434 - pan/bi: Add missing sr_count to pseudo-atomics
435 - pan/bi: Don't reference uninit source in ATOM_C1
436 - pan/bi: Add simple constant folding pass
437 - pan/bi: Don't reference nir_lower_mediump_outputs
438 - pan/bi: Simplify Python expression
439 - pan/bi: Union modifiers from across variants
440 - pan/bi: Support 16-bit load_interpolated_input
441 - pan/bi: Emit int CSEL instead of float by default
442 - pan/bi: Implement vectorized f32_to_f16
443 - pan/bi: Fix 16-bit fsat
444 - pan/bi: Improve assert for vector size errors
445 - pan/bi: Implement vectorized int downcasts
446 - pan/bi: Fix loads and stores smaller than 32 bits
447 - pan/bi: Lower swizzles on CLPER
448 - pan/bi: Add and use bi_negzero helper
449 - pan/bi: Don't schedule clamps to +FADD.v2f16
450 - pan/bi: Workaround \*V2F32_TO_V2F16 erratum
453 - pan/bi: Track dual-src blend type
454 - pan/bi: Handle different sizes of LD_TILE
455 - pan/bi: Add single-component 8-bit mkvec lowering
456 - pan/bi: Handle swizzles in i2i8
457 - pan/bi: Lower 8-bit fragment input
482 - pan/bi: Add imm_uintN helper
483 - pan/bi: Handle integer min/max ourselves
484 - pan/bi: Handle ineg
485 - pan/bi: Handle b2f ourselves
486 - pan/bi: Handle b2i8/16
487 - pan/bi: Track scalarness of 16-bit ALU
488 - pan/bi: Don't swizzle scalars
489 - pan/bi: Switch to 1-bit bools
490 - pan/bi: Use nir_lower_to_bit_size
497 - pan/bi: Change swizzled scalars to identity
498 - pan/bi: Adapt branching for 1-bit bools
499 - pan/bi: Handle make_vec with 1-bit bools
500 - pan/bi: Temporarily switch back to 0/~0 bools
501 - pan/bi: Enable NIR vectorization
502 - pan/bi: Fix int<-->float size converts
503 - pan/bi: Copyprop constants
504 - pan/bi: Garbage collect bifrost_nir.h
505 - pan/bi: Enable mediump BLEND lowering
511 - pan/bi: Drop load_sampler_lod_parameters_pan
512 - pan/bi: Map load_subgroup_invocation to FAU
513 - pan/bi: Add "lanes per warp" accessor
514 - pan/bi: Add divergent intrinsic lowering pass
534 - pan/bi: Add first_vertex to vertex ID
725 - pan/bi: Pull out bi_count_write_registers
726 - pan/bi: Use TEXS_2D for rect textures
727 - pan/bi: Simplify TEXC codegen for sr_count=0
728 - pan/bi: Fix bi_rewrite_passthrough ordering
729 - pan/bi: Bundle after RA
730 - pan/bi: Add post-RA optimizer
731 - pan/bi: Track liveness while scheduling
732 - pan/bi: Allow IADD.u32 on FMA as \*IADDC
733 - pan/bi: Use explicit affinities in RA
734 - pan/bi: Inline spilling in RA
735 - pan/bi: Explicit zero reg_live_{in, out} when needed
736 - pan/bi: Model interference with preloaded regs
737 - pan/bi: Allow move/sink in blend shaders
738 - pan/bi: Don't restrict the register file in non-blend shaders
739 - pan/bi: Model +BLEND clobbering of r48
740 - pan/bi: Handle images in vertex shaders
741 - pan/bi: Lower loads with component > 0
742 - pan/bi: Lower stores with component != 0
743 - pan/bi: Lower 64-bit ints again
744 - pan/bi: Emit a dummy ATEST if needed
745 - pan/bi: Simplify spill code
746 - pan/bi: Track words instead of bytes in RA
747 - pan/bi: Don't allocate past the end of the reg file
761 - pan/bi: Force u32 for flat varyings
828 - pan/bi: Include modifier info in opcode table
829 - pan/bi: Move bi_word_node to common code
830 - pan/bi: Move typesize to common code
831 - pan/bi: Track instruction size in opcode table
832 - pan/bi: Handle fsat_signed and fclamp_pos
833 - pan/bi: Report tuples, not nops, in shader-db
834 - pan/bi: Propagate fabs/neg/sat
835 - pan/bi: Add back custom algebraic opts
836 - pan/bi: Fuse fclamp_pos and fsat_signed
837 - pan/bi: Schedule FCMP.v2f16 with abs modifier
838 - pan/bi: Fuse abs into FCMP/FMIN/FMAX.v2f16
927 - pan/bi: Improve clause printing
928 - pan/bi: Fix skip/lod_mode aliasing with VAR_TEX
929 - pan/bi: Add bi_foreach_instr_global_rev_safe helper
930 - pan/bi: Pack staging_barrier for the -next- clause
931 - pan/bi: Try to hit full occupancy on v7
932 - pan/bi: Only spill nodes that could progress in RA
933 - pan/bi: Report cycle counts
934 - pan/bi: Track LOD mode even for TEXC
935 - pan/bi: Analyze helper invocations
936 - pan/bi: Fuse LD_VAR+TEXS_2D -> VAR_TEX
937 - pan/bi: Add a constant subexpression elimination pass
938 - pan/bi: Workaround widen restrictions on +FADD.f32
939 - pan/bi: Simplify cube map descriptor generation
940 - pan/bi: Comment the fexp2 implementation
941 - pan/bi: Factor out exp2/log2 code
942 - pan/bi: Don't lower fpow
998 - pan/bi: Refuse to CSE non-SSA sources
999 - pan/bi: Make bi_foreach_instr_in_tuple safer
1000 - pan/bi: Update ins->link after scheduling
1001 - pan/bi: Do helper termination analysis on clauses
1002 - pan/bi: Handle multiple destinations in scheduler
1003 - pan/bi: Add bi_before_tuple convenience method
1004 - pan/bi: Handle 4-src instructions in scheduler
1005 - pan/bi: Calculate dependency graph when bundling
1006 - pan/bi: Add a bundling heuristic
1013 - pan/bi: Restrict swizzles on same cycle temporaries
1014 - pan/bi: Remove incorrect errata workaround
1175 - pan/bi: Don't set the EOS flag if there's at least one successor
1179 - pan/bi: Expand pseudo instructions when nosched is set
1207 - pan/bi: Add support for gl_{BaseVertex,BaseInstance}
1208 - pan/bi: Add support for gl_DrawID
2744 - pan/bi: Skip nir_opt_move/sink for blend shaders
2752 - pan/bi: Add "lane_dest" modifier
2753 - pan/bi: Replace lane0 modifier with lane_dest for load instructions
2756 - pan/bi: Add two tuples to a clause when needed with NOSCHED
2769 - pan/bi: Create a mask of UBOs that need to be uploaded
2777 - pan/bi: Create a nop clause when the shader starts with ATEST
3645 - android: pan/bi: add bi_opt_constant_fold.c to Makefile.sources

Indexes created Sun Dec 14 06:09:32 GMT 2025