## Top method improvements
-19 (-3.11 % of base) - System.Number:DiyFp128Log1p(System.Number+DiyFp128,bool,byref):System.Number+DiyFp128
```diff
; Assembly listing for method System.Number:DiyFp128Log1p(System.Number+DiyFp128,bool,byref):System.Number+DiyFp128 (FullOpts)
; Emitting BLENDED_CODE for x64 + VEX + EVEX on Unix
; FullOpts code
; optimized code
; rbp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 4 single block inlinees; 1 inlinees without PGO data
+; 0 inlinees with PGO data; 7 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 RetBuf [V00,T00] ( 5, 4 ) long -> rbx single-def
; V01 arg0 [V01 ] ( 6, 4 ) struct (24) [rbp+0x10] do-not-enreg[XSF] addr-exposed ld-addr-op single-def
; V02 arg1 [V02,T02] ( 4, 3 ) ubyte -> r14 single-def
; V03 arg2 [V03,T01] ( 4, 3 ) byref -> r15 single-def
; V04 loc0 [V04,T03] ( 3, 2.50) int -> rdi
-; V05 loc1 [V05,T06] ( 2, 1.50) int -> rax
+; V05 loc1 [V05,T05] ( 2, 1.50) int -> rax
;* V06 loc2 [V06 ] ( 0, 0 ) ubyte -> zero-ref
-; V07 loc3 [V07,T05] ( 3, 1.50) long -> rdi
-; V08 loc4 [V08 ] ( 3, 1.50) struct (24) [rbp-0x38] do-not-enreg[XS] addr-exposed ld-addr-op
-; V09 loc5 [V09 ] ( 3, 1.50) struct (24) [rbp-0x50] do-not-enreg[XS] addr-exposed ld-addr-op
-; V10 loc6 [V10 ] ( 8, 4 ) struct (24) [rbp-0x68] do-not-enreg[XS] addr-exposed ld-addr-op
-; V11 loc7 [V11 ] ( 4, 2 ) struct (24) [rbp-0x80] do-not-enreg[XDAS] addr-exposed defined-via-address ld-addr-op
-; V12 loc8 [V12 ] ( 2, 1 ) struct (24) [rbp-0x98] do-not-enreg[XS] addr-exposed ld-addr-op
-; V13 loc9 [V13 ] ( 3, 1.50) struct (24) [rbp-0xB0] do-not-enreg[XS] addr-exposed ld-addr-op
+; V07 loc3 [V07,T04] ( 3, 1.50) long -> rdi
+; V08 loc4 [V08 ] ( 3, 1.50) struct (24) [rbp-0x30] do-not-enreg[XS] addr-exposed ld-addr-op
+; V09 loc5 [V09 ] ( 3, 1.50) struct (24) [rbp-0x48] do-not-enreg[XS] addr-exposed ld-addr-op
+; V10 loc6 [V10 ] ( 8, 4 ) struct (24) [rbp-0x60] do-not-enreg[XS] addr-exposed ld-addr-op
+; V11 loc7 [V11 ] ( 4, 2 ) struct (24) [rbp-0x78] do-not-enreg[XDAS] addr-exposed defined-via-address ld-addr-op
+; V12 loc8 [V12 ] ( 2, 1 ) struct (24) [rbp-0x90] do-not-enreg[XS] addr-exposed ld-addr-op
+; V13 loc9 [V13 ] ( 3, 1.50) struct (24) [rbp-0xA8] do-not-enreg[XS] addr-exposed ld-addr-op
; V14 OutArgs [V14 ] ( 1, 1 ) struct (24) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
-; V15 tmp1 [V15,T08] ( 2, 1 ) long -> rcx
+; V15 tmp1 [V15,T06] ( 2, 1 ) long -> rcx
;* V16 tmp2 [V16 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
;* V17 tmp3 [V17 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
;* V18 tmp4 [V18 ] ( 0, 0 ) struct (16) zero-ref "spilled call-like call argument"
-;* V19 tmp5 [V19 ] ( 0, 0 ) ref -> zero-ref class-hnd exact "Inlining Arg"
-;* V20 tmp6 [V20 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
-; V21 tmp7 [V21,T04] ( 2, 2 ) ref -> r13 class-hnd exact single-def "Inlining Arg"
-; V22 tmp8 [V22 ] ( 3, 1.50) struct (24) [rbp-0xC8] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
-; V23 tmp9 [V23 ] ( 4, 2 ) struct (24) [rbp-0xE0] do-not-enreg[XS] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
-;* V24 tmp10 [V24 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
-; V25 tmp11 [V25 ] ( 2, 1 ) struct (24) [rbp-0xF8] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
-; V26 tmp12 [V26 ] ( 6, 3 ) int -> [rbp-0x68] do-not-enreg[X] addr-exposed "field V10._sign (fldOffset=0x0)" P-DEP
-; V27 tmp13 [V27 ] ( 8, 4 ) int -> [rbp-0x64] do-not-enreg[X] addr-exposed "field V10._exponent (fldOffset=0x4)" P-DEP
-; V28 tmp14 [V28 ] ( 6, 3 ) long -> [rbp-0x60] do-not-enreg[X] addr-exposed "field V10._hi (fldOffset=0x8)" P-DEP
-; V29 tmp15 [V29 ] ( 6, 3 ) long -> [rbp-0x58] do-not-enreg[X] addr-exposed "field V10._lo (fldOffset=0x10)" P-DEP
-;* V30 tmp16 [V30,T09] ( 0, 0 ) byref -> zero-ref single-def "field V16._reference (fldOffset=0x0)" P-INDEP
-;* V31 tmp17 [V31 ] ( 0, 0 ) int -> zero-ref "field V16._length (fldOffset=0x8)" P-INDEP
-;* V32 tmp18 [V32,T10] ( 0, 0 ) byref -> zero-ref single-def "field V17._reference (fldOffset=0x0)" P-INDEP
-;* V33 tmp19 [V33 ] ( 0, 0 ) int -> zero-ref "field V17._length (fldOffset=0x8)" P-INDEP
-;* V34 tmp20 [V34 ] ( 0, 0 ) byref -> zero-ref single-def "field V18._reference (fldOffset=0x0)" P-INDEP
-;* V35 tmp21 [V35 ] ( 0, 0 ) int -> zero-ref "field V18._length (fldOffset=0x8)" P-INDEP
-; V36 tmp22 [V36,T07] ( 2, 1 ) byref -> r13 "field V20._reference (fldOffset=0x0)" P-INDEP
-;* V37 tmp23 [V37,T11] ( 0, 0 ) int -> zero-ref "field V20._length (fldOffset=0x8)" P-INDEP
-; V38 tmp24 [V38 ] ( 3, 1.50) int -> [rbp-0xE0] do-not-enreg[X] addr-exposed "field V23._sign (fldOffset=0x0)" P-DEP
-; V39 tmp25 [V39 ] ( 4, 2 ) int -> [rbp-0xDC] do-not-enreg[X] addr-exposed "field V23._exponent (fldOffset=0x4)" P-DEP
-; V40 tmp26 [V40 ] ( 3, 1.50) long -> [rbp-0xD8] do-not-enreg[X] addr-exposed "field V23._hi (fldOffset=0x8)" P-DEP
-; V41 tmp27 [V41 ] ( 3, 1.50) long -> [rbp-0xD0] do-not-enreg[X] addr-exposed "field V23._lo (fldOffset=0x10)" P-DEP
+;* V19 tmp5 [V19 ] ( 0, 0 ) struct (16) zero-ref "ReadOnlySpan for CreateSpan"
+;* V20 tmp6 [V20 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V21 tmp7 [V21 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V22 tmp8 [V22 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V23 tmp9 [V23 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V24 tmp10 [V24 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V25 tmp11 [V25 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
+;* V26 tmp12 [V26 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V27 tmp13 [V27 ] ( 0, 0 ) byref -> zero-ref "Inlining Arg"
