Skip to content

A lightweight WebAssembly runtime that is fast, secure, and standards-compliant

Cranelift: use SP-offset amodes for `stack_addr`+load/store. (#11727)

Chris Fallin62dfbd60a653
diff --git a/cranelift/codegen/src/isa/aarch64/inst.isle b/cranelift/codegen/src/isa/aarch64/inst.isle
index a9b3bdc236..9251f4d8b1 100644
--- a/cranelift/codegen/src/isa/aarch64/inst.isle
+++ b/cranelift/codegen/src/isa/aarch64/inst.isle
@@ -3899,6 +3899,11 @@
(if-let new_offset (i32_checked_add x offset))
(amode_no_more_iconst ty y new_offset))
+(rule 3
+ (amode ty (stack_addr slot offset1) offset2)
+ (AMode.SlotOffset
+ (abi_stackslot_offset_into_slot_region slot offset1 offset2)))
+
(decl amode_no_more_iconst (Type Value i32) AMode)
;; Base case: move the `offset` into a register and add it to `val` via the
;; amode
diff --git a/cranelift/codegen/src/isa/x64/inst.isle b/cranelift/codegen/src/isa/x64/inst.isle
index 242db5a0e3..454a08637c 100644
--- a/cranelift/codegen/src/isa/x64/inst.isle
+++ b/cranelift/codegen/src/isa/x64/inst.isle
@@ -403,6 +403,15 @@
(decl amode_to_synthetic_amode (Amode) SyntheticAmode)
(extern constructor amode_to_synthetic_amode amode_to_synthetic_amode)
+(decl synthetic_amode_slot (i32) SyntheticAmode)
+(extern constructor synthetic_amode_slot synthetic_amode_slot)
+
+;; Helper for loads/stores to/from stackslots.
+(decl stackslot_amode (StackSlot Offset32 Offset32) SyntheticAmode)
+(rule (stackslot_amode slot offset1 offset2)
+ (let ((slot_offset i32 (abi_stackslot_offset_into_slot_region slot offset1 offset2)))
+ (synthetic_amode_slot slot_offset)))
+
;; An `Amode` represents a possible addressing mode that can be used
;; in instructions. These denote a 64-bit value only.
(type Amode (enum
@@ -494,12 +503,16 @@
(provide (= result (concat flags (bvadd val (sign_ext 64 offset)))))
(require
(= (widthof val) 64)))
-(decl to_amode (MemFlags Value Offset32) Amode)
+(decl to_amode (MemFlags Value Offset32) SyntheticAmode)
(rule 0 (to_amode flags base offset)
(amode_imm_reg flags base offset))
(rule 1 (to_amode flags (iadd x y) offset)
(to_amode_add flags x y offset))
+(rule 2
+ (to_amode flags (stack_addr slot offset1) offset2)
+ (stackslot_amode slot offset1 offset2))
+
;; Same as `to_amode`, except that the base address is computed via the addition
;; of the two `Value` arguments provided.
;;
@@ -572,7 +585,7 @@
;; Offsetting an Amode. Used when we need to do consecutive
;; loads/stores to adjacent addresses.
-(decl amode_offset (Amode i32) Amode)
+(decl amode_offset (SyntheticAmode i32) SyntheticAmode)
(extern constructor amode_offset amode_offset)
;; Return a zero offset as an `Offset32`.
@@ -1368,7 +1381,7 @@
(rule 0 (x64_load (multi_lane _bits _lanes) addr _ext_kind)
(x64_movdqu_load addr))
-(decl x64_mov (Amode) Reg)
+(decl x64_mov (SyntheticAmode) Reg)
(spec (x64_mov addr)
(provide (= result (conv_to 64 (load_effect (extract 79 64 addr) 64 (extract 63 0 addr))))))
(rule (x64_mov addr) (x64_movq_rm addr))
@@ -2701,25 +2714,25 @@
(decl x64_pextrb (Xmm u8) Gpr)
(rule (x64_pextrb src lane) (x64_pextrb_a_or_avx src lane))
-(decl x64_pextrb_store (Amode Xmm u8) SideEffectNoResult)
+(decl x64_pextrb_store (SyntheticAmode Xmm u8) SideEffectNoResult)
(rule (x64_pextrb_store addr src lane) (x64_pextrb_a_mem_or_avx addr src lane))
(decl x64_pextrw (Xmm u8) Gpr)
(rule (x64_pextrw src lane) (x64_pextrw_a_or_avx src lane))
-(decl x64_pextrw_store (Amode Xmm u8) SideEffectNoResult)
+(decl x64_pextrw_store (SyntheticAmode Xmm u8) SideEffectNoResult)
(rule (x64_pextrw_store addr src lane) (x64_pextrw_b_mem_or_avx addr src lane))
(decl x64_pextrd (Xmm u8) Gpr)
(rule (x64_pextrd src lane) (x64_pextrd_a_or_avx src lane))
-(decl x64_pextrd_store (Amode Xmm u8) SideEffectNoResult)
+(decl x64_pextrd_store (SyntheticAmode Xmm u8) SideEffectNoResult)
(rule (x64_pextrd_store addr src lane) (x64_pextrd_a_mem_or_avx addr src lane))
(decl x64_pextrq (Xmm u8) Gpr)
(rule (x64_pextrq src lane) (x64_pextrq_a_or_avx src lane))
-(decl x64_pextrq_store (Amode Xmm u8) SideEffectNoResult)
+(decl x64_pextrq_store (SyntheticAmode Xmm u8) SideEffectNoResult)
(rule (x64_pextrq_store addr src lane) (x64_pextrq_a_mem_or_avx addr src lane))
;; Helper for creating `pmovmskb` instructions.
@@ -3181,7 +3194,7 @@
)
(x64_por low_gt_and_high_eq high_halves_gt)))
-(decl x64_add_mem (Type Amode Value) SideEffectNoResult)
+(decl x64_add_mem (Type SyntheticAmode Value) SideEffectNoResult)
(spec (x64_add_mem ty addr val)
(provide (= result (store_effect
(extract 79 64 addr)
@@ -3207,7 +3220,7 @@
(rule 2 (x64_add_mem $I32 addr (i8_from_iconst val)) (x64_addl_mi_sxb_mem addr val))
(rule 2 (x64_add_mem $I64 addr (i8_from_iconst val)) (x64_addq_mi_sxb_mem addr val))
-(decl x64_sub_mem (Type Amode Value) SideEffectNoResult)
+(decl x64_sub_mem (Type SyntheticAmode Value) SideEffectNoResult)
;; `sub mem, reg`
(rule 0 (x64_sub_mem $I8 addr val) (x64_subb_mr_mem addr val))
@@ -3223,7 +3236,7 @@
(rule 2 (x64_sub_mem $I32 addr (i8_from_iconst val)) (x64_subl_mi_sxb_mem addr val))
(rule 2 (x64_sub_mem $I64 addr (i8_from_iconst val)) (x64_subq_mi_sxb_mem addr val))
-(decl x64_and_mem (Type Amode Value) SideEffectNoResult)
+(decl x64_and_mem (Type SyntheticAmode Value) SideEffectNoResult)
;; `and mem, imm`
(rule (x64_and_mem $I8 addr val) (x64_andb_mr_mem addr val))
@@ -3241,7 +3254,7 @@
(rule 2 (x64_and_mem $I32 addr (i8_from_iconst val)) (x64_andl_mi_sxb_mem addr val))
(rule 2 (x64_and_mem $I64 addr (i8_from_iconst val)) (x64_andq_mi_sxb_mem addr val))
-(decl x64_or_mem (Type Amode Value) SideEffectNoResult)
+(decl x64_or_mem (Type SyntheticAmode Value) SideEffectNoResult)
;; `or mem, reg`
(rule 0 (x64_or_mem $I8 addr val) (x64_orb_mr_mem addr val))
@@ -3259,7 +3272,7 @@
(rule 2 (x64_or_mem $I32 addr (i8_from_iconst val)) (x64_orl_mi_sxb_mem addr val))
(rule 2 (x64_or_mem $I64 addr (i8_from_iconst val)) (x64_orq_mi_sxb_mem addr val))
-(decl x64_xor_mem (Type Amode Value) SideEffectNoResult)
