|// Low-level VM code for PowerPC 32 bit or 32on64 bit mode.
|// Bytecode interpreter, fast functions and helper functions.
|// Copyright (C) 2005-2023 Mike Pall. See Copyright Notice in luajit.h
|
|.arch ppc
|.section code_op, code_sub
|
|.actionlist build_actionlist
|.globals GLOB_
|.globalnames globnames
|.externnames extnames
|
|// Note: The ragged indentation of the instructions is intentional.
|//       The starting columns indicate data dependencies.
|
|//-----------------------------------------------------------------------
|
|// DynASM defines used by the PPC port:
|//
|// P64     64 bit pointers (only for GPR64 testing).
|// GPR64   64 bit registers (but possibly 32 bit pointers, e.g. PS3).
|//         Affects reg saves, stack layout, carry/overflow/dot flags etc.
|// FRAME32 Use 32 bit frame layout, even with GPR64 (Xbox 360).
|// TOC     Need table of contents (64 bit or 32 bit variant, e.g. PS3).
|//         Function pointers are really a struct: code, TOC, env (optional).
|// TOCENV  Function pointers have an environment pointer, too (not on PS3).
|// PPE     Power Processor Element of Cell (PS3) or Xenon (Xbox 360).
|//         Must avoid (slow) micro-coded instructions.
|
|.if P64
|.define TOC, 1
|.define TOCENV, 1
|.macro lpx, a, b, c; ldx a, b, c; .endmacro
|.macro lp, a, b; ld a, b; .endmacro
|.macro stp, a, b; std a, b; .endmacro
|.define decode_OPP, decode_OP8
|.if FFI
|// Missing: Calling conventions, 64 bit regs, TOC.
|.error lib_ffi not yet implemented for PPC64
|.endif
|.else
|.macro lpx, a, b, c; lwzx a, b, c; .endmacro
|.macro lp, a, b; lwz a, b; .endmacro
|.macro stp, a, b; stw a, b; .endmacro
|.define decode_OPP, decode_OP4
|.endif
|
|// Convenience macros for TOC handling.
|.if TOC
|// Linker needs a TOC patch area for every external call relocation.
|.macro blex, target; bl extern target@plt; nop; .endmacro
|.macro .toc, a, b; a, b; .endmacro
|.if P64
|.define TOC_OFS,	 8
|.define ENV_OFS,	16
|.else
|.define TOC_OFS,	4
|.define ENV_OFS,	8
|.endif
|.else  // No TOC.
|.macro blex, target; bl extern target@plt; .endmacro
|.macro .toc, a, b; .endmacro
|.endif
|.macro .tocenv, a, b; .if TOCENV; a, b; .endif; .endmacro
|
|.macro .gpr64, a, b; .if GPR64; a, b; .endif; .endmacro
|
|.macro andix., y, a, i
|.if PPE
|  rlwinm y, a, 0, 31-lj_fls(i), 31-lj_ffs(i)
|  cmpwi y, 0
|.else
|  andi. y, a, i
|.endif
|.endmacro
|
|.macro clrso, reg
|.if PPE
|  li reg, 0
|  mtxer reg
|.else
|  mcrxr cr0
|.endif
|.endmacro
|
|.macro checkov, reg, noov
|.if PPE
|  mfxer reg
|  add reg, reg, reg
|  cmpwi reg, 0
|   li reg, 0
|   mtxer reg
|  bgey noov
|.else
|  mcrxr cr0
|  bley noov
|.endif
|.endmacro
|
|//-----------------------------------------------------------------------
|
|// Fixed register assignments for the interpreter.
|// Don't use: r1 = sp, r2 and r13 = reserved (TOC, TLS or SDATA)
|
|.macro .FPU, a, b
|.if FPU
|  a, b
|.endif
|.endmacro
|
|.macro .FPU, a, b, c
|.if FPU
|  a, b, c
|.endif
|.endmacro
|
|// The following must be C callee-save (but BASE is often refetched).
|.define BASE,		r14	// Base of current Lua stack frame.
|.define KBASE,		r15	// Constants of current Lua function.
|.define PC,		r16	// Next PC.
|.define DISPATCH,	r17	// Opcode dispatch table.
|.define LREG,		r18	// Register holding lua_State (also in SAVE_L).
|.define MULTRES,	r19	// Size of multi-result: (nresults+1)*8.
|.define JGL,		r31	// On-trace: global_State + 32768.
|
|// Constants for type-comparisons, stores and conversions. C callee-save.
|.define TISNUM,	r22
|.define TISNIL,	r23
|.define ZERO,		r24
|.if FPU
|.define TOBIT,		f30	// 2^52 + 2^51.
|.define TONUM,		f31	// 2^52 + 2^51 + 2^31.
|.endif
|
|// The following temporaries are not saved across C calls, except for RA.
|.define RA,		r20	// Callee-save.
|.define RB,		r10
|.define RC,		r11
|.define RD,		r12
|.define INS,		r7	// Overlaps CARG5.
|
|.define TMP0,		r0
|.define TMP1,		r8
|.define TMP2,		r9
|.define TMP3,		r6	// Overlaps CARG4.
|
|// Saved temporaries.
|.define SAVE0,		r21
|.define SAVE1,		r25
|
|// Calling conventions.
|.define CARG1,		r3
|.define CARG2,		r4
|.define CARG3,		r5
|.define CARG4,		r6	// Overlaps TMP3.
|.define CARG5,		r7	// Overlaps INS.
|
|.if FPU
|.define FARG1,		f1
|.define FARG2,		f2
|.endif
|
|.define CRET1,		r3
|.define CRET2,		r4
|
|.define TOCREG,	r2	// TOC register (only used by C code).
|.define ENVREG,	r11	// Environment pointer (nested C functions).
|
|// Stack layout while in interpreter. Must match with lj_frame.h.
|.if GPR64
|.if FRAME32
|
|//			456(sp) // \ 32/64 bit C frame info
|.define TONUM_LO,	452(sp) // |
|.define TONUM_HI,	448(sp) // |
|.define TMPD_LO,	444(sp) // |
|.define TMPD_HI,	440(sp) // |
|.define SAVE_CR,	432(sp) // | 64 bit CR save.
|.define SAVE_ERRF,	424(sp) //  > Parameter save area.
|.define SAVE_NRES,	420(sp) // |
|.define SAVE_L,	416(sp) // |
|.define SAVE_PC,	412(sp) // |
|.define SAVE_MULTRES,	408(sp) // |
|.define SAVE_CFRAME,	400(sp) // / 64 bit C frame chain.
|//			392(sp) // Reserved.
|.define CFRAME_SPACE,	384     // Delta for sp.
|// Back chain for sp:	384(sp) <-- sp entering interpreter
|.define SAVE_LR,	376(sp) // 32 bit LR stored in hi-part.
|.define SAVE_GPR_,	232     // .. 232+18*8: 64 bit GPR saves.
|.define SAVE_FPR_,	88      // .. 88+18*8: 64 bit FPR saves.
|//			80(sp) // Needed for 16 byte stack frame alignment.
|//			16(sp)  // Callee parameter save area (ABI mandated).
|//			8(sp)   // Reserved
|// Back chain for sp:	0(sp)   <-- sp while in interpreter
|// 32 bit sp stored in hi-part of 0(sp).
|
|.define TMPD_BLO,	447(sp)
|.define TMPD,		TMPD_HI
|.define TONUM_D,	TONUM_HI
|
|.else
|
|//			508(sp) // \ 32 bit C frame info.
|.define SAVE_ERRF,	472(sp) // |
|.define SAVE_NRES,	468(sp) // |
|.define SAVE_L,	464(sp) //  > Parameter save area.
|.define SAVE_PC,	460(sp) // |
|.define SAVE_MULTRES,	456(sp) // |
|.define SAVE_CFRAME,	448(sp) // / 64 bit C frame chain.
|.define SAVE_LR,	416(sp)
|.define CFRAME_SPACE,	400     // Delta for sp.
|// Back chain for sp:	400(sp) <-- sp entering interpreter
|.define SAVE_FPR_,	256     // .. 256+18*8: 64 bit FPR saves.
|.define SAVE_GPR_,	112     // .. 112+18*8: 64 bit GPR saves.
|//			48(sp)  // Callee parameter save area (ABI mandated).
|.define SAVE_TOC,	40(sp)  // TOC save area.
|.define TMPD_LO,	36(sp)  // \ Link editor temp (ABI mandated).
|.define TMPD_HI,	32(sp)  // /
|.define TONUM_LO,	28(sp)  // \ Compiler temp (ABI mandated).
|.define TONUM_HI,	24(sp)  // /
|// Next frame lr:	16(sp)
|.define SAVE_CR,	8(sp)  // 64 bit CR save.
|// Back chain for sp:	0(sp)	<-- sp while in interpreter
|
|.define TMPD_BLO,	39(sp)
|.define TMPD,		TMPD_HI
|.define TONUM_D,	TONUM_HI
|
|.endif
|.else
|
|.if FPU
|.define SAVE_LR,	276(sp)
|.define CFRAME_SPACE,	272     // Delta for sp.
|// Back chain for sp:	272(sp) <-- sp entering interpreter
|.define SAVE_FPR_,	128     // .. 128+18*8: 64 bit FPR saves.
|.else
|.define SAVE_LR,	132(sp)
|.define CFRAME_SPACE,	128     // Delta for sp.
|// Back chain for sp:	128(sp) <-- sp entering interpreter
|.endif
|.define SAVE_GPR_,	56      // .. 56+18*4: 32 bit GPR saves.
|.define SAVE_CR,	52(sp)  // 32 bit CR save.
|.define SAVE_ERRF,	48(sp)  // 32 bit C frame info.
|.define SAVE_NRES,	44(sp)
|.define SAVE_CFRAME,	40(sp)
|.define SAVE_L,	36(sp)
|.define SAVE_PC,	32(sp)
|.define SAVE_MULTRES,	28(sp)
|.define UNUSED1,	24(sp)
|.if FPU
|.define TMPD_LO,	20(sp)
|.define TMPD_HI,	16(sp)
|.define TONUM_LO,	12(sp)
|.define TONUM_HI,	8(sp)
|.else
|.define SFSAVE_4,	20(sp)
|.define SFSAVE_3,	16(sp)
|.define SFSAVE_2,	12(sp)
|.define SFSAVE_1,	8(sp)
|.endif
|// Next frame lr:	4(sp)
|// Back chain for sp:	0(sp)	<-- sp while in interpreter
|
|.if FPU
|.define TMPD_BLO,	23(sp)
|.define TMPD,		TMPD_HI
|.define TONUM_D,	TONUM_HI
|.endif
|
|.endif
|
|.macro save_, reg
|.if GPR64
|  std r..reg, SAVE_GPR_+(reg-14)*8(sp)
|.else
|  stw r..reg, SAVE_GPR_+(reg-14)*4(sp)
|.endif
|  .FPU stfd f..reg, SAVE_FPR_+(reg-14)*8(sp)
|.endmacro
|.macro rest_, reg
|.if GPR64
|  ld r..reg, SAVE_GPR_+(reg-14)*8(sp)
|.else
|  lwz r..reg, SAVE_GPR_+(reg-14)*4(sp)
|.endif
|  .FPU lfd f..reg, SAVE_FPR_+(reg-14)*8(sp)
|.endmacro
|
|.macro saveregs
|.if GPR64 and not FRAME32
|  stdu sp, -CFRAME_SPACE(sp)
|.else
|  stwu sp, -CFRAME_SPACE(sp)
|.endif
|  save_ 14; save_ 15; save_ 16
|  mflr r0
|  save_ 17; save_ 18; save_ 19; save_ 20; save_ 21; save_ 22
|.if GPR64 and not FRAME32
|  std r0, SAVE_LR
|.else
|  stw r0, SAVE_LR
|.endif
|  save_ 23; save_ 24; save_ 25
|  mfcr r0
|  save_ 26; save_ 27; save_ 28; save_ 29; save_ 30; save_ 31
|.if GPR64
|  std r0, SAVE_CR
|.else
|  stw r0, SAVE_CR
|.endif
|  .toc std TOCREG, SAVE_TOC
|.endmacro
|
|.macro restoreregs
|.if GPR64 and not FRAME32
|  ld r0, SAVE_LR
|.else
|  lwz r0, SAVE_LR
|.endif
|.if GPR64
|  ld r12, SAVE_CR
|.else
|  lwz r12, SAVE_CR
|.endif
|  rest_ 14; rest_ 15; rest_ 16; rest_ 17; rest_ 18; rest_ 19
|  mtlr r0;
|.if PPE; mtocrf 0x20, r12; .else; mtcrf 0x38, r12; .endif
|  rest_ 20; rest_ 21; rest_ 22; rest_ 23; rest_ 24; rest_ 25
|.if PPE; mtocrf 0x10, r12; .endif
|  rest_ 26; rest_ 27; rest_ 28; rest_ 29; rest_ 30; rest_ 31
|.if PPE; mtocrf 0x08, r12; .endif
|  addi sp, sp, CFRAME_SPACE
|.endmacro
|
|// Type definitions. Some of these are only used for documentation.
|.type L,		lua_State,	LREG
|.type GL,		global_State
|.type TVALUE,		TValue
|.type GCOBJ,		GCobj
|.type STR,		GCstr
|.type TAB,		GCtab
|.type LFUNC,		GCfuncL
|.type CFUNC,		GCfuncC
|.type PROTO,		GCproto
|.type UPVAL,		GCupval
|.type NODE,		Node
|.type NARGS8,		int
|.type TRACE,		GCtrace
|.type SBUF,		SBuf
|
|//-----------------------------------------------------------------------
|
|// Trap for not-yet-implemented parts.
|.macro NYI; tw 4, sp, sp; .endmacro
|
|.if FPU
|// int/FP conversions.
|.macro tonum_i, freg, reg
|  xoris reg, reg, 0x8000
|  stw reg, TONUM_LO
|  lfd freg, TONUM_D
|  fsub freg, freg, TONUM
|.endmacro
|
|.macro tonum_u, freg, reg
|  stw reg, TONUM_LO
|  lfd freg, TONUM_D
|  fsub freg, freg, TOBIT
|.endmacro
|
|.macro toint, reg, freg, tmpfreg
|  fctiwz tmpfreg, freg
|  stfd tmpfreg, TMPD
|  lwz reg, TMPD_LO
|.endmacro
|
|.macro toint, reg, freg
|  toint reg, freg, freg
|.endmacro
|.endif
|
|//-----------------------------------------------------------------------
|
|// Access to frame relative to BASE.
|.define FRAME_PC,	-8
|.define FRAME_FUNC,	-4
|
|// Instruction decode.
|.macro decode_OP4, dst, ins; rlwinm dst, ins, 2, 22, 29; .endmacro
|.macro decode_OP8, dst, ins; rlwinm dst, ins, 3, 21, 28; .endmacro
|.macro decode_RA8, dst, ins; rlwinm dst, ins, 27, 21, 28; .endmacro
|.macro decode_RB8, dst, ins; rlwinm dst, ins, 11, 21, 28; .endmacro
|.macro decode_RC8, dst, ins; rlwinm dst, ins, 19, 21, 28; .endmacro
|.macro decode_RD8, dst, ins; rlwinm dst, ins, 19, 13, 28; .endmacro
|
|.macro decode_OP1, dst, ins; rlwinm dst, ins, 0, 24, 31; .endmacro
|.macro decode_RD4, dst, ins; rlwinm dst, ins, 18, 14, 29; .endmacro
|
|// Instruction fetch.
|.macro ins_NEXT1
|  lwz INS, 0(PC)
|   addi PC, PC, 4
|.endmacro
|// Instruction decode+dispatch. Note: optimized for e300!
|.macro ins_NEXT2
|  decode_OPP TMP1, INS
|  lpx TMP0, DISPATCH, TMP1
|  mtctr TMP0
|   decode_RB8 RB, INS
|   decode_RD8 RD, INS
|   decode_RA8 RA, INS
|   decode_RC8 RC, INS
|  bctr
|.endmacro
|.macro ins_NEXT
|  ins_NEXT1
|  ins_NEXT2
|.endmacro
|
|// Instruction footer.
|.if 1
|  // Replicated dispatch. Less unpredictable branches, but higher I-Cache use.
|  .define ins_next, ins_NEXT
|  .define ins_next_, ins_NEXT
|  .define ins_next1, ins_NEXT1
|  .define ins_next2, ins_NEXT2
|.else
|  // Common dispatch. Lower I-Cache use, only one (very) unpredictable branch.
|  // Affects only certain kinds of benchmarks (and only with -j off).
|  .macro ins_next
|    b ->ins_next
|  .endmacro
|  .macro ins_next1
|  .endmacro
|  .macro ins_next2
|    b ->ins_next
|  .endmacro
|  .macro ins_next_
|  ->ins_next:
|    ins_NEXT
|  .endmacro
|.endif
|
|// Call decode and dispatch.
|.macro ins_callt
|  // BASE = new base, RB = LFUNC/CFUNC, RC = nargs*8, FRAME_PC(BASE) = PC
|  lwz PC, LFUNC:RB->pc
|  lwz INS, 0(PC)
|   addi PC, PC, 4
|  decode_OPP TMP1, INS
|   decode_RA8 RA, INS
|  lpx TMP0, DISPATCH, TMP1
|   add RA, RA, BASE
|  mtctr TMP0
|  bctr
|.endmacro
|
|.macro ins_call
|  // BASE = new base, RB = LFUNC/CFUNC, RC = nargs*8, PC = caller PC
|  stw PC, FRAME_PC(BASE)
|  ins_callt
|.endmacro
|
|//-----------------------------------------------------------------------
|
|// Macros to test operand types.
|.macro checknum, reg; cmplw reg, TISNUM; .endmacro
|.macro checknum, cr, reg; cmplw cr, reg, TISNUM; .endmacro
|.macro checkstr, reg; cmpwi reg, LJ_TSTR; .endmacro
|.macro checktab, reg; cmpwi reg, LJ_TTAB; .endmacro
|.macro checkfunc, reg; cmpwi reg, LJ_TFUNC; .endmacro
|.macro checknil, reg; cmpwi reg, LJ_TNIL; .endmacro
|
|.macro branch_RD
|  srwi TMP0, RD, 1
|  addis PC, PC, -(BCBIAS_J*4 >> 16)
|  add PC, PC, TMP0
|.endmacro
|
|// Assumes DISPATCH is relative to GL.
#define DISPATCH_GL(field)	(GG_DISP2G + (int)offsetof(global_State, field))
#define DISPATCH_J(field)	(GG_DISP2J + (int)offsetof(jit_State, field))
|
#define PC2PROTO(field)  ((int)offsetof(GCproto, field)-(int)sizeof(GCproto))
|
|.macro hotcheck, delta, target
|  rlwinm TMP1, PC, 31, 25, 30
|  addi TMP1, TMP1, GG_DISP2HOT
|  lhzx TMP2, DISPATCH, TMP1
|  addic. TMP2, TMP2, -delta
|  sthx TMP2, DISPATCH, TMP1
|  blt target
|.endmacro
|
|.macro hotloop
|  hotcheck HOTCOUNT_LOOP, ->vm_hotloop
|.endmacro
|
|.macro hotcall
|  hotcheck HOTCOUNT_CALL, ->vm_hotcall
|.endmacro
|
|// Set current VM state. Uses TMP0.
|.macro li_vmstate, st; li TMP0, ~LJ_VMST_..st; .endmacro
|.macro st_vmstate; stw TMP0, DISPATCH_GL(vmstate)(DISPATCH); .endmacro
|
|// Move table write barrier back. Overwrites mark and tmp.
|.macro barrierback, tab, mark, tmp
|  lwz tmp, DISPATCH_GL(gc.grayagain)(DISPATCH)
|  // Assumes LJ_GC_BLACK is 0x04.
|   rlwinm mark, mark, 0, 30, 28		// black2gray(tab)
|  stw tab, DISPATCH_GL(gc.grayagain)(DISPATCH)
|   stb mark, tab->marked
|  stw tmp, tab->gclist
|.endmacro
|
|//-----------------------------------------------------------------------