+; V28 tmp14 [V28 ] ( 3, 1.50) struct (24) [rbp-0xC0] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
+; V29 tmp15 [V29 ] ( 4, 2 ) struct (24) [rbp-0xD8] do-not-enreg[XS] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+;* V30 tmp16 [V30 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
+; V31 tmp17 [V31 ] ( 2, 1 ) struct (24) [rbp-0xF0] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
+; V32 tmp18 [V32 ] ( 6, 3 ) int -> [rbp-0x60] do-not-enreg[X] addr-exposed "field V10._sign (fldOffset=0x0)" P-DEP
+; V33 tmp19 [V33 ] ( 8, 4 ) int -> [rbp-0x5C] do-not-enreg[X] addr-exposed "field V10._exponent (fldOffset=0x4)" P-DEP
+; V34 tmp20 [V34 ] ( 6, 3 ) long -> [rbp-0x58] do-not-enreg[X] addr-exposed "field V10._hi (fldOffset=0x8)" P-DEP
+; V35 tmp21 [V35 ] ( 6, 3 ) long -> [rbp-0x50] do-not-enreg[X] addr-exposed "field V10._lo (fldOffset=0x10)" P-DEP
+;* V36 tmp22 [V36,T07] ( 0, 0 ) byref -> zero-ref single-def "field V16._reference (fldOffset=0x0)" P-INDEP
+;* V37 tmp23 [V37 ] ( 0, 0 ) int -> zero-ref "field V16._length (fldOffset=0x8)" P-INDEP
+;* V38 tmp24 [V38,T08] ( 0, 0 ) byref -> zero-ref single-def "field V17._reference (fldOffset=0x0)" P-INDEP
+;* V39 tmp25 [V39 ] ( 0, 0 ) int -> zero-ref "field V17._length (fldOffset=0x8)" P-INDEP
+;* V40 tmp26 [V40 ] ( 0, 0 ) byref -> zero-ref single-def "field V18._reference (fldOffset=0x0)" P-INDEP
+;* V41 tmp27 [V41 ] ( 0, 0 ) int -> zero-ref "field V18._length (fldOffset=0x8)" P-INDEP
+;* V42 tmp28 [V42 ] ( 0, 0 ) byref -> zero-ref single-def "field V19._reference (fldOffset=0x0)" P-INDEP
+;* V43 tmp29 [V43 ] ( 0, 0 ) int -> zero-ref "field V19._length (fldOffset=0x8)" P-INDEP
+;* V44 tmp30 [V44 ] ( 0, 0 ) byref -> zero-ref single-def "field V22._reference (fldOffset=0x0)" P-INDEP
+;* V45 tmp31 [V45 ] ( 0, 0 ) int -> zero-ref "field V22._length (fldOffset=0x8)" P-INDEP
+;* V46 tmp32 [V46 ] ( 0, 0 ) byref -> zero-ref single-def "field V25._reference (fldOffset=0x0)" P-INDEP
+;* V47 tmp33 [V47 ] ( 0, 0 ) int -> zero-ref "field V25._length (fldOffset=0x8)" P-INDEP
+;* V48 tmp34 [V48 ] ( 0, 0 ) byref -> zero-ref single-def "field V26._reference (fldOffset=0x0)" P-INDEP
+;* V49 tmp35 [V49 ] ( 0, 0 ) int -> zero-ref "field V26._length (fldOffset=0x8)" P-INDEP
+; V50 tmp36 [V50 ] ( 3, 1.50) int -> [rbp-0xD8] do-not-enreg[X] addr-exposed "field V29._sign (fldOffset=0x0)" P-DEP
+; V51 tmp37 [V51 ] ( 4, 2 ) int -> [rbp-0xD4] do-not-enreg[X] addr-exposed "field V29._exponent (fldOffset=0x4)" P-DEP
+; V52 tmp38 [V52 ] ( 3, 1.50) long -> [rbp-0xD0] do-not-enreg[X] addr-exposed "field V29._hi (fldOffset=0x8)" P-DEP
+; V53 tmp39 [V53 ] ( 3, 1.50) long -> [rbp-0xC8] do-not-enreg[X] addr-exposed "field V29._lo (fldOffset=0x10)" P-DEP
;
-; Lcl frame size = 240
+; Lcl frame size = 248
G_M34379_IG01:
push rbp
push r15
push r14
- push r13
push rbx
- sub rsp, 240
+ sub rsp, 248
lea rbp, [rsp+0x110]
mov rbx, rdi
mov r14d, esi
mov r15, rdx
- ;; size=32 bbWeight=1 PerfScore 6.50
+ ;; size=30 bbWeight=1 PerfScore 5.50
G_M34379_IG02:
mov edi, dword ptr [rbp+0x14]
mov eax, dword ptr [rbp+0x10]
test edi, edi
jge SHORT G_M34379_IG04
;; size=10 bbWeight=1 PerfScore 3.25
G_M34379_IG03:
cmp edi, -2
jle G_M34379_IG05
mov rdi, qword ptr [rbp+0x18]
shr rdi, 2
mov rcx, rdi
neg rcx
test eax, eax
cmove rcx, rdi
mov rdi, 0xD1FFAB1E
add rdi, rcx
mov rax, 0xD1FFAB1E
cmp rdi, rax
jb SHORT G_M34379_IG05
;; size=57 bbWeight=0.50 PerfScore 3.00
G_M34379_IG04:
vxorps xmm0, xmm0, xmm0
- vmovdqu32 xmmword ptr [rbp-0xB0], xmm0
vmovdqu xmmword ptr [rbp-0xA8], xmm0
- lea rdi, [rbp-0x80]
+ vmovdqu32 xmmword ptr [rbp-0xA0], xmm0
+ lea rdi, [rbp-0x78]
mov rax, 0xD1FFAB1E ; code for System.Number:get_DiyFp128One():System.Number+DiyFp128
call [rax]System.Number:get_DiyFp128One():System.Number+DiyFp128
- lea rcx, bword ptr [rbp-0xB0]
+ lea rcx, bword ptr [rbp-0xA8]
mov r8d, 1
lea rsi, [rbp+0x10]
- lea rdi, [rbp-0x80]
+ lea rdi, [rbp-0x78]
xor edx, edx
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128AddSub(byref,byref,int,System.Span`1[System.Number+DiyFp128])
call [rax]System.Number:DiyFp128AddSub(byref,byref,int,System.Span`1[System.Number+DiyFp128])
- vmovdqu32 xmm0, xmmword ptr [rbp-0xB0]
+ vmovdqu xmm0, xmmword ptr [rbp-0xA8]
vmovdqu xmmword ptr [rsp], xmm0
- mov rsi, qword ptr [rbp-0xA0]
+ mov rsi, qword ptr [rbp-0x98]
mov qword ptr [rsp+0x10], rsi
movzx rsi, r14b
mov rdi, rbx
mov rdx, r15
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Log(System.Number+DiyFp128,bool,byref):System.Number+DiyFp128
call [rax]System.Number:DiyFp128Log(System.Number+DiyFp128,bool,byref):System.Number+DiyFp128
jmp G_M34379_IG08
- ;; size=121 bbWeight=0.50 PerfScore 10.67
+ ;; size=122 bbWeight=0.50 PerfScore 10.67
G_M34379_IG05:
vxorps xmm0, xmm0, xmm0
- vmovdqu xmmword ptr [rbp-0x38], xmm0
vmovdqu xmmword ptr [rbp-0x30], xmm0
- lea rdi, [rbp-0x80]
+ vmovdqu xmmword ptr [rbp-0x28], xmm0
+ lea rdi, [rbp-0x78]
mov rax, 0xD1FFAB1E ; code for System.Number:get_LogTwo():System.Number+DiyFp128
call [rax]System.Number:get_LogTwo():System.Number+DiyFp128
- lea rcx, bword ptr [rbp-0x38]
+ lea rcx, bword ptr [rbp-0x30]
mov r8d, 1
lea rsi, [rbp+0x10]
- lea rdi, [rbp-0x80]
+ lea rdi, [rbp-0x78]
xor edx, edx
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128AddSub(byref,byref,int,System.Span`1[System.Number+DiyFp128])
call [rax]System.Number:DiyFp128AddSub(byref,byref,int,System.Span`1[System.Number+DiyFp128])
lea rdi, [rbp+0x10]
- lea rsi, [rbp-0x38]
- lea rcx, [rbp-0x50]
+ lea rsi, [rbp-0x30]
+ lea rcx, [rbp-0x48]
mov edx, 2
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Divide(byref,byref,int,byref)
call [rax]System.Number:DiyFp128Divide(byref,byref,int,byref)
