[JitDiff X64] [hez2010] Always inline number conversions #974
Comments
Top method regressions22 (169.23 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromChecked[long](long,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M34483_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M34483_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, rdi
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=29 bbWeight=1 PerfScore 10.83
G_M34483_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=be48794c) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 35, prolog size 1, PerfScore 13.58, instruction count 10, allocated bytes for code 35 (MethodHash=be48794c) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; ============================================================22 (169.23 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromSaturating[long](long,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M2048_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M2048_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, rdi
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=29 bbWeight=1 PerfScore 10.83
G_M2048_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=64a2f7ff) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 35, prolog size 1, PerfScore 13.58, instruction count 10, allocated bytes for code 35 (MethodHash=64a2f7ff) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; ============================================================22 (169.23 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromTruncating[long](long,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M19455_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M19455_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, rdi
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=29 bbWeight=1 PerfScore 10.83
G_M19455_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=97f9b400) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 35, prolog size 1, PerfScore 13.58, instruction count 10, allocated bytes for code 35 (MethodHash=97f9b400) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
; ============================================================21 (161.54 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromChecked[int](int,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[int](int,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) int -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M12403_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M12403_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[int](int,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, edi
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=28 bbWeight=1 PerfScore 10.83
G_M12403_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[int](int,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=89c2cf8c) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[int](int,byref):ubyte (FullOpts)
+; Total bytes of code 34, prolog size 1, PerfScore 13.58, instruction count 10, allocated bytes for code 34 (MethodHash=89c2cf8c) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[int](int,byref):ubyte (FullOpts)
; ============================================================21 (123.53 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromChecked[short](short,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[short](short,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) short -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) short -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M45331_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M45331_IG02:
- movsx rdi, di
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[short](short,byref):ubyte
- ;; size=14 bbWeight=1 PerfScore 0.50
+ movsx rax, di
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, eax
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=32 bbWeight=1 PerfScore 11.08
G_M45331_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[short](short,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 17, prolog size 0, PerfScore 2.50, instruction count 3, allocated bytes for code 17 (MethodHash=ea824eec) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[short](short,byref):ubyte (FullOpts)
+; Total bytes of code 38, prolog size 1, PerfScore 13.83, instruction count 11, allocated bytes for code 38 (MethodHash=ea824eec) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[short](short,byref):ubyte (FullOpts)
; ============================================================21 (161.54 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromSaturating[int](int,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[int](int,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) int -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M19072_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M19072_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[int](int,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, edi
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=28 bbWeight=1 PerfScore 10.83
G_M19072_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[int](int,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=0a23b57f) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[int](int,byref):ubyte (FullOpts)
+; Total bytes of code 34, prolog size 1, PerfScore 13.58, instruction count 10, allocated bytes for code 34 (MethodHash=0a23b57f) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[int](int,byref):ubyte (FullOpts)
; ============================================================21 (123.53 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromSaturating[short](short,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[short](short,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) short -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) short -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M59872_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M59872_IG02:
- movsx rdi, di
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[short](short,byref):ubyte
- ;; size=14 bbWeight=1 PerfScore 0.50
+ movsx rax, di
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, eax
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=32 bbWeight=1 PerfScore 11.08
G_M59872_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[short](short,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 17, prolog size 0, PerfScore 2.50, instruction count 3, allocated bytes for code 17 (MethodHash=8069161f) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[short](short,byref):ubyte (FullOpts)
+; Total bytes of code 38, prolog size 1, PerfScore 13.83, instruction count 11, allocated bytes for code 38 (MethodHash=8069161f) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[short](short,byref):ubyte (FullOpts)
; ============================================================21 (161.54 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromTruncating[int](int,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[int](int,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) int -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M54975_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M54975_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[int](int,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, edi
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=28 bbWeight=1 PerfScore 10.83
G_M54975_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[int](int,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=9e9a2940) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[int](int,byref):ubyte (FullOpts)
+; Total bytes of code 34, prolog size 1, PerfScore 13.58, instruction count 10, allocated bytes for code 34 (MethodHash=9e9a2940) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[int](int,byref):ubyte (FullOpts)
; ============================================================21 (123.53 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromTruncating[short](short,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[short](short,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 1 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) short -> rdi single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) short -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M4959_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rsi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M4959_IG02:
- movsx rdi, di
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[short](short,byref):ubyte
