WebAssembly: use .skip instead of .zero directive

.zero is confusing when used with two arguments. Documentation:

  This directive emits SIZE 0-valued bytes.  SIZE must be an absolute
  expression.  This directive is actually an alias for the '.skip'
  directive so in can take an optional second argument of the value to
  store in the bytes instead of zero.  Using '.zero' in this way would be
  confusing however.

Ref: https://sourceware.org/bugzilla/show_bug.cgi?id=18353

Hexagon and Sparc do the same, and it's all the same to WebAssembly so
let's pick the less confusing of the two.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@257111 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCAsmInfo.cpp b/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCAsmInfo.cpp
index b1ab3ef29d86814d7d46f74cf6259639427d7e0b..02c717a9210173d8e673d2ec3737eba45d06e557 100644 (file)
--- a/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCAsmInfo.cpp
+++ b/lib/Target/WebAssembly/MCTargetDesc/WebAssemblyMCAsmInfo.cpp
@@ -29,6 +29,10 @@ WebAssemblyMCAsmInfo::WebAssemblyMCAsmInfo(const Triple &T) {
 
   UseDataRegionDirectives = true;
 
+  // Use .skip instead of .zero because .zero is confusing when used with two
+  // arguments (it doesn't actually zero things out).
+  ZeroDirective = "\t.skip\t";
+
   Data8bitsDirective = "\t.int8\t";
   Data16bitsDirective = "\t.int16\t";
   Data32bitsDirective = "\t.int32\t";

diff --git a/test/CodeGen/WebAssembly/global.ll b/test/CodeGen/WebAssembly/global.ll
index 29c47e1e8e920dcf2a6d473c74b27c932b07e571..f2bb8e7deac8921ae9aecb91d3983c8fc0a0d2f2 100644 (file)
--- a/test/CodeGen/WebAssembly/global.ll
+++ b/test/CodeGen/WebAssembly/global.ll
@@ -37,7 +37,7 @@ define i8* @call_memcpy(i8* %p, i8* nocapture readonly %q, i32 %n) {
 @g = private global i32 1337
 
 ; CHECK-LABEL: ud:
-; CHECK-NEXT: .zero 4{{$}}
+; CHECK-NEXT: .skip 4{{$}}
 ; CHECK-NEXT: .size ud, 4{{$}}
 @ud = internal global i32 undef
 
@@ -73,7 +73,7 @@ define i8* @call_memcpy(i8* %p, i8* nocapture readonly %q, i32 %n) {
 ; CHECK: .type ud64,@object
 ; CHECK: .align 3{{$}}
 ; CHECK-NEXT: ud64:
-; CHECK-NEXT: .zero 8{{$}}
+; CHECK-NEXT: .skip 8{{$}}
 ; CHECK-NEXT: .size ud64, 8{{$}}
 @ud64 = internal global i64 undef
 
@@ -102,7 +102,7 @@ define i8* @call_memcpy(i8* %p, i8* nocapture readonly %q, i32 %n) {
 ; CHECK: .type f32ud,@object
 ; CHECK: .align 2{{$}}
 ; CHECK-NEXT: f32ud:
-; CHECK-NEXT: .zero 4{{$}}
+; CHECK-NEXT: .skip 4{{$}}
 ; CHECK-NEXT: .size f32ud, 4{{$}}
 @f32ud = internal global float undef
 
@@ -131,7 +131,7 @@ define i8* @call_memcpy(i8* %p, i8* nocapture readonly %q, i32 %n) {
 ; CHECK: .type f64ud,@object
 ; CHECK: .align 3{{$}}
 ; CHECK-NEXT: f64ud:
-; CHECK-NEXT: .zero 8{{$}}
+; CHECK-NEXT: .skip 8{{$}}
 ; CHECK-NEXT: .size f64ud, 8{{$}}
 @f64ud = internal global double undef
 
@@ -172,6 +172,6 @@ define i8* @call_memcpy(i8* %p, i8* nocapture readonly %q, i32 %n) {
 ; CHECK: .globl   rom{{$}}
 ; CHECK: .align   4{{$}}
 ; CHECK: rom:
-; CHECK: .zero    512{{$}}
+; CHECK: .skip    512{{$}}
 ; CHECK: .size    rom, 512{{$}}
 @rom = constant [128 x i32] zeroinitializer, align 16