[JitDiff X64] [xtqqczze] Optimise bounds check in `Replace.ReplaceComparer`

[MihuBot]

Job completed in 16 minutes 35 seconds (remote runner delay: 44 seconds).
dotnet/runtime#119560

Diffs

Found 82 files with textual diffs.

Summary of Code Size diffs:
(Lower is better)

Total bytes of base: 41766169
Total bytes of diff: 41764582
Total bytes of delta: -1587 (-0.00 % of base)
Total relative delta: -5.90
    diff is an improvement.
    relative diff is an improvement.


Top file improvements (bytes):
       -1587 : System.Private.CoreLib.dasm (-0.02 % of base)

1 total files with Code Size differences (1 improved, 0 regressed), 259 unchanged.

Top method improvements (bytes):
        -149 (-39.31 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon) (FullOpts)
        -129 (-29.79 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon,System.Collections.Generic.IEqualityComparer`1[System.__Canon]) (FullOpts)
        -119 (-27.80 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:Replace[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]) (FullOpts)
        -116 (-44.44 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int]) (FullOpts)
        -111 (-31.81 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float],System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]) (FullOpts)
        -110 (-29.97 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]) (FullOpts)
        -108 (-31.12 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[short](System.ReadOnlySpan`1[short],System.Span`1[short],short,short,System.Collections.Generic.IEqualityComparer`1[short]) (FullOpts)
        -104 (-30.77 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[long](System.ReadOnlySpan`1[long],System.Span`1[long],long,long,System.Collections.Generic.IEqualityComparer`1[long]) (FullOpts)
         -99 (-29.73 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[int](System.ReadOnlySpan`1[int],System.Span`1[int],int,int,System.Collections.Generic.IEqualityComparer`1[int]) (FullOpts)
         -96 (-28.32 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[byte](System.ReadOnlySpan`1[byte],System.Span`1[byte],byte,byte,System.Collections.Generic.IEqualityComparer`1[byte]) (FullOpts)
         -91 (-28.00 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[double](System.ReadOnlySpan`1[double],System.Span`1[double],double,double,System.Collections.Generic.IEqualityComparer`1[double]) (FullOpts)
         -86 (-40.00 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float]) (FullOpts)
         -68 (-40.72 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[double](System.ReadOnlySpan`1[double],System.Span`1[double],double,double) (FullOpts)
         -49 (-41.18 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[byte](System.ReadOnlySpan`1[byte],System.Span`1[byte],byte,byte) (FullOpts)
         -46 (-39.66 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[short](System.ReadOnlySpan`1[short],System.Span`1[short],short,short) (FullOpts)
         -46 (-12.81 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:Replace[double](System.ReadOnlySpan`1[double],System.Span`1[double],double,double,System.Collections.Generic.IEqualityComparer`1[double]) (FullOpts)
         -30 (-32.26 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[int](System.ReadOnlySpan`1[int],System.Span`1[int],int,int) (FullOpts)
         -30 (-32.26 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[long](System.ReadOnlySpan`1[long],System.Span`1[long],long,long) (FullOpts)

Top method improvements (percentages):
        -116 (-44.44 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int]) (FullOpts)
         -49 (-41.18 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[byte](System.ReadOnlySpan`1[byte],System.Span`1[byte],byte,byte) (FullOpts)
         -68 (-40.72 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[double](System.ReadOnlySpan`1[double],System.Span`1[double],double,double) (FullOpts)
         -86 (-40.00 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float]) (FullOpts)
         -46 (-39.66 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[short](System.ReadOnlySpan`1[short],System.Span`1[short],short,short) (FullOpts)
        -149 (-39.31 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon) (FullOpts)
         -30 (-32.26 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[int](System.ReadOnlySpan`1[int],System.Span`1[int],int,int) (FullOpts)
         -30 (-32.26 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[long](System.ReadOnlySpan`1[long],System.Span`1[long],long,long) (FullOpts)
        -111 (-31.81 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float],System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]) (FullOpts)
        -108 (-31.12 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[short](System.ReadOnlySpan`1[short],System.Span`1[short],short,short,System.Collections.Generic.IEqualityComparer`1[short]) (FullOpts)
        -104 (-30.77 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[long](System.ReadOnlySpan`1[long],System.Span`1[long],long,long,System.Collections.Generic.IEqualityComparer`1[long]) (FullOpts)
        -110 (-29.97 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]) (FullOpts)
        -129 (-29.79 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon,System.Collections.Generic.IEqualityComparer`1[System.__Canon]) (FullOpts)
         -99 (-29.73 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[int](System.ReadOnlySpan`1[int],System.Span`1[int],int,int,System.Collections.Generic.IEqualityComparer`1[int]) (FullOpts)
         -96 (-28.32 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[byte](System.ReadOnlySpan`1[byte],System.Span`1[byte],byte,byte,System.Collections.Generic.IEqualityComparer`1[byte]) (FullOpts)
         -91 (-28.00 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[double](System.ReadOnlySpan`1[double],System.Span`1[double],double,double,System.Collections.Generic.IEqualityComparer`1[double]) (FullOpts)
        -119 (-27.80 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:Replace[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]) (FullOpts)
         -46 (-12.81 % of base) : System.Private.CoreLib.dasm - System.MemoryExtensions:Replace[double](System.ReadOnlySpan`1[double],System.Span`1[double],double,double,System.Collections.Generic.IEqualityComparer`1[double]) (FullOpts)

18 total methods with Code Size differences (18 improved, 0 regressed), 250993 unchanged.

--------------------------------------------------------------------------------

Artifacts:

• build-clr-checked-main.zip (87 MB)
• build-artifacts-main.zip (682 MB)
• build-clr-checked-pr.zip (87 MB)
• build-artifacts-pr.zip (683 MB)
• jit-diffs-main.zip (139 MB)
• jit-diffs-pr.zip (138 MB)
• diff-frameworks.txt (10 KB)
• ShortDiffsImprovements.md (58 KB)
• LongDiffsImprovements.md (206 KB)

[MihuBot]

Top method improvements

-149 (-39.31 % of base) - System.MemoryExtensions:g__ReplaceDefaultComparer|176_0[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon)