/* Generate subroutines used by opcodes and other parts of the VM. */
/* The .code_sub section should be last to help static branch prediction. */
static void build_subroutines(BuildCtx *ctx)
{
  |.code_sub
  |
  |//-----------------------------------------------------------------------
  |//-- Return handling ----------------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |->vm_returnp:
  |  // See vm_return. Also: TMP2 = previous base.
  |  andix. TMP0, PC, FRAME_P
  |   li TMP1, LJ_TTRUE
  |  beq ->cont_dispatch
  |
  |  // Return from pcall or xpcall fast func.
  |  lwz PC, FRAME_PC(TMP2)		// Fetch PC of previous frame.
  |  mr BASE, TMP2			// Restore caller base.
  |  // Prepending may overwrite the pcall frame, so do it at the end.
  |   stwu TMP1, FRAME_PC(RA)		// Prepend true to results.
  |
  |->vm_returnc:
  |  addi RD, RD, 8			// RD = (nresults+1)*8.
  |   andix. TMP0, PC, FRAME_TYPE
  |  cmpwi cr1, RD, 0
  |  li CRET1, LUA_YIELD
  |  beq cr1, ->vm_unwind_c_eh
  |  mr MULTRES, RD
  |   beq ->BC_RET_Z			// Handle regular return to Lua.
  |
  |->vm_return:
  |  // BASE = base, RA = resultptr, RD/MULTRES = (nresults+1)*8, PC = return
  |  // TMP0 = PC & FRAME_TYPE
  |  cmpwi TMP0, FRAME_C
  |   rlwinm TMP2, PC, 0, 0, 28
  |    li_vmstate C
  |   sub TMP2, BASE, TMP2		// TMP2 = previous base.
  |  bney ->vm_returnp
  |
  |  addic. TMP1, RD, -8
  |   stp TMP2, L->base
  |   lwz TMP2, SAVE_NRES
  |    subi BASE, BASE, 8
  |    st_vmstate
  |   slwi TMP2, TMP2, 3
  |  beq >2
  |1:
  |  addic. TMP1, TMP1, -8
  |.if FPU
  |   lfd f0, 0(RA)
  |.else
  |   lwz CARG1, 0(RA)
  |   lwz CARG2, 4(RA)
  |.endif
  |    addi RA, RA, 8
  |.if FPU
  |   stfd f0, 0(BASE)
  |.else
  |   stw CARG1, 0(BASE)
  |   stw CARG2, 4(BASE)
  |.endif
  |    addi BASE, BASE, 8
  |  bney <1
  |
  |2:
  |  cmpw TMP2, RD			// More/less results wanted?
  |  bne >6
  |3:
  |  stp BASE, L->top			// Store new top.
  |
  |->vm_leave_cp:
  |  lp TMP0, SAVE_CFRAME		// Restore previous C frame.
  |   li CRET1, 0			// Ok return status for vm_pcall.
  |  stp TMP0, L->cframe
  |
  |->vm_leave_unw:
  |  restoreregs
  |  blr
  |
  |6:
  |  ble >7				// Less results wanted?
  |  // More results wanted. Check stack size and fill up results with nil.
  |  lwz TMP1, L->maxstack
  |  cmplw BASE, TMP1
  |  bge >8
  |  stw TISNIL, 0(BASE)
  |  addi RD, RD, 8
  |  addi BASE, BASE, 8
  |  b <2
  |
  |7:  // Less results wanted.
  |  subfic TMP3, TMP2, 0		// LUA_MULTRET+1 case?
  |   sub TMP0, RD, TMP2
  |  subfe TMP1, TMP1, TMP1		// TMP1 = TMP2 == 0 ? 0 : -1
  |   and TMP0, TMP0, TMP1
  |  sub BASE, BASE, TMP0		// Either keep top or shrink it.
  |  b <3
  |
  |8:  // Corner case: need to grow stack for filling up results.
  |  // This can happen if:
  |  // - A C function grows the stack (a lot).
  |  // - The GC shrinks the stack in between.
  |  // - A return back from a lua_call() with (high) nresults adjustment.
  |  stp BASE, L->top			// Save current top held in BASE (yes).
  |   mr SAVE0, RD
  |  srwi CARG2, TMP2, 3
  |  mr CARG1, L
  |  bl extern lj_state_growstack	// (lua_State *L, int n)
  |    lwz TMP2, SAVE_NRES
  |   mr RD, SAVE0
  |    slwi TMP2, TMP2, 3
  |  lp BASE, L->top			// Need the (realloced) L->top in BASE.
  |  b <2
  |
  |->vm_unwind_c:			// Unwind C stack, return from vm_pcall.
  |  // (void *cframe, int errcode)
  |  mr sp, CARG1
  |  mr CRET1, CARG2
  |->vm_unwind_c_eh:			// Landing pad for external unwinder.
  |  lwz L, SAVE_L
  |  .toc ld TOCREG, SAVE_TOC
  |   li TMP0, ~LJ_VMST_C
  |  lwz GL:TMP1, L->glref
  |   stw TMP0, GL:TMP1->vmstate
  |  b ->vm_leave_unw
  |
  |->vm_unwind_ff:			// Unwind C stack, return from ff pcall.
  |  // (void *cframe)
  |.if GPR64
  |  rldicr sp, CARG1, 0, 61
  |.else
  |  rlwinm sp, CARG1, 0, 0, 29
  |.endif
  |->vm_unwind_ff_eh:			// Landing pad for external unwinder.
  |  lwz L, SAVE_L
  |  .toc ld TOCREG, SAVE_TOC
  |     li TISNUM, LJ_TISNUM		// Setup type comparison constants.
  |  lp BASE, L->base
  |     .FPU lus TMP3, 0x59c0		// TOBIT = 2^52 + 2^51 (float).
  |   lwz DISPATCH, L->glref		// Setup pointer to dispatch table.
  |     li ZERO, 0
  |     .FPU stw TMP3, TMPD
  |  li TMP1, LJ_TFALSE
  |     .FPU ori TMP3, TMP3, 0x0004	// TONUM = 2^52 + 2^51 + 2^31 (float).
  |     li TISNIL, LJ_TNIL
  |    li_vmstate INTERP
  |     .FPU lfs TOBIT, TMPD
  |  lwz PC, FRAME_PC(BASE)		// Fetch PC of previous frame.
  |  la RA, -8(BASE)			// Results start at BASE-8.
  |     .FPU stw TMP3, TMPD
  |   addi DISPATCH, DISPATCH, GG_G2DISP
  |  stw TMP1, 0(RA)			// Prepend false to error message.
  |  li RD, 16				// 2 results: false + error message.
  |    st_vmstate
  |     .FPU lfs TONUM, TMPD
  |  b ->vm_returnc
  |
  |//-----------------------------------------------------------------------
  |//-- Grow stack for calls -----------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |->vm_growstack_c:			// Grow stack for C function.
  |  li CARG2, LUA_MINSTACK
  |  b >2
  |
  |->vm_growstack_l:			// Grow stack for Lua function.
  |  // BASE = new base, RA = BASE+framesize*8, RC = nargs*8, PC = first PC
  |  add RC, BASE, RC
  |   sub RA, RA, BASE
  |  stp BASE, L->base
  |   addi PC, PC, 4			// Must point after first instruction.
  |  stp RC, L->top
  |   srwi CARG2, RA, 3
  |2:
  |  // L->base = new base, L->top = top
  |   stw PC, SAVE_PC
  |  mr CARG1, L
  |  bl extern lj_state_growstack	// (lua_State *L, int n)
  |  lp BASE, L->base
  |  lp RC, L->top
  |  lwz LFUNC:RB, FRAME_FUNC(BASE)
  |  sub RC, RC, BASE
  |  // BASE = new base, RB = LFUNC/CFUNC, RC = nargs*8, FRAME_PC(BASE) = PC
  |  ins_callt				// Just retry the call.
  |
  |//-----------------------------------------------------------------------
  |//-- Entry points into the assembler VM ---------------------------------
  |//-----------------------------------------------------------------------
  |
  |->vm_resume:				// Setup C frame and resume thread.
  |  // (lua_State *L, TValue *base, int nres1 = 0, ptrdiff_t ef = 0)
  |  saveregs
  |  mr L, CARG1
  |    lwz DISPATCH, L->glref		// Setup pointer to dispatch table.
  |  mr BASE, CARG2
  |    lbz TMP1, L->status
  |   stw L, SAVE_L
  |  li PC, FRAME_CP
  |  addi TMP0, sp, CFRAME_RESUME
  |    addi DISPATCH, DISPATCH, GG_G2DISP
  |   stw CARG3, SAVE_NRES
  |    cmplwi TMP1, 0
  |   stw CARG3, SAVE_ERRF
  |   stp CARG3, SAVE_CFRAME
  |   stw CARG1, SAVE_PC		// Any value outside of bytecode is ok.
  |  stp TMP0, L->cframe
  |    beq >3
  |
  |  // Resume after yield (like a return).
  |  stw L, DISPATCH_GL(cur_L)(DISPATCH)
  |  mr RA, BASE
  |   lp BASE, L->base
  |     li TISNUM, LJ_TISNUM		// Setup type comparison constants.
  |   lp TMP1, L->top
  |  lwz PC, FRAME_PC(BASE)
  |     .FPU lus TMP3, 0x59c0		// TOBIT = 2^52 + 2^51 (float).
  |    stb CARG3, L->status
  |     .FPU stw TMP3, TMPD
  |     .FPU ori TMP3, TMP3, 0x0004	// TONUM = 2^52 + 2^51 + 2^31 (float).
  |     .FPU lfs TOBIT, TMPD
  |   sub RD, TMP1, BASE
  |     .FPU stw TMP3, TMPD
  |     .FPU lus TMP0, 0x4338		// Hiword of 2^52 + 2^51 (double)
  |   addi RD, RD, 8
  |     .FPU stw TMP0, TONUM_HI
  |    li_vmstate INTERP
  |     li ZERO, 0
  |    st_vmstate
  |  andix. TMP0, PC, FRAME_TYPE
  |   mr MULTRES, RD
  |     .FPU lfs TONUM, TMPD
  |     li TISNIL, LJ_TNIL
  |  beq ->BC_RET_Z
  |  b ->vm_return
  |
  |->vm_pcall:				// Setup protected C frame and enter VM.
  |  // (lua_State *L, TValue *base, int nres1, ptrdiff_t ef)
  |  saveregs
  |  li PC, FRAME_CP
  |  stw CARG4, SAVE_ERRF
  |  b >1
  |
  |->vm_call:				// Setup C frame and enter VM.
  |  // (lua_State *L, TValue *base, int nres1)
  |  saveregs
  |  li PC, FRAME_C
  |
  |1:  // Entry point for vm_pcall above (PC = ftype).
  |  lp TMP1, L:CARG1->cframe
  |    mr L, CARG1
  |   stw CARG3, SAVE_NRES
  |    lwz DISPATCH, L->glref		// Setup pointer to dispatch table.
  |   stw CARG1, SAVE_L
  |     mr BASE, CARG2
  |    addi DISPATCH, DISPATCH, GG_G2DISP
  |   stw CARG1, SAVE_PC		// Any value outside of bytecode is ok.
  |  stp TMP1, SAVE_CFRAME
  |  stp sp, L->cframe			// Add our C frame to cframe chain.
  |
  |3:  // Entry point for vm_cpcall/vm_resume (BASE = base, PC = ftype).
  |  stw L, DISPATCH_GL(cur_L)(DISPATCH)
  |  lp TMP2, L->base			// TMP2 = old base (used in vmeta_call).
  |     li TISNUM, LJ_TISNUM		// Setup type comparison constants.
  |   lp TMP1, L->top
  |     .FPU lus TMP3, 0x59c0		// TOBIT = 2^52 + 2^51 (float).
  |  add PC, PC, BASE
  |     .FPU stw TMP3, TMPD
  |     li ZERO, 0
  |     .FPU ori TMP3, TMP3, 0x0004	// TONUM = 2^52 + 2^51 + 2^31 (float).
  |     .FPU lfs TOBIT, TMPD
  |  sub PC, PC, TMP2			// PC = frame delta + frame type
  |     .FPU stw TMP3, TMPD
  |     .FPU lus TMP0, 0x4338		// Hiword of 2^52 + 2^51 (double)
  |   sub NARGS8:RC, TMP1, BASE
  |     .FPU stw TMP0, TONUM_HI
  |    li_vmstate INTERP
  |     .FPU lfs TONUM, TMPD
  |     li TISNIL, LJ_TNIL
  |    st_vmstate
  |
  |->vm_call_dispatch:
  |  // TMP2 = old base, BASE = new base, RC = nargs*8, PC = caller PC
  |  lwz TMP0, FRAME_PC(BASE)
  |   lwz LFUNC:RB, FRAME_FUNC(BASE)
  |  checkfunc TMP0; bne ->vmeta_call
  |
  |->vm_call_dispatch_f:
  |  ins_call
  |  // BASE = new base, RB = func, RC = nargs*8, PC = caller PC
  |
  |->vm_cpcall:				// Setup protected C frame, call C.
  |  // (lua_State *L, lua_CFunction func, void *ud, lua_CPFunction cp)
  |  saveregs
  |  mr L, CARG1
  |   lwz TMP0, L:CARG1->stack
  |  stw CARG1, SAVE_L
  |   lp TMP1, L->top
  |     lwz DISPATCH, L->glref		// Setup pointer to dispatch table.
  |  stw CARG1, SAVE_PC			// Any value outside of bytecode is ok.
  |   sub TMP0, TMP0, TMP1		// Compute -savestack(L, L->top).
  |    lp TMP1, L->cframe
  |     addi DISPATCH, DISPATCH, GG_G2DISP
  |  .toc lp CARG4, 0(CARG4)
  |  li TMP2, 0
  |   stw TMP0, SAVE_NRES		// Neg. delta means cframe w/o frame.
  |  stw TMP2, SAVE_ERRF		// No error function.
  |    stp TMP1, SAVE_CFRAME
  |    stp sp, L->cframe		// Add our C frame to cframe chain.
  |     stw L, DISPATCH_GL(cur_L)(DISPATCH)
  |  mtctr CARG4
  |  bctrl			// (lua_State *L, lua_CFunction func, void *ud)
  |.if PPE
  |  mr BASE, CRET1
  |  cmpwi CRET1, 0
  |.else
  |  mr. BASE, CRET1
  |.endif
  |   li PC, FRAME_CP
  |  bne <3				// Else continue with the call.
  |  b ->vm_leave_cp			// No base? Just remove C frame.
  |
  |//-----------------------------------------------------------------------
  |//-- Metamethod handling ------------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |// The lj_meta_* functions (except for lj_meta_cat) don't reallocate the
  |// stack, so BASE doesn't need to be reloaded across these calls.
  |
  |//-- Continuation dispatch ----------------------------------------------
  |
  |->cont_dispatch:
  |  // BASE = meta base, RA = resultptr, RD = (nresults+1)*8
  |  lwz TMP0, -12(BASE)		// Continuation.
  |   mr RB, BASE
  |   mr BASE, TMP2			// Restore caller BASE.
  |    lwz LFUNC:TMP1, FRAME_FUNC(TMP2)
  |.if FFI
  |  cmplwi TMP0, 1
  |.endif
  |     lwz PC, -16(RB)			// Restore PC from [cont|PC].
  |   subi TMP2, RD, 8
  |   stwx TISNIL, RA, TMP2		// Ensure one valid arg.
  |.if FFI
  |  ble >1
  |.endif
  |    lwz TMP1, LFUNC:TMP1->pc
  |    lwz KBASE, PC2PROTO(k)(TMP1)
  |  // BASE = base, RA = resultptr, RB = meta base
  |  mtctr TMP0
  |  bctr				// Jump to continuation.
  |
  |.if FFI
  |1:
  |  beq ->cont_ffi_callback		// cont = 1: return from FFI callback.
  |  // cont = 0: tailcall from C function.
  |  subi TMP1, RB, 16
  |  sub RC, TMP1, BASE
  |  b ->vm_call_tail
  |.endif
  |
  |->cont_cat:				// RA = resultptr, RB = meta base
  |  lwz INS, -4(PC)
  |   subi CARG2, RB, 16
  |  decode_RB8 SAVE0, INS
  |.if FPU
  |   lfd f0, 0(RA)
  |.else
  |   lwz TMP2, 0(RA)
  |   lwz TMP3, 4(RA)
  |.endif
  |  add TMP1, BASE, SAVE0
  |   stp BASE, L->base
  |  cmplw TMP1, CARG2
  |   sub CARG3, CARG2, TMP1
  |  decode_RA8 RA, INS
  |.if FPU
  |   stfd f0, 0(CARG2)
  |.else
  |   stw TMP2, 0(CARG2)
  |   stw TMP3, 4(CARG2)
  |.endif
  |  bney ->BC_CAT_Z
  |.if FPU
  |   stfdx f0, BASE, RA
  |.else
  |   stwux TMP2, RA, BASE
  |   stw TMP3, 4(RA)
  |.endif
  |  b ->cont_nop
  |
  |//-- Table indexing metamethods -----------------------------------------
  |
  |->vmeta_tgets1:
  |  la CARG3, DISPATCH_GL(tmptv)(DISPATCH)
  |  li TMP0, LJ_TSTR
  |   decode_RB8 RB, INS
  |  stw STR:RC, 4(CARG3)
  |   add CARG2, BASE, RB
  |  stw TMP0, 0(CARG3)
  |  b >1
  |
  |->vmeta_tgets:
  |  la CARG2, DISPATCH_GL(tmptv)(DISPATCH)
  |  li TMP0, LJ_TTAB
  |  stw TAB:RB, 4(CARG2)
  |   la CARG3, DISPATCH_GL(tmptv2)(DISPATCH)
  |  stw TMP0, 0(CARG2)
  |   li TMP1, LJ_TSTR
  |   stw STR:RC, 4(CARG3)
  |   stw TMP1, 0(CARG3)
  |  b >1
  |
  |->vmeta_tgetb:			// TMP0 = index
  |.if not DUALNUM
  |  tonum_u f0, TMP0
  |.endif
  |   decode_RB8 RB, INS
  |  la CARG3, DISPATCH_GL(tmptv)(DISPATCH)
  |   add CARG2, BASE, RB
  |.if DUALNUM
  |  stw TISNUM, 0(CARG3)
  |  stw TMP0, 4(CARG3)
  |.else
  |  stfd f0, 0(CARG3)
  |.endif
  |  b >1
  |
  |->vmeta_tgetv:
  |  decode_RB8 RB, INS
  |   decode_RC8 RC, INS
  |  add CARG2, BASE, RB
  |   add CARG3, BASE, RC
  |1:
  |  stp BASE, L->base
  |  mr CARG1, L
  |  stw PC, SAVE_PC
  |  bl extern lj_meta_tget		// (lua_State *L, TValue *o, TValue *k)
  |  // Returns TValue * (finished) or NULL (metamethod).
  |  cmplwi CRET1, 0
  |  beq >3
  |.if FPU
  |   lfd f0, 0(CRET1)
  |.else
  |   lwz TMP0, 0(CRET1)
  |   lwz TMP1, 4(CRET1)
  |.endif
  |  ins_next1
  |.if FPU
  |   stfdx f0, BASE, RA
  |.else
  |   stwux TMP0, RA, BASE
  |   stw TMP1, 4(RA)
  |.endif
  |  ins_next2
  |
  |3:  // Call __index metamethod.
  |  // BASE = base, L->top = new base, stack = cont/func/t/k
  |  subfic TMP1, BASE, FRAME_CONT
  |  lp BASE, L->top
  |  stw PC, -16(BASE)			// [cont|PC]
  |   add PC, TMP1, BASE
  |  lwz LFUNC:RB, FRAME_FUNC(BASE)	// Guaranteed to be a function here.
  |   li NARGS8:RC, 16			// 2 args for func(t, k).
  |  b ->vm_call_dispatch_f
  |
  |->vmeta_tgetr:
  |  bl extern lj_tab_getinth		// (GCtab *t, int32_t key)
  |  // Returns cTValue * or NULL.
  |  cmplwi CRET1, 0
  |  beq >1
  |.if FPU
  |  lfd f14, 0(CRET1)
  |.else
  |  lwz SAVE0, 0(CRET1)
  |  lwz SAVE1, 4(CRET1)
  |.endif
  |  b ->BC_TGETR_Z
  |1:
  |  stwx TISNIL, BASE, RA
  |  b ->cont_nop
  |
  |//-----------------------------------------------------------------------
  |
  |->vmeta_tsets1:
  |  la CARG3, DISPATCH_GL(tmptv)(DISPATCH)
  |  li TMP0, LJ_TSTR
  |   decode_RB8 RB, INS
  |  stw STR:RC, 4(CARG3)
  |   add CARG2, BASE, RB
  |  stw TMP0, 0(CARG3)
  |  b >1
  |
  |->vmeta_tsets:
  |  la CARG2, DISPATCH_GL(tmptv)(DISPATCH)
  |  li TMP0, LJ_TTAB
  |  stw TAB:RB, 4(CARG2)
  |   la CARG3, DISPATCH_GL(tmptv2)(DISPATCH)
  |  stw TMP0, 0(CARG2)
  |   li TMP1, LJ_TSTR
  |   stw STR:RC, 4(CARG3)
  |   stw TMP1, 0(CARG3)
  |  b >1
  |
  |->vmeta_tsetb:			// TMP0 = index
  |.if not DUALNUM
  |  tonum_u f0, TMP0
  |.endif
  |   decode_RB8 RB, INS
  |  la CARG3, DISPATCH_GL(tmptv)(DISPATCH)
  |   add CARG2, BASE, RB
  |.if DUALNUM
  |  stw TISNUM, 0(CARG3)
  |  stw TMP0, 4(CARG3)
  |.else
  |  stfd f0, 0(CARG3)
  |.endif
  |  b >1
  |
  |->vmeta_tsetv:
  |  decode_RB8 RB, INS
  |   decode_RC8 RC, INS
  |  add CARG2, BASE, RB
  |   add CARG3, BASE, RC
  |1:
  |  stp BASE, L->base
  |  mr CARG1, L
  |  stw PC, SAVE_PC
  |  bl extern lj_meta_tset		// (lua_State *L, TValue *o, TValue *k)
  |  // Returns TValue * (finished) or NULL (metamethod).
  |  cmplwi CRET1, 0
  |.if FPU
  |   lfdx f0, BASE, RA
  |.else
  |   lwzux TMP2, RA, BASE
  |   lwz TMP3, 4(RA)
  |.endif
  |  beq >3
  |  // NOBARRIER: lj_meta_tset ensures the table is not black.
  |  ins_next1
  |.if FPU
  |   stfd f0, 0(CRET1)
  |.else
  |   stw TMP2, 0(CRET1)
  |   stw TMP3, 4(CRET1)
  |.endif
  |  ins_next2
  |
  |3:  // Call __newindex metamethod.
  |  // BASE = base, L->top = new base, stack = cont/func/t/k/(v)
  |  subfic TMP1, BASE, FRAME_CONT
  |  lp BASE, L->top
  |  stw PC, -16(BASE)			// [cont|PC]
  |   add PC, TMP1, BASE
  |  lwz LFUNC:RB, FRAME_FUNC(BASE)	// Guaranteed to be a function here.
  |   li NARGS8:RC, 24			// 3 args for func(t, k, v)
  |.if FPU
  |  stfd f0, 16(BASE)			// Copy value to third argument.
  |.else
  |  stw TMP2, 16(BASE)
  |  stw TMP3, 20(BASE)
  |.endif
  |  b ->vm_call_dispatch_f
  |
  |->vmeta_tsetr:
  |  stp BASE, L->base
  |  mr CARG1, L
  |  stw PC, SAVE_PC
  |  bl extern lj_tab_setinth  // (lua_State *L, GCtab *t, int32_t key)
  |  // Returns TValue *.
  |.if FPU
  |  stfd f14, 0(CRET1)
  |.else
  |  stw SAVE0, 0(CRET1)
  |  stw SAVE1, 4(CRET1)
  |.endif
  |  b ->cont_nop
  |
  |//-- Comparison metamethods ---------------------------------------------
  |
  |->vmeta_comp:
  |  mr CARG1, L
  |   subi PC, PC, 4
  |.if DUALNUM
  |  mr CARG2, RA
  |.else
  |  add CARG2, BASE, RA
  |.endif
  |   stw PC, SAVE_PC
  |.if DUALNUM
  |  mr CARG3, RD
  |.else
  |  add CARG3, BASE, RD
  |.endif
  |   stp BASE, L->base
  |  decode_OP1 CARG4, INS
  |  bl extern lj_meta_comp  // (lua_State *L, TValue *o1, *o2, int op)
  |  // Returns 0/1 or TValue * (metamethod).
  |3:
  |  cmplwi CRET1, 1
  |  bgt ->vmeta_binop
  |  subfic CRET1, CRET1, 0
  |4:
  |  lwz INS, 0(PC)
  |   addi PC, PC, 4
  |  decode_RD4 TMP2, INS
  |  addis TMP2, TMP2, -(BCBIAS_J*4 >> 16)
  |  and TMP2, TMP2, CRET1
  |  add PC, PC, TMP2
  |->cont_nop:
  |  ins_next
  |
  |->cont_ra:				// RA = resultptr
  |  lwz INS, -4(PC)
  |.if FPU
  |   lfd f0, 0(RA)
  |.else
  |   lwz CARG1, 0(RA)
  |   lwz CARG2, 4(RA)
  |.endif
  |  decode_RA8 TMP1, INS
  |.if FPU
  |   stfdx f0, BASE, TMP1
  |.else
  |   stwux CARG1, TMP1, BASE
  |   stw CARG2, 4(TMP1)
  |.endif
  |  b ->cont_nop
  |
  |->cont_condt:			// RA = resultptr
  |  lwz TMP0, 0(RA)
  |  .gpr64 extsw TMP0, TMP0
  |  subfic TMP0, TMP0, LJ_TTRUE	// Branch if result is true.
  |  subfe CRET1, CRET1, CRET1
  |  not CRET1, CRET1
  |  b <4
  |
  |->cont_condf:			// RA = resultptr
  |  lwz TMP0, 0(RA)
  |  .gpr64 extsw TMP0, TMP0
  |  subfic TMP0, TMP0, LJ_TTRUE	// Branch if result is false.
  |  subfe CRET1, CRET1, CRET1
  |  b <4
  |
  |->vmeta_equal:
  |  // CARG2, CARG3, CARG4 are already set by BC_ISEQV/BC_ISNEV.
  |  subi PC, PC, 4
  |   stp BASE, L->base
  |  mr CARG1, L
  |   stw PC, SAVE_PC
  |  bl extern lj_meta_equal  // (lua_State *L, GCobj *o1, *o2, int ne)
  |  // Returns 0/1 or TValue * (metamethod).
  |  b <3
  |
  |->vmeta_equal_cd:
  |.if FFI
  |  mr CARG2, INS
  |  subi PC, PC, 4
  |   stp BASE, L->base
  |  mr CARG1, L
  |   stw PC, SAVE_PC
  |  bl extern lj_meta_equal_cd		// (lua_State *L, BCIns op)
  |  // Returns 0/1 or TValue * (metamethod).
  |  b <3
  |.endif
  |
  |->vmeta_istype:
  |  subi PC, PC, 4
  |   stp BASE, L->base
  |   srwi CARG2, RA, 3
  |   mr CARG1, L
  |   srwi CARG3, RD, 3
  |  stw PC, SAVE_PC
  |  bl extern lj_meta_istype  // (lua_State *L, BCReg ra, BCReg tp)
  |  b ->cont_nop
  |
  |//-- Arithmetic metamethods ---------------------------------------------
  |
  |->vmeta_arith_nv:
  |  add CARG3, KBASE, RC
  |  add CARG4, BASE, RB
  |  b >1
  |->vmeta_arith_nv2:
  |.if DUALNUM
  |  mr CARG3, RC
  |  mr CARG4, RB
  |  b >1
  |.endif
  |
  |->vmeta_unm:
  |  mr CARG3, RD
  |  mr CARG4, RD
  |  b >1
  |
  |->vmeta_arith_vn:
  |  add CARG3, BASE, RB
  |  add CARG4, KBASE, RC
  |  b >1
  |
  |->vmeta_arith_vv:
  |  add CARG3, BASE, RB
  |  add CARG4, BASE, RC
  |.if DUALNUM
  |  b >1
  |.endif
  |->vmeta_arith_vn2:
  |->vmeta_arith_vv2:
  |.if DUALNUM
  |  mr CARG3, RB
  |  mr CARG4, RC
  |.endif
  |1:
  |  add CARG2, BASE, RA
  |   stp BASE, L->base
  |  mr CARG1, L
  |   stw PC, SAVE_PC
  |  decode_OP1 CARG5, INS		// Caveat: CARG5 overlaps INS.
  |  bl extern lj_meta_arith  // (lua_State *L, TValue *ra,*rb,*rc, BCReg op)
  |  // Returns NULL (finished) or TValue * (metamethod).
  |  cmplwi CRET1, 0
  |  beq ->cont_nop
  |
  |  // Call metamethod for binary op.
  |->vmeta_binop:
  |  // BASE = old base, CRET1 = new base, stack = cont/func/o1/o2
  |  sub TMP1, CRET1, BASE
  |   stw PC, -16(CRET1)		// [cont|PC]
  |   mr TMP2, BASE
  |  addi PC, TMP1, FRAME_CONT
  |   mr BASE, CRET1
  |  li NARGS8:RC, 16			// 2 args for func(o1, o2).
  |  b ->vm_call_dispatch
  |
  |->vmeta_len:
#if LJ_52
  |  mr SAVE0, CARG1
#endif
  |  mr CARG2, RD
  |   stp BASE, L->base
  |  mr CARG1, L
  |   stw PC, SAVE_PC
  |  bl extern lj_meta_len		// (lua_State *L, TValue *o)
  |  // Returns NULL (retry) or TValue * (metamethod base).
#if LJ_52
  |  cmplwi CRET1, 0
  |  bne ->vmeta_binop			// Binop call for compatibility.
  |  mr CARG1, SAVE0
  |  b ->BC_LEN_Z
#else
  |  b ->vmeta_binop			// Binop call for compatibility.
#endif
  |
  |//-- Call metamethod ----------------------------------------------------
  |
  |->vmeta_call:			// Resolve and call __call metamethod.
  |  // TMP2 = old base, BASE = new base, RC = nargs*8
  |  mr CARG1, L
  |   stp TMP2, L->base			// This is the callers base!
  |  subi CARG2, BASE, 8
  |   stw PC, SAVE_PC
  |  add CARG3, BASE, RC
  |   mr SAVE0, NARGS8:RC
  |  bl extern lj_meta_call	// (lua_State *L, TValue *func, TValue *top)
  |  lwz LFUNC:RB, FRAME_FUNC(BASE)	// Guaranteed to be a function here.
  |   addi NARGS8:RC, SAVE0, 8		// Got one more argument now.
  |  ins_call
  |
  |->vmeta_callt:			// Resolve __call for BC_CALLT.
  |  // BASE = old base, RA = new base, RC = nargs*8
  |  mr CARG1, L
  |   stp BASE, L->base
  |  subi CARG2, RA, 8
  |   stw PC, SAVE_PC
  |  add CARG3, RA, RC
  |   mr SAVE0, NARGS8:RC
  |  bl extern lj_meta_call	// (lua_State *L, TValue *func, TValue *top)
  |  lwz TMP1, FRAME_PC(BASE)
  |   addi NARGS8:RC, SAVE0, 8		// Got one more argument now.
  |   lwz LFUNC:RB, FRAME_FUNC(RA)	// Guaranteed to be a function here.
  |  b ->BC_CALLT_Z
  |
  |//-- Argument coercion for 'for' statement ------------------------------
  |
  |->vmeta_for:
  |  mr CARG1, L
  |   stp BASE, L->base
  |  mr CARG2, RA
  |   stw PC, SAVE_PC
  |  mr SAVE0, INS
  |  bl extern lj_meta_for	// (lua_State *L, TValue *base)
  |.if JIT
  |   decode_OP1 TMP0, SAVE0
  |.endif
  |  decode_RA8 RA, SAVE0
  |.if JIT
  |   cmpwi TMP0, BC_JFORI
  |.endif
  |  decode_RD8 RD, SAVE0
  |.if JIT
  |   beqy =>BC_JFORI
  |.endif
  |  b =>BC_FORI
  |
  |//-----------------------------------------------------------------------
  |//-- Fast functions -----------------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |.macro .ffunc, name
  |->ff_ .. name:
  |.endmacro
  |
  |.macro .ffunc_1, name
  |->ff_ .. name:
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG3, 0(BASE)
  |    lwz CARG1, 4(BASE)
  |  blt ->fff_fallback
  |.endmacro
  |
  |.macro .ffunc_2, name
  |->ff_ .. name:
  |  cmplwi NARGS8:RC, 16
  |   lwz CARG3, 0(BASE)
  |    lwz CARG4, 8(BASE)
  |   lwz CARG1, 4(BASE)
  |    lwz CARG2, 12(BASE)
  |  blt ->fff_fallback
  |.endmacro
  |
  |.macro .ffunc_n, name
  |->ff_ .. name:
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG1, 0(BASE)
  |.if FPU
  |    lfd FARG1, 0(BASE)
  |.else
  |    lwz CARG2, 4(BASE)
  |.endif
  |  blt ->fff_fallback
  |  checknum CARG1; bge ->fff_fallback
  |.endmacro
  |
  |.macro .ffunc_nn, name
  |->ff_ .. name:
  |  cmplwi NARGS8:RC, 16
  |   lwz CARG1, 0(BASE)
  |.if FPU
  |    lfd FARG1, 0(BASE)
  |   lwz CARG3, 8(BASE)
  |    lfd FARG2, 8(BASE)
  |.else
  |    lwz CARG2, 4(BASE)
  |   lwz CARG3, 8(BASE)
  |    lwz CARG4, 12(BASE)
  |.endif
  |  blt ->fff_fallback
  |  checknum CARG1; bge ->fff_fallback
  |  checknum CARG3; bge ->fff_fallback
  |.endmacro
  |
  |// Inlined GC threshold check. Caveat: uses TMP0 and TMP1.
  |.macro ffgccheck
  |  lwz TMP0, DISPATCH_GL(gc.total)(DISPATCH)
  |  lwz TMP1, DISPATCH_GL(gc.threshold)(DISPATCH)
  |  cmplw TMP0, TMP1
  |  bgel ->fff_gcstep
  |.endmacro
  |
  |//-- Base library: checks -----------------------------------------------
  |
  |.ffunc_1 assert
  |  li TMP1, LJ_TFALSE
  |   la RA, -8(BASE)
  |  cmplw cr1, CARG3, TMP1
  |    lwz PC, FRAME_PC(BASE)
  |  bge cr1, ->fff_fallback
  |   stw CARG3, 0(RA)
  |  addi RD, NARGS8:RC, 8		// Compute (nresults+1)*8.
  |  addi TMP1, BASE, 8
  |  add TMP2, RA, NARGS8:RC
  |   stw CARG1, 4(RA)
  |  beq ->fff_res			// Done if exactly 1 argument.
  |1:
  |  cmplw TMP1, TMP2
  |.if FPU
  |   lfd f0, 0(TMP1)
  |   stfd f0, 0(TMP1)
  |.else
  |   lwz CARG1, 0(TMP1)
  |   lwz CARG2, 4(TMP1)
  |   stw CARG1, -8(TMP1)
  |   stw CARG2, -4(TMP1)
  |.endif
  |    addi TMP1, TMP1, 8
  |  bney <1
  |  b ->fff_res
  |
