This is a patch to the assembler frontend to detect when aligning a text

section with TextAlignFillValue and calls EmitCodeAlignment() instead of
calling EmitValueToAlignment().  This allows x86 assembly code to be aligned
with optimal nops.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97158 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/MC/MCParser/AsmParser.cpp b/lib/MC/MCParser/AsmParser.cpp
index 51ad5d18f3bfce6524869c5e86fa67801a9568cf..1b22166f6a46d61539ded36cbd7f04ad1c34331c 100644 (file)
--- a/lib/MC/MCParser/AsmParser.cpp
+++ b/lib/MC/MCParser/AsmParser.cpp
@@ -146,7 +146,7 @@ bool AsmParser::Run() {
   // FIXME: Target hook & command line option for initial section.
   Out.SwitchSection(getMachOSection("__TEXT", "__text",
                                     MCSectionMachO::S_ATTR_PURE_INSTRUCTIONS,
-                                    0, SectionKind()));
+                                    0, SectionKind::getText()));
 
 
   // Prime the lexer.
@@ -1237,8 +1237,14 @@ bool AsmParser::ParseDirectiveAlign(bool IsPow2, unsigned ValueSize) {
     }
   }
 
-  // FIXME: Target specific behavior about how the "extra" bytes are filled.
-  Out.EmitValueToAlignment(Alignment, FillExpr, ValueSize, MaxBytesToFill);
+  // FIXME: hard code the parser to use EmitCodeAlignment for text when using
+  // the TextAlignFillValue.
+  if(Out.getCurrentSection()->getKind().isText() && 
+     Lexer.getMAI().getTextAlignFillValue() == FillExpr)
+    Out.EmitCodeAlignment(Alignment, MaxBytesToFill);
+  else
+    // FIXME: Target specific behavior about how the "extra" bytes are filled.
+    Out.EmitValueToAlignment(Alignment, FillExpr, ValueSize, MaxBytesToFill);
 
   return false;
 }

diff --git a/test/MC/MachO/Darwin/optimal_nop.s b/test/MC/MachO/Darwin/optimal_nop.s
new file mode 100644 (file)
index 0000000..29cb073
--- /dev/null
+++ b/test/MC/MachO/Darwin/optimal_nop.s
@@ -0,0 +1,156 @@
+// Validate that we can assemble this file exactly like the platform
+// assembler.
+//
+// RUN: llvm-mc -filetype=obj -triple i386-apple-darwin10 -o %t.mc.o %s
+// RUN: as -arch i386 -o %t.as.o %s
+// RUN: diff %t.mc.o %t.as.o
+
+# 1 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        # nop
+        # 0x90
+        .align 1, 0x90
+        ret
+# 2 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        # xchg %ax,%ax
+        # 0x66, 0x90
+        .align 2, 0x90
+        ret
+# 3 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        # nopl (%[re]ax)
+        # 0x0f, 0x1f, 0x00
+        .align 2, 0x90
+        ret
+# 4 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        ret
+        # nopl 0(%[re]ax)
+        # 0x0f, 0x1f, 0x40, 0x00
+        .align 3, 0x90
+        ret
+# 5 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        # nopl 0(%[re]ax,%[re]ax,1)
+        # 0x0f, 0x1f, 0x44, 0x00, 0x00
+        .align 3, 0x90
+        ret
+# 6 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        # nopw 0(%[re]ax,%[re]ax,1)
+        # 0x66, 0x0f, 0x1f, 0x44, 0x00, 0x00
+        .align 3, 0x90
+        ret
+# 7 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        # nopl 0L(%[re]ax)
+        # 0x0f, 0x1f, 0x80, 0x00, 0x00, 0x00, 0x00
+        .align 3, 0x90
+        ret
+# 8 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        ret
+        ret
+        ret
+        ret
+        ret
+        # nopl 0L(%[re]ax,%[re]ax,1)
+        # 0x0f, 0x1f, 0x84, 0x00, 0x00, 0x00, 0x00, 0x00
+        .align 3, 0x90
+        ret
+# 9 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        ret
+        ret
+        ret
+        ret
+        # nopw 0L(%[re]ax,%[re]ax,1)
+        # 0x66, 0x0f, 0x1f, 0x84, 0x00, 0x00, 0x00, 0x00, 0x00
+        .align 4, 0x90
+        ret
+# 10 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        ret
+        ret
+        ret
+        ret
+        # nopw %cs:0L(%[re]ax,%[re]ax,1)
+        # 0x66, 0x2e, 0x0f, 0x1f, 0x84, 0x00, 0x00, 0x00, 0x00, 0x00
+        .align 4, 0x90
+        ret
+# 11 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        ret
+        ret
+        # nopw %cs:0L(%[re]ax,%[re]ax,1)
+        # 0x66, 0x2e, 0x0f, 0x1f, 0x84, 0x00, 0x00, 0x00, 0x00, 0x00
+        .align 4, 0x90
+        ret
+# 12 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        ret
+        # nopw 0(%[re]ax,%[re]ax,1)
+        # nopw 0(%[re]ax,%[re]ax,1)
+        # 0x66, 0x0f, 0x1f, 0x44, 0x00, 0x00,
+        # 0x66, 0x0f, 0x1f, 0x44, 0x00, 0x00
+        .align 4, 0x90
+        ret
+# 13 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        ret
+        # nopw 0(%[re]ax,%[re]ax,1)
+        # nopl 0L(%[re]ax)
+        # 0x66, 0x0f, 0x1f, 0x44, 0x00, 0x00,
+        # 0x0f, 0x1f, 0x80, 0x00, 0x00, 0x00, 0x00
+        .align 4, 0x90
+        ret
+# 14 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        ret
+        # nopl 0L(%[re]ax)
+        # nopl 0L(%[re]ax)
+        # 0x0f, 0x1f, 0x80, 0x00, 0x00, 0x00, 0x00,
+        # 0x0f, 0x1f, 0x80, 0x00, 0x00, 0x00, 0x00
+        .align 4, 0x90
+        ret
+# 15 byte nop test
+        .align 4, 0 # start with 16 byte alignment filled with zeros
+        ret
+        # nopl 0L(%[re]ax)
+        # nopl 0L(%[re]ax,%[re]ax,1)
+        # 0x0f, 0x1f, 0x80, 0x00, 0x00, 0x00, 0x00,
+        # 0x0f, 0x1f, 0x84, 0x00, 0x00, 0x00, 0x00, 0x00
+        .align 4, 0x90
+        ret