- ;; size=14 bbWeight=1 PerfScore 0.50
+ movsx rax, di
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, eax
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(float):System.Half
+ call [rax]System.Half:op_Explicit(float):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=32 bbWeight=1 PerfScore 11.08
G_M4959_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[short](short,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 17, prolog size 0, PerfScore 2.50, instruction count 3, allocated bytes for code 17 (MethodHash=3fefeca0) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[short](short,byref):ubyte (FullOpts)
+; Total bytes of code 38, prolog size 1, PerfScore 13.83, instruction count 11, allocated bytes for code 38 (MethodHash=3fefeca0) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[short](short,byref):ubyte (FullOpts)
; ============================================================13 (100.00 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromChecked[double](double,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[double](double,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) double -> mm0 single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rdi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M41267_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rdi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M41267_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[double](double,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(double):System.Half
+ call [rax]System.Half:op_Explicit(double):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=20 bbWeight=1 PerfScore 4.50
G_M41267_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[double](double,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=3e5a5ecc) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[double](double,byref):ubyte (FullOpts)
+; Total bytes of code 26, prolog size 1, PerfScore 7.25, instruction count 8, allocated bytes for code 26 (MethodHash=3e5a5ecc) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromChecked[double](double,byref):ubyte (FullOpts)
; ============================================================13 (100.00 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromSaturating[double](double,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[double](double,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) double -> mm0 single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rdi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M54784_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rdi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M54784_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[double](double,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(double):System.Half
+ call [rax]System.Half:op_Explicit(double):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=20 bbWeight=1 PerfScore 4.50
G_M54784_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[double](double,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=c4e229ff) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[double](double,byref):ubyte (FullOpts)
+; Total bytes of code 26, prolog size 1, PerfScore 7.25, instruction count 8, allocated bytes for code 26 (MethodHash=c4e229ff) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromSaturating[double](double,byref):ubyte (FullOpts)
; ============================================================13 (100.00 % of base) - System.Half:System.Numerics.INumberBase.TryConvertFromTruncating[double](double,byref):ubyte ; Assembly listing for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[double](double,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) double -> mm0 single-def
-; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rdi single-def
+; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rbx single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M59519_IG01:
- ;; size=0 bbWeight=1 PerfScore 0.00
+ push rbx
+ mov rbx, rdi
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M59519_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Half:TryConvertFrom[double](double,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ mov rax, 0xD1FFAB1E ; code for System.Half:op_Explicit(double):System.Half
+ call [rax]System.Half:op_Explicit(double):System.Half
+ mov word ptr [rbx], ax
+ mov eax, 1
+ ;; size=20 bbWeight=1 PerfScore 4.50
G_M59519_IG03:
- tail.jmp [rax]System.Half:TryConvertFrom[double](double,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ pop rbx
+ ret
+ ;; size=2 bbWeight=1 PerfScore 1.50
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=7f9d1780) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[double](double,byref):ubyte (FullOpts)
+; Total bytes of code 26, prolog size 1, PerfScore 7.25, instruction count 8, allocated bytes for code 26 (MethodHash=7f9d1780) for method System.Half:System.Numerics.INumberBase<System.Half>.TryConvertFromTruncating[double](double,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Double:System.Numerics.INumberBase.TryConvertFromChecked[long](long,byref):ubyte ; Assembly listing for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M2227_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M2227_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2sd xmm0, xmm0, rdi
+ vmovsd qword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M2227_IG03:
- tail.jmp [rax]System.Double:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=75a8f74c) for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=75a8f74c) for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Double:System.Numerics.INumberBase.TryConvertFromSaturating[long](long,byref):ubyte ; Assembly listing for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M18944_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M18944_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2sd xmm0, xmm0, rdi
+ vmovsd qword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M18944_IG03:
- tail.jmp [rax]System.Double:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=a51ab5ff) for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=a51ab5ff) for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Double:System.Numerics.INumberBase.TryConvertFromTruncating[long](long,byref):ubyte ; Assembly listing for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M36351_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M36351_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2sd xmm0, xmm0, rdi
+ vmovsd qword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M36351_IG03:
- tail.jmp [rax]System.Double:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=d8717200) for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=d8717200) for method System.Double:System.Numerics.INumberBase<System.Double>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase.TryConvertFromChecked[long](long,byref):ubyte ; Assembly listing for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
+;* V05 tmp3 [V05 ] ( 0, 0 ) struct ( 8) zero-ref ld-addr-op "NewObj constructor temp" <System.Runtime.InteropServices.NFloat>
+;* V06 tmp4 [V06 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+; V07 tmp5 [V07,T02] ( 2, 2 ) double -> mm0 "field V05._value (fldOffset=0x0)" P-INDEP
;
; Lcl frame size = 0
G_M42803_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M42803_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Runtime.InteropServices.NFloat:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2sd xmm0, xmm0, rdi
+ vmovsd qword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M42803_IG03:
- tail.jmp [rax]System.Runtime.InteropServices.NFloat:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=846858cc) for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=846858cc) for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase.TryConvertFromSaturating[long](long,byref):ubyte ; Assembly listing for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
+;* V05 tmp3 [V05 ] ( 0, 0 ) struct ( 8) zero-ref ld-addr-op "NewObj constructor temp" <System.Runtime.InteropServices.NFloat>
+;* V06 tmp4 [V06 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+; V07 tmp5 [V07,T02] ( 2, 2 ) double -> mm0 "field V05._value (fldOffset=0x0)" P-INDEP
;
; Lcl frame size = 0
G_M22656_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M22656_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Runtime.InteropServices.NFloat:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2sd xmm0, xmm0, rdi