  |.ffunc type
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG1, 0(BASE)
  |  blt ->fff_fallback
  |  .gpr64 extsw CARG1, CARG1
  |  subfc TMP0, TISNUM, CARG1
  |  subfe TMP2, CARG1, CARG1
  |  orc TMP1, TMP2, TMP0
  |  addi TMP1, TMP1, ~LJ_TISNUM+1
  |  slwi TMP1, TMP1, 3
  |.if FPU
  |   la TMP2, CFUNC:RB->upvalue
  |  lfdx FARG1, TMP2, TMP1
  |.else
  |  add TMP1, CFUNC:RB, TMP1
  |  lwz CARG1, CFUNC:TMP1->upvalue[0].u32.hi
  |  lwz CARG2, CFUNC:TMP1->upvalue[0].u32.lo
  |.endif
  |  b ->fff_resn
  |
  |//-- Base library: getters and setters ---------------------------------
  |
  |.ffunc_1 getmetatable
  |  checktab CARG3; bne >6
  |1:  // Field metatable must be at same offset for GCtab and GCudata!
  |  lwz TAB:CARG1, TAB:CARG1->metatable
  |2:
  |  li CARG3, LJ_TNIL
  |   cmplwi TAB:CARG1, 0
  |  lwz STR:RC, DISPATCH_GL(gcroot[GCROOT_MMNAME+MM_metatable])(DISPATCH)
  |   beq ->fff_restv
  |  lwz TMP0, TAB:CARG1->hmask
  |   li CARG3, LJ_TTAB			// Use metatable as default result.
  |  lwz TMP1, STR:RC->sid
  |  lwz NODE:TMP2, TAB:CARG1->node
  |  and TMP1, TMP1, TMP0		// idx = str->sid & tab->hmask
  |  slwi TMP0, TMP1, 5
  |  slwi TMP1, TMP1, 3
  |  sub TMP1, TMP0, TMP1
  |  add NODE:TMP2, NODE:TMP2, TMP1	// node = tab->node + (idx*32-idx*8)
  |3:  // Rearranged logic, because we expect _not_ to find the key.
  |  lwz CARG4, NODE:TMP2->key
  |   lwz TMP0, 4+offsetof(Node, key)(NODE:TMP2)
  |    lwz CARG2, NODE:TMP2->val
  |     lwz TMP1, 4+offsetof(Node, val)(NODE:TMP2)
  |  checkstr CARG4; bne >4
  |   cmpw TMP0, STR:RC; beq >5
  |4:
  |  lwz NODE:TMP2, NODE:TMP2->next
  |  cmplwi NODE:TMP2, 0
  |  beq ->fff_restv			// Not found, keep default result.
  |  b <3
  |5:
  |  checknil CARG2
  |  beq ->fff_restv			// Ditto for nil value.
  |  mr CARG3, CARG2			// Return value of mt.__metatable.
  |  mr CARG1, TMP1
  |  b ->fff_restv
  |
  |6:
  |  cmpwi CARG3, LJ_TUDATA; beq <1
  |  .gpr64 extsw CARG3, CARG3
  |  subfc TMP0, TISNUM, CARG3
  |  subfe TMP2, CARG3, CARG3
  |  orc TMP1, TMP2, TMP0
  |  addi TMP1, TMP1, ~LJ_TISNUM+1
  |  slwi TMP1, TMP1, 2
  |   la TMP2, DISPATCH_GL(gcroot[GCROOT_BASEMT])(DISPATCH)
  |  lwzx TAB:CARG1, TMP2, TMP1
  |  b <2
  |
  |.ffunc_2 setmetatable
  |  // Fast path: no mt for table yet and not clearing the mt.
  |   checktab CARG3; bne ->fff_fallback
  |  lwz TAB:TMP1, TAB:CARG1->metatable
  |   checktab CARG4; bne ->fff_fallback
  |  cmplwi TAB:TMP1, 0
  |   lbz TMP3, TAB:CARG1->marked
  |  bne ->fff_fallback
  |   andix. TMP0, TMP3, LJ_GC_BLACK	// isblack(table)
  |    stw TAB:CARG2, TAB:CARG1->metatable
  |   beq ->fff_restv
  |  barrierback TAB:CARG1, TMP3, TMP0
  |  b ->fff_restv
  |
  |.ffunc rawget
  |  cmplwi NARGS8:RC, 16
  |   lwz CARG4, 0(BASE)
  |    lwz TAB:CARG2, 4(BASE)
  |  blt ->fff_fallback
  |  checktab CARG4; bne ->fff_fallback
  |   la CARG3, 8(BASE)
  |   mr CARG1, L
  |  bl extern lj_tab_get  // (lua_State *L, GCtab *t, cTValue *key)
  |  // Returns cTValue *.
  |.if FPU
  |  lfd FARG1, 0(CRET1)
  |.else
  |  lwz CARG2, 4(CRET1)
  |  lwz CARG1, 0(CRET1)	// Caveat: CARG1 == CRET1.
  |.endif
  |  b ->fff_resn
  |
  |//-- Base library: conversions ------------------------------------------
  |
  |.ffunc tonumber
  |  // Only handles the number case inline (without a base argument).
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG1, 0(BASE)
  |.if FPU
  |    lfd FARG1, 0(BASE)
  |.else
  |    lwz CARG2, 4(BASE)
  |.endif
  |  bne ->fff_fallback			// Exactly one argument.
  |   checknum CARG1; bgt ->fff_fallback
  |  b ->fff_resn
  |
  |.ffunc_1 tostring
  |  // Only handles the string or number case inline.
  |  checkstr CARG3
  |  // A __tostring method in the string base metatable is ignored.
  |  beq ->fff_restv			// String key?
  |  // Handle numbers inline, unless a number base metatable is present.
  |  lwz TMP0, DISPATCH_GL(gcroot[GCROOT_BASEMT_NUM])(DISPATCH)
  |  checknum CARG3
  |  cmplwi cr1, TMP0, 0
  |   stp BASE, L->base			// Add frame since C call can throw.
  |  crorc 4*cr0+eq, 4*cr0+gt, 4*cr1+eq
  |   stw PC, SAVE_PC			// Redundant (but a defined value).
  |  beq ->fff_fallback
  |  ffgccheck
  |  mr CARG1, L
  |  mr CARG2, BASE
  |.if DUALNUM
  |  bl extern lj_strfmt_number		// (lua_State *L, cTValue *o)
  |.else
  |  bl extern lj_strfmt_num		// (lua_State *L, lua_Number *np)
  |.endif
  |  // Returns GCstr *.
  |  li CARG3, LJ_TSTR
  |  b ->fff_restv
  |
  |//-- Base library: iterators -------------------------------------------
  |
  |.ffunc_1 next
  |   stwx TISNIL, BASE, NARGS8:RC	// Set missing 2nd arg to nil.
  |  checktab CARG3
  |   lwz PC, FRAME_PC(BASE)
  |  bne ->fff_fallback
  |  la CARG2, 8(BASE)
  |  la CARG3, -8(BASE)
  |  bl extern lj_tab_next		// (GCtab *t, cTValue *key, TValue *o)
  |  // Returns 1=found, 0=end, -1=error.
  |  cmpwi CRET1, 0
  |   la RA, -8(BASE)
  |   li RD, (2+1)*8
  |  bgt ->fff_res			// Found key/value.
  |   li CARG3, LJ_TNIL
  |  beq ->fff_restv			// End of traversal: return nil.
  |   lwz CFUNC:RB, FRAME_FUNC(BASE)
  |   li NARGS8:RC, 2*8
  |  b ->fff_fallback			// Invalid key.
  |
  |.ffunc_1 pairs
  |  checktab CARG3
  |   lwz PC, FRAME_PC(BASE)
  |  bne ->fff_fallback
#if LJ_52
  |   lwz TAB:TMP2, TAB:CARG1->metatable
  |.if FPU
  |  lfd f0, CFUNC:RB->upvalue[0]
  |.else
  |  lwz TMP0, CFUNC:RB->upvalue[0].u32.hi
  |  lwz TMP1, CFUNC:RB->upvalue[0].u32.lo
  |.endif
  |   cmplwi TAB:TMP2, 0
  |  la RA, -8(BASE)
  |   bne ->fff_fallback
#else
  |.if FPU
  |  lfd f0, CFUNC:RB->upvalue[0]
  |.else
  |  lwz TMP0, CFUNC:RB->upvalue[0].u32.hi
  |  lwz TMP1, CFUNC:RB->upvalue[0].u32.lo
  |.endif
  |  la RA, -8(BASE)
#endif
  |   stw TISNIL, 8(BASE)
  |  li RD, (3+1)*8
  |.if FPU
  |  stfd f0, 0(RA)
  |.else
  |  stw TMP0, 0(RA)
  |  stw TMP1, 4(RA)
  |.endif
  |  b ->fff_res
  |
  |.ffunc ipairs_aux
  |  cmplwi NARGS8:RC, 16
  |   lwz CARG3, 0(BASE)
  |    lwz TAB:CARG1, 4(BASE)
  |   lwz CARG4, 8(BASE)
  |.if DUALNUM
  |    lwz TMP2, 12(BASE)
  |.else
  |    lfd FARG2, 8(BASE)
  |.endif
  |  blt ->fff_fallback
  |  checktab CARG3
  |  checknum cr1, CARG4
  |   lwz PC, FRAME_PC(BASE)
  |.if DUALNUM
  |  bne ->fff_fallback
  |  bne cr1, ->fff_fallback
  |.else
  |    lus TMP0, 0x3ff0
  |    stw ZERO, TMPD_LO
  |  bne ->fff_fallback
  |    stw TMP0, TMPD_HI
  |  bge cr1, ->fff_fallback
  |    lfd FARG1, TMPD
  |  toint TMP2, FARG2, f0
  |.endif
  |   lwz TMP0, TAB:CARG1->asize
  |   lwz TMP1, TAB:CARG1->array
  |.if not DUALNUM
  |  fadd FARG2, FARG2, FARG1
  |.endif
  |  addi TMP2, TMP2, 1
  |   la RA, -8(BASE)
  |  cmplw TMP0, TMP2
  |.if DUALNUM
  |  stw TISNUM, 0(RA)
  |   slwi TMP3, TMP2, 3
  |  stw TMP2, 4(RA)
  |.else
  |   slwi TMP3, TMP2, 3
  |  stfd FARG2, 0(RA)
  |.endif
  |  ble >2				// Not in array part?
  |.if FPU
  |  lwzx TMP2, TMP1, TMP3
  |  lfdx f0, TMP1, TMP3
  |.else
  |  lwzux TMP2, TMP1, TMP3
  |  lwz TMP3, 4(TMP1)
  |.endif
  |1:
  |  checknil TMP2
  |   li RD, (0+1)*8
  |  beq ->fff_res			// End of iteration, return 0 results.
  |   li RD, (2+1)*8
  |.if FPU
  |  stfd f0, 8(RA)
  |.else
  |  stw TMP2, 8(RA)
  |  stw TMP3, 12(RA)
  |.endif
  |  b ->fff_res
  |2:  // Check for empty hash part first. Otherwise call C function.
  |  lwz TMP0, TAB:CARG1->hmask
  |  cmplwi TMP0, 0
  |   li RD, (0+1)*8
  |  beq ->fff_res
  |   mr CARG2, TMP2
  |  bl extern lj_tab_getinth		// (GCtab *t, int32_t key)
  |  // Returns cTValue * or NULL.
  |  cmplwi CRET1, 0
  |   li RD, (0+1)*8
  |  beq ->fff_res
  |  lwz TMP2, 0(CRET1)
  |.if FPU
  |  lfd f0, 0(CRET1)
  |.else
  |  lwz TMP3, 4(CRET1)
  |.endif
  |  b <1
  |
  |.ffunc_1 ipairs
  |  checktab CARG3
  |   lwz PC, FRAME_PC(BASE)
  |  bne ->fff_fallback
#if LJ_52
  |   lwz TAB:TMP2, TAB:CARG1->metatable
  |.if FPU
  |  lfd f0, CFUNC:RB->upvalue[0]
  |.else
  |  lwz TMP0, CFUNC:RB->upvalue[0].u32.hi
  |  lwz TMP1, CFUNC:RB->upvalue[0].u32.lo
  |.endif
  |   cmplwi TAB:TMP2, 0
  |  la RA, -8(BASE)
  |   bne ->fff_fallback
#else
  |.if FPU
  |  lfd f0, CFUNC:RB->upvalue[0]
  |.else
  |  lwz TMP0, CFUNC:RB->upvalue[0].u32.hi
  |  lwz TMP1, CFUNC:RB->upvalue[0].u32.lo
  |.endif
  |  la RA, -8(BASE)
#endif
  |.if DUALNUM
  |  stw TISNUM, 8(BASE)
  |.else
  |  stw ZERO, 8(BASE)
  |.endif
  |   stw ZERO, 12(BASE)
  |  li RD, (3+1)*8
  |.if FPU
  |  stfd f0, 0(RA)
  |.else
  |  stw TMP0, 0(RA)
  |  stw TMP1, 4(RA)
  |.endif
  |  b ->fff_res
  |
  |//-- Base library: catch errors ----------------------------------------
  |
  |.ffunc pcall
  |  cmplwi NARGS8:RC, 8
  |   lbz TMP3, DISPATCH_GL(hookmask)(DISPATCH)
  |  blt ->fff_fallback
  |   mr TMP2, BASE
  |   la BASE, 8(BASE)
  |  // Remember active hook before pcall.
  |  rlwinm TMP3, TMP3, 32-HOOK_ACTIVE_SHIFT, 31, 31
  |   subi NARGS8:RC, NARGS8:RC, 8
  |  addi PC, TMP3, 8+FRAME_PCALL
  |  b ->vm_call_dispatch
  |
  |.ffunc xpcall
  |  cmplwi NARGS8:RC, 16
  |   lwz CARG3, 8(BASE)
  |.if FPU
  |    lfd FARG2, 8(BASE)
  |    lfd FARG1, 0(BASE)
  |.else
  |    lwz CARG1, 0(BASE)
  |    lwz CARG2, 4(BASE)
  |    lwz CARG4, 12(BASE)
  |.endif
  |  blt ->fff_fallback
  |  lbz TMP1, DISPATCH_GL(hookmask)(DISPATCH)
  |   mr TMP2, BASE
  |  checkfunc CARG3; bne ->fff_fallback  // Traceback must be a function.
  |   la BASE, 16(BASE)
  |  // Remember active hook before pcall.
  |  rlwinm TMP1, TMP1, 32-HOOK_ACTIVE_SHIFT, 31, 31
  |.if FPU
  |    stfd FARG2, 0(TMP2)		// Swap function and traceback.
  |    stfd FARG1, 8(TMP2)
  |.else
  |    stw CARG3, 0(TMP2)
  |    stw CARG4, 4(TMP2)
  |    stw CARG1, 8(TMP2)
  |    stw CARG2, 12(TMP2)
  |.endif
  |  subi NARGS8:RC, NARGS8:RC, 16
  |  addi PC, TMP1, 16+FRAME_PCALL
  |  b ->vm_call_dispatch
  |
  |//-- Coroutine library --------------------------------------------------
  |
  |.macro coroutine_resume_wrap, resume
  |.if resume
  |.ffunc_1 coroutine_resume
  |  cmpwi CARG3, LJ_TTHREAD; bne ->fff_fallback
  |.else
  |.ffunc coroutine_wrap_aux
  |  lwz L:CARG1, CFUNC:RB->upvalue[0].gcr
  |.endif
  |  lbz TMP0, L:CARG1->status
  |   lp TMP1, L:CARG1->cframe
  |    lp CARG2, L:CARG1->top
  |  cmplwi cr0, TMP0, LUA_YIELD
  |    lp TMP2, L:CARG1->base
  |   cmplwi cr1, TMP1, 0
  |   lwz TMP0, L:CARG1->maxstack
  |    cmplw cr7, CARG2, TMP2
  |   lwz PC, FRAME_PC(BASE)
  |  crorc 4*cr6+lt, 4*cr0+gt, 4*cr1+eq		// st>LUA_YIELD || cframe!=0
  |   add TMP2, CARG2, NARGS8:RC
  |  crandc 4*cr6+gt, 4*cr7+eq, 4*cr0+eq	// base==top && st!=LUA_YIELD
  |   cmplw cr1, TMP2, TMP0
  |  cror 4*cr6+lt, 4*cr6+lt, 4*cr6+gt
  |   stw PC, SAVE_PC
  |  cror 4*cr6+lt, 4*cr6+lt, 4*cr1+gt		// cond1 || cond2 || stackov
  |   stp BASE, L->base
  |  blt cr6, ->fff_fallback
  |1:
  |.if resume
  |  addi BASE, BASE, 8			// Keep resumed thread in stack for GC.
  |  subi NARGS8:RC, NARGS8:RC, 8
  |  subi TMP2, TMP2, 8
  |.endif
  |  stp TMP2, L:CARG1->top
  |  li TMP1, 0
  |  stp BASE, L->top
  |2:  // Move args to coroutine.
  |  cmpw TMP1, NARGS8:RC
  |.if FPU
  |   lfdx f0, BASE, TMP1
  |.else
  |   add CARG3, BASE, TMP1
  |   lwz TMP2, 0(CARG3)
  |   lwz TMP3, 4(CARG3)
  |.endif
  |  beq >3
  |.if FPU
  |   stfdx f0, CARG2, TMP1
  |.else
  |   add CARG3, CARG2, TMP1
  |   stw TMP2, 0(CARG3)
  |   stw TMP3, 4(CARG3)
  |.endif
  |  addi TMP1, TMP1, 8
  |  b <2
  |3:
  |  li CARG3, 0
  |   mr L:SAVE0, L:CARG1
  |  li CARG4, 0
  |  bl ->vm_resume			// (lua_State *L, TValue *base, 0, 0)
  |  // Returns thread status.
  |4:
  |  lp TMP2, L:SAVE0->base
  |   cmplwi CRET1, LUA_YIELD
  |  lp TMP3, L:SAVE0->top
  |    li_vmstate INTERP
  |  lp BASE, L->base
  |    stw L, DISPATCH_GL(cur_L)(DISPATCH)
  |    st_vmstate
  |   bgt >8
  |  sub RD, TMP3, TMP2
  |   lwz TMP0, L->maxstack
  |  cmplwi RD, 0
  |   add TMP1, BASE, RD
  |  beq >6				// No results?
  |  cmplw TMP1, TMP0
  |   li TMP1, 0
  |  bgt >9				// Need to grow stack?
  |
  |  subi TMP3, RD, 8
  |   stp TMP2, L:SAVE0->top		// Clear coroutine stack.
  |5:  // Move results from coroutine.
  |  cmplw TMP1, TMP3
  |.if FPU
  |   lfdx f0, TMP2, TMP1
  |   stfdx f0, BASE, TMP1
  |.else
  |   add CARG3, TMP2, TMP1
  |   lwz CARG1, 0(CARG3)
  |   lwz CARG2, 4(CARG3)
  |   add CARG3, BASE, TMP1
  |   stw CARG1, 0(CARG3)
  |   stw CARG2, 4(CARG3)
  |.endif
  |    addi TMP1, TMP1, 8
  |  bne <5
  |6:
  |  andix. TMP0, PC, FRAME_TYPE
  |.if resume
  |  li TMP1, LJ_TTRUE
  |   la RA, -8(BASE)
  |  stw TMP1, -8(BASE)			// Prepend true to results.
  |  addi RD, RD, 16
  |.else
  |  mr RA, BASE
  |  addi RD, RD, 8
  |.endif
  |7:
  |    stw PC, SAVE_PC
  |   mr MULTRES, RD
  |  beq ->BC_RET_Z
  |  b ->vm_return
  |
  |8:  // Coroutine returned with error (at co->top-1).
  |.if resume
  |  andix. TMP0, PC, FRAME_TYPE
  |  la TMP3, -8(TMP3)
  |   li TMP1, LJ_TFALSE
  |.if FPU
  |  lfd f0, 0(TMP3)
  |.else
  |  lwz CARG1, 0(TMP3)
  |  lwz CARG2, 4(TMP3)
  |.endif
  |   stp TMP3, L:SAVE0->top		// Remove error from coroutine stack.
  |    li RD, (2+1)*8
  |   stw TMP1, -8(BASE)		// Prepend false to results.
  |    la RA, -8(BASE)
  |.if FPU
  |  stfd f0, 0(BASE)			// Copy error message.
  |.else
  |  stw CARG1, 0(BASE)			// Copy error message.
  |  stw CARG2, 4(BASE)
  |.endif
  |  b <7
  |.else
  |  mr CARG1, L
  |  mr CARG2, L:SAVE0
  |  bl extern lj_ffh_coroutine_wrap_err  // (lua_State *L, lua_State *co)
  |.endif
  |
  |9:  // Handle stack expansion on return from yield.
  |  mr CARG1, L
  |  srwi CARG2, RD, 3
  |  bl extern lj_state_growstack	// (lua_State *L, int n)
  |  li CRET1, 0
  |  b <4
  |.endmacro
  |
  |  coroutine_resume_wrap 1		// coroutine.resume
  |  coroutine_resume_wrap 0		// coroutine.wrap
  |
  |.ffunc coroutine_yield
  |  lp TMP0, L->cframe
  |   add TMP1, BASE, NARGS8:RC
  |   stp BASE, L->base
  |  andix. TMP0, TMP0, CFRAME_RESUME
  |   stp TMP1, L->top
  |    li CRET1, LUA_YIELD
  |  beq ->fff_fallback
  |   stp ZERO, L->cframe
  |    stb CRET1, L->status
  |  b ->vm_leave_unw
  |
  |//-- Math library -------------------------------------------------------
  |
  |.ffunc_1 math_abs
  |  checknum CARG3
  |.if DUALNUM
  |  bne >2
  |  srawi TMP1, CARG1, 31
  |  xor TMP2, TMP1, CARG1
  |.if GPR64
  |  lus TMP0, 0x8000
  |  sub CARG1, TMP2, TMP1
  |  cmplw CARG1, TMP0
  |  beq >1
  |.else
  |  sub. CARG1, TMP2, TMP1
  |  blt >1
  |.endif
  |->fff_resi:
  |  lwz PC, FRAME_PC(BASE)
  |  la RA, -8(BASE)
  |  stw TISNUM, -8(BASE)
  |  stw CRET1, -4(BASE)
  |  b ->fff_res1
  |1:
  |  lus CARG3, 0x41e0	// 2^31.
  |  li CARG1, 0
  |  b ->fff_restv
  |2:
  |.endif
  |  bge ->fff_fallback
  |  rlwinm CARG3, CARG3, 0, 1, 31
  |  // Fallthrough.
  |
  |->fff_restv:
  |  // CARG3/CARG1 = TValue result.
  |  lwz PC, FRAME_PC(BASE)
  |   stw CARG3, -8(BASE)
  |  la RA, -8(BASE)
  |   stw CARG1, -4(BASE)
  |->fff_res1:
  |  // RA = results, PC = return.
  |  li RD, (1+1)*8
  |->fff_res:
  |  // RA = results, RD = (nresults+1)*8, PC = return.
  |  andix. TMP0, PC, FRAME_TYPE
  |   mr MULTRES, RD
  |  bney ->vm_return
  |  lwz INS, -4(PC)
  |  decode_RB8 RB, INS
  |5:
  |  cmplw RB, RD			// More results expected?
  |   decode_RA8 TMP0, INS
  |  bgt >6
  |  ins_next1
  |  // Adjust BASE. KBASE is assumed to be set for the calling frame.
  |   sub BASE, RA, TMP0
  |  ins_next2
  |
  |6:  // Fill up results with nil.
  |  subi TMP1, RD, 8
  |   addi RD, RD, 8
  |  stwx TISNIL, RA, TMP1
  |  b <5
  |
  |.macro math_extern, func
  |  .ffunc_n math_ .. func
  |  blex func
  |  b ->fff_resn
  |.endmacro
  |
  |.macro math_extern2, func
  |  .ffunc_nn math_ .. func
  |  blex func
  |  b ->fff_resn
  |.endmacro
  |
  |.macro math_round, func
  |  .ffunc_1 math_ .. func
  |   checknum CARG3; beqy ->fff_restv
  |  rlwinm TMP2, CARG3, 12, 21, 31
  |   bge ->fff_fallback
  |  addic. TMP2, TMP2, -1023		// exp = exponent(x) - 1023
  |  cmplwi cr1, TMP2, 31		// 0 <= exp < 31?
  |   subfic TMP0, TMP2, 31
  |  blt >3
  |  slwi TMP1, CARG3, 11
  |   srwi TMP3, CARG1, 21
  |  oris TMP1, TMP1, 0x8000
  |   addi TMP2, TMP2, 1
  |  or TMP1, TMP1, TMP3
  |   slwi CARG2, CARG1, 11
  |  bge cr1, >4
  |   slw TMP3, TMP1, TMP2
  |  srw RD, TMP1, TMP0
  |   or TMP3, TMP3, CARG2
  |  srawi TMP2, CARG3, 31
  |.if "func" == "floor"
  |  and TMP1, TMP3, TMP2
  |  addic TMP0, TMP1, -1
  |  subfe TMP1, TMP0, TMP1
  |  add CARG1, RD, TMP1
  |  xor CARG1, CARG1, TMP2
  |  sub CARG1, CARG1, TMP2
  |  b ->fff_resi
  |.else
  |  andc TMP1, TMP3, TMP2
  |  addic TMP0, TMP1, -1
  |  subfe TMP1, TMP0, TMP1
  |  add CARG1, RD, TMP1
  |  cmpw CARG1, RD
  |  xor CARG1, CARG1, TMP2
  |  sub CARG1, CARG1, TMP2
  |  bge ->fff_resi
  |  // Overflow to 2^31.
  |  lus CARG3, 0x41e0			// 2^31.
  |  li CARG1, 0
  |  b ->fff_restv
  |.endif
  |3:  // |x| < 1
  |  slwi TMP2, CARG3, 1
  |   srawi TMP1, CARG3, 31
  |  or TMP2, CARG1, TMP2		// ztest = (hi+hi) | lo
  |.if "func" == "floor"
  |  and TMP1, TMP2, TMP1		// (ztest & sign) == 0 ? 0 : -1
  |  subfic TMP2, TMP1, 0
  |  subfe CARG1, CARG1, CARG1
  |.else
  |  andc TMP1, TMP2, TMP1		// (ztest & ~sign) == 0 ? 0 : 1
  |  addic TMP2, TMP1, -1
  |  subfe CARG1, TMP2, TMP1
  |.endif
  |  b ->fff_resi
  |4:  // exp >= 31. Check for -(2^31).
  |  xoris TMP1, TMP1, 0x8000
  |  srawi TMP2, CARG3, 31
  |.if "func" == "floor"
  |  or TMP1, TMP1, CARG2
  |.endif
  |.if PPE
  |  orc TMP1, TMP1, TMP2
  |  cmpwi TMP1, 0
  |.else
  |  orc. TMP1, TMP1, TMP2
  |.endif
  |  crand 4*cr0+eq, 4*cr0+eq, 4*cr1+eq
  |  lus CARG1, 0x8000			// -(2^31).
  |  beqy ->fff_resi
  |5:
  |.if FPU
  |  lfd FARG1, 0(BASE)
  |.else
  |  lwz CARG1, 0(BASE)
  |  lwz CARG2, 4(BASE)
  |.endif
  |  blex func
  |  b ->fff_resn
  |.endmacro
  |
  |.if DUALNUM
  |  math_round floor
  |  math_round ceil
  |.else
  |  // NYI: use internal implementation.
  |  math_extern floor
  |  math_extern ceil
  |.endif
  |
  |.if SQRT
  |.ffunc_n math_sqrt
  |  fsqrt FARG1, FARG1
  |  b ->fff_resn
  |.else
  |  math_extern sqrt
  |.endif
  |
  |.ffunc math_log
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG1, 0(BASE)
  |  bne ->fff_fallback			// Need exactly 1 argument.
  |  checknum CARG1; bge ->fff_fallback
  |.if FPU
  |  lfd FARG1, 0(BASE)
  |.else
  |  lwz CARG2, 4(BASE)
  |.endif
  |  blex log
  |  b ->fff_resn
  |
  |  math_extern log10
  |  math_extern exp
  |  math_extern sin
  |  math_extern cos
  |  math_extern tan
  |  math_extern asin
  |  math_extern acos
  |  math_extern atan
  |  math_extern sinh
  |  math_extern cosh
  |  math_extern tanh
  |  math_extern2 pow
  |  math_extern2 atan2
  |  math_extern2 fmod
  |
  |.if DUALNUM
  |.ffunc math_ldexp
  |  cmplwi NARGS8:RC, 16
  |   lwz TMP0, 0(BASE)
  |.if FPU
  |    lfd FARG1, 0(BASE)
  |.else
  |    lwz CARG1, 0(BASE)
  |    lwz CARG2, 4(BASE)
  |.endif
  |   lwz TMP1, 8(BASE)
  |.if GPR64
  |    lwz CARG2, 12(BASE)
  |.elif FPU
  |    lwz CARG1, 12(BASE)
  |.else
  |    lwz CARG3, 12(BASE)
  |.endif
  |  blt ->fff_fallback
  |  checknum TMP0; bge ->fff_fallback
  |  checknum TMP1; bne ->fff_fallback
  |.else
  |.ffunc_nn math_ldexp
  |.if GPR64
  |  toint CARG2, FARG2
  |.else
  |  toint CARG1, FARG2
  |.endif
  |.endif
  |  blex ldexp
  |  b ->fff_resn
  |
  |.ffunc_n math_frexp
  |.if GPR64
  |  la CARG2, DISPATCH_GL(tmptv)(DISPATCH)
  |.elif FPU
  |  la CARG1, DISPATCH_GL(tmptv)(DISPATCH)
  |.else
  |  la CARG3, DISPATCH_GL(tmptv)(DISPATCH)
  |.endif
  |   lwz PC, FRAME_PC(BASE)
  |  blex frexp
  |   lwz TMP1, DISPATCH_GL(tmptv)(DISPATCH)
  |   la RA, -8(BASE)
  |.if not DUALNUM
  |   tonum_i FARG2, TMP1
  |.endif
  |.if FPU
  |  stfd FARG1, 0(RA)
  |.else
  |  stw CRET1, 0(RA)
  |  stw CRET2, 4(RA)
  |.endif
  |  li RD, (2+1)*8
  |.if DUALNUM
  |   stw TISNUM, 8(RA)
  |   stw TMP1, 12(RA)
  |.else
  |   stfd FARG2, 8(RA)
  |.endif
  |  b ->fff_res
  |
  |.ffunc_n math_modf
  |.if GPR64
  |  la CARG2, -8(BASE)
  |.elif FPU
  |  la CARG1, -8(BASE)
  |.else
  |  la CARG3, -8(BASE)
  |.endif
  |   lwz PC, FRAME_PC(BASE)
  |  blex modf
  |   la RA, -8(BASE)
  |.if FPU
  |  stfd FARG1, 0(BASE)
  |.else
  |  stw CRET1, 0(BASE)
  |  stw CRET2, 4(BASE)
  |.endif
  |  li RD, (2+1)*8
  |  b ->fff_res
  |
  |.macro math_minmax, name, ismax
  |.if DUALNUM
  |  .ffunc_1 name
  |  checknum CARG3
  |   addi SAVE0, BASE, 8
  |   add SAVE1, BASE, NARGS8:RC
  |  bne >4
  |1:  // Handle integers.
  |  lwz CARG4, 0(SAVE0)
  |   cmplw cr1, SAVE0, SAVE1
  |  lwz CARG2, 4(SAVE0)
  |   bge cr1, ->fff_resi
  |  checknum CARG4
  |   xoris TMP0, CARG1, 0x8000
  |   xoris TMP3, CARG2, 0x8000
  |  bne >3
  |  subfc TMP3, TMP3, TMP0
  |  subfe TMP0, TMP0, TMP0
  |.if ismax
  |  andc TMP3, TMP3, TMP0
  |.else
  |  and TMP3, TMP3, TMP0
  |.endif
  |  add CARG1, TMP3, CARG2
  |.if GPR64
  |  rldicl CARG1, CARG1, 0, 32
  |.endif
  |   addi SAVE0, SAVE0, 8
  |  b <1
  |3:
  |  bge ->fff_fallback
  |  // Convert intermediate result to number and continue below.
  |.if FPU
  |  tonum_i FARG1, CARG1
  |  lfd FARG2, 0(SAVE0)
  |.else
  |  mr CARG2, CARG1
  |  bl ->vm_sfi2d_1
  |  lwz CARG3, 0(SAVE0)
  |  lwz CARG4, 4(SAVE0)
  |.endif
  |  b >6
  |4:
  |.if FPU
  |   lfd FARG1, 0(BASE)
  |.else
  |   lwz CARG1, 0(BASE)
  |   lwz CARG2, 4(BASE)
  |.endif
  |  bge ->fff_fallback
  |5:  // Handle numbers.
  |  lwz CARG3, 0(SAVE0)
  |   cmplw cr1, SAVE0, SAVE1
  |.if FPU
  |  lfd FARG2, 0(SAVE0)
  |.else
  |  lwz CARG4, 4(SAVE0)
  |.endif
  |   bge cr1, ->fff_resn
  |  checknum CARG3; bge >7
  |6:
  |   addi SAVE0, SAVE0, 8
  |.if FPU
  |.if ismax
  |  fsub f0, FARG1, FARG2
  |.else
  |  fsub f0, FARG2, FARG1
  |.endif
  |  fsel FARG1, f0, FARG1, FARG2
  |.else
  |  stw CARG1, SFSAVE_1
  |  stw CARG2, SFSAVE_2
  |  stw CARG3, SFSAVE_3
  |  stw CARG4, SFSAVE_4
  |  blex __ledf2
  |  cmpwi CRET1, 0
  |.if ismax
  |  blt >8
  |.else
  |  bge >8
  |.endif
  |  lwz CARG1, SFSAVE_1
  |  lwz CARG2, SFSAVE_2
  |  b <5
  |8:
  |  lwz CARG1, SFSAVE_3
  |  lwz CARG2, SFSAVE_4
  |.endif
  |  b <5
  |7:  // Convert integer to number and continue above.
  |   lwz CARG3, 4(SAVE0)
  |  bne ->fff_fallback
  |.if FPU
  |  tonum_i FARG2, CARG3
  |.else
  |  bl ->vm_sfi2d_2
  |.endif
  |  b <6
  |.else
  |  .ffunc_n name
  |  li TMP1, 8
  |1:
  |   lwzx CARG2, BASE, TMP1
  |   lfdx FARG2, BASE, TMP1
  |  cmplw cr1, TMP1, NARGS8:RC
  |   checknum CARG2
  |  bge cr1, ->fff_resn
  |   bge ->fff_fallback
  |.if ismax
  |  fsub f0, FARG1, FARG2
  |.else
  |  fsub f0, FARG2, FARG1
  |.endif
  |   addi TMP1, TMP1, 8
  |  fsel FARG1, f0, FARG1, FARG2
  |  b <1
  |.endif
  |.endmacro
  |
  |  math_minmax math_min, 0
  |  math_minmax math_max, 1
  |
  |//-- String library -----------------------------------------------------
  |
  |.ffunc string_byte			// Only handle the 1-arg case here.
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG3, 0(BASE)
  |    lwz STR:CARG1, 4(BASE)
  |  bne ->fff_fallback			// Need exactly 1 argument.
  |   checkstr CARG3
  |   bne ->fff_fallback
  |  lwz TMP0, STR:CARG1->len
  |.if DUALNUM
  |   lbz CARG1, STR:CARG1[1]		// Access is always ok (NUL at end).
  |   li RD, (0+1)*8
  |   lwz PC, FRAME_PC(BASE)
  |  cmplwi TMP0, 0
  |   la RA, -8(BASE)
  |  beqy ->fff_res
  |  b ->fff_resi
  |.else
  |   lbz TMP1, STR:CARG1[1]		// Access is always ok (NUL at end).
  |  addic TMP3, TMP0, -1		// RD = ((str->len != 0)+1)*8
  |  subfe RD, TMP3, TMP0
  |   stw TMP1, TONUM_LO		// Inlined tonum_u f0, TMP1.
  |  addi RD, RD, 1
  |   lfd f0, TONUM_D
  |  la RA, -8(BASE)
  |  lwz PC, FRAME_PC(BASE)
  |   fsub f0, f0, TOBIT
  |  slwi RD, RD, 3
  |   stfd f0, 0(RA)
  |  b ->fff_res
  |.endif
  |
  |.ffunc string_char			// Only handle the 1-arg case here.
  |  ffgccheck
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG3, 0(BASE)
  |.if DUALNUM
  |    lwz TMP0, 4(BASE)
  |  bne ->fff_fallback			// Exactly 1 argument.
  |  checknum CARG3; bne ->fff_fallback
  |   la CARG2, 7(BASE)
  |.else
  |    lfd FARG1, 0(BASE)
  |  bne ->fff_fallback			// Exactly 1 argument.
  |  checknum CARG3; bge ->fff_fallback
  |  toint TMP0, FARG1
  |   la CARG2, TMPD_BLO
  |.endif
  |   li CARG3, 1
  |  cmplwi TMP0, 255; bgt ->fff_fallback
  |->fff_newstr:
  |  mr CARG1, L
  |  stp BASE, L->base
  |  stw PC, SAVE_PC
  |  bl extern lj_str_new		// (lua_State *L, char *str, size_t l)
  |->fff_resstr:
  |  // Returns GCstr *.
  |  lp BASE, L->base
  |  li CARG3, LJ_TSTR
  |  b ->fff_restv
  |
  |.ffunc string_sub
  |  ffgccheck
  |  cmplwi NARGS8:RC, 16
  |   lwz CARG3, 16(BASE)
  |.if not DUALNUM
  |    lfd f0, 16(BASE)
  |.endif
  |   lwz TMP0, 0(BASE)
  |    lwz STR:CARG1, 4(BASE)
  |  blt ->fff_fallback
  |   lwz CARG2, 8(BASE)
  |.if DUALNUM
  |    lwz TMP1, 12(BASE)
