A GameCube/Wii emulator and debugger written in Rust.
asdasdasdasd
Layleae29f4afe1ef
diff --git a/crates/gecko/src/flipper/gx/jit/attr.rs b/crates/gecko/src/flipper/gx/jit/attr.rsindex 9a77284..00c0d26 100644--- a/crates/gecko/src/flipper/gx/jit/attr.rs+++ b/crates/gecko/src/flipper/gx/jit/attr.rs@@ -2,7 +2,7 @@ use cranelift_codegen::ir::{self, InstBuilder};use cranelift_frontend::FunctionBuilder;use super::VtxKey;-use super::builder::{MEMFLAGS, MEMFLAGS_RO, array_offset, offset, xf_byte_off};+use super::builder::{MEMFLAGS, MEMFLAGS_RO, MEMFLAGS_RO_MOVABLE, array_offset, offset, xf_byte_off};use crate::flipper::gx::constants::*;use crate::flipper::gx::regs::{AttributeType, ColorCount, ColorFormat, ComponentFormat, NrmCount, PosCount, TexCount};@@ -113,7 +113,7 @@ fn read_pos_mtx_idx(ctx: &mut AttrCtx, vcd_lo: crate::flipper::gx::regs::VcdLo)// xf_mem[XF_MATRIX_INDEX_A].pos_mtx_idx() -> bits 0..=5let cell = ctx.bd.ins().load(ir::types::I32,- MEMFLAGS_RO,+ MEMFLAGS_RO_MOVABLE,ctx.xf_mem_ptr,xf_byte_off(XF_MATRIX_INDEX_A),);@@ -136,13 +136,13 @@ fn read_tex_mtx_indices(ctx: &mut AttrCtx, vcd_lo: crate::flipper::gx::regs::Vcd// starting at bit 6) and XF_MATRIX_INDEX_B (tex4..tex7).let cell_a = ctx.bd.ins().load(ir::types::I32,- MEMFLAGS_RO,+ MEMFLAGS_RO_MOVABLE,ctx.xf_mem_ptr,xf_byte_off(XF_MATRIX_INDEX_A),);let cell_b = ctx.bd.ins().load(ir::types::I32,- MEMFLAGS_RO,+ MEMFLAGS_RO_MOVABLE,ctx.xf_mem_ptr,xf_byte_off(XF_MATRIX_INDEX_B),);@@ -546,7 +546,7 @@ fn decode_texcoord(fn xf_transform_3x4(ctx: &mut AttrCtx, pos_mtx_idx: ir::Value, pos: &[ir::Value; 3]) -> [ir::Value; 3] {let base_addr = self::pos_mtx_base_addr(ctx, pos_mtx_idx);- let m = self::load_matrix12(ctx, base_addr);+ let m = self::load_matrix12(ctx, base_addr, MEMFLAGS_RO_MOVABLE);let r0 = self::dot3_affine(ctx, [m[0], m[1], m[2], m[3]], *pos);let r1 = self::dot3_affine(ctx, [m[4], m[5], m[6], m[7]], *pos);@@ -562,9 +562,12 @@ fn transform_and_normalize_normal(ctx: &mut AttrCtx, pos_mtx_idx: ir::Value, nrmlet base = ctx.bd.ins().iadd(ctx.xf_mem_ptr, cell_off);let nm = std::array::from_fn::<ir::Value, 9, _>(|i| {- ctx.bd- .ins()- .load(ir::types::F32, MEMFLAGS_RO, base, (XF_NRM_MTX_BASE * 4 + i * 4) as i32)+ ctx.bd.ins().load(+ ir::types::F32,+ MEMFLAGS_RO_MOVABLE,+ base,+ (XF_NRM_MTX_BASE * 4 + i * 4) as i32,+ )});let row = |i: usize, ctx: &mut AttrCtx| -> ir::Value {@@ -611,14 +614,16 @@ fn emit_texgens(tex_mtx_idx: &[ir::Value; 8],raw_st: &[Option<[ir::Value; 2]>; 8],) {- let num_texgens = ctx- .bd- .ins()- .load(ir::types::I32, MEMFLAGS_RO, ctx.xf_mem_ptr, xf_byte_off(XF_NUM_TEXGENS));+ let num_texgens = ctx.bd.ins().load(+ ir::types::I32,+ MEMFLAGS_RO_MOVABLE,+ ctx.xf_mem_ptr,+ xf_byte_off(XF_NUM_TEXGENS),+ );let dual_cell = ctx.bd.ins().load(ir::types::I32,- MEMFLAGS_RO,+ MEMFLAGS_RO_MOVABLE,ctx.xf_mem_ptr,xf_byte_off(XF_DUAL_TEX_ENABLE),);@@ -693,7 +698,7 @@ fn emit_texgen_active(let tg_cell = ctx.bd.ins().load(ir::types::I32,- MEMFLAGS_RO,+ MEMFLAGS_RO_MOVABLE,ctx.xf_mem_ptr,xf_byte_off(XF_TEXGEN_BASE + tg),);@@ -708,7 +713,7 @@ fn emit_texgen_active(let input = [src[0], src[1], input2, one];let base_addr = self::pos_mtx_base_addr(ctx, tex_mtx_idx[tg]);- let m = self::load_matrix12(ctx, base_addr);+ let m = self::load_matrix12(ctx, base_addr, MEMFLAGS_RO_MOVABLE);let s = self::dot4(ctx, [m[0], m[1], m[2], m[3]], input);let t = self::dot4(ctx, [m[4], m[5], m[6], m[7]], input);@@ -732,43 +737,74 @@ fn texgen_source(nrm: &[ir::Value; 3],raw_st: &[Option<[ir::Value; 2]>; 8],) -> [ir::Value; 3] {- use ir::condcodes::IntCC;+ use cranelift_frontend::Switch;let zero = ctx.bd.ins().f32const(0.0);let one = ctx.bd.ins().f32const(1.0);- let is0 = ctx.bd.ins().icmp_imm(IntCC::Equal, src_row, 0);- let is1 = ctx.bd.ins().icmp_imm(IntCC::Equal, src_row, 1);- let ge2 = ctx.bd.ins().icmp_imm(IntCC::UnsignedGreaterThanOrEqual, src_row, 2);- let le4 = ctx.bd.ins().icmp_imm(IntCC::UnsignedLessThanOrEqual, src_row, 4);- let is_dead = ctx.bd.ins().band(ge2, le4);+ let merge = ctx.bd.create_block();+ ctx.bd.append_block_param(merge, ir::types::F32);+ ctx.bd.append_block_param(merge, ir::types::F32);+ ctx.bd.append_block_param(merge, ir::types::F32);++ let pos_blk = ctx.bd.create_block();+ let nrm_blk = ctx.bd.create_block();+ let dead_blk = ctx.bd.create_block();+ let tex_blk: [ir::Block; 8] = std::array::from_fn(|k| {+ if raw_st[k].is_some() {+ ctx.bd.create_block()+ } else {+ dead_blk+ }+ });- let mut is_tex = [is0; 8];+ let mut switch = Switch::new();+ switch.set_entry(0, pos_blk);+ switch.set_entry(1, nrm_blk);+ switch.set_entry(2, dead_blk);+ switch.set_entry(3, dead_blk);+ switch.set_entry(4, dead_blk);for k in 0..8 {- is_tex[k] = ctx.bd.ins().icmp_imm(IntCC::Equal, src_row, (5 + k) as i64);+ switch.set_entry((5 + k) as u128, tex_blk[k]);}+ // src_row >= 13 aliases to tex7 (TexGenSrc::from_raw last-variant fallback).+ switch.emit(ctx.bd, src_row, tex_blk[7]);- // one tex component: present ? (i<2 ? s/t : 1) : (i<2 ? 