 ; Assembly listing for method System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon) (FullOpts)
 ; Emitting BLENDED_CODE for generic X64 + VEX + EVEX on Unix
 ; FullOpts code
 ; optimized code
 ; rbp based frame
 ; partially interruptible
 ; No PGO data
-; 0 inlinees with PGO data; 1 single block inlinees; 0 inlinees without PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 1 inlinees without PGO data
 ; Final local variable assignments
 ;
-;  V00 TypeCtx      [V00,T09] (  7,  3.65)    long  ->  r13         single-def
+;  V00 TypeCtx      [V00,T10] (  5,  3.60)    long  ->  rdi         single-def
 ;* V01 arg0         [V01    ] (  0,  0   )  struct (16) zero-ref    multireg-arg ld-addr-op single-def <System.ReadOnlySpan`1[System.__Canon]>
 ;* V02 arg1         [V02    ] (  0,  0   )  struct (16) zero-ref    multireg-arg ld-addr-op single-def <System.Span`1[System.__Canon]>
-;  V03 arg2         [V03,T06] (  4,  6   )     ref  ->  r15         class-hnd single-def <System.__Canon>
-;  V04 arg3         [V04,T16] (  2,  2   )     ref  ->  [rbp+0x10]  class-hnd single-def <System.__Canon>
-;  V05 loc0         [V05,T17] (  8,  1.26)     int  ->  [rbp-0x34] 
+;  V03 arg2         [V03,T07] (  3,  5.96)     ref  ->  r14         class-hnd single-def <System.__Canon>
+;  V04 arg3         [V04,T16] (  1,  1.98)     ref  ->  r12         class-hnd single-def <System.__Canon>
+;* V05 loc0         [V05,T18] (  0,  0   )     int  ->  zero-ref   
 ;# V06 OutArgs      [V06    ] (  1,  1   )  struct ( 0) [rsp+0x00]  do-not-enreg[XS] addr-exposed "OutgoingArgSpace" <Empty>
-;  V07 tmp1         [V07,T02] (  4, 16   )    long  ->  rdi         "spilling helperCall"
-;* V08 tmp2         [V08    ] (  0,  0   )   byref  ->  zero-ref    "impAppendStmt"
-;* V09 tmp3         [V09    ] (  0,  0   )   byref  ->  zero-ref   
-;  V10 tmp4         [V10,T07] (  4,  8   )   byref  ->  [rbp-0x48] 
-;  V11 tmp5         [V11,T05] (  6,  8   )     ref  ->  rsi        
-;* V12 tmp6         [V12    ] (  0,  0   )    long  ->  zero-ref    "Inlining Arg"
-;  V13 tmp7         [V13,T04] (  5,  7   )   byref  ->  rbx         single-def "field V01._reference (fldOffset=0x0)" P-INDEP
-;  V14 tmp8         [V14,T10] (  5,  3.04)     int  ->  r12         single-def "field V01._length (fldOffset=0x8)" P-INDEP
-;  V15 tmp9         [V15,T08] (  3,  5   )   byref  ->  r14         single-def "field V02._reference (fldOffset=0x0)" P-INDEP
-;  V16 tmp10        [V16,T13] (  3,  1.54)     int  ->  [rbp-0x38]  single-def "field V02._length (fldOffset=0x8)" P-INDEP
-;  V17 tmp11        [V17,T00] (  6, 24   )     ref  ->  rdi         "argument with side effect"
-;  V18 cse0         [V18,T11] (  4,  4.54)    long  ->  registers   hoist multi-def "CSE #01: moderate"
-;* V19 cse1         [V19,T12] (  0,  0   )    long  ->  zero-ref    "CSE #02: aggressive"
-;  V20 rat0         [V20,T01] (  6, 18.32)    long  ->  [rbp-0x40]  "Strength reduced derived IV"
-;  V21 rat1         [V21,T03] (  4, 12.38)     int  ->  r12         "Trip count IV"
-;  V22 rat2         [V22,T15] (  3,  2   )    long  ->  r13         "runtime lookup"
-;  V23 rat3         [V23,T14] (  3,  2.80)    long  ->  rax         "fgMakeTemp is creating a new local variable"
-;  V24 rat4         [V24,T19] (  3,  0.16)    long  ->   r8         "runtime lookup"
-;  V25 rat5         [V25,T18] (  3,  0.22)    long  ->   r8         "fgMakeTemp is creating a new local variable"
+;* V07 tmp1         [V07    ] (  0,  0   )    long  ->  zero-ref    "spilling helperCall"
+;  V08 tmp2         [V08,T02] (  2, 15.84)    long  ->  rdi         "spilling helperCall"
+;* V09 tmp3         [V09    ] (  0,  0   )   byref  ->  zero-ref    "impAppendStmt"
+;* V10 tmp4         [V10    ] (  0,  0   )   byref  ->  zero-ref   
+;  V11 tmp5         [V11,T08] (  2,  7.92)   byref  ->  [rbp-0x48]  spill-single-def
+;  V12 tmp6         [V12,T06] (  3,  7.92)     ref  ->  rsi        
+;* V13 tmp7         [V13    ] (  0,  0   )     int  ->  zero-ref    "Inlining Arg"
+;* V14 tmp8         [V14    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "NewObj constructor temp" <System.Span`1[System.__Canon]>
+;* V15 tmp9         [V15    ] (  0,  0   )    long  ->  zero-ref    "Inlining Arg"
+;* V16 tmp10        [V16    ] (  0,  0   )   byref  ->  zero-ref    "Inlining Arg"
+;* V17 tmp11        [V17    ] (  0,  0   )    long  ->  zero-ref    "Inlining Arg"
+;  V18 tmp12        [V18,T04] (  3,  6.94)   byref  ->  rbx         single-def "field V01._reference (fldOffset=0x0)" P-INDEP
+;  V19 tmp13        [V19,T11] (  4,  3.50)     int  ->  r13         single-def "field V01._length (fldOffset=0x8)" P-INDEP
+;  V20 tmp14        [V20,T05] (  4,  6.96)   byref  ->  r15         "field V02._reference (fldOffset=0x0)" P-INDEP
+;  V21 tmp15        [V21,T09] (  4,  4   )     int  ->   r8         "field V02._length (fldOffset=0x8)" P-INDEP
+;  V22 tmp16        [V22,T17] (  2,  2   )   byref  ->  r15         single-def "field V14._reference (fldOffset=0x0)" P-INDEP
+;* V23 tmp17        [V23    ] (  0,  0   )     int  ->  zero-ref    "field V14._length (fldOffset=0x8)" P-INDEP
+;  V24 tmp18        [V24,T00] (  3, 23.76)     ref  ->  rdi         "argument with side effect"
+;  V25 cse0         [V25,T12] (  2,  4.46)    long  ->  [rbp-0x38]  spill-single-def hoist "CSE #01: moderate"
+;* V26 cse1         [V26,T13] (  0,  0   )    long  ->  zero-ref    "CSE #02: aggressive"
+;  V27 rat0         [V27,T01] (  6, 18.32)    long  ->  [rbp-0x40]  "Strength reduced derived IV"
+;  V28 rat1         [V28,T03] (  4, 12.38)     int  ->  r13         "Trip count IV"
+;  V29 rat2         [V29,T15] (  3,  2   )    long  ->  rax         "runtime lookup"
+;  V30 rat3         [V30,T14] (  3,  2.80)    long  ->  rax         "fgMakeTemp is creating a new local variable"
 ;
 ; Lcl frame size = 40
 
 G_M54835_IG01:
        push     rbp
        push     r15
        push     r14
        push     r13
        push     r12
        push     rbx
        sub      rsp, 40
        lea      rbp, [rsp+0x50]
        mov      qword ptr [rbp-0x30], rdi
-       mov      r13, rdi
        mov      rbx, rsi
-       mov      r12d, edx
-       mov      r14, rcx
-       mov      eax, r8d
-       mov      r15, r9
-						;; size=41 bbWeight=1 PerfScore 9.25
+       mov      r13d, edx
+       mov      r15, rcx
+       mov      r14, r9
+       mov      r12, gword ptr [rbp+0x10]
+						;; size=39 bbWeight=1 PerfScore 9.75
 G_M54835_IG02:
-       xor      edx, edx
-       mov      dword ptr [rbp-0x34], edx
-       test     r12d, r12d
+       cmp      r13d, r8d
+       jg       G_M54835_IG12
+       mov      r8d, r13d
+       test     r8d, r8d
        jle      G_M54835_IG11
-						;; size=14 bbWeight=1 PerfScore 2.50
+						;; size=21 bbWeight=1 PerfScore 2.75
 G_M54835_IG03:
-       mov      dword ptr [rbp-0x38], eax
-       cmp      r12d, eax
-       jg       G_M54835_IG14
-       mov      rdi, qword ptr [r13+0x18]
-       mov      rax, qword ptr [rdi+0x10]
+       mov      rsi, qword ptr [rdi+0x18]
+       mov      rax, qword ptr [rsi+0x18]
        test     rax, rax
        je       SHORT G_M54835_IG08
-						;; size=25 bbWeight=0.50 PerfScore 3.75
+						;; size=13 bbWeight=0.50 PerfScore 2.62
 G_M54835_IG04:
-       mov      r13, rax
-						;; size=3 bbWeight=0.40 PerfScore 0.10
+       mov      qword ptr [rbp-0x38], rax
+       xor      ecx, ecx
+       jmp      SHORT G_M54835_IG06
+						;; size=8 bbWeight=0.50 PerfScore 1.62
 G_M54835_IG05:
-       xor      eax, eax
-						;; size=2 bbWeight=0.50 PerfScore 0.12
+       mov      rax, qword ptr [rbp-0x38]
+						;; size=4 bbWeight=1.98 PerfScore 1.98
 G_M54835_IG06:
-       mov      rdi, r13
-       mov      qword ptr [rbp-0x40], rax
-       lea      rdx, bword ptr [r14+rax]
+       mov      rdi, rax
+       mov      qword ptr [rbp-0x40], rcx
+       lea      rdx, bword ptr [r15+rcx]
        mov      bword ptr [rbp-0x48], rdx
        mov      rsi, 0xD1FFAB1E      ; code for CORINFO_HELP_GET_GCSTATIC_BASE
        call     [rsi]CORINFO_HELP_GET_GCSTATIC_BASE
        mov      rdi, gword ptr [rax]
        mov      rax, qword ptr [rbp-0x40]
        mov      rsi, gword ptr [rbx+rax]
-       mov      rdx, r15
+       mov      rdx, r14
        mov      rcx, qword ptr [rdi]
        mov      rcx, qword ptr [rcx+0x40]
        call     [rcx+0x20]System.Collections.Generic.EqualityComparer`1[System.__Canon]:Equals(System.__Canon,System.__Canon):bool:this
        test     eax, eax
        jne      SHORT G_M54835_IG09
 						;; size=55 bbWeight=3.96 PerfScore 77.22
 G_M54835_IG07:
-       mov      rcx, qword ptr [rbp-0x40]
-       mov      rsi, gword ptr [rbx+rcx]
+       mov      rax, qword ptr [rbp-0x40]
+       mov      rsi, gword ptr [rbx+rax]
        jmp      SHORT G_M54835_IG10
 						;; size=10 bbWeight=1.98 PerfScore 9.90
 G_M54835_IG08:
-       mov      rdi, r13
        mov      rsi, 0xD1FFAB1E      ; global ptr
        mov      rax, 0xD1FFAB1E      ; code for CORINFO_HELP_RUNTIMEHANDLE_METHOD
        call     [rax]CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       mov      r13, rax
-       jmp      SHORT G_M54835_IG05
-						;; size=30 bbWeight=0.10 PerfScore 0.60
+       jmp      SHORT G_M54835_IG04
+						;; size=24 bbWeight=0.10 PerfScore 0.55
 G_M54835_IG09:
-       mov      rax, gword ptr [rbp+0x10]
-       mov      rsi, rax
-						;; size=7 bbWeight=1.98 PerfScore 2.48
+       mov      rsi, r12
+						;; size=3 bbWeight=1.98 PerfScore 0.49
 G_M54835_IG10:
        mov      rdi, bword ptr [rbp-0x48]
        call     CORINFO_HELP_CHECKED_ASSIGN_REF
-       mov      rcx, qword ptr [rbp-0x40]
-       add      rcx, 8
-       dec      r12d
-       mov      rax, rcx
-       jne      SHORT G_M54835_IG06
+       mov      rax, qword ptr [rbp-0x40]
+       add      rax, 8
+       dec      r13d
+       mov      rcx, rax
+       jne      SHORT G_M54835_IG05
 						;; size=25 bbWeight=3.96 PerfScore 18.81
 G_M54835_IG11:
        add      rsp, 40
        pop      rbx
        pop      r12
        pop      r13
        pop      r14
        pop      r15
        pop      rbp
        ret      
 						;; size=15 bbWeight=1 PerfScore 4.25
 G_M54835_IG12:
-       mov      esi, dword ptr [rbp-0x34]
-       mov      rsi, gword ptr [rbx+8*rsi]
-						;; size=7 bbWeight=0.02 PerfScore 0.06
-G_M54835_IG13:
-       mov      rdi, bword ptr [rbp-0x48]
-       call     CORINFO_HELP_CHECKED_ASSIGN_REF
-       mov      ecx, dword ptr [rbp-0x34]
-       inc      ecx
-       cmp      ecx, r12d
-       mov      dword ptr [rbp-0x34], ecx
-       jge      SHORT G_M54835_IG11
-						;; size=22 bbWeight=0.04 PerfScore 0.22
-G_M54835_IG14:
-       mov      rdi, qword ptr [r13+0x18]
-       mov      r8, qword ptr [rdi+0x10]
-       test     r8, r8
-       je       SHORT G_M54835_IG17
-						;; size=13 bbWeight=0.04 PerfScore 0.21
-G_M54835_IG15:
-       mov      rdi, r8
-       mov      edx, dword ptr [rbp-0x34]
-       cmp      edx, dword ptr [rbp-0x38]
-       jae      SHORT G_M54835_IG18
-       mov      dword ptr [rbp-0x34], edx
-       mov      esi, edx
-       lea      rsi, bword ptr [r14+8*rsi]
-       mov      bword ptr [rbp-0x48], rsi
-       mov      r8, 0xD1FFAB1E      ; code for CORINFO_HELP_GET_GCSTATIC_BASE
-       call     [r8]CORINFO_HELP_GET_GCSTATIC_BASE
-       mov      rdi, gword ptr [rax]
-       mov      esi, dword ptr [rbp-0x34]
-       mov      rsi, gword ptr [rbx+8*rsi]
-       mov      rdx, r15
-       mov      rcx, qword ptr [rdi]
-       mov      rcx, qword ptr [rcx+0x40]
-       call     [rcx+0x20]System.Collections.Generic.EqualityComparer`1[System.__Canon]:Equals(System.__Canon,System.__Canon):bool:this
-       test     eax, eax
-       je       SHORT G_M54835_IG12
-						;; size=64 bbWeight=0.04 PerfScore 0.95
-G_M54835_IG16:
-       mov      rax, gword ptr [rbp+0x10]
-       mov      rsi, rax
-       jmp      SHORT G_M54835_IG13
-						;; size=9 bbWeight=0.02 PerfScore 0.07
-G_M54835_IG17:
-       mov      rdi, r13
-       mov      rsi, 0xD1FFAB1E      ; global ptr
-       mov      r8, 0xD1FFAB1E      ; code for CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       call     [r8]CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       mov      r8, rax
-       jmp      SHORT G_M54835_IG15
-						;; size=31 bbWeight=0.01 PerfScore 0.05
-G_M54835_IG18:
-       call     CORINFO_HELP_RNGCHKFAIL
+       mov      rax, 0xD1FFAB1E      ; code for System.ThrowHelper:ThrowArgumentOutOfRangeException()
+       call     [rax]System.ThrowHelper:ThrowArgumentOutOfRangeException()
        int3     
-						;; size=6 bbWeight=0 PerfScore 0.00
+						;; size=13 bbWeight=0 PerfScore 0.00
 
