Functions — github.com/anthropics/claudes-c-compiler

hub / github.com/anthropics/claudes-c-compiler / functions

Functions6,057 in github.com/anthropics/claudes-c-compiler

⨍ Functions 6,057◇ Types & classes 346

↓ 4 callersMethodassign_scratch_reg

(&mut self, kind: &AsmOperandKind, excluded: &[String])

src/backend/arm/codegen/asm_emitter.rs:149

↓ 4 callersMethodassign_scratch_reg

(&mut self, kind: &AsmOperandKind, excluded: &[String])

src/backend/riscv/codegen/asm_emitter.rs:181

↓ 4 callersMethodbuild_asm_extra_args

Build extra assembler arguments for RISC-V ABI/arch overrides. When -mabi= or -march= are specified on the CLI, these override the defaults hardcoded

src/driver/external_tools.rs:270

↓ 4 callersFunctionbuild_jump_table

Build a jump table from switch cases: maps each index in [min..max] to a BlockId. Returns (table, min_val, range).

src/backend/traits.rs:1419

↓ 4 callersFunctioncalculate_stack_space_common

Shared stack space calculation: iterates over all instructions, assigns stack slots for allocas and value results. Arch-specific offset direction is h

src/backend/stack_layout/mod.rs:119

↓ 4 callersFunctioncc_from_mnemonic

Map condition code suffix to encoding.

src/backend/x86/assembler/encoder/registers.rs:276

↓ 4 callersFunctionclassify_params_full

Classify all parameters of a function for callee-side store emission. Uses the same `CallAbiConfig` as `classify_call_args` to ensure caller and call

src/backend/call_abi.rs:645

↓ 4 callersMethodcollect_refs_from_compound

Collect function name references from a compound statement.

src/ir/lowering/ref_collection.rs:119

↓ 4 callersMethodcollect_scope_cleanup_vars_above_depth

Collect cleanup variables from scopes above `target_depth` (for break/continue). This collects from the innermost scope down to (but not including) th

src/ir/lowering/lower.rs:412

↓ 4 callersFunctioncompute_live_intervals

Compute live intervals for all non-alloca values in a function. Uses backward dataflow analysis to correctly handle loops: - live_in[B] = gen[B] ∪ (l

src/backend/liveness.rs:153

↓ 4 callersFunctioncompute_unsigned_magic_64

Compute the magic number for unsigned 64-bit division by `d`. Returns (magic_number, post_shift, needs_add) where: - If !needs_add: result = mulhi(x,

src/passes/div_by_const.rs:865

↓ 4 callersMethodconfigure_preprocessor

Configure the preprocessor with CLI-defined macros and target.

src/driver/pipeline.rs:777

↓ 4 callersFunctionconstraint_is_immediate_only

Shared inline assembly constraint classification utilities. Used by the inline pass (symbol resolution), cfg_simplify (dead block detection), and the

src/common/asm_constraints.rs:14

↓ 4 callersFunctionconstraint_is_memory_only

Check whether a constraint is memory-only (has memory alternative but no register alternative). For constraints like "rm", "qm", "g" that allow both r

src/backend/inline_asm.rs:273

↓ 4 callersMethodconvert_struct_fields

Convert struct field declarations to StructField for layout computation.

src/frontend/sema/const_eval.rs:1057

↓ 4 callersMethodctype_align

(&self, ctype: &CType)

src/ir/lowering/pointer_analysis.rs:256

↓ 4 callersMethodcurrent_section_mut

(&mut self)

src/backend/elf_writer_common.rs:359

↓ 4 callersMethoddeclare

(&mut self, symbol: Symbol)

src/common/symbol_table.rs:45

↓ 4 callersFunctiondecompose_symbol_addend

Decompose a symbol name that may contain an embedded addend. For example, `"cgroup_bpf_enabled_key+144"` -> `("cgroup_bpf_enabled_key", 144)`. If the

src/backend/riscv/assembler/elf_writer.rs:262

↓ 4 callersMethoddrill_designators

Drill through a slice of designators to find the target type and byte offset. This resolves chains like [Field("b"), Index(2), Field("x")] by walking