0 : 1)- let tex_comp = |i: usize, k: usize| match raw_st[k] {- Some(st) if i < 2 => st[i],- _ if i < 2 => zero,- _ => one,+ let jump3 = |ctx: &mut AttrCtx, blk: ir::Block, v: [ir::Value; 3]| {+ ctx.bd.ins().jump(+ blk,+ &[+ ir::BlockArg::Value(v[0]),+ ir::BlockArg::Value(v[1]),+ ir::BlockArg::Value(v[2]),+ ],+ );};- let mut out = [zero; 3];- for i in 0..3 {- let mut acc = tex_comp(i, 7); // default: src_row >= 13 aliases to tex7- for k in 0..8 {- let tc = tex_comp(i, k);- acc = ctx.bd.ins().select(is_tex[k], tc, acc);+ ctx.bd.switch_to_block(pos_blk);+ ctx.bd.seal_block(pos_blk);+ jump3(ctx, merge, *pos);++ ctx.bd.switch_to_block(nrm_blk);+ ctx.bd.seal_block(nrm_blk);+ jump3(ctx, merge, *nrm);++ ctx.bd.switch_to_block(dead_blk);+ ctx.bd.seal_block(dead_blk);+ jump3(ctx, merge, [zero, zero, one]);++ for k in 0..8 {+ if let Some(st) = raw_st[k] {+ ctx.bd.switch_to_block(tex_blk[k]);+ ctx.bd.seal_block(tex_blk[k]);+ jump3(ctx, merge, [st[0], st[1], one]);}- let dead_c = if i < 2 { zero } else { one };- acc = ctx.bd.ins().select(is_dead, dead_c, acc);- acc = ctx.bd.ins().select(is1, nrm[i], acc);- acc = ctx.bd.ins().select(is0, pos[i], acc);- out[i] = acc;}- out++ ctx.bd.switch_to_block(merge);+ ctx.bd.seal_block(merge);+ let p = ctx.bd.block_params(merge);+ [p[0], p[1], p[2]]}// Test a single config bit: (word >> shift) & 1 != 0.@@ -787,12 +823,8 @@ fn pos_mtx_base_addr(ctx: &mut AttrCtx, idx: ir::Value) -> ir::Value {}// Load the 12 f32 cells of a 3x4 matrix.-fn load_matrix12(ctx: &mut AttrCtx, base_addr: ir::Value) -> [ir::Value; 12] {- std::array::from_fn(|i| {- ctx.bd- .ins()- .load(ir::types::F32, MEMFLAGS_RO, base_addr, (i * 4) as i32)- })+fn load_matrix12(ctx: &mut AttrCtx, base_addr: ir::Value, flags: ir::MemFlagsData) -> [ir::Value; 12] {+ std::array::from_fn(|i| ctx.bd.ins().load(ir::types::F32, flags, base_addr, (i * 4) as i32))}// (m0*v0 + m1*v1) + m2*v2 + m3*v3, no fma.@@ -870,7 +902,7 @@ fn emit_dual(let post_bytes = ctx.bd.ins().iadd_imm(post_bytes, (XF_POST_MTX_BASE * 4) as i64);let post_bytes = ctx.bd.ins().uextend(ctx.pointer_ty, post_bytes);let post_addr = ctx.bd.ins().iadd(ctx.xf_mem_ptr, post_bytes);- let pm = self::load_matrix12(ctx, post_addr);+ let pm = self::load_matrix12(ctx, post_addr, MEMFLAGS_RO);let nsv = [ns, nt, nq];let ps = self::dot3_affine(ctx, [pm[0], pm[1], pm[2], pm[3]], nsv);diff --git a/crates/gecko/src/flipper/gx/jit/builder.rs b/crates/gecko/src/flipper/gx/jit/builder.rsindex ac8be41..06c3dc8 100644--- a/crates/gecko/src/flipper/gx/jit/builder.rs+++ b/crates/gecko/src/flipper/gx/jit/builder.rs@@ -9,6 +9,7 @@ use crate::host::DrawVertex;pub(crate) const MEMFLAGS: ir::MemFlagsData = ir::MemFlagsData::new().with_notrap();pub(crate) const MEMFLAGS_RO: ir::MemFlagsData = ir::MemFlagsData::new().with_notrap().with_readonly();+pub(crate) const MEMFLAGS_RO_MOVABLE: ir::MemFlagsData = MEMFLAGS_RO.with_can_move();pub fn build_parser(ctx: &mut Context, fn_ctx: &mut FunctionBuilderContext, pointer_ty: ir::Type, key: VtxKey) {let mut bd = FunctionBuilder::new(&mut ctx.func, fn_ctx);