+(decl x64_xor_mem (Type SyntheticAmode Value) SideEffectNoResult)
;; `xor mem, reg`
(rule 0 (x64_xor_mem $I8 addr val) (x64_xorb_mr_mem addr val))
@@ -3644,31 +3657,31 @@
(rule (x64_xchg $I32 addr operand) (x64_xchgl_rm operand addr))
(rule (x64_xchg $I64 addr operand) (x64_xchgq_rm operand addr))
-(decl x64_lock_add (OperandSize Amode Gpr) SideEffectNoResult)
+(decl x64_lock_add (OperandSize SyntheticAmode Gpr) SideEffectNoResult)
(rule (x64_lock_add (OperandSize.Size8) addr reg) (x64_lock_addb_mr_mem addr reg))
(rule (x64_lock_add (OperandSize.Size16) addr reg) (x64_lock_addw_mr_mem addr reg))
(rule (x64_lock_add (OperandSize.Size32) addr reg) (x64_lock_addl_mr_mem addr reg))
(rule (x64_lock_add (OperandSize.Size64) addr reg) (x64_lock_addq_mr_mem addr reg))
-(decl x64_lock_sub (OperandSize Amode Gpr) SideEffectNoResult)
+(decl x64_lock_sub (OperandSize SyntheticAmode Gpr) SideEffectNoResult)
(rule (x64_lock_sub (OperandSize.Size8) addr reg) (x64_lock_subb_mr_mem addr reg))
(rule (x64_lock_sub (OperandSize.Size16) addr reg) (x64_lock_subw_mr_mem addr reg))
(rule (x64_lock_sub (OperandSize.Size32) addr reg) (x64_lock_subl_mr_mem addr reg))
(rule (x64_lock_sub (OperandSize.Size64) addr reg) (x64_lock_subq_mr_mem addr reg))
-(decl x64_lock_and (OperandSize Amode Gpr) SideEffectNoResult)
+(decl x64_lock_and (OperandSize SyntheticAmode Gpr) SideEffectNoResult)
(rule (x64_lock_and (OperandSize.Size8) addr reg) (x64_lock_andb_mr_mem addr reg))
(rule (x64_lock_and (OperandSize.Size16) addr reg) (x64_lock_andw_mr_mem addr reg))
(rule (x64_lock_and (OperandSize.Size32) addr reg) (x64_lock_andl_mr_mem addr reg))
(rule (x64_lock_and (OperandSize.Size64) addr reg) (x64_lock_andq_mr_mem addr reg))
-(decl x64_lock_or (OperandSize Amode Gpr) SideEffectNoResult)
+(decl x64_lock_or (OperandSize SyntheticAmode Gpr) SideEffectNoResult)
(rule (x64_lock_or (OperandSize.Size8) addr reg) (x64_lock_orb_mr_mem addr reg))
(rule (x64_lock_or (OperandSize.Size16) addr reg) (x64_lock_orw_mr_mem addr reg))
(rule (x64_lock_or (OperandSize.Size32) addr reg) (x64_lock_orl_mr_mem addr reg))
(rule (x64_lock_or (OperandSize.Size64) addr reg) (x64_lock_orq_mr_mem addr reg))
-(decl x64_lock_xor (OperandSize Amode Gpr) SideEffectNoResult)
+(decl x64_lock_xor (OperandSize SyntheticAmode Gpr) SideEffectNoResult)
(rule (x64_lock_xor (OperandSize.Size8) addr reg) (x64_lock_xorb_mr_mem addr reg))
(rule (x64_lock_xor (OperandSize.Size16) addr reg) (x64_lock_xorw_mr_mem addr reg))
(rule (x64_lock_xor (OperandSize.Size32) addr reg) (x64_lock_xorl_mr_mem addr reg))
diff --git a/cranelift/codegen/src/isa/x64/inst/args.rs b/cranelift/codegen/src/isa/x64/inst/args.rs
index ad71ae4da3..c34983643e 100644
--- a/cranelift/codegen/src/isa/x64/inst/args.rs
+++ b/cranelift/codegen/src/isa/x64/inst/args.rs
@@ -573,6 +573,21 @@ impl SyntheticAmode {
| SyntheticAmode::ConstantOffset { .. } => true,
}
}
+
+ /// Offset the synthetic amode by a fixed offset.
+ pub(crate) fn offset(&self, offset: i32) -> Self {
+ let mut ret = self.clone();
+ match &mut ret {
+ SyntheticAmode::Real(amode) => *amode = amode.offset(offset),
+ SyntheticAmode::SlotOffset { simm32 } => *simm32 += offset,
+ // `amode_offset` is used only in i128.load/store which
+ // takes a synthetic amode from `to_amode`; `to_amode` can
+ // only produce Real or SlotOffset amodes, never
+ // IncomingArg or ConstantOffset.
+ _ => panic!("Cannot offset SyntheticAmode: {self:?}"),
+ }
+ ret
+ }
}
impl From<Amode> for SyntheticAmode {
diff --git a/cranelift/codegen/src/isa/x64/lower.isle b/cranelift/codegen/src/isa/x64/lower.isle
index efbd8fe849..a7a8104de9 100644
--- a/cranelift/codegen/src/isa/x64/lower.isle
+++ b/cranelift/codegen/src/isa/x64/lower.isle
@@ -3056,8 +3056,8 @@
;; We can load an I128 by doing two 64-bit loads.
(rule -3 (lower (has_type $I128
(load (little_or_native_endian flags) address offset)))
- (let ((addr_lo Amode (to_amode flags address offset))
- (addr_hi Amode (amode_offset addr_lo 8))
+ (let ((addr_lo SyntheticAmode (to_amode flags address offset))
+ (addr_hi SyntheticAmode (amode_offset addr_lo 8))
(value_lo Reg (x64_mov addr_lo))
(value_hi Reg (x64_mov addr_hi)))
(value_regs value_lo value_hi)))
@@ -3187,8 +3187,8 @@
(let ((value_reg ValueRegs value)
(value_lo Gpr (value_regs_get_gpr value_reg 0))
(value_hi Gpr (value_regs_get_gpr value_reg 1))
- (addr_lo Amode (to_amode flags address offset))
- (addr_hi Amode (amode_offset addr_lo 8)))
+ (addr_lo SyntheticAmode (to_amode flags address offset))
+ (addr_hi SyntheticAmode (amode_offset addr_lo 8)))
(side_effect
(side_effect_concat
(x64_movrm $I64 addr_lo value_lo)
diff --git a/cranelift/codegen/src/isa/x64/lower/isle.rs b/cranelift/codegen/src/isa/x64/lower/isle.rs
index 823656d890..78262fce53 100644
--- a/cranelift/codegen/src/isa/x64/lower/isle.rs
+++ b/cranelift/codegen/src/isa/x64/lower/isle.rs
@@ -409,6 +409,11 @@ impl Context for IsleContext<'_, '_, MInst, X64Backend> {
amode.clone().into()
}
+ #[inline]
+ fn synthetic_amode_slot(&mut self, offset: i32) -> SyntheticAmode {
+ SyntheticAmode::SlotOffset { simm32: offset }
+ }
+
#[inline]
fn const_to_synthetic_amode(&mut self, c: VCodeConstant) -> SyntheticAmode {
SyntheticAmode::ConstantOffset(c)
@@ -633,7 +638,7 @@ impl Context for IsleContext<'_, '_, MInst, X64Backend> {
}
#[inline]
- fn amode_offset(&mut self, addr: &Amode, offset: i32) -> Amode {
+ fn amode_offset(&mut self, addr: &SyntheticAmode, offset: i32) -> SyntheticAmode {
addr.offset(offset)
}
diff --git a/cranelift/codegen/src/machinst/abi.rs b/cranelift/codegen/src/machinst/abi.rs
index 217a69ab58..cc13a6d49d 100644
--- a/cranelift/codegen/src/machinst/abi.rs
+++ b/cranelift/codegen/src/machinst/abi.rs
@@ -2139,6 +2139,11 @@ impl<M: ABIMachineSpec> Callee<M> {
}
}
+ /// Get the raw offset of a sized stackslot in the slot region.
+ pub fn sized_stackslot_offset(&self, slot: StackSlot) -> u32 {
+ self.sized_stackslots[slot]
+ }
+
/// Produce an instruction that computes a sized stackslot address.