-; Total bytes of code 379, prolog size 23, PerfScore 130.53, instruction count 112, allocated bytes for code 379 (MethodHash=4b2729cc) for method System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon) (FullOpts)
+; Total bytes of code 230, prolog size 23, PerfScore 129.95, instruction count 68, allocated bytes for code 230 (MethodHash=4b2729cc) for method System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon) (FullOpts)
 ; ============================================================

-129 (-29.79 % of base) - System.MemoryExtensions:g__ReplaceComparer|176_1[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon,System.Collections.Generic.IEqualityComparer`1[System.__Canon])

 ; Assembly listing for method System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon,System.Collections.Generic.IEqualityComparer`1[System.__Canon]) (FullOpts)
 ; Emitting BLENDED_CODE for generic X64 + VEX + EVEX on Unix
 ; FullOpts code
 ; optimized code
 ; rbp based frame
 ; partially interruptible
 ; No PGO data
-; 0 inlinees with PGO data; 1 single block inlinees; 0 inlinees without PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 1 inlinees without PGO data
 ; Final local variable assignments
 ;
-;  V00 TypeCtx      [V00,T08] (  9,  4.25)    long  ->  r13         single-def
+;  V00 TypeCtx      [V00,T08] (  7,  4.20)    long  ->  r13         single-def
 ;* V01 arg0         [V01    ] (  0,  0   )  struct (16) zero-ref    multireg-arg ld-addr-op single-def <System.ReadOnlySpan`1[System.__Canon]>
 ;* V02 arg1         [V02    ] (  0,  0   )  struct (16) zero-ref    multireg-arg ld-addr-op single-def <System.Span`1[System.__Canon]>
-;  V03 arg2         [V03,T05] (  4,  6   )     ref  ->  r15         class-hnd single-def <System.__Canon>
-;  V04 arg3         [V04,T18] (  2,  2   )     ref  ->  [rbp+0x10]  class-hnd single-def <System.__Canon>
-;  V05 arg4         [V05,T09] (  4,  5.50)     ref  ->  r12         class-hnd <System.Collections.Generic.IEqualityComparer`1[System.__Canon]>
-;  V06 loc0         [V06,T19] (  8,  1.26)     int  ->  [rbp-0x34] 
+;  V03 arg2         [V03,T06] (  3,  5.96)     ref  ->  r14         class-hnd single-def <System.__Canon>
+;  V04 arg3         [V04,T18] (  1,  1.98)     ref  ->  [rbp+0x10]  class-hnd single-def <System.__Canon>
+;  V05 arg4         [V05,T10] (  3,  5.46)     ref  ->  [rbp+0x18]  class-hnd <System.Collections.Generic.IEqualityComparer`1[System.__Canon]>
+;* V06 loc0         [V06,T20] (  0,  0   )     int  ->  zero-ref   
 ;# V07 OutArgs      [V07    ] (  1,  1   )  struct ( 0) [rsp+0x00]  do-not-enreg[XS] addr-exposed "OutgoingArgSpace" <Empty>
-;  V08 tmp1         [V08,T00] (  6, 24   )    long  ->  r11         "spilling helperCall"
-;* V09 tmp2         [V09    ] (  0,  0   )   byref  ->  zero-ref    "non-inline candidate call"
-;* V10 tmp3         [V10    ] (  0,  0   )   byref  ->  zero-ref   
-;  V11 tmp4         [V11,T06] (  4,  8   )   byref  ->  [rbp-0x58] 
-;  V12 tmp5         [V12,T04] (  6,  8   )     ref  ->  rsi        
-;* V13 tmp6         [V13    ] (  0,  0   )    long  ->  zero-ref    "spilling helperCall"
-;* V14 tmp7         [V14    ] (  0,  0   )    long  ->  zero-ref    "Inlining Arg"
-;  V15 tmp8         [V15,T03] (  5,  7   )   byref  ->  rbx         single-def "field V01._reference (fldOffset=0x0)" P-INDEP
-;  V16 tmp9         [V16,T10] (  5,  3.04)     int  ->  [rbp-0x38]  single-def "field V01._length (fldOffset=0x8)" P-INDEP
-;  V17 tmp10        [V17,T07] (  3,  5   )   byref  ->  r14         single-def "field V02._reference (fldOffset=0x0)" P-INDEP
-;  V18 tmp11        [V18,T13] (  3,  1.54)     int  ->  [rbp-0x3C]  single-def "field V02._length (fldOffset=0x8)" P-INDEP
-;  V19 cse0         [V19,T11] (  4,  4.54)    long  ->  registers   hoist multi-def "CSE #01: moderate"
-;* V20 cse1         [V20,T12] (  0,  0   )    long  ->  zero-ref    "CSE #02: aggressive"
-;  V21 rat0         [V21,T01] (  6, 18.32)    long  ->  [rbp-0x48]  "Strength reduced derived IV"
-;  V22 rat1         [V22,T02] (  4, 12.38)     int  ->  [rbp-0x4C]  "Trip count IV"
-;  V23 rat2         [V23,T16] (  3,  2   )    long  ->  rdi         "runtime lookup"
-;  V24 rat3         [V24,T14] (  3,  2.80)    long  ->  rdi         "fgMakeTemp is creating a new local variable"
-;  V25 rat4         [V25,T17] (  3,  2   )    long  ->  r13         "runtime lookup"
-;  V26 rat5         [V26,T15] (  3,  2.80)    long  ->  rcx         "fgMakeTemp is creating a new local variable"
-;  V27 rat6         [V27,T21] (  3,  0.16)    long  ->   r9         "runtime lookup"
-;  V28 rat7         [V28,T20] (  3,  0.22)    long  ->   r9         "fgMakeTemp is creating a new local variable"
+;* V08 tmp1         [V08    ] (  0,  0   )    long  ->  zero-ref    "spilling helperCall"
+;  V09 tmp2         [V09,T00] (  3, 23.76)    long  ->  r11         "spilling helperCall"
+;* V10 tmp3         [V10    ] (  0,  0   )   byref  ->  zero-ref    "non-inline candidate call"
+;* V11 tmp4         [V11    ] (  0,  0   )   byref  ->  zero-ref   
+;  V12 tmp5         [V12,T07] (  2,  7.92)   byref  ->  [rbp-0x40]  spill-single-def
+;  V13 tmp6         [V13,T05] (  3,  7.92)     ref  ->  rsi        
+;* V14 tmp7         [V14    ] (  0,  0   )    long  ->  zero-ref    "spilling helperCall"
+;* V15 tmp8         [V15    ] (  0,  0   )     int  ->  zero-ref    "Inlining Arg"
+;* V16 tmp9         [V16    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "NewObj constructor temp" <System.Span`1[System.__Canon]>
+;* V17 tmp10        [V17    ] (  0,  0   )    long  ->  zero-ref    "Inlining Arg"
+;* V18 tmp11        [V18    ] (  0,  0   )   byref  ->  zero-ref    "Inlining Arg"
+;* V19 tmp12        [V19    ] (  0,  0   )    long  ->  zero-ref    "Inlining Arg"
+;  V20 tmp13        [V20,T03] (  3,  6.94)   byref  ->  rbx         single-def "field V01._reference (fldOffset=0x0)" P-INDEP
+;  V21 tmp14        [V21,T11] (  4,  3.50)     int  ->  [rbp-0x34]  single-def "field V01._length (fldOffset=0x8)" P-INDEP
+;  V22 tmp15        [V22,T04] (  4,  6.96)   byref  ->  r15         "field V02._reference (fldOffset=0x0)" P-INDEP
+;  V23 tmp16        [V23,T09] (  4,  4   )     int  ->  r12         "field V02._length (fldOffset=0x8)" P-INDEP
+;  V24 tmp17        [V24,T19] (  2,  2   )   byref  ->  r15         single-def "field V16._reference (fldOffset=0x0)" P-INDEP
+;* V25 tmp18        [V25    ] (  0,  0   )     int  ->  zero-ref    "field V16._length (fldOffset=0x8)" P-INDEP
+;  V26 cse0         [V26,T12] (  2,  4.46)    long  ->  r12         hoist "CSE #01: moderate"
+;* V27 cse1         [V27,T13] (  0,  0   )    long  ->  zero-ref    "CSE #02: aggressive"
+;  V28 rat0         [V28,T01] (  6, 18.32)    long  ->  r13         "Strength reduced derived IV"
+;  V29 rat1         [V29,T02] (  4, 12.38)     int  ->  [rbp-0x38]  "Trip count IV"
+;  V30 rat2         [V30,T16] (  3,  2   )    long  ->  rdi         "runtime lookup"
+;  V31 rat3         [V31,T14] (  3,  2.80)    long  ->  rdi         "fgMakeTemp is creating a new local variable"
+;  V32 rat4         [V32,T17] (  3,  2   )    long  ->  r12         "runtime lookup"
+;  V33 rat5         [V33,T15] (  3,  2.80)    long  ->  r12         "fgMakeTemp is creating a new local variable"
 ;
-; Lcl frame size = 56
+; Lcl frame size = 24
 