- mov rdi, 0xD1FFAB1E ; const ptr
- mov r13, gword ptr [rdi]
- add r13, 16
- vmovdqu xmm0, xmmword ptr [rbp-0x50]
- vmovdqu xmmword ptr [rbp-0xC8], xmm0
- mov rdi, qword ptr [rbp-0x40]
- mov qword ptr [rbp-0xB8], rdi
- lea rdi, [rbp-0xC8]
- lea rsi, [rbp-0xC8]
- lea rdx, [rbp-0xE0]
+ vmovdqu xmm0, xmmword ptr [rbp-0x48]
+ vmovdqu32 xmmword ptr [rbp-0xC0], xmm0
+ mov rdi, qword ptr [rbp-0x38]
+ mov qword ptr [rbp-0xB0], rdi
+ lea rdi, [rbp-0xC0]
+ lea rsi, [rbp-0xC0]
+ lea rdx, [rbp-0xD8]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Multiply(byref,byref,byref)
call [rax]System.Number:DiyFp128Multiply(byref,byref,byref)
- lea rdi, [rbp-0xE0]
+ lea rdi, [rbp-0xD8]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Normalize(byref)
call [rax]System.Number:DiyFp128Normalize(byref)
- lea rsi, [rbp-0x68]
+ lea rsi, [rbp-0x60]
mov qword ptr [rsp], rsi
- movsxd rsi, dword ptr [rbp-0xDC]
+ movsxd rsi, dword ptr [rbp-0xD4]
imul rsi, rsi, -17
- mov rdx, r13
+ mov rdx, 0xD1FFAB1E ; static handle
mov ecx, 18
- lea rdi, [rbp-0xE0]
+ lea rdi, [rbp-0xD8]
xor r8d, r8d
mov r9d, 17
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128EvaluatePositivePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
- ;; size=237 bbWeight=0.50 PerfScore 18.79
-G_M34379_IG06:
call [rax]System.Number:DiyFp128EvaluatePositivePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
- vmovdqu xmm0, xmmword ptr [rbp-0x50]
- vmovdqu xmmword ptr [rbp-0xF8], xmm0
- mov rdi, qword ptr [rbp-0x40]
- mov qword ptr [rbp-0xE8], rdi
- lea rdi, [rbp-0xF8]
- lea rsi, [rbp-0x68]
- lea rdx, [rbp-0x68]
+ vmovdqu xmm0, xmmword ptr [rbp-0x48]
+ ;; size=233 bbWeight=0.50 PerfScore 19.54
+G_M34379_IG06:
+ vmovdqu32 xmmword ptr [rbp-0xF0], xmm0
+ mov rdi, qword ptr [rbp-0x38]
+ mov qword ptr [rbp-0xE0], rdi
+ lea rdi, [rbp-0xF0]
+ lea rsi, [rbp-0x60]
+ lea rdx, [rbp-0x60]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Multiply(byref,byref,byref)
call [rax]System.Number:DiyFp128Multiply(byref,byref,byref)
- mov edi, dword ptr [rbp-0x64]
+ mov edi, dword ptr [rbp-0x5C]
add edi, 2
- mov dword ptr [rbp-0x64], edi
+ mov dword ptr [rbp-0x5C], edi
test r14b, r14b
je SHORT G_M34379_IG07
vmovdqu xmm0, xmmword ptr [r15]
- vmovdqu xmmword ptr [rbp-0x98], xmm0
+ vmovdqu32 xmmword ptr [rbp-0x90], xmm0
mov rdi, qword ptr [r15+0x10]
- mov qword ptr [rbp-0x88], rdi
- lea rdi, [rbp-0x68]
- lea rsi, [rbp-0x98]
- lea rdx, [rbp-0x68]
+ mov qword ptr [rbp-0x80], rdi
+ lea rdi, [rbp-0x60]
+ lea rsi, [rbp-0x90]
+ lea rdx, [rbp-0x60]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Multiply(byref,byref,byref)
call [rax]System.Number:DiyFp128Multiply(byref,byref,byref)
- ;; size=118 bbWeight=0.50 PerfScore 13.00
+ ;; size=106 bbWeight=0.50 PerfScore 11.00
G_M34379_IG07:
- vmovdqu xmm0, xmmword ptr [rbp-0x68]
+ vmovdqu xmm0, xmmword ptr [rbp-0x60]
vmovdqu xmmword ptr [rbx], xmm0
- mov rax, qword ptr [rbp-0x58]
+ mov rax, qword ptr [rbp-0x50]
mov qword ptr [rbx+0x10], rax
;; size=17 bbWeight=0.50 PerfScore 2.00
G_M34379_IG08:
mov rax, rbx
;; size=3 bbWeight=1 PerfScore 0.25
G_M34379_IG09:
- add rsp, 240
+ add rsp, 248
pop rbx
- pop r13
pop r14
pop r15
pop rbp
ret
- ;; size=16 bbWeight=1 PerfScore 3.75
+ ;; size=14 bbWeight=1 PerfScore 3.25
-; Total bytes of code 611, prolog size 23, PerfScore 61.21, instruction count 130, allocated bytes for code 611 (MethodHash=819c79b4) for method System.Number:DiyFp128Log1p(System.Number+DiyFp128,bool,byref):System.Number+DiyFp128 (FullOpts)
+; Total bytes of code 592, prolog size 21, PerfScore 58.46, instruction count 125, allocated bytes for code 592 (MethodHash=819c79b4) for method System.Number:DiyFp128Log1p(System.Number+DiyFp128,bool,byref):System.Number+DiyFp128 (FullOpts)
; ============================================================
```
-19 (-5.60 % of base) - System.Number:DiyFp128Tan(byref):System.Number+DiyFp128
```diff
; Assembly listing for method System.Number:DiyFp128Tan(byref):System.Number+DiyFp128 (FullOpts)
; Emitting BLENDED_CODE for x64 + VEX + EVEX on Unix
; FullOpts code
; optimized code
; rbp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 7 single block inlinees; 3 inlinees without PGO data
+; 0 inlinees with PGO data; 13 single block inlinees; 3 inlinees without PGO data
; Final local variable assignments
;
; V00 RetBuf [V00,T01] ( 5, 4 ) long -> rbx single-def
; V01 arg0 [V01,T00] ( 8, 5.50) byref -> rax single-def
;* V02 loc0 [V02 ] ( 0, 0 ) int -> zero-ref
; V03 loc1 [V03 ] ( 6, 4 ) struct (24) [rbp-0x28] do-not-enreg[XS] addr-exposed ld-addr-op
;* V04 loc2 [V04 ] ( 0, 0 ) int -> zero-ref
-; V05 loc3 [V05,T07] ( 2, 1 ) int -> rsi
+; V05 loc3 [V05,T03] ( 2, 1 ) int -> rsi
;* V06 loc4 [V06 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op
; V07 loc5 [V07 ] ( 7, 3.50) struct (48) [rbp-0x58] do-not-enreg[XS] addr-exposed ld-addr-op
; V08 OutArgs [V08 ] ( 1, 1 ) struct (40) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
-; V09 tmp1 [V09,T08] ( 2, 1 ) int -> rsi
+; V09 tmp1 [V09,T04] ( 2, 1 ) int -> rsi
;* V10 tmp2 [V10 ] ( 0, 0 ) struct (16) zero-ref "impAppendStmt"
;* V11 tmp3 [V11 ] ( 0, 0 ) struct (16) zero-ref "spilled call-like call argument"
;* V12 tmp4 [V12 ] ( 0, 0 ) byref -> zero-ref single-def "dup spill"
; V13 tmp5 [V13,T02] ( 4, 2 ) int -> r15 "Inline return value spill temp"
;* V14 tmp6 [V14 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "NewObj constructor temp"
-;* V15 tmp7 [V15 ] ( 0, 0 ) ref -> zero-ref class-hnd exact "Inlining Arg"
-;* V16 tmp8 [V16 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
-; V17 tmp9 [V17,T03] ( 2, 2 ) ref -> rdx class-hnd exact single-def "Inlining Arg"