  |.else
  |    lfd f1, 8(BASE)
  |.endif
  |   li TMP2, -1
  |  beq >1
  |.if DUALNUM
  |  checknum CARG3
  |   lwz TMP2, 20(BASE)
  |  bne ->fff_fallback
  |1:
  |  checknum CARG2; bne ->fff_fallback
  |.else
  |  checknum CARG3; bge ->fff_fallback
  |  toint TMP2, f0
  |1:
  |  checknum CARG2; bge ->fff_fallback
  |.endif
  |  checkstr TMP0; bne ->fff_fallback
  |.if not DUALNUM
  |   toint TMP1, f1
  |.endif
  |   lwz TMP0, STR:CARG1->len
  |  cmplw TMP0, TMP2			// len < end? (unsigned compare)
  |   addi TMP3, TMP2, 1
  |  blt >5
  |2:
  |  cmpwi TMP1, 0			// start <= 0?
  |   add TMP3, TMP1, TMP0
  |  ble >7
  |3:
  |  sub CARG3, TMP2, TMP1
  |    addi CARG2, STR:CARG1, #STR-1
  |  srawi TMP0, CARG3, 31
  |   addi CARG3, CARG3, 1
  |    add CARG2, CARG2, TMP1
  |  andc CARG3, CARG3, TMP0
  |.if GPR64
  |  rldicl CARG2, CARG2, 0, 32
  |  rldicl CARG3, CARG3, 0, 32
  |.endif
  |  b ->fff_newstr
  |
  |5:  // Negative end or overflow.
  |  cmpw TMP0, TMP2			// len >= end? (signed compare)
  |   add TMP2, TMP0, TMP3		// Negative end: end = end+len+1.
  |  bge <2
  |   mr TMP2, TMP0			// Overflow: end = len.
  |  b <2
  |
  |7:  // Negative start or underflow.
  |  .gpr64 extsw TMP1, TMP1
  |  addic CARG3, TMP1, -1
  |  subfe CARG3, CARG3, CARG3
  |   srawi CARG2, TMP3, 31		// Note: modifies carry.
  |  andc TMP3, TMP3, CARG3
  |   andc TMP1, TMP3, CARG2
  |  addi TMP1, TMP1, 1			// start = 1 + (start ? start+len : 0)
  |  b <3
  |
  |.macro ffstring_op, name
  |  .ffunc string_ .. name
  |  ffgccheck
  |  cmplwi NARGS8:RC, 8
  |   lwz CARG3, 0(BASE)
  |    lwz STR:CARG2, 4(BASE)
  |  blt ->fff_fallback
  |  checkstr CARG3
  |   la SBUF:CARG1, DISPATCH_GL(tmpbuf)(DISPATCH)
  |  bne ->fff_fallback
  |   lwz TMP0, SBUF:CARG1->b
  |  stw L, SBUF:CARG1->L
  |  stp BASE, L->base
  |  stw PC, SAVE_PC
  |   stw TMP0, SBUF:CARG1->w
  |  bl extern lj_buf_putstr_ .. name
  |  bl extern lj_buf_tostr
  |  b ->fff_resstr
  |.endmacro
  |
  |ffstring_op reverse
  |ffstring_op lower
  |ffstring_op upper
  |
  |//-- Bit library --------------------------------------------------------
  |
  |.macro .ffunc_bit, name
  |.if DUALNUM
  |  .ffunc_1 bit_..name
  |  checknum CARG3; bnel ->fff_tobit_fb
  |.else
  |  .ffunc_n bit_..name
  |  fadd FARG1, FARG1, TOBIT
  |  stfd FARG1, TMPD
  |  lwz CARG1, TMPD_LO
  |.endif
  |.endmacro
  |
  |.macro .ffunc_bit_op, name, ins
  |  .ffunc_bit name
  |  addi SAVE0, BASE, 8
  |  add SAVE1, BASE, NARGS8:RC
  |1:
  |  lwz CARG4, 0(SAVE0)
  |   cmplw cr1, SAVE0, SAVE1
  |.if DUALNUM
  |  lwz CARG2, 4(SAVE0)
  |.else
  |  lfd FARG1, 0(SAVE0)
  |.endif
  |   bgey cr1, ->fff_resi
  |  checknum CARG4
  |.if DUALNUM
  |.if FPU
  |  bnel ->fff_bitop_fb
  |.else
  |  beq >3
  |  stw CARG1, SFSAVE_1
  |  bl ->fff_bitop_fb
  |  mr CARG2, CARG1
  |  lwz CARG1, SFSAVE_1
  |3:
  |.endif
  |.else
  |  fadd FARG1, FARG1, TOBIT
  |  bge ->fff_fallback
  |  stfd FARG1, TMPD
  |  lwz CARG2, TMPD_LO
  |.endif
  |  ins CARG1, CARG1, CARG2
  |   addi SAVE0, SAVE0, 8
  |  b <1
  |.endmacro
  |
  |.ffunc_bit_op band, and
  |.ffunc_bit_op bor, or
  |.ffunc_bit_op bxor, xor
  |
  |.ffunc_bit bswap
  |  rotlwi TMP0, CARG1, 8
  |  rlwimi TMP0, CARG1, 24, 0, 7
  |  rlwimi TMP0, CARG1, 24, 16, 23
  |  mr CRET1, TMP0
  |  b ->fff_resi
  |
  |.ffunc_bit bnot
  |  not CRET1, CARG1
  |  b ->fff_resi
  |
  |.macro .ffunc_bit_sh, name, ins, shmod
  |.if DUALNUM
  |  .ffunc_2 bit_..name
  |.if FPU
  |  checknum CARG3; bnel ->fff_tobit_fb
  |.else
  |  checknum CARG3; beq >1
  |  bl ->fff_tobit_fb
  |  lwz CARG2, 12(BASE)	// Conversion polluted CARG2.
  |1:
  |.endif
  |  // Note: no inline conversion from number for 2nd argument!
  |  checknum CARG4; bne ->fff_fallback
  |.else
  |  .ffunc_nn bit_..name
  |  fadd FARG1, FARG1, TOBIT
  |  fadd FARG2, FARG2, TOBIT
  |  stfd FARG1, TMPD
  |  lwz CARG1, TMPD_LO
  |  stfd FARG2, TMPD
  |  lwz CARG2, TMPD_LO
  |.endif
  |.if shmod == 1
  |  rlwinm CARG2, CARG2, 0, 27, 31
  |.elif shmod == 2
  |  neg CARG2, CARG2
  |.endif
  |  ins CRET1, CARG1, CARG2
  |  b ->fff_resi
  |.endmacro
  |
  |.ffunc_bit_sh lshift, slw, 1
  |.ffunc_bit_sh rshift, srw, 1
  |.ffunc_bit_sh arshift, sraw, 1
  |.ffunc_bit_sh rol, rotlw, 0
  |.ffunc_bit_sh ror, rotlw, 2
  |
  |.ffunc_bit tobit
  |.if DUALNUM
  |  b ->fff_resi
  |.else
  |->fff_resi:
  |  tonum_i FARG1, CRET1
  |.endif
  |->fff_resn:
  |  lwz PC, FRAME_PC(BASE)
  |  la RA, -8(BASE)
  |.if FPU
  |  stfd FARG1, -8(BASE)
  |.else
  |  stw CARG1, -8(BASE)
  |  stw CARG2, -4(BASE)
  |.endif
  |  b ->fff_res1
  |
  |// Fallback FP number to bit conversion.
  |->fff_tobit_fb:
  |.if DUALNUM
  |.if FPU
  |  lfd FARG1, 0(BASE)
  |  bgt ->fff_fallback
  |  fadd FARG1, FARG1, TOBIT
  |  stfd FARG1, TMPD
  |  lwz CARG1, TMPD_LO
  |  blr
  |.else
  |  bgt ->fff_fallback
  |  mr CARG2, CARG1
  |  mr CARG1, CARG3
  |// Modifies: CARG1, CARG2, TMP0, TMP1, TMP2.
  |->vm_tobit:
  |  slwi TMP2, CARG1, 1
  |  addis TMP2, TMP2, 0x0020
  |  cmpwi TMP2, 0
  |  bge >2
  |   li TMP1, 0x3e0
  |  srawi TMP2, TMP2, 21
  |   not TMP1, TMP1
  |  sub. TMP2, TMP1, TMP2
  |    cmpwi cr7, CARG1, 0
  |  blt >1
  |   slwi TMP1, CARG1, 11
  |    srwi TMP0, CARG2, 21
  |   oris TMP1, TMP1, 0x8000
  |   or TMP1, TMP1, TMP0
  |   srw CARG1, TMP1, TMP2
  |  bclr 4, 28			// Return if cr7[lt] == 0, no hint.
  |   neg CARG1, CARG1
  |  blr
  |1:
  |  addi TMP2, TMP2, 21
  |  srw TMP1, CARG2, TMP2
  |   slwi CARG2, CARG1, 12
  |  subfic TMP2, TMP2, 20
  |   slw TMP0, CARG2, TMP2
  |   or CARG1, TMP1, TMP0
  |  bclr 4, 28			// Return if cr7[lt] == 0, no hint.
  |   neg CARG1, CARG1
  |  blr
  |2:
  |  li CARG1, 0
  |  blr
  |.endif
  |.endif
  |->fff_bitop_fb:
  |.if DUALNUM
  |.if FPU
  |  lfd FARG1, 0(SAVE0)
  |  bgt ->fff_fallback
  |  fadd FARG1, FARG1, TOBIT
  |  stfd FARG1, TMPD
  |  lwz CARG2, TMPD_LO
  |  blr
  |.else
  |  bgt ->fff_fallback
  |  mr CARG1, CARG4
  |  b ->vm_tobit
  |.endif
  |.endif
  |
  |//-----------------------------------------------------------------------
  |
  |->fff_fallback:			// Call fast function fallback handler.
  |  // BASE = new base, RB = CFUNC, RC = nargs*8
  |  lp TMP3, CFUNC:RB->f
  |    add TMP1, BASE, NARGS8:RC
  |   lwz PC, FRAME_PC(BASE)		// Fallback may overwrite PC.
  |    addi TMP0, TMP1, 8*LUA_MINSTACK
  |     lwz TMP2, L->maxstack
  |   stw PC, SAVE_PC			// Redundant (but a defined value).
  |  .toc lp TMP3, 0(TMP3)
  |  cmplw TMP0, TMP2
  |     stp BASE, L->base
  |    stp TMP1, L->top
  |   mr CARG1, L
  |  bgt >5				// Need to grow stack.
  |  mtctr TMP3
  |  bctrl				// (lua_State *L)
  |  // Either throws an error, or recovers and returns -1, 0 or nresults+1.
  |  lp BASE, L->base
  |  cmpwi CRET1, 0
  |   slwi RD, CRET1, 3
  |   la RA, -8(BASE)
  |  bgt ->fff_res			// Returned nresults+1?
  |1:  // Returned 0 or -1: retry fast path.
  |  lp TMP0, L->top
  |   lwz LFUNC:RB, FRAME_FUNC(BASE)
  |  sub NARGS8:RC, TMP0, BASE
  |  bne ->vm_call_tail			// Returned -1?
  |  ins_callt				// Returned 0: retry fast path.
  |
  |// Reconstruct previous base for vmeta_call during tailcall.
  |->vm_call_tail:
  |  andix. TMP0, PC, FRAME_TYPE
  |   rlwinm TMP1, PC, 0, 0, 28
  |  bne >3
  |  lwz INS, -4(PC)
  |  decode_RA8 TMP1, INS
  |  addi TMP1, TMP1, 8
  |3:
  |  sub TMP2, BASE, TMP1
  |  b ->vm_call_dispatch		// Resolve again for tailcall.
  |
  |5:  // Grow stack for fallback handler.
  |  li CARG2, LUA_MINSTACK
  |  bl extern lj_state_growstack	// (lua_State *L, int n)
  |  lp BASE, L->base
  |  cmpw TMP0, TMP0			// Set 4*cr0+eq to force retry.
  |  b <1
  |
  |->fff_gcstep:			// Call GC step function.
  |  // BASE = new base, RC = nargs*8
  |  mflr SAVE0
  |   stp BASE, L->base
  |  add TMP0, BASE, NARGS8:RC
  |   stw PC, SAVE_PC			// Redundant (but a defined value).
  |  stp TMP0, L->top
  |  mr CARG1, L
  |  bl extern lj_gc_step		// (lua_State *L)
  |   lp BASE, L->base
  |  mtlr SAVE0
  |    lp TMP0, L->top
  |   sub NARGS8:RC, TMP0, BASE
  |   lwz CFUNC:RB, FRAME_FUNC(BASE)
  |  blr
  |
  |//-----------------------------------------------------------------------
  |//-- Special dispatch targets -------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |->vm_record:				// Dispatch target for recording phase.
  |.if JIT
  |  lbz TMP3, DISPATCH_GL(hookmask)(DISPATCH)
  |  andix. TMP0, TMP3, HOOK_VMEVENT	// No recording while in vmevent.
  |  bne >5
  |  // Decrement the hookcount for consistency, but always do the call.
  |   lwz TMP2, DISPATCH_GL(hookcount)(DISPATCH)
  |  andix. TMP0, TMP3, HOOK_ACTIVE
  |  bne >1
  |   subi TMP2, TMP2, 1
  |  andi. TMP0, TMP3, LUA_MASKLINE|LUA_MASKCOUNT
  |  beqy >1
  |   stw TMP2, DISPATCH_GL(hookcount)(DISPATCH)
  |  b >1
  |.endif
  |
  |->vm_rethook:			// Dispatch target for return hooks.
  |  lbz TMP3, DISPATCH_GL(hookmask)(DISPATCH)
  |  andix. TMP0, TMP3, HOOK_ACTIVE	// Hook already active?
  |  beq >1
  |5:  // Re-dispatch to static ins.
  |  addi TMP1, TMP1, GG_DISP2STATIC	// Assumes decode_OPP TMP1, INS.
  |  lpx TMP0, DISPATCH, TMP1
  |  mtctr TMP0
  |  bctr
  |
  |->vm_inshook:			// Dispatch target for instr/line hooks.
  |  lbz TMP3, DISPATCH_GL(hookmask)(DISPATCH)
  |  lwz TMP2, DISPATCH_GL(hookcount)(DISPATCH)
  |  andix. TMP0, TMP3, HOOK_ACTIVE	// Hook already active?
  |   rlwinm TMP0, TMP3, 31-LUA_HOOKLINE, 31, 0
  |  bne <5
  |
  |   cmpwi cr1, TMP0, 0
  |  addic. TMP2, TMP2, -1
  |   beq cr1, <5
  |  stw TMP2, DISPATCH_GL(hookcount)(DISPATCH)
  |  beq >1
  |   bge cr1, <5
  |1:
  |  mr CARG1, L
  |   stw MULTRES, SAVE_MULTRES
  |  mr CARG2, PC
  |   stp BASE, L->base
  |  // SAVE_PC must hold the _previous_ PC. The callee updates it with PC.
  |  bl extern lj_dispatch_ins		// (lua_State *L, const BCIns *pc)
  |3:
  |  lp BASE, L->base
  |4:  // Re-dispatch to static ins.
  |  lwz INS, -4(PC)
  |  decode_OPP TMP1, INS
  |   decode_RB8 RB, INS
  |  addi TMP1, TMP1, GG_DISP2STATIC
  |   decode_RD8 RD, INS
  |  lpx TMP0, DISPATCH, TMP1
  |   decode_RA8 RA, INS
  |   decode_RC8 RC, INS
  |  mtctr TMP0
  |  bctr
  |
  |->cont_hook:				// Continue from hook yield.
  |  addi PC, PC, 4
  |  lwz MULTRES, -20(RB)		// Restore MULTRES for *M ins.
  |  b <4
  |
  |->vm_hotloop:			// Hot loop counter underflow.
  |.if JIT
  |  lwz LFUNC:TMP1, FRAME_FUNC(BASE)
  |   addi CARG1, DISPATCH, GG_DISP2J
  |   stw PC, SAVE_PC
  |  lwz TMP1, LFUNC:TMP1->pc
  |   mr CARG2, PC
  |   stw L, DISPATCH_J(L)(DISPATCH)
  |  lbz TMP1, PC2PROTO(framesize)(TMP1)
  |   stp BASE, L->base
  |  slwi TMP1, TMP1, 3
  |  add TMP1, BASE, TMP1
  |  stp TMP1, L->top
  |  bl extern lj_trace_hot		// (jit_State *J, const BCIns *pc)
  |  b <3
  |.endif
  |
  |->vm_callhook:			// Dispatch target for call hooks.
  |  mr CARG2, PC
  |.if JIT
  |  b >1
  |.endif
  |
  |->vm_hotcall:			// Hot call counter underflow.
  |.if JIT
  |  ori CARG2, PC, 1
  |1:
  |.endif
  |  add TMP0, BASE, RC
  |   stw PC, SAVE_PC
  |  mr CARG1, L
  |   stp BASE, L->base
  |  sub RA, RA, BASE
  |   stp TMP0, L->top
  |  bl extern lj_dispatch_call		// (lua_State *L, const BCIns *pc)
  |  // Returns ASMFunction.
  |  lp BASE, L->base
  |   lp TMP0, L->top
  |   stw ZERO, SAVE_PC			// Invalidate for subsequent line hook.
  |  sub NARGS8:RC, TMP0, BASE
  |  add RA, BASE, RA
  |  lwz LFUNC:RB, FRAME_FUNC(BASE)
  |  lwz INS, -4(PC)
  |  mtctr CRET1
  |  bctr
  |
  |->cont_stitch:			// Trace stitching.
  |.if JIT
  |  // RA = resultptr, RB = meta base
  |  lwz INS, -4(PC)
  |    lwz TRACE:TMP2, -20(RB)		// Save previous trace.
  |   addic. TMP1, MULTRES, -8
  |  decode_RA8 RC, INS			// Call base.
  |   beq >2
  |1:  // Move results down.
  |.if FPU
  |  lfd f0, 0(RA)
  |.else
  |  lwz CARG1, 0(RA)
  |  lwz CARG2, 4(RA)
  |.endif
  |   addic. TMP1, TMP1, -8
  |    addi RA, RA, 8
  |.if FPU
  |  stfdx f0, BASE, RC
  |.else
  |  add CARG3, BASE, RC
  |  stw CARG1, 0(CARG3)
  |  stw CARG2, 4(CARG3)
  |.endif
  |    addi RC, RC, 8
  |   bne <1
  |2:
  |   decode_RA8 RA, INS
  |   decode_RB8 RB, INS
  |   add RA, RA, RB
  |3:
  |   cmplw RA, RC
  |   bgt >9				// More results wanted?
  |
  |  lhz TMP3, TRACE:TMP2->traceno
  |  lhz RD, TRACE:TMP2->link
  |  cmpw RD, TMP3
  |   cmpwi cr1, RD, 0
  |  beq ->cont_nop			// Blacklisted.
  |    slwi RD, RD, 3
  |   bne cr1, =>BC_JLOOP		// Jump to stitched trace.
  |
  |  // Stitch a new trace to the previous trace.
  |  stw TMP3, DISPATCH_J(exitno)(DISPATCH)
  |  stp L, DISPATCH_J(L)(DISPATCH)
  |  stp BASE, L->base
  |  addi CARG1, DISPATCH, GG_DISP2J
  |  mr CARG2, PC
  |  bl extern lj_dispatch_stitch	// (jit_State *J, const BCIns *pc)
  |  lp BASE, L->base
  |  b ->cont_nop
  |
  |9:
  |  stwx TISNIL, BASE, RC
  |  addi RC, RC, 8
  |  b <3
  |.endif
  |
  |->vm_profhook:			// Dispatch target for profiler hook.
#if LJ_HASPROFILE
  |  mr CARG1, L
  |   stw MULTRES, SAVE_MULTRES
  |  mr CARG2, PC
  |   stp BASE, L->base
  |  bl extern lj_dispatch_profile	// (lua_State *L, const BCIns *pc)
  |  // HOOK_PROFILE is off again, so re-dispatch to dynamic instruction.
  |  lp BASE, L->base
  |  subi PC, PC, 4
  |  b ->cont_nop
#endif
  |
  |//-----------------------------------------------------------------------
  |//-- Trace exit handler -------------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |.macro savex_, a, b, c, d
  |.if FPU
  |  stfd f..a, 16+a*8(sp)
  |  stfd f..b, 16+b*8(sp)
  |  stfd f..c, 16+c*8(sp)
  |  stfd f..d, 16+d*8(sp)
  |.endif
  |.endmacro
  |
  |->vm_exit_handler:
  |.if JIT
  |  addi sp, sp, -(16+32*8+32*4)
  |  stmw r2, 16+32*8+2*4(sp)
  |    addi DISPATCH, JGL, -GG_DISP2G-32768
  |    li CARG2, ~LJ_VMST_EXIT
  |   lwz CARG1, 16+32*8+32*4(sp)	// Get stack chain.
  |    stw CARG2, DISPATCH_GL(vmstate)(DISPATCH)
  |  savex_ 0,1,2,3
  |   stw CARG1, 0(sp)			// Store extended stack chain.
  |   clrso TMP1
  |  savex_ 4,5,6,7
  |   addi CARG2, sp, 16+32*8+32*4	// Recompute original value of sp.
  |  savex_ 8,9,10,11
  |   stw CARG2, 16+32*8+1*4(sp)	// Store sp in RID_SP.
  |  savex_ 12,13,14,15
  |   mflr CARG3
  |   li TMP1, 0
  |  savex_ 16,17,18,19
  |   stw TMP1, 16+32*8+0*4(sp)		// Clear RID_TMP.
  |  savex_ 20,21,22,23
  |   lhz CARG4, 2(CARG3)		// Load trace number.
  |  savex_ 24,25,26,27
  |  lwz L, DISPATCH_GL(cur_L)(DISPATCH)
  |  savex_ 28,29,30,31
  |   sub CARG3, TMP0, CARG3		// Compute exit number.
  |  lp BASE, DISPATCH_GL(jit_base)(DISPATCH)
  |   srwi CARG3, CARG3, 2
  |  stp L, DISPATCH_J(L)(DISPATCH)
  |   subi CARG3, CARG3, 2
  |  stp BASE, L->base
  |   stw CARG4, DISPATCH_J(parent)(DISPATCH)
  |  stw TMP1, DISPATCH_GL(jit_base)(DISPATCH)
  |  addi CARG1, DISPATCH, GG_DISP2J
  |   stw CARG3, DISPATCH_J(exitno)(DISPATCH)
  |  addi CARG2, sp, 16
  |  bl extern lj_trace_exit		// (jit_State *J, ExitState *ex)
  |  // Returns MULTRES (unscaled) or negated error code.
  |  lp TMP1, L->cframe
  |  lwz TMP2, 0(sp)
  |   lp BASE, L->base
  |.if GPR64
  |  rldicr sp, TMP1, 0, 61
  |.else
  |  rlwinm sp, TMP1, 0, 0, 29
  |.endif
  |   lwz PC, SAVE_PC			// Get SAVE_PC.
  |  stw TMP2, 0(sp)
  |  stw L, SAVE_L			// Set SAVE_L (on-trace resume/yield).
  |  b >1
  |.endif
  |->vm_exit_interp:
  |.if JIT
  |  // CARG1 = MULTRES or negated error code, BASE, PC and JGL set.
  |  lwz L, SAVE_L
  |  addi DISPATCH, JGL, -GG_DISP2G-32768
  |  stp BASE, L->base
  |1:
  |  li TMP2, -LUA_ERRERR
  |  cmplw CARG1, TMP2
  |  bge >9				// Check for error from exit.
  |  lwz LFUNC:RB, FRAME_FUNC(BASE)
  |   slwi MULTRES, CARG1, 3
  |    li TMP2, 0
  |   stw MULTRES, SAVE_MULTRES
  |  lwz TMP1, LFUNC:RB->pc
  |    stw TMP2, DISPATCH_GL(jit_base)(DISPATCH)
  |  lwz KBASE, PC2PROTO(k)(TMP1)
  |  // Setup type comparison constants.
  |  li TISNUM, LJ_TISNUM
  |  .FPU lus TMP3, 0x59c0		// TOBIT = 2^52 + 2^51 (float).
  |  .FPU stw TMP3, TMPD
  |  li ZERO, 0
  |  .FPU ori TMP3, TMP3, 0x0004	// TONUM = 2^52 + 2^51 + 2^31 (float).
  |  .FPU lfs TOBIT, TMPD
  |  .FPU stw TMP3, TMPD
  |  .FPU lus TMP0, 0x4338			// Hiword of 2^52 + 2^51 (double)
  |    li TISNIL, LJ_TNIL
  |  .FPU stw TMP0, TONUM_HI
  |  .FPU lfs TONUM, TMPD
  |  // Modified copy of ins_next which handles function header dispatch, too.
  |  lwz INS, 0(PC)
  |   addi PC, PC, 4
  |    // Assumes TISNIL == ~LJ_VMST_INTERP == -1.
  |    stw TISNIL, DISPATCH_GL(vmstate)(DISPATCH)
  |  cmpwi CARG1, -17			// Static dispatch?
  |  beq >5
  |  decode_OPP TMP1, INS
  |   decode_RA8 RA, INS
  |  lpx TMP0, DISPATCH, TMP1
  |  mtctr TMP0
  |  cmplwi TMP1, BC_FUNCF*4		// Function header?
  |  bge >2
  |   decode_RB8 RB, INS
  |   decode_RD8 RD, INS
  |   decode_RC8 RC, INS
  |  bctr
  |2:
  |  cmplwi TMP1, (BC_FUNCC+2)*4	// Fast function?
  |  blt >3
  |  // Check frame below fast function.
  |  lwz TMP1, FRAME_PC(BASE)
  |  andix. TMP0, TMP1, FRAME_TYPE
  |  bney >3				// Trace stitching continuation?
  |  // Otherwise set KBASE for Lua function below fast function.
  |  lwz TMP2, -4(TMP1)
  |  decode_RA8 TMP0, TMP2
  |  sub TMP1, BASE, TMP0
  |  lwz LFUNC:TMP2, -12(TMP1)
  |  lwz TMP1, LFUNC:TMP2->pc
  |  lwz KBASE, PC2PROTO(k)(TMP1)
  |3:
  |   subi RC, MULTRES, 8
  |   add RA, RA, BASE
  |  bctr
  |
  |5:  // Dispatch to static entry of original ins replaced by BC_JLOOP.
  |  lwz TMP1, DISPATCH_J(trace)(DISPATCH)
  |  decode_RD4 RD, INS
  |  lwzx TRACE:TMP1, TMP1, RD
  |  lwz INS, TRACE:TMP1->startins
  |  decode_OPP TMP1, INS
  |  addi TMP1, TMP1, GG_DISP2STATIC
  |  lpx TMP0, DISPATCH, TMP1
  |  mtctr TMP0
  |   decode_RB8 RB, INS
  |   decode_RD8 RD, INS
  |   decode_RA8 RA, INS
  |   decode_RC8 RC, INS
  |  bctr
  |
  |9:  // Rethrow error from the right C frame.
  |  neg CARG2, CARG1
  |  mr CARG1, L
  |  bl extern lj_err_trace		// (lua_State *L, int errcode)
  |.endif
  |
  |//-----------------------------------------------------------------------
  |//-- Math helper functions ----------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |// NYI: Use internal implementations of floor, ceil, trunc, sfcmp.
  |
  |.macro sfi2d, AHI, ALO
  |.if not FPU
  |  mr. AHI, ALO
  |  bclr 12, 2				// Handle zero first.
  |  srawi TMP0, ALO, 31
  |  xor TMP1, ALO, TMP0
  |  sub TMP1, TMP1, TMP0		// Absolute value in TMP1.
  |  cntlzw AHI, TMP1
  |  andix. TMP0, TMP0, 0x800		// Mask sign bit.
  |  slw TMP1, TMP1, AHI		// Align mantissa left with leading 1.
  |  subfic AHI, AHI, 0x3ff+31-1	// Exponent -1 in AHI.
  |  slwi ALO, TMP1, 21
  |  or AHI, AHI, TMP0			// Sign | Exponent.
  |  srwi TMP1, TMP1, 11
  |  slwi AHI, AHI, 20			// Align left.
  |  add AHI, AHI, TMP1			// Add mantissa, increment exponent.
  |  blr
  |.endif
  |.endmacro
  |
  |// Input: CARG2. Output: CARG1, CARG2. Temporaries: TMP0, TMP1.
  |->vm_sfi2d_1:
  |  sfi2d CARG1, CARG2
  |
  |// Input: CARG4. Output: CARG3, CARG4. Temporaries: TMP0, TMP1.
  |->vm_sfi2d_2:
  |  sfi2d CARG3, CARG4
  |
  |->vm_modi:
  |  divwo. TMP0, CARG1, CARG2
  |  bso >1
  |.if GPR64
  |   xor CARG3, CARG1, CARG2
  |   cmpwi CARG3, 0
  |.else
  |   xor. CARG3, CARG1, CARG2
  |.endif
  |  mullw TMP0, TMP0, CARG2
  |  sub CARG1, CARG1, TMP0
  |   bgelr
  |  cmpwi CARG1, 0; beqlr
  |  add CARG1, CARG1, CARG2
  |  blr
  |1:
  |  cmpwi CARG2, 0
  |   li CARG1, 0
  |  beqlr
  |  clrso TMP0			// Clear SO for -2147483648 % -1 and return 0.
  |  blr
  |
  |//-----------------------------------------------------------------------
  |//-- Miscellaneous functions --------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |// void lj_vm_cachesync(void *start, void *end)
  |// Flush D-Cache and invalidate I-Cache. Assumes 32 byte cache line size.
  |// This is a good lower bound, except for very ancient PPC models.
  |->vm_cachesync:
  |.if JIT or FFI
  |  // Compute start of first cache line and number of cache lines.
  |  rlwinm CARG1, CARG1, 0, 0, 26
  |  sub CARG2, CARG2, CARG1
  |  addi CARG2, CARG2, 31
  |  rlwinm. CARG2, CARG2, 27, 5, 31
  |  beqlr
  |  mtctr CARG2
  |  mr CARG3, CARG1
  |1:  // Flush D-Cache.
  |  dcbst r0, CARG1
  |  addi CARG1, CARG1, 32
  |  bdnz <1
  |  sync
  |  mtctr CARG2
  |1:  // Invalidate I-Cache.
  |  icbi r0, CARG3
  |  addi CARG3, CARG3, 32
  |  bdnz <1
  |  isync
  |  blr
  |.endif
  |
  |->vm_next:
  |.if JIT
  |  NYI  // On big-endian.
  |.endif
  |
  |//-----------------------------------------------------------------------
  |//-- FFI helper functions -----------------------------------------------
  |//-----------------------------------------------------------------------
  |
  |// Handler for callback functions. Callback slot number in r11, g in r12.
  |->vm_ffi_callback:
  |.if FFI
  |.type CTSTATE, CTState, PC
  |  saveregs
  |  lwz CTSTATE, GL:r12->ctype_state
  |   addi DISPATCH, r12, GG_G2DISP
  |  stw r11, CTSTATE->cb.slot
  |  stw r3, CTSTATE->cb.gpr[0]
  |   .FPU stfd f1, CTSTATE->cb.fpr[0]
  |  stw r4, CTSTATE->cb.gpr[1]
  |   .FPU stfd f2, CTSTATE->cb.fpr[1]
  |  stw r5, CTSTATE->cb.gpr[2]
  |   .FPU stfd f3, CTSTATE->cb.fpr[2]
  |  stw r6, CTSTATE->cb.gpr[3]
  |   .FPU stfd f4, CTSTATE->cb.fpr[3]
  |  stw r7, CTSTATE->cb.gpr[4]
  |   .FPU stfd f5, CTSTATE->cb.fpr[4]
  |  stw r8, CTSTATE->cb.gpr[5]
  |   .FPU stfd f6, CTSTATE->cb.fpr[5]
  |  stw r9, CTSTATE->cb.gpr[6]
  |   .FPU stfd f7, CTSTATE->cb.fpr[6]
  |  stw r10, CTSTATE->cb.gpr[7]
  |   .FPU stfd f8, CTSTATE->cb.fpr[7]
  |  addi TMP0, sp, CFRAME_SPACE+8
  |  stw TMP0, CTSTATE->cb.stack
  |   mr CARG1, CTSTATE
  |  stw CTSTATE, SAVE_PC		// Any value outside of bytecode is ok.
  |   mr CARG2, sp
  |  bl extern lj_ccallback_enter	// (CTState *cts, void *cf)
  |  // Returns lua_State *.
  |  lp BASE, L:CRET1->base
  |     li TISNUM, LJ_TISNUM		// Setup type comparison constants.
  |  lp RC, L:CRET1->top
  |     .FPU lus TMP3, 0x59c0		// TOBIT = 2^52 + 2^51 (float).
  |     li ZERO, 0
  |   mr L, CRET1
  |     .FPU stw TMP3, TMPD
  |     .FPU lus TMP0, 0x4338		// Hiword of 2^52 + 2^51 (double)
  |  lwz LFUNC:RB, FRAME_FUNC(BASE)
  |     .FPU ori TMP3, TMP3, 0x0004	// TONUM = 2^52 + 2^51 + 2^31 (float).
  |     .FPU stw TMP0, TONUM_HI
  |     li TISNIL, LJ_TNIL
  |    li_vmstate INTERP
  |     .FPU lfs TOBIT, TMPD
  |     .FPU stw TMP3, TMPD
  |  sub RC, RC, BASE
  |    st_vmstate
  |     .FPU lfs TONUM, TMPD
  |  ins_callt
  |.endif
  |
  |->cont_ffi_callback:			// Return from FFI callback.
  |.if FFI
  |  lwz CTSTATE, DISPATCH_GL(ctype_state)(DISPATCH)
  |   stp BASE, L->base
  |   stp RB, L->top
  |  stp L, CTSTATE->L
  |  mr CARG1, CTSTATE
  |  mr CARG2, RA
  |  bl extern lj_ccallback_leave	// (CTState *cts, TValue *o)
  |  lwz CRET1, CTSTATE->cb.gpr[0]
  |  .FPU lfd FARG1, CTSTATE->cb.fpr[0]
  |  lwz CRET2, CTSTATE->cb.gpr[1]
  |  b ->vm_leave_unw
  |.endif
  |
  |->vm_ffi_call:			// Call C function via FFI.
  |  // Caveat: needs special frame unwinding, see below.
  |.if FFI
  |  .type CCSTATE, CCallState, CARG1
  |  lwz TMP1, CCSTATE->spadj
  |    mflr TMP0
  |   lbz CARG2, CCSTATE->nsp
  |   lbz CARG3, CCSTATE->nfpr
  |  neg TMP1, TMP1
  |    stw TMP0, 4(sp)
  |   cmpwi cr1, CARG3, 0
  |  mr TMP2, sp
  |   addic. CARG2, CARG2, -1
  |  stwux sp, sp, TMP1
  |   crnot 4*cr1+eq, 4*cr1+eq		// For vararg calls.
  |  stw r14, -4(TMP2)
  |  stw CCSTATE, -8(TMP2)
  |  mr r14, TMP2
  |  la TMP1, CCSTATE->stack
  |   slwi CARG2, CARG2, 2
  |   blty >2
  |  la TMP2, 8(sp)
  |1:
  |  lwzx TMP0, TMP1, CARG2
  |  stwx TMP0, TMP2, CARG2
  |   addic. CARG2, CARG2, -4
  |  bge <1
  |2:
  |  bney cr1, >3
  |  .FPU lfd f1, CCSTATE->fpr[0]
  |  .FPU lfd f2, CCSTATE->fpr[1]
  |  .FPU lfd f3, CCSTATE->fpr[2]
  |  .FPU lfd f4, CCSTATE->fpr[3]
  |  .FPU lfd f5, CCSTATE->fpr[4]
  |  .FPU lfd f6, CCSTATE->fpr[5]
  |  .FPU lfd f7, CCSTATE->fpr[6]
  |  .FPU lfd f8, CCSTATE->fpr[7]
  |3:
  |   lp TMP0, CCSTATE->func
  |  lwz CARG2, CCSTATE->gpr[1]
  |  lwz CARG3, CCSTATE->gpr[2]
  |  lwz CARG4, CCSTATE->gpr[3]
  |  lwz CARG5, CCSTATE->gpr[4]
  |   mtctr TMP0
  |  lwz r8, CCSTATE->gpr[5]
  |  lwz r9, CCSTATE->gpr[6]
  |  lwz r10, CCSTATE->gpr[7]
  |  lwz CARG1, CCSTATE->gpr[0]		// Do this last, since CCSTATE is CARG1.
  |   bctrl
  |  lwz CCSTATE:TMP1, -8(r14)
  |  lwz TMP2, -4(r14)
  |   lwz TMP0, 4(r14)
  |  stw CARG1, CCSTATE:TMP1->gpr[0]
  |  .FPU stfd FARG1, CCSTATE:TMP1->fpr[0]
  |  stw CARG2, CCSTATE:TMP1->gpr[1]
  |   mtlr TMP0
  |  stw CARG3, CCSTATE:TMP1->gpr[2]
  |   mr sp, r14
  |  stw CARG4, CCSTATE:TMP1->gpr[3]
  |   mr r14, TMP2
  |  blr
  |.endif
  |// Note: vm_ffi_call must be the last function in this object file!
  |
  |//-----------------------------------------------------------------------
}