 G_M8431_IG01:
        push     rbp
        push     r15
        push     r14
        push     r13
        push     r12
        push     rbx
-       sub      rsp, 56
-       lea      rbp, [rsp+0x60]
+       sub      rsp, 24
+       lea      rbp, [rsp+0x40]
        mov      qword ptr [rbp-0x30], rdi
-       mov      dword ptr [rbp-0x38], edx
-       mov      dword ptr [rbp-0x3C], r8d
        mov      r13, rdi
        mov      rbx, rsi
-       mov      r14, rcx
-       mov      r15, r9
-       mov      r12, gword ptr [rbp+0x18]
-						;; size=46 bbWeight=1 PerfScore 11.75
+       mov      r15, rcx
+       mov      ecx, edx
+       mov      r12d, r8d
+       mov      r14, r9
+       mov      rax, gword ptr [rbp+0x18]
+						;; size=44 bbWeight=1 PerfScore 10.25
 G_M8431_IG02:
-       test     r12, r12
+       cmp      ecx, r12d
+       jg       G_M8431_IG15
+       mov      dword ptr [rbp-0x34], ecx
+       mov      r12d, ecx
+       test     rax, rax
        jne      SHORT G_M8431_IG05
-						;; size=5 bbWeight=1 PerfScore 1.25
+						;; size=20 bbWeight=1 PerfScore 3.75
 G_M8431_IG03:
        mov      rdi, qword ptr [r13+0x18]
-       mov      rdi, qword ptr [rdi+0x18]
+       mov      rdi, qword ptr [rdi+0x20]
        test     rdi, rdi
-       je       SHORT G_M8431_IG11
+       je       SHORT G_M8431_IG10
 						;; size=13 bbWeight=0.50 PerfScore 2.62
 G_M8431_IG04:
-       mov      rsi, 0xD1FFAB1E      ; code for CORINFO_HELP_GET_GCSTATIC_BASE
-       call     [rsi]CORINFO_HELP_GET_GCSTATIC_BASE
-       mov      r12, gword ptr [rax]
-						;; size=15 bbWeight=0.50 PerfScore 2.62
+       mov      rax, 0xD1FFAB1E      ; code for CORINFO_HELP_GET_GCSTATIC_BASE
+       call     [rax]CORINFO_HELP_GET_GCSTATIC_BASE
+       mov      rax, gword ptr [rax]
+       mov      gword ptr [rbp+0x18], rax
+       mov      rax, gword ptr [rbp+0x18]
+						;; size=23 bbWeight=0.50 PerfScore 3.62
 G_M8431_IG05:
-       xor      r8d, r8d
-       mov      dword ptr [rbp-0x34], r8d
-       mov      eax, dword ptr [rbp-0x38]
-       test     eax, eax
-       jle      G_M8431_IG15
-						;; size=18 bbWeight=1 PerfScore 3.50
+       test     r12d, r12d
+       jle      G_M8431_IG14
+						;; size=9 bbWeight=1 PerfScore 1.25
 G_M8431_IG06:
-       mov      dword ptr [rbp-0x38], eax
-       cmp      eax, dword ptr [rbp-0x3C]
-       jg       G_M8431_IG18
        mov      rdi, qword ptr [r13+0x18]
-       mov      rcx, qword ptr [rdi+0x10]
-       test     rcx, rcx
-       je       SHORT G_M8431_IG12
-						;; size=25 bbWeight=0.50 PerfScore 4.62
+       mov      r12, qword ptr [rdi+0x18]
+       test     r12, r12
+       je       SHORT G_M8431_IG11
+						;; size=13 bbWeight=0.50 PerfScore 2.62
 G_M8431_IG07:
-       mov      r13, rcx
-						;; size=3 bbWeight=0.40 PerfScore 0.10
+       xor      r13d, r13d
+       mov      ecx, dword ptr [rbp-0x34]
+       mov      dword ptr [rbp-0x38], ecx
+						;; size=9 bbWeight=0.50 PerfScore 1.12
 G_M8431_IG08:
-       xor      ecx, ecx
-       mov      eax, dword ptr [rbp-0x38]
-       mov      dword ptr [rbp-0x4C], eax
-						;; size=8 bbWeight=0.50 PerfScore 1.12
-G_M8431_IG09:
-       mov      r11, r13
-       lea      r8, bword ptr [r14+rcx]
-       mov      bword ptr [rbp-0x58], r8
-       mov      qword ptr [rbp-0x48], rcx
-       mov      rsi, gword ptr [rbx+rcx]
-       mov      rdi, r12
-       mov      rdx, r15
+       mov      r11, r12
+       lea      r8, bword ptr [r15+r13]
+       mov      bword ptr [rbp-0x40], r8
+       mov      rsi, gword ptr [rbx+r13]
+       mov      gword ptr [rbp+0x18], rax
+       mov      rdi, rax
+       mov      rdx, r14
        call     [r11]
        test     eax, eax
-       jne      SHORT G_M8431_IG13
+       jne      SHORT G_M8431_IG12
 						;; size=32 bbWeight=3.96 PerfScore 37.62
+G_M8431_IG09:
+       mov      rsi, gword ptr [rbx+r13]
+       jmp      SHORT G_M8431_IG13
+						;; size=6 bbWeight=1.98 PerfScore 7.92
 G_M8431_IG10:
-       mov      rcx, qword ptr [rbp-0x48]
-       mov      rsi, gword ptr [rbx+rcx]
-       jmp      SHORT G_M8431_IG14
-						;; size=10 bbWeight=1.98 PerfScore 9.90
+       mov      rdi, r13
+       mov      rsi, 0xD1FFAB1E      ; global ptr
+       mov      rax, 0xD1FFAB1E      ; code for CORINFO_HELP_RUNTIMEHANDLE_METHOD
+       call     [rax]CORINFO_HELP_RUNTIMEHANDLE_METHOD
+       mov      rdi, rax
+       jmp      SHORT G_M8431_IG04
+						;; size=30 bbWeight=0.10 PerfScore 0.60
 G_M8431_IG11:
+       mov      gword ptr [rbp+0x18], rax
        mov      rdi, r13
        mov      rsi, 0xD1FFAB1E      ; global ptr
        mov      r8, 0xD1FFAB1E      ; code for CORINFO_HELP_RUNTIMEHANDLE_METHOD
        call     [r8]CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       mov      rdi, rax
-       jmp      G_M8431_IG04
-						;; size=34 bbWeight=0.10 PerfScore 0.60
+       mov      r12, rax
+       mov      rax, gword ptr [rbp+0x18]
+       jmp      SHORT G_M8431_IG07
+						;; size=39 bbWeight=0.10 PerfScore 0.80
 G_M8431_IG12:
-       mov      rdi, r13
-       mov      rsi, 0xD1FFAB1E      ; global ptr
-       mov      rcx, 0xD1FFAB1E      ; code for CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       call     [rcx]CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       mov      r13, rax
-       jmp      SHORT G_M8431_IG08
-						;; size=30 bbWeight=0.10 PerfScore 0.60
-G_M8431_IG13:
        mov      rax, gword ptr [rbp+0x10]
        mov      rsi, rax
 						;; size=7 bbWeight=1.98 PerfScore 2.48
-G_M8431_IG14:
-       mov      rdi, bword ptr [rbp-0x58]
+G_M8431_IG13:
+       mov      rdi, bword ptr [rbp-0x40]
        call     CORINFO_HELP_CHECKED_ASSIGN_REF
-       mov      rcx, qword ptr [rbp-0x48]
-       add      rcx, 8
-       mov      eax, dword ptr [rbp-0x4C]
-       dec      eax
-       mov      dword ptr [rbp-0x4C], eax
-       jne      G_M8431_IG09
+       add      r13, 8
+       mov      ecx, dword ptr [rbp-0x38]
+       dec      ecx
+       mov      dword ptr [rbp-0x38], ecx
+       mov      rax, gword ptr [rbp+0x18]
+       jne      G_M8431_IG08
 						;; size=31 bbWeight=3.96 PerfScore 25.74
-G_M8431_IG15:
-       add      rsp, 56
+G_M8431_IG14:
+       add      rsp, 24
        pop      rbx
        pop      r12
        pop      r13
        pop      r14
        pop      r15
        pop      rbp
        ret      
 						;; size=15 bbWeight=1 PerfScore 4.25
-G_M8431_IG16:
-       mov      esi, dword ptr [rbp-0x34]
-       mov      rsi, gword ptr [rbx+8*rsi]
-						;; size=7 bbWeight=0.02 PerfScore 0.06
-G_M8431_IG17:
-       mov      rdi, bword ptr [rbp-0x58]
-       call     CORINFO_HELP_CHECKED_ASSIGN_REF
-       mov      ecx, dword ptr [rbp-0x34]
-       inc      ecx
-       mov      eax, dword ptr [rbp-0x38]
-       cmp      ecx, eax
-       mov      dword ptr [rbp-0x38], eax
-       mov      dword ptr [rbp-0x34], ecx
-       jge      SHORT G_M8431_IG15
-						;; size=27 bbWeight=0.04 PerfScore 0.30
-G_M8431_IG18:
-       mov      rdi, qword ptr [r13+0x18]
-       mov      r9, qword ptr [rdi+0x10]
-       test     r9, r9
-       je       SHORT G_M8431_IG21
-						;; size=13 bbWeight=0.04 PerfScore 0.21
-G_M8431_IG19:
-       mov      r11, r9
-       mov      r8d, dword ptr [rbp-0x34]
-       cmp      r8d, dword ptr [rbp-0x3C]
-       jae      SHORT G_M8431_IG22
-       mov      esi, r8d
-       lea      rdi, bword ptr [r14+8*rsi]
-       mov      bword ptr [rbp-0x58], rdi
-       mov      dword ptr [rbp-0x34], r8d
-       mov      esi, r8d
-       mov      rsi, gword ptr [rbx+8*rsi]
-       mov      rdi, r12
-       mov      rdx, r15
-       call     [r11]
-       test     eax, eax
-       je       SHORT G_M8431_IG16
-						;; size=48 bbWeight=0.04 PerfScore 0.56
-G_M8431_IG20:
-       mov      rax, gword ptr [rbp+0x10]
-       mov      rsi, rax
-       jmp      SHORT G_M8431_IG17
-						;; size=9 bbWeight=0.02 PerfScore 0.07
-G_M8431_IG21:
-       mov      rdi, r13
-       mov      rsi, 0xD1FFAB1E      ; global ptr
-       mov      r9, 0xD1FFAB1E      ; code for CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       call     [r9]CORINFO_HELP_RUNTIMEHANDLE_METHOD
-       mov      r9, rax
-       jmp      SHORT G_M8431_IG19
-						;; size=31 bbWeight=0.01 PerfScore 0.05
-G_M8431_IG22:
-       call     CORINFO_HELP_RNGCHKFAIL
+G_M8431_IG15:
+       mov      rax, 0xD1FFAB1E      ; code for System.ThrowHelper:ThrowArgumentOutOfRangeException()
+       call     [rax]System.ThrowHelper:ThrowArgumentOutOfRangeException()
        int3     
-						;; size=6 bbWeight=0 PerfScore 0.00
+						;; size=13 bbWeight=0 PerfScore 0.00
 