+ vmovsd qword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M22656_IG03:
- tail.jmp [rax]System.Runtime.InteropServices.NFloat:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=e4f0a77f) for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=e4f0a77f) for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase.TryConvertFromTruncating[long](long,byref):ubyte ; Assembly listing for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
+;* V05 tmp3 [V05 ] ( 0, 0 ) struct ( 8) zero-ref ld-addr-op "NewObj constructor temp" <System.Runtime.InteropServices.NFloat>
+;* V06 tmp4 [V06 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+; V07 tmp5 [V07,T02] ( 2, 2 ) double -> mm0 "field V05._value (fldOffset=0x0)" P-INDEP
;
; Lcl frame size = 0
G_M40063_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M40063_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Runtime.InteropServices.NFloat:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2sd xmm0, xmm0, rdi
+ vmovsd qword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M40063_IG03:
- tail.jmp [rax]System.Runtime.InteropServices.NFloat:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=18476380) for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=18476380) for method System.Runtime.InteropServices.NFloat:System.Numerics.INumberBase<System.Runtime.InteropServices.NFloat>.TryConvertFromTruncating[long](long,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Single:System.Numerics.INumberBase.TryConvertFromChecked[long](long,byref):ubyte ; Assembly listing for method System.Single:System.Numerics.INumberBase<System.Single>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M43059_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M43059_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Single:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, rdi
+ vmovss dword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M43059_IG03:
- tail.jmp [rax]System.Single:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=188457cc) for method System.Single:System.Numerics.INumberBase<System.Single>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=188457cc) for method System.Single:System.Numerics.INumberBase<System.Single>.TryConvertFromChecked[long](long,byref):ubyte (FullOpts)
; ============================================================6 (46.15 % of base) - System.Single:System.Numerics.INumberBase.TryConvertFromSaturating[long](long,byref):ubyte ; Assembly listing for method System.Single:System.Numerics.INumberBase<System.Single>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rsp based frame
-; fully interruptible
+; partially interruptible
; No PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 1 inlinees without PGO data
; Final local variable assignments
;
; V00 arg0 [V00,T01] ( 3, 3 ) long -> rdi single-def
; V01 arg1 [V01,T00] ( 3, 3 ) byref -> rsi single-def
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
+;* V03 tmp1 [V03 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V04 tmp2 [V04 ] ( 0, 0 ) long -> zero-ref "Inline stloc first use temp"
;
; Lcl frame size = 0
G_M47488_IG01:
;; size=0 bbWeight=1 PerfScore 0.00
G_M47488_IG02:
- mov rax, 0xD1FFAB1E ; code for System.Single:TryConvertFrom[long](long,byref):ubyte
- ;; size=10 bbWeight=1 PerfScore 0.25
+ vxorps xmm0, xmm0, xmm0
+ vcvtsi2ss xmm0, xmm0, rdi
+ vmovss dword ptr [rsi], xmm0
+ mov eax, 1
+ ;; size=18 bbWeight=1 PerfScore 8.58
G_M47488_IG03:
- tail.jmp [rax]System.Single:TryConvertFrom[long](long,byref):ubyte
- ;; size=3 bbWeight=1 PerfScore 2.00
+ ret
+ ;; size=1 bbWeight=1 PerfScore 1.00
-; Total bytes of code 13, prolog size 0, PerfScore 2.25, instruction count 2, allocated bytes for code 13 (MethodHash=5300467f) for method System.Single:System.Numerics.INumberBase<System.Single>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
+; Total bytes of code 19, prolog size 0, PerfScore 9.58, instruction count 5, allocated bytes for code 19 (MethodHash=5300467f) for method System.Single:System.Numerics.INumberBase<System.Single>.TryConvertFromSaturating[long](long,byref):ubyte (FullOpts)
; ============================================================ |
Top method improvements-157 (-77.72 % of base) - System.Security.Cryptography.PemEncoding:g__IsLabelChar|7_0[double](double):ubyte ; Assembly listing for method System.Security.Cryptography.PemEncoding:<IsValidLabel>g__IsLabelChar|7_0[double](double):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
-; rbp based frame
+; rsp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 3 single block inlinees; 6 inlinees without PGO data
+; 0 inlinees with PGO data; 3 single block inlinees; 9 inlinees without PGO data
; Final local variable assignments
;
-; V00 arg0 [V00,T00] ( 4, 3.50) double -> [rbp-0x08] single-def
+; V00 arg0 [V00,T00] ( 4, 3.50) double -> mm0 single-def
;# V01 OutArgs [V01 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
-; V02 tmp1 [V02,T01] ( 2, 4 ) double -> [rbp-0x10] spill-single-def "impAppendStmt"
-; V03 tmp2 [V03 ] ( 3, 2.50) double -> [rbp-0x18] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+; V02 tmp1 [V02,T01] ( 2, 4 ) double -> mm1 "impAppendStmt"
+;* V03 tmp2 [V03 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V04 tmp3 [V04 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-;* V05 tmp4 [V05 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
-; V06 tmp5 [V06,T02] ( 2, 4 ) double -> mm0 "Inlining Arg"
-; V07 tmp6 [V07 ] ( 3, 2.50) double -> [rbp-0x20] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
-;* V08 tmp7 [V08 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-;* V09 tmp8 [V09 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
-; V10 tmp9 [V10,T03] ( 2, 4 ) double -> mm0 "Inlining Arg"
-; V11 tmp10 [V11 ] ( 3, 1.50) double -> [rbp-0x28] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
-;* V12 tmp11 [V12 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-;* V13 tmp12 [V13 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
-; V14 tmp13 [V14,T04] ( 2, 2 ) double -> mm0 "Inlining Arg"
+;* V05 tmp4 [V05 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V06 tmp5 [V06 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
+;* V07 tmp6 [V07 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+;* V08 tmp7 [V08 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
+;* V09 tmp8 [V09 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V10 tmp9 [V10 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V11 tmp10 [V11 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
+;* V12 tmp11 [V12 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+;* V13 tmp12 [V13 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
+;* V14 tmp13 [V14 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V15 tmp14 [V15 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V16 tmp15 [V16 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
+;* V17 tmp16 [V17 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
;
-; Lcl frame size = 48
+; Lcl frame size = 0
G_M46973_IG01:
- push rbp
- sub rsp, 48
- lea rbp, [rsp+0x30]
- vmovsd qword ptr [rbp-0x08], xmm0
- ;; size=15 bbWeight=1 PerfScore 2.75
+ ;; size=0 bbWeight=1 PerfScore 0.00
G_M46973_IG02:
- lea rsi, [rbp-0x18]
- mov edi, 33
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[uint](uint,byref):ubyte
- call [rax]System.Double:TryConvertFrom[uint](uint,byref):ubyte
- test eax, eax
- jne SHORT G_M46973_IG04
- ;; size=25 bbWeight=1 PerfScore 5.25
+ vsubsd xmm1, xmm0, qword ptr [reloc @RWD00]
+ vmovsd xmm2, qword ptr [reloc @RWD08]
+ vucomisd xmm2, xmm1
+ jae SHORT G_M46973_IG05
+ ;; size=22 bbWeight=1 PerfScore 11.00
G_M46973_IG03:
- mov rsi, 0xD1FFAB1E
- mov qword ptr [rbp-0x18], rsi
- ;; size=14 bbWeight=0.50 PerfScore 0.62
-G_M46973_IG04:
- vmovsd xmm0, qword ptr [rbp-0x18]
- vmovsd xmm1, qword ptr [rbp-0x08]
- vsubsd xmm0, xmm1, xmm0
- vmovsd qword ptr [rbp-0x10], xmm0
- lea rsi, [rbp-0x20]
- mov edi, 93
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[uint](uint,byref):ubyte
- call [rax]System.Double:TryConvertFrom[uint](uint,byref):ubyte
- test eax, eax
- jne SHORT G_M46973_IG06
- ;; size=44 bbWeight=1 PerfScore 15.25
-G_M46973_IG05:
- mov rax, 0xD1FFAB1E
- mov qword ptr [rbp-0x20], rax
- ;; size=14 bbWeight=0.50 PerfScore 0.62
-G_M46973_IG06:
- vmovsd xmm0, qword ptr [rbp-0x20]
- vucomisd xmm0, qword ptr [rbp-0x10]
- jae SHORT G_M46973_IG09
- ;; size=12 bbWeight=1 PerfScore 8.00
-G_M46973_IG07:
xor eax, eax
;; size=2 bbWeight=0.50 PerfScore 0.12