pub fn sized_stackslot_addr(
&self,
diff --git a/cranelift/codegen/src/machinst/isle.rs b/cranelift/codegen/src/machinst/isle.rs
index 5f2f1b24c2..4f9a37a406 100644
--- a/cranelift/codegen/src/machinst/isle.rs
+++ b/cranelift/codegen/src/machinst/isle.rs
@@ -507,6 +507,22 @@ macro_rules! isle_lower_prelude_methods {
.into()
}
+ fn abi_stackslot_offset_into_slot_region(
+ &mut self,
+ stack_slot: StackSlot,
+ offset1: Offset32,
+ offset2: Offset32,
+ ) -> i32 {
+ let offset1 = i32::from(offset1);
+ let offset2 = i32::from(offset2);
+ i32::try_from(self.lower_ctx.abi().sized_stackslot_offset(stack_slot))
+ .expect("Stack slot region cannot be larger than 2GiB")
+ .checked_add(offset1)
+ .expect("Stack slot region cannot be larger than 2GiB")
+ .checked_add(offset2)
+ .expect("Stack slot region cannot be larger than 2GiB")
+ }
+
fn abi_dynamic_stackslot_addr(
&mut self,
dst: WritableReg,
diff --git a/cranelift/codegen/src/prelude_lower.isle b/cranelift/codegen/src/prelude_lower.isle
index 6bbd3624f4..3c30a8bf34 100644
--- a/cranelift/codegen/src/prelude_lower.isle
+++ b/cranelift/codegen/src/prelude_lower.isle
@@ -1119,6 +1119,10 @@
(decl abi_stackslot_addr (WritableReg StackSlot Offset32) MInst)
(extern constructor abi_stackslot_addr abi_stackslot_addr)
+;; StackSlot raw offset into slot region
+(decl abi_stackslot_offset_into_slot_region (StackSlot Offset32 Offset32) i32)
+(extern constructor abi_stackslot_offset_into_slot_region abi_stackslot_offset_into_slot_region)
+
;; DynamicStackSlot addr
(decl abi_dynamic_stackslot_addr (WritableReg DynamicStackSlot) MInst)
(extern constructor abi_dynamic_stackslot_addr abi_dynamic_stackslot_addr)
diff --git a/cranelift/filetests/filetests/isa/aarch64/dynamic-slot.clif b/cranelift/filetests/filetests/isa/aarch64/dynamic-slot.clif
index d43c607a23..39c26cdf10 100644
--- a/cranelift/filetests/filetests/isa/aarch64/dynamic-slot.clif
+++ b/cranelift/filetests/filetests/isa/aarch64/dynamic-slot.clif
@@ -16,9 +16,8 @@ block0:
; mov fp, sp
; sub sp, sp, #16
; block0:
-; movz x1, #1
-; mov x2, sp
-; str x1, [x2]
+; movz x0, #1
+; str x0, [sp]
; add sp, sp, #16
; ldp fp, lr, [sp], #16
; ret
@@ -29,9 +28,8 @@ block0:
; mov x29, sp
; sub sp, sp, #0x10
; block1: ; offset 0xc
-; mov x1, #1
-; mov x2, sp
-; str x1, [x2]
+; mov x0, #1
+; stur x0, [sp]
; add sp, sp, #0x10
; ldp x29, x30, [sp], #0x10
; ret
@@ -51,9 +49,8 @@ block0:
; mov fp, sp
; sub sp, sp, #16
; block0:
-; movz x1, #1
-; mov x2, sp
-; str x1, [x2]
+; movz x0, #1
+; str x0, [sp]
; add sp, sp, #16
; ldp fp, lr, [sp], #16
; ret
@@ -64,9 +61,8 @@ block0:
; mov x29, sp
; sub sp, sp, #0x10
; block1: ; offset 0xc
-; mov x1, #1
-; mov x2, sp
-; str x1, [x2]
+; mov x0, #1
+; stur x0, [sp]
; add sp, sp, #0x10
; ldp x29, x30, [sp], #0x10
; ret
diff --git a/cranelift/filetests/filetests/isa/aarch64/stack.clif b/cranelift/filetests/filetests/isa/aarch64/stack.clif
index 12b9788ec1..167f62922b 100644
--- a/cranelift/filetests/filetests/isa/aarch64/stack.clif
+++ b/cranelift/filetests/filetests/isa/aarch64/stack.clif
@@ -82,8 +82,7 @@ block0:
; mov fp, sp
; sub sp, sp, #16
; block0:
-; mov x1, sp
-; ldr x0, [x1]
+; ldr x0, [sp]
; add sp, sp, #16
; ldp fp, lr, [sp], #16
; ret
@@ -94,8 +93,7 @@ block0:
; mov x29, sp
; sub sp, sp, #0x10
; block1: ; offset 0xc
-; mov x1, sp
-; ldr x0, [x1]
+; ldur x0, [sp]
; add sp, sp, #0x10
; ldp x29, x30, [sp], #0x10
; ret
@@ -116,8 +114,7 @@ block0:
; movk w16, w16, #1, LSL #16
; sub sp, sp, x16, UXTX
; block0:
-; mov x1, sp
-; ldr x0, [x1]
+; ldr x0, [sp]
; movz w16, #34480
; movk w16, w16, #1, LSL #16
; add sp, sp, x16, UXTX
@@ -132,8 +129,7 @@ block0:
; movk w16, #1, lsl #16
; sub sp, sp, x16
; block1: ; offset 0x14
-; mov x1, sp
-; ldr x0, [x1]
+; ldur x0, [sp]
; mov w16, #0x86b0
; movk w16, #1, lsl #16
; add sp, sp, x16
@@ -153,8 +149,7 @@ block0(v0: i64):
; mov fp, sp
; sub sp, sp, #16
; block0:
-; mov x2, sp
-; str x0, [x2]
+; str x0, [sp]
; add sp, sp, #16
; ldp fp, lr, [sp], #16
; ret
@@ -165,8 +160,7 @@ block0(v0: i64):
; mov x29, sp
; sub sp, sp, #0x10
; block1: ; offset 0xc
-; mov x2, sp
-; str x0, [x2]
+; stur x0, [sp]
; add sp, sp, #0x10
; ldp x29, x30, [sp], #0x10
; ret
@@ -187,8 +181,7 @@ block0(v0: i64):
; movk w16, w16, #1, LSL #16
; sub sp, sp, x16, UXTX
; block0:
-; mov x2, sp
-; str x0, [x2]
+; str x0, [sp]
; movz w16, #34480
; movk w16, w16, #1, LSL #16
; add sp, sp, x16, UXTX
@@ -203,8 +196,7 @@ block0(v0: i64):
; movk w16, #1, lsl #16
; sub sp, sp, x16
; block1: ; offset 0x14
-; mov x2, sp
-; str x0, [x2]
+; stur x0, [sp]
; mov w16, #0x86b0
; movk w16, #1, lsl #16
; add sp, sp, x16
diff --git a/cranelift/filetests/filetests/isa/aarch64/stackslot.clif b/cranelift/filetests/filetests/isa/aarch64/stackslot.clif
new file mode 100644
index 0000000000..4d6c29a1f1
--- /dev/null
+++ b/cranelift/filetests/filetests/isa/aarch64/stackslot.clif
@@ -0,0 +1,60 @@
+test compile precise-output
+set enable_multi_ret_implicit_sret
+target aarch64
+
+function %f1(i32, i64, f32, f64, i8x16) -> i32, i64, f32, f64, i8x16 {
+ ss0 = explicit_slot 40
+
+block0(v0: i32, v1: i64, v2: f32, v3: f64, v4: i8x16):
+ stack_store v0, ss0+0
+ stack_store v1, ss0+4
+ stack_store v2, ss0+12
+ stack_store v3, ss0+16
+ stack_store v4, ss0+24
+ v5 = stack_load.i32 ss0+0
+ v6 = stack_load.i64 ss0+4
+ v7 = stack_load.f32 ss0+12
+ v8 = stack_load.f64 ss0+16
+ v9 = stack_load.i8x16 ss0+24
+ return v5, v6, v7, v8, v9
+}
+
+; VCode:
+; stp fp, lr, [sp, #-16]!
+; mov fp, sp
+; sub sp, sp, #48
+; block0:
+; str w0, [sp]
+; str x1, [sp, #4]
+; str s0, [sp, #12]
+; str d1, [sp, #16]
+; str q2, [sp, #24]
+; ldr w0, [sp]
+; ldr x1, [sp, #4]
+; ldr s0, [sp, #12]
+; ldr d1, [sp, #16]
+; ldr q2, [sp, #24]
+; add sp, sp, #48
+; ldp fp, lr, [sp], #16
+; ret
+;
+; Disassembled:
+; block0: ; offset 0x0
+; stp x29, x30, [sp, #-0x10]!
+; mov x29, sp
+; sub sp, sp, #0x30
+; block1: ; offset 0xc
+; stur w0, [sp]
+; stur x1, [sp, #4]
+; stur s0, [sp, #0xc]
+; stur d1, [sp, #0x10]
+; stur q2, [sp, #0x18]
+; ldur w0, [sp]
+; ldur x1, [sp, #4]
+; ldur s0, [sp, #0xc]
+; ldur d1, [sp, #0x10]
+; ldur q2, [sp, #0x18]
+; add sp, sp, #0x30
+; ldp x29, x30, [sp], #0x10
+; ret
+
diff --git a/cranelift/filetests/filetests/isa/aarch64/user_stack_maps.clif b/cranelift/filetests/filetests/isa/aarch64/user_stack_maps.clif
index 531ecb748a..469086bed3 100644
--- a/cranelift/filetests/filetests/isa/aarch64/user_stack_maps.clif
+++ b/cranelift/filetests/filetests/isa/aarch64/user_stack_maps.clif
@@ -34,39 +34,33 @@ block0:
; VCode:
; stp fp, lr, [sp, #-16]!
; mov fp, sp
-; str x28, [sp, #-16]!
-; stp x19, x21, [sp, #-16]!
+; str x24, [sp, #-16]!
+; stp x22, x23, [sp, #-16]!
; sub sp, sp, #16
; block0:
-; movz w19, #0
-; movz w28, #1
-; movz w21, #2
-; mov x8, sp
-; str w19, [x8]
-; add x9, sp, #4
-; str w28, [x9]
-; add x10, sp, #8
-; str w21, [x10]
-; mov x0, x19
+; movz w24, #0
+; movz w22, #1
+; movz w23, #2
+; str w24, [sp]
+; str w22, [sp, #4]
+; str w23, [sp, #8]
+; mov x0, x24
; bl 0
; ; UserStackMap { by_type: [(types::I32, CompoundBitSet {0, 4, 8})], sp_to_sized_stack_slots: None }
-; mov x12, sp
-; str w28, [x12]
-; add x13, sp, #4
-; str w21, [x13]
-; mov x0, x19
+; str w22, [sp]
+; str w23, [sp, #4]
+; mov x0, x24
; bl 0
; ; UserStackMap { by_type: [(types::I32, CompoundBitSet {0, 4})], sp_to_sized_stack_slots: None }
-; mov x15, sp
-; str w21, [x15]
-; mov x0, x28
+; str w23, [sp]
+; mov x0, x22
; bl 0
; ; UserStackMap { by_type: [(types::I32, CompoundBitSet {0})], sp_to_sized_stack_slots: None }
-; mov x0, x21
+; mov x0, x23
; bl 0
; add sp, sp, #16
-; ldp x19, x21, [sp], #16
-; ldr x28, [sp], #16
+; ldp x22, x23, [sp], #16
+; ldr x24, [sp], #16
; ldp fp, lr, [sp], #16
; ret
;
@@ -74,36 +68,30 @@ block0:
; block0: ; offset 0x0
; stp x29, x30, [sp, #-0x10]!
