mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-10-19 19:12:56 +02:00
02f7297367
Utilizing the 8 and 16 bit comparison instructions, even when an input can be folded into the comparison instruction itself, is typically not worth it. There are too many partial register stalls as a result, leading to significant slowdowns. By always performing comparisons on at least 32-bit registers, performance of the calculation chain leading to the comparison improves. Continue to use the smaller comparisons when minimizing size, as that allows better folding of loads into the comparison instructions. rdar://15386341 llvm-svn: 195496
117 lines
3.6 KiB
LLVM
117 lines
3.6 KiB
LLVM
; RUN: llc < %s -mtriple=x86_64-linux | FileCheck %s
|
|
; RUN: llc < %s -disable-simplify-libcalls -mtriple=x86_64-linux | FileCheck %s --check-prefix=NOBUILTIN
|
|
; RUN: llc < %s -mtriple=x86_64-win32 | FileCheck %s
|
|
|
|
; This tests codegen time inlining/optimization of memcmp
|
|
; rdar://6480398
|
|
|
|
@.str = private constant [23 x i8] c"fooooooooooooooooooooo\00", align 1 ; <[23 x i8]*> [#uses=1]
|
|
|
|
declare i32 @memcmp(...)
|
|
|
|
define void @memcmp2(i8* %X, i8* %Y, i32* nocapture %P) nounwind {
|
|
entry:
|
|
%0 = tail call i32 (...)* @memcmp(i8* %X, i8* %Y, i32 2) nounwind ; <i32> [#uses=1]
|
|
%1 = icmp eq i32 %0, 0 ; <i1> [#uses=1]
|
|
br i1 %1, label %return, label %bb
|
|
|
|
bb: ; preds = %entry
|
|
store i32 4, i32* %P, align 4
|
|
ret void
|
|
|
|
return: ; preds = %entry
|
|
ret void
|
|
; CHECK-LABEL: memcmp2:
|
|
; CHECK: movzwl
|
|
; CHECK-NEXT: movzwl
|
|
; CHECK-NEXT: cmpl
|
|
; NOBUILTIN-LABEL: memcmp2:
|
|
; NOBUILTIN: callq
|
|
}
|
|
|
|
define void @memcmp2a(i8* %X, i32* nocapture %P) nounwind {
|
|
entry:
|
|
%0 = tail call i32 (...)* @memcmp(i8* %X, i8* getelementptr inbounds ([23 x i8]* @.str, i32 0, i32 1), i32 2) nounwind ; <i32> [#uses=1]
|
|
%1 = icmp eq i32 %0, 0 ; <i1> [#uses=1]
|
|
br i1 %1, label %return, label %bb
|
|
|
|
bb: ; preds = %entry
|
|
store i32 4, i32* %P, align 4
|
|
ret void
|
|
|
|
return: ; preds = %entry
|
|
ret void
|
|
; CHECK-LABEL: memcmp2a:
|
|
; CHECK: movzwl
|
|
; CHECK-NEXT: cmpl $28527,
|
|
}
|
|
|
|
|
|
define void @memcmp4(i8* %X, i8* %Y, i32* nocapture %P) nounwind {
|
|
entry:
|
|
%0 = tail call i32 (...)* @memcmp(i8* %X, i8* %Y, i32 4) nounwind ; <i32> [#uses=1]
|
|
%1 = icmp eq i32 %0, 0 ; <i1> [#uses=1]
|
|
br i1 %1, label %return, label %bb
|
|
|
|
bb: ; preds = %entry
|
|
store i32 4, i32* %P, align 4
|
|
ret void
|
|
|
|
return: ; preds = %entry
|
|
ret void
|
|
; CHECK-LABEL: memcmp4:
|
|
; CHECK: movl
|
|
; CHECK-NEXT: cmpl
|
|
}
|
|
|
|
define void @memcmp4a(i8* %X, i32* nocapture %P) nounwind {
|
|
entry:
|
|
%0 = tail call i32 (...)* @memcmp(i8* %X, i8* getelementptr inbounds ([23 x i8]* @.str, i32 0, i32 1), i32 4) nounwind ; <i32> [#uses=1]
|
|
%1 = icmp eq i32 %0, 0 ; <i1> [#uses=1]
|
|
br i1 %1, label %return, label %bb
|
|
|
|
bb: ; preds = %entry
|
|
store i32 4, i32* %P, align 4
|
|
ret void
|
|
|
|
return: ; preds = %entry
|
|
ret void
|
|
; CHECK-LABEL: memcmp4a:
|
|
; CHECK: cmpl $1869573999,
|
|
}
|
|
|
|
define void @memcmp8(i8* %X, i8* %Y, i32* nocapture %P) nounwind {
|
|
entry:
|
|
%0 = tail call i32 (...)* @memcmp(i8* %X, i8* %Y, i32 8) nounwind ; <i32> [#uses=1]
|
|
%1 = icmp eq i32 %0, 0 ; <i1> [#uses=1]
|
|
br i1 %1, label %return, label %bb
|
|
|
|
bb: ; preds = %entry
|
|
store i32 4, i32* %P, align 4
|
|
ret void
|
|
|
|
return: ; preds = %entry
|
|
ret void
|
|
; CHECK-LABEL: memcmp8:
|
|
; CHECK: movq
|
|
; CHECK: cmpq
|
|
}
|
|
|
|
define void @memcmp8a(i8* %X, i32* nocapture %P) nounwind {
|
|
entry:
|
|
%0 = tail call i32 (...)* @memcmp(i8* %X, i8* getelementptr inbounds ([23 x i8]* @.str, i32 0, i32 0), i32 8) nounwind ; <i32> [#uses=1]
|
|
%1 = icmp eq i32 %0, 0 ; <i1> [#uses=1]
|
|
br i1 %1, label %return, label %bb
|
|
|
|
bb: ; preds = %entry
|
|
store i32 4, i32* %P, align 4
|
|
ret void
|
|
|
|
return: ; preds = %entry
|
|
ret void
|
|
; CHECK-LABEL: memcmp8a:
|
|
; CHECK: movabsq $8029759185026510694,
|
|
; CHECK: cmpq
|
|
}
|
|
|