-G_M46973_IG08:
- add rsp, 48
- pop rbp
+G_M46973_IG04:
ret
- ;; size=6 bbWeight=0.50 PerfScore 0.88
-G_M46973_IG09:
- lea rsi, [rbp-0x28]
- mov edi, 45
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[ushort](ushort,byref):ubyte
- call [rax]System.Double:TryConvertFrom[ushort](ushort,byref):ubyte
- test eax, eax
- jne SHORT G_M46973_IG10
- mov rax, 0xD1FFAB1E
- mov qword ptr [rbp-0x28], rax
- ;; size=39 bbWeight=0.50 PerfScore 3.25
-G_M46973_IG10:
- vmovsd xmm0, qword ptr [rbp-0x28]
- vmovsd xmm1, qword ptr [rbp-0x08]
- vucomisd xmm1, xmm0
+ ;; size=1 bbWeight=0.50 PerfScore 0.50
+G_M46973_IG05:
+ vucomisd xmm0, qword ptr [reloc @RWD16]
setp al
- jp SHORT G_M46973_IG11
+ jp SHORT G_M46973_IG06
setne al
- ;; size=22 bbWeight=0.50 PerfScore 5.50
-G_M46973_IG11:
+ ;; size=16 bbWeight=0.50 PerfScore 3.50
+G_M46973_IG06:
movzx rax, al
;; size=3 bbWeight=0.50 PerfScore 0.12
-G_M46973_IG12:
- add rsp, 48
- pop rbp
+G_M46973_IG07:
ret
- ;; size=6 bbWeight=0.50 PerfScore 0.88
+ ;; size=1 bbWeight=0.50 PerfScore 0.50
+RWD00 dq 4040800000000000h ; 33
+RWD08 dq 4057400000000000h ; 93
+RWD16 dq 4046800000000000h ; 45
-; Total bytes of code 202, prolog size 10, PerfScore 43.25, instruction count 49, allocated bytes for code 202 (MethodHash=7c174882) for method System.Security.Cryptography.PemEncoding:<IsValidLabel>g__IsLabelChar|7_0[double](double):ubyte (FullOpts)
+
+; Total bytes of code 45, prolog size 0, PerfScore 15.75, instruction count 12, allocated bytes for code 45 (MethodHash=7c174882) for method System.Security.Cryptography.PemEncoding:<IsValidLabel>g__IsLabelChar|7_0[double](double):ubyte (FullOpts)
; ============================================================-149 (-56.02 % of base) - System.Formats.Tar.TarHelpers:ParseOctal[double](System.ReadOnlySpan`1[ubyte]):double ; Assembly listing for method System.Formats.Tar.TarHelpers:ParseOctal[double](System.ReadOnlySpan`1[ubyte]):double (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rbp based frame
-; partially interruptible
+; fully interruptible
; No PGO data
-; 0 inlinees with PGO data; 6 single block inlinees; 4 inlinees without PGO data
+; 0 inlinees with PGO data; 6 single block inlinees; 6 inlinees without PGO data
; Final local variable assignments
;
;* V00 arg0 [V00 ] ( 0, 0 ) struct (16) zero-ref multireg-arg multireg-ret ld-addr-op <System.ReadOnlySpan`1[ubyte]>
-; V01 loc0 [V01,T10] ( 2, 4.25) double -> [rbp-0x28] spill-single-def
-; V02 loc1 [V02,T09] ( 4, 8.75) double -> mm1
+; V01 loc0 [V01,T10] ( 2, 4.25) double -> mm0
+; V02 loc1 [V02,T08] ( 4, 8.75) double -> mm1
;* V03 loc2 [V03 ] ( 0, 0 ) struct (16) zero-ref ld-addr-op <System.ReadOnlySpan`1[ubyte]>
;* V04 loc3 [V04,T06] ( 0, 0 ) int -> zero-ref
-; V05 loc4 [V05,T01] ( 4, 14 ) int -> r13
+; V05 loc4 [V05,T01] ( 3, 12 ) int -> rcx
;# V06 OutArgs [V06 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
-; V07 tmp1 [V07,T07] ( 2, 16 ) double -> [rbp-0x30] spill-single-def "impAppendStmt"
-; V08 tmp2 [V08 ] ( 3, 1.25) double -> [rbp-0x38] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+; V07 tmp1 [V07,T07] ( 2, 16 ) double -> mm1 "impAppendStmt"
+;* V08 tmp2 [V08 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V09 tmp3 [V09 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-;* V10 tmp4 [V10 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
-; V11 tmp5 [V11 ] ( 3, 10 ) double -> [rbp-0x40] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
-;* V12 tmp6 [V12 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-;* V13 tmp7 [V13 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
-; V14 tmp8 [V14,T08] ( 2, 16 ) double -> mm0 "Inlining Arg"
-; V15 tmp9 [V15,T03] ( 6, 5.25) byref -> r15 single-def "field V00._reference (fldOffset=0x0)" P-INDEP
-; V16 tmp10 [V16,T02] ( 7, 6.25) int -> rbx single-def "field V00._length (fldOffset=0x8)" P-INDEP
-; V17 tmp11 [V17,T04] ( 2, 4.25) byref -> r15 single-def "field V03._reference (fldOffset=0x0)" P-INDEP
-; V18 tmp12 [V18,T05] ( 2, 4.25) int -> rbx "field V03._length (fldOffset=0x8)" P-INDEP
-; V19 rat0 [V19,T00] ( 5, 16.25) long -> r14 "Widened IV V04"
+;* V10 tmp4 [V10 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V11 tmp5 [V11 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
+; V12 tmp6 [V12,T09] ( 2, 8 ) double -> mm2 ld-addr-op "Inline ldloca(s) first use temp"
+;* V13 tmp7 [V13 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V14 tmp8 [V14 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V15 tmp9 [V15 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
+;* V16 tmp10 [V16 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+; V17 tmp11 [V17,T03] ( 6, 5.25) byref -> rdi single-def "field V00._reference (fldOffset=0x0)" P-INDEP
+; V18 tmp12 [V18,T02] ( 7, 6.25) int -> rsi single-def "field V00._length (fldOffset=0x8)" P-INDEP
+; V19 tmp13 [V19,T04] ( 2, 4.25) byref -> rdi single-def "field V03._reference (fldOffset=0x0)" P-INDEP
+; V20 tmp14 [V20,T05] ( 2, 4.25) int -> rsi "field V03._length (fldOffset=0x8)" P-INDEP
+; V21 rat0 [V21,T00] ( 5, 16.25) long -> rax "Widened IV V04"
;
-; Lcl frame size = 32
+; Lcl frame size = 0
G_M18170_IG01:
push rbp
- push r15
- push r14
- push r13
- push rbx
- sub rsp, 32
- lea rbp, [rsp+0x40]
- mov r15, rdi
- mov ebx, esi
- ;; size=22 bbWeight=1 PerfScore 6.25
+ mov rbp, rsp
+ ;; size=4 bbWeight=1 PerfScore 1.25
G_M18170_IG02:
- mov rdi, r15
- mov esi, ebx
mov rax, 0xD1FFAB1E ; code for System.Formats.Tar.TarHelpers:TrimEndingNullsAndSpaces(System.ReadOnlySpan`1[ubyte]):System.ReadOnlySpan`1[ubyte]
call [rax]System.Formats.Tar.TarHelpers:TrimEndingNullsAndSpaces(System.ReadOnlySpan`1[ubyte]):System.ReadOnlySpan`1[ubyte]
- mov r15, rax
- mov ebx, edx
- mov rdi, r15
- mov esi, ebx
+ mov rdi, rax
+ mov esi, edx
mov rax, 0xD1FFAB1E ; code for System.Formats.Tar.TarHelpers:TrimLeadingNullsAndSpaces(System.ReadOnlySpan`1[ubyte]):System.ReadOnlySpan`1[ubyte]
call [rax]System.Formats.Tar.TarHelpers:TrimLeadingNullsAndSpaces(System.ReadOnlySpan`1[ubyte]):System.ReadOnlySpan`1[ubyte]
- mov r15, rax
- mov ebx, edx
- test ebx, ebx
- je G_M18170_IG11
- ;; size=52 bbWeight=1 PerfScore 9.75
+ mov rdi, rax
+ mov esi, edx
+ test esi, esi
+ je SHORT G_M18170_IG08
+ ;; size=38 bbWeight=1 PerfScore 8.75
G_M18170_IG03:
- lea rsi, [rbp-0x38]
- mov edi, 8
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[uint](uint,byref):ubyte
- call [rax]System.Double:TryConvertFrom[uint](uint,byref):ubyte
- test eax, eax
- jne SHORT G_M18170_IG04
- mov rsi, 0xD1FFAB1E
- mov qword ptr [rbp-0x38], rsi
- ;; size=39 bbWeight=0.50 PerfScore 3.25
-G_M18170_IG04:
- vmovsd xmm0, qword ptr [rbp-0x38]
- vmovsd qword ptr [rbp-0x28], xmm0
+ vmovsd xmm0, qword ptr [reloc @RWD00]
vxorps xmm1, xmm1, xmm1
- xor r14d, r14d
- ;; size=17 bbWeight=0.25 PerfScore 1.15
-G_M18170_IG05:
- movzx r13, byte ptr [r15+r14]
- add r13d, -48
- cmp r13d, 8
- jae SHORT G_M18170_IG10
+ xor eax, eax
+ align [0 bytes for IG04]
+ ;; size=14 bbWeight=0.25 PerfScore 0.90
+G_M18170_IG04:
+ movzx rcx, byte ptr [rdi+rax]
+ add ecx, -48
+ cmp ecx, 8
+ jae SHORT G_M18170_IG07
vmulsd xmm1, xmm1, xmm0
- vmovsd qword ptr [rbp-0x30], xmm1
- lea rsi, [rbp-0x40]
- mov edi, r13d
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[uint](uint,byref):ubyte
- call [rax]System.Double:TryConvertFrom[uint](uint,byref):ubyte
- test eax, eax
- jne SHORT G_M18170_IG07
- ;; size=47 bbWeight=4 PerfScore 51.00
-G_M18170_IG06:
- vxorps xmm0, xmm0, xmm0
- vcvtusi2sd xmm0, xmm0, r13d
- vmovsd qword ptr [rbp-0x40], xmm0
- ;; size=15 bbWeight=2 PerfScore 10.67
-G_M18170_IG07:
- vmovsd xmm0, qword ptr [rbp-0x40]
- vmovsd xmm1, qword ptr [rbp-0x30]
- vaddsd xmm1, xmm1, xmm0
- inc r14d
- cmp r14d, ebx
- vmovsd xmm0, qword ptr [rbp-0x28]
- jl SHORT G_M18170_IG05
- ;; size=27 bbWeight=4 PerfScore 54.00
-G_M18170_IG08:
+ vxorps xmm2, xmm2, xmm2
+ vcvtusi2sd xmm2, xmm2, ecx
+ vaddsd xmm1, xmm1, xmm2
+ inc eax
+ cmp eax, esi
+ jl SHORT G_M18170_IG04
+ ;; size=36 bbWeight=4 PerfScore 61.33
+G_M18170_IG05:
vmovaps xmm0, xmm1
;; size=4 bbWeight=0.50 PerfScore 0.12
-G_M18170_IG09:
- add rsp, 32
- pop rbx
- pop r13
- pop r14
- pop r15
+G_M18170_IG06:
pop rbp
ret
- ;; size=13 bbWeight=0.50 PerfScore 1.88
-G_M18170_IG10:
+ ;; size=2 bbWeight=0.50 PerfScore 0.75
+G_M18170_IG07:
mov rax, 0xD1FFAB1E ; code for System.Formats.Tar.TarHelpers:ThrowInvalidNumber()
call [rax]System.Formats.Tar.TarHelpers:ThrowInvalidNumber()
int3
;; size=13 bbWeight=0.50 PerfScore 1.75
-G_M18170_IG11:
+G_M18170_IG08:
vxorps xmm0, xmm0, xmm0
;; size=4 bbWeight=0.50 PerfScore 0.17
