Improve and enable folding of conditional branches with tail calls. 1. Make it so that conditional tail calls can be emitted even when there are multiple predecessors. 2. Don't guard the transformation behind -Os. The rationale for guarding it was static-prediction can be affected by whether the branch is forward of backward. This is no longer true for almost any X86 cpus (anything newer than `SnB`) so is no longer a meaningful concern. Reviewed By: pengfei Differential Revision: https://reviews.llvm.org/D140931
149 lines
3.6 KiB
LLVM
149 lines
3.6 KiB
LLVM
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
|
|
; RUN: llc < %s -O1 -mcpu=atom -mtriple=i686-linux | FileCheck %s
|
|
|
|
declare void @external_function(...)
|
|
|
|
define i32 @test_return_val(i32 %a) nounwind {
|
|
; CHECK-LABEL: test_return_val:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: retl
|
|
ret i32 %a
|
|
}
|
|
|
|
define i32 @test_optsize(i32 %a) nounwind optsize {
|
|
; CHECK-LABEL: test_optsize:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: retl
|
|
ret i32 %a
|
|
}
|
|
|
|
define i32 @test_minsize(i32 %a) nounwind minsize {
|
|
; CHECK-LABEL: test_minsize:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: retl
|
|
ret i32 %a
|
|
}
|
|
|
|
define i32 @test_pgso(i32 %a) nounwind !prof !14 {
|
|
; CHECK-LABEL: test_pgso:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: retl
|
|
ret i32 %a
|
|
}
|
|
|
|
define i32 @test_add(i32 %a, i32 %b) nounwind {
|
|
; CHECK-LABEL: test_add:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: addl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: retl
|
|
%result = add i32 %a, %b
|
|
ret i32 %result
|
|
}
|
|
|
|
define i32 @test_multiple_ret(i32 %a, i32 %b, i1 %c) nounwind {
|
|
; CHECK-LABEL: test_multiple_ret:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: testb $1, {{[0-9]+}}(%esp)
|
|
; CHECK-NEXT: je .LBB5_2
|
|
; CHECK-NEXT: # %bb.1: # %bb1
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: retl
|
|
; CHECK-NEXT: .LBB5_2: # %bb2
|
|
; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: retl
|
|
|
|
|
|
|
|
br i1 %c, label %bb1, label %bb2
|
|
|
|
bb1:
|
|
ret i32 %a
|
|
|
|
bb2:
|
|
ret i32 %b
|
|
}
|
|
|
|
define void @test_call_others(i32 %x) nounwind {
|
|
; CHECK-LABEL: test_call_others:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: cmpl $0, {{[0-9]+}}(%esp)
|
|
; CHECK-NEXT: jne external_function@PLT # TAILCALL
|
|
; CHECK-NEXT: # %bb.1: # %if.end
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: retl
|
|
%tobool = icmp eq i32 %x, 0
|
|
br i1 %tobool, label %if.end, label %true.case
|
|
|
|
true.case:
|
|
tail call void @external_function() nounwind
|
|
br label %if.end
|
|
|
|
if.end:
|
|
ret void
|
|
|
|
}
|
|
|
|
define void @test_branch_to_same_bb(i32 %x, i32 %y) nounwind {
|
|
; CHECK-LABEL: test_branch_to_same_bb:
|
|
; CHECK: # %bb.0:
|
|
; CHECK-NEXT: cmpl $0, {{[0-9]+}}(%esp)
|
|
; CHECK-NEXT: jle .LBB7_2
|
|
; CHECK-NEXT: .p2align 4, 0x90
|
|
; CHECK-NEXT: .LBB7_1: # %while.cond
|
|
; CHECK-NEXT: # =>This Inner Loop Header: Depth=1
|
|
; CHECK-NEXT: jmp .LBB7_1
|
|
; CHECK-NEXT: .LBB7_2: # %while.end
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: nop
|
|
; CHECK-NEXT: retl
|
|
%cmp = icmp sgt i32 %x, 0
|
|
br i1 %cmp, label %while.cond, label %while.end
|
|
|
|
while.cond:
|
|
br label %while.cond
|
|
|
|
while.end:
|
|
ret void
|
|
}
|
|
|
|
!llvm.module.flags = !{!0}
|
|
!0 = !{i32 1, !"ProfileSummary", !1}
|
|
!1 = !{!2, !3, !4, !5, !6, !7, !8, !9}
|
|
!2 = !{!"ProfileFormat", !"InstrProf"}
|
|
!3 = !{!"TotalCount", i64 10000}
|
|
!4 = !{!"MaxCount", i64 10}
|
|
!5 = !{!"MaxInternalCount", i64 1}
|
|
!6 = !{!"MaxFunctionCount", i64 1000}
|
|
!7 = !{!"NumCounts", i64 3}
|
|
!8 = !{!"NumFunctions", i64 3}
|
|
!9 = !{!"DetailedSummary", !10}
|
|
!10 = !{!11, !12, !13}
|
|
!11 = !{i32 10000, i64 100, i32 1}
|
|
!12 = !{i32 999000, i64 100, i32 1}
|
|
!13 = !{i32 999999, i64 1, i32 2}
|
|
!14 = !{!"function_entry_count", i64 0}
|