-;* V18 tmp10 [V18 ] ( 0, 0 ) ref -> zero-ref class-hnd exact "Inlining Arg"
-;* V19 tmp11 [V19 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
-; V20 tmp12 [V20,T04] ( 2, 2 ) ref -> r8 class-hnd exact single-def "Inlining Arg"
-; V21 tmp13 [V21 ] ( 4, 2 ) int -> [rbp-0x28] do-not-enreg[X] addr-exposed "field V03._sign (fldOffset=0x0)" P-DEP
-; V22 tmp14 [V22 ] ( 4, 2 ) int -> [rbp-0x24] do-not-enreg[X] addr-exposed "field V03._exponent (fldOffset=0x4)" P-DEP
-; V23 tmp15 [V23 ] ( 5, 3 ) long -> [rbp-0x20] do-not-enreg[X] addr-exposed "field V03._hi (fldOffset=0x8)" P-DEP
-; V24 tmp16 [V24 ] ( 5, 3 ) long -> [rbp-0x18] do-not-enreg[X] addr-exposed "field V03._lo (fldOffset=0x10)" P-DEP
-;* V25 tmp17 [V25 ] ( 0, 0 ) byref -> zero-ref single-def "field V06._reference (fldOffset=0x0)" P-INDEP
-;* V26 tmp18 [V26 ] ( 0, 0 ) int -> zero-ref "field V06._length (fldOffset=0x8)" P-INDEP
-;* V27 tmp19 [V27 ] ( 0, 0 ) byref -> zero-ref single-def "field V10._reference (fldOffset=0x0)" P-INDEP
-;* V28 tmp20 [V28 ] ( 0, 0 ) int -> zero-ref "field V10._length (fldOffset=0x8)" P-INDEP
-;* V29 tmp21 [V29 ] ( 0, 0 ) byref -> zero-ref "field V11._reference (fldOffset=0x0)" P-INDEP
-;* V30 tmp22 [V30 ] ( 0, 0 ) int -> zero-ref "field V11._length (fldOffset=0x8)" P-INDEP
-;* V31 tmp23 [V31,T09] ( 0, 0 ) byref -> zero-ref single-def "field V14._reference (fldOffset=0x0)" P-INDEP
-;* V32 tmp24 [V32 ] ( 0, 0 ) int -> zero-ref "field V14._length (fldOffset=0x8)" P-INDEP
-; V33 tmp25 [V33,T05] ( 2, 1 ) byref -> rdx "field V16._reference (fldOffset=0x0)" P-INDEP
-;* V34 tmp26 [V34,T10] ( 0, 0 ) int -> zero-ref "field V16._length (fldOffset=0x8)" P-INDEP
-; V35 tmp27 [V35,T06] ( 2, 1 ) byref -> r8 "field V19._reference (fldOffset=0x0)" P-INDEP
-;* V36 tmp28 [V36,T11] ( 0, 0 ) int -> zero-ref "field V19._length (fldOffset=0x8)" P-INDEP
-;* V37 tmp29 [V37 ] ( 0, 0 ) byref -> zero-ref single-def "Zero offset field obj"
+;* V15 tmp7 [V15 ] ( 0, 0 ) struct (16) zero-ref "ReadOnlySpan for CreateSpan"
+;* V16 tmp8 [V16 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V17 tmp9 [V17 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V18 tmp10 [V18 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V19 tmp11 [V19 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V20 tmp12 [V20 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V21 tmp13 [V21 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
+;* V22 tmp14 [V22 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V23 tmp15 [V23 ] ( 0, 0 ) byref -> zero-ref "Inlining Arg"
+;* V24 tmp16 [V24 ] ( 0, 0 ) struct (16) zero-ref "ReadOnlySpan for CreateSpan"
+;* V25 tmp17 [V25 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V26 tmp18 [V26 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V27 tmp19 [V27 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V28 tmp20 [V28 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V29 tmp21 [V29 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V30 tmp22 [V30 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
+;* V31 tmp23 [V31 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V32 tmp24 [V32 ] ( 0, 0 ) byref -> zero-ref "Inlining Arg"
+; V33 tmp25 [V33 ] ( 4, 2 ) int -> [rbp-0x28] do-not-enreg[X] addr-exposed "field V03._sign (fldOffset=0x0)" P-DEP
+; V34 tmp26 [V34 ] ( 4, 2 ) int -> [rbp-0x24] do-not-enreg[X] addr-exposed "field V03._exponent (fldOffset=0x4)" P-DEP
+; V35 tmp27 [V35 ] ( 5, 3 ) long -> [rbp-0x20] do-not-enreg[X] addr-exposed "field V03._hi (fldOffset=0x8)" P-DEP
+; V36 tmp28 [V36 ] ( 5, 3 ) long -> [rbp-0x18] do-not-enreg[X] addr-exposed "field V03._lo (fldOffset=0x10)" P-DEP
+;* V37 tmp29 [V37 ] ( 0, 0 ) byref -> zero-ref single-def "field V06._reference (fldOffset=0x0)" P-INDEP
+;* V38 tmp30 [V38 ] ( 0, 0 ) int -> zero-ref "field V06._length (fldOffset=0x8)" P-INDEP
+;* V39 tmp31 [V39 ] ( 0, 0 ) byref -> zero-ref single-def "field V10._reference (fldOffset=0x0)" P-INDEP
+;* V40 tmp32 [V40 ] ( 0, 0 ) int -> zero-ref "field V10._length (fldOffset=0x8)" P-INDEP
+;* V41 tmp33 [V41 ] ( 0, 0 ) byref -> zero-ref "field V11._reference (fldOffset=0x0)" P-INDEP
+;* V42 tmp34 [V42 ] ( 0, 0 ) int -> zero-ref "field V11._length (fldOffset=0x8)" P-INDEP
+;* V43 tmp35 [V43,T05] ( 0, 0 ) byref -> zero-ref single-def "field V14._reference (fldOffset=0x0)" P-INDEP
+;* V44 tmp36 [V44 ] ( 0, 0 ) int -> zero-ref "field V14._length (fldOffset=0x8)" P-INDEP
+;* V45 tmp37 [V45 ] ( 0, 0 ) byref -> zero-ref single-def "field V15._reference (fldOffset=0x0)" P-INDEP
+;* V46 tmp38 [V46 ] ( 0, 0 ) int -> zero-ref "field V15._length (fldOffset=0x8)" P-INDEP
+;* V47 tmp39 [V47 ] ( 0, 0 ) byref -> zero-ref single-def "field V18._reference (fldOffset=0x0)" P-INDEP
+;* V48 tmp40 [V48 ] ( 0, 0 ) int -> zero-ref "field V18._length (fldOffset=0x8)" P-INDEP
+;* V49 tmp41 [V49 ] ( 0, 0 ) byref -> zero-ref single-def "field V21._reference (fldOffset=0x0)" P-INDEP
+;* V50 tmp42 [V50 ] ( 0, 0 ) int -> zero-ref "field V21._length (fldOffset=0x8)" P-INDEP
+;* V51 tmp43 [V51 ] ( 0, 0 ) byref -> zero-ref single-def "field V22._reference (fldOffset=0x0)" P-INDEP
+;* V52 tmp44 [V52 ] ( 0, 0 ) int -> zero-ref "field V22._length (fldOffset=0x8)" P-INDEP
+;* V53 tmp45 [V53 ] ( 0, 0 ) byref -> zero-ref single-def "field V24._reference (fldOffset=0x0)" P-INDEP
+;* V54 tmp46 [V54 ] ( 0, 0 ) int -> zero-ref "field V24._length (fldOffset=0x8)" P-INDEP
+;* V55 tmp47 [V55 ] ( 0, 0 ) byref -> zero-ref single-def "field V27._reference (fldOffset=0x0)" P-INDEP
+;* V56 tmp48 [V56 ] ( 0, 0 ) int -> zero-ref "field V27._length (fldOffset=0x8)" P-INDEP