; mov x29, sp
-; str x28, [sp, #-0x10]!
-; stp x19, x21, [sp, #-0x10]!
+; str x24, [sp, #-0x10]!
+; stp x22, x23, [sp, #-0x10]!
; sub sp, sp, #0x10
; block1: ; offset 0x14
-; mov w19, #0
-; mov w28, #1
-; mov w21, #2
-; mov x8, sp
-; str w19, [x8]
-; add x9, sp, #4
-; str w28, [x9]
-; add x10, sp, #8
-; str w21, [x10]
-; mov x0, x19
-; bl #0x3c ; reloc_external Call u0:0 0
-; mov x12, sp
-; str w28, [x12]
-; add x13, sp, #4
-; str w21, [x13]
-; mov x0, x19
+; mov w24, #0
+; mov w22, #1
+; mov w23, #2
+; stur w24, [sp]
+; stur w22, [sp, #4]
+; stur w23, [sp, #8]
+; mov x0, x24
+; bl #0x30 ; reloc_external Call u0:0 0
+; stur w22, [sp]
+; stur w23, [sp, #4]
+; mov x0, x24
+; bl #0x40 ; reloc_external Call u0:0 0
+; stur w23, [sp]
+; mov x0, x22
+; bl #0x4c ; reloc_external Call u0:0 0
+; mov x0, x23
; bl #0x54 ; reloc_external Call u0:0 0
-; mov x15, sp
-; str w21, [x15]
-; mov x0, x28
-; bl #0x64 ; reloc_external Call u0:0 0
-; mov x0, x21
-; bl #0x6c ; reloc_external Call u0:0 0
; add sp, sp, #0x10
-; ldp x19, x21, [sp], #0x10
-; ldr x28, [sp], #0x10
+; ldp x22, x23, [sp], #0x10
+; ldr x24, [sp], #0x10
; ldp x29, x30, [sp], #0x10
; ret
@@ -130,39 +118,33 @@ block0(v0: i8, v1: i16, v2: i32, v3: i64, v4: f32, v5: f64):
; VCode:
; stp fp, lr, [sp, #-16]!
; mov fp, sp
-; stp x23, x26, [sp, #-16]!
-; stp x19, x21, [sp, #-16]!
+; stp x24, x27, [sp, #-16]!
+; stp x19, x23, [sp, #-16]!
; sub sp, sp, #128
; block0:
-; mov x12, sp
-; strb w0, [x12]
-; mov x21, x0
-; add x13, sp, #8
-; strh w1, [x13]
-; mov x26, x1
-; add x14, sp, #16
-; str w2, [x14]
-; mov x23, x2
-; add x15, sp, #20
-; str s0, [x15]
+; strb w0, [sp]
+; mov x23, x0
+; strh w1, [sp, #8]
+; mov x19, x1
+; str w2, [sp, #16]
+; mov x27, x2
+; str s0, [sp, #20]
; str q0, [sp, #96]
-; add x0, sp, #24
-; str x3, [x0]
-; mov x19, x3
-; add x1, sp, #32
-; str d1, [x1]
+; str x3, [sp, #24]
+; mov x24, x3
+; str d1, [sp, #32]
; str q1, [sp, #112]
; bl 0
; ; UserStackMap { by_type: [(types::I8, CompoundBitSet {0}), (types::I16, CompoundBitSet {8}), (types::I32, CompoundBitSet {16}), (types::F32, CompoundBitSet {20}), (types::I64, CompoundBitSet {24}), (types::F64, CompoundBitSet {32})], sp_to_sized_stack_slots: None }
-; mov x0, x21
-; mov x1, x26
-; mov x2, x23
-; mov x3, x19
+; mov x0, x23
+; mov x1, x19
+; mov x2, x27
+; mov x3, x24
; ldr q0, [sp, #96]
; ldr q1, [sp, #112]
; add sp, sp, #128
-; ldp x19, x21, [sp], #16
-; ldp x23, x26, [sp], #16
+; ldp x19, x23, [sp], #16
+; ldp x24, x27, [sp], #16
; ldp fp, lr, [sp], #16
; ret
;
@@ -170,38 +152,32 @@ block0(v0: i8, v1: i16, v2: i32, v3: i64, v4: f32, v5: f64):
; block0: ; offset 0x0
; stp x29, x30, [sp, #-0x10]!
; mov x29, sp
-; stp x23, x26, [sp, #-0x10]!
-; stp x19, x21, [sp, #-0x10]!
+; stp x24, x27, [sp, #-0x10]!
+; stp x19, x23, [sp, #-0x10]!
; sub sp, sp, #0x80
; block1: ; offset 0x14
-; mov x12, sp
-; strb w0, [x12]
-; mov x21, x0
-; add x13, sp, #8
-; strh w1, [x13]
-; mov x26, x1
-; add x14, sp, #0x10
-; str w2, [x14]
-; mov x23, x2
-; add x15, sp, #0x14
-; str s0, [x15]
+; sturb w0, [sp]
+; mov x23, x0
+; sturh w1, [sp, #8]
+; mov x19, x1
+; stur w2, [sp, #0x10]
+; mov x27, x2
+; stur s0, [sp, #0x14]
; stur q0, [sp, #0x60]
-; add x0, sp, #0x18
-; str x3, [x0]
-; mov x19, x3
-; add x1, sp, #0x20
-; str d1, [x1]
+; stur x3, [sp, #0x18]
+; mov x24, x3
+; stur d1, [sp, #0x20]
; stur q1, [sp, #0x70]
-; bl #0x5c ; reloc_external Call u0:0 0
-; mov x0, x21
-; mov x1, x26
-; mov x2, x23
-; mov x3, x19
+; bl #0x44 ; reloc_external Call u0:0 0
+; mov x0, x23
+; mov x1, x19
+; mov x2, x27
+; mov x3, x24
; ldur q0, [sp, #0x60]
; ldur q1, [sp, #0x70]
; add sp, sp, #0x80
-; ldp x19, x21, [sp], #0x10
-; ldp x23, x26, [sp], #0x10
+; ldp x19, x23, [sp], #0x10
+; ldp x24, x27, [sp], #0x10
; ldp x29, x30, [sp], #0x10
; ret
diff --git a/cranelift/filetests/filetests/isa/x64/fmsub-inst.clif b/cranelift/filetests/filetests/isa/x64/fmsub-inst.clif
index cdeb5a7a8c..98975b85a0 100644
--- a/cranelift/filetests/filetests/isa/x64/fmsub-inst.clif
+++ b/cranelift/filetests/filetests/isa/x64/fmsub-inst.clif
@@ -458,7 +458,7 @@ block0(v0: f64, v1: f64, v2: f64):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; vmovsd %xmm2, (%r8)
+; vmovsd %xmm2, <offset:1>+(%rsp)
; vfnmsub213sd (%r8), %xmm1, %xmm0
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -472,7 +472,7 @@ block0(v0: f64, v1: f64, v2: f64):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; vmovsd %xmm2, (%r8) ; trap: heap_oob
+; vmovsd %xmm2, (%rsp)
; vfnmsub213sd (%r8), %xmm1, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -496,7 +496,7 @@ block0(v0: f64, v1: f64, v2: f64):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; vmovsd %xmm2, (%r8)
+; vmovsd %xmm2, <offset:1>+(%rsp)
; vfmsub213sd (%r8), %xmm1, %xmm0
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -510,7 +510,7 @@ block0(v0: f64, v1: f64, v2: f64):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; vmovsd %xmm2, (%r8) ; trap: heap_oob
+; vmovsd %xmm2, (%rsp)
; vfmsub213sd (%r8), %xmm1, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -534,7 +534,7 @@ block0(v0: f32, v1: f32, v2: f32):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; vmovss %xmm1, (%r8)
+; vmovss %xmm1, <offset:1>+(%rsp)
; vfmsub132ss (%r8), %xmm2, %xmm0
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -548,7 +548,7 @@ block0(v0: f32, v1: f32, v2: f32):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; vmovss %xmm1, (%r8) ; trap: heap_oob
+; vmovss %xmm1, (%rsp)
; vfmsub132ss (%r8), %xmm2, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -574,7 +574,7 @@ block0(v0: f32, v1: f32, v2: f32):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; vmovss %xmm1, (%r8)
+; vmovss %xmm1, <offset:1>+(%rsp)
; vfnmsub132ss (%r8), %xmm2, %xmm0
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -588,7 +588,7 @@ block0(v0: f32, v1: f32, v2: f32):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; vmovss %xmm1, (%r8) ; trap: heap_oob
+; vmovss %xmm1, (%rsp)
; vfnmsub132ss (%r8), %xmm2, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -620,7 +620,7 @@ block0(v0: f32, v1: f32, v2: f32):
; movl $0x80000000, %r9d
; vmovd %r9d, %xmm3
; vxorps %xmm3, %xmm1, %xmm3
-; vmovss %xmm3, (%r11)
