1 | -5/+10 |
* | aco: Split opcodes for GFX6 and GFX10 emulated bpermute. | Timur Kristóf | 2022-12-14 | 1 | -3/+3 |
* | nir: Eliminate nir_op_i2b | Ian Romanick | 2022-12-14 | 1 | -2/+1 |
* | nir: replace IS_SWIZZLED flag with ACCESS_IS_SWIZZLED_AMD | Marek Olšák | 2022-12-13 | 1 | -2/+2 |
* | nir: remove redundant SLC_AMD in favor of ACCESS_STREAM_CACHE_POLICY | Marek Olšák | 2022-12-13 | 1 | -2/+2 |
* | radv,aco: move radv_ps_epilog_key to the graphics pipeline key | Samuel Pitoiset | 2022-12-08 | 1 | -6/+7 |
* | radv,aco: rename color output related fields for consistency | Samuel Pitoiset | 2022-12-08 | 1 | -3/+3 |
* | aco: apply NUW to additions for scratch access | Rhys Perry | 2022-12-06 | 1 | -1/+4 |
* | aco: fix missing uses of MRT output flags | Samuel Pitoiset | 2022-12-05 | 1 | -0/+3 |
* | radv,aco: stop lowering FS outputs in NIR | Samuel Pitoiset | 2022-12-05 | 1 | -62/+63 |
* | aco: fix indexing MRT0 alpha channel for alpha-to-coverage via MRTZ on GFX11 | Samuel Pitoiset | 2022-12-05 | 1 | -1/+1 |
* | aco: always use 32-bit for exporting alpha-to-coverage via MRTZ on GFX11 | Samuel Pitoiset | 2022-12-05 | 1 | -20/+5 |
* | aco: Use wave size specific opcode for s_or in cube map coord code. | Georg Lehmann | 2022-12-01 | 1 | -2/+1 |
* | aco: Don't prematurely emit s_andn2. | Georg Lehmann | 2022-11-30 | 1 | -18/+17 |
* | radv,aco: remove old streamout code | Rhys Perry | 2022-11-29 | 1 | -113/+0 |