JIT: optRedundantBranches misses an opportunity

[EgorBo]

Extracted & simplified from #36649 (comment) to a separate issue

void PrintN(object o)
{
    if (o is int n)
        Console.WriteLine(n);
}

This simple method emits a lot of codegen:

; Method testout1:PrintN(System.Object):this
G_M58869_IG01:              ;; offset=0000H
       56                   push     rsi
       4883EC20             sub      rsp, 32
       488BF2               mov      rsi, rdx
G_M58869_IG02:              ;; offset=0008H
       488BD6               mov      rdx, rsi
       4885D2               test     rdx, rdx
       7442                 je       SHORT G_M58869_IG08
G_M58869_IG03:              ;; offset=0010H
       488B12               mov      rdx, qword ptr [rdx]
       48B9B86A32F4FA7F0000 mov      rcx, 0x7FFAF4326AB8      ; System.Int32
       483BD1               cmp      rdx, rcx
       7530                 jne      SHORT G_M58869_IG08
G_M58869_IG04:              ;; offset=0022H
       48B9B86A32F4FA7F0000 mov      rcx, 0x7FFAF4326AB8      ; System.Int32
       483BD1               cmp      rdx, rcx
       7413                 je       SHORT G_M58869_IG06
G_M58869_IG05:              ;; offset=0031H
       488BD6               mov      rdx, rsi
       48B9B86A32F4FA7F0000 mov      rcx, 0x7FFAF4326AB8      ; System.Int32
       FF15FC2BFDFF         call     [CORINFO_HELP_UNBOX]
G_M58869_IG06:              ;; offset=0044H
       8B4E08               mov      ecx, dword ptr [rsi+8]
G_M58869_IG07:              ;; offset=0047H
       4883C420             add      rsp, 32
       5E                   pop      rsi
       FF2586E73900         tail.jmp [System.Console:WriteLine(int)]
G_M58869_IG08:              ;; offset=0052H
       4883C420             add      rsp, 32
       5E                   pop      rsi
       C3                   ret      
; Total bytes of code: 88

From this flowgraph (built after "Redundant branch opts" phase) it's obvious that BB06 is dominated by BB02 with exactly the same condition which means BB06 can be folded (and BB07 becomes dead as well).

A not very TP-wise solution is basically run a sort of "UpdateFlowGraph" phase to drop unreachable blocks and recompute doms and then run optRedundantBranches again

cc @AndyAyersMS

[ghost]

Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch
See info in area-owners.md if you want to be subscribed.

Issue Details

---

Extracted & simplified from #36649 (comment) to a separate issue

void PrintN(object o)
{
    if (o is int n)
        Console.WriteLine(n);
}

This simple method emits a lot of codegen:

; Method testout1:PrintN(System.Object):this
G_M58869_IG01:              ;; offset=0000H
       56                   push     rsi
       4883EC20             sub      rsp, 32
       488BF2               mov      rsi, rdx
G_M58869_IG02:              ;; offset=0008H
       488BD6               mov      rdx, rsi
       4885D2               test     rdx, rdx
       7442                 je       SHORT G_M58869_IG08
G_M58869_IG03:              ;; offset=0010H
       488B12               mov      rdx, qword ptr [rdx]
       48B9B86A32F4FA7F0000 mov      rcx, 0x7FFAF4326AB8      ; System.Int32
       483BD1               cmp      rdx, rcx
       7530                 jne      SHORT G_M58869_IG08
G_M58869_IG04:              ;; offset=0022H
       48B9B86A32F4FA7F0000 mov      rcx, 0x7FFAF4326AB8      ; System.Int32
       483BD1               cmp      rdx, rcx
       7413                 je       SHORT G_M58869_IG06
G_M58869_IG05:              ;; offset=0031H
       488BD6               mov      rdx, rsi
       48B9B86A32F4FA7F0000 mov      rcx, 0x7FFAF4326AB8      ; System.Int32
       FF15FC2BFDFF         call     [CORINFO_HELP_UNBOX]
G_M58869_IG06:              ;; offset=0044H
       8B4E08               mov      ecx, dword ptr [rsi+8]
G_M58869_IG07:              ;; offset=0047H
       4883C420             add      rsp, 32
       5E                   pop      rsi
       FF2586E73900         tail.jmp [System.Console:WriteLine(int)]
G_M58869_IG08:              ;; offset=0052H
       4883C420             add      rsp, 32
       5E                   pop      rsi
       C3                   ret      
; Total bytes of code: 88

From this flowgraph (built after "Redundant branch opts" phase) it's obvious that BB06 is dominated by BB02 with exactly the same condition which means BB06 can be folded.

A not very TP-wise solution is basically run a sort of "UpdateFlowGraph" phase to drop unreachable blocks and recompute doms and then run optRedundantBranches again

cc @AndyAyersMS

Author:	EgorBo
Assignees:	-
Labels:	tenet-performance, area-CodeGen-coreclr
Milestone:	-

[Wraith2]

Isn't IG04 reachable from exiting IG03 with an equality on the comparison?

[EgorBo]

Isn't IG04 reachable from exiting IG03 with an equality on the comparison?

BB03 is BBJ_ALWAYS to BB09, so no

[AndyAyersMS]

There is a bit of precedent in the JIT for updating dominator tree links as we optimize, so it's possible that we could look into something like that. But even with this, RBO might not visit the branches in the right order, so might need the ability to revisit cases that were already looked at...