src/ir/lowering/global_init_bytes.rs:200

↓ 4 callersFunctioneliminate_dead_code

Eliminate dead code in a single function using use-count-based worklist DCE. Algorithm: 1. Build use-counts for all values (single pass) 2. Build a d

src/passes/dce.rs:30

↓ 4 callersMethodemit_add_offset_to_addr_reg_impl

(&mut self, offset: i64)

src/backend/arm/codegen/memory.rs:124

↓ 4 callersMethodemit_add_offset_to_addr_reg_impl

(&mut self, offset: i64)

src/backend/x86/codegen/memory.rs:213

↓ 4 callersMethodemit_add_offset_to_addr_reg_impl

(&mut self, offset: i64)

src/backend/riscv/codegen/memory.rs:113

↓ 4 callersMethodemit_branch_nonzero

(&mut self, label: &str)

src/backend/i686/codegen/emit.rs:1469

↓ 4 callersMethodemit_complex_expr_to_offset

Emit a complex expression to a memory location at the given offset. Handles integer-to-complex conversion properly by using lower_expr_to_complex and

src/ir/lowering/stmt.rs:591

↓ 4 callersMethodemit_copy_value

emit_copy_value: handles F128, wide (F64/I64/U64), and 32-bit copies.

src/backend/i686/codegen/emit.rs:1680

↓ 4 callersMethodemit_data_integer

Emit a plain integer value for .byte (size=1), .short (size=2), .long (size=4) or .quad (size=8).

src/backend/elf/writer_base.rs:570

↓ 4 callersMethodemit_data_values

(&mut self, vals: &[DataValue], size: usize)

src/backend/elf_writer_common.rs:664

↓ 4 callersMethodemit_entry_alloca

Emit an alloca into the entry block buffer. Used for local variable declarations so that variables whose declarations are skipped by `goto` still have

src/ir/lowering/lower.rs:1096

↓ 4 callersMethodemit_expr_to_compound

Emit a single expression as a compound element. When `coerce_ty` is Some, tries const eval with type coercion first (for scalar fields with known typ

src/ir/lowering/global_init_compound.rs:300

↓ 4 callersFunctionemit_inline_asm_common

Shared inline assembly emission logic. All four backends call this from their `emit_inline_asm` implementation, providing an `InlineAsmEmitter` to han

src/backend/inline_asm.rs:403

↓ 4 callersMethodemit_instr_imm_mem

(&mut self, mnemonic: &str, imm: i64, offset: i64, base: &str)

src/backend/common.rs:1007

↓ 4 callersFunctionemit_int_data

Emit an integer constant at the width specified by `ty`. Truncates or sign-extends `val` (an i64) as needed to match the target width.

src/backend/common.rs:1650

↓ 4 callersMethodemit_save_acc

(&mut self)

src/backend/arm/codegen/emit.rs:1827

↓ 4 callersMethodemit_segment_prefix

Encode ModR/M + SIB + displacement for a memory operand. For i686 REL relocations, the relocation offset must point to the displacement field (where

src/backend/i686/assembler/encoder/core.rs:31

↓ 4 callersMethodemit_struct_field_init_compound

Emit a single struct field initialization into the byte buffer and ptr_ranges. Handles pointer fields, pointer array fields, bitfields, nested structs

src/ir/lowering/global_init_compound_ptrs.rs:279

↓ 4 callersMethodemit_ternary_branch

Shared helper for ternary branch patterns (conditional and GNU conditional). Evaluates `then_fn` in the true branch and `else_fn` in the false branch,

src/ir/lowering/expr_ops.rs:623

↓ 4 callersMethodencode_avx_3op_0f_imm8

Encode AVX 3-operand in 0F map (mm=1) with imm8 (vshufps, vshufpd, etc.)