-; Total bytes of code 433, prolog size 23, PerfScore 110.03, instruction count 125, allocated bytes for code 433 (MethodHash=8393df10) for method System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon,System.Collections.Generic.IEqualityComparer`1[System.__Canon]) (FullOpts)
+; Total bytes of code 304, prolog size 23, PerfScore 104.65, instruction count 87, allocated bytes for code 304 (MethodHash=8393df10) for method System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.__Canon](System.ReadOnlySpan`1[System.__Canon],System.Span`1[System.__Canon],System.__Canon,System.__Canon,System.Collections.Generic.IEqualityComparer`1[System.__Canon]) (FullOpts)
 ; ============================================================

-119 (-27.80 % of base) - System.MemoryExtensions:Replace[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]])

 ; Assembly listing for method System.MemoryExtensions:Replace[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]) (FullOpts)
 ; Emitting BLENDED_CODE for generic X64 + VEX + EVEX on Unix
 ; FullOpts code
 ; optimized code
 ; rbp based frame
 ; fully interruptible
 ; No PGO data
-; 0 inlinees with PGO data; 10 single block inlinees; 3 inlinees without PGO data
+; 0 inlinees with PGO data; 4 single block inlinees; 0 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[System.Nullable`1[int]]>
 ;* V01 arg1         [V01    ] (  0,  0   )  struct (16) zero-ref    multireg-arg ld-addr-op single-def <System.Span`1[System.Nullable`1[int]]>
-;* V02 arg2         [V02    ] (  0,  0   )  struct ( 8) zero-ref    do-not-enreg[SF] single-def <System.Nullable`1[int]>
-;* V03 arg3         [V03    ] (  0,  0   )  struct ( 8) zero-ref    do-not-enreg[SF] single-def <System.Nullable`1[int]>
-;  V04 arg4         [V04,T27] (  3,  1.50)     ref  ->  r12         class-hnd single-def <System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]>
+;  V02 arg2         [V02,T02] (  4,  3   )  struct ( 8) [rbp-0x30]  do-not-enreg[SF] single-def <System.Nullable`1[int]>
+;  V03 arg3         [V03,T03] (  4,  3   )  struct ( 8) [rbp-0x38]  do-not-enreg[SF] single-def <System.Nullable`1[int]>
+;  V04 arg4         [V04,T07] (  3,  1.50)     ref  ->  r12         class-hnd single-def <System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]>
 ;* V05 loc0         [V05    ] (  0,  0   )     int  ->  zero-ref   
 ;* V06 loc1         [V06    ] (  0,  0   )   byref  ->  zero-ref    single-def
 ;* V07 loc2         [V07    ] (  0,  0   )   byref  ->  zero-ref   
-;  V08 loc3         [V08,T26] (  4,  2   )    long  ->  rdi        
+;  V08 loc3         [V08,T06] (  4,  2   )    long  ->  rdi        
 ;# V09 OutArgs      [V09    ] (  1,  1   )  struct ( 0) [rsp+0x00]  do-not-enreg[XS] addr-exposed "OutgoingArgSpace" <Empty>
 ;* V10 tmp1         [V10    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "Inlining Arg" <System.ReadOnlySpan`1[System.Nullable`1[int]]>
 ;* V11 tmp2         [V11    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "Inlining Arg" <System.Span`1[System.Nullable`1[int]]>
-;* V12 tmp3         [V12,T28] (  0,  0   )     int  ->  zero-ref    "Inline stloc first use temp"
-;* V13 tmp4         [V13    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "Inlining Arg" <System.ReadOnlySpan`1[System.Nullable`1[int]]>
-;* V14 tmp5         [V14    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "Inlining Arg" <System.Span`1[System.Nullable`1[int]]>
-;  V15 tmp6         [V15,T00] (  3, 23.76)   byref  ->   r9         "impAppendStmt"
-;* V16 tmp7         [V16    ] (  0,  0   )   byref  ->  zero-ref   
-;* V17 tmp8         [V17    ] (  0,  0   )   byref  ->  zero-ref   
-;* V18 tmp9         [V18    ] (  0,  0   )  struct ( 8) zero-ref    <System.Nullable`1[int]>
-;  V19 tmp10        [V19,T03] (  2,  7.92)   ubyte  ->  r11         "Inline return value spill temp"
-;* V20 tmp11        [V20    ] (  0,  0   )  struct ( 8) zero-ref    ld-addr-op "Inlining Arg" <System.Nullable`1[int]>
-;* V21 tmp12        [V21    ] (  0,  0   )  struct ( 8) zero-ref    ld-addr-op "Inlining Arg" <System.Nullable`1[int]>
-;* V22 tmp13        [V22    ] (  0,  0   )   ubyte  ->  zero-ref    "Inline return value spill temp"
-;* V23 tmp14        [V23    ] (  0,  0   )     int  ->  zero-ref    ld-addr-op "Inlining Arg"
-;* V24 tmp15        [V24    ] (  0,  0   )     int  ->  zero-ref    "Inlining Arg"
-;  V25 tmp16        [V25,T16] (  4,  2.50)   byref  ->  r14         single-def "field V00._reference (fldOffset=0x0)" P-INDEP
-;  V26 tmp17        [V26,T12] (  6,  4   )     int  ->  rbx         single-def "field V00._length (fldOffset=0x8)" P-INDEP
-;  V27 tmp18        [V27,T17] (  4,  2.50)   byref  ->  r13         single-def "field V01._reference (fldOffset=0x0)" P-INDEP
-;  V28 tmp19        [V28,T15] (  4,  2.50)     int  ->  r15         single-def "field V01._length (fldOffset=0x8)" P-INDEP
-;* V29 tmp20        [V29    ] (  0,  0   )   byref  ->  zero-ref    single-def "field V10._reference (fldOffset=0x0)" P-INDEP
-;* V30 tmp21        [V30    ] (  0,  0   )     int  ->  zero-ref    "field V10._length (fldOffset=0x8)" P-INDEP
-;* V31 tmp22        [V31    ] (  0,  0   )   byref  ->  zero-ref    single-def "field V11._reference (fldOffset=0x0)" P-INDEP
-;* V32 tmp23        [V32    ] (  0,  0   )     int  ->  zero-ref    "field V11._length (fldOffset=0x8)" P-INDEP
-;  V33 tmp24        [V33,T18] (  2,  4.46)   byref  ->  r14         single-def "field V13._reference (fldOffset=0x0)" P-INDEP
-;  V34 tmp25        [V34,T20] (  2,  4.46)     int  ->  rbx         "field V13._length (fldOffset=0x8)" P-INDEP
-;  V35 tmp26        [V35,T19] (  2,  4.46)   byref  ->  r13         single-def "field V14._reference (fldOffset=0x0)" P-INDEP
-;* V36 tmp27        [V36,T29] (  0,  0   )     int  ->  zero-ref    "field V14._length (fldOffset=0x8)" P-INDEP
-;  V37 tmp28        [V37,T08] (  3,  7.92)   ubyte  ->  r11         "field V18.hasValue (fldOffset=0x0)" P-INDEP
-;  V38 tmp29        [V38,T09] (  3,  7.92)     int  ->  r15         "field V18.value (fldOffset=0x4)" P-INDEP
-;  V39 tmp30        [V39,T10] (  2,  7.92)   ubyte  ->  r11         "field V20.hasValue (fldOffset=0x0)" P-INDEP
-;  V40 tmp31        [V40,T11] (  2,  7.92)     int  ->  r15         "field V20.value (fldOffset=0x4)" P-INDEP
-;* V41 tmp32        [V41    ] (  0,  0   )   ubyte  ->  zero-ref    "field V21.hasValue (fldOffset=0x0)" P-INDEP
-;* V42 tmp33        [V42    ] (  0,  0   )     int  ->  zero-ref    "field V21.value (fldOffset=0x4)" P-INDEP
-;  V43 tmp34        [V43,T13] (  4,  5.46)   ubyte  ->  [rbp-0x2C]  spill-single-def "V02.[000..001)"
-;  V44 tmp35        [V44,T14] (  3,  5.46)     int  ->  [rbp-0x30]  spill-single-def "V02.[004..008)"
-;  V45 tmp36        [V45,T24] (  3,  3.48)   ubyte  ->  [rbp-0x34]  spill-single-def "V03.[000..001)"
-;  V46 tmp37        [V46,T25] (  3,  3.48)     int  ->  [rbp-0x38]  spill-single-def "V03.[004..008)"
-;  V47 tmp38        [V47,T01] (  3, 23.76)   byref  ->   r8         "BlockOp address local"
-;* V48 tmp39        [V48,T23] (  0,  0   )   byref  ->  zero-ref    "BlockOp address local"
-;  V49 cse0         [V49,T05] (  2,  7.92)   byref  ->   r8         "CSE #03: aggressive"
-;  V50 cse1         [V50,T04] (  3, 11.88)    long  ->   r8         "CSE #02: aggressive"
-;  V51 cse2         [V51,T06] (  3,  9.90)     int  ->  r10         "CSE #04: aggressive"
-;  V52 cse3         [V52,T07] (  3,  9.90)     int  ->   r8         "CSE #05: aggressive"
-;  V53 rat0         [V53,T02] (  5, 16.34)    long  ->  rdi         "Widened IV V12"
-;  V54 rat1         [V54,T21] (  4,  4   )    long  ->   r8         "V02.r8"
-;  V55 rat2         [V55,T22] (  4,  4   )    long  ->   r9         "V03.r9"
+;  V12 tmp3         [V12,T04] (  4,  2.50)   byref  ->  r14         single-def "field V00._reference (fldOffset=0x0)" P-INDEP
+;  V13 tmp4         [V13,T00] (  6,  4   )     int  ->  rbx         single-def "field V00._length (fldOffset=0x8)" P-INDEP
+;  V14 tmp5         [V14,T05] (  4,  2.50)   byref  ->  r13         single-def "field V01._reference (fldOffset=0x0)" P-INDEP
+;  V15 tmp6         [V15,T01] (  5,  3   )     int  ->  r15         single-def "field V01._length (fldOffset=0x8)" P-INDEP
+;* V16 tmp7         [V16    ] (  0,  0   )   byref  ->  zero-ref    single-def "field V10._reference (fldOffset=0x0)" P-INDEP
+;* V17 tmp8         [V17    ] (  0,  0   )     int  ->  zero-ref    "field V10._length (fldOffset=0x8)" P-INDEP
+;* V18 tmp9         [V18    ] (  0,  0   )   byref  ->  zero-ref    single-def "field V11._reference (fldOffset=0x0)" P-INDEP
+;* V19 tmp10        [V19    ] (  0,  0   )     int  ->  zero-ref    "field V11._length (fldOffset=0x8)" P-INDEP
 ;
 ; Lcl frame size = 24
 