/* Generate the code for a single instruction. */
static void build_ins(BuildCtx *ctx, BCOp op, int defop)
{
  int vk = 0;
  |=>defop:

  switch (op) {

  /* -- Comparison ops ---------------------------------------------------- */

  /* Remember: all ops branch for a true comparison, fall through otherwise. */

  case BC_ISLT: case BC_ISGE: case BC_ISLE: case BC_ISGT:
    |  // RA = src1*8, RD = src2*8, JMP with RD = target
    |.if DUALNUM
    |  lwzux CARG1, RA, BASE
    |    addi PC, PC, 4
    |   lwz CARG2, 4(RA)
    |  lwzux CARG3, RD, BASE
    |    lwz TMP2, -4(PC)
    |  checknum cr0, CARG1
    |   lwz CARG4, 4(RD)
    |    decode_RD4 TMP2, TMP2
    |  checknum cr1, CARG3
    |    addis SAVE0, TMP2, -(BCBIAS_J*4 >> 16)
    |  bne cr0, >7
    |  bne cr1, >8
    |   cmpw CARG2, CARG4
    if (op == BC_ISLT) {
      |  bge >2
    } else if (op == BC_ISGE) {
      |  blt >2
    } else if (op == BC_ISLE) {
      |  bgt >2
    } else {
      |  ble >2
    }
    |1:
    |  add PC, PC, SAVE0
    |2:
    |  ins_next
    |
    |7:  // RA is not an integer.
    |  bgt cr0, ->vmeta_comp
    |  // RA is a number.
    |   .FPU lfd f0, 0(RA)
    |  bgt cr1, ->vmeta_comp
    |  blt cr1, >4
    |  // RA is a number, RD is an integer.
    |.if FPU
    |  tonum_i f1, CARG4
    |.else
    |  bl ->vm_sfi2d_2
    |.endif
    |  b >5
    |
    |8: // RA is an integer, RD is not an integer.
    |  bgt cr1, ->vmeta_comp
    |  // RA is an integer, RD is a number.
    |.if FPU
    |  tonum_i f0, CARG2
    |.else
    |  bl ->vm_sfi2d_1
    |.endif
    |4:
    |  .FPU lfd f1, 0(RD)
    |5:
    |.if FPU
    |  fcmpu cr0, f0, f1
    |.else
    |  blex __ledf2
    |  cmpwi CRET1, 0
    |.endif
    if (op == BC_ISLT) {
      |  bge <2
    } else if (op == BC_ISGE) {
      |  blt <2
    } else if (op == BC_ISLE) {
      |  cror 4*cr0+lt, 4*cr0+lt, 4*cr0+eq
      |  bge <2
    } else {
      |  cror 4*cr0+lt, 4*cr0+lt, 4*cr0+eq
      |  blt <2
    }
    |  b <1
    |.else
    |  lwzx TMP0, BASE, RA
    |    addi PC, PC, 4
    |   lfdx f0, BASE, RA
    |  lwzx TMP1, BASE, RD
    |  checknum cr0, TMP0
    |    lwz TMP2, -4(PC)
    |   lfdx f1, BASE, RD
    |  checknum cr1, TMP1
    |    decode_RD4 TMP2, TMP2
    |  bge cr0, ->vmeta_comp
    |    addis TMP2, TMP2, -(BCBIAS_J*4 >> 16)
    |  bge cr1, ->vmeta_comp
    |  fcmpu cr0, f0, f1
    if (op == BC_ISLT) {
      |  bge >1
    } else if (op == BC_ISGE) {
      |  blt >1
    } else if (op == BC_ISLE) {
      |  cror 4*cr0+lt, 4*cr0+lt, 4*cr0+eq
      |  bge >1
    } else {
      |  cror 4*cr0+lt, 4*cr0+lt, 4*cr0+eq
      |  blt >1
    }
    |  add PC, PC, TMP2
    |1:
    |  ins_next
    |.endif
    break;

  case BC_ISEQV: case BC_ISNEV:
    vk = op == BC_ISEQV;
    |  // RA = src1*8, RD = src2*8, JMP with RD = target
    |.if DUALNUM
    |  lwzux CARG1, RA, BASE
    |    addi PC, PC, 4
    |   lwz CARG2, 4(RA)
    |  lwzux CARG3, RD, BASE
    |  checknum cr0, CARG1
    |    lwz SAVE0, -4(PC)
    |  checknum cr1, CARG3
    |    decode_RD4 SAVE0, SAVE0
    |   lwz CARG4, 4(RD)
    |  cror 4*cr7+gt, 4*cr0+gt, 4*cr1+gt
    |    addis SAVE0, SAVE0, -(BCBIAS_J*4 >> 16)
    if (vk) {
      |  ble cr7, ->BC_ISEQN_Z
    } else {
      |  ble cr7, ->BC_ISNEN_Z
    }
    |.else
    |  lwzux CARG1, RA, BASE
    |   lwz SAVE0, 0(PC)
    |    lfd f0, 0(RA)
    |   addi PC, PC, 4
    |  lwzux CARG3, RD, BASE
    |  checknum cr0, CARG1
    |   decode_RD4 SAVE0, SAVE0
    |    lfd f1, 0(RD)
    |  checknum cr1, CARG3
    |   addis SAVE0, SAVE0, -(BCBIAS_J*4 >> 16)
    |  bge cr0, >5
    |  bge cr1, >5
    |  fcmpu cr0, f0, f1
    if (vk) {
      |  bne >1
      |  add PC, PC, SAVE0
    } else {
      |  beq >1
      |  add PC, PC, SAVE0
    }
    |1:
    |  ins_next
    |.endif
    |5:  // Either or both types are not numbers.
    |.if not DUALNUM
    |    lwz CARG2, 4(RA)
    |    lwz CARG4, 4(RD)
    |.endif
    |.if FFI
    |  cmpwi cr7, CARG1, LJ_TCDATA
    |  cmpwi cr5, CARG3, LJ_TCDATA
    |.endif
    |   not TMP2, CARG1
    |  cmplw CARG1, CARG3
    |   cmplwi cr1, TMP2, ~LJ_TISPRI		// Primitive?
    |.if FFI
    |  cror 4*cr7+eq, 4*cr7+eq, 4*cr5+eq
    |.endif
    |   cmplwi cr6, TMP2, ~LJ_TISTABUD		// Table or userdata?
    |.if FFI
    |  beq cr7, ->vmeta_equal_cd
    |.endif
    |    cmplw cr5, CARG2, CARG4
    |  crandc 4*cr0+gt, 4*cr0+eq, 4*cr1+gt	// 2: Same type and primitive.
    |  crorc 4*cr0+lt, 4*cr5+eq, 4*cr0+eq	// 1: Same tv or different type.
    |  crand 4*cr0+eq, 4*cr0+eq, 4*cr5+eq	// 0: Same type and same tv.
    |   mr SAVE1, PC
    |  cror 4*cr0+eq, 4*cr0+eq, 4*cr0+gt	// 0 or 2.
    |  cror 4*cr0+lt, 4*cr0+lt, 4*cr0+gt	// 1 or 2.
    if (vk) {
      |  bne cr0, >6
      |  add PC, PC, SAVE0
      |6:
    } else {
      |  beq cr0, >6
      |  add PC, PC, SAVE0
      |6:
    }
    |.if DUALNUM
    |  bge cr0, >2			// Done if 1 or 2.
    |1:
    |  ins_next
    |2:
    |.else
    |  blt cr0, <1			// Done if 1 or 2.
    |.endif
    |  blt cr6, <1			// Done if not tab/ud.
    |
    |  // Different tables or userdatas. Need to check __eq metamethod.
    |  // Field metatable must be at same offset for GCtab and GCudata!
    |   mr CARG3, CARG4
    |  lwz TAB:TMP2, TAB:CARG2->metatable
    |   li CARG4, 1-vk			// ne = 0 or 1.
    |  cmplwi TAB:TMP2, 0
    |  beq <1				// No metatable?
    |  lbz TMP2, TAB:TMP2->nomm
    |  andix. TMP2, TMP2, 1<<MM_eq
    |  bne <1				// Or 'no __eq' flag set?
    |  mr PC, SAVE1			// Restore old PC.
    |  b ->vmeta_equal			// Handle __eq metamethod.
    break;

  case BC_ISEQS: case BC_ISNES:
    vk = op == BC_ISEQS;
    |  // RA = src*8, RD = str_const*8 (~), JMP with RD = target
    |  lwzux TMP0, RA, BASE
    |   srwi RD, RD, 1
    |  lwz STR:TMP3, 4(RA)
    |    lwz TMP2, 0(PC)
    |   subfic RD, RD, -4
    |    addi PC, PC, 4
    |.if FFI
    |  cmpwi TMP0, LJ_TCDATA
    |.endif
    |   lwzx STR:TMP1, KBASE, RD	// KBASE-4-str_const*4
    |  .gpr64 extsw TMP0, TMP0
    |  subfic TMP0, TMP0, LJ_TSTR
    |.if FFI
    |  beq ->vmeta_equal_cd
    |.endif
    |  sub TMP1, STR:TMP1, STR:TMP3
    |  or TMP0, TMP0, TMP1
    |    decode_RD4 TMP2, TMP2
    |  subfic TMP0, TMP0, 0
    |    addis TMP2, TMP2, -(BCBIAS_J*4 >> 16)
    |  subfe TMP1, TMP1, TMP1
    if (vk) {
      |  andc TMP2, TMP2, TMP1
    } else {
      |  and TMP2, TMP2, TMP1
    }
    |  add PC, PC, TMP2
    |  ins_next
    break;

  case BC_ISEQN: case BC_ISNEN:
    vk = op == BC_ISEQN;
    |  // RA = src*8, RD = num_const*8, JMP with RD = target
    |.if DUALNUM
    |  lwzux CARG1, RA, BASE
    |    addi PC, PC, 4
    |   lwz CARG2, 4(RA)
    |  lwzux CARG3, RD, KBASE
    |  checknum cr0, CARG1
    |    lwz SAVE0, -4(PC)
    |  checknum cr1, CARG3
    |    decode_RD4 SAVE0, SAVE0
    |   lwz CARG4, 4(RD)
    |    addis SAVE0, SAVE0, -(BCBIAS_J*4 >> 16)
    if (vk) {
      |->BC_ISEQN_Z:
    } else {
      |->BC_ISNEN_Z:
    }
    |  bne cr0, >7
    |  bne cr1, >8
    |   cmpw CARG2, CARG4
    |4:
    |.else
    if (vk) {
      |->BC_ISEQN_Z:  // Dummy label.
    } else {
      |->BC_ISNEN_Z:  // Dummy label.
    }
    |  lwzx CARG1, BASE, RA
    |    addi PC, PC, 4
    |   lfdx f0, BASE, RA
    |    lwz SAVE0, -4(PC)
    |  lfdx f1, KBASE, RD
    |    decode_RD4 SAVE0, SAVE0
    |  checknum CARG1
    |    addis SAVE0, SAVE0, -(BCBIAS_J*4 >> 16)
    |  bge >3
    |  fcmpu cr0, f0, f1
    |.endif
    if (vk) {
      |  bne >1
      |  add PC, PC, SAVE0
      |1:
      |.if not FFI
      |3:
      |.endif
    } else {
      |  beq >2
      |1:
      |.if not FFI
      |3:
      |.endif
      |  add PC, PC, SAVE0
      |2:
    }
    |  ins_next
    |.if FFI
    |3:
    |  cmpwi CARG1, LJ_TCDATA
    |  beq ->vmeta_equal_cd
    |  b <1
    |.endif
    |.if DUALNUM
    |7:  // RA is not an integer.
    |  bge cr0, <3
    |  // RA is a number.
    |   .FPU lfd f0, 0(RA)
    |  blt cr1, >1
    |  // RA is a number, RD is an integer.
    |.if FPU
    |  tonum_i f1, CARG4
    |.else
    |  bl ->vm_sfi2d_2
    |.endif
    |  b >2
    |
    |8: // RA is an integer, RD is a number.
    |.if FPU
    |  tonum_i f0, CARG2
    |.else
    |  bl ->vm_sfi2d_1
    |.endif
    |1:
    |  .FPU lfd f1, 0(RD)
    |2:
    |.if FPU
    |  fcmpu cr0, f0, f1
    |.else
    |  blex __ledf2
    |  cmpwi CRET1, 0
    |.endif
    |  b <4
    |.endif
    break;

  case BC_ISEQP: case BC_ISNEP:
    vk = op == BC_ISEQP;
    |  // RA = src*8, RD = primitive_type*8 (~), JMP with RD = target
    |  lwzx TMP0, BASE, RA
    |   srwi TMP1, RD, 3
    |    lwz TMP2, 0(PC)
    |   not TMP1, TMP1
    |    addi PC, PC, 4
    |.if FFI
    |  cmpwi TMP0, LJ_TCDATA
    |.endif
    |  sub TMP0, TMP0, TMP1
    |.if FFI
    |  beq ->vmeta_equal_cd
    |.endif
    |    decode_RD4 TMP2, TMP2
    |  .gpr64 extsw TMP0, TMP0
    |  addic TMP0, TMP0, -1
    |    addis TMP2, TMP2, -(BCBIAS_J*4 >> 16)
    |  subfe TMP1, TMP1, TMP1
    if (vk) {
      |  and TMP2, TMP2, TMP1
    } else {
      |  andc TMP2, TMP2, TMP1
    }
    |  add PC, PC, TMP2
    |  ins_next
    break;