src/backend/x86/assembler/encoder/avx.rs:305

↓ 4 callersMethodencode_avx_mov_np

Encode AVX vmovaps/vmovapd/vmovups/vmovupd (no mandatory prefix, or 66 prefix)

src/backend/x86/assembler/encoder/avx.rs:190

↓ 4 callersMethodencode_bit_scan

(&mut self, ops: &[Operand], mnemonic: &str, opcode2: u8)

src/backend/x86/assembler/encoder/x87_misc.rs:182

↓ 4 callersMethodencode_crc32

Encode CRC32 instruction. crc32b: F2 0F 38 F0 /r (8-bit source) crc32w: 66 F2 0F 38 F1 /r (16-bit source) crc32l: F2 0F 38 F1 /r (32-bit source) crc32

src/backend/x86/assembler/encoder/sse.rs:539

↓ 4 callersMethodencode_evex_rotate_imm

Encode EVEX rotate-by-immediate instructions (vprold, vprolq, vprord, vprorq). AT&T syntax: `vprold $imm8, %src, %dst` ops[0] = imm8 (rotation count)

src/backend/x86/assembler/encoder/avx.rs:115

↓ 4 callersFunctionencode_fp_arith_d

(operands: &[Operand], funct7: u32)

src/backend/riscv/assembler/encoder/float.rs:77

↓ 4 callersMethodencode_imul

(&mut self, ops: &[Operand], size: u8)

src/backend/x86/assembler/encoder/gp_integer.rs:641

↓ 4 callersFunctionencode_instruction

Encode an AArch64 instruction from its mnemonic and parsed operands.

src/backend/arm/assembler/encoder/mod.rs:192

↓ 4 callersFunctionencode_ldtr_sized

Encode LDTR/STTR with explicit size (for ldtrh, ldtrb, etc.)

src/backend/arm/assembler/encoder/load_store.rs:292

↓ 4 callersFunctionencode_ldur_stur

Encode LDUR/STUR (unscaled immediate offset load/store) Format: size 111 V 00 opc 0 imm9 00 Rn Rt

src/backend/arm/assembler/encoder/load_store.rs:248

↓ 4 callersFunctionencode_ldxp_stxp

Encode LDXP/STXP/LDAXP/STLXP (exclusive pair) instructions. LDXP Xt1, Xt2, [Xn] : sz 001000 0 1 1 11111 0 Rt2 Rn Rt LDAXP Xt1, Xt2, [Xn] : sz 0010

src/backend/arm/assembler/encoder/load_store.rs:604

↓ 4 callersMethodencode_mov

(&mut self, ops: &[Operand], size: u8)

src/backend/i686/assembler/encoder/gp_integer.rs:11

↓ 4 callersMethodencode_mov_seg

(&mut self, ops: &[Operand])

src/backend/x86/assembler/encoder/x87_misc.rs:211

↓ 4 callersFunctionencode_neon_across

Encode NEON across-vector instructions: UMAXV, UMINV, SMAXV, SMINV Format: 0 Q U 01110 size 11000 opcode 10 Rn Rd `u_bit`: 0 for signed, 1 for unsig

src/backend/arm/assembler/encoder/neon.rs:445

↓ 4 callersFunctionencode_neon_aes

Encode NEON AES instructions (AESE, AESD, AESMC, AESIMC)

src/backend/arm/assembler/encoder/neon.rs:1146

↓ 4 callersFunctionencode_neon_shll

── NEON SSHLL/USHLL (shift left long) ─────────────────────────────────── Format: 0 Q U 011110 immh immb 10100 1 Rn Rd

src/backend/arm/assembler/encoder/neon.rs:1472

↓ 4 callersFunctionencode_neon_shrn

── NEON shift right narrow (SHRN/RSHRN) ───────────────────────────────── Format: 0 Q 0 01111 0 immh immb opcode 1 Rn Rd SHRN opcode=10000, RSHRN opco

src/backend/arm/assembler/encoder/neon.rs:1436

↓ 4 callersFunctionencode_neon_sqshrun

Encode NEON SQSHRUN/SQSHRUN2: Signed saturating shift right unsigned narrow Format: 0 Q 1 011110 immh immb 100011 Rn Rd

src/backend/arm/assembler/encoder/neon.rs:119

↓ 4 callersFunctionencode_neon_xtl

Encode NEON UXTL/SXTL (unsigned/signed extend long). These are aliases for USHLL/SSHLL with shift #0. Format: 0 Q U 011110 immh immb 10100 1 Rn Rd

src/backend/arm/assembler/encoder/neon.rs:163

↓ 4 callersMethodencode_prefetch

Encode prefetch instructions (0F 18 /hint)