 G_M10641_IG01:
        push     rbp
        push     r15
        push     r14
        push     r13
        push     r12
        push     rbx
        sub      rsp, 24
        lea      rbp, [rsp+0x40]
+       mov      qword ptr [rbp-0x30], r8
+       mov      qword ptr [rbp-0x38], r9
        mov      r14, rdi
        mov      ebx, esi
        mov      r13, rdx
        mov      r15d, ecx
        mov      r12, gword ptr [rbp+0x10]
-						;; size=34 bbWeight=1 PerfScore 8.75
+						;; size=42 bbWeight=1 PerfScore 10.75
 G_M10641_IG02:
-       movzx    rax, r8b
-       mov      dword ptr [rbp-0x2C], eax
-       shr      r8, 32
-       mov      ecx, r8d
-       mov      dword ptr [rbp-0x30], ecx
-       movzx    rdx, r9b
-       mov      dword ptr [rbp-0x34], edx
-       shr      r9, 32
-       mov      esi, r9d
-       mov      dword ptr [rbp-0x38], esi
        test     ebx, ebx
-       je       G_M10641_IG17
-						;; size=42 bbWeight=1 PerfScore 7.25
+       je       G_M10641_IG10
+						;; size=8 bbWeight=1 PerfScore 1.25
 G_M10641_IG03:
        cmp      ebx, r15d
-       jg       G_M10641_IG18
+       jg       G_M10641_IG11
        mov      rdi, r13
        sub      rdi, r14
        je       SHORT G_M10641_IG04
-       mov      r8d, ebx
-       shl      r8, 3
-       cmp      rdi, r8
-       jb       G_M10641_IG19
-       mov      r8d, r15d
-       shl      r8, 3
-       neg      r8
-       cmp      r8, rdi
-       jb       G_M10641_IG19
-						;; size=52 bbWeight=0.50 PerfScore 3.50
+       mov      eax, ebx
+       shl      rax, 3
+       cmp      rdi, rax
+       jb       G_M10641_IG12
+       mov      eax, r15d
+       shl      rax, 3
+       neg      rax
+       cmp      rax, rdi
+       jb       G_M10641_IG12
+						;; size=51 bbWeight=0.50 PerfScore 3.50
 G_M10641_IG04:
        test     r12, r12
        je       SHORT G_M10641_IG08
        mov      rdi, 0xD1FFAB1E      ; global ptr
        test     byte  ptr [rdi], 1
-       je       G_M10641_IG20
+       je       G_M10641_IG13
 						;; size=24 bbWeight=0.50 PerfScore 2.75
 G_M10641_IG05:
-       mov      r8, 0xD1FFAB1E      ; data for System.Collections.Generic.EqualityComparer`1[System.Nullable`1[int]]:<Default>k__BackingField
-       cmp      r12, gword ptr [r8]
+       mov      rdi, 0xD1FFAB1E      ; data for System.Collections.Generic.EqualityComparer`1[System.Nullable`1[int]]:<Default>k__BackingField
+       cmp      r12, gword ptr [rdi]
        je       SHORT G_M10641_IG08
 						;; size=15 bbWeight=0.50 PerfScore 2.12
 G_M10641_IG06:
        mov      gword ptr [rbp+0x10], r12
-       mov      r8d, dword ptr [rbp-0x30]
-       shl      r8, 32
-       movzx    r9, byte  ptr [rbp-0x2C]
-       or       r8, r9
-       mov      r9d, dword ptr [rbp-0x38]
-       shl      r9, 32
-       movzx    rdi, byte  ptr [rbp-0x34]
-       or       r9, rdi
        mov      rdi, r14
        mov      esi, ebx
        mov      rdx, r13
        mov      ecx, r15d
+       mov      r8, qword ptr [rbp-0x30]
+       mov      r9, qword ptr [rbp-0x38]
        mov      rax, 0xD1FFAB1E      ; code for System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]])
-						;; size=57 bbWeight=0.50 PerfScore 3.88
+						;; size=33 bbWeight=0.50 PerfScore 2.12
 G_M10641_IG07:
        add      rsp, 24
        pop      rbx
        pop      r12
        pop      r13
        pop      r14
        pop      r15
        pop      rbp
        tail.jmp [rax]System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]])
 						;; size=17 bbWeight=0.50 PerfScore 3.62
 G_M10641_IG08:
-       xor      edi, edi
-       align    [0 bytes for IG09]
-						;; size=2 bbWeight=0.50 PerfScore 0.12
+       mov      rdi, r14
+       mov      esi, ebx
+       mov      rdx, r13
+       mov      ecx, r15d
+       mov      r8, qword ptr [rbp-0x30]
+       mov      r9, qword ptr [rbp-0x38]
+       mov      rax, 0xD1FFAB1E      ; code for System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int])
+						;; size=29 bbWeight=0.50 PerfScore 1.62
 G_M10641_IG09:
-       lea      r8, [8*rdi]
-       lea      r9, bword ptr [r8+r13]
-       add      r8, r14
-       movzx    r10, byte  ptr [r8]
-       movzx    r11, r10b
-       mov      r8d, dword ptr [r8+0x04]
-       mov      r15d, r8d
-       test     r11d, r11d
-       jne      SHORT G_M10641_IG11
-						;; size=35 bbWeight=3.96 PerfScore 27.72
+       add      rsp, 24
+       pop      rbx
+       pop      r12
+       pop      r13
+       pop      r14
+       pop      r15
+       pop      rbp
+       tail.jmp [rax]System.MemoryExtensions:<Replace>g__ReplaceDefaultComparer|176_0[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int])
+						;; size=17 bbWeight=0.50 PerfScore 3.62
 G_M10641_IG10:
-       mov      eax, dword ptr [rbp-0x2C]
-       test     eax, eax
-       mov      ecx, dword ptr [rbp-0x30]
-       jne      SHORT G_M10641_IG13
-       jmp      SHORT G_M10641_IG15
-						;; size=12 bbWeight=1.98 PerfScore 10.39
-G_M10641_IG11:
-       mov      eax, dword ptr [rbp-0x2C]
-       test     eax, eax
-       jne      SHORT G_M10641_IG14
-						;; size=7 bbWeight=1.98 PerfScore 4.46
-G_M10641_IG12:
-       mov      ecx, dword ptr [rbp-0x30]
-						;; size=3 bbWeight=0.99 PerfScore 0.99
-G_M10641_IG13:
-       movzx    r11, r10b
-       mov      r15d, r8d
-       jmp      SHORT G_M10641_IG16
-						;; size=9 bbWeight=1.98 PerfScore 4.95
-G_M10641_IG14:
-       mov      ecx, dword ptr [rbp-0x30]
-       cmp      r15d, ecx
-       sete     r11b
-       movzx    r11, r11b
-       test     r11d, r11d
-       je       SHORT G_M10641_IG13
-						;; size=19 bbWeight=3.96 PerfScore 14.85
-G_M10641_IG15:
-       mov      edx, dword ptr [rbp-0x34]
-       movzx    r11, dl
-       mov      esi, dword ptr [rbp-0x38]
-       mov      r15d, esi
-						;; size=13 bbWeight=1.98 PerfScore 4.95
-G_M10641_IG16:
-       mov      byte  ptr [r9], r11b
-       mov      dword ptr [r9+0x04], r15d
-       inc      edi
-       cmp      edi, ebx
-       jl       SHORT G_M10641_IG09
-						;; size=13 bbWeight=3.96 PerfScore 13.86
-G_M10641_IG17:
        add      rsp, 24
        pop      rbx
        pop      r12
        pop      r13
        pop      r14
        pop      r15
        pop      rbp
        ret      
 						;; size=15 bbWeight=0.50 PerfScore 2.12
-G_M10641_IG18:
+G_M10641_IG11:
        mov      rax, 0xD1FFAB1E      ; code for System.ThrowHelper:ThrowArgumentException_DestinationTooShort()
        call     [rax]System.ThrowHelper:ThrowArgumentException_DestinationTooShort()
        int3     
 						;; size=13 bbWeight=0 PerfScore 0.00
-G_M10641_IG19:
+G_M10641_IG12:
        mov      edi, 72
        mov      rax, 0xD1FFAB1E      ; code for System.ThrowHelper:ThrowArgumentException(int)
        call     [rax]System.ThrowHelper:ThrowArgumentException(int)
        int3     
 						;; size=18 bbWeight=0 PerfScore 0.00
-G_M10641_IG20:
+G_M10641_IG13:
        mov      rdi, 0xD1FFAB1E      ; System.Collections.Generic.EqualityComparer`1[System.Nullable`1[int]]
-       mov      r8, 0xD1FFAB1E      ; code for CORINFO_HELP_GET_GCSTATIC_BASE
-       call     [r8]CORINFO_HELP_GET_GCSTATIC_BASE
+       mov      rax, 0xD1FFAB1E      ; code for CORINFO_HELP_GET_GCSTATIC_BASE
+       call     [rax]CORINFO_HELP_GET_GCSTATIC_BASE
        jmp      G_M10641_IG05
-						;; size=28 bbWeight=0 PerfScore 0.00
+						;; size=27 bbWeight=0 PerfScore 0.00
 