+; vmovss %xmm3, <offset:1>+(%rsp)
; vfmsub132ss (%r11), %xmm2, %xmm0
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -637,7 +637,7 @@ block0(v0: f32, v1: f32, v2: f32):
; movl $0x80000000, %r9d
; vmovd %r9d, %xmm3
; vxorps %xmm3, %xmm1, %xmm3
-; vmovss %xmm3, (%r11) ; trap: heap_oob
+; vmovss %xmm3, (%rsp)
; vfmsub132ss (%r11), %xmm2, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
; movq %rbp, %rsp
@@ -661,7 +661,7 @@ block0(v0: f32x4, v1: f32x4, v2: f32x4):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; vmovups %xmm0, (%r8)
+; vmovups %xmm0, <offset:1>+(%rsp)
; movdqa %xmm1, %xmm0
; vfmsub132ps (%r8), %xmm2, %xmm0
; addq $0x10, %rsp
@@ -676,7 +676,7 @@ block0(v0: f32x4, v1: f32x4, v2: f32x4):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; vmovups %xmm0, (%r8) ; trap: heap_oob
+; vmovups %xmm0, (%rsp)
; movdqa %xmm1, %xmm0
; vfmsub132ps (%r8), %xmm2, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
diff --git a/cranelift/filetests/filetests/isa/x64/fpromote-demote-avx.clif b/cranelift/filetests/filetests/isa/x64/fpromote-demote-avx.clif
index 9fa1b5c135..66fbe4a2ce 100644
--- a/cranelift/filetests/filetests/isa/x64/fpromote-demote-avx.clif
+++ b/cranelift/filetests/filetests/isa/x64/fpromote-demote-avx.clif
@@ -46,7 +46,7 @@ block0(v1: i64, v2: f32):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; vmovss %xmm0, (%r8)
+; vmovss %xmm0, <offset:1>+(%rsp)
; uninit %xmm4
; vxorpd %xmm4, %xmm4, %xmm6
; vcvtss2sd (%r8), %xmm6, %xmm0
@@ -62,7 +62,7 @@ block0(v1: i64, v2: f32):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; vmovss %xmm0, (%r8) ; trap: heap_oob
+; vmovss %xmm0, (%rsp)
; vxorpd %xmm4, %xmm4, %xmm6
; vcvtss2sd (%r8), %xmm6, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
@@ -115,7 +115,7 @@ block0(v1: i64, v2: f64):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; vmovsd %xmm0, (%r8)
+; vmovsd %xmm0, <offset:1>+(%rsp)
; uninit %xmm4
; vxorps %xmm4, %xmm4, %xmm6
; vcvtsd2ss (%r8), %xmm6, %xmm0
@@ -131,7 +131,7 @@ block0(v1: i64, v2: f64):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; vmovsd %xmm0, (%r8) ; trap: heap_oob
+; vmovsd %xmm0, (%rsp)
; vxorps %xmm4, %xmm4, %xmm6
; vcvtsd2ss (%r8), %xmm6, %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
diff --git a/cranelift/filetests/filetests/isa/x64/fpromote-demote.clif b/cranelift/filetests/filetests/isa/x64/fpromote-demote.clif
index 2959372545..a7f89bcbbb 100644
--- a/cranelift/filetests/filetests/isa/x64/fpromote-demote.clif
+++ b/cranelift/filetests/filetests/isa/x64/fpromote-demote.clif
@@ -50,7 +50,7 @@ block0(v1: i64, v2: f32):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; movss %xmm0, (%r8)
+; movss %xmm0, <offset:1>+(%rsp)
; uninit %xmm0
; xorpd %xmm0, %xmm0
; cvtss2sd (%r8), %xmm0
@@ -66,7 +66,7 @@ block0(v1: i64, v2: f32):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; movss %xmm0, (%r8) ; trap: heap_oob
+; movss %xmm0, (%rsp)
; xorpd %xmm0, %xmm0
; cvtss2sd (%r8), %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
@@ -123,7 +123,7 @@ block0(v1: i64, v2: f64):
; subq $0x10, %rsp
; block0:
; leaq <offset:1>+(%rsp), %r8
-; movsd %xmm0, (%r8)
+; movsd %xmm0, <offset:1>+(%rsp)
; uninit %xmm0
; xorps %xmm0, %xmm0
; cvtsd2ss (%r8), %xmm0
@@ -139,7 +139,7 @@ block0(v1: i64, v2: f64):
; subq $0x10, %rsp
; block1: ; offset 0x8
; leaq (%rsp), %r8
-; movsd %xmm0, (%r8) ; trap: heap_oob
+; movsd %xmm0, (%rsp)
; xorps %xmm0, %xmm0
; cvtsd2ss (%r8), %xmm0 ; trap: heap_oob
; addq $0x10, %rsp
diff --git a/cranelift/filetests/filetests/isa/x64/stackslot.clif b/cranelift/filetests/filetests/isa/x64/stackslot.clif
index 3529747237..baf869fbb3 100644
--- a/cranelift/filetests/filetests/isa/x64/stackslot.clif
+++ b/cranelift/filetests/filetests/isa/x64/stackslot.clif
@@ -60,3 +60,63 @@ block0(v0: i64):
; retq
; ud2 ; trap: stk_ovf
+function %f1(i32, i64, f32, f64, i8x16) -> i32, i64, f32, f64, i8x16 {
+ ss0 = explicit_slot 40
+
+block0(v0: i32, v1: i64, v2: f32, v3: f64, v4: i8x16):
+ stack_store v0, ss0+0
+ stack_store v1, ss0+4
+ stack_store v2, ss0+12
+ stack_store v3, ss0+16
+ stack_store v4, ss0+24
+ v5 = stack_load.i32 ss0+0
+ v6 = stack_load.i64 ss0+4
+ v7 = stack_load.f32 ss0+12
+ v8 = stack_load.f64 ss0+16
+ v9 = stack_load.i8x16 ss0+24
+ return v5, v6, v7, v8, v9
+}
+
+; VCode:
+; pushq %rbp
+; movq %rsp, %rbp
+; subq $0x30, %rsp
+; block0:
+; movl %esi, <offset:1>+(%rsp)
+; movq %rdx, <offset:1>+4(%rsp)
+; movss %xmm0, <offset:1>+0xc(%rsp)
+; movsd %xmm1, <offset:1>+0x10(%rsp)
+; movdqu %xmm2, <offset:1>+0x18(%rsp)
+; movl <offset:1>+(%rsp), %eax
+; movq <offset:1>+4(%rsp), %rdx
+; movss <offset:1>+0xc(%rsp), %xmm0
+; movsd <offset:1>+0x10(%rsp), %xmm1
+; movdqu <offset:1>+0x18(%rsp), %xmm2
+; movdqu %xmm2, (%rdi)
+; addq $0x30, %rsp
+; movq %rbp, %rsp
+; popq %rbp
+; retq
+;
+; Disassembled:
+; block0: ; offset 0x0
+; pushq %rbp
+; movq %rsp, %rbp
+; subq $0x30, %rsp
+; block1: ; offset 0x8
+; movl %esi, (%rsp)
+; movq %rdx, 4(%rsp)
+; movss %xmm0, 0xc(%rsp)
+; movsd %xmm1, 0x10(%rsp)
+; movdqu %xmm2, 0x18(%rsp)
+; movl (%rsp), %eax
+; movq 4(%rsp), %rdx
+; movss 0xc(%rsp), %xmm0
+; movsd 0x10(%rsp), %xmm1
+; movdqu 0x18(%rsp), %xmm2
+; movdqu %xmm2, (%rdi)
+; addq $0x30, %rsp
+; movq %rbp, %rsp
+; popq %rbp
+; retq
+
diff --git a/cranelift/filetests/filetests/isa/x64/user_stack_maps.clif b/cranelift/filetests/filetests/isa/x64/user_stack_maps.clif
index f1b7fa8b36..a9e15e1c9c 100644
--- a/cranelift/filetests/filetests/isa/x64/user_stack_maps.clif
+++ b/cranelift/filetests/filetests/isa/x64/user_stack_maps.clif
@@ -42,31 +42,24 @@ block0:
; block0:
; uninit %rdi
; xorl %edi, %edi
-; movq %rdi, %r14
-; movl $0x1, %r15d
-; movl $0x2, %ebx
-; leaq <offset:1>+(%rsp), %rsi
-; movl $0x0, (%rsi)
-; leaq <offset:1>+4(%rsp), %rdi
-; movl $0x1, (%rdi)
-; leaq <offset:1>+8(%rsp), %rax
-; movl $0x2, (%rax)
-; movq %r14, %rdi
+; movq %rdi, %rbx
+; movl $0x1, %r14d
+; movl $0x2, %r15d
+; movl $0x0, <offset:1>+(%rsp)
+; movl $0x1, <offset:1>+4(%rsp)
+; movl $0x2, <offset:1>+8(%rsp)
; call User(userextname0)
; ; UserStackMap { by_type: [(types::I32, CompoundBitSet {0, 4, 8})], sp_to_sized_stack_slots: None }
-; leaq <offset:1>+(%rsp), %rdx
-; movl $0x1, (%rdx)