src/backend/i686/assembler/encoder/system.rs:11

↓ 4 callersFunctionencode_shift

(operands: &[Operand], shift_type: u32)

src/backend/arm/assembler/encoder/data_processing.rs:796

↓ 4 callersMethodencode_sse_cvt_gp_to_xmm

(&mut self, ops: &[Operand], opcode: &[u8], gp_size: u8)

src/backend/x86/assembler/encoder/sse.rs:282

↓ 4 callersMethodencode_sse_extract

Encode SSE/SSE4.1 extract instructions (pextrw, pextrd, pextrb, pextrq). `swap_reg_rm`: if true, the GP register goes in the ModRM "reg" field and the

src/backend/x86/assembler/encoder/sse.rs:416

↓ 4 callersMethodencode_sse_insert

(&mut self, ops: &[Operand], opcode: &[u8], rex_w: bool)

src/backend/x86/assembler/encoder/sse.rs:374

↓ 4 callersFunctionencode_store

(operands: &[Operand], funct3: u32)

src/backend/riscv/assembler/encoder/base.rs:194

↓ 4 callersMethodencode_suffixless_unary

(&mut self, ops: &[Operand], op_ext: u8)

src/backend/x86/assembler/encoder/x87_misc.rs:373

↓ 4 callersMethodencode_x87_arith_reg

Encode x87 register-register arithmetic (fadd, fmul, fsub, fdiv). In AT&T syntax: fadd %st(i), %st -> D8 (base_st0 + i) -- st(0) = st(0) op st(i)

src/backend/x86/assembler/encoder/avx.rs:1026

↓ 4 callersMethodencode_x87_arith_reg

Encode x87 register-register arithmetic (fadd/fmul/fsub/fdiv with st(i) operands).

src/backend/i686/assembler/encoder/x87.rs:230

↓ 4 callersMethodends_with_funclike_macro

Check if a line ends with an identifier that is a defined function-like macro. This is used to detect cases where the macro arguments '(' might be on

src/frontend/preprocessor/pipeline.rs:577

↓ 4 callersMethodensure_section

(&mut self)

src/backend/elf_writer_common.rs:607

↓ 4 callersMethodentry_addr

Get the address of a GOT entry for a given symbol key.

src/backend/arm/linker/reloc.rs:52

↓ 4 callersMethodeval_const_expr_as_usize

(&self, expr: &Expr)

src/frontend/sema/analysis.rs:2012

↓ 4 callersMethodeval_string_literal_with_offset

Helper: intern a string literal + offset => GlobalAddr or GlobalAddrOffset.

src/ir/lowering/global_init.rs:1102

↓ 4 callersFunctioneval_unary

(tokens: &[ExprToken], pos: &mut usize)

src/backend/asm_expr.rs:200

↓ 4 callersFunctionexpand_sdiv64

Expand `dest = x /s C` for native 64-bit signed x using 128-bit multiply.

src/passes/div_by_const.rs:1097

↓ 4 callersFunctionexpand_srem64

Expand `dest = x %s C` for 64-bit signed x.

src/passes/div_by_const.rs:1266

↓ 4 callersMethodexpand_text

Recursively expand macros in text, tracking which macros are currently being expanded to prevent infinite recursion. Operates on bytes for performanc

src/frontend/preprocessor/macro_defs.rs:362

↓ 4 callersFunctionextract_jump_target

Extract the jump target label from a jump/branch instruction.

src/backend/x86/codegen/peephole/passes/helpers.rs:202

↓ 4 callersMethodextract_symbol_from_global_addr_expr

Try to resolve an expression to a global symbol name with optional offset, using the global address expression evaluator. Returns strings like `"boot_

src/ir/lowering/stmt_asm.rs:496

↓ 4 callersFunctionextract_trailing_ident

Extract the trailing identifier from a string, if it ends with one. Returns Some(ident) if the string ends with an identifier (skipping trailing white

src/frontend/preprocessor/macro_defs.rs:1175

↓ 4 callersMethodf128_is_alloca

(&self, val_id: u32)

src/backend/arm/codegen/f128.rs:310

↓ 4 callersMethodf128_load_indirect_ptr_to_addr_reg

(&mut self, slot: StackSlot, _val_id: u32)