-; Total bytes of code 428, prolog size 34, PerfScore 116.29, instruction count 126, allocated bytes for code 428 (MethodHash=b80cd66e) for method System.MemoryExtensions:Replace[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]) (FullOpts)
+; Total bytes of code 309, prolog size 42, PerfScore 33.50, instruction count 89, allocated bytes for code 309 (MethodHash=b80cd66e) for method System.MemoryExtensions:Replace[System.Nullable`1[int]](System.ReadOnlySpan`1[System.Nullable`1[int]],System.Span`1[System.Nullable`1[int]],System.Nullable`1[int],System.Nullable`1[int],System.Collections.Generic.IEqualityComparer`1[System.Nullable`1[int]]) (FullOpts)
 ; ============================================================

-111 (-31.81 % of base) - System.MemoryExtensions:g__ReplaceComparer|176_1[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float],System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]])

 ; Assembly listing for method System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float],System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]) (FullOpts)
 ; Emitting BLENDED_CODE for generic X64 + VEX + EVEX on Unix
 ; FullOpts code
 ; optimized code
 ; rbp based frame
 ; partially interruptible
 ; No PGO data
-; 0 inlinees with PGO data; 1 single block inlinees; 0 inlinees without PGO data
+; 0 inlinees with PGO data; 2 single block inlinees; 1 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[System.Numerics.Vector`1[float]]>
 ;* V01 arg1         [V01    ] (  0,  0   )  struct (16) zero-ref    multireg-arg ld-addr-op single-def <System.Span`1[System.Numerics.Vector`1[float]]>
-;  V02 arg2         [V02,T12] (  2,  4   )  simd32  ->  [rbp+0x10]  single-def <System.Numerics.Vector`1[float]>
-;  V03 arg3         [V03,T13] (  2,  2   )  simd32  ->  [rbp+0x30]  single-def <System.Numerics.Vector`1[float]>
-;  V04 arg4         [V04,T02] (  6,  7.50)     ref  ->  rbx         class-hnd <System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]>
-;  V05 loc0         [V05,T09] (  7,  1.24)     int  ->  [rbp-0x2C] 
+;  V02 arg2         [V02,T12] (  1,  3.96)  simd32  ->  [rbp+0x10]  single-def <System.Numerics.Vector`1[float]>
+;  V03 arg3         [V03,T13] (  1,  1.98)  simd32  ->  [rbp+0x30]  single-def <System.Numerics.Vector`1[float]>
+;  V04 arg4         [V04,T02] (  5,  7.46)     ref  ->  rbx         class-hnd <System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]>
+;* V05 loc0         [V05,T10] (  0,  0   )     int  ->  zero-ref   
 ;  V06 OutArgs      [V06    ] (  1,  1   )  struct (64) [rsp+0x00]  do-not-enreg[XS] addr-exposed "OutgoingArgSpace" <UNNAMED>
 ;* V07 tmp1         [V07    ] (  0,  0   )   byref  ->  zero-ref    "non-inline candidate call"
 ;* V08 tmp2         [V08    ] (  0,  0   )   byref  ->  zero-ref   
-;  V09 tmp3         [V09,T04] (  4,  8   )   byref  ->  [rbp-0x38] 
-;  V10 tmp4         [V10,T11] (  6,  8   )  simd32  ->  mm0        
-;  V11 tmp5         [V11,T03] (  4,  6.98)   byref  ->  r15         single-def "field V00._reference (fldOffset=0x0)" P-INDEP
-;  V12 tmp6         [V12,T06] (  5,  3.04)     int  ->  r13         single-def "field V00._length (fldOffset=0x8)" P-INDEP
-;  V13 tmp7         [V13,T05] (  3,  5   )   byref  ->  r14         single-def "field V01._reference (fldOffset=0x0)" P-INDEP
-;  V14 tmp8         [V14,T08] (  3,  1.54)     int  ->  r12         single-def "field V01._length (fldOffset=0x8)" P-INDEP
-;  V15 cse0         [V15,T10] (  3,  0.10)   byref  ->  [rbp-0x40]  spill-single-def "CSE #05: conservative"
-;* V16 cse1         [V16,T07] (  0,  0   )    long  ->  zero-ref    "CSE #02: aggressive"
-;  V17 rat0         [V17,T00] (  6, 18.32)    long  ->  r12         "Strength reduced derived IV"
-;  V18 rat1         [V18,T01] (  4, 12.38)     int  ->  r13         "Trip count IV"
+;  V09 tmp3         [V09,T05] (  2,  7.92)   byref  ->  [rbp-0x30]  spill-single-def
+;  V10 tmp4         [V10,T11] (  3,  7.92)  simd32  ->  mm0        
+;* V11 tmp5         [V11    ] (  0,  0   )     int  ->  zero-ref    "Inlining Arg"
+;* V12 tmp6         [V12    ] (  0,  0   )  struct (16) zero-ref    ld-addr-op "NewObj constructor temp" <System.Span`1[System.Numerics.Vector`1[float]]>
+;* V13 tmp7         [V13    ] (  0,  0   )   byref  ->  zero-ref    "Inlining Arg"
+;  V14 tmp8         [V14,T03] (  3,  6.94)   byref  ->  r15         single-def "field V00._reference (fldOffset=0x0)" P-INDEP
+;  V15 tmp9         [V15,T07] (  4,  3.50)     int  ->  r12         single-def "field V00._length (fldOffset=0x8)" P-INDEP
+;  V16 tmp10        [V16,T04] (  4,  6.96)   byref  ->  r14         "field V01._reference (fldOffset=0x0)" P-INDEP
+;  V17 tmp11        [V17,T06] (  4,  4   )     int  ->  r13         "field V01._length (fldOffset=0x8)" P-INDEP
+;  V18 tmp12        [V18,T09] (  2,  2   )   byref  ->  r14         single-def "field V12._reference (fldOffset=0x0)" P-INDEP
+;* V19 tmp13        [V19    ] (  0,  0   )     int  ->  zero-ref    "field V12._length (fldOffset=0x8)" P-INDEP
+;* V20 cse0         [V20,T08] (  0,  0   )    long  ->  zero-ref    "CSE #02: aggressive"
+;  V21 rat0         [V21,T00] (  6, 18.32)    long  ->  r13         "Strength reduced derived IV"
+;  V22 rat1         [V22,T01] (  4, 12.38)     int  ->  r12         "Trip count IV"
 ;
-; Lcl frame size = 88
+; Lcl frame size = 72
 