-G_M18170_IG12:
- add rsp, 32
- pop rbx
- pop r13
- pop r14
- pop r15
+G_M18170_IG09:
pop rbp
ret
- ;; size=13 bbWeight=0.50 PerfScore 1.88
+ ;; size=2 bbWeight=0.50 PerfScore 0.75
+RWD00 dq 4020000000000000h ; 8
-; Total bytes of code 266, prolog size 17, PerfScore 141.85, instruction count 76, allocated bytes for code 266 (MethodHash=2fc0b905) for method System.Formats.Tar.TarHelpers:ParseOctal[double](System.ReadOnlySpan`1[ubyte]):double (FullOpts)
+
+; Total bytes of code 117, prolog size 4, PerfScore 75.77, instruction count 36, allocated bytes for code 117 (MethodHash=2fc0b905) for method System.Formats.Tar.TarHelpers:ParseOctal[double](System.ReadOnlySpan`1[ubyte]):double (FullOpts)
; ============================================================-100 (-65.36 % of base) - System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]:.ctor(int):this ; Assembly listing for method System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]:.ctor(int):this (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rbp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 0 single block inlinees; 4 inlinees without PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 6 inlinees without PGO data
; Final local variable assignments
;
; V00 this [V00,T00] ( 4, 4 ) byref -> rbx this single-def
-; V01 arg1 [V01,T01] ( 4, 4 ) int -> r15 single-def
+; V01 arg1 [V01,T01] ( 4, 4 ) int -> rsi single-def
;* V02 loc0 [V02 ] ( 0, 0 ) double -> zero-ref ld-addr-op
;# V03 OutArgs [V03 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
;* V04 tmp1 [V04 ] ( 0, 0 ) byref -> zero-ref
;* V05 tmp2 [V05 ] ( 0, 0 ) byref -> zero-ref single-def
-; V06 tmp3 [V06,T02] ( 2, 4 ) double -> [rbp-0x18] spill-single-def "impAppendStmt"
+;* V06 tmp3 [V06 ] ( 0, 0 ) double -> zero-ref "impAppendStmt"
;* V07 tmp4 [V07 ] ( 0, 0 ) byref -> zero-ref single-def
-; V08 tmp5 [V08,T03] ( 2, 2 ) double -> mm0 single-def
-; V09 tmp6 [V09 ] ( 3, 2.50) double -> [rbp-0x20] do-not-enreg[X] must-init addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+; V08 tmp5 [V08,T02] ( 2, 2 ) double -> mm0 single-def
+;* V09 tmp6 [V09 ] ( 0, 0 ) double -> zero-ref ld-addr-op single-def "Inline ldloca(s) first use temp"
;* V10 tmp7 [V10 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-; V11 tmp8 [V11 ] ( 3, 2.50) double -> [rbp-0x28] do-not-enreg[X] must-init addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+;* V11 tmp8 [V11 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
;* V12 tmp9 [V12 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+; V13 tmp10 [V13,T03] ( 2, 2 ) double -> mm1 ld-addr-op single-def "Inline ldloca(s) first use temp"
+;* V14 tmp11 [V14 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V15 tmp12 [V15 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V16 tmp13 [V16 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
;
-; Lcl frame size = 32
+; Lcl frame size = 8
G_M63659_IG01:
push rbp
- push r15
push rbx
- sub rsp, 32
- lea rbp, [rsp+0x30]
- xor eax, eax
- mov qword ptr [rbp-0x20], rax
- mov qword ptr [rbp-0x28], rax
+ push rax
+ lea rbp, [rsp+0x10]
mov rbx, rdi
- mov r15d, esi
- ;; size=29 bbWeight=1 PerfScore 6.50
+ ;; size=11 bbWeight=1 PerfScore 3.75
G_M63659_IG02:
- mov dword ptr [rbx], r15d
- lea rsi, [rbp-0x20]
- mov edi, 2
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Double:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M63659_IG05
- vmovsd xmm0, qword ptr [rbp-0x20]
- vmovsd qword ptr [rbp-0x18], xmm0
- lea rsi, [rbp-0x28]
- mov edi, r15d
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Double:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M63659_IG04
- vmovsd xmm0, qword ptr [rbp-0x18]
- vmovsd xmm1, qword ptr [rbp-0x28]
+ mov dword ptr [rbx], esi
+ vxorps xmm1, xmm1, xmm1
+ vcvtsi2sd xmm1, xmm1, esi
+ vmovsd xmm0, qword ptr [reloc @RWD00]
mov rax, 0xD1FFAB1E ; code for System.Double:Pow(double,double):double
call [rax]System.Double:Pow(double,double):double
vmovsd qword ptr [rbx+0x08], xmm0
- ;; size=88 bbWeight=1 PerfScore 26.75
+ ;; size=35 bbWeight=1 PerfScore 15.58
G_M63659_IG03:
- add rsp, 32
+ add rsp, 8
pop rbx
- pop r15
pop rbp
ret
- ;; size=9 bbWeight=1 PerfScore 2.75
-G_M63659_IG04:
- xor eax, eax
- mov qword ptr [rbp-0x28], rax
- jmp SHORT G_M63659_IG06
- ;; size=8 bbWeight=0.50 PerfScore 1.62
-G_M63659_IG05:
- xor eax, eax
- mov qword ptr [rbp-0x20], rax
- ;; size=6 bbWeight=0.50 PerfScore 0.62
-G_M63659_IG06:
- mov rax, 0xD1FFAB1E ; code for System.ThrowHelper:ThrowNotSupportedException()
- call [rax]System.ThrowHelper:ThrowNotSupportedException()
- int3
- ;; size=13 bbWeight=0.50 PerfScore 1.75
+ ;; size=7 bbWeight=1 PerfScore 2.25
+RWD00 dq 4000000000000000h ; 2
-; Total bytes of code 153, prolog size 23, PerfScore 40.00, instruction count 43, allocated bytes for code 153 (MethodHash=7d060754) for method System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]:.ctor(int):this (FullOpts)
+
+; Total bytes of code 53, prolog size 8, PerfScore 21.58, instruction count 16, allocated bytes for code 53 (MethodHash=7d060754) for method System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]:.ctor(int):this (FullOpts)
; ============================================================-93 (-34.57 % of base) - System.Security.Cryptography.PemEncoding:IsValidLabel[double](System.ReadOnlySpan`1[double]):ubyte ; Assembly listing for method System.Security.Cryptography.PemEncoding:IsValidLabel[double](System.ReadOnlySpan`1[double]):ubyte (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
; rbp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 3 single block inlinees; 4 inlinees without PGO data
+; 0 inlinees with PGO data; 3 single block inlinees; 6 inlinees without PGO data
; Final local variable assignments
;
;* V00 arg0 [V00 ] ( 0, 0 ) struct (16) zero-ref multireg-arg ld-addr-op single-def <System.ReadOnlySpan`1[double]>
; V01 loc0 [V01,T03] ( 5, 6.75) ubyte -> r14
;* V02 loc1 [V02,T05] ( 0, 0 ) int -> zero-ref
; V03 loc2 [V03,T06] ( 4, 12 ) double -> [rbp-0x28] spill-single-def
;# V04 OutArgs [V04 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
; V05 tmp1 [V05,T04] ( 2, 4 ) int -> rax
-; V06 tmp2 [V06 ] ( 3, 6 ) double -> [rbp-0x30] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+;* V06 tmp2 [V06 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V07 tmp3 [V07 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-;* V08 tmp4 [V08 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
-; V09 tmp5 [V09,T07] ( 2, 8 ) double -> mm0 "Inlining Arg"
-; V10 tmp6 [V10 ] ( 3, 6 ) double -> [rbp-0x38] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
-;* V11 tmp7 [V11 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-;* V12 tmp8 [V12 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
-; V13 tmp9 [V13,T08] ( 2, 8 ) double -> mm0 "Inlining Arg"
-; V14 tmp10 [V14,T02] ( 2, 5 ) byref -> r15 single-def "field V00._reference (fldOffset=0x0)" P-INDEP
-; V15 tmp11 [V15,T01] ( 3, 6 ) int -> rbx single-def "field V00._length (fldOffset=0x8)" P-INDEP
-; V16 rat0 [V16,T00] ( 5, 16.25) long -> r13 "Widened IV V02"
+;* V08 tmp4 [V08 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V09 tmp5 [V09 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
+;* V10 tmp6 [V10 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+;* V11 tmp7 [V11 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
+;* V12 tmp8 [V12 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V13 tmp9 [V13 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V14 tmp10 [V14 ] ( 0, 0 ) double -> zero-ref "Inline stloc first use temp"
+;* V15 tmp11 [V15 ] ( 0, 0 ) double -> zero-ref "Inlining Arg"
+; V16 tmp12 [V16,T02] ( 2, 5 ) byref -> r15 single-def "field V00._reference (fldOffset=0x0)" P-INDEP
+; V17 tmp13 [V17,T01] ( 3, 6 ) int -> rbx single-def "field V00._length (fldOffset=0x8)" P-INDEP
+; V18 rat0 [V18,T00] ( 5, 16.25) long -> r13 "Widened IV V02"
;
-; Lcl frame size = 32
+; Lcl frame size = 16
G_M42638_IG01:
push rbp
push r15
push r14
push r13
push rbx
- sub rsp, 32
- lea rbp, [rsp+0x40]
+ sub rsp, 16
+ lea rbp, [rsp+0x30]
mov r15, rdi
mov ebx, esi
;; size=22 bbWeight=1 PerfScore 6.25
G_M42638_IG02:
test ebx, ebx
- je G_M42638_IG17
+ je G_M42638_IG15
;; size=8 bbWeight=1 PerfScore 1.25
G_M42638_IG03:
xor r14d, r14d
xor r13d, r13d
;; size=6 bbWeight=0.25 PerfScore 0.12
G_M42638_IG04:
vmovsd xmm0, qword ptr [r15+8*r13]
vmovsd qword ptr [rbp-0x28], xmm0
mov rax, 0xD1FFAB1E ; code for System.Security.Cryptography.PemEncoding:<IsValidLabel>g__IsLabelChar|7_0[double](double):ubyte