+;* V57 tmp49 [V57 ] ( 0, 0 ) byref -> zero-ref single-def "field V30._reference (fldOffset=0x0)" P-INDEP
+;* V58 tmp50 [V58 ] ( 0, 0 ) int -> zero-ref "field V30._length (fldOffset=0x8)" P-INDEP
+;* V59 tmp51 [V59 ] ( 0, 0 ) byref -> zero-ref single-def "field V31._reference (fldOffset=0x0)" P-INDEP
+;* V60 tmp52 [V60 ] ( 0, 0 ) int -> zero-ref "field V31._length (fldOffset=0x8)" P-INDEP
+;* V61 tmp53 [V61 ] ( 0, 0 ) byref -> zero-ref single-def "Zero offset field obj"
;
; Lcl frame size = 112
G_M64545_IG01:
push rbp
push r15
push rbx
sub rsp, 112
lea rbp, [rsp+0x80]
mov rbx, rdi
mov rax, rsi
;; size=22 bbWeight=1 PerfScore 4.25
G_M64545_IG02:
cmp byte ptr [rax+0x38], 0
jne SHORT G_M64545_IG04
;; size=6 bbWeight=1 PerfScore 3.00
G_M64545_IG03:
lea rdx, [rbp-0x28]
mov rdi, rax
xor esi, esi
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128RadianReduce(byref,int,byref):int
call [rax]System.Number:DiyFp128RadianReduce(byref,int,byref):int
mov r15d, eax
jmp SHORT G_M64545_IG05
;; size=26 bbWeight=0.50 PerfScore 3.25
G_M64545_IG04:
mov rsi, qword ptr [rax+0x20]
mov rdx, qword ptr [rax+0x28]
mov edi, dword ptr [rax+0x34]
mov ecx, dword ptr [rax+0x30]
lea r9, [rbp-0x28]
xor r8d, r8d
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128DecimalRadianReduce(uint,System.UInt128,int,int,byref):int
call [rax]System.Number:DiyFp128DecimalRadianReduce(uint,System.UInt128,int,int,byref):int
mov r15d, eax
;; size=36 bbWeight=0.50 PerfScore 6.12
G_M64545_IG05:
mov rsi, qword ptr [rbp-0x20]
or rsi, qword ptr [rbp-0x18]
je G_M64545_IG08
;; size=14 bbWeight=1 PerfScore 3.00
G_M64545_IG06:
mov esi, 512
xor edx, edx
test r15b, 1
cmove esi, edx
or esi, 206
vxorps ymm0, ymm0, ymm0
vmovdqu ymmword ptr [rbp-0x58], ymm0
vmovdqu xmmword ptr [rbp-0x38], xmm0
vxorps xmm0, xmm0, xmm0
vmovdqu xmmword ptr [rbp-0x58], xmm0
vmovdqu xmmword ptr [rbp-0x50], xmm0
vxorps xmm0, xmm0, xmm0
vmovdqu xmmword ptr [rbp-0x40], xmm0
vmovdqu xmmword ptr [rbp-0x38], xmm0
- mov rdx, 0xD1FFAB1E ; const ptr
- mov rdx, gword ptr [rdx]
- add rdx, 16
- mov r8, 0xD1FFAB1E ; const ptr
- mov r8, gword ptr [r8]
- add r8, 16
mov dword ptr [rsp], 1
mov dword ptr [rsp+0x08], 7
mov dword ptr [rsp+0x10], esi
lea rsi, bword ptr [rbp-0x58]
- mov r9d, 2
+ mov edx, 2
mov bword ptr [rsp+0x18], rsi
- mov dword ptr [rsp+0x20], r9d
- mov rsi, rdx
+ mov dword ptr [rsp+0x20], edx
+ mov rsi, 0xD1FFAB1E ; static handle
mov edx, 8
+ mov r8, 0xD1FFAB1E ; static handle
mov r9d, 8
lea rdi, [rbp-0x28]
mov ecx, 1
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128EvaluateRational(byref,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,int,int,System.Span`1[System.Number+DiyFp128])
call [rax]System.Number:DiyFp128EvaluateRational(byref,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,int,int,System.Span`1[System.Number+DiyFp128])
test r15b, 1
je SHORT G_M64545_IG07
mov eax, dword ptr [rbp-0x58]
xor eax, 0xD1FFAB1E
mov dword ptr [rbp-0x58], eax
- ;; size=187 bbWeight=0.50 PerfScore 13.75
+ ;; size=168 bbWeight=0.50 PerfScore 11.38
G_M64545_IG07:
vmovdqu xmm0, xmmword ptr [rbp-0x58]
vmovdqu xmmword ptr [rbx], xmm0
mov rax, qword ptr [rbp-0x48]
mov qword ptr [rbx+0x10], rax
jmp SHORT G_M64545_IG09
;; size=19 bbWeight=0.50 PerfScore 3.00
G_M64545_IG08:
vmovdqu xmm0, xmmword ptr [rbp-0x28]
vmovdqu xmmword ptr [rbx], xmm0
mov rax, qword ptr [rbp-0x18]
mov qword ptr [rbx+0x10], rax
;; size=17 bbWeight=0.50 PerfScore 2.00
G_M64545_IG09:
mov rax, rbx
;; size=3 bbWeight=1 PerfScore 0.25
G_M64545_IG10:
add rsp, 112
pop rbx
pop r15
pop rbp
ret
;; size=9 bbWeight=1 PerfScore 2.75
-; Total bytes of code 339, prolog size 16, PerfScore 41.38, instruction count 82, allocated bytes for code 339 (MethodHash=04d503de) for method System.Number:DiyFp128Tan(byref):System.Number+DiyFp128 (FullOpts)
+; Total bytes of code 320, prolog size 16, PerfScore 39.00, instruction count 77, allocated bytes for code 320 (MethodHash=04d503de) for method System.Number:DiyFp128Tan(byref):System.Number+DiyFp128 (FullOpts)
; ============================================================
```
-14 (-8.00 % of base) - System.Number:DiyFp128EvaluatePow2Polynomial(byref,byref)
```diff
; Assembly listing for method System.Number:DiyFp128EvaluatePow2Polynomial(byref,byref) (FullOpts)
; Emitting BLENDED_CODE for x64 + VEX + EVEX on Unix
; FullOpts code
; optimized code
; rbp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 2 single block inlinees; 2 inlinees without PGO data
+; 0 inlinees with PGO data; 8 single block inlinees; 2 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) byref -> rdi single-def
; V01 arg1 [V01,T00] ( 6, 5 ) byref -> rbx single-def
; V02 loc0 [V02 ] ( 6, 5 ) struct (24) [rbp-0x20] do-not-enreg[XS] addr-exposed ld-addr-op
; V03 loc1 [V03,T02] ( 3, 2 ) long -> rsi
; V04 OutArgs [V04 ] ( 1, 1 ) struct ( 8) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
;* V05 tmp1 [V05 ] ( 0, 0 ) struct (16) zero-ref "spilled call-like call argument"
;* V06 tmp2 [V06 ] ( 0, 0 ) byref -> zero-ref "dup spill"
;* V07 tmp3 [V07 ] ( 0, 0 ) struct (16) zero-ref "spilled call-like call argument"
-;* V08 tmp4 [V08 ] ( 0, 0 ) ref -> zero-ref class-hnd exact "Inlining Arg"
-;* V09 tmp5 [V09 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
-; V10 tmp6 [V10,T03] ( 2, 2 ) ref -> rdx class-hnd exact single-def "Inlining Arg"
-;* V11 tmp7 [V11 ] ( 0, 0 ) ref -> zero-ref class-hnd exact "Inlining Arg"
-;* V12 tmp8 [V12 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
-; V13 tmp9 [V13,T04] ( 2, 2 ) ref -> rdx class-hnd exact single-def "Inlining Arg"
-; V14 tmp10 [V14 ] ( 5, 4 ) int -> [rbp-0x20] do-not-enreg[X] addr-exposed "field V02._sign (fldOffset=0x0)" P-DEP
