RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-24 03:33:20 +01:00
Code Issues Projects Releases Wiki Activity

[SelectionDAG] try to convert funnel shift directly to rotate if legal

Browse Source 
If the DAGCombiner's rotate matching was working as expected, 
I don't think we'd see any test diffs here. 

This sidesteps the issue of custom lowering for rotates raised in PR38243:
https://bugs.llvm.org/show_bug.cgi?id=38243
...by only dealing with legal operations.

llvm-svn: 337966
This commit is contained in:
Sanjay Patel 2018-07-25 21:38:30 +00:00
parent cc01a9a147
commit a9aafee89e
5 changed files with 18 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
include/llvm/CodeGen/ISDOpcodes.h
View File

@ -377,6 +377,8 @@ namespace ISD {
/// When the 1st operand is a vector, the shift amount must be in the same
/// type. (TLI.getShiftAmountTy() will return the same type when the input
/// type is a vector.)
/// For rotates, the shift amount is treated as an unsigned amount modulo
/// the element size of the first operand.
SHL, SRA, SRL, ROTL, ROTR,
/// Byte Swap and Counting operators.

11
lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp
View File

@ -5672,7 +5672,16 @@ SelectionDAGBuilder::visitIntrinsicCall(const CallInst &I, unsigned Intrinsic) {
SDValue Z = getValue(I.getArgOperand(2));
EVT VT = X.getValueType();
// TODO: When X == Y, this is rotate. Create the node directly if legal.
// When X == Y, this is rotate. Create the node directly if legal.
// TODO: This should also be done if the operation is custom, but we have
// to make sure targets are handling the modulo shift amount as expected.
// TODO: If the rotate direction (left or right) corresponding to the shift
// is not available, adjust the shift value and invert the direction.
auto RotateOpcode = IsFSHL ? ISD::ROTL : ISD::ROTR;
if (X == Y && TLI.isOperationLegal(RotateOpcode, VT)) {
setValue(&I, DAG.getNode(RotateOpcode, sdl, VT, X, Z));
return nullptr;
}
// Get the shift amount and inverse shift amount, modulo the bit-width.
SDValue BitWidthC = DAG.getConstant(VT.getScalarSizeInBits(), sdl, VT);

6
test/CodeGen/AArch64/funnel-shift-rot.ll
View File

@ -163,11 +163,7 @@ define i32 @rotr_i32(i32 %x, i32 %z) {
define i64 @rotr_i64(i64 %x, i64 %z) {
; CHECK-LABEL: rotr_i64:
; CHECK: // %bb.0:
; CHECK-NEXT: orr w9, wzr, #0x40
; CHECK-NEXT: sub w9, w9, w1
; CHECK-NEXT: lsr x8, x0, x1
; CHECK-NEXT: lsl x9, x0, x9
; CHECK-NEXT: orr x0, x9, x8
; CHECK-NEXT: ror x0, x0, x1
; CHECK-NEXT: ret
%f = call i64 @llvm.fshr.i64(i64 %x, i64 %x, i64 %z)
ret i64 %f

4
test/CodeGen/PowerPC/funnel-shift-rot.ll
View File

@ -55,7 +55,6 @@ define i16 @rotl_i16(i16 %x, i16 %z) {
define i32 @rotl_i32(i32 %x, i32 %z) {
; CHECK-LABEL: rotl_i32:
; CHECK: # %bb.0:
; CHECK-NEXT: rlwinm 4, 4, 0, 27, 31
; CHECK-NEXT: rlwnm 3, 3, 4, 0, 31
; CHECK-NEXT: blr
%f = call i32 @llvm.fshl.i32(i32 %x, i32 %x, i32 %z)
@ -65,8 +64,7 @@ define i32 @rotl_i32(i32 %x, i32 %z) {
define i64 @rotl_i64(i64 %x, i64 %z) {
; CHECK-LABEL: rotl_i64:
; CHECK: # %bb.0:
; CHECK-NEXT: rlwinm 4, 4, 0, 26, 31
; CHECK-NEXT: rotld 3, 3, 4
; CHECK-NEXT: rldcl 3, 3, 4, 0
; CHECK-NEXT: blr
%f = call i64 @llvm.fshl.i64(i64 %x, i64 %x, i64 %z)
ret i64 %f

8
test/CodeGen/X86/funnel-shift-rot.ll
View File

@ -169,12 +169,12 @@ define i8 @rotr_i8_const_shift(i8 %x) nounwind {
; X32-SSE2-LABEL: rotr_i8_const_shift:
; X32-SSE2: # %bb.0:
; X32-SSE2-NEXT: movb {{[0-9]+}}(%esp), %al
; X32-SSE2-NEXT: rolb $5, %al
; X32-SSE2-NEXT: rorb $3, %al
; X32-SSE2-NEXT: retl
;
; X64-AVX2-LABEL: rotr_i8_const_shift:
; X64-AVX2: # %bb.0:
; X64-AVX2-NEXT: rolb $5, %dil
; X64-AVX2-NEXT: rorb $3, %dil
; X64-AVX2-NEXT: movl %edi, %eax
; X64-AVX2-NEXT: retq
%f = call i8 @llvm.fshr.i8(i8 %x, i8 %x, i8 3)
@ -185,12 +185,12 @@ define i32 @rotr_i32_const_shift(i32 %x) nounwind {
; X32-SSE2-LABEL: rotr_i32_const_shift:
; X32-SSE2: # %bb.0:
; X32-SSE2-NEXT: movl {{[0-9]+}}(%esp), %eax
; X32-SSE2-NEXT: roll $29, %eax
; X32-SSE2-NEXT: rorl $3, %eax
; X32-SSE2-NEXT: retl
;
; X64-AVX2-LABEL: rotr_i32_const_shift:
; X64-AVX2: # %bb.0:
; X64-AVX2-NEXT: roll $29, %edi
; X64-AVX2-NEXT: rorl $3, %edi
; X64-AVX2-NEXT: movl %edi, %eax
; X64-AVX2-NEXT: retq
%f = call i32 @llvm.fshr.i32(i32 %x, i32 %x, i32 3)