call [rax]System.Security.Cryptography.PemEncoding:<IsValidLabel>g__IsLabelChar|7_0[double](double):ubyte
test eax, eax
- jne G_M42638_IG11
- ;; size=31 bbWeight=4 PerfScore 38.00
+ jne SHORT G_M42638_IG09
+ ;; size=27 bbWeight=4 PerfScore 38.00
G_M42638_IG05:
- lea rsi, [rbp-0x30]
- mov edi, 32
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[ushort](ushort,byref):ubyte
- call [rax]System.Double:TryConvertFrom[ushort](ushort,byref):ubyte
- test eax, eax
- jne SHORT G_M42638_IG06
- mov rsi, 0xD1FFAB1E
- mov qword ptr [rbp-0x30], rsi
- ;; size=39 bbWeight=2 PerfScore 13.00
+ vmovsd xmm0, qword ptr [rbp-0x28]
+ vucomisd xmm0, qword ptr [reloc @RWD00]
+ jp SHORT G_M42638_IG06
+ je SHORT G_M42638_IG08
+ ;; size=17 bbWeight=2 PerfScore 18.00
G_M42638_IG06:
- vmovsd xmm0, qword ptr [rbp-0x30]
- vmovsd xmm1, qword ptr [rbp-0x28]
- vucomisd xmm1, xmm0
- jp SHORT G_M42638_IG07
- je SHORT G_M42638_IG10
- ;; size=18 bbWeight=2 PerfScore 20.00
-G_M42638_IG07:
- lea rsi, [rbp-0x38]
- mov edi, 45
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[ushort](ushort,byref):ubyte
- call [rax]System.Double:TryConvertFrom[ushort](ushort,byref):ubyte
- test eax, eax
- jne SHORT G_M42638_IG08
- mov rax, 0xD1FFAB1E
- mov qword ptr [rbp-0x38], rax
- ;; size=39 bbWeight=2 PerfScore 13.00
-G_M42638_IG08:
- vmovsd xmm0, qword ptr [rbp-0x38]
- vmovsd xmm1, qword ptr [rbp-0x28]
- vucomisd xmm1, xmm0
+ vucomisd xmm0, qword ptr [reloc @RWD08]
setnp al
- jp SHORT G_M42638_IG09
+ jp SHORT G_M42638_IG07
sete al
- ;; size=22 bbWeight=2 PerfScore 22.00
-G_M42638_IG09:
+ ;; size=16 bbWeight=2 PerfScore 14.00
+G_M42638_IG07:
movzx rax, al
test eax, eax
- je SHORT G_M42638_IG15
+ je SHORT G_M42638_IG13
;; size=7 bbWeight=2 PerfScore 3.00
-G_M42638_IG10:
+G_M42638_IG08:
test r14d, r14d
- je SHORT G_M42638_IG15
+ je SHORT G_M42638_IG13
xor r14d, r14d
- jmp SHORT G_M42638_IG12
+ jmp SHORT G_M42638_IG10
;; size=10 bbWeight=2 PerfScore 7.00
-G_M42638_IG11:
+G_M42638_IG09:
mov r14d, 1
;; size=6 bbWeight=2 PerfScore 0.50
-G_M42638_IG12:
+G_M42638_IG10:
inc r13d
cmp r13d, ebx
- jl G_M42638_IG04
- ;; size=12 bbWeight=4 PerfScore 6.00
-G_M42638_IG13:
+ jl SHORT G_M42638_IG04
+ ;; size=8 bbWeight=4 PerfScore 6.00
+G_M42638_IG11:
mov eax, r14d
;; size=3 bbWeight=0.50 PerfScore 0.12
+G_M42638_IG12:
+ add rsp, 16
+ pop rbx
+ pop r13
+ pop r14
+ pop r15
+ pop rbp
+ ret
+ ;; size=13 bbWeight=0.50 PerfScore 1.88
+G_M42638_IG13:
+ xor eax, eax
+ ;; size=2 bbWeight=0.50 PerfScore 0.12
G_M42638_IG14:
- add rsp, 32
+ add rsp, 16
pop rbx
pop r13
pop r14
pop r15
pop rbp
ret
;; size=13 bbWeight=0.50 PerfScore 1.88
G_M42638_IG15:
- xor eax, eax
- ;; size=2 bbWeight=0.50 PerfScore 0.12
-G_M42638_IG16:
- add rsp, 32
- pop rbx
- pop r13
- pop r14
- pop r15
- pop rbp
- ret
- ;; size=13 bbWeight=0.50 PerfScore 1.88
-G_M42638_IG17:
mov eax, 1
;; size=5 bbWeight=0.50 PerfScore 0.12
-G_M42638_IG18:
- add rsp, 32
+G_M42638_IG16:
+ add rsp, 16
pop rbx
pop r13
pop r14
pop r15
pop rbp
ret
;; size=13 bbWeight=0.50 PerfScore 1.88
+RWD00 dq 4040000000000000h ; 32
+RWD08 dq 4046800000000000h ; 45
-; Total bytes of code 269, prolog size 17, PerfScore 136.12, instruction count 81, allocated bytes for code 269 (MethodHash=79345971) for method System.Security.Cryptography.PemEncoding:IsValidLabel[double](System.ReadOnlySpan`1[double]):ubyte (FullOpts)
+
+; Total bytes of code 176, prolog size 17, PerfScore 100.12, instruction count 62, allocated bytes for code 176 (MethodHash=79345971) for method System.Security.Cryptography.PemEncoding:IsValidLabel[double](System.ReadOnlySpan`1[double]):ubyte (FullOpts)
; ============================================================-87 (-42.65 % of base) - System.Numerics.Tensors.TensorPrimitives:ScaleB[double](System.ReadOnlySpan`1[double],int,System.Span`1[double]) ; Assembly listing for method System.Numerics.Tensors.TensorPrimitives:ScaleB[double](System.ReadOnlySpan`1[double],int,System.Span`1[double]) (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
-; rbp based frame
-; partially interruptible
+; rsp based frame
+; fully interruptible
; No PGO data
-; 0 inlinees with PGO data; 0 single block inlinees; 5 inlinees without PGO data
+; 0 inlinees with PGO data; 0 single block inlinees; 7 inlinees without PGO data
; Final local variable assignments
;
;* V00 arg0 [V00 ] ( 0, 0 ) struct (16) zero-ref multireg-arg single-def <System.ReadOnlySpan`1[double]>
-; V01 arg1 [V01,T00] ( 4, 4 ) int -> rbx single-def
+; V01 arg1 [V01,T00] ( 4, 4 ) int -> rdx single-def
;* V02 arg2 [V02 ] ( 0, 0 ) struct (16) zero-ref multireg-arg single-def <System.Span`1[double]>
;# V03 OutArgs [V03 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
-; V04 tmp1 [V04 ] ( 5, 10 ) struct (16) [rbp-0x38] do-not-enreg[SFA] multireg-arg ld-addr-op "NewObj constructor temp" <System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]>
+; V04 tmp1 [V04 ] ( 5, 10 ) struct (16) [rsp+0x08] do-not-enreg[SFA] multireg-arg ld-addr-op "NewObj constructor temp" <System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]>
;* V05 tmp2 [V05 ] ( 0, 0 ) byref -> zero-ref
;* V06 tmp3 [V06 ] ( 0, 0 ) byref -> zero-ref
-; V07 tmp4 [V07,T07] ( 2, 4 ) double -> [rbp-0x40] spill-single-def "impAppendStmt"
+;* V07 tmp4 [V07 ] ( 0, 0 ) double -> zero-ref "impAppendStmt"
;* V08 tmp5 [V08 ] ( 0, 0 ) byref -> zero-ref
;* V09 tmp6 [V09 ] ( 0, 0 ) double -> zero-ref
-; V10 tmp7 [V10 ] ( 3, 2.50) double -> [rbp-0x48] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+;* V10 tmp7 [V10 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V11 tmp8 [V11 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-; V12 tmp9 [V12 ] ( 3, 2.50) double -> [rbp-0x50] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+;* V12 tmp9 [V12 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
;* V13 tmp10 [V13 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
-; V14 tmp11 [V14,T02] ( 2, 2 ) byref -> r15 single-def "field V00._reference (fldOffset=0x0)" P-INDEP
-; V15 tmp12 [V15,T04] ( 2, 2 ) int -> r13 single-def "field V00._length (fldOffset=0x8)" P-INDEP
-; V16 tmp13 [V16,T03] ( 2, 2 ) byref -> r14 single-def "field V02._reference (fldOffset=0x0)" P-INDEP
-; V17 tmp14 [V17,T05] ( 2, 2 ) int -> r12 single-def "field V02._length (fldOffset=0x8)" P-INDEP
-; V18 tmp15 [V18,T01] ( 4, 6 ) int -> [rbp-0x38] do-not-enreg[] "field V04._n (fldOffset=0x0)" P-DEP
-; V19 tmp16 [V19,T06] ( 3, 5 ) double -> [rbp-0x30] do-not-enreg[] "field V04._pow2n (fldOffset=0x8)" P-DEP
+; V14 tmp11 [V14,T07] ( 2, 2 ) double -> mm1 ld-addr-op "Inline ldloca(s) first use temp"
+;* V15 tmp12 [V15 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V16 tmp13 [V16 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V17 tmp14 [V17 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+; V18 tmp15 [V18,T02] ( 2, 2 ) byref -> rbx single-def "field V00._reference (fldOffset=0x0)" P-INDEP
+; V19 tmp16 [V19,T04] ( 2, 2 ) int -> r14 single-def "field V00._length (fldOffset=0x8)" P-INDEP
+; V20 tmp17 [V20,T03] ( 2, 2 ) byref -> r15 single-def "field V02._reference (fldOffset=0x0)" P-INDEP
+; V21 tmp18 [V21,T05] ( 2, 2 ) int -> r13 single-def "field V02._length (fldOffset=0x8)" P-INDEP
+; V22 tmp19 [V22,T01] ( 4, 6 ) int -> [rsp+0x08] do-not-enreg[] "field V04._n (fldOffset=0x0)" P-DEP
+; V23 tmp20 [V23,T06] ( 3, 5 ) double -> [rsp+0x10] do-not-enreg[] "field V04._pow2n (fldOffset=0x8)" P-DEP
;
-; Lcl frame size = 40
+; Lcl frame size = 24
G_M43791_IG01:
- push rbp
push r15
push r14
push r13
- push r12
push rbx
- sub rsp, 40
- lea rbp, [rsp+0x50]
- mov r15, rdi
- mov r13d, esi
- mov ebx, edx
- mov r14, rcx
- mov r12d, r8d
- ;; size=33 bbWeight=1 PerfScore 8.00
+ sub rsp, 24
+ mov rbx, rdi
+ mov r14d, esi
+ mov r15, rcx
+ mov r13d, r8d
+ ;; size=23 bbWeight=1 PerfScore 5.25
G_M43791_IG02:
vxorps xmm0, xmm0, xmm0
- vmovups xmmword ptr [rbp-0x38], xmm0
- mov dword ptr [rbp-0x38], ebx
- lea rsi, [rbp-0x48]
- mov edi, 2
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Double:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M43791_IG05
- vmovsd xmm0, qword ptr [rbp-0x48]
- vmovsd qword ptr [rbp-0x40], xmm0
- lea rsi, [rbp-0x50]
- mov edi, ebx
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Double:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M43791_IG04
- vmovsd xmm0, qword ptr [rbp-0x40]
- vmovsd xmm1, qword ptr [rbp-0x50]
+ vmovups xmmword ptr [rsp+0x08], xmm0
+ mov dword ptr [rsp+0x08], edx
+ vxorps xmm1, xmm1, xmm1
+ vcvtsi2sd xmm1, xmm1, edx
+ vmovsd xmm0, qword ptr [reloc @RWD00]
