mirror of
https://github.com/RPCS3/llvm-mirror.git
synced 2024-11-24 19:52:54 +01:00
6c1440cf27
Original commit message: Defer some shl transforms to DAGCombine. The shl instruction is used to represent multiplication by a constant power of two as well as bitwise left shifts. Some InstCombine transformations would turn an shl instruction into a bit mask operation, making it difficult for later analysis passes to recognize the constsnt multiplication. Disable those shl transformations, deferring them to DAGCombine time. An 'shl X, C' instruction is now treated mostly the same was as 'mul X, C'. These transformations are deferred: (X >>? C) << C --> X & (-1 << C) (When X >> C has multiple uses) (X >>? C1) << C2 --> X << (C2-C1) & (-1 << C2) (When C2 > C1) (X >>? C1) << C2 --> X >>? (C1-C2) & (-1 << C2) (When C1 > C2) The corresponding exact transformations are preserved, just like div-exact + mul: (X >>?,exact C) << C --> X (X >>?,exact C1) << C2 --> X << (C2-C1) (X >>?,exact C1) << C2 --> X >>?,exact (C1-C2) The disabled transformations could also prevent the instruction selector from recognizing rotate patterns in hash functions and cryptographic primitives. I have a test case for that, but it is too fragile. llvm-svn: 155362
65 lines
2.3 KiB
LLVM
65 lines
2.3 KiB
LLVM
; RUN: opt -O3 -S -analyze -scalar-evolution %s | FileCheck %s
|
|
;
|
|
; This file contains phase ordering tests for scalar evolution.
|
|
; Test that the standard passes don't obfuscate the IR so scalar evolution can't
|
|
; recognize expressions.
|
|
|
|
; CHECK: test1
|
|
; The loop body contains two increments by %div.
|
|
; Make sure that 2*%div is recognizable, and not expressed as a bit mask of %d.
|
|
; CHECK: --> {%p,+,(2 * (%d /u 4) * sizeof(i32))}
|
|
define void @test1(i64 %d, i32* %p) nounwind uwtable ssp {
|
|
entry:
|
|
%div = udiv i64 %d, 4
|
|
br label %for.cond
|
|
|
|
for.cond: ; preds = %for.inc, %entry
|
|
%p.addr.0 = phi i32* [ %p, %entry ], [ %add.ptr1, %for.inc ]
|
|
%i.0 = phi i32 [ 0, %entry ], [ %inc, %for.inc ]
|
|
%cmp = icmp ne i32 %i.0, 64
|
|
br i1 %cmp, label %for.body, label %for.end
|
|
|
|
for.body: ; preds = %for.cond
|
|
store i32 0, i32* %p.addr.0, align 4
|
|
%add.ptr = getelementptr inbounds i32* %p.addr.0, i64 %div
|
|
store i32 1, i32* %add.ptr, align 4
|
|
%add.ptr1 = getelementptr inbounds i32* %add.ptr, i64 %div
|
|
br label %for.inc
|
|
|
|
for.inc: ; preds = %for.body
|
|
%inc = add i32 %i.0, 1
|
|
br label %for.cond
|
|
|
|
for.end: ; preds = %for.cond
|
|
ret void
|
|
}
|
|
|
|
; CHECK: test1a
|
|
; Same thing as test1, but it is even more tempting to fold 2 * (%d /u 2)
|
|
; CHECK: --> {%p,+,(2 * (%d /u 2) * sizeof(i32))}
|
|
define void @test1a(i64 %d, i32* %p) nounwind uwtable ssp {
|
|
entry:
|
|
%div = udiv i64 %d, 2
|
|
br label %for.cond
|
|
|
|
for.cond: ; preds = %for.inc, %entry
|
|
%p.addr.0 = phi i32* [ %p, %entry ], [ %add.ptr1, %for.inc ]
|
|
%i.0 = phi i32 [ 0, %entry ], [ %inc, %for.inc ]
|
|
%cmp = icmp ne i32 %i.0, 64
|
|
br i1 %cmp, label %for.body, label %for.end
|
|
|
|
for.body: ; preds = %for.cond
|
|
store i32 0, i32* %p.addr.0, align 4
|
|
%add.ptr = getelementptr inbounds i32* %p.addr.0, i64 %div
|
|
store i32 1, i32* %add.ptr, align 4
|
|
%add.ptr1 = getelementptr inbounds i32* %add.ptr, i64 %div
|
|
br label %for.inc
|
|
|
|
for.inc: ; preds = %for.body
|
|
%inc = add i32 %i.0, 1
|
|
br label %for.cond
|
|
|
|
for.end: ; preds = %for.cond
|
|
ret void
|
|
}
|