/aosp_15_r20/external/mesa3d/docs/relnotes/ |
H A D | 7.10.rst | 36 - New fragment shader back-end for i965-class hardware. 37 - Support for Sandybridge chipset in i965 DRI driver. 81 [i965 gles2c bisected] OpenGL ES 2.0 conformance 108 [i965] brw_fs.cpp:1461: void 143 - i965: Update renderer strings for sandybridge 1034 - i965: Share the KIL_NV implementation between glsl and non-glsl. 1035 - i965: Also enable CC statistics when doing OQs. 1036 - i965: Track the windowizer's dispatch for kill pixel, promoted, and 1043 - i965: Fix the vector/expression splitting for the write_mask change. 1044 - i965: When splitting vector variable assignment, ignore unset [all …]
|
H A D | 10.2.5.rst | 41 - i965/fs: Refactor check for potential copy propagated instructions. 42 - i965/fs: skip copy-propate for logical instructions with negated src 44 - i965/vec4: skip copy-propate for logical instructions with negated 53 - i965: Fix z_offset computation in intel_miptree_unmap_depthstencil() 79 - i965: Generalize the pixel_x/y workaround for all UW types. 102 - i965: Add auxiliary surface field #defines for Broadwell. 111 - i965: Don't copy propagate abs into Broadwell logic instructions. 112 - i965: Set execution size to 8 for instructions with force_sechalf 114 - i965/fs: Set force_uncompressed and force_sechalf on samplepos setup. 115 - i965/fs: Use WE_all for gl_SampleID header register munging. [all …]
|
H A D | 10.2.4.rst | 42 - i965/fs: Refactor check for potential copy propagated instructions. 43 - i965/fs: skip copy-propate for logical instructions with negated src 45 - i965/vec4: skip copy-propate for logical instructions with negated 61 - i965: Generalize the pixel_x/y workaround for all UW types. 72 - i965: Add auxiliary surface field #defines for Broadwell. 76 - i965: Don't copy propagate abs into Broadwell logic instructions. 77 - i965: Set execution size to 8 for instructions with force_sechalf 79 - i965/fs: Set force_uncompressed and force_sechalf on samplepos setup. 80 - i965/fs: Use WE_all for gl_SampleID header register munging. 81 - i965: Add plumbing for Broadwell's auxiliary surface support. [all …]
|
H A D | 9.0.3.rst | 42 [bisected i965]Oglc shad-compiler(advanced.TestLessThani) regressed 141 - i965/disasm: Fix horizontal stride of dest registers 145 - i965/fs: Fix the gen6-specific if handling for 80ecb8f15b9ad7d6edc 146 - i965/fs: Don't generate saturates over existing variable values. 147 - i965: Actually add support for GL_ANY_SAMPLES_PASSED from GL_ARB_oq2. 148 - i965/vs: Try again when we've successfully spilled a reg. 149 - i965/gen7: Set up all samplers even if samplers are sparsely used. 171 - i965: Fix primitive restart on Haswell. 172 - i965: Refactor texture swizzle generation into a helper. 173 - i965: Do texture swizzling in hardware on Haswell. [all …]
|
H A D | 17.3.4.rst | 44 - [GEN9+] 2-3% perf drop in GfxBench Manhattan 3.1 from "i965: 66 - i965: perform 2 uploads with dual slot \*64*PASSTHRU formats on gen<8 111 - cherry-ignore: i965: Accept CONTEXT_ATTRIB_PRIORITY for 124 - cherry-ignore: add i965 shader cache fixes 160 - i965: Call brw_cache_flush_for_render in predraw_resolve_framebuffer 161 - i965: Add more precise cache tracking helpers 162 - i965/blorp: Add more destination flushing 163 - i965: Track the depth and render caches separately 164 - i965: Track format and aux usage in the render cache 166 - i965/miptree: Refactor CCS_E and CCS_D cases in render_aux_usage [all …]
|
H A D | 10.5.1.rst | 69 [HSW, regression, bisected] i965/fs: Emit MAD instructions when 72 [HSW, regression, bisected] i965: Add LINTERP/CINTERP to 84 - i965/vec4: Don't lose the saturate modifier in copy propagation. 88 - i965/gs: Check newly-generated GS-out VUE map against correct stage 108 - i965: Fix out-of-bounds accesses into pull_constant_loc array 112 - i965/fs/nir: Use emit_math for nir_op_fpow 128 - i965: Split Gen4-5 BlitFramebuffer code; prefer BLT over Meta. 131 - i965/fs: Set force_writemask_all on shader_time instructions. 132 - i965/fs: Set smear on shader_time diff register. 133 - i965/fs: Make emit_shader_time_write return rather than emit. [all …]