-; leaq <offset:1>+4(%rsp), %r8
-; movl $0x2, (%r8)
-; movq %r14, %rdi
+; movl $0x1, <offset:1>+(%rsp)
+; movl $0x2, <offset:1>+4(%rsp)
+; movq %rbx, %rdi
; call User(userextname0)
; ; UserStackMap { by_type: [(types::I32, CompoundBitSet {0, 4})], sp_to_sized_stack_slots: None }
-; leaq <offset:1>+(%rsp), %r10
-; movl $0x2, (%r10)
-; movq %r15, %rdi
+; movl $0x2, <offset:1>+(%rsp)
+; movq %r14, %rdi
; call User(userextname0)
; ; UserStackMap { by_type: [(types::I32, CompoundBitSet {0})], sp_to_sized_stack_slots: None }
-; movq %rbx, %rdi
+; movq %r15, %rdi
; call User(userextname0)
; movq 0x10(%rsp), %rbx
; movq 0x18(%rsp), %r14
@@ -86,29 +79,22 @@ block0:
; movq %r15, 0x20(%rsp)
; block1: ; offset 0x17
; xorl %edi, %edi
-; movq %rdi, %r14
-; movl $1, %r15d
-; movl $2, %ebx
-; leaq (%rsp), %rsi
-; movl $0, (%rsi)
-; leaq 4(%rsp), %rdi
-; movl $1, (%rdi)
-; leaq 8(%rsp), %rax
-; movl $2, (%rax)
-; movq %r14, %rdi
-; callq 0x4f ; reloc_external CallPCRel4 u0:0 -4
-; leaq (%rsp), %rdx
-; movl $1, (%rdx)
-; leaq 4(%rsp), %r8
-; movl $2, (%r8)
+; movq %rdi, %rbx
+; movl $1, %r14d
+; movl $2, %r15d
+; movl $0, (%rsp)
+; movl $1, 4(%rsp)
+; movl $2, 8(%rsp)
+; callq 0x44 ; reloc_external CallPCRel4 u0:0 -4
+; movl $1, (%rsp)
+; movl $2, 4(%rsp)
+; movq %rbx, %rdi
+; callq 0x5b ; reloc_external CallPCRel4 u0:0 -4
+; movl $2, (%rsp)
; movq %r14, %rdi
-; callq 0x6d ; reloc_external CallPCRel4 u0:0 -4
-; leaq (%rsp), %r10
-; movl $2, (%r10)
+; callq 0x6a ; reloc_external CallPCRel4 u0:0 -4
; movq %r15, %rdi
-; callq 0x80 ; reloc_external CallPCRel4 u0:0 -4
-; movq %rbx, %rdi
-; callq 0x88 ; reloc_external CallPCRel4 u0:0 -4
+; callq 0x72 ; reloc_external CallPCRel4 u0:0 -4
; movq 0x10(%rsp), %rbx
; movq 0x18(%rsp), %r14
; movq 0x20(%rsp), %r15
@@ -147,34 +133,28 @@ block0(v0: i8, v1: i16, v2: i32, v3: i64, v4: f32, v5: f64):
; movq %r14, 0x98(%rsp)
; movq %r15, 0xa0(%rsp)
; block0:
-; movq %rdi, %r13
-; leaq <offset:1>+(%rsp), %r9
-; movb %sil, (%r9)
-; movq %rsi, %r15
-; leaq <offset:1>+8(%rsp), %r9
-; movw %dx, (%r9)
-; movq %rdx, %r12
-; leaq <offset:1>+0x10(%rsp), %r9
-; movl %ecx, (%r9)
-; movq %rcx, %rbx
-; leaq <offset:1>+0x14(%rsp), %r10
-; movss %xmm0, (%r10)
+; movq %rdi, %r12
+; movb %sil, <offset:1>+(%rsp)
+; movq %rsi, %r14
+; movw %dx, <offset:1>+8(%rsp)
+; movq %rdx, %rbx
+; movl %ecx, <offset:1>+0x10(%rsp)
+; movq %rcx, %r15
+; movss %xmm0, <offset:1>+0x14(%rsp)
; movdqu %xmm0, <offset:1>+0x60(%rsp)
-; leaq <offset:1>+0x18(%rsp), %r11
-; movq %r8, (%r11)
-; movq %r8, %r14
-; leaq <offset:1>+0x20(%rsp), %rsi
-; movsd %xmm1, (%rsi)
+; movq %r8, <offset:1>+0x18(%rsp)
+; movq %r8, %r13
+; movsd %xmm1, <offset:1>+0x20(%rsp)
; movdqu %xmm1, <offset:1>+0x70(%rsp)
; call User(userextname0)
; ; UserStackMap { by_type: [(types::I8, CompoundBitSet {0}), (types::I16, CompoundBitSet {8}), (types::I32, CompoundBitSet {16}), (types::F32, CompoundBitSet {20}), (types::I64, CompoundBitSet {24}), (types::F64, CompoundBitSet {32})], sp_to_sized_stack_slots: None }
-; movq %rbx, %rcx
-; movq %r13, %rdi
+; movq %r15, %rcx
+; movq %r12, %rdi
; movl %ecx, (%rdi)
-; movq %r14, %r8
+; movq %r13, %r8
; movq %r8, 8(%rdi)
-; movq %r15, %rax
-; movq %r12, %rdx
+; movq %r14, %rax
+; movq %rbx, %rdx
; movdqu <offset:1>+0x60(%rsp), %xmm0
; movdqu <offset:1>+0x70(%rsp), %xmm1
; movq 0x80(%rsp), %rbx
@@ -198,33 +178,27 @@ block0(v0: i8, v1: i16, v2: i32, v3: i64, v4: f32, v5: f64):
; movq %r14, 0x98(%rsp)
; movq %r15, 0xa0(%rsp)
; block1: ; offset 0x33
-; movq %rdi, %r13
-; leaq (%rsp), %r9
-; movb %sil, (%r9)
-; movq %rsi, %r15
-; leaq 8(%rsp), %r9
-; movw %dx, (%r9)
-; movq %rdx, %r12
-; leaq 0x10(%rsp), %r9
-; movl %ecx, (%r9)
-; movq %rcx, %rbx
-; leaq 0x14(%rsp), %r10
-; movss %xmm0, (%r10)
+; movq %rdi, %r12
+; movb %sil, (%rsp)
+; movq %rsi, %r14
+; movw %dx, 8(%rsp)
+; movq %rdx, %rbx
+; movl %ecx, 0x10(%rsp)
+; movq %rcx, %r15
+; movss %xmm0, 0x14(%rsp)
; movdqu %xmm0, 0x60(%rsp)
-; leaq 0x18(%rsp), %r11
-; movq %r8, (%r11)
-; movq %r8, %r14
-; leaq 0x20(%rsp), %rsi
-; movsd %xmm1, (%rsi)
+; movq %r8, 0x18(%rsp)
+; movq %r8, %r13
+; movsd %xmm1, 0x20(%rsp)
; movdqu %xmm1, 0x70(%rsp)
-; callq 0x86 ; reloc_external CallPCRel4 u0:0 -4
-; movq %rbx, %rcx
-; movq %r13, %rdi
+; callq 0x71 ; reloc_external CallPCRel4 u0:0 -4
+; movq %r15, %rcx
+; movq %r12, %rdi
; movl %ecx, (%rdi)
-; movq %r14, %r8
+; movq %r13, %r8
; movq %r8, 8(%rdi)
-; movq %r15, %rax
-; movq %r12, %rdx
+; movq %r14, %rax
+; movq %rbx, %rdx
; movdqu 0x60(%rsp), %xmm0
; movdqu 0x70(%rsp), %xmm1
; movq 0x80(%rsp), %rbx
diff --git a/tests/disas/exceptions.wat b/tests/disas/exceptions.wat
index 23129927fa..c125c8b60c 100644
--- a/tests/disas/exceptions.wat
+++ b/tests/disas/exceptions.wat
@@ -33,14 +33,14 @@
;; movq %rdi, %r12
;; movq %rcx, %r13
;; movq %rdx, %r15
-;; callq 0x3c2
+;; callq 0x3b2
;; movq %rax, %r14
;; movl $0x4000000, %esi
;; movl $3, %edx
;; movl $0x30, %ecx
;; movl $8, %r8d
;; movq %r12, %rdi
-;; callq 0x35f
+;; callq 0x34f
;; movq 8(%r12), %r8
;; movq 0x18(%r8), %r8
;; movl %eax, %r9d
@@ -54,7 +54,7 @@
;; movq %rax, %rsi
;; movq %r12, %rdi
;; movq %r12, (%rsp)
-;; callq 0x3ee
+;; callq 0x3de
;; ud2
;; ud2
;;
diff --git a/tests/disas/gc/struct-new-stack-map.wat b/tests/disas/gc/struct-new-stack-map.wat
index a4baa3fa6d..a58654548d 100644
--- a/tests/disas/gc/struct-new-stack-map.wat
+++ b/tests/disas/gc/struct-new-stack-map.wat
@@ -16,52 +16,47 @@
;; movq %rsp, %rbp
;; movq 8(%rdi), %r10
;; movq 0x10(%r10), %r10
-;; addq $0x50, %r10
+;; addq $0x40, %r10
;; cmpq %rsp, %r10
-;; ja 0xc7
-;; 19: subq $0x40, %rsp
-;; movq %r12, 0x20(%rsp)
-;; movq %r13, 0x28(%rsp)
-;; movq %r14, 0x30(%rsp)
-;; movq %rdx, %r12
+;; ja 0xb5
+;; 19: subq $0x30, %rsp
+;; movq %r13, 0x20(%rsp)
+;; movq %r14, 0x28(%rsp)
+;; movq %rdx, %r14
;; movdqu %xmm0, 8(%rsp)
-;; leaq (%rsp), %r14
-;; movl %ecx, (%r14)
+;; movl %ecx, (%rsp)
;; movl $0xb0000000, %esi
;; xorl %edx, %edx
;; movl $0x28, %ecx
;; movl $8, %r8d
;; movq %rdi, %r13
-;; callq 0x14b
-;; movq 8(%r13), %r8
-;; ╰─╼ stack_map: frame_size=64, frame_offsets=[0]
-;; movq 0x18(%r8), %r8
-;; movq %rax, %r10
-;; movl %r10d, %r9d
+;; callq 0x12f
+;; movq 8(%r13), %rdx
+;; ╰─╼ stack_map: frame_size=48, frame_offsets=[0]
+;; movq 0x18(%rdx), %rdx
+;; movl %eax, %r8d
;; movdqu 8(%rsp), %xmm0
-;; movss %xmm0, 0x18(%r8, %r9)
-;; movq %r12, %rdx
-;; movb %dl, 0x1c(%r8, %r9)
-;; movl (%r14), %r11d
-;; movq %r11, %rdx
-;; andl $1, %edx
-;; testl %r11d, %r11d
-;; sete %sil
-;; movzbl %sil, %esi
-;; orl %esi, %edx
-;; testl %edx, %edx
-;; jne 0xa4
-;; 97: movl %r11d, %ecx
-;; leaq (%r8, %rcx), %rax
-;; addq $1, 8(%r8, %rcx)
-;; movl (%r14), %edx
-;; movl %edx, 0x20(%r8, %r9)
-;; movq %r10, %rax
-;; movq 0x20(%rsp), %r12
-;; movq 0x28(%rsp), %r13
-;; movq 0x30(%rsp), %r14
-;; addq $0x40, %rsp
+;; movss %xmm0, 0x18(%rdx, %r8)
+;; movq %r14, %r9
+;; movb %r9b, 0x1c(%rdx, %r8)
+;; movl (%rsp), %r9d
+;; movq %r9, %rcx
+;; andl $1, %ecx
+;; testl %r9d, %r9d
+;; sete %r10b
+;; movzbl %r10b, %r10d
+;; orl %r10d, %ecx
+;; testl %ecx, %ecx
+;; jne 0x9a
+;; 8d: movl %r9d, %edi
+;; leaq (%rdx, %rdi), %rcx
+;; addq $1, 8(%rdx, %rdi)
+;; movl (%rsp), %ecx
+;; movl %ecx, 0x20(%rdx, %r8)
+;; movq 0x20(%rsp), %r13
+;; movq 0x28(%rsp), %r14
+;; addq $0x30, %rsp
;; movq %rbp, %rsp
;; popq %rbp
;; retq
-;; c7: ud2
+;; b5: ud2
diff --git a/tests/disas/trunc.wat b/tests/disas/trunc.wat
index 74e4f4f7ae..e6acff281e 100644
--- a/tests/disas/trunc.wat
+++ b/tests/disas/trunc.wat
@@ -24,7 +24,7 @@
;; jne 0x101
;; 39: movq %r14, %rdi
;; movdqu (%rsp), %xmm0
-;; callq 0x253
+;; callq 0x245
;; movabsq $13830554455654793216, %rax
;; movq %rax, %xmm6
;; ucomisd %xmm0, %xmm6
@@ -55,27 +55,27 @@
;; retq
;; d3: movl $6, %esi
;; d8: movq %r14, %rdi
-;; db: callq 0x27f
+;; db: callq 0x271
;; e0: movq %r14, %rdi
-;; e3: callq 0x2af
+;; e3: callq 0x2a1
;; e8: ud2
;; ea: movl $6, %esi
;; ef: movq %r14, %rdi
-;; f2: callq 0x27f
+;; f2: callq 0x271
;; f7: movq %r14, %rdi
-;; fa: callq 0x2af
+;; fa: callq 0x2a1
;; ff: ud2
;; 101: movl $8, %esi
;; 106: movq %r14, %rdi
-;; 109: callq 0x27f
+;; 109: callq 0x271
;; 10e: movq %r14, %rdi
-;; 111: callq 0x2af
+;; 111: callq 0x2a1
;; 116: ud2
;; 118: xorl %esi, %esi
;; 11a: movq %r14, %rdi
-;; 11d: callq 0x27f
+;; 11d: callq 0x271
;; 122: movq %r14, %rdi
-;; 125: callq 0x2af
+;; 125: callq 0x2a1
;; 12a: ud2
;; 12c: ud2
;; 12e: ud2
diff --git a/tests/disas/trunc32.wat b/tests/disas/trunc32.wat
index c184255e64..8e5e2e8a63 100644
--- a/tests/disas/trunc32.wat
+++ b/tests/disas/trunc32.wat
@@ -26,7 +26,7 @@
;; jp 0xf6
;; jne 0xf6
;; 46: movq %r12, %rdi
-;; callq 0x252
+;; callq 0x243
;; movabsq $13830554455654793216, %r8
;; movq %r8, %xmm1
;; ucomisd %xmm0, %xmm1
@@ -56,27 +56,27 @@
;; retq
;; c8: movl $6, %esi
;; cd: movq %r12, %rdi
-;; d0: callq 0x27e
+;; d0: callq 0x26f
;; d5: movq %r12, %rdi
-;; d8: callq 0x2ae
+;; d8: callq 0x29f
;; dd: ud2
;; df: movl $6, %esi
;; e4: movq %r12, %rdi
-;; e7: callq 0x27e
+;; e7: callq 0x26f
;; ec: movq %r12, %rdi
-;; ef: callq 0x2ae
+;; ef: callq 0x29f
;; f4: ud2
;; f6: movl $8, %esi
;; fb: movq %r12, %rdi
-;; fe: callq 0x27e
+;; fe: callq 0x26f
;; 103: movq %r12, %rdi
-;; 106: callq 0x2ae
+;; 106: callq 0x29f
;; 10b: ud2
;; 10d: xorl %esi, %esi
;; 10f: movq %r12, %rdi
-;; 112: callq 0x27e
+;; 112: callq 0x26f
;; 117: movq %r12, %rdi
-;; 11a: callq 0x2ae
+;; 11a: callq 0x29f
;; 11f: ud2
;; 121: ud2
;; 123: ud2
diff --git a/tests/disas/winch/x64/atomic/notify/notify.wat b/tests/disas/winch/x64/atomic/notify/notify.wat
index 581a89e20c..4f8e4e0e3d 100644
--- a/tests/disas/winch/x64/atomic/notify/notify.wat
+++ b/tests/disas/winch/x64/atomic/notify/notify.wat
@@ -27,7 +27,7 @@
;; movl $0, %esi
;; movq 8(%rsp), %rdx
;; movl 4(%rsp), %ecx
-;; callq 0x177
+;; callq 0x175
;; addq $4, %rsp
;; addq $0xc, %rsp
;; movq 8(%rsp), %r14
diff --git a/tests/disas/winch/x64/atomic/notify/notify_offset.wat b/tests/disas/winch/x64/atomic/notify/notify_offset.wat
index b6b1238188..5f5fd6c8d4 100644
--- a/tests/disas/winch/x64/atomic/notify/notify_offset.wat
+++ b/tests/disas/winch/x64/atomic/notify/notify_offset.wat
@@ -28,7 +28,7 @@
;; movl $0, %esi
;; movq 8(%rsp), %rdx
;; movl 4(%rsp), %ecx
-;; callq 0x17e
+;; callq 0x17c
;; addq $4, %rsp
;; addq $0xc, %rsp
;; movq 8(%rsp), %r14
diff --git a/tests/disas/winch/x64/atomic/wait/wait32.wat b/tests/disas/winch/x64/atomic/wait/wait32.wat
index 5c4ff09208..2d018f2f70 100644
--- a/tests/disas/winch/x64/atomic/wait/wait32.wat
+++ b/tests/disas/winch/x64/atomic/wait/wait32.wat
@@ -30,7 +30,7 @@
;; movq 0x18(%rsp), %rdx
;; movl 0x14(%rsp), %ecx
;; movq 0xc(%rsp), %r8
-;; callq 0x184
+;; callq 0x182
;; addq $0xc, %rsp
;; addq $0x14, %rsp
;; movq 8(%rsp), %r14
diff --git a/tests/disas/winch/x64/atomic/wait/wait32_offset.wat b/tests/disas/winch/x64/atomic/wait/wait32_offset.wat
index bbb4601c48..bca69bb45e 100644
--- a/tests/disas/winch/x64/atomic/wait/wait32_offset.wat
+++ b/tests/disas/winch/x64/atomic/wait/wait32_offset.wat
@@ -34,7 +34,7 @@
;; movq 0x18(%rsp), %rdx
;; movl 0x14(%rsp), %ecx
;; movq 0xc(%rsp), %r8
-;; callq 0x18b
+;; callq 0x189
;; addq $0xc, %rsp
;; addq $0x14, %rsp
;; movq 8(%rsp), %r14
diff --git a/tests/disas/winch/x64/atomic/wait/wait64.wat b/tests/disas/winch/x64/atomic/wait/wait64.wat
index 36e5672349..92e9279f34 100644
--- a/tests/disas/winch/x64/atomic/wait/wait64.wat
+++ b/tests/disas/winch/x64/atomic/wait/wait64.wat
@@ -29,7 +29,7 @@
;; movq 0x18(%rsp), %rdx
;; movq 0x10(%rsp), %rcx
;; movq 8(%rsp), %r8
-;; callq 0x17c
+;; callq 0x17a
;; addq $8, %rsp
;; addq $0x18, %rsp
;; movq 8(%rsp), %r14
diff --git a/tests/disas/winch/x64/atomic/wait/wait64_offset.wat b/tests/disas/winch/x64/atomic/wait/wait64_offset.wat
index 83da4af2f4..60278a7855 100644
--- a/tests/disas/winch/x64/atomic/wait/wait64_offset.wat
+++ b/tests/disas/winch/x64/atomic/wait/wait64_offset.wat
@@ -33,7 +33,7 @@
;; movq 0x18(%rsp), %rdx
;; movq 0x10(%rsp), %rcx
;; movq 8(%rsp), %r8
-;; callq 0x183
+;; callq 0x181
;; addq $8, %rsp
;; addq $0x18, %rsp
;; movq 8(%rsp), %r14
diff --git a/tests/disas/winch/x64/call_indirect/call_indirect.wat b/tests/disas/winch/x64/call_indirect/call_indirect.wat
index 662c529014..d6ac1776bd 100644
--- a/tests/disas/winch/x64/call_indirect/call_indirect.wat
+++ b/tests/disas/winch/x64/call_indirect/call_indirect.wat
@@ -76,7 +76,7 @@
;; movq %r14, %rdi
;; movl $0, %esi
;; movl 8(%rsp), %edx
-;; callq 0x335
+;; callq 0x337