-; V15 tmp11 [V15 ] ( 5, 4 ) int -> [rbp-0x1C] do-not-enreg[X] addr-exposed "field V02._exponent (fldOffset=0x4)" P-DEP
-; V16 tmp12 [V16 ] ( 4, 3 ) long -> [rbp-0x18] do-not-enreg[X] addr-exposed "field V02._hi (fldOffset=0x8)" P-DEP
-; V17 tmp13 [V17 ] ( 4, 3 ) long -> [rbp-0x10] do-not-enreg[X] addr-exposed "field V02._lo (fldOffset=0x10)" P-DEP
-;* V18 tmp14 [V18 ] ( 0, 0 ) byref -> zero-ref "field V05._reference (fldOffset=0x0)" P-INDEP
-;* V19 tmp15 [V19 ] ( 0, 0 ) int -> zero-ref "field V05._length (fldOffset=0x8)" P-INDEP
-;* V20 tmp16 [V20 ] ( 0, 0 ) byref -> zero-ref "field V07._reference (fldOffset=0x0)" P-INDEP
-;* V21 tmp17 [V21 ] ( 0, 0 ) int -> zero-ref "field V07._length (fldOffset=0x8)" P-INDEP
-; V22 tmp18 [V22,T05] ( 2, 1 ) byref -> rdx "field V09._reference (fldOffset=0x0)" P-INDEP
-;* V23 tmp19 [V23,T07] ( 0, 0 ) int -> zero-ref "field V09._length (fldOffset=0x8)" P-INDEP
-; V24 tmp20 [V24,T06] ( 2, 1 ) byref -> rdx "field V12._reference (fldOffset=0x0)" P-INDEP
-;* V25 tmp21 [V25,T08] ( 0, 0 ) int -> zero-ref "field V12._length (fldOffset=0x8)" P-INDEP
+;* V08 tmp4 [V08 ] ( 0, 0 ) struct (16) zero-ref "ReadOnlySpan for CreateSpan"
+;* V09 tmp5 [V09 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V10 tmp6 [V10 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V11 tmp7 [V11 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V12 tmp8 [V12 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V13 tmp9 [V13 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V14 tmp10 [V14 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
+;* V15 tmp11 [V15 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V16 tmp12 [V16 ] ( 0, 0 ) byref -> zero-ref "Inlining Arg"
+;* V17 tmp13 [V17 ] ( 0, 0 ) struct (16) zero-ref "ReadOnlySpan for CreateSpan"
+;* V18 tmp14 [V18 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V19 tmp15 [V19 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V20 tmp16 [V20 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V21 tmp17 [V21 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V22 tmp18 [V22 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V23 tmp19 [V23 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
+;* V24 tmp20 [V24 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V25 tmp21 [V25 ] ( 0, 0 ) byref -> zero-ref "Inlining Arg"
+; V26 tmp22 [V26 ] ( 5, 4 ) int -> [rbp-0x20] do-not-enreg[X] addr-exposed "field V02._sign (fldOffset=0x0)" P-DEP
+; V27 tmp23 [V27 ] ( 5, 4 ) int -> [rbp-0x1C] do-not-enreg[X] addr-exposed "field V02._exponent (fldOffset=0x4)" P-DEP
+; V28 tmp24 [V28 ] ( 4, 3 ) long -> [rbp-0x18] do-not-enreg[X] addr-exposed "field V02._hi (fldOffset=0x8)" P-DEP
+; V29 tmp25 [V29 ] ( 4, 3 ) long -> [rbp-0x10] do-not-enreg[X] addr-exposed "field V02._lo (fldOffset=0x10)" P-DEP
+;* V30 tmp26 [V30 ] ( 0, 0 ) byref -> zero-ref "field V05._reference (fldOffset=0x0)" P-INDEP
+;* V31 tmp27 [V31 ] ( 0, 0 ) int -> zero-ref "field V05._length (fldOffset=0x8)" P-INDEP
+;* V32 tmp28 [V32 ] ( 0, 0 ) byref -> zero-ref "field V07._reference (fldOffset=0x0)" P-INDEP
+;* V33 tmp29 [V33 ] ( 0, 0 ) int -> zero-ref "field V07._length (fldOffset=0x8)" P-INDEP
+;* V34 tmp30 [V34 ] ( 0, 0 ) byref -> zero-ref single-def "field V08._reference (fldOffset=0x0)" P-INDEP
+;* V35 tmp31 [V35 ] ( 0, 0 ) int -> zero-ref "field V08._length (fldOffset=0x8)" P-INDEP
+;* V36 tmp32 [V36 ] ( 0, 0 ) byref -> zero-ref single-def "field V11._reference (fldOffset=0x0)" P-INDEP
+;* V37 tmp33 [V37 ] ( 0, 0 ) int -> zero-ref "field V11._length (fldOffset=0x8)" P-INDEP
+;* V38 tmp34 [V38 ] ( 0, 0 ) byref -> zero-ref single-def "field V14._reference (fldOffset=0x0)" P-INDEP
+;* V39 tmp35 [V39 ] ( 0, 0 ) int -> zero-ref "field V14._length (fldOffset=0x8)" P-INDEP
+;* V40 tmp36 [V40 ] ( 0, 0 ) byref -> zero-ref single-def "field V15._reference (fldOffset=0x0)" P-INDEP
+;* V41 tmp37 [V41 ] ( 0, 0 ) int -> zero-ref "field V15._length (fldOffset=0x8)" P-INDEP
+;* V42 tmp38 [V42 ] ( 0, 0 ) byref -> zero-ref single-def "field V17._reference (fldOffset=0x0)" P-INDEP
+;* V43 tmp39 [V43 ] ( 0, 0 ) int -> zero-ref "field V17._length (fldOffset=0x8)" P-INDEP
+;* V44 tmp40 [V44 ] ( 0, 0 ) byref -> zero-ref single-def "field V20._reference (fldOffset=0x0)" P-INDEP
+;* V45 tmp41 [V45 ] ( 0, 0 ) int -> zero-ref "field V20._length (fldOffset=0x8)" P-INDEP
+;* V46 tmp42 [V46 ] ( 0, 0 ) byref -> zero-ref single-def "field V23._reference (fldOffset=0x0)" P-INDEP
+;* V47 tmp43 [V47 ] ( 0, 0 ) int -> zero-ref "field V23._length (fldOffset=0x8)" P-INDEP
+;* V48 tmp44 [V48 ] ( 0, 0 ) byref -> zero-ref single-def "field V24._reference (fldOffset=0x0)" P-INDEP
+;* V49 tmp45 [V49 ] ( 0, 0 ) int -> zero-ref "field V24._length (fldOffset=0x8)" P-INDEP
;
; Lcl frame size = 40
G_M6143_IG01:
push rbp
push rbx
sub rsp, 40
lea rbp, [rsp+0x30]
mov rbx, rsi
;; size=14 bbWeight=1 PerfScore 3.00
G_M6143_IG02:
vmovdqu xmm0, xmmword ptr [rdi]
vmovdqu xmmword ptr [rbp-0x20], xmm0
mov rax, qword ptr [rdi+0x10]
mov qword ptr [rbp-0x10], rax
lea rdi, [rbp-0x20]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Normalize(byref)
call [rax]System.Number:DiyFp128Normalize(byref)
movsxd rdx, dword ptr [rbp-0x1C]
imul rsi, rdx, -22
cmp dword ptr [rbp-0x20], 0
jne SHORT G_M6143_IG04
;; size=47 bbWeight=1 PerfScore 14.75
G_M6143_IG03:
- mov rdx, 0xD1FFAB1E ; const ptr
- mov rdx, gword ptr [rdx]
- add rdx, 16
mov bword ptr [rsp], rbx
+ mov rdx, 0xD1FFAB1E ; static handle
mov ecx, 23
lea rdi, [rbp-0x20]
xor r8d, r8d
mov r9d, 22
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128EvaluatePositivePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