|
H A D | 11.0.8.rst | 71 - cherry-ignore: don't pick a specific i965 formats patch 72 - Revert "i965/nir: Remove unused indirect handling" 73 - Revert "i965/state: Get rid of dword_pitch arguments to buffer 75 - Revert "i965/vec4: Use a stride of 1 and byte offsets for UBOs" 76 - Revert "i965/fs: Use a stride of 1 and byte offsets for UBOs" 77 - Revert "i965/vec4: Use byte offsets for UBO pulls on Sandy Bridge" 82 - i965: Resolve color and flush for all active shader images in 112 - i965/vec4: Use byte offsets for UBO pulls on Sandy Bridge 113 - i965/fs: Use a stride of 1 and byte offsets for UBOs 114 - i965/vec4: Use a stride of 1 and byte offsets for UBOs [all …]
|
H A D | 9.1.4.rst | 34 [bisected i965] Bus error (core dumped) on oglc texdecaltile 119 - i965: Fix glColorPointer(GL_FIXED) 147 - i965: fix problem with constant out of bounds access (v3) 154 - i965/fs: Bake regs_written into the IR instead of recomputing it 156 - i965/vs: Fix implied_mrf_writes() for integer division pre-gen6. 162 - i965: Shut up the last release build warning. 184 - mesa: Add i965 varying index patches to .cherry-ignore. 185 - i965: Turn brw->urb.vs_size and gs_size into local variables. 186 - i965: Use a variable for the push constant size in kB. 187 - i965: Update URB partitioning code for Haswell's GT3 variant. [all …]
|
H A D | 17.1.4.rst | 33 Stellaris - colored overlay of sectors doesn't render on i965 63 - i965: Add and initialize l3_banks field for gen7+ 64 - i965: Fix broxton 2x6 l3 config 107 - i965: update MaxTextureRectSize to match PRMs and comply with OpenGL 117 - i965: Flush around state base address 118 - i965: Take a uint64_t immediate in emit_pipe_control_write 119 - i965: Unify the two emit_pipe_control functions 120 - i965: Do an end-of-pipe sync prior to STATE_BASE_ADDRESS 121 - i965/blorp: Do an end-of-pipe sync around CCS ops 122 - i965: Do an end-of-pipe sync after flushes [all …]
|
H A D | 9.1.3.rst | 56 [i965 Bisected]Ogles1conform/Ogles2conform/Ogles3conform cases 104 - i965/fs: Don't try to use bogus interpolation modes pre-Gen6. 114 - i965/fs: Remove creation of a MOV instruction that's never used. 115 - i965/fs: Move varying uniform offset compuation into the helper func. 116 - i965: Make the constant surface interface take a normal byte size. 117 - i965/fs: Avoid inappropriate optimization with regs_written > 1. 118 - i965/fs: Do CSE on gen7's varying-index pull constant loads. 119 - i965/fs: Clean up the setup of gen4 simd16 message destinations. 120 - i965/gen7: Skip resetting SOL offsets at batch start with HW 122 - i965/gen6: Reduce updates of transform feedback offsets with HW [all …]
|
H A D | 7.10.1.rst | 39 - Fix an i965 shader bug where the negative absolute value was 55 [i965] After updating to Mesa 7.9, Civilization IV starts to show 219 - i965/fs: When producing ir_unop_abs of an operand, strip negate. 220 - i965/vs: When MOVing to produce ABS, strip negate of the operand. 221 - i965/fs: Do flat shading when appropriate. 222 - i965: Avoid double-negation of immediate values in the VS. 224 - i965: Fix dead pointers to fp->Parameters->ParameterValues[] after 226 - docs: Add a relnote for the Civ IV on i965. 233 - i965: Fix a bug in i965 compute-to-MRF. 234 - i965/fs: Add a helper function for detecting math opcodes. [all …]
|
H A D | 8.0.5.rst | 157 - i965/gen7: Reduce GT1 WM thread count according to updated BSpec. 158 - i965/fs: Invalidate live intervals in passes that remove an 160 - i965: Fix bug in the old FS backend's projtex() calculation. 161 - i965: Add support for GL_SKIP_DECODE_EXT on other SRGB formats. 162 - i965/vs: Convert EdgeFlagPointer values appropriately for the VS on 164 - i965: Fix accumulator_contains() test to also reject swizzles of the 168 - i965: Drop the confusing saturate argument to math instruction setup. 173 - Revert "i965: Avoid unnecessary recompiles for shaders that don't use 175 - i965: Fix regression in depth texture rendering on pre-SNB 198 - i965/fs: Initialize output_components[] by filling it with zeros. [all …]
|