;; addq $8, %rsp
;; addq $4, %rsp
;; movq 0x1c(%rsp), %r14
@@ -128,7 +128,7 @@
;; movq %r14, %rdi
;; movl $0, %esi
;; movl 4(%rsp), %edx
-;; callq 0x335
+;; callq 0x337
;; addq $4, %rsp
;; addq $4, %rsp
;; movq 0x20(%rsp), %r14
diff --git a/tests/disas/winch/x64/call_indirect/local_arg.wat b/tests/disas/winch/x64/call_indirect/local_arg.wat
index 2e0ceb0fe8..a28a35c29d 100644
--- a/tests/disas/winch/x64/call_indirect/local_arg.wat
+++ b/tests/disas/winch/x64/call_indirect/local_arg.wat
@@ -72,7 +72,7 @@
;; movq %r14, %rdi
;; movl $0, %esi
;; movl 8(%rsp), %edx
-;; callq 0x32a
+;; callq 0x32b
;; addq $8, %rsp
;; addq $4, %rsp
;; movq 0x1c(%rsp), %r14
diff --git a/tests/disas/winch/x64/f32_ceil/f32_ceil_param.wat b/tests/disas/winch/x64/f32_ceil/f32_ceil_param.wat
index ba73bd867c..69d565b6d8 100644
--- a/tests/disas/winch/x64/f32_ceil/f32_ceil_param.wat
+++ b/tests/disas/winch/x64/f32_ceil/f32_ceil_param.wat
@@ -26,7 +26,7 @@
;; subq $0xc, %rsp
;; movq %r14, %rdi
;; movss 0xc(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $0xc, %rsp
;; addq $4, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/f32_floor/f32_floor_param.wat b/tests/disas/winch/x64/f32_floor/f32_floor_param.wat
index 2dc210ed53..f9db481766 100644
--- a/tests/disas/winch/x64/f32_floor/f32_floor_param.wat
+++ b/tests/disas/winch/x64/f32_floor/f32_floor_param.wat
@@ -26,7 +26,7 @@
;; subq $0xc, %rsp
;; movq %r14, %rdi
;; movss 0xc(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $0xc, %rsp
;; addq $4, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/f32_nearest/f32_nearest_param.wat b/tests/disas/winch/x64/f32_nearest/f32_nearest_param.wat
index b58d858558..adc7fa0af7 100644
--- a/tests/disas/winch/x64/f32_nearest/f32_nearest_param.wat
+++ b/tests/disas/winch/x64/f32_nearest/f32_nearest_param.wat
@@ -26,7 +26,7 @@
;; subq $0xc, %rsp
;; movq %r14, %rdi
;; movss 0xc(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $0xc, %rsp
;; addq $4, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/f32_trunc/f32_trunc_param.wat b/tests/disas/winch/x64/f32_trunc/f32_trunc_param.wat
index 6da754bd28..9aebb3411a 100644
--- a/tests/disas/winch/x64/f32_trunc/f32_trunc_param.wat
+++ b/tests/disas/winch/x64/f32_trunc/f32_trunc_param.wat
@@ -26,7 +26,7 @@
;; subq $0xc, %rsp
;; movq %r14, %rdi
;; movss 0xc(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $0xc, %rsp
;; addq $4, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/f64_ceil/f64_ceil_param.wat b/tests/disas/winch/x64/f64_ceil/f64_ceil_param.wat
index 196eb051e3..cdc1b87b18 100644
--- a/tests/disas/winch/x64/f64_ceil/f64_ceil_param.wat
+++ b/tests/disas/winch/x64/f64_ceil/f64_ceil_param.wat
@@ -26,7 +26,7 @@
;; subq $8, %rsp
;; movq %r14, %rdi
;; movsd 8(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $8, %rsp
;; addq $8, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/f64_floor/f64_floor_param.wat b/tests/disas/winch/x64/f64_floor/f64_floor_param.wat
index a4fb3a0efd..2b4debf2fd 100644
--- a/tests/disas/winch/x64/f64_floor/f64_floor_param.wat
+++ b/tests/disas/winch/x64/f64_floor/f64_floor_param.wat
@@ -26,7 +26,7 @@
;; subq $8, %rsp
;; movq %r14, %rdi
;; movsd 8(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $8, %rsp
;; addq $8, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/f64_nearest/f64_nearest_param.wat b/tests/disas/winch/x64/f64_nearest/f64_nearest_param.wat
index ba87e47ed7..e78d9a01fe 100644
--- a/tests/disas/winch/x64/f64_nearest/f64_nearest_param.wat
+++ b/tests/disas/winch/x64/f64_nearest/f64_nearest_param.wat
@@ -26,7 +26,7 @@
;; subq $8, %rsp
;; movq %r14, %rdi
;; movsd 8(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $8, %rsp
;; addq $8, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/f64_trunc/f64_trunc_param.wat b/tests/disas/winch/x64/f64_trunc/f64_trunc_param.wat
index 7907a6752f..138fff2647 100644
--- a/tests/disas/winch/x64/f64_trunc/f64_trunc_param.wat
+++ b/tests/disas/winch/x64/f64_trunc/f64_trunc_param.wat
@@ -26,7 +26,7 @@
;; subq $8, %rsp
;; movq %r14, %rdi
;; movsd 8(%rsp), %xmm0
-;; callq 0xe1
+;; callq 0xdc
;; addq $8, %rsp
;; addq $8, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/load/grow_load.wat b/tests/disas/winch/x64/load/grow_load.wat
index 5df39a344d..f3ed3d2d78 100644
--- a/tests/disas/winch/x64/load/grow_load.wat
+++ b/tests/disas/winch/x64/load/grow_load.wat
@@ -65,7 +65,7 @@
;; movq %r14, %rdi
;; movl 0xc(%rsp), %esi
;; movl $0, %edx
-;; callq 0x2db
+;; callq 0x2e2
;; addq $0xc, %rsp
;; addq $4, %rsp
;; movq 0x58(%rsp), %r14
diff --git a/tests/disas/winch/x64/table/fill.wat b/tests/disas/winch/x64/table/fill.wat
index 93ec192302..a58d41451a 100644
--- a/tests/disas/winch/x64/table/fill.wat
+++ b/tests/disas/winch/x64/table/fill.wat
@@ -113,7 +113,7 @@
;; movq %r14, %rdi
;; movl $0, %esi
;; movl 0xc(%rsp), %edx
-;; callq 0x4eb
+;; callq 0x4ee
;; addq $0xc, %rsp
;; addq $4, %rsp
;; movq 0x28(%rsp), %r14
@@ -133,7 +133,7 @@
;; movl 0xc(%rsp), %edx
;; movq 4(%rsp), %rcx
;; movl (%rsp), %r8d
-;; callq 0x517
+;; callq 0x51a
;; addq $0x10, %rsp
;; movq 0x28(%rsp), %r14
;; addq $0x30, %rsp
diff --git a/tests/disas/winch/x64/table/get.wat b/tests/disas/winch/x64/table/get.wat
index 99579ba396..03febe2ed8 100644
--- a/tests/disas/winch/x64/table/get.wat
+++ b/tests/disas/winch/x64/table/get.wat
@@ -65,7 +65,7 @@
;; movq %r14, %rdi
;; movl $0, %esi
;; movl 0xc(%rsp), %edx
-;; callq 0x2ec
+;; callq 0x2ef
;; addq $0xc, %rsp
;; addq $4, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/table/grow.wat b/tests/disas/winch/x64/table/grow.wat
index 3b7a274e4a..a6dfa4e1a5 100644
--- a/tests/disas/winch/x64/table/grow.wat
+++ b/tests/disas/winch/x64/table/grow.wat
@@ -30,7 +30,7 @@
;; movl $0, %esi
;; movl $0xa, %edx
;; movq 8(%rsp), %rcx
-;; callq 0x176
+;; callq 0x178
;; addq $8, %rsp
;; addq $8, %rsp
;; movq 0x18(%rsp), %r14
diff --git a/tests/disas/winch/x64/table/set.wat b/tests/disas/winch/x64/table/set.wat
index 44fe068696..2ae1255aee 100644
--- a/tests/disas/winch/x64/table/set.wat
+++ b/tests/disas/winch/x64/table/set.wat
@@ -109,7 +109,7 @@
;; movq %r14, %rdi
;; movl $0, %esi
;; movl 8(%rsp), %edx
-;; callq 0x4b6
+;; callq 0x4ba
;; addq $8, %rsp
;; addq $4, %rsp
;; movq 0x1c(%rsp), %r14