  /* -- Unary test and copy ops ------------------------------------------- */

  case BC_ISTC: case BC_ISFC: case BC_IST: case BC_ISF:
    |  // RA = dst*8 or unused, RD = src*8, JMP with RD = target
    |  lwzx TMP0, BASE, RD
    |   lwz INS, 0(PC)
    |   addi PC, PC, 4
    if (op == BC_IST || op == BC_ISF) {
      |  .gpr64 extsw TMP0, TMP0
      |  subfic TMP0, TMP0, LJ_TTRUE
      |   decode_RD4 TMP2, INS
      |  subfe TMP1, TMP1, TMP1
      |   addis TMP2, TMP2, -(BCBIAS_J*4 >> 16)
      if (op == BC_IST) {
	|  andc TMP2, TMP2, TMP1
      } else {
	|  and TMP2, TMP2, TMP1
      }
      |  add PC, PC, TMP2
    } else {
      |  li TMP1, LJ_TFALSE
      |.if FPU
      |   lfdx f0, BASE, RD
      |.else
      |   lwzux CARG1, RD, BASE
      |   lwz CARG2, 4(RD)
      |.endif
      |  cmplw TMP0, TMP1
      if (op == BC_ISTC) {
	|  bge >1
      } else {
	|  blt >1
      }
      |  addis PC, PC, -(BCBIAS_J*4 >> 16)
      |  decode_RD4 TMP2, INS
      |.if FPU
      |   stfdx f0, BASE, RA
      |.else
      |   stwux CARG1, RA, BASE
      |   stw CARG2, 4(RA)
      |.endif
      |  add PC, PC, TMP2
      |1:
    }
    |  ins_next
    break;

  case BC_ISTYPE:
    |  // RA = src*8, RD = -type*8
    |  lwzx TMP0, BASE, RA
    |  srwi TMP1, RD, 3
    |  ins_next1
    |.if not PPE and not GPR64
    |  add. TMP0, TMP0, TMP1
    |.else
    |  neg TMP1, TMP1
    |  cmpw TMP0, TMP1
    |.endif
    |  bne ->vmeta_istype
    |  ins_next2
    break;
  case BC_ISNUM:
    |  // RA = src*8, RD = -(TISNUM-1)*8
    |  lwzx TMP0, BASE, RA
    |  ins_next1
    |  checknum TMP0
    |  bge ->vmeta_istype
    |  ins_next2
    break;

  /* -- Unary ops --------------------------------------------------------- */

  case BC_MOV:
    |  // RA = dst*8, RD = src*8
    |  ins_next1
    |.if FPU
    |  lfdx f0, BASE, RD
    |  stfdx f0, BASE, RA
    |.else
    |  lwzux TMP0, RD, BASE
    |  lwz TMP1, 4(RD)
    |  stwux TMP0, RA, BASE
    |  stw TMP1, 4(RA)
    |.endif
    |  ins_next2
    break;
  case BC_NOT:
    |  // RA = dst*8, RD = src*8
    |  ins_next1
    |  lwzx TMP0, BASE, RD
    |  .gpr64 extsw TMP0, TMP0
    |  subfic TMP1, TMP0, LJ_TTRUE
    |  adde TMP0, TMP0, TMP1
    |  stwx TMP0, BASE, RA
    |  ins_next2
    break;
  case BC_UNM:
    |  // RA = dst*8, RD = src*8
    |  lwzux TMP1, RD, BASE
    |   lwz TMP0, 4(RD)
    |  checknum TMP1
    |.if DUALNUM
    |  bne >5
    |.if GPR64
    |  lus TMP2, 0x8000
    |  neg TMP0, TMP0
    |  cmplw TMP0, TMP2
    |  beq >4
    |.else
    |  nego. TMP0, TMP0
    |  bso >4
    |1:
    |.endif
    |  ins_next1
    |  stwux TISNUM, RA, BASE
    |   stw TMP0, 4(RA)
    |3:
    |  ins_next2
    |4:
    |.if not GPR64
    |  // Potential overflow.
    |  checkov TMP1, <1			// Ignore unrelated overflow.
    |.endif
    |  lus TMP1, 0x41e0			// 2^31.
    |  li TMP0, 0
    |  b >7
    |.endif
    |5:
    |  bge ->vmeta_unm
    |  xoris TMP1, TMP1, 0x8000
    |7:
    |  ins_next1
    |  stwux TMP1, RA, BASE
    |   stw TMP0, 4(RA)
    |.if DUALNUM
    |  b <3
    |.else
    |  ins_next2
    |.endif
    break;
  case BC_LEN:
    |  // RA = dst*8, RD = src*8
    |  lwzux TMP0, RD, BASE
    |   lwz CARG1, 4(RD)
    |  checkstr TMP0; bne >2
    |  lwz CRET1, STR:CARG1->len
    |1:
    |.if DUALNUM
    |  ins_next1
    |  stwux TISNUM, RA, BASE
    |   stw CRET1, 4(RA)
    |.else
    |  tonum_u f0, CRET1		// Result is a non-negative integer.
    |  ins_next1
    |  stfdx f0, BASE, RA
    |.endif
    |  ins_next2
    |2:
    |  checktab TMP0; bne ->vmeta_len
#if LJ_52
    |  lwz TAB:TMP2, TAB:CARG1->metatable
    |  cmplwi TAB:TMP2, 0
    |  bne >9
    |3:
#endif
    |->BC_LEN_Z:
    |  bl extern lj_tab_len		// (GCtab *t)
    |  // Returns uint32_t (but less than 2^31).
    |  b <1
#if LJ_52
    |9:
    |  lbz TMP0, TAB:TMP2->nomm
    |  andix. TMP0, TMP0, 1<<MM_len
    |  bne <3				// 'no __len' flag set: done.
    |  b ->vmeta_len
#endif
    break;

  /* -- Binary ops -------------------------------------------------------- */

    |.macro ins_arithpre
    |  // RA = dst*8, RB = src1*8, RC = src2*8 | num_const*8
    ||vk = ((int)op - BC_ADDVN) / (BC_ADDNV-BC_ADDVN);
    ||switch (vk) {
    ||case 0:
    |   lwzx CARG1, BASE, RB
    |   .if DUALNUM
    |     lwzx CARG3, KBASE, RC
    |   .endif
    |   .if FPU
    |    lfdx f14, BASE, RB
    |    lfdx f15, KBASE, RC
    |   .else
    |    add TMP1, BASE, RB
    |    add TMP2, KBASE, RC
    |    lwz CARG2, 4(TMP1)
    |    lwz CARG4, 4(TMP2)
    |   .endif
    |   .if DUALNUM
    |     checknum cr0, CARG1
    |     checknum cr1, CARG3
    |     crand 4*cr0+lt, 4*cr0+lt, 4*cr1+lt
    |     bge ->vmeta_arith_vn
    |   .else
    |     checknum CARG1; bge ->vmeta_arith_vn
    |   .endif
    ||  break;
    ||case 1:
    |   lwzx CARG1, BASE, RB
    |   .if DUALNUM
    |     lwzx CARG3, KBASE, RC
    |   .endif
    |   .if FPU
    |    lfdx f15, BASE, RB
    |    lfdx f14, KBASE, RC
    |   .else
    |    add TMP1, BASE, RB
    |    add TMP2, KBASE, RC
    |    lwz CARG2, 4(TMP1)
    |    lwz CARG4, 4(TMP2)
    |   .endif
    |   .if DUALNUM
    |     checknum cr0, CARG1
    |     checknum cr1, CARG3
    |     crand 4*cr0+lt, 4*cr0+lt, 4*cr1+lt
    |     bge ->vmeta_arith_nv
    |   .else
    |     checknum CARG1; bge ->vmeta_arith_nv
    |   .endif
    ||  break;
    ||default:
    |   lwzx CARG1, BASE, RB
    |   lwzx CARG3, BASE, RC
    |   .if FPU
    |    lfdx f14, BASE, RB
    |    lfdx f15, BASE, RC
    |   .else
    |    add TMP1, BASE, RB
    |    add TMP2, BASE, RC
    |    lwz CARG2, 4(TMP1)
    |    lwz CARG4, 4(TMP2)
    |   .endif
    |   checknum cr0, CARG1
    |   checknum cr1, CARG3
    |   crand 4*cr0+lt, 4*cr0+lt, 4*cr1+lt
    |   bge ->vmeta_arith_vv
    ||  break;
    ||}
    |.endmacro
    |
    |.macro ins_arithfallback, ins
    ||switch (vk) {
    ||case 0:
    |   ins ->vmeta_arith_vn2
    ||  break;
    ||case 1:
    |   ins ->vmeta_arith_nv2
    ||  break;
    ||default:
    |   ins ->vmeta_arith_vv2
    ||  break;
    ||}
    |.endmacro
    |
    |.macro intmod, a, b, c
    |  bl ->vm_modi
    |.endmacro
    |
    |.macro fpmod, a, b, c
    |->BC_MODVN_Z:
    |  fdiv FARG1, b, c
    |  // NYI: Use internal implementation of floor.
    |  blex floor			// floor(b/c)
    |  fmul a, FARG1, c
    |  fsub a, b, a			// b - floor(b/c)*c
    |.endmacro
    |
    |.macro sfpmod
    |->BC_MODVN_Z:
    |  stw CARG1, SFSAVE_1
    |  stw CARG2, SFSAVE_2
    |  mr SAVE0, CARG3
    |  mr SAVE1, CARG4
    |  blex __divdf3
    |  blex floor
    |  mr CARG3, SAVE0
    |  mr CARG4, SAVE1
    |  blex __muldf3
    |  mr CARG3, CRET1
    |  mr CARG4, CRET2
    |  lwz CARG1, SFSAVE_1
    |  lwz CARG2, SFSAVE_2
    |  blex __subdf3
    |.endmacro
    |
    |.macro ins_arithfp, fpins
    |  ins_arithpre
    |.if "fpins" == "fpmod_"
    |  b ->BC_MODVN_Z			// Avoid 3 copies. It's slow anyway.
    |.elif FPU
    |  fpins f0, f14, f15
    |  ins_next1
    |  stfdx f0, BASE, RA
    |  ins_next2
    |.else
    |  blex __divdf3			// Only soft-float div uses this macro.
    |  ins_next1
    |  stwux CRET1, RA, BASE
    |  stw CRET2, 4(RA)
    |  ins_next2
    |.endif
    |.endmacro
    |
    |.macro ins_arithdn, intins, fpins, fpcall
    |  // RA = dst*8, RB = src1*8, RC = src2*8 | num_const*8
    ||vk = ((int)op - BC_ADDVN) / (BC_ADDNV-BC_ADDVN);
    ||switch (vk) {
    ||case 0:
    |   lwzux CARG1, RB, BASE
    |   lwzux CARG3, RC, KBASE
    |    lwz CARG2, 4(RB)
    |   checknum cr0, CARG1
    |    lwz CARG4, 4(RC)
    |   checknum cr1, CARG3
    ||  break;
    ||case 1:
    |   lwzux CARG3, RB, BASE
    |   lwzux CARG1, RC, KBASE
    |    lwz CARG4, 4(RB)
    |   checknum cr0, CARG3
    |    lwz CARG2, 4(RC)
    |   checknum cr1, CARG1
    ||  break;
    ||default:
    |   lwzux CARG1, RB, BASE
    |   lwzux CARG3, RC, BASE
    |    lwz CARG2, 4(RB)
    |   checknum cr0, CARG1
    |    lwz CARG4, 4(RC)
    |   checknum cr1, CARG3
    ||  break;
    ||}
    |  bne >5
    |  bne cr1, >5
    |.if "intins" == "intmod"
    |  mr CARG1, CARG2
    |  mr CARG2, CARG4
    |.endif
    |  intins CARG1, CARG2, CARG4
    |  bso >4
    |1:
    |  ins_next1
    |  stwux TISNUM, RA, BASE
    |  stw CARG1, 4(RA)
    |2:
    |  ins_next2
    |4:  // Overflow.
    |  checkov TMP0, <1			// Ignore unrelated overflow.
    |  ins_arithfallback b
    |5:  // FP variant.
    |.if FPU
    ||if (vk == 1) {
    |  lfd f15, 0(RB)
    |  lfd f14, 0(RC)
    ||} else {
    |  lfd f14, 0(RB)
    |  lfd f15, 0(RC)
    ||}
    |.endif
    |  crand 4*cr0+lt, 4*cr0+lt, 4*cr1+lt
    |   ins_arithfallback bge
    |.if "fpins" == "fpmod_"
    |  b ->BC_MODVN_Z			// Avoid 3 copies. It's slow anyway.
    |.else
    |.if FPU
    |  fpins f0, f14, f15
    |  stfdx f0, BASE, RA
    |.else
    |.if "fpcall" == "sfpmod"
    |  sfpmod
    |.else
    |  blex fpcall
    |.endif
    |  stwux CRET1, RA, BASE
    |  stw CRET2, 4(RA)
    |.endif
    |  ins_next1
    |  b <2
    |.endif
    |.endmacro
    |
    |.macro ins_arith, intins, fpins, fpcall
    |.if DUALNUM
    |  ins_arithdn intins, fpins, fpcall
    |.else
    |  ins_arithfp fpins
    |.endif
    |.endmacro

  case BC_ADDVN: case BC_ADDNV: case BC_ADDVV:
    |.if GPR64
    |.macro addo32., y, a, b
    |  // Need to check overflow for (a<<32) + (b<<32).
    |  rldicr TMP0, a, 32, 31
    |  rldicr TMP1, b, 32, 31
    |  addo. TMP0, TMP0, TMP1
    |  add y, a, b
    |.endmacro
    |  ins_arith addo32., fadd, __adddf3
    |.else
    |  ins_arith addo., fadd, __adddf3
    |.endif
    break;
  case BC_SUBVN: case BC_SUBNV: case BC_SUBVV:
    |.if GPR64
    |.macro subo32., y, a, b
    |  // Need to check overflow for (a<<32) - (b<<32).
    |  rldicr TMP0, a, 32, 31
    |  rldicr TMP1, b, 32, 31
    |  subo. TMP0, TMP0, TMP1
    |  sub y, a, b
    |.endmacro
    |  ins_arith subo32., fsub, __subdf3
    |.else
    |  ins_arith subo., fsub, __subdf3
    |.endif
    break;
  case BC_MULVN: case BC_MULNV: case BC_MULVV:
    |  ins_arith mullwo., fmul, __muldf3
    break;
  case BC_DIVVN: case BC_DIVNV: case BC_DIVVV:
    |  ins_arithfp fdiv
    break;
  case BC_MODVN:
    |  ins_arith intmod, fpmod, sfpmod
    break;
  case BC_MODNV: case BC_MODVV:
    |  ins_arith intmod, fpmod_, sfpmod
    break;
  case BC_POW:
    |  // NYI: (partial) integer arithmetic.
    |  lwzx CARG1, BASE, RB
    |  lwzx CARG3, BASE, RC
    |.if FPU
    |   lfdx FARG1, BASE, RB
    |   lfdx FARG2, BASE, RC
    |.else
    |   add TMP1, BASE, RB
    |   add TMP2, BASE, RC
    |   lwz CARG2, 4(TMP1)
    |   lwz CARG4, 4(TMP2)
    |.endif
    |  checknum cr0, CARG1
    |  checknum cr1, CARG3
    |  crand 4*cr0+lt, 4*cr0+lt, 4*cr1+lt
    |  bge ->vmeta_arith_vv
    |  blex pow
    |  ins_next1
    |.if FPU
    |  stfdx FARG1, BASE, RA
    |.else
    |  stwux CARG1, RA, BASE
    |  stw CARG2, 4(RA)
    |.endif
    |  ins_next2
    break;

  case BC_CAT:
    |  // RA = dst*8, RB = src_start*8, RC = src_end*8
    |  sub CARG3, RC, RB
    |   stp BASE, L->base
    |  add CARG2, BASE, RC
    |  mr SAVE0, RB
    |->BC_CAT_Z:
    |   stw PC, SAVE_PC
    |  mr CARG1, L
    |  srwi CARG3, CARG3, 3
    |  bl extern lj_meta_cat		// (lua_State *L, TValue *top, int left)
    |  // Returns NULL (finished) or TValue * (metamethod).
    |  cmplwi CRET1, 0
    |   lp BASE, L->base
    |  bne ->vmeta_binop
    |  ins_next1
    |.if FPU
    |  lfdx f0, BASE, SAVE0		// Copy result from RB to RA.
    |  stfdx f0, BASE, RA
    |.else
    |  lwzux TMP0, SAVE0, BASE
    |  lwz TMP1, 4(SAVE0)
    |  stwux TMP0, RA, BASE
    |  stw TMP1, 4(RA)
    |.endif
    |  ins_next2
    break;

  /* -- Constant ops ------------------------------------------------------ */

  case BC_KSTR:
    |  // RA = dst*8, RD = str_const*8 (~)
    |  srwi TMP1, RD, 1
    |  subfic TMP1, TMP1, -4
    |  ins_next1
    |  lwzx TMP0, KBASE, TMP1		// KBASE-4-str_const*4
    |  li TMP2, LJ_TSTR
    |  stwux TMP2, RA, BASE
    |  stw TMP0, 4(RA)
    |  ins_next2
    break;
  case BC_KCDATA:
    |.if FFI
    |  // RA = dst*8, RD = cdata_const*8 (~)
    |  srwi TMP1, RD, 1
    |  subfic TMP1, TMP1, -4
    |  ins_next1
    |  lwzx TMP0, KBASE, TMP1		// KBASE-4-cdata_const*4
    |  li TMP2, LJ_TCDATA
    |  stwux TMP2, RA, BASE
    |  stw TMP0, 4(RA)
    |  ins_next2
    |.endif
    break;
  case BC_KSHORT:
    |  // RA = dst*8, RD = int16_literal*8
    |.if DUALNUM
    |  slwi RD, RD, 13
    |  srawi RD, RD, 16
    |  ins_next1
    |   stwux TISNUM, RA, BASE
    |   stw RD, 4(RA)
    |  ins_next2
    |.else
    |  // The soft-float approach is faster.
    |  slwi RD, RD, 13
    |  srawi TMP1, RD, 31
    |  xor TMP2, TMP1, RD
    |  sub TMP2, TMP2, TMP1		// TMP2 = abs(x)
    |  cntlzw TMP3, TMP2
    |  subfic TMP1, TMP3, 0x40d		// TMP1 = exponent-1
    |   slw TMP2, TMP2, TMP3		// TMP2 = left aligned mantissa
    |    subfic TMP3, RD, 0
    |  slwi TMP1, TMP1, 20
    |   rlwimi RD, TMP2, 21, 1, 31	// hi = sign(x) | (mantissa>>11)
    |    subfe TMP0, TMP0, TMP0
    |   add RD, RD, TMP1		// hi = hi + exponent-1
    |    and RD, RD, TMP0		// hi = x == 0 ? 0 : hi
    |  ins_next1
    |    stwux RD, RA, BASE
    |    stw ZERO, 4(RA)
    |  ins_next2
    |.endif
    break;
  case BC_KNUM:
    |  // RA = dst*8, RD = num_const*8
    |  ins_next1
    |.if FPU
    |  lfdx f0, KBASE, RD
    |  stfdx f0, BASE, RA
    |.else
    |  lwzux TMP0, RD, KBASE
    |  lwz TMP1, 4(RD)
    |  stwux TMP0, RA, BASE
    |  stw TMP1, 4(RA)
    |.endif
    |  ins_next2
    break;
  case BC_KPRI:
    |  // RA = dst*8, RD = primitive_type*8 (~)
    |  srwi TMP1, RD, 3
    |  not TMP0, TMP1
    |  ins_next1
    |  stwx TMP0, BASE, RA
    |  ins_next2
    break;
  case BC_KNIL:
    |  // RA = base*8, RD = end*8
    |  stwx TISNIL, BASE, RA
    |   addi RA, RA, 8
    |1:
    |  stwx TISNIL, BASE, RA
    |  cmpw RA, RD
    |   addi RA, RA, 8
    |  blt <1
    |  ins_next_
    break;

  /* -- Upvalue and function ops ------------------------------------------ */

  case BC_UGET:
    |  // RA = dst*8, RD = uvnum*8
    |  lwz LFUNC:RB, FRAME_FUNC(BASE)
    |   srwi RD, RD, 1
    |   addi RD, RD, offsetof(GCfuncL, uvptr)
    |  lwzx UPVAL:RB, LFUNC:RB, RD
    |  ins_next1
    |  lwz TMP1, UPVAL:RB->v
    |.if FPU
    |  lfd f0, 0(TMP1)
    |  stfdx f0, BASE, RA
    |.else
    |  lwz TMP2, 0(TMP1)
    |  lwz TMP3, 4(TMP1)
    |  stwux TMP2, RA, BASE
    |  stw TMP3, 4(RA)
    |.endif
    |  ins_next2
    break;
  case BC_USETV:
    |  // RA = uvnum*8, RD = src*8
    |  lwz LFUNC:RB, FRAME_FUNC(BASE)
    |    srwi RA, RA, 1
    |    addi RA, RA, offsetof(GCfuncL, uvptr)
    |.if FPU
    |   lfdux f0, RD, BASE
    |.else
    |   lwzux CARG1, RD, BASE
    |   lwz CARG3, 4(RD)
    |.endif
    |  lwzx UPVAL:RB, LFUNC:RB, RA
    |  lbz TMP3, UPVAL:RB->marked
    |   lwz CARG2, UPVAL:RB->v
    |  andix. TMP3, TMP3, LJ_GC_BLACK	// isblack(uv)
    |    lbz TMP0, UPVAL:RB->closed
    |   lwz TMP2, 0(RD)
    |.if FPU
    |   stfd f0, 0(CARG2)
    |.else
    |   stw CARG1, 0(CARG2)
    |   stw CARG3, 4(CARG2)
    |.endif
    |    cmplwi cr1, TMP0, 0
    |   lwz TMP1, 4(RD)
    |  cror 4*cr0+eq, 4*cr0+eq, 4*cr1+eq
    |   subi TMP2, TMP2, (LJ_TNUMX+1)
    |  bne >2				// Upvalue is closed and black?
    |1:
    |  ins_next
    |
    |2:  // Check if new value is collectable.
    |  cmplwi TMP2, LJ_TISGCV - (LJ_TNUMX+1)
    |  bge <1				// tvisgcv(v)
    |  lbz TMP3, GCOBJ:TMP1->gch.marked
    |  andix. TMP3, TMP3, LJ_GC_WHITES	// iswhite(v)
    |   la CARG1, GG_DISP2G(DISPATCH)
    |  // Crossed a write barrier. Move the barrier forward.
    |  beq <1
    |  bl extern lj_gc_barrieruv	// (global_State *g, TValue *tv)
    |  b <1
    break;
  case BC_USETS:
    |  // RA = uvnum*8, RD = str_const*8 (~)
    |  lwz LFUNC:RB, FRAME_FUNC(BASE)
    |   srwi TMP1, RD, 1
    |    srwi RA, RA, 1
    |   subfic TMP1, TMP1, -4
    |    addi RA, RA, offsetof(GCfuncL, uvptr)
    |   lwzx STR:TMP1, KBASE, TMP1	// KBASE-4-str_const*4
    |  lwzx UPVAL:RB, LFUNC:RB, RA
    |  lbz TMP3, UPVAL:RB->marked
    |   lwz CARG2, UPVAL:RB->v
    |  andix. TMP3, TMP3, LJ_GC_BLACK	// isblack(uv)
    |   lbz TMP3, STR:TMP1->marked
    |   lbz TMP2, UPVAL:RB->closed
    |   li TMP0, LJ_TSTR
    |   stw STR:TMP1, 4(CARG2)
    |   stw TMP0, 0(CARG2)
    |  bne >2
    |1:
    |  ins_next
    |
    |2:  // Check if string is white and ensure upvalue is closed.
    |  andix. TMP3, TMP3, LJ_GC_WHITES	// iswhite(str)
    |   cmplwi cr1, TMP2, 0
    |  cror 4*cr0+eq, 4*cr0+eq, 4*cr1+eq
    |   la CARG1, GG_DISP2G(DISPATCH)
    |  // Crossed a write barrier. Move the barrier forward.
    |  beq <1
    |  bl extern lj_gc_barrieruv	// (global_State *g, TValue *tv)
    |  b <1
    break;
  case BC_USETN:
    |  // RA = uvnum*8, RD = num_const*8
    |  lwz LFUNC:RB, FRAME_FUNC(BASE)
    |   srwi RA, RA, 1
    |   addi RA, RA, offsetof(GCfuncL, uvptr)
    |.if FPU
    |    lfdx f0, KBASE, RD
    |.else
    |    lwzux TMP2, RD, KBASE
    |    lwz TMP3, 4(RD)
    |.endif
    |  lwzx UPVAL:RB, LFUNC:RB, RA
    |  ins_next1
    |  lwz TMP1, UPVAL:RB->v
    |.if FPU
    |  stfd f0, 0(TMP1)
    |.else
    |  stw TMP2, 0(TMP1)
    |  stw TMP3, 4(TMP1)
    |.endif
    |  ins_next2
    break;
  case BC_USETP:
    |  // RA = uvnum*8, RD = primitive_type*8 (~)
    |  lwz LFUNC:RB, FRAME_FUNC(BASE)
    |   srwi RA, RA, 1
    |    srwi TMP0, RD, 3
    |   addi RA, RA, offsetof(GCfuncL, uvptr)
    |    not TMP0, TMP0
    |  lwzx UPVAL:RB, LFUNC:RB, RA
    |  ins_next1
    |  lwz TMP1, UPVAL:RB->v
    |  stw TMP0, 0(TMP1)
    |  ins_next2
    break;

  case BC_UCLO:
    |  // RA = level*8, RD = target
    |  lwz TMP1, L->openupval
    |  branch_RD			// Do this first since RD is not saved.
    |   stp BASE, L->base
    |  cmplwi TMP1, 0
    |   mr CARG1, L
    |  beq >1
    |   add CARG2, BASE, RA
    |  bl extern lj_func_closeuv	// (lua_State *L, TValue *level)
    |  lp BASE, L->base
    |1:
    |  ins_next
    break;

  case BC_FNEW:
    |  // RA = dst*8, RD = proto_const*8 (~) (holding function prototype)
    |  srwi TMP1, RD, 1
    |   stp BASE, L->base
    |  subfic TMP1, TMP1, -4
    |   stw PC, SAVE_PC
    |  lwzx CARG2, KBASE, TMP1		// KBASE-4-tab_const*4
    |   mr CARG1, L
    |  lwz CARG3, FRAME_FUNC(BASE)
    |  // (lua_State *L, GCproto *pt, GCfuncL *parent)
    |  bl extern lj_func_newL_gc
    |  // Returns GCfuncL *.
    |  lp BASE, L->base
    |   li TMP0, LJ_TFUNC
    |  stwux TMP0, RA, BASE
    |  stw LFUNC:CRET1, 4(RA)
    |  ins_next
    break;