H A D | 12.0.2.rst | 75 [i965 on HD4600 Haswell] xcom switch to ingame cinematics cause 114 - i965: Fix miptree layout for EGLImage-based renderbuffers 115 - i965: Respect miptree offsets in intel_readpixels_tiled_memcpy() 139 - i965: store reference to the context within struct brw_fence (v2) 149 - Revert "i965/miptree: Set logical_depth0 == 6 for cube maps" 162 - i965: Emit SKL VF cache invalidation W/A from 164 - i965: Make room in the batch epilogue for three more pipe controls. 165 - i965: Fix remaining flush vs invalidate race conditions in 221 - i965/miptree: Enforce that height == 1 for 1-D array textures 222 - i965/miptree: Set logical_depth0 == 6 for cube maps [all …]
|
H A D | 10.2.2.rst | 38 [i965 Bisected]Piglit/glx_glx-make-glxdrawable-current fails 47 [i965] glClear on a multisample texture doesn't work 119 - i965: Don't use the head sentinel as an fs_inst in Gen4 workaround 121 - i965: Invalidate live intervals when inserting Gen4 SEND workarounds. 122 - i965/vec4: Fix dead code elimination for VGRFs of size > 1. 123 - i965: Add missing MOCS setup for 3DSTATE_INDEX_BUFFER on Broadwell. 124 - i965: Drop Broadwell perf_debugs about missing MOCS that aren't 126 - i965: Add missing newlines to a few perf_debug messages. 127 - i965/vec4: Use the sampler for pull constant loads on Broadwell. 128 - i965: Use 8x4 aligned rectangles for HiZ operations on Broadwell. [all …]
|
H A D | 13.0.6.rst | 117 - i965: move brw_define.h ifndef guard to the top 162 - i965: Use a better guardband calculation. 164 - i965/fs: Remove the inline pack_double_2x32 optimization 185 - i965: Fix fast depth clears for surfaces with a dimension of 16384. 186 - i965: Use a UW source type for CS_OPCODE_CS_TERMINATE. 187 - i965: Fix check for negative pitch in can_do_fast_copy_blit(). 188 - i965: Support the force_glsl_version driconf option. 189 - i965: Combine the Gen6 SF and Clip viewport atoms. 197 - i965/fs: fix uninitialized memory access 244 - i965/fs: mark last DF uniform array element as 64 bit live one [all …]
|
H A D | 19.0.0.rst | 39 GL_ARB_shader_ballot on i965. 59 - [i965] Downward causes GPU hangs and misrendering on Haswell 83 - i965 incorrectly calculates the number of layers for texture views 118 - i965 regressions from EXT_texture_sRGB_R8 253 - i965: Lift restriction in external textures for EGLImage support 314 - i965: avoid 'unused variable' warnings 315 - i965/batch: avoid reverting batch buffer if saved state is an empty 320 - i965: re-emit index buffer state on a reset option change. 324 - i965/icl: Set Error Detection Behavior Control Bit in L3CNTLREG 327 - i965/icl: Fix L3 configurations [all …]
|
H A D | 9.1.2.rst | 41 [i965 Bisected]Oglc fbblit(advanced.blitFb-3d-cube.mirror.both) fails 116 - i965: Avoid segfault in gen6_upload_state 120 - i965/vs: Fix Gen4/5 VUE map inconsistency with gl_ClipVertex 131 - i965/fs: Fix register allocation for uniform pull constants in 133 - i965/fs: Fix broken rendering in large shaders with UBO loads. 134 - i965/fs: Also do the gen4 SEND dependency workaround against other 136 - i965: Add definitions for gen7+ data cache messages. 161 - i965: Fix INTEL_DEBUG=shader_time for Haswell. 162 - i965: Specialize SURFACE_STATE creation for shader time. 163 - i965: Make INTEL_DEBUG=shader_time use the RAW surface format. [all …]
|
H A D | 13.0.0.rst | 28 - OpenGL ES 3.1 on i965/hsw 29 - OpenGL ES 3.2 on i965/gen9+ (Skylake and later) 30 - GL_ARB_ES3_1_compatibility on i965 31 - GL_ARB_ES3_2_compatibility on i965/gen8+ 35 - GL_ARB_enhanced_layouts on i965, nv50, nvc0, radeonsi, llvmpipe, 41 - GL_ARB_shader_viewport_layer_array on i965/gen6+ 42 - GL_ARB_stencil_texturing on i965/hsw 43 - GL_ARB_texture_stencil8 on i965/hsw 45 - GL_KHR_blend_equation_advanced on i965 47 - GL_KHR_texture_compression_astc_sliced_3d on i965 [all …]
|
H A D | 8.0.3.rst | 51 [bisected i965] oglc sRGB(Mipmap.1D_textures) regressed 53 [i965]oglc max_values(negative.textureSize.textureCube) segfaults 159 - i965: fix typo 163 - i965/fs: Jump from discard statements to the end of the program when 167 - i965/fs: Move GL_CLAMP handling to coordinate setup. 168 - i965/fs: Implement GL_CLAMP behavior on texture rectangles on gen6+. 181 - i965/vs: Fix up swizzle for dereference_array of matrices. 232 - i965: Actually upload sampler state pointers for the VS unit on Gen6. 233 - i965/fs: Fix FB writes that tried to use the non-existent m16 239 - i965: Fix GPU hangs in the dummy fragment shader. [all …]