call [rax]System.Number:DiyFp128EvaluatePositivePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
jmp SHORT G_M6143_IG05
- ;; size=53 bbWeight=0.50 PerfScore 5.00
+ ;; size=46 bbWeight=0.50 PerfScore 3.88
G_M6143_IG04:
- mov rdx, 0xD1FFAB1E ; const ptr
- mov rdx, gword ptr [rdx]
- add rdx, 16
mov bword ptr [rsp], rbx
+ mov rdx, 0xD1FFAB1E ; static handle
mov ecx, 23
lea rdi, [rbp-0x20]
xor r8d, r8d
mov r9d, 22
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128EvaluateNegativePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
call [rax]System.Number:DiyFp128EvaluateNegativePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
- ;; size=51 bbWeight=0.50 PerfScore 4.00
+ ;; size=44 bbWeight=0.50 PerfScore 2.88
G_M6143_IG05:
inc dword ptr [rbx+0x04]
;; size=3 bbWeight=1 PerfScore 3.00
G_M6143_IG06:
add rsp, 40
pop rbx
pop rbp
ret
;; size=7 bbWeight=1 PerfScore 2.25
-; Total bytes of code 175, prolog size 11, PerfScore 32.00, instruction count 42, allocated bytes for code 175 (MethodHash=288de800) for method System.Number:DiyFp128EvaluatePow2Polynomial(byref,byref) (FullOpts)
+; Total bytes of code 161, prolog size 11, PerfScore 29.75, instruction count 38, allocated bytes for code 161 (MethodHash=288de800) for method System.Number:DiyFp128EvaluatePow2Polynomial(byref,byref) (FullOpts)
; ============================================================
```
-13 (-4.64 % of base) - System.Number:DiyFp128LogPoly(byref):System.Number+DiyFp128
```diff
; Assembly listing for method System.Number:DiyFp128LogPoly(byref):System.Number+DiyFp128 (FullOpts)
; Emitting BLENDED_CODE for x64 + VEX + EVEX on Unix
; FullOpts code
; optimized code
; rbp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 2 single block inlinees; 1 inlinees without PGO data
+; 0 inlinees with PGO data; 5 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 RetBuf [V00,T01] ( 4, 4 ) long -> r15 single-def
; V01 arg0 [V01,T00] ( 4, 4 ) byref -> rbx single-def
-; V02 loc0 [V02 ] ( 8, 8 ) struct (24) [rbp-0x30] do-not-enreg[XS] addr-exposed ld-addr-op
-; V03 loc1 [V03 ] ( 2, 2 ) struct (24) [rbp-0x48] do-not-enreg[XDAS] addr-exposed defined-via-address ld-addr-op
+; V02 loc0 [V02 ] ( 8, 8 ) struct (24) [rbp-0x28] do-not-enreg[XS] addr-exposed ld-addr-op
+; V03 loc1 [V03 ] ( 2, 2 ) struct (24) [rbp-0x40] do-not-enreg[XDAS] addr-exposed defined-via-address ld-addr-op
; V04 OutArgs [V04 ] ( 1, 1 ) struct ( 8) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
;* V05 tmp1 [V05 ] ( 0, 0 ) struct (16) zero-ref "spilled call-like call argument"
-;* V06 tmp2 [V06 ] ( 0, 0 ) ref -> zero-ref class-hnd exact "Inlining Arg"
-;* V07 tmp3 [V07 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
-; V08 tmp4 [V08,T02] ( 2, 4 ) ref -> r14 class-hnd exact single-def "Inlining Arg"
-; V09 tmp5 [V09 ] ( 3, 3 ) struct (24) [rbp-0x60] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
-; V10 tmp6 [V10 ] ( 4, 4 ) struct (24) [rbp-0x78] do-not-enreg[XS] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
-;* V11 tmp7 [V11 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
-; V12 tmp8 [V12 ] ( 2, 2 ) struct (24) [rbp-0x90] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
-; V13 tmp9 [V13 ] ( 6, 6 ) int -> [rbp-0x30] do-not-enreg[X] addr-exposed "field V02._sign (fldOffset=0x0)" P-DEP
-; V14 tmp10 [V14 ] ( 8, 8 ) int -> [rbp-0x2C] do-not-enreg[X] addr-exposed "field V02._exponent (fldOffset=0x4)" P-DEP
-; V15 tmp11 [V15 ] ( 6, 6 ) long -> [rbp-0x28] do-not-enreg[X] addr-exposed "field V02._hi (fldOffset=0x8)" P-DEP
-; V16 tmp12 [V16 ] ( 6, 6 ) long -> [rbp-0x20] do-not-enreg[X] addr-exposed "field V02._lo (fldOffset=0x10)" P-DEP
-;* V17 tmp13 [V17 ] ( 0, 0 ) byref -> zero-ref single-def "field V05._reference (fldOffset=0x0)" P-INDEP
-;* V18 tmp14 [V18 ] ( 0, 0 ) int -> zero-ref "field V05._length (fldOffset=0x8)" P-INDEP
-; V19 tmp15 [V19,T03] ( 2, 2 ) byref -> r14 "field V07._reference (fldOffset=0x0)" P-INDEP
-;* V20 tmp16 [V20,T04] ( 0, 0 ) int -> zero-ref "field V07._length (fldOffset=0x8)" P-INDEP
-; V21 tmp17 [V21 ] ( 3, 3 ) int -> [rbp-0x78] do-not-enreg[X] addr-exposed "field V10._sign (fldOffset=0x0)" P-DEP
-; V22 tmp18 [V22 ] ( 4, 4 ) int -> [rbp-0x74] do-not-enreg[X] addr-exposed "field V10._exponent (fldOffset=0x4)" P-DEP
-; V23 tmp19 [V23 ] ( 3, 3 ) long -> [rbp-0x70] do-not-enreg[X] addr-exposed "field V10._hi (fldOffset=0x8)" P-DEP
-; V24 tmp20 [V24 ] ( 3, 3 ) long -> [rbp-0x68] do-not-enreg[X] addr-exposed "field V10._lo (fldOffset=0x10)" P-DEP
+;* V06 tmp2 [V06 ] ( 0, 0 ) struct (16) zero-ref "ReadOnlySpan for CreateSpan"
+;* V07 tmp3 [V07 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V08 tmp4 [V08 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V09 tmp5 [V09 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V10 tmp6 [V10 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V11 tmp7 [V11 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V12 tmp8 [V12 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op "NewObj constructor temp"
+;* V13 tmp9 [V13 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op "Inlining Arg"
+;* V14 tmp10 [V14 ] ( 0, 0 ) byref -> zero-ref "Inlining Arg"
+; V15 tmp11 [V15 ] ( 3, 3 ) struct (24) [rbp-0x58] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
+; V16 tmp12 [V16 ] ( 4, 4 ) struct (24) [rbp-0x70] do-not-enreg[XS] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+;* V17 tmp13 [V17 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
+; V18 tmp14 [V18 ] ( 2, 2 ) struct (24) [rbp-0x88] do-not-enreg[XS] addr-exposed ld-addr-op "Inline stloc first use temp"
