From edcaae782eddc1b820f5e58d5405c541665c87be Mon Sep 17 00:00:00 2001
From: Vasileios Kalintiris <Vasileios.Kalintiris@imgtec.com>
Date: Tue, 4 Aug 2015 14:26:35 +0000
Subject: [PATCH] Revert r229675 - [mips] Avoid redundant sign extension of the
 result of binary bitwise instructions.

It introduced two regressions on 64-bit big-endian targets running under N32
(MultiSource/Benchmarks/tramp3d-v4/tramp3d-v4, and
MultiSource/Applications/kimwitu++/kc) The issue is that on 64-bit targets
comparisons such as BEQ compare the whole GPR64 but incorrectly tell the
instruction selector that they operate on GPR32's. This leads to the
elimination of i32->i64 extensions that are actually required by
comparisons to work correctly.

There's currently a patch under review that fixes this problem.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@243984 91177308-0d34-0410-b5e6-96231b3b80d8
---
 lib/Target/Mips/Mips64InstrInfo.td   | 8 --------
 test/CodeGen/Mips/delay-slot-kill.ll | 2 ++
 test/CodeGen/Mips/llvm-ir/and.ll     | 5 ++++-
 test/CodeGen/Mips/llvm-ir/or.ll      | 6 +++++-
 test/CodeGen/Mips/llvm-ir/xor.ll     | 5 ++++-
 5 files changed, 15 insertions(+), 11 deletions(-)

diff --git a/lib/Target/Mips/Mips64InstrInfo.td b/lib/Target/Mips/Mips64InstrInfo.td
index c37cf95cadc..f917ecad4a5 100644
--- a/lib/Target/Mips/Mips64InstrInfo.td
+++ b/lib/Target/Mips/Mips64InstrInfo.td
@@ -500,14 +500,6 @@ def : MipsPat<(trunc (assertzext GPR64:$src)),
 def : MipsPat<(i32 (trunc GPR64:$src)),
               (SLL (EXTRACT_SUBREG GPR64:$src, sub_32), 0)>;
 
-// Bypass trunc nodes for bitwise ops.
-def : MipsPat<(i32 (trunc (and GPR64:$lhs, GPR64:$rhs))),
-              (EXTRACT_SUBREG (AND64 GPR64:$lhs, GPR64:$rhs), sub_32)>;
-def : MipsPat<(i32 (trunc (or GPR64:$lhs, GPR64:$rhs))),
-              (EXTRACT_SUBREG (OR64 GPR64:$lhs, GPR64:$rhs), sub_32)>;
-def : MipsPat<(i32 (trunc (xor GPR64:$lhs, GPR64:$rhs))),
-              (EXTRACT_SUBREG (XOR64 GPR64:$lhs, GPR64:$rhs), sub_32)>;
-
 // variable shift instructions patterns
 def : MipsPat<(shl GPR64:$rt, (i32 (trunc GPR64:$rs))),
               (DSLLV GPR64:$rt, (EXTRACT_SUBREG GPR64:$rs, sub_32))>;
diff --git a/test/CodeGen/Mips/delay-slot-kill.ll b/test/CodeGen/Mips/delay-slot-kill.ll
index 57b630303c2..5e301441fd2 100644
--- a/test/CodeGen/Mips/delay-slot-kill.ll
+++ b/test/CodeGen/Mips/delay-slot-kill.ll
@@ -1,4 +1,6 @@
 ; RUN: llc < %s -march=mips64 -mcpu=mips3 | FileCheck %s
+; We have to XFAIL this temporarily because of the reversion of r229675.
+; XFAIL: *
 
 ; Currently, the following IR assembly generates a KILL instruction between
 ; the bitwise-and instruction and the return instruction. We verify that the
diff --git a/test/CodeGen/Mips/llvm-ir/and.ll b/test/CodeGen/Mips/llvm-ir/and.ll
index 8ebcfe4a3f6..c4121701ec1 100644
--- a/test/CodeGen/Mips/llvm-ir/and.ll
+++ b/test/CodeGen/Mips/llvm-ir/and.ll
@@ -59,7 +59,10 @@ define signext i32 @and_i32(i32 signext %a, i32 signext %b) {
 entry:
 ; ALL-LABEL: and_i32:
 
-  ; ALL:          and     $2, $4, $5
+  ; GP32:         and     $2, $4, $5
+
+  ; GP64:         and     $[[T0:[0-9]+]], $4, $5
+  ; GP64:         sll     $2, $[[T0]], 0
 
   %r = and i32 %a, %b
   ret i32 %r
diff --git a/test/CodeGen/Mips/llvm-ir/or.ll b/test/CodeGen/Mips/llvm-ir/or.ll
index 6215e403632..8509d6ce93f 100644
--- a/test/CodeGen/Mips/llvm-ir/or.ll
+++ b/test/CodeGen/Mips/llvm-ir/or.ll
@@ -59,7 +59,11 @@ define signext i32 @or_i32(i32 signext %a, i32 signext %b) {
 entry:
 ; ALL-LABEL: or_i32:
 
-  ; ALL:          or     $2, $4, $5
+  ; GP32:         or     $2, $4, $5
+
+  ; GP64:         or     $[[T0:[0-9]+]], $4, $5
+  ; FIXME: The sll instruction below is redundant.
+  ; GP64:         sll     $2, $[[T0]], 0
 
   %r = or i32 %a, %b
   ret i32 %r
diff --git a/test/CodeGen/Mips/llvm-ir/xor.ll b/test/CodeGen/Mips/llvm-ir/xor.ll
index 89af99981a3..d3cc5748489 100644
--- a/test/CodeGen/Mips/llvm-ir/xor.ll
+++ b/test/CodeGen/Mips/llvm-ir/xor.ll
@@ -59,7 +59,10 @@ define signext i32 @xor_i32(i32 signext %a, i32 signext %b) {
 entry:
 ; ALL-LABEL: xor_i32:
 
-  ; ALL:          xor     $2, $4, $5
+  ; GP32:         xor     $2, $4, $5
+
+  ; GP64:         xor     $[[T0:[0-9]+]], $4, $5
+  ; GP64:         sll     $2, $[[T0]], 0
 
   %r = xor i32 %a, %b
   ret i32 %r
-- 
2.34.1