 G_M45263_IG01:
        push     rbp
        push     r15
        push     r14
        push     r13
        push     r12
        push     rbx
-       sub      rsp, 88
-       lea      rbp, [rsp+0x80]
+       sub      rsp, 72
+       lea      rbp, [rsp+0x70]
        mov      r15, rdi
-       mov      r13d, esi
+       mov      r12d, esi
        mov      r14, rdx
-       mov      r12d, ecx
+       mov      r13d, ecx
        mov      rbx, r8
-						;; size=37 bbWeight=1 PerfScore 8.00
+						;; size=34 bbWeight=1 PerfScore 8.00
 G_M45263_IG02:
+       cmp      r12d, r13d
+       jg       G_M45263_IG12
+       mov      r13d, r12d
        test     rbx, rbx
        jne      SHORT G_M45263_IG05
-						;; size=5 bbWeight=1 PerfScore 1.25
+						;; size=17 bbWeight=1 PerfScore 2.75
 G_M45263_IG03:
        mov      rdi, 0xD1FFAB1E      ; global ptr
        test     byte  ptr [rdi], 1
-       je       G_M45263_IG16
+       je       G_M45263_IG13
 						;; size=19 bbWeight=0.50 PerfScore 2.12
 G_M45263_IG04:
        mov      rdi, 0xD1FFAB1E      ; data for System.Collections.Generic.EqualityComparer`1[System.Numerics.Vector`1[float]]:<Default>k__BackingField
        mov      rbx, gword ptr [rdi]
 						;; size=13 bbWeight=0.50 PerfScore 1.12
 G_M45263_IG05:
-       xor      eax, eax
        test     r13d, r13d
        jle      SHORT G_M45263_IG11
-						;; size=7 bbWeight=1 PerfScore 1.50
+						;; size=5 bbWeight=1 PerfScore 1.25
 G_M45263_IG06:
-       cmp      r13d, r12d
-       jg       G_M45263_IG14
-       xor      r12d, r12d
-						;; size=12 bbWeight=0.50 PerfScore 0.75
+       xor      r13d, r13d
+						;; size=3 bbWeight=0.50 PerfScore 0.12
 G_M45263_IG07:
-       lea      rcx, bword ptr [r14+r12]
-       mov      bword ptr [rbp-0x38], rcx
-       vmovups  ymm2, ymmword ptr [r15+r12]
+       lea      rax, bword ptr [r14+r13]
+       mov      bword ptr [rbp-0x30], rax
+       vmovups  ymm2, ymmword ptr [r15+r13]
        vmovups  ymmword ptr [rsp], ymm2
        vmovups  ymm0, ymmword ptr [rbp+0x10]
        vmovups  ymmword ptr [rsp+0x20], ymm0
        mov      rdi, rbx
        mov      r11, 0xD1FFAB1E      ; code for System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]:Equals(System.Numerics.Vector`1[float],System.Numerics.Vector`1[float]):bool:this
        call     [r11]System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]:Equals(System.Numerics.Vector`1[float],System.Numerics.Vector`1[float]):bool:this
        test     eax, eax
        je       SHORT G_M45263_IG09
 						;; size=50 bbWeight=3.96 PerfScore 68.31
 G_M45263_IG08:
        vmovups  ymm1, ymmword ptr [rbp+0x30]
        vmovaps  ymm0, ymm1
+       vmovups  ymmword ptr [rbp+0x30], ymm1
        jmp      SHORT G_M45263_IG10
-						;; size=11 bbWeight=1.98 PerfScore 12.38
+						;; size=16 bbWeight=1.98 PerfScore 14.36
 G_M45263_IG09:
-       vmovups  ymm0, ymmword ptr [r15+r12]
+       vmovups  ymm0, ymmword ptr [r15+r13]
 						;; size=6 bbWeight=1.98 PerfScore 9.90
 G_M45263_IG10:
-       mov      rax, bword ptr [rbp-0x38]
+       mov      rax, bword ptr [rbp-0x30]
        vmovups  ymmword ptr [rax], ymm0
-       add      r12, 32
-       dec      r13d
+       add      r13, 32
+       dec      r12d
        jne      SHORT G_M45263_IG07
 						;; size=17 bbWeight=3.96 PerfScore 17.82
 G_M45263_IG11:
        vzeroupper 
-       add      rsp, 88
+       add      rsp, 72
        pop      rbx
        pop      r12
        pop      r13
        pop      r14
        pop      r15
        pop      rbp
        ret      
 						;; size=18 bbWeight=1 PerfScore 5.25
 G_M45263_IG12:
-       mov      rdx, bword ptr [rbp-0x40]
-       vmovups  ymm0, ymmword ptr [rdx]
-       vmovups  ymm1, ymmword ptr [rbp+0x30]
-						;; size=13 bbWeight=0.02 PerfScore 0.20
+       mov      rax, 0xD1FFAB1E      ; code for System.ThrowHelper:ThrowArgumentOutOfRangeException()
+       call     [rax]System.ThrowHelper:ThrowArgumentOutOfRangeException()
+       int3     
+						;; size=13 bbWeight=0 PerfScore 0.00
 G_M45263_IG13:
-       mov      rcx, bword ptr [rbp-0x38]
-       vmovups  ymmword ptr [rcx], ymm0
-       mov      eax, dword ptr [rbp-0x2C]
-       inc      eax
-       cmp      eax, r13d
-       vmovups  ymmword ptr [rbp+0x30], ymm1
-       jge      SHORT G_M45263_IG11
-						;; size=23 bbWeight=0.04 PerfScore 0.26
-G_M45263_IG14:
-       cmp      eax, r12d
-       jae      SHORT G_M45263_IG17
-       mov      edi, eax
-       shl      rdi, 5
-       add      rdi, r14
-       mov      bword ptr [rbp-0x38], rdi
-       mov      dword ptr [rbp-0x2C], eax
-       mov      edx, eax
-       shl      rdx, 5
-       add      rdx, r15
-       mov      bword ptr [rbp-0x40], rdx
-       vmovups  ymm2, ymmword ptr [rdx]
-       vmovups  ymmword ptr [rsp], ymm2
-       vmovups  ymm0, ymmword ptr [rbp+0x10]
-       vmovups  ymmword ptr [rsp+0x20], ymm0
-       mov      rdi, rbx
-       mov      r11, 0xD1FFAB1E      ; function address
-       call     [r11]System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]:Equals(System.Numerics.Vector`1[float],System.Numerics.Vector`1[float]):bool:this
-       test     eax, eax
-       je       SHORT G_M45263_IG12
-						;; size=74 bbWeight=0.04 PerfScore 0.88
-G_M45263_IG15:
-       vmovups  ymm1, ymmword ptr [rbp+0x30]
-       vmovaps  ymm0, ymm1
-       jmp      SHORT G_M45263_IG13
-						;; size=11 bbWeight=0.02 PerfScore 0.13
-G_M45263_IG16:
        mov      rdi, 0xD1FFAB1E      ; System.Collections.Generic.EqualityComparer`1[System.Numerics.Vector`1[float]]
        mov      rax, 0xD1FFAB1E      ; code for CORINFO_HELP_GET_GCSTATIC_BASE
        call     [rax]CORINFO_HELP_GET_GCSTATIC_BASE
        jmp      G_M45263_IG04
 						;; size=27 bbWeight=0 PerfScore 0.00
-G_M45263_IG17:
-       call     CORINFO_HELP_RNGCHKFAIL
-       int3     
-						;; size=6 bbWeight=0 PerfScore 0.00
 
-; Total bytes of code 349, prolog size 22, PerfScore 129.87, instruction count 94, allocated bytes for code 349 (MethodHash=a8f14f30) for method System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float],System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]) (FullOpts)
+; Total bytes of code 238, prolog size 19, PerfScore 131.01, instruction count 63, allocated bytes for code 238 (MethodHash=a8f14f30) for method System.MemoryExtensions:<Replace>g__ReplaceComparer|176_1[System.Numerics.Vector`1[float]](System.ReadOnlySpan`1[System.Numerics.Vector`1[float]],System.Span`1[System.Numerics.Vector`1[float]],System.Numerics.Vector`1[float],System.Numerics.Vector`1[float],System.Collections.Generic.IEqualityComparer`1[System.Numerics.Vector`1[float]]) (FullOpts)
 ; ============================================================

Larger list of diffs: https://gist.github.com/MihuBot/e22845c5d975989af933af19fa63a51b

[MihuBot]

@xtqqczze