+; V19 tmp15 [V19 ] ( 6, 6 ) int -> [rbp-0x28] do-not-enreg[X] addr-exposed "field V02._sign (fldOffset=0x0)" P-DEP
+; V20 tmp16 [V20 ] ( 8, 8 ) int -> [rbp-0x24] do-not-enreg[X] addr-exposed "field V02._exponent (fldOffset=0x4)" P-DEP
+; V21 tmp17 [V21 ] ( 6, 6 ) long -> [rbp-0x20] do-not-enreg[X] addr-exposed "field V02._hi (fldOffset=0x8)" P-DEP
+; V22 tmp18 [V22 ] ( 6, 6 ) long -> [rbp-0x18] do-not-enreg[X] addr-exposed "field V02._lo (fldOffset=0x10)" P-DEP
+;* V23 tmp19 [V23 ] ( 0, 0 ) byref -> zero-ref single-def "field V05._reference (fldOffset=0x0)" P-INDEP
+;* V24 tmp20 [V24 ] ( 0, 0 ) int -> zero-ref "field V05._length (fldOffset=0x8)" P-INDEP
+;* V25 tmp21 [V25 ] ( 0, 0 ) byref -> zero-ref single-def "field V06._reference (fldOffset=0x0)" P-INDEP
+;* V26 tmp22 [V26 ] ( 0, 0 ) int -> zero-ref "field V06._length (fldOffset=0x8)" P-INDEP
+;* V27 tmp23 [V27 ] ( 0, 0 ) byref -> zero-ref single-def "field V09._reference (fldOffset=0x0)" P-INDEP
+;* V28 tmp24 [V28 ] ( 0, 0 ) int -> zero-ref "field V09._length (fldOffset=0x8)" P-INDEP
+;* V29 tmp25 [V29 ] ( 0, 0 ) byref -> zero-ref single-def "field V12._reference (fldOffset=0x0)" P-INDEP
+;* V30 tmp26 [V30 ] ( 0, 0 ) int -> zero-ref "field V12._length (fldOffset=0x8)" P-INDEP
+;* V31 tmp27 [V31 ] ( 0, 0 ) byref -> zero-ref single-def "field V13._reference (fldOffset=0x0)" P-INDEP
+;* V32 tmp28 [V32 ] ( 0, 0 ) int -> zero-ref "field V13._length (fldOffset=0x8)" P-INDEP
+; V33 tmp29 [V33 ] ( 3, 3 ) int -> [rbp-0x70] do-not-enreg[X] addr-exposed "field V16._sign (fldOffset=0x0)" P-DEP
+; V34 tmp30 [V34 ] ( 4, 4 ) int -> [rbp-0x6C] do-not-enreg[X] addr-exposed "field V16._exponent (fldOffset=0x4)" P-DEP
+; V35 tmp31 [V35 ] ( 3, 3 ) long -> [rbp-0x68] do-not-enreg[X] addr-exposed "field V16._hi (fldOffset=0x8)" P-DEP
+; V36 tmp32 [V36 ] ( 3, 3 ) long -> [rbp-0x60] do-not-enreg[X] addr-exposed "field V16._lo (fldOffset=0x10)" P-DEP
;
-; Lcl frame size = 136
+; Lcl frame size = 128
G_M56148_IG01:
push rbp
push r15
- push r14
push rbx
- sub rsp, 136
- lea rbp, [rsp+0xA0]
+ sub rsp, 128
+ lea rbp, [rsp+0x90]
mov r15, rdi
mov rbx, rsi
- ;; size=27 bbWeight=1 PerfScore 5.25
+ ;; size=25 bbWeight=1 PerfScore 4.25
G_M56148_IG02:
- mov rdi, 0xD1FFAB1E ; const ptr
- mov r14, gword ptr [rdi]
- add r14, 16
vmovdqu xmm0, xmmword ptr [rbx]
- vmovdqu xmmword ptr [rbp-0x60], xmm0
+ vmovdqu xmmword ptr [rbp-0x58], xmm0
mov rdi, qword ptr [rbx+0x10]
- mov qword ptr [rbp-0x50], rdi
- lea rdi, [rbp-0x60]
- lea rsi, [rbp-0x60]
- lea rdx, [rbp-0x78]
+ mov qword ptr [rbp-0x48], rdi
+ lea rdi, [rbp-0x58]
+ lea rsi, [rbp-0x58]
+ lea rdx, [rbp-0x70]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Multiply(byref,byref,byref)
call [rax]System.Number:DiyFp128Multiply(byref,byref,byref)
- lea rdi, [rbp-0x78]
+ lea rdi, [rbp-0x70]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Normalize(byref)
call [rax]System.Number:DiyFp128Normalize(byref)
- lea rsi, [rbp-0x30]
+ lea rsi, [rbp-0x28]
mov qword ptr [rsp], rsi
- movsxd rsi, dword ptr [rbp-0x74]
+ movsxd rsi, dword ptr [rbp-0x6C]
imul rsi, rsi, -17
- mov rdx, r14
+ mov rdx, 0xD1FFAB1E ; static handle
mov ecx, 18
- lea rdi, [rbp-0x78]
+ lea rdi, [rbp-0x70]
xor r8d, r8d
mov r9d, 17
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128EvaluatePositivePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
call [rax]System.Number:DiyFp128EvaluatePositivePolynomial(byref,long,System.ReadOnlySpan`1[System.Number+DiyFp128FixedCoefficient],int,long,byref)
vmovdqu xmm0, xmmword ptr [rbx]
- vmovdqu32 xmmword ptr [rbp-0x90], xmm0
+ vmovdqu xmmword ptr [rbp-0x88], xmm0
mov rdi, qword ptr [rbx+0x10]
- mov qword ptr [rbp-0x80], rdi
- lea rdi, [rbp-0x90]
- lea rsi, [rbp-0x30]
- lea rdx, [rbp-0x30]
+ mov qword ptr [rbp-0x78], rdi
+ lea rdi, [rbp-0x88]
+ lea rsi, [rbp-0x28]
+ lea rdx, [rbp-0x28]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Multiply(byref,byref,byref)
call [rax]System.Number:DiyFp128Multiply(byref,byref,byref)
- mov edi, dword ptr [rbp-0x2C]
+ mov edi, dword ptr [rbp-0x24]
add edi, 2
- mov dword ptr [rbp-0x2C], edi
- lea rdi, [rbp-0x48]
+ mov dword ptr [rbp-0x24], edi
+ lea rdi, [rbp-0x40]
mov rax, 0xD1FFAB1E ; code for System.Number:get_LogLn2():System.Number+DiyFp128
call [rax]System.Number:get_LogLn2():System.Number+DiyFp128
- lea rdi, [rbp-0x30]
- lea rsi, [rbp-0x48]
- lea rdx, [rbp-0x30]
+ lea rdi, [rbp-0x28]
+ lea rsi, [rbp-0x40]
+ lea rdx, [rbp-0x28]
mov rax, 0xD1FFAB1E ; code for System.Number:DiyFp128Multiply(byref,byref,byref)
call [rax]System.Number:DiyFp128Multiply(byref,byref,byref)
- vmovdqu xmm0, xmmword ptr [rbp-0x30]
+ vmovdqu xmm0, xmmword ptr [rbp-0x28]
vmovdqu xmmword ptr [r15], xmm0
- mov rax, qword ptr [rbp-0x20]
- ;; size=232 bbWeight=1 PerfScore 50.75
-G_M56148_IG03:
+ mov rax, qword ptr [rbp-0x18]
mov qword ptr [r15+0x10], rax
mov rax, r15
- ;; size=7 bbWeight=1 PerfScore 1.25
-G_M56148_IG04:
- add rsp, 136
+ ;; size=230 bbWeight=1 PerfScore 49.50
+G_M56148_IG03:
+ add rsp, 128
pop rbx
- pop r14
pop r15
pop rbp
ret
- ;; size=14 bbWeight=1 PerfScore 3.25
+ ;; size=12 bbWeight=1 PerfScore 2.75
-; Total bytes of code 280, prolog size 21, PerfScore 60.50, instruction count 65, allocated bytes for code 280 (MethodHash=8bea24ab) for method System.Number:DiyFp128LogPoly(byref):System.Number+DiyFp128 (FullOpts)
+; Total bytes of code 267, prolog size 19, PerfScore 56.50, instruction count 60, allocated bytes for code 267 (MethodHash=8bea24ab) for method System.Number:DiyFp128LogPoly(byref):System.Number+DiyFp128 (FullOpts)
; ============================================================
```
Note: some changes were skipped as they were too large to fit into a comment.