From 65eea557aee10ddf0dc35be069597f34554de308 Mon Sep 17 00:00:00 2001
From: Renato Golin <renato.golin@linaro.org>
Date: Tue, 10 Jun 2014 16:39:21 +0000
Subject: [PATCH] Fix a bug in the Thumb1 ARM Load/Store optimizer

Previously, the basic block was searched for future uses of the base register,
and if necessary any writeback to the base register was reset using a SUB
instruction (e.g. before calling a function) just before such a use. However,
this step happened *before* the merged LDM/STM instruction was built. So if
there was (e.g.) a function call directly after the not-yet-formed LDM/STM,
the pass would first insert a SUB instruction to reset the base register,
and then (at the same location, incorrectly) insert the LDM/STM itself.

This patch fixes PR19972. Patch by Moritz Roth.

llvm-svn: 210542
---
 llvm/lib/Target/ARM/ARMLoadStoreOptimizer.cpp           | 14 +++++++-------
 .../CodeGen/Thumb/2014-06-10-thumb1-ldst-opt-bug.ll     | 17 +++++++++++++++++
 2 files changed, 24 insertions(+), 7 deletions(-)
 create mode 100644 llvm/test/CodeGen/Thumb/2014-06-10-thumb1-ldst-opt-bug.ll

diff --git a/llvm/lib/Target/ARM/ARMLoadStoreOptimizer.cpp b/llvm/lib/Target/ARM/ARMLoadStoreOptimizer.cpp
index ee7df54..62fd059 100644
--- a/llvm/lib/Target/ARM/ARMLoadStoreOptimizer.cpp
+++ b/llvm/lib/Target/ARM/ARMLoadStoreOptimizer.cpp
@@ -505,7 +505,7 @@ ARMLoadStoreOpt::MergeOps(MachineBasicBlock &MBB,
 
   // Exception: If the base register is in the input reglist, Thumb1 LDM is
   // non-writeback. Check for this.
-  if (Opcode == ARM::tLDRi && isThumb1)
+  if (Opcode == ARM::tLDMIA && isThumb1)
     for (unsigned I = 0; I < NumRegs; ++I)
       if (Base == Regs[I].first) {
         Writeback = false;
@@ -519,17 +519,17 @@ ARMLoadStoreOpt::MergeOps(MachineBasicBlock &MBB,
       // Update tLDMIA with writeback if necessary.
       Opcode = ARM::tLDMIA_UPD;
 
-    // The base isn't dead after a merged instruction with writeback. Update
-    // future uses of the base with the added offset (if possible), or reset
-    // the base register as necessary.
-    if (!BaseKill)
-      UpdateBaseRegUses(MBB, MBBI, dl, Base, NumRegs, Pred, PredReg);
-
     MIB = BuildMI(MBB, MBBI, dl, TII->get(Opcode));
 
     // Thumb1: we might need to set base writeback when building the MI.
     MIB.addReg(Base, getDefRegState(true))
        .addReg(Base, getKillRegState(BaseKill));
+
+    // The base isn't dead after a merged instruction with writeback. Update
+    // future uses of the base with the added offset (if possible), or reset
+    // the base register as necessary.
+    if (!BaseKill)
+      UpdateBaseRegUses(MBB, MBBI, dl, Base, NumRegs, Pred, PredReg);
   } else {
     // No writeback, simply build the MachineInstr.
     MIB = BuildMI(MBB, MBBI, dl, TII->get(Opcode));
diff --git a/llvm/test/CodeGen/Thumb/2014-06-10-thumb1-ldst-opt-bug.ll b/llvm/test/CodeGen/Thumb/2014-06-10-thumb1-ldst-opt-bug.ll
new file mode 100644
index 0000000..3c43383
--- /dev/null
+++ b/llvm/test/CodeGen/Thumb/2014-06-10-thumb1-ldst-opt-bug.ll
@@ -0,0 +1,17 @@
+; RUN: llc < %s -mtriple=thumbv6m-eabi -o - | FileCheck %s
+
+define void @foo(i32* %A) #0 {
+entry:
+; CHECK-LABEL: foo:
+; CHECK: push {r7, lr}
+; CHECK: ldm [[REG0:r[0-9]]]!,
+; CHECK-NEXT: subs [[REG0]]
+; CHECK-NEXT: bl
+  %0 = load i32* %A, align 4
+  %arrayidx1 = getelementptr inbounds i32* %A, i32 1
+  %1 = load i32* %arrayidx1, align 4
+  tail call void @bar(i32* %A, i32 %0, i32 %1) #2
+  ret void
+}
+
+declare void @bar(i32*, i32, i32) #1
-- 
2.7.4