|
H A D | 7.11.1.rst | 193 - i965/fs: Respect ARB_color_buffer_float clamping. 194 - i965: Add missing \_NEW_POLYGON flag to polygon stipple upload. 195 - i965: Fix polygon stipple offset state flagging. 198 - glsl: Fix gl_NormalMatrix swizzle setup to match i965's invariants. 256 - i965/gen5+: Fix incorrect miptree layout for non-power-of-two 258 - i965: Use proper texture alignment units for cubemaps on Gen5+. 259 - i965: Fix incorrect maximum PS thread count shift on Ivybridge. 260 - i965: Emit depth stalls and flushes before changing depth state on 262 - i965/fs: Allow SIMD16 with control flow on Ivybridge. 263 - i965: Allow SIMD16 color writes on Ivybridge. [all …]
|
H A D | 11.1.3.rst | 80 - i965: Fix assert conditions for src/dst x/y offsets 84 - i965: Make sure we blit a full compressed block 85 - i965/skl: Add two missing device IDs 94 - i965/blorp: Fix hiz ops on MSAA surfaces 139 - i965: Reupload push and pull constants when we get new shader image 141 - i965/fs: Add missing analysis invalidation in opt_sampler_eot(). 142 - i965/fs: Add missing analysis invalidation in fixup_3src_null_dest(). 143 - i965/vec4: Consider removal of no-op MOVs as progress during register 182 - i965: Only magnify depth for 3D textures, not array textures. 200 - i965/vec4: don't copy ATTR into 3src instructions with complex [all …]
|
H A D | 9.2.1.rst | 83 - i965/fs: Gen4: Zero out extra coordinates when using shadow compare 84 - i965: Fix cube array coordinate normalization 85 - i965: fix bogus swizzle in brw_cubemap_normalize 108 - i965/gen4: Fix fragment program rectangle texture shadow compares. 109 - i965: Reenable glBitmap() after the sRGB winsys enabling. 136 - i965/vs: Detect GRF sources in split_virtual_grfs send-from-GRF code. 137 - i965/fs: Detect GRF sources in split_virtual_grfs send-from-GRF code. 138 - i965/vec4: Only zero out unused message components when there are 140 - i965: Fix brw_vs_prog_data_compare to actually check field members. 156 - i965: Initialize inout_offset parameter to brw_search_cache().
|
H A D | 8.0.1.rst | 94 - i965: Rewrite the HiZ op 95 - i965: Remove file i965/junk, accidentally added in 7b36c68 103 - i965: Fix HiZ change compiler warning. 104 - i965: Report the failure message when failing to compile the fragment 106 - i965/fs: Enable register spilling on gen7 too. 117 - i965: Fix border color on Ironlake. 118 - i965/fs: Add a new fs_inst::regs_written function. 119 - i965/fs: Take # of components into account in try_rewrite_rhs_to_dst. 120 - i965: Emit Ivybridge VS workaround flushes.
|
H A D | 17.2.6.rst | 78 - cherry-ignore: i965: Mark BOs as external when we export their handle 88 - i965: Program DWord Length in MI_FLUSH_DW 89 - i965/gen8+: Fix the number of dwords programmed in MI_FLUSH_DW 129 - i965: Add stencil buffers to cache set regardless of stencil 134 - i965: properly initialize brw->cs.base.stage to MESA_SHADER_COMPUTE 135 - i965: Make L3 configuration atom listen for TCS/TES program updates. 137 - i965: Implement another VF cache invalidate workaround on Gen8+. 138 - i965: Upload invariant state once at the start of the batch on 143 - i965/fs: Fix extract_i8/u8 to a 64-bit destination 144 - i965/fs: Split all 32->64-bit MOVs on CHV, BXT, GLK
|
H A D | 10.6.2.rst | 92 - i965/fs: Fix ir_txs in emit_texture_gen4_simd16(). 93 - i965: Reserve more batch space to accomodate Gen6 perfmonitors. 94 - i965/vs: Fix matNxM vertex attributes where M != 4. 96 - Revert "i965: Delete linked GLSL IR when using NIR." 113 - i965/fs: Don't mess up stride for uniform integer multiplication. 123 - i965: allocate at least 1 BLEND_STATE element 127 - i965/skl: Set the pulls bary bit in 3DSTATE_PS_EXTRA 130 - i965: Don't try to print the GLSL IR if it has been freed 135 - i965: Delete linked GLSL IR when using NIR. 138 - i965: use EmitNoIndirectSampler for gen < 7
|