mov rax, 0xD1FFAB1E ; code for System.Double:Pow(double,double):double
call [rax]System.Double:Pow(double,double):double
- vmovsd qword ptr [rbp-0x30], xmm0
- mov rdi, r15
- mov esi, r13d
- mov edx, dword ptr [rbp-0x38]
- vmovsd xmm0, qword ptr [rbp-0x30]
- mov rcx, r14
- mov r8d, r12d
+ vmovsd qword ptr [rsp+0x10], xmm0
+ mov rdi, rbx
+ mov esi, r14d
+ mov edx, dword ptr [rsp+0x08]
+ vmovsd xmm0, qword ptr [rsp+0x10]
+ mov rcx, r15
+ mov r8d, r13d
mov rax, 0xD1FFAB1E ; code for System.Numerics.Tensors.TensorPrimitives:InvokeSpanIntoSpan[double,System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]](System.ReadOnlySpan`1[double],System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double],System.Span`1[double])
- call [rax]System.Numerics.Tensors.TensorPrimitives:InvokeSpanIntoSpan[double,System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]](System.ReadOnlySpan`1[double],System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double],System.Span`1[double])
- nop
- ;; size=129 bbWeight=1 PerfScore 35.58
+ ;; size=80 bbWeight=1 PerfScore 21.17
G_M43791_IG03:
- add rsp, 40
+ add rsp, 24
pop rbx
- pop r12
pop r13
pop r14
pop r15
- pop rbp
- ret
- ;; size=15 bbWeight=1 PerfScore 4.25
-G_M43791_IG04:
- xor eax, eax
- mov qword ptr [rbp-0x50], rax
- jmp SHORT G_M43791_IG06
- ;; size=8 bbWeight=0.50 PerfScore 1.62
-G_M43791_IG05:
- xor eax, eax
- mov qword ptr [rbp-0x48], rax
- ;; size=6 bbWeight=0.50 PerfScore 0.62
-G_M43791_IG06:
- mov rax, 0xD1FFAB1E ; code for System.ThrowHelper:ThrowNotSupportedException()
- call [rax]System.ThrowHelper:ThrowNotSupportedException()
- int3
- ;; size=13 bbWeight=0.50 PerfScore 1.75
+ tail.jmp [rax]System.Numerics.Tensors.TensorPrimitives:InvokeSpanIntoSpan[double,System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double]](System.ReadOnlySpan`1[double],System.Numerics.Tensors.TensorPrimitives+ScaleBOperator`1[double],System.Span`1[double])
+ ;; size=14 bbWeight=1 PerfScore 4.25
+RWD00 dq 4000000000000000h ; 2
-; Total bytes of code 204, prolog size 19, PerfScore 51.83, instruction count 60, allocated bytes for code 204 (MethodHash=3e7754f0) for method System.Numerics.Tensors.TensorPrimitives:ScaleB[double](System.ReadOnlySpan`1[double],int,System.Span`1[double]) (FullOpts)
+
+; Total bytes of code 117, prolog size 23, PerfScore 30.67, instruction count 31, allocated bytes for code 117 (MethodHash=3e7754f0) for method System.Numerics.Tensors.TensorPrimitives:ScaleB[double](System.ReadOnlySpan`1[double],int,System.Span`1[double]) (FullOpts)
; ============================================================-83 (-65.35 % of base) - System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] ; Assembly listing for method System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
-; rbp based frame
+; rsp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 2 single block inlinees; 3 inlinees without PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 4 inlinees without PGO data
; Final local variable assignments
;
-; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rbx single-def
-; V01 arg0 [V01,T02] ( 1, 1 ) simd64 -> [rbp+0x10] single-def <System.Runtime.Intrinsics.Vector512`1[double]>
+; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rdi single-def
+; V01 arg0 [V01,T02] ( 1, 1 ) simd64 -> [rsp+0x08] single-def <System.Runtime.Intrinsics.Vector512`1[double]>
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
;* V03 tmp1 [V03 ] ( 0, 0 ) simd64 -> zero-ref "spilled call-like call argument"
;* V04 tmp2 [V04 ] ( 0, 0 ) simd64 -> zero-ref "Inline return value spill temp" <System.Runtime.Intrinsics.Vector512`1[double]>
-; V05 tmp3 [V05,T01] ( 2, 4 ) simd64 -> [rbp-0x70] spill-single-def "impAppendStmt"
-; V06 tmp4 [V06 ] ( 3, 2.50) double -> [rbp-0x78] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+; V05 tmp3 [V05,T01] ( 2, 4 ) simd64 -> mm0 "impAppendStmt"
+;* V06 tmp4 [V06 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V07 tmp5 [V07 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V08 tmp6 [V08 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V09 tmp7 [V09 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
;
-; Lcl frame size = 120
+; Lcl frame size = 0
G_M49678_IG01:
- push rbp
- push rbx
- sub rsp, 120
- lea rbp, [rsp+0x80]
- mov rbx, rdi
- ;; size=17 bbWeight=1 PerfScore 3.00
+ ;; size=0 bbWeight=1 PerfScore 0.00
G_M49678_IG02:
- vmovups zmm0, zmmword ptr [rbp+0x10]
+ vmovups zmm0, zmmword ptr [rsp+0x08]
vmulpd zmm0, zmm0, qword ptr [reloc @RWD00] {1to8}
- vmovups zmmword ptr [rbp-0x70], zmm0
- lea rsi, [rbp-0x78]
- mov edi, 180
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Double:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M49678_IG04
- vmovups zmm0, zmmword ptr [rbp-0x70]
- vdivpd zmm0, zmm0, qword ptr [rbp-0x78] {1to8}
- vmovups zmmword ptr [rbx], zmm0
- mov rax, rbx
- ;; size=81 bbWeight=1 PerfScore 33.50
+ vdivpd zmm0, zmm0, qword ptr [reloc @RWD08] {1to8}
+ vmovups zmmword ptr [rdi], zmm0
+ mov rax, rdi
+ ;; size=40 bbWeight=1 PerfScore 24.25
G_M49678_IG03:
vzeroupper
- add rsp, 120
- pop rbx
- pop rbp
ret
- ;; size=10 bbWeight=1 PerfScore 3.25
-G_M49678_IG04:
- xor eax, eax
- mov qword ptr [rbp-0x78], rax
- mov rax, 0xD1FFAB1E ; code for System.ThrowHelper:ThrowNotSupportedException()
- call [rax]System.ThrowHelper:ThrowNotSupportedException()
- int3
- ;; size=19 bbWeight=0.50 PerfScore 2.38
+ ;; size=4 bbWeight=1 PerfScore 2.00
RWD00 dq 400921FB54442D18h ; 3.14159265
+RWD08 dq 4066800000000000h ; 180
-; Total bytes of code 127, prolog size 14, PerfScore 42.12, instruction count 28, allocated bytes for code 130 (MethodHash=35cb3df1) for method System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] (FullOpts)
+; Total bytes of code 44, prolog size 0, PerfScore 26.25, instruction count 7, allocated bytes for code 44 (MethodHash=35cb3df1) for method System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] (FullOpts)
; ============================================================-83 (-65.35 % of base) - System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] ; Assembly listing for method System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
-; rbp based frame
+; rsp 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; 2 single block inlinees; 3 inlinees without PGO data
; Final local variable assignments
;
-; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rbx single-def
-; V01 arg0 [V01,T02] ( 1, 1 ) simd64 -> [rbp+0x10] single-def <System.Runtime.Intrinsics.Vector512`1[double]>
+; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rdi single-def
+; V01 arg0 [V01,T02] ( 1, 1 ) simd64 -> [rsp+0x08] single-def <System.Runtime.Intrinsics.Vector512`1[double]>
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
-; V03 tmp1 [V03,T01] ( 2, 4 ) simd64 -> [rbp-0x70] spill-single-def "impAppendStmt"
-; V04 tmp2 [V04 ] ( 3, 2.50) double -> [rbp-0x78] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+; V03 tmp1 [V03,T01] ( 2, 4 ) simd64 -> mm0 "impAppendStmt"
+;* V04 tmp2 [V04 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V05 tmp3 [V05 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V06 tmp4 [V06 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V07 tmp5 [V07 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
;
-; Lcl frame size = 120
+; Lcl frame size = 0
G_M17681_IG01:
- push rbp
- push rbx
- sub rsp, 120
- lea rbp, [rsp+0x80]
- mov rbx, rdi
- ;; size=17 bbWeight=1 PerfScore 3.00
+ ;; size=0 bbWeight=1 PerfScore 0.00
G_M17681_IG02:
- vmovups zmm0, zmmword ptr [rbp+0x10]
+ vmovups zmm0, zmmword ptr [rsp+0x08]
vmulpd zmm0, zmm0, qword ptr [reloc @RWD00] {1to8}
- vmovups zmmword ptr [rbp-0x70], zmm0
- lea rsi, [rbp-0x78]
- mov edi, 180
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Double:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M17681_IG04
- vmovups zmm0, zmmword ptr [rbp-0x70]
- vdivpd zmm0, zmm0, qword ptr [rbp-0x78] {1to8}
- vmovups zmmword ptr [rbx], zmm0
- mov rax, rbx
- ;; size=81 bbWeight=1 PerfScore 33.50
+ vdivpd zmm0, zmm0, qword ptr [reloc @RWD08] {1to8}
+ vmovups zmmword ptr [rdi], zmm0
+ mov rax, rdi
+ ;; size=40 bbWeight=1 PerfScore 24.25
G_M17681_IG03:
vzeroupper
- add rsp, 120
- pop rbx
- pop rbp
ret
- ;; size=10 bbWeight=1 PerfScore 3.25
-G_M17681_IG04:
- xor eax, eax
- mov qword ptr [rbp-0x78], rax
- mov rax, 0xD1FFAB1E ; code for System.ThrowHelper:ThrowNotSupportedException()
- call [rax]System.ThrowHelper:ThrowNotSupportedException()
- int3
- ;; size=19 bbWeight=0.50 PerfScore 2.38
