mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-10-29 23:12:55 +01:00
e3e67d4a0a
This changes the SelectionDAG scheduling preference to source order. Soon, the SelectionDAG scheduler can be bypassed saving a nice chunk of compile time. Performance differences that result from this change are often a consequence of register coalescing. The register coalescer is far from perfect. Bugs can be filed for deficiencies. On x86 SandyBridge/Haswell, the source order schedule is often preserved, particularly for small blocks. Register pressure is generally improved over the SD scheduler's ILP mode. However, we are still able to handle large blocks that require latency hiding, unlike the SD scheduler's BURR mode. MI scheduler also attempts to discover the critical path in single-block loops and adjust heuristics accordingly. The MI scheduler relies on the new machine model. This is currently unimplemented for AVX, so we may not be generating the best code yet. Unit tests are updated so they don't depend on SD scheduling heuristics. llvm-svn: 192750
27 lines
1.0 KiB
LLVM
27 lines
1.0 KiB
LLVM
; RUN: llc < %s -combiner-alias-analysis -march=x86-64 -mcpu=core2 | FileCheck %s
|
|
|
|
target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64"
|
|
target triple = "x86_64-apple-darwin10.4"
|
|
declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture, i8* nocapture, i64, i32, i1) nounwind
|
|
|
|
define fastcc i32 @cli_magic_scandesc(i8* %in) nounwind ssp {
|
|
entry:
|
|
%a = alloca [64 x i8]
|
|
%b = getelementptr inbounds [64 x i8]* %a, i64 0, i32 0
|
|
%c = getelementptr inbounds [64 x i8]* %a, i64 0, i32 30
|
|
%d = load i8* %b, align 8
|
|
%e = load i8* %c, align 8
|
|
%f = bitcast [64 x i8]* %a to i8*
|
|
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %f, i8* %in, i64 64, i32 8, i1 false) nounwind
|
|
store i8 %d, i8* %b, align 8
|
|
store i8 %e, i8* %c, align 8
|
|
ret i32 0
|
|
}
|
|
|
|
; CHECK: movq ___stack_chk_guard@GOTPCREL(%rip)
|
|
; CHECK: movb (%rsp), [[R1:%.+]]
|
|
; CHECK: movb 30(%rsp), [[R0:%.+]]
|
|
; CHECK: movb [[R1]], (%rsp)
|
|
; CHECK: movb [[R0]], 30(%rsp)
|
|
; CHECK: callq ___stack_chk_fail
|