src/backend/arm/codegen/f128.rs:227

↓ 4 callersMethodf128_move_aligned_to_addr_reg

Move the computed aligned address into the address register. ARM needs `mov x17, x9` because x9 is the alloca addr register and x17 is the F128 addr r

src/backend/f128_softfloat.rs:201

↓ 4 callersMethodf128_move_callee_reg_to_addr_reg

(&mut self, val_id: u32)

src/backend/arm/codegen/f128.rs:212

↓ 4 callersMethodf128_set_dyn_alloca

(&mut self, val: bool)

src/backend/arm/codegen/f128.rs:206

↓ 4 callersFunctionf128_to_abs_uint

Core f128-to-unsigned-integer conversion. Extracts the absolute value of an f128 float as u128, with truncation toward zero. `max_bits` limits the out

src/common/long_double.rs:1054

↓ 4 callersFunctionf64_to_f128_bytes

Convert an f64 value to IEEE 754 binary128 (quad-precision) encoding (16 bytes, little-endian). Quad format: 1 sign bit, 15 exponent bits (bias 16383)

src/ir/constants.rs:48

↓ 4 callersFunctionf64_to_f128_bytes_lossless

Convert f64 to f128 bytes (widening, zero-fills extra mantissa bits).

src/common/long_double.rs:1028

↓ 4 callersMethodfill_multidim_array_field

Fill a multi-dimensional array field from a braced initializer list. Handles arrays like `int a[2][3]`, `struct S a[2][2][2]`, `union U a[3][4]`, etc.

src/ir/lowering/global_init_bytes.rs:617

↓ 4 callersFunctionfilter_available_regs

Filter a callee-saved register list by removing ASM-clobbered entries. Returns the filtered list suitable for passing to `run_regalloc_and_merge_clobb

src/backend/stack_layout/regalloc_helpers.rs:54

↓ 4 callersFunctionfind_gp_scratch_for_output

Find a GP scratch register for storing inline asm output values that doesn't conflict with any output register or the current output being stored. Us

src/backend/riscv/codegen/asm_emitter.rs:50

↓ 4 callersMethodfinish

(&self)

src/common/fx_hash.rs:78

↓ 4 callersMethodfinish_int_literal

Common integer literal finish: parse suffix, return token.

src/frontend/lexer/scan.rs:319

↓ 4 callersMethodflatten_global_init_item_bool

Flatten a single initializer item with _Bool awareness.

src/ir/lowering/global_init.rs:1906

↓ 4 callersMethodfloat_operand_to_xmm0

Load a float operand into %xmm0. Handles both Value operands (from stack) and float constants (loaded via their bit pattern into rax first).

src/backend/x86/codegen/intrinsics.rs:25

↓ 4 callersMethodfor_each_set_bit

Iterate over all set bits, calling f(bit_index) for each.

src/backend/liveness.rs:109

↓ 4 callersFunctiongcc_cc_to_x86

Map GCC inline asm condition code suffix to x86 SETcc/Jcc suffix. GCC's `=@cc<cond>` output constraint maps directly to x86 condition codes. This map

src/backend/x86_common.rs:63

↓ 4 callersFunctiongenerate_module_with_debug

Generate assembly for a module using the given architecture's codegen. Generate assembly for an IR module with debug info support. Sets `debug_info` o

src/backend/generation.rs:488

↓ 4 callersFunctionget_symbol

(operands: &[Operand], idx: usize)

src/backend/riscv/assembler/encoder/mod.rs:394

↓ 4 callersMethodget_va_args

Get variadic arguments (__VA_ARGS__) as a comma-separated string.

src/frontend/preprocessor/macro_defs.rs:1121

↓ 4 callersFunctiongot_key

Build a GOT key for a symbol reference in a relocation. Local symbols must be scoped to their object to avoid collisions (e.g., `.LANCHOR3` in differe

src/backend/arm/linker/reloc.rs:95

↓ 4 callersFunctiongsf_invalidate_all

Invalidate all slot mappings.

src/backend/riscv/codegen/peephole.rs:704

← previous next →801–900 of 6,057, ranked by callers