+ ;; size=4 bbWeight=1 PerfScore 2.00
RWD00 dq 400921FB54442D18h ; 3.14159265
+RWD08 dq 4066800000000000h ; 180
-; Total bytes of code 127, prolog size 14, PerfScore 42.12, instruction count 28, allocated bytes for code 130 (MethodHash=32a3baee) for method System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] (FullOpts)
+; Total bytes of code 44, prolog size 0, PerfScore 26.25, instruction count 7, allocated bytes for code 44 (MethodHash=32a3baee) for method System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[double]):System.Runtime.Intrinsics.Vector512`1[double] (FullOpts)
; ============================================================-82 (-65.08 % of base) - System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[float]):System.Runtime.Intrinsics.Vector512`1[float] ; Assembly listing for method System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[float]):System.Runtime.Intrinsics.Vector512`1[float] (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
-; rbp based frame
+; rsp 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; 2 single block inlinees; 3 inlinees without PGO data
; Final local variable assignments
;
-; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rbx single-def
-; V01 arg0 [V01,T02] ( 1, 1 ) simd64 -> [rbp+0x10] single-def <System.Runtime.Intrinsics.Vector512`1[float]>
+; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rdi single-def
+; V01 arg0 [V01,T02] ( 1, 1 ) simd64 -> [rsp+0x08] single-def <System.Runtime.Intrinsics.Vector512`1[float]>
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
-; V03 tmp1 [V03,T01] ( 2, 4 ) simd64 -> [rbp-0x70] spill-single-def "impAppendStmt"
-; V04 tmp2 [V04 ] ( 3, 2.50) float -> [rbp-0x74] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+; V03 tmp1 [V03,T01] ( 2, 4 ) simd64 -> mm0 "impAppendStmt"
+;* V04 tmp2 [V04 ] ( 0, 0 ) float -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V05 tmp3 [V05 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V06 tmp4 [V06 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V07 tmp5 [V07 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
;
-; Lcl frame size = 120
+; Lcl frame size = 0
G_M24241_IG01:
- push rbp
- push rbx
- sub rsp, 120
- lea rbp, [rsp+0x80]
- mov rbx, rdi
- ;; size=17 bbWeight=1 PerfScore 3.00
+ ;; size=0 bbWeight=1 PerfScore 0.00
G_M24241_IG02:
- vmovups zmm0, zmmword ptr [rbp+0x10]
+ vmovups zmm0, zmmword ptr [rsp+0x08]
vmulps zmm0, zmm0, dword ptr [reloc @RWD00] {1to16}
- vmovups zmmword ptr [rbp-0x70], zmm0
- lea rsi, [rbp-0x74]
- mov edi, 180
- mov rax, 0xD1FFAB1E ; code for System.Single:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Single:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M24241_IG04
- vmovups zmm0, zmmword ptr [rbp-0x70]
- vdivps zmm0, zmm0, dword ptr [rbp-0x74] {1to16}
- vmovups zmmword ptr [rbx], zmm0
- mov rax, rbx
- ;; size=81 bbWeight=1 PerfScore 31.50
+ vdivps zmm0, zmm0, dword ptr [reloc @RWD04] {1to16}
+ vmovups zmmword ptr [rdi], zmm0
+ mov rax, rdi
+ ;; size=40 bbWeight=1 PerfScore 22.25
G_M24241_IG03:
vzeroupper
- add rsp, 120
- pop rbx
- pop rbp
ret
- ;; size=10 bbWeight=1 PerfScore 3.25
-G_M24241_IG04:
- xor eax, eax
- mov dword ptr [rbp-0x74], eax
- mov rax, 0xD1FFAB1E ; code for System.ThrowHelper:ThrowNotSupportedException()
- call [rax]System.ThrowHelper:ThrowNotSupportedException()
- int3
- ;; size=18 bbWeight=0.50 PerfScore 2.38
+ ;; size=4 bbWeight=1 PerfScore 2.00
RWD00 dd 40490FDBh ; 3.14159
+RWD04 dd 43340000h ; 180
-; Total bytes of code 126, prolog size 14, PerfScore 40.12, instruction count 28, allocated bytes for code 129 (MethodHash=b42fa14e) for method System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[float]):System.Runtime.Intrinsics.Vector512`1[float] (FullOpts)
+; Total bytes of code 44, prolog size 0, PerfScore 24.25, instruction count 7, allocated bytes for code 44 (MethodHash=b42fa14e) for method System.Runtime.Intrinsics.Vector512:DegreesToRadians(System.Runtime.Intrinsics.Vector512`1[float]):System.Runtime.Intrinsics.Vector512`1[float] (FullOpts)
; ============================================================-70 (-67.31 % of base) - System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector128`1[double]):System.Runtime.Intrinsics.Vector128`1[double] ; Assembly listing for method System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector128`1[double]):System.Runtime.Intrinsics.Vector128`1[double] (FullOpts)
; Emitting BLENDED_CODE for X64 with AVX512 - Unix
; FullOpts code
; optimized code
-; rbp based frame
+; rsp based frame
; partially interruptible
; No PGO data
-; 0 inlinees with PGO data; 2 single block inlinees; 3 inlinees without PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 4 inlinees without PGO data
; Final local variable assignments
;
-; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rbx single-def
-; V01 arg0 [V01,T02] ( 1, 1 ) simd16 -> [rbp+0x10] single-def <System.Runtime.Intrinsics.Vector128`1[double]>
+; V00 RetBuf [V00,T00] ( 4, 4 ) byref -> rdi single-def
+; V01 arg0 [V01,T02] ( 1, 1 ) simd16 -> [rsp+0x08] single-def <System.Runtime.Intrinsics.Vector128`1[double]>
;# V02 OutArgs [V02 ] ( 1, 1 ) struct ( 0) [rsp+0x00] do-not-enreg[XS] addr-exposed "OutgoingArgSpace"
;* V03 tmp1 [V03 ] ( 0, 0 ) simd16 -> zero-ref "spilled call-like call argument"
;* V04 tmp2 [V04 ] ( 0, 0 ) simd16 -> zero-ref "Inline return value spill temp" <System.Runtime.Intrinsics.Vector128`1[double]>
-; V05 tmp3 [V05,T01] ( 2, 4 ) simd16 -> [rbp-0x20] spill-single-def "impAppendStmt"
-; V06 tmp4 [V06 ] ( 3, 2.50) double -> [rbp-0x28] do-not-enreg[X] addr-exposed ld-addr-op "Inline ldloca(s) first use temp"
+; V05 tmp3 [V05,T01] ( 2, 4 ) simd16 -> mm0 "impAppendStmt"
+;* V06 tmp4 [V06 ] ( 0, 0 ) double -> zero-ref ld-addr-op "Inline ldloca(s) first use temp"
;* V07 tmp5 [V07 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
+;* V08 tmp6 [V08 ] ( 0, 0 ) int -> zero-ref "Inline stloc first use temp"
+;* V09 tmp7 [V09 ] ( 0, 0 ) ubyte -> zero-ref "Inline return value spill temp"
;
-; Lcl frame size = 40
+; Lcl frame size = 0
G_M11950_IG01:
- push rbp
- push rbx
- sub rsp, 40
- lea rbp, [rsp+0x30]
- mov rbx, rdi
- ;; size=14 bbWeight=1 PerfScore 3.00
+ ;; size=0 bbWeight=1 PerfScore 0.00
G_M11950_IG02:
- vmovaps xmm0, xmmword ptr [rbp+0x10]
+ vmovaps xmm0, xmmword ptr [rsp+0x08]
vmulpd xmm0, xmm0, qword ptr [reloc @RWD00] {1to2}
- vmovaps xmmword ptr [rbp-0x20], xmm0
- lea rsi, [rbp-0x28]
- mov edi, 180
- mov rax, 0xD1FFAB1E ; code for System.Double:TryConvertFrom[int](int,byref):ubyte
- call [rax]System.Double:TryConvertFrom[int](int,byref):ubyte
- test eax, eax
- je SHORT G_M11950_IG04
- vmovaps xmm0, xmmword ptr [rbp-0x20]
- vdivpd xmm0, xmm0, qword ptr [rbp-0x28] {1to2}
- vmovups xmmword ptr [rbx], xmm0
- mov rax, rbx
- ;; size=64 bbWeight=1 PerfScore 33.50
+ vdivpd xmm0, xmm0, qword ptr [reloc @RWD08] {1to2}
+ vmovups xmmword ptr [rdi], xmm0
+ mov rax, rdi
+ ;; size=33 bbWeight=1 PerfScore 24.25
G_M11950_IG03:
- add rsp, 40
- pop rbx
- pop rbp
ret
- ;; size=7 bbWeight=1 PerfScore 2.25
-G_M11950_IG04:
- xor eax, eax
- mov qword ptr [rbp-0x28], rax
- mov rax, 0xD1FFAB1E ; code for System.ThrowHelper:ThrowNotSupportedException()
- call [rax]System.ThrowHelper:ThrowNotSupportedException()
- int3
- ;; size=19 bbWeight=0.50 PerfScore 2.38
+ ;; size=1 bbWeight=1 PerfScore 1.00
RWD00 dq 400921FB54442D18h ; 3.14159265
+RWD08 dq 4066800000000000h ; 180
-; Total bytes of code 104, prolog size 11, PerfScore 41.12, instruction count 27, allocated bytes for code 107 (MethodHash=39d7d151) for method System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector128`1[double]):System.Runtime.Intrinsics.Vector128`1[double] (FullOpts)
+; Total bytes of code 34, prolog size 0, PerfScore 25.25, instruction count 6, allocated bytes for code 34 (MethodHash=39d7d151) for method System.Numerics.Tensors.TensorPrimitives+DegreesToRadiansOperator`1[double]:Invoke(System.Runtime.Intrinsics.Vector128`1[double]):System.Runtime.Intrinsics.Vector128`1[double] (FullOpts)
; ============================================================Note: some changes were skipped as they were too large to fit into a comment. Larger list of diffs: https://gist.github.com/MihuBot/4504d8744eb8302da53125904ac4ac5a |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Job completed in 15 minutes 36 seconds (remote runner delay: 1 minute 21 seconds).
dotnet/runtime#112061
Diffs
Diffs
Artifacts:
The text was updated successfully, but these errors were encountered: