[GlobalMerge] Fix GlobalMerge on bss external global variables.

[huihuiz]

Summary

Global variables that are external and zero initialized are
supposed to be merged with global variables in the bss section
rather than the data section.

[efriedma]

LGTM

llvm/trunk/lib/CodeGen/GlobalMerge.cpp

for (auto &GV : M.globals()) {
 
     // Ignore all "required" globals:
     if (isMustKeepGlobalVariable(&GV))
       continue;
 
     Type *Ty = GV.getValueType();
     if (DL.getTypeAllocSize(Ty) < MaxOffset) {
       if (TM &&
-          TargetLoweringObjectFile::getKindForGlobal(&GV, *TM).isBSSLocal())
+          TargetLoweringObjectFile::getKindForGlobal(&GV, *TM).isBSS())
         BSSGlobals[{AddressSpace, Section}].push_back(&GV);
       else if (GV.isConstant())
         ConstGlobals[{AddressSpace, Section}].push_back(&GV);
       else
         Globals[{AddressSpace, Section}].push_back(&GV);
     }
   }
 

llvm/trunk/test/CodeGen/AArch64/global-merge-3.ll

 ; RUN: llc %s -mtriple=aarch64-none-linux-gnu -aarch64-enable-global-merge -global-merge-on-external -disable-post-ra -o - | FileCheck %s
 ; RUN: llc %s -mtriple=aarch64-linux-gnuabi -aarch64-enable-global-merge -global-merge-on-external -disable-post-ra -o - | FileCheck %s
 ; RUN: llc %s -mtriple=aarch64-apple-ios -aarch64-enable-global-merge -global-merge-on-external -disable-post-ra -o - | FileCheck %s --check-prefix=CHECK-APPLE-IOS
 
-@x = global [1000 x i32] zeroinitializer, align 1
-@y = global [1000 x i32] zeroinitializer, align 1
+@x = global [100 x i32] zeroinitializer, align 1
+@y = global [100 x i32] zeroinitializer, align 1
 @z = internal global i32 1, align 4
 
 define void @f1(i32 %a1, i32 %a2, i32 %a3) {
-;CHECK-APPLE-IOS: adrp  x8, __MergedGlobals_x@PAGE
+;CHECK-APPLE-IOS: adrp    x8, __MergedGlobals_x@PAGE+12
 ;CHECK-APPLE-IOS-NOT: adrp
-;CHECK-APPLE-IOS: add x8, x8, __MergedGlobals_x@PAGEOFF
-;CHECK-APPLE-IOS: adrp  x9, _y@PAGE+12
-;CHECK-APPLE-IOS: str w1, [x9]
-  %x3 = getelementptr inbounds [1000 x i32], [1000 x i32]* @x, i32 0, i64 3
-  %y3 = getelementptr inbounds [1000 x i32], [1000 x i32]* @y, i32 0, i64 3
+;CHECK-APPLE-IOS: add   x8, x8, __MergedGlobals_x@PAGEOFF+12
+;CHECK-APPLE-IOS: adrp    x9, _z@PAGE
+;CHECK-APPLE-IOS: str   w0, [x8]
+;CHECK-APPLE-IOS: str   w1, [x8, #400]
+;CHECK-APPLE-IOS: str     w2, [x9, _z@PAGEOFF]
+;CHECK: adrp    x8, .L_MergedGlobals+12
+;CHECK: add     x8, x8, :lo12:.L_MergedGlobals+12
+;CHECK: adrp    x9, z
+;CHECK: str     w0, [x8]
+;CHECK: str     w1, [x8, #400]
+;CHECK: str     w2, [x9, :lo12:z]
+  %x3 = getelementptr inbounds [100 x i32], [100 x i32]* @x, i32 0, i64 3
+  %y3 = getelementptr inbounds [100 x i32], [100 x i32]* @y, i32 0, i64 3
   store i32 %a1, i32* %x3, align 4
   store i32 %a2, i32* %y3, align 4
   store i32 %a3, i32* @z, align 4
   ret void
 }
 
-;CHECK: .type .L_MergedGlobals,@object // @_MergedGlobals
-;CHECK: .p2align  2
-;CHECK: .L_MergedGlobals:
-;CHECK: .size .L_MergedGlobals, 4004
-
-;CHECK-APPLE-IOS: .zerofill __DATA,__common,_y,4000,2
-
-;CHECK-APPLE-IOS: .p2align  2
-;CHECK-APPLE-IOS:  __MergedGlobals_x:
-;CHECK-APPLE-IOS: .long 1
-;CHECK-APPLE-IOS: .space  4000
+;CHECK-APPLE-IOS: .globl  __MergedGlobals_x
+;CHECK-APPLE-IOS: .zerofill __DATA,__common,__MergedGlobals_x,800,2
+;CHECK-APPLE-IOS: .set _x, __MergedGlobals_x
+;CHECK-APPLE-IOS: .set _y, __MergedGlobals_x+400
 
-
-;CHECK: .set z, .L_MergedGlobals
-;CHECK: .globl  x
-;CHECK: .set x, .L_MergedGlobals+4
-;CHECK: .size x, 4000
-
-;CHECK-APPLE-IOS-NOT: .set _z, __MergedGlobals_x
-;CHECK-APPLE-IOS:.globl _x
-;CHECK-APPLE-IOS:.set _x, __MergedGlobals_x+4
+;CHECK: .type   .L_MergedGlobals,@object // @_MergedGlobals
+;CHECK: .local  .L_MergedGlobals
+;CHECK: .comm   .L_MergedGlobals,800,4
+;CHECK: globl  x
+;CHECK: .set x, .L_MergedGlobals
+;CHECK: globl  y
+;CHECK: .set y, .L_MergedGlobals+400
+;CHECK-NOT: .set z, .L_MergedGlobals

llvm/trunk/test/CodeGen/ARM/global-merge-external-2.ll

+; RUN: llc < %s -mtriple=arm-eabi  -arm-global-merge                                 | FileCheck %s --check-prefixes=CHECK,CHECK-MERGE
+; RUN: llc < %s -mtriple=arm-eabi  -arm-global-merge -global-merge-on-external=true  | FileCheck %s --check-prefixes=CHECK,CHECK-MERGE
+; RUN: llc < %s -mtriple=arm-eabi  -arm-global-merge -global-merge-on-external=false | FileCheck %s --check-prefixes=CHECK,CHECK-NO-MERGE
+; RUN: llc < %s -mtriple=arm-macho -arm-global-merge                                 | FileCheck %s --check-prefixes=CHECK,CHECK-NO-MERGE
+; RUN: llc < %s -mtriple=arm-eabi  -arm-global-merge -relocation-model=pic           | FileCheck %s --check-prefixes=CHECK,CHECK-NO-MERGE
+; RUN: llc < %s -mtriple=thumbv7-win32 -arm-global-merge                             | FileCheck %s --check-prefixes=CHECK-WIN32
+
+@x = global i32 0, align 4
+@y = global i32 0, align 4
+@z = internal global i32 1, align 4
+
+define void @f1(i32 %a1, i32 %a2) {
+;CHECK:          f1:
+;CHECK:          ldr {{r[0-9]+}}, [[LABEL1:\.?LCPI[0-9]+_[0-9]]]
+;CHECK:          [[LABEL1]]:
+;CHECK-MERGE:    .long .L_MergedGlobals
+;CHECK-NO-MERGE: .long {{_?x}}
+;CHECK-WIN32:    f1:
+;CHECK-WIN32:    movw [[REG1:r[0-9]+]], :lower16:.L_MergedGlobals
+;CHECK-WIN32:    movt [[REG1]], :upper16:.L_MergedGlobals
+  store i32 %a1, i32* @x, align 4
+  store i32 %a2, i32* @y, align 4
+  ret void
+}
+
+define void @g1(i32 %a1, i32 %a2) {
+;CHECK:          g1:
+;CHECK:          ldr {{r[0-9]+}}, [[LABEL2:\.?LCPI[0-9]+_[0-9]]]
+;CHECK:          ldr {{r[0-9]+}}, [[LABEL3:\.?LCPI[0-9]+_[0-9]]]
+;CHECK:          [[LABEL2]]:
+;CHECK-MERGE:    .long {{_?z}}
+;CHECK:          [[LABEL3]]:
+;CHECK-MERGE:    .long .L_MergedGlobals
+;CHECK-NO-MERGE: .long {{_?y}}
+;CHECK-WIN32:    g1:
+;CHECK-WIN32:    movw    [[REG2:r[0-9]+]], :lower16:z
+;CHECK-WIN32:    movt    [[REG2]], :upper16:z
+;CHECK-WIN32:    movw [[REG3:r[0-9]+]], :lower16:.L_MergedGlobals
+;CHECK-WIN32:    movt [[REG3]], :upper16:.L_MergedGlobals
+  store i32 %a1, i32* @y, align 4
+  store i32 %a2, i32* @z, align 4
+  ret void
+}
+
+;CHECK-NO-MERGE-NOT: .globl .L_MergedGlobals
+
+;CHECK-MERGE:   .type   .L_MergedGlobals,%object
+;CHECK-MERGE:   .local  .L_MergedGlobals
+;CHECK-MERGE:   .comm   .L_MergedGlobals,8,4
+;CHECK-WIN32:   .lcomm  .L_MergedGlobals,8,4
+
+;CHECK-MERGE:   .globl  x
+;CHECK-MERGE: .set x, .L_MergedGlobals
+;CHECK-MERGE: .size x, 4
+;CHECK-MERGE:   .globl  y
+;CHECK-MERGE: .set y, .L_MergedGlobals+4
+;CHECK-MERGE: .size y, 4
+;CHECK-MERGE-NOT: .set z, .L_MergedGlobals+8
+
+
+;CHECK-WIN32:   .globl  x
+;CHECK-WIN32: .set x, .L_MergedGlobals
+;CHECK-WIN32:   .globl  y
+;CHECK-WIN32: .set y, .L_MergedGlobals+4
+;CHECK-WIN32-NOT: .set z, .L_MergedGlobals+8