  /* -- Table ops --------------------------------------------------------- */

  case BC_TNEW:
  case BC_TDUP:
    |  // RA = dst*8, RD = (hbits|asize)*8 | tab_const*8 (~)
    |  lwz TMP0, DISPATCH_GL(gc.total)(DISPATCH)
    |   mr CARG1, L
    |  lwz TMP1, DISPATCH_GL(gc.threshold)(DISPATCH)
    |   stp BASE, L->base
    |  cmplw TMP0, TMP1
    |   stw PC, SAVE_PC
    |  bge >5
    |1:
    if (op == BC_TNEW) {
      |  rlwinm CARG2, RD, 29, 21, 31
      |  rlwinm CARG3, RD, 18, 27, 31
      |  cmpwi CARG2, 0x7ff; beq >3
      |2:
      |  bl extern lj_tab_new  // (lua_State *L, int32_t asize, uint32_t hbits)
      |  // Returns Table *.
    } else {
      |  srwi TMP1, RD, 1
      |  subfic TMP1, TMP1, -4
      |  lwzx CARG2, KBASE, TMP1		// KBASE-4-tab_const*4
      |  bl extern lj_tab_dup  // (lua_State *L, Table *kt)
      |  // Returns Table *.
    }
    |  lp BASE, L->base
    |   li TMP0, LJ_TTAB
    |  stwux TMP0, RA, BASE
    |  stw TAB:CRET1, 4(RA)
    |  ins_next
    if (op == BC_TNEW) {
      |3:
      |  li CARG2, 0x801
      |  b <2
    }
    |5:
    |  mr SAVE0, RD
    |  bl extern lj_gc_step_fixtop  // (lua_State *L)
    |  mr RD, SAVE0
    |  mr CARG1, L
    |  b <1
    break;

  case BC_GGET:
    |  // RA = dst*8, RD = str_const*8 (~)
  case BC_GSET:
    |  // RA = src*8, RD = str_const*8 (~)
    |  lwz LFUNC:TMP2, FRAME_FUNC(BASE)
    |   srwi TMP1, RD, 1
    |  lwz TAB:RB, LFUNC:TMP2->env
    |   subfic TMP1, TMP1, -4
    |   lwzx STR:RC, KBASE, TMP1	// KBASE-4-str_const*4
    if (op == BC_GGET) {
      |  b ->BC_TGETS_Z
    } else {
      |  b ->BC_TSETS_Z
    }
    break;

  case BC_TGETV:
    |  // RA = dst*8, RB = table*8, RC = key*8
    |  lwzux CARG1, RB, BASE
    |  lwzux CARG2, RC, BASE
    |   lwz TAB:RB, 4(RB)
    |.if DUALNUM
    |   lwz RC, 4(RC)
    |.else
    |   lfd f0, 0(RC)
    |.endif
    |  checktab CARG1
    |   checknum cr1, CARG2
    |  bne ->vmeta_tgetv
    |.if DUALNUM
    |  lwz TMP0, TAB:RB->asize
    |   bne cr1, >5
    |   lwz TMP1, TAB:RB->array
    |  cmplw TMP0, RC
    |   slwi TMP2, RC, 3
    |.else
    |   bge cr1, >5
    |  // Convert number key to integer, check for integerness and range.
    |  fctiwz f1, f0
    |    fadd f2, f0, TOBIT
    |  stfd f1, TMPD
    |   lwz TMP0, TAB:RB->asize
    |    fsub f2, f2, TOBIT
    |  lwz TMP2, TMPD_LO
    |   lwz TMP1, TAB:RB->array
    |    fcmpu cr1, f0, f2
    |  cmplw cr0, TMP0, TMP2
    |  crand 4*cr0+gt, 4*cr0+gt, 4*cr1+eq
    |   slwi TMP2, TMP2, 3
    |.endif
    |  ble ->vmeta_tgetv		// Integer key and in array part?
    |  lwzx TMP0, TMP1, TMP2
    |.if FPU
    |   lfdx f14, TMP1, TMP2
    |.else
    |   lwzux SAVE0, TMP1, TMP2
    |   lwz SAVE1, 4(TMP1)
    |.endif
    |  checknil TMP0; beq >2
    |1:
    |  ins_next1
    |.if FPU
    |   stfdx f14, BASE, RA
    |.else
    |   stwux SAVE0, RA, BASE
    |   stw SAVE1, 4(RA)
    |.endif
    |  ins_next2
    |
    |2:  // Check for __index if table value is nil.
    |  lwz TAB:TMP2, TAB:RB->metatable
    |  cmplwi TAB:TMP2, 0
    |  beq <1				// No metatable: done.
    |  lbz TMP0, TAB:TMP2->nomm
    |  andix. TMP0, TMP0, 1<<MM_index
    |  bne <1				// 'no __index' flag set: done.
    |  b ->vmeta_tgetv
    |
    |5:
    |  checkstr CARG2; bne ->vmeta_tgetv
    |.if not DUALNUM
    |  lwz STR:RC, 4(RC)
    |.endif
    |  b ->BC_TGETS_Z			// String key?
    break;
  case BC_TGETS:
    |  // RA = dst*8, RB = table*8, RC = str_const*8 (~)
    |  lwzux CARG1, RB, BASE
    |   srwi TMP1, RC, 1
    |    lwz TAB:RB, 4(RB)
    |   subfic TMP1, TMP1, -4
    |  checktab CARG1
    |   lwzx STR:RC, KBASE, TMP1	// KBASE-4-str_const*4
    |  bne ->vmeta_tgets1
    |->BC_TGETS_Z:
    |  // TAB:RB = GCtab *, STR:RC = GCstr *, RA = dst*8
    |  lwz TMP0, TAB:RB->hmask
    |  lwz TMP1, STR:RC->sid
    |  lwz NODE:TMP2, TAB:RB->node
    |  and TMP1, TMP1, TMP0		// idx = str->sid & tab->hmask
    |  slwi TMP0, TMP1, 5
    |  slwi TMP1, TMP1, 3
    |  sub TMP1, TMP0, TMP1
    |  add NODE:TMP2, NODE:TMP2, TMP1	// node = tab->node + (idx*32-idx*8)
    |1:
    |  lwz CARG1, NODE:TMP2->key
    |   lwz TMP0, 4+offsetof(Node, key)(NODE:TMP2)
    |    lwz CARG2, NODE:TMP2->val
    |     lwz TMP1, 4+offsetof(Node, val)(NODE:TMP2)
    |  checkstr CARG1; bne >4
    |   cmpw TMP0, STR:RC; bne >4
    |    checknil CARG2; beq >5		// Key found, but nil value?
    |3:
    |    stwux CARG2, RA, BASE
    |     stw TMP1, 4(RA)
    |  ins_next
    |
    |4:  // Follow hash chain.
    |  lwz NODE:TMP2, NODE:TMP2->next
    |  cmplwi NODE:TMP2, 0
    |  bne <1
    |  // End of hash chain: key not found, nil result.
    |   li CARG2, LJ_TNIL
    |
    |5:  // Check for __index if table value is nil.
    |  lwz TAB:TMP2, TAB:RB->metatable
    |  cmplwi TAB:TMP2, 0
    |  beq <3				// No metatable: done.
    |  lbz TMP0, TAB:TMP2->nomm
    |  andix. TMP0, TMP0, 1<<MM_index
    |  bne <3				// 'no __index' flag set: done.
    |  b ->vmeta_tgets
    break;
  case BC_TGETB:
    |  // RA = dst*8, RB = table*8, RC = index*8
    |  lwzux CARG1, RB, BASE
    |   srwi TMP0, RC, 3
    |   lwz TAB:RB, 4(RB)
    |  checktab CARG1; bne ->vmeta_tgetb
    |  lwz TMP1, TAB:RB->asize
    |   lwz TMP2, TAB:RB->array
    |  cmplw TMP0, TMP1; bge ->vmeta_tgetb
    |.if FPU
    |  lwzx TMP1, TMP2, RC
    |   lfdx f0, TMP2, RC
    |.else
    |  lwzux TMP1, TMP2, RC
    |   lwz TMP3, 4(TMP2)
    |.endif
    |  checknil TMP1; beq >5
    |1:
    |  ins_next1
    |.if FPU
    |   stfdx f0, BASE, RA
    |.else
    |   stwux TMP1, RA, BASE
    |   stw TMP3, 4(RA)
    |.endif
    |  ins_next2
    |
    |5:  // Check for __index if table value is nil.
    |  lwz TAB:TMP2, TAB:RB->metatable
    |  cmplwi TAB:TMP2, 0
    |  beq <1				// No metatable: done.
    |  lbz TMP2, TAB:TMP2->nomm
    |  andix. TMP2, TMP2, 1<<MM_index
    |  bne <1				// 'no __index' flag set: done.
    |  b ->vmeta_tgetb			// Caveat: preserve TMP0!
    break;
  case BC_TGETR:
    |  // RA = dst*8, RB = table*8, RC = key*8
    |  add RB, BASE, RB
    |  lwz TAB:CARG1, 4(RB)
    |.if DUALNUM
    |  add RC, BASE, RC
    |  lwz TMP0, TAB:CARG1->asize
    |  lwz CARG2, 4(RC)
    |   lwz TMP1, TAB:CARG1->array
    |.else
    |  lfdx f0, BASE, RC
    |  lwz TMP0, TAB:CARG1->asize
    |  toint CARG2, f0
    |   lwz TMP1, TAB:CARG1->array
    |.endif
    |  cmplw TMP0, CARG2
    |   slwi TMP2, CARG2, 3
    |  ble ->vmeta_tgetr		// In array part?
    |.if FPU
    |   lfdx f14, TMP1, TMP2
    |.else
    |   lwzux SAVE0, TMP2, TMP1
    |   lwz SAVE1, 4(TMP2)
    |.endif
    |->BC_TGETR_Z:
    |  ins_next1
    |.if FPU
    |   stfdx f14, BASE, RA
    |.else
    |   stwux SAVE0, RA, BASE
    |   stw SAVE1, 4(RA)
    |.endif
    |  ins_next2
    break;

  case BC_TSETV:
    |  // RA = src*8, RB = table*8, RC = key*8
    |  lwzux CARG1, RB, BASE
    |  lwzux CARG2, RC, BASE
    |   lwz TAB:RB, 4(RB)
    |.if DUALNUM
    |   lwz RC, 4(RC)
    |.else
    |   lfd f0, 0(RC)
    |.endif
    |  checktab CARG1
    |   checknum cr1, CARG2
    |  bne ->vmeta_tsetv
    |.if DUALNUM
    |  lwz TMP0, TAB:RB->asize
    |   bne cr1, >5
    |   lwz TMP1, TAB:RB->array
    |  cmplw TMP0, RC
    |   slwi TMP0, RC, 3
    |.else
    |   bge cr1, >5
    |  // Convert number key to integer, check for integerness and range.
    |  fctiwz f1, f0
    |    fadd f2, f0, TOBIT
    |  stfd f1, TMPD
    |   lwz TMP0, TAB:RB->asize
    |    fsub f2, f2, TOBIT
    |  lwz TMP2, TMPD_LO
    |   lwz TMP1, TAB:RB->array
    |    fcmpu cr1, f0, f2
    |  cmplw cr0, TMP0, TMP2
    |  crand 4*cr0+gt, 4*cr0+gt, 4*cr1+eq
    |   slwi TMP0, TMP2, 3
    |.endif
    |  ble ->vmeta_tsetv		// Integer key and in array part?
    |   lwzx TMP2, TMP1, TMP0
    |  lbz TMP3, TAB:RB->marked
    |.if FPU
    |    lfdx f14, BASE, RA
    |.else
    |    add SAVE1, BASE, RA
    |    lwz SAVE0, 0(SAVE1)
    |    lwz SAVE1, 4(SAVE1)
    |.endif
    |   checknil TMP2; beq >3
    |1:
    |  andix. TMP2, TMP3, LJ_GC_BLACK	// isblack(table)
    |.if FPU
    |    stfdx f14, TMP1, TMP0
    |.else
    |    stwux SAVE0, TMP1, TMP0
    |    stw SAVE1, 4(TMP1)
    |.endif
    |  bne >7
    |2:
    |  ins_next
    |
    |3:  // Check for __newindex if previous value is nil.
    |  lwz TAB:TMP2, TAB:RB->metatable
    |  cmplwi TAB:TMP2, 0
    |  beq <1				// No metatable: done.
    |  lbz TMP2, TAB:TMP2->nomm
    |  andix. TMP2, TMP2, 1<<MM_newindex
    |  bne <1				// 'no __newindex' flag set: done.
    |  b ->vmeta_tsetv
    |
    |5:
    |  checkstr CARG2; bne ->vmeta_tsetv
    |.if not DUALNUM
    |  lwz STR:RC, 4(RC)
    |.endif
    |  b ->BC_TSETS_Z			// String key?
    |
    |7:  // Possible table write barrier for the value. Skip valiswhite check.
    |  barrierback TAB:RB, TMP3, TMP0
    |  b <2
    break;
  case BC_TSETS:
    |  // RA = src*8, RB = table*8, RC = str_const*8 (~)
    |  lwzux CARG1, RB, BASE
    |   srwi TMP1, RC, 1
    |    lwz TAB:RB, 4(RB)
    |   subfic TMP1, TMP1, -4
    |  checktab CARG1
    |   lwzx STR:RC, KBASE, TMP1	// KBASE-4-str_const*4
    |  bne ->vmeta_tsets1
    |->BC_TSETS_Z:
    |  // TAB:RB = GCtab *, STR:RC = GCstr *, RA = src*8
    |  lwz TMP0, TAB:RB->hmask
    |  lwz TMP1, STR:RC->sid
    |  lwz NODE:TMP2, TAB:RB->node
    |    stb ZERO, TAB:RB->nomm		// Clear metamethod cache.
    |  and TMP1, TMP1, TMP0		// idx = str->sid & tab->hmask
    |.if FPU
    |    lfdx f14, BASE, RA
    |.else
    |    add CARG2, BASE, RA
    |    lwz SAVE0, 0(CARG2)
    |    lwz SAVE1, 4(CARG2)
    |.endif
    |  slwi TMP0, TMP1, 5
    |  slwi TMP1, TMP1, 3
    |  sub TMP1, TMP0, TMP1
    |    lbz TMP3, TAB:RB->marked
    |  add NODE:TMP2, NODE:TMP2, TMP1	// node = tab->node + (idx*32-idx*8)
    |1:
    |  lwz CARG1, NODE:TMP2->key
    |   lwz TMP0, 4+offsetof(Node, key)(NODE:TMP2)
    |    lwz CARG2, NODE:TMP2->val
    |     lwz NODE:TMP1, NODE:TMP2->next
    |  checkstr CARG1; bne >5
    |   cmpw TMP0, STR:RC; bne >5
    |    checknil CARG2; beq >4		// Key found, but nil value?
    |2:
    |  andix. TMP0, TMP3, LJ_GC_BLACK	// isblack(table)
    |.if FPU
    |    stfd f14, NODE:TMP2->val
    |.else
    |    stw SAVE0, NODE:TMP2->val.u32.hi
    |    stw SAVE1, NODE:TMP2->val.u32.lo
    |.endif
    |  bne >7
    |3:
    |  ins_next
    |
    |4:  // Check for __newindex if previous value is nil.
    |  lwz TAB:TMP1, TAB:RB->metatable
    |  cmplwi TAB:TMP1, 0
    |  beq <2				// No metatable: done.
    |  lbz TMP0, TAB:TMP1->nomm
    |  andix. TMP0, TMP0, 1<<MM_newindex
    |  bne <2				// 'no __newindex' flag set: done.
    |  b ->vmeta_tsets
    |
    |5:  // Follow hash chain.
    |  cmplwi NODE:TMP1, 0
    |   mr NODE:TMP2, NODE:TMP1
    |  bne <1
    |  // End of hash chain: key not found, add a new one.
    |
    |  // But check for __newindex first.
    |  lwz TAB:TMP1, TAB:RB->metatable
    |   la CARG3, DISPATCH_GL(tmptv)(DISPATCH)
    |   stw PC, SAVE_PC
    |   mr CARG1, L
    |  cmplwi TAB:TMP1, 0
    |   stp BASE, L->base
    |  beq >6				// No metatable: continue.
    |  lbz TMP0, TAB:TMP1->nomm
    |  andix. TMP0, TMP0, 1<<MM_newindex
    |  beq ->vmeta_tsets		// 'no __newindex' flag NOT set: check.
    |6:
    |  li TMP0, LJ_TSTR
    |   stw STR:RC, 4(CARG3)
    |   mr CARG2, TAB:RB
    |  stw TMP0, 0(CARG3)
    |  bl extern lj_tab_newkey		// (lua_State *L, GCtab *t, TValue *k)
    |  // Returns TValue *.
    |  lp BASE, L->base
    |.if FPU
    |  stfd f14, 0(CRET1)
    |.else
    |  stw SAVE0, 0(CRET1)
    |  stw SAVE1, 4(CRET1)
    |.endif
    |  b <3				// No 2nd write barrier needed.
    |
    |7:  // Possible table write barrier for the value. Skip valiswhite check.
    |  barrierback TAB:RB, TMP3, TMP0
    |  b <3
    break;
  case BC_TSETB:
    |  // RA = src*8, RB = table*8, RC = index*8
    |  lwzux CARG1, RB, BASE
    |   srwi TMP0, RC, 3
    |   lwz TAB:RB, 4(RB)
    |  checktab CARG1; bne ->vmeta_tsetb
    |  lwz TMP1, TAB:RB->asize
    |   lwz TMP2, TAB:RB->array
    |    lbz TMP3, TAB:RB->marked
    |  cmplw TMP0, TMP1
    |.if FPU
    |   lfdx f14, BASE, RA
    |.else
    |   add CARG2, BASE, RA
    |   lwz SAVE0, 0(CARG2)
    |   lwz SAVE1, 4(CARG2)
    |.endif
    |  bge ->vmeta_tsetb
    |  lwzx TMP1, TMP2, RC
    |  checknil TMP1; beq >5
    |1:
    |  andix. TMP0, TMP3, LJ_GC_BLACK	// isblack(table)
    |.if FPU
    |   stfdx f14, TMP2, RC
    |.else
    |   stwux SAVE0, RC, TMP2
    |   stw SAVE1, 4(RC)
    |.endif
    |  bne >7
    |2:
    |  ins_next
    |
    |5:  // Check for __newindex if previous value is nil.
    |  lwz TAB:TMP1, TAB:RB->metatable
    |  cmplwi TAB:TMP1, 0
    |  beq <1				// No metatable: done.
    |  lbz TMP1, TAB:TMP1->nomm
    |  andix. TMP1, TMP1, 1<<MM_newindex
    |  bne <1				// 'no __newindex' flag set: done.
    |  b ->vmeta_tsetb			// Caveat: preserve TMP0!
    |
    |7:  // Possible table write barrier for the value. Skip valiswhite check.
    |  barrierback TAB:RB, TMP3, TMP0
    |  b <2
    break;
  case BC_TSETR:
    |  // RA = dst*8, RB = table*8, RC = key*8
    |  add RB, BASE, RB
    |  lwz TAB:CARG2, 4(RB)
    |.if DUALNUM
    |  add RC, BASE, RC
    |    lbz TMP3, TAB:CARG2->marked
    |  lwz TMP0, TAB:CARG2->asize
    |  lwz CARG3, 4(RC)
    |   lwz TMP1, TAB:CARG2->array
    |.else
    |  lfdx f0, BASE, RC
    |    lbz TMP3, TAB:CARG2->marked
    |  lwz TMP0, TAB:CARG2->asize
    |  toint CARG3, f0
    |   lwz TMP1, TAB:CARG2->array
    |.endif
    |  andix. TMP2, TMP3, LJ_GC_BLACK	// isblack(table)
    |  bne >7
    |2:
    |  cmplw TMP0, CARG3
    |   slwi TMP2, CARG3, 3
    |.if FPU
    |   lfdx f14, BASE, RA
    |.else
    |  lwzux SAVE0, RA, BASE
    |  lwz SAVE1, 4(RA)
    |.endif
    |  ble ->vmeta_tsetr		// In array part?
    |  ins_next1
    |.if FPU
    |   stfdx f14, TMP1, TMP2
    |.else
    |   stwux SAVE0, TMP1, TMP2
    |   stw SAVE1, 4(TMP1)
    |.endif
    |  ins_next2
    |
    |7:  // Possible table write barrier for the value. Skip valiswhite check.
    |  barrierback TAB:CARG2, TMP3, TMP2
    |  b <2
    break;


  case BC_TSETM:
    |  // RA = base*8 (table at base-1), RD = num_const*8 (start index)
    |  add RA, BASE, RA
    |1:
    |   add TMP3, KBASE, RD
    |  lwz TAB:CARG2, -4(RA)		// Guaranteed to be a table.
    |    addic. TMP0, MULTRES, -8
    |   lwz TMP3, 4(TMP3)		// Integer constant is in lo-word.
    |    srwi CARG3, TMP0, 3
    |    beq >4				// Nothing to copy?
    |  add CARG3, CARG3, TMP3
    |  lwz TMP2, TAB:CARG2->asize
    |   slwi TMP1, TMP3, 3
    |    lbz TMP3, TAB:CARG2->marked
    |  cmplw CARG3, TMP2
    |   add TMP2, RA, TMP0
    |   lwz TMP0, TAB:CARG2->array
    |  bgt >5
    |   add TMP1, TMP1, TMP0
    |    andix. TMP0, TMP3, LJ_GC_BLACK	// isblack(table)
    |3:  // Copy result slots to table.
    |.if FPU
    |   lfd f0, 0(RA)
    |.else
    |   lwz SAVE0, 0(RA)
    |   lwz SAVE1, 4(RA)
    |.endif
    |  addi RA, RA, 8
    |  cmpw cr1, RA, TMP2
    |.if FPU
    |   stfd f0, 0(TMP1)
    |.else
    |   stw SAVE0, 0(TMP1)
    |   stw SAVE1, 4(TMP1)
    |.endif
    |    addi TMP1, TMP1, 8
    |  blt cr1, <3
    |  bne >7
    |4:
    |  ins_next
    |
    |5:  // Need to resize array part.
    |   stp BASE, L->base
    |  mr CARG1, L
    |   stw PC, SAVE_PC
    |  mr SAVE0, RD
    |  bl extern lj_tab_reasize		// (lua_State *L, GCtab *t, int nasize)
    |  // Must not reallocate the stack.
    |  mr RD, SAVE0
    |  b <1
    |
    |7:  // Possible table write barrier for any value. Skip valiswhite check.
    |  barrierback TAB:CARG2, TMP3, TMP0
    |  b <4
    break;

  /* -- Calls and vararg handling ----------------------------------------- */

  case BC_CALLM:
    |  // RA = base*8, (RB = (nresults+1)*8,) RC = extra_nargs*8
    |  add NARGS8:RC, NARGS8:RC, MULTRES
    |  // Fall through. Assumes BC_CALL follows.
    break;
  case BC_CALL:
    |  // RA = base*8, (RB = (nresults+1)*8,) RC = (nargs+1)*8
    |  mr TMP2, BASE
    |  lwzux TMP0, BASE, RA
    |   lwz LFUNC:RB, 4(BASE)
    |    subi NARGS8:RC, NARGS8:RC, 8
    |   addi BASE, BASE, 8
    |  checkfunc TMP0; bne ->vmeta_call
    |  ins_call
    break;

  case BC_CALLMT:
    |  // RA = base*8, (RB = 0,) RC = extra_nargs*8
    |  add NARGS8:RC, NARGS8:RC, MULTRES
    |  // Fall through. Assumes BC_CALLT follows.
    break;
  case BC_CALLT:
    |  // RA = base*8, (RB = 0,) RC = (nargs+1)*8
    |  lwzux TMP0, RA, BASE
    |   lwz LFUNC:RB, 4(RA)
    |    subi NARGS8:RC, NARGS8:RC, 8
    |    lwz TMP1, FRAME_PC(BASE)
    |  checkfunc TMP0
    |   addi RA, RA, 8
    |  bne ->vmeta_callt
    |->BC_CALLT_Z:
    |  andix. TMP0, TMP1, FRAME_TYPE	// Caveat: preserve cr0 until the crand.
    |   lbz TMP3, LFUNC:RB->ffid
    |    xori TMP2, TMP1, FRAME_VARG
    |    cmplwi cr1, NARGS8:RC, 0
    |  bne >7
    |1:
    |  stw LFUNC:RB, FRAME_FUNC(BASE)	// Copy function down, but keep PC.
    |  li TMP2, 0
    |   cmplwi cr7, TMP3, 1		// (> FF_C) Calling a fast function?
    |    beq cr1, >3
    |2:
    |  addi TMP3, TMP2, 8
    |.if FPU
    |   lfdx f0, RA, TMP2
    |.else
    |   add CARG3, RA, TMP2
    |   lwz CARG1, 0(CARG3)
    |   lwz CARG2, 4(CARG3)
    |.endif
    |  cmplw cr1, TMP3, NARGS8:RC
    |.if FPU
    |   stfdx f0, BASE, TMP2
    |.else
    |   stwux CARG1, TMP2, BASE
    |   stw CARG2, 4(TMP2)
    |.endif
    |  mr TMP2, TMP3
    |  bne cr1, <2
    |3:
    |  crand 4*cr0+eq, 4*cr0+eq, 4*cr7+gt
    |  beq >5
    |4:
    |  ins_callt
    |
    |5:  // Tailcall to a fast function with a Lua frame below.
    |  lwz INS, -4(TMP1)
    |  decode_RA8 RA, INS
    |  sub TMP1, BASE, RA
    |  lwz LFUNC:TMP1, FRAME_FUNC-8(TMP1)
    |  lwz TMP1, LFUNC:TMP1->pc
    |  lwz KBASE, PC2PROTO(k)(TMP1)	// Need to prepare KBASE.
    |  b <4
    |
    |7:  // Tailcall from a vararg function.
    |  andix. TMP0, TMP2, FRAME_TYPEP
    |  bne <1				// Vararg frame below?
    |  sub BASE, BASE, TMP2		// Relocate BASE down.
    |  lwz TMP1, FRAME_PC(BASE)
    |  andix. TMP0, TMP1, FRAME_TYPE
    |  b <1
    break;

  case BC_ITERC:
    |  // RA = base*8, (RB = (nresults+1)*8, RC = (nargs+1)*8 ((2+1)*8))
    |  mr TMP2, BASE
    |  add BASE, BASE, RA
    |  lwz TMP1, -24(BASE)
    |   lwz LFUNC:RB, -20(BASE)
    |.if FPU
    |    lfd f1, -8(BASE)
    |    lfd f0, -16(BASE)
    |.else
    |    lwz CARG1, -8(BASE)
    |    lwz CARG2, -4(BASE)
    |    lwz CARG3, -16(BASE)
    |    lwz CARG4, -12(BASE)
    |.endif
    |  stw TMP1, 0(BASE)		// Copy callable.
    |   stw LFUNC:RB, 4(BASE)
    |  checkfunc TMP1
    |     li NARGS8:RC, 16		// Iterators get 2 arguments.
    |.if FPU
    |    stfd f1, 16(BASE)		// Copy control var.
    |    stfdu f0, 8(BASE)		// Copy state.
    |.else
    |    stw CARG1, 16(BASE)		// Copy control var.
    |    stw CARG2, 20(BASE)
    |    stwu CARG3, 8(BASE)		// Copy state.
    |    stw CARG4, 4(BASE)
    |.endif
    |  bne ->vmeta_call
    |  ins_call
    break;

  case BC_ITERN:
    |  // RA = base*8, (RB = (nresults+1)*8, RC = (nargs+1)*8 (2+1)*8)
    |.if JIT
    |  // NYI on big-endian
    |.endif
    |->vm_IITERN:
    |  add RA, BASE, RA
    |  lwz TAB:RB, -12(RA)
    |  lwz RC, -4(RA)			// Get index from control var.
    |  lwz TMP0, TAB:RB->asize
    |  lwz TMP1, TAB:RB->array
    |   addi PC, PC, 4
    |1:  // Traverse array part.
    |  cmplw RC, TMP0
    |   slwi TMP3, RC, 3
    |  bge >5				// Index points after array part?
    |  lwzx TMP2, TMP1, TMP3
    |.if FPU
    |   lfdx f0, TMP1, TMP3
    |.else
    |   lwzux CARG1, TMP3, TMP1
    |   lwz CARG2, 4(TMP3)
    |.endif
    |  checknil TMP2
    |     lwz INS, -4(PC)
    |  beq >4
    |.if DUALNUM
    |   stw RC, 4(RA)
    |   stw TISNUM, 0(RA)
    |.else
    |   tonum_u f1, RC
    |.endif
    |    addi RC, RC, 1
    |     addis TMP3, PC, -(BCBIAS_J*4 >> 16)
    |.if FPU
    |  stfd f0, 8(RA)
    |.else
    |  stw CARG1, 8(RA)
    |  stw CARG2, 12(RA)
    |.endif
    |     decode_RD4 TMP1, INS
    |    stw RC, -4(RA)			// Update control var.
    |     add PC, TMP1, TMP3
    |.if not DUALNUM
    |   stfd f1, 0(RA)
    |.endif
    |3:
    |  ins_next
    |
    |4:  // Skip holes in array part.
    |  addi RC, RC, 1
    |  b <1
    |
    |5:  // Traverse hash part.
    |  lwz TMP1, TAB:RB->hmask
    |  sub RC, RC, TMP0
    |   lwz TMP2, TAB:RB->node
    |6:
    |  cmplw RC, TMP1			// End of iteration? Branch to ITERL+1.
    |   slwi TMP3, RC, 5
    |  bgty <3
    |   slwi RB, RC, 3
    |   sub TMP3, TMP3, RB
    |  lwzx RB, TMP2, TMP3
    |.if FPU
    |  lfdx f0, TMP2, TMP3
    |.else
    |  add CARG3, TMP2, TMP3
    |  lwz CARG1, 0(CARG3)
    |  lwz CARG2, 4(CARG3)
    |.endif
    |   add NODE:TMP3, TMP2, TMP3
    |  checknil RB
    |     lwz INS, -4(PC)
    |  beq >7
    |.if FPU
    |   lfd f1, NODE:TMP3->key
    |.else
    |   lwz CARG3, NODE:TMP3->key.u32.hi
    |   lwz CARG4, NODE:TMP3->key.u32.lo
    |.endif
    |     addis TMP2, PC, -(BCBIAS_J*4 >> 16)
    |.if FPU
    |  stfd f0, 8(RA)
    |.else
    |  stw CARG1, 8(RA)
    |  stw CARG2, 12(RA)
    |.endif
    |    add RC, RC, TMP0
    |     decode_RD4 TMP1, INS
    |.if FPU
    |   stfd f1, 0(RA)
    |.else
    |   stw CARG3, 0(RA)
    |   stw CARG4, 4(RA)
    |.endif
    |    addi RC, RC, 1
    |     add PC, TMP1, TMP2
    |    stw RC, -4(RA)			// Update control var.
    |  b <3
    |
    |7:  // Skip holes in hash part.
    |  addi RC, RC, 1
    |  b <6
    break;

  case BC_ISNEXT:
    |  // RA = base*8, RD = target (points to ITERN)
    |  add RA, BASE, RA
    |  lwz TMP0, -24(RA)
    |  lwz CFUNC:TMP1, -20(RA)
    |   lwz TMP2, -16(RA)
    |    lwz TMP3, -8(RA)
    |   cmpwi cr0, TMP2, LJ_TTAB
    |  cmpwi cr1, TMP0, LJ_TFUNC
    |    cmpwi cr6, TMP3, LJ_TNIL
    |  bne cr1, >5
    |  lbz TMP1, CFUNC:TMP1->ffid
    |   crand 4*cr0+eq, 4*cr0+eq, 4*cr6+eq
    |  cmpwi cr7, TMP1, FF_next_N
    |    srwi TMP0, RD, 1
    |  crand 4*cr0+eq, 4*cr0+eq, 4*cr7+eq
    |    add TMP3, PC, TMP0
    |  bne cr0, >5
    |  lus TMP1, (LJ_KEYINDEX >> 16)
    |  ori TMP1, TMP1, (LJ_KEYINDEX & 0xffff)
    |  stw ZERO, -4(RA)			// Initialize control var.
    |  stw TMP1, -8(RA)
    |    addis PC, TMP3, -(BCBIAS_J*4 >> 16)
    |1:
    |  ins_next
    |5:  // Despecialize bytecode if any of the checks fail.
    |  li TMP0, BC_JMP
    |   li TMP1, BC_ITERC
    |  stb TMP0, -1(PC)
    |    addis PC, TMP3, -(BCBIAS_J*4 >> 16)
    |  // NYI on big-endian: unpatch JLOOP.
    |   stb TMP1, 3(PC)
    |  b <1
    break;

  case BC_VARG:
    |  // RA = base*8, RB = (nresults+1)*8, RC = numparams*8
    |  lwz TMP0, FRAME_PC(BASE)
    |  add RC, BASE, RC
    |   add RA, BASE, RA
    |  addi RC, RC, FRAME_VARG
    |   add TMP2, RA, RB
    |  subi TMP3, BASE, 8		// TMP3 = vtop
    |  sub RC, RC, TMP0			// RC = vbase
    |  // Note: RC may now be even _above_ BASE if nargs was < numparams.
    |  cmplwi cr1, RB, 0
    |.if PPE
    |   sub TMP1, TMP3, RC
    |   cmpwi TMP1, 0
    |.else
    |   sub. TMP1, TMP3, RC
    |.endif
    |  beq cr1, >5			// Copy all varargs?
    |   subi TMP2, TMP2, 16
    |   ble >2				// No vararg slots?
    |1:  // Copy vararg slots to destination slots.
    |.if FPU
    |  lfd f0, 0(RC)
    |.else
    |  lwz CARG1, 0(RC)
    |  lwz CARG2, 4(RC)
    |.endif
    |   addi RC, RC, 8
    |.if FPU
    |  stfd f0, 0(RA)
    |.else
    |  stw CARG1, 0(RA)
    |  stw CARG2, 4(RA)
    |.endif
    |  cmplw RA, TMP2
    |   cmplw cr1, RC, TMP3
    |  bge >3				// All destination slots filled?
    |    addi RA, RA, 8
    |   blt cr1, <1			// More vararg slots?
    |2:  // Fill up remainder with nil.
    |  stw TISNIL, 0(RA)
    |  cmplw RA, TMP2
    |   addi RA, RA, 8
    |  blt <2
    |3:
    |  ins_next
    |
    |5:  // Copy all varargs.
    |  lwz TMP0, L->maxstack
    |   li MULTRES, 8			// MULTRES = (0+1)*8
    |  bley <3				// No vararg slots?
    |  add TMP2, RA, TMP1
    |  cmplw TMP2, TMP0
    |   addi MULTRES, TMP1, 8
    |  bgt >7
    |6:
    |.if FPU
    |  lfd f0, 0(RC)
    |.else
    |  lwz CARG1, 0(RC)
    |  lwz CARG2, 4(RC)
    |.endif
    |   addi RC, RC, 8
    |.if FPU
    |  stfd f0, 0(RA)
    |.else
    |  stw CARG1, 0(RA)
    |  stw CARG2, 4(RA)
    |.endif
    |  cmplw RC, TMP3
    |   addi RA, RA, 8
    |  blt <6				// More vararg slots?
    |  b <3
    |
    |7:  // Grow stack for varargs.
    |  mr CARG1, L
    |   stp RA, L->top
    |  sub SAVE0, RC, BASE		// Need delta, because BASE may change.
    |   stp BASE, L->base
    |  sub RA, RA, BASE
    |   stw PC, SAVE_PC
    |  srwi CARG2, TMP1, 3
    |  bl extern lj_state_growstack	// (lua_State *L, int n)
    |  lp BASE, L->base
    |  add RA, BASE, RA
    |  add RC, BASE, SAVE0
    |  subi TMP3, BASE, 8
    |  b <6
    break;

  /* -- Returns ----------------------------------------------------------- */

  case BC_RETM:
    |  // RA = results*8, RD = extra_nresults*8
    |  add RD, RD, MULTRES		// MULTRES >= 8, so RD >= 8.
    |  // Fall through. Assumes BC_RET follows.
    break;

  case BC_RET:
    |  // RA = results*8, RD = (nresults+1)*8
    |  lwz PC, FRAME_PC(BASE)
    |   add RA, BASE, RA
    |    mr MULTRES, RD
    |1:
    |  andix. TMP0, PC, FRAME_TYPE
    |   xori TMP1, PC, FRAME_VARG
    |  bne ->BC_RETV_Z
    |
    |->BC_RET_Z:
    |  // BASE = base, RA = resultptr, RD = (nresults+1)*8, PC = return
    |   lwz INS, -4(PC)
    |  cmpwi RD, 8
    |   subi TMP2, BASE, 8
    |   subi RC, RD, 8
    |   decode_RB8 RB, INS
    |  beq >3
    |   li TMP1, 0
    |2:
    |  addi TMP3, TMP1, 8
    |.if FPU
    |   lfdx f0, RA, TMP1
    |.else
    |   add CARG3, RA, TMP1
    |   lwz CARG1, 0(CARG3)
    |   lwz CARG2, 4(CARG3)
    |.endif
    |  cmpw TMP3, RC
    |.if FPU
    |   stfdx f0, TMP2, TMP1
    |.else
    |   add CARG3, TMP2, TMP1
    |   stw CARG1, 0(CARG3)
    |   stw CARG2, 4(CARG3)
    |.endif
    |  beq >3
    |  addi TMP1, TMP3, 8
    |.if FPU
    |   lfdx f1, RA, TMP3
    |.else
    |   add CARG3, RA, TMP3
    |   lwz CARG1, 0(CARG3)
    |   lwz CARG2, 4(CARG3)
    |.endif
    |  cmpw TMP1, RC
    |.if FPU
    |   stfdx f1, TMP2, TMP3
    |.else
    |   add CARG3, TMP2, TMP3
    |   stw CARG1, 0(CARG3)
    |   stw CARG2, 4(CARG3)
    |.endif
    |  bne <2
    |3:
    |5:
    |  cmplw RB, RD
    |   decode_RA8 RA, INS
    |  bgt >6
    |   sub BASE, TMP2, RA
    |  lwz LFUNC:TMP1, FRAME_FUNC(BASE)
    |  ins_next1
    |  lwz TMP1, LFUNC:TMP1->pc
    |  lwz KBASE, PC2PROTO(k)(TMP1)
    |  ins_next2
    |
    |6:  // Fill up results with nil.
    |  subi TMP1, RD, 8
    |   addi RD, RD, 8
    |  stwx TISNIL, TMP2, TMP1
    |  b <5
    |
    |->BC_RETV_Z:  // Non-standard return case.
    |  andix. TMP2, TMP1, FRAME_TYPEP
    |  bne ->vm_return
    |  // Return from vararg function: relocate BASE down.
    |  sub BASE, BASE, TMP1
    |  lwz PC, FRAME_PC(BASE)
    |  b <1
    break;

  case BC_RET0: case BC_RET1:
    |  // RA = results*8, RD = (nresults+1)*8
    |  lwz PC, FRAME_PC(BASE)
    |   add RA, BASE, RA
    |    mr MULTRES, RD
    |  andix. TMP0, PC, FRAME_TYPE
    |   xori TMP1, PC, FRAME_VARG
    |  bney ->BC_RETV_Z
    |
    |  lwz INS, -4(PC)
    |   subi TMP2, BASE, 8
    |  decode_RB8 RB, INS
    if (op == BC_RET1) {
      |.if FPU
      |  lfd f0, 0(RA)
      |  stfd f0, 0(TMP2)
      |.else
      |  lwz CARG1, 0(RA)
      |  lwz CARG2, 4(RA)
      |  stw CARG1, 0(TMP2)
      |  stw CARG2, 4(TMP2)
      |.endif
    }
    |5:
    |  cmplw RB, RD
    |   decode_RA8 RA, INS
    |  bgt >6
    |   sub BASE, TMP2, RA
    |  lwz LFUNC:TMP1, FRAME_FUNC(BASE)
    |  ins_next1
    |  lwz TMP1, LFUNC:TMP1->pc
    |  lwz KBASE, PC2PROTO(k)(TMP1)
    |  ins_next2
    |
    |6:  // Fill up results with nil.
    |  subi TMP1, RD, 8
    |   addi RD, RD, 8
    |  stwx TISNIL, TMP2, TMP1
    |  b <5
    break;

  /* -- Loops and branches ------------------------------------------------ */

  case BC_FORL:
    |.if JIT
    |  hotloop
    |.endif
    |  // Fall through. Assumes BC_IFORL follows.
    break;

  case BC_JFORI:
  case BC_JFORL:
#if !LJ_HASJIT
    break;
#endif
  case BC_FORI:
  case BC_IFORL:
    |  // RA = base*8, RD = target (after end of loop or start of loop)
    vk = (op == BC_IFORL || op == BC_JFORL);
    |.if DUALNUM
    |  // Integer loop.
    |  lwzux TMP1, RA, BASE
    |   lwz CARG1, FORL_IDX*8+4(RA)
    |  cmplw cr0, TMP1, TISNUM
    if (vk) {
      |   lwz CARG3, FORL_STEP*8+4(RA)
      |  bne >9
      |.if GPR64
      |  // Need to check overflow for (a<<32) + (b<<32).
      |  rldicr TMP0, CARG1, 32, 31
      |  rldicr TMP2, CARG3, 32, 31
      |  add CARG1, CARG1, CARG3
      |  addo. TMP0, TMP0, TMP2
      |.else
      |  addo. CARG1, CARG1, CARG3
      |.endif
      |    cmpwi cr6, CARG3, 0
      |   lwz CARG2, FORL_STOP*8+4(RA)
      |  bso >6
      |4:
      |  stw CARG1, FORL_IDX*8+4(RA)
    } else {
      |  lwz SAVE0, FORL_STEP*8(RA)
      |   lwz CARG3, FORL_STEP*8+4(RA)
      |  lwz TMP2, FORL_STOP*8(RA)
      |   lwz CARG2, FORL_STOP*8+4(RA)
      |  cmplw cr7, SAVE0, TISNUM
      |  cmplw cr1, TMP2, TISNUM
      |  crand 4*cr0+eq, 4*cr0+eq, 4*cr7+eq
      |  crand 4*cr0+eq, 4*cr0+eq, 4*cr1+eq
      |    cmpwi cr6, CARG3, 0
      |  bne >9
    }
    |    blt cr6, >5
    |  cmpw CARG1, CARG2
    |1:
    |   stw TISNUM, FORL_EXT*8(RA)
    if (op != BC_JFORL) {
      |  srwi RD, RD, 1
    }
    |   stw CARG1, FORL_EXT*8+4(RA)
    if (op != BC_JFORL) {
      |  add RD, PC, RD
    }
    if (op == BC_FORI) {
      |  bgt >3  // See FP loop below.
    } else if (op == BC_JFORI) {
      |  addis PC, RD, -(BCBIAS_J*4 >> 16)
      |  bley >7
    } else if (op == BC_IFORL) {
      |  bgt >2
      |  addis PC, RD, -(BCBIAS_J*4 >> 16)
    } else {
      |  bley =>BC_JLOOP
    }
    |2:
    |  ins_next
    |5:  // Invert check for negative step.
    |  cmpw CARG2, CARG1
    |  b <1
    if (vk) {
      |6:  // Potential overflow.
      |  checkov TMP0, <4		// Ignore unrelated overflow.
      |  b <2
    }
    |.endif
    if (vk) {
      |.if DUALNUM
      |9:  // FP loop.
      |.if FPU
      |  lfd f1, FORL_IDX*8(RA)
      |.else
      |  lwz CARG1, FORL_IDX*8(RA)
      |  lwz CARG2, FORL_IDX*8+4(RA)
      |.endif
      |.else
      |  lfdux f1, RA, BASE
      |.endif
      |.if FPU
      |  lfd f3, FORL_STEP*8(RA)
      |  lfd f2, FORL_STOP*8(RA)
      |  fadd f1, f1, f3
      |  stfd f1, FORL_IDX*8(RA)
      |.else
      |  lwz CARG3, FORL_STEP*8(RA)
      |  lwz CARG4, FORL_STEP*8+4(RA)
      |  mr SAVE1, RD
      |  blex __adddf3
      |  mr RD, SAVE1
      |  stw CRET1, FORL_IDX*8(RA)
      |  stw CRET2, FORL_IDX*8+4(RA)
      |  lwz CARG3, FORL_STOP*8(RA)
      |  lwz CARG4, FORL_STOP*8+4(RA)
      |.endif
      |   lwz SAVE0, FORL_STEP*8(RA)
    } else {
      |.if DUALNUM
      |9:  // FP loop.
      |.else
      |  lwzux TMP1, RA, BASE
      |  lwz SAVE0, FORL_STEP*8(RA)
      |  lwz TMP2, FORL_STOP*8(RA)
      |  cmplw cr0, TMP1, TISNUM
      |  cmplw cr7, SAVE0, TISNUM
      |  cmplw cr1, TMP2, TISNUM
      |.endif
      |.if FPU
      |   lfd f1, FORL_IDX*8(RA)
      |.else
      |   lwz CARG1, FORL_IDX*8(RA)
      |   lwz CARG2, FORL_IDX*8+4(RA)
      |.endif
      |  crand 4*cr0+lt, 4*cr0+lt, 4*cr7+lt
      |  crand 4*cr0+lt, 4*cr0+lt, 4*cr1+lt
      |.if FPU
      |   lfd f2, FORL_STOP*8(RA)
      |.else
      |   lwz CARG3, FORL_STOP*8(RA)
      |   lwz CARG4, FORL_STOP*8+4(RA)
      |.endif
      |  bge ->vmeta_for
    }
    |  cmpwi cr6, SAVE0, 0
    if (op != BC_JFORL) {
      |  srwi RD, RD, 1
    }
    |.if FPU
    |   stfd f1, FORL_EXT*8(RA)
    |.else
    |   stw CARG1, FORL_EXT*8(RA)
    |   stw CARG2, FORL_EXT*8+4(RA)
    |.endif
    if (op != BC_JFORL) {
      |  add RD, PC, RD
    }
    |.if FPU
    |  fcmpu cr0, f1, f2
    |.else
    |  mr SAVE1, RD
    |  blex __ledf2
    |  cmpwi CRET1, 0
    |  mr RD, SAVE1
    |.endif
    if (op == BC_JFORI) {
      |  addis PC, RD, -(BCBIAS_J*4 >> 16)
    }
    |  blt cr6, >5
    if (op == BC_FORI) {
      |  bgt >3
    } else if (op == BC_IFORL) {
      |.if DUALNUM
      |  bgty <2
      |.else
      |  bgt >2
      |.endif
      |1:
      |  addis PC, RD, -(BCBIAS_J*4 >> 16)
    } else if (op == BC_JFORI) {
      |  bley >7
    } else {
      |  bley =>BC_JLOOP
    }
    |.if DUALNUM
    |  b <2
    |.else
    |2:
    |  ins_next
    |.endif
    |5:  // Negative step.
    if (op == BC_FORI) {
      |  bge <2
      |3:  // Used by integer loop, too.
      |  addis PC, RD, -(BCBIAS_J*4 >> 16)
    } else if (op == BC_IFORL) {
      |  bgey <1
    } else if (op == BC_JFORI) {
      |  bgey >7
    } else {
      |  bgey =>BC_JLOOP
    }
    |  b <2
    if (op == BC_JFORI) {
      |7:
      |  lwz INS, -4(PC)
      |  decode_RD8 RD, INS
      |  b =>BC_JLOOP
    }
    break;

  case BC_ITERL:
    |.if JIT
    |  hotloop
    |.endif
    |  // Fall through. Assumes BC_IITERL follows.
    break;

  case BC_JITERL:
#if !LJ_HASJIT
    break;
#endif
  case BC_IITERL:
    |  // RA = base*8, RD = target
    |  lwzux TMP1, RA, BASE
    |   lwz TMP2, 4(RA)
    |  checknil TMP1; beq >1		// Stop if iterator returned nil.
    if (op == BC_JITERL) {
      |  stw TMP1, -8(RA)
      |   stw TMP2, -4(RA)
      |  b =>BC_JLOOP
    } else {
      |  branch_RD			// Otherwise save control var + branch.
      |  stw TMP1, -8(RA)
      |   stw TMP2, -4(RA)
    }
    |1:
    |  ins_next
    break;

  case BC_LOOP:
    |  // RA = base*8, RD = target (loop extent)
    |  // Note: RA/RD is only used by trace recorder to determine scope/extent
    |  // This opcode does NOT jump, it's only purpose is to detect a hot loop.
    |.if JIT
    |  hotloop
    |.endif
    |  // Fall through. Assumes BC_ILOOP follows.
    break;

  case BC_ILOOP:
    |  // RA = base*8, RD = target (loop extent)
    |  ins_next
    break;

  case BC_JLOOP:
    |.if JIT
    |  // RA = base*8 (ignored), RD = traceno*8
    |  lwz TMP1, DISPATCH_J(trace)(DISPATCH)
    |  srwi RD, RD, 1
    |  // Traces on PPC don't store the trace number, so use 0.
    |   stw ZERO, DISPATCH_GL(vmstate)(DISPATCH)
    |  lwzx TRACE:TMP2, TMP1, RD
    |  clrso TMP1
    |  lp TMP2, TRACE:TMP2->mcode
    |   stw BASE, DISPATCH_GL(jit_base)(DISPATCH)
    |  mtctr TMP2
    |   addi JGL, DISPATCH, GG_DISP2G+32768
    |   stw L, DISPATCH_GL(tmpbuf.L)(DISPATCH)
    |  bctr
    |.endif
    break;

  case BC_JMP:
    |  // RA = base*8 (only used by trace recorder), RD = target
    |  branch_RD
    |  ins_next
    break;

  /* -- Function headers -------------------------------------------------- */

  case BC_FUNCF:
    |.if JIT
    |  hotcall
    |.endif
  case BC_FUNCV:  /* NYI: compiled vararg functions. */
    |  // Fall through. Assumes BC_IFUNCF/BC_IFUNCV follow.
    break;

  case BC_JFUNCF:
#if !LJ_HASJIT
    break;
#endif
  case BC_IFUNCF:
    |  // BASE = new base, RA = BASE+framesize*8, RB = LFUNC, RC = nargs*8
    |  lwz TMP2, L->maxstack
    |   lbz TMP1, -4+PC2PROTO(numparams)(PC)
    |    lwz KBASE, -4+PC2PROTO(k)(PC)
    |  cmplw RA, TMP2
    |   slwi TMP1, TMP1, 3
    |  bgt ->vm_growstack_l
    if (op != BC_JFUNCF) {
      |  ins_next1
    }
    |2:
    |  cmplw NARGS8:RC, TMP1		// Check for missing parameters.
    |  blt >3
    if (op == BC_JFUNCF) {
      |  decode_RD8 RD, INS
      |  b =>BC_JLOOP
    } else {
      |  ins_next2
    }
    |
    |3:  // Clear missing parameters.
    |  stwx TISNIL, BASE, NARGS8:RC
    |  addi NARGS8:RC, NARGS8:RC, 8
    |  b <2
    break;

  case BC_JFUNCV:
#if !LJ_HASJIT
    break;
#endif
    |  NYI  // NYI: compiled vararg functions
    break;  /* NYI: compiled vararg functions. */

  case BC_IFUNCV:
    |  // BASE = new base, RA = BASE+framesize*8, RB = LFUNC, RC = nargs*8
    |  lwz TMP2, L->maxstack
    |   add TMP1, BASE, RC
    |  add TMP0, RA, RC
    |   stw LFUNC:RB, 4(TMP1)		// Store copy of LFUNC.
    |   addi TMP3, RC, 8+FRAME_VARG
    |    lwz KBASE, -4+PC2PROTO(k)(PC)
    |  cmplw TMP0, TMP2
    |   stw TMP3, 0(TMP1)		// Store delta + FRAME_VARG.
    |  bge ->vm_growstack_l
    |  lbz TMP2, -4+PC2PROTO(numparams)(PC)
    |   mr RA, BASE
    |   mr RC, TMP1
    |  ins_next1
    |  cmpwi TMP2, 0
    |   addi BASE, TMP1, 8
    |  beq >3
    |1:
    |  cmplw RA, RC			// Less args than parameters?
    |   lwz TMP0, 0(RA)
    |   lwz TMP3, 4(RA)
    |  bge >4
    |    stw TISNIL, 0(RA)		// Clear old fixarg slot (help the GC).
    |    addi RA, RA, 8
    |2:
    |  addic. TMP2, TMP2, -1
    |   stw TMP0, 8(TMP1)
    |   stw TMP3, 12(TMP1)
    |    addi TMP1, TMP1, 8
    |  bne <1
    |3:
    |  ins_next2
    |
    |4:  // Clear missing parameters.
    |  li TMP0, LJ_TNIL
    |  b <2
    break;

  case BC_FUNCC:
  case BC_FUNCCW:
    |  // BASE = new base, RA = BASE+framesize*8, RB = CFUNC, RC = nargs*8
    if (op == BC_FUNCC) {
      |  lp RD, CFUNC:RB->f
    } else {
      |  lp RD, DISPATCH_GL(wrapf)(DISPATCH)
    }
    |   add TMP1, RA, NARGS8:RC
    |   lwz TMP2, L->maxstack
    |  .toc lp TMP3, 0(RD)
    |    add RC, BASE, NARGS8:RC
    |   stp BASE, L->base
    |   cmplw TMP1, TMP2
    |    stp RC, L->top
    |     li_vmstate C
    |.if TOC
    |  mtctr TMP3
    |.else
    |  mtctr RD
    |.endif
    if (op == BC_FUNCCW) {
      |  lp CARG2, CFUNC:RB->f
    }
    |  mr CARG1, L
    |   bgt ->vm_growstack_c		// Need to grow stack.
    |  .toc lp TOCREG, TOC_OFS(RD)
    |  .tocenv lp ENVREG, ENV_OFS(RD)
    |     st_vmstate
    |  bctrl				// (lua_State *L [, lua_CFunction f])
    |  // Returns nresults.
    |  lp BASE, L->base
    |  .toc ld TOCREG, SAVE_TOC
    |   slwi RD, CRET1, 3
    |  lp TMP1, L->top
    |    li_vmstate INTERP
    |  lwz PC, FRAME_PC(BASE)		// Fetch PC of caller.
    |    stw L, DISPATCH_GL(cur_L)(DISPATCH)
    |   sub RA, TMP1, RD		// RA = L->top - nresults*8
    |    st_vmstate
    |  b ->vm_returnc
    break;

  /* ---------------------------------------------------------------------- */

  default:
    fprintf(stderr, "Error: undefined opcode BC_%s\n", bc_names[op]);
    exit(2);
    break;
  }
}

static int build_backend(BuildCtx *ctx)
{
  int op;

  dasm_growpc(Dst, BC__MAX);

  build_subroutines(ctx);

  |.code_op
  for (op = 0; op < BC__MAX; op++)
    build_ins(ctx, (BCOp)op, op);

  return BC__MAX;
}

/* Emit pseudo frame-info for all assembler functions. */
static void emit_asm_debug(BuildCtx *ctx)
{
  int fcofs = (int)((uint8_t *)ctx->glob[GLOB_vm_ffi_call] - ctx->code);
  int i;
  switch (ctx->mode) {
  case BUILD_elfasm:
    fprintf(ctx->fp, "\t.section .debug_frame,\"\",@progbits\n");
    fprintf(ctx->fp,
	".Lframe0:\n"
	"\t.long .LECIE0-.LSCIE0\n"
	".LSCIE0:\n"
	"\t.long 0xffffffff\n"
	"\t.byte 0x1\n"
	"\t.string \"\"\n"
	"\t.uleb128 0x1\n"
	"\t.sleb128 -4\n"
	"\t.byte 65\n"
	"\t.byte 0xc\n\t.uleb128 1\n\t.uleb128 0\n"
	"\t.align 2\n"
	".LECIE0:\n\n");
    fprintf(ctx->fp,
	".LSFDE0:\n"
	"\t.long .LEFDE0-.LASFDE0\n"
	".LASFDE0:\n"
	"\t.long .Lframe0\n"
	"\t.long .Lbegin\n"
	"\t.long %d\n"
	"\t.byte 0xe\n\t.uleb128 %d\n"
	"\t.byte 0x11\n\t.uleb128 65\n\t.sleb128 -1\n"
	"\t.byte 0x5\n\t.uleb128 70\n\t.uleb128 55\n",
	fcofs, CFRAME_SIZE);
    for (i = 14; i <= 31; i++)
      fprintf(ctx->fp,
	"\t.byte %d\n\t.uleb128 %d\n"
	"\t.byte %d\n\t.uleb128 %d\n",
	0x80+i, 37+(31-i), 0x80+32+i, 2+2*(31-i));
    fprintf(ctx->fp,
	"\t.align 2\n"
	".LEFDE0:\n\n");
#if LJ_HASFFI
    fprintf(ctx->fp,
	".LSFDE1:\n"
	"\t.long .LEFDE1-.LASFDE1\n"
	".LASFDE1:\n"
	"\t.long .Lframe0\n"
#if LJ_TARGET_PS3
	"\t.long .lj_vm_ffi_call\n"
#else
	"\t.long lj_vm_ffi_call\n"
#endif
	"\t.long %d\n"
	"\t.byte 0x11\n\t.uleb128 65\n\t.sleb128 -1\n"
	"\t.byte 0x8e\n\t.uleb128 2\n"
	"\t.byte 0xd\n\t.uleb128 0xe\n"
	"\t.align 2\n"
	".LEFDE1:\n\n", (int)ctx->codesz - fcofs);
#endif
#if !LJ_NO_UNWIND
    fprintf(ctx->fp, "\t.section .eh_frame,\"a\",@progbits\n");
    fprintf(ctx->fp,
	".Lframe1:\n"
	"\t.long .LECIE1-.LSCIE1\n"
	".LSCIE1:\n"
	"\t.long 0\n"
	"\t.byte 0x1\n"
	"\t.string \"zPR\"\n"
	"\t.uleb128 0x1\n"
	"\t.sleb128 -4\n"
	"\t.byte 65\n"
	"\t.uleb128 6\n"			/* augmentation length */
	"\t.byte 0x1b\n"			/* pcrel|sdata4 */
	"\t.long lj_err_unwind_dwarf-.\n"
	"\t.byte 0x1b\n"			/* pcrel|sdata4 */
	"\t.byte 0xc\n\t.uleb128 1\n\t.uleb128 0\n"
	"\t.align 2\n"
	".LECIE1:\n\n");
    fprintf(ctx->fp,
	".LSFDE2:\n"
	"\t.long .LEFDE2-.LASFDE2\n"
	".LASFDE2:\n"
	"\t.long .LASFDE2-.Lframe1\n"
	"\t.long .Lbegin-.\n"
	"\t.long %d\n"
	"\t.uleb128 0\n"			/* augmentation length */
	"\t.byte 0xe\n\t.uleb128 %d\n"
	"\t.byte 0x11\n\t.uleb128 65\n\t.sleb128 -1\n"
	"\t.byte 0x5\n\t.uleb128 70\n\t.uleb128 55\n",
	fcofs, CFRAME_SIZE);
    for (i = 14; i <= 31; i++)
      fprintf(ctx->fp,
	"\t.byte %d\n\t.uleb128 %d\n"
	"\t.byte %d\n\t.uleb128 %d\n",
	0x80+i, 37+(31-i), 0x80+32+i, 2+2*(31-i));
    fprintf(ctx->fp,
	"\t.align 2\n"
	".LEFDE2:\n\n");
#if LJ_HASFFI
    fprintf(ctx->fp,
	".Lframe2:\n"
	"\t.long .LECIE2-.LSCIE2\n"
	".LSCIE2:\n"
	"\t.long 0\n"
	"\t.byte 0x1\n"
	"\t.string \"zR\"\n"
	"\t.uleb128 0x1\n"
	"\t.sleb128 -4\n"
	"\t.byte 65\n"
	"\t.uleb128 1\n"			/* augmentation length */
	"\t.byte 0x1b\n"			/* pcrel|sdata4 */
	"\t.byte 0xc\n\t.uleb128 1\n\t.uleb128 0\n"
	"\t.align 2\n"
	".LECIE2:\n\n");
    fprintf(ctx->fp,
	".LSFDE3:\n"
	"\t.long .LEFDE3-.LASFDE3\n"
	".LASFDE3:\n"
	"\t.long .LASFDE3-.Lframe2\n"
	"\t.long lj_vm_ffi_call-.\n"
	"\t.long %d\n"
	"\t.uleb128 0\n"			/* augmentation length */
	"\t.byte 0x11\n\t.uleb128 65\n\t.sleb128 -1\n"
	"\t.byte 0x8e\n\t.uleb128 2\n"
	"\t.byte 0xd\n\t.uleb128 0xe\n"
	"\t.align 2\n"
	".LEFDE3:\n\n", (int)ctx->codesz - fcofs);
#endif
#endif
    break;
  default:
    break;
  }
}