From: Evan Cheng <evan.cheng@apple.com>
Date: Mon, 8 Mar 2010 21:05:02 +0000 (+0000)
Subject: Add documentation on sibling call optimization. Rename tailcall2.ll test to sibcall.ll.
X-Git-Url: http://demsky.eecs.uci.edu/git/?a=commitdiff_plain;h=dc444e9a236c9e5a9a3a5e6da75f29f7000208d3;p=oota-llvm.git

Add documentation on sibling call optimization. Rename tailcall2.ll test to sibcall.ll.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@97980 91177308-0d34-0410-b5e6-96231b3b80d8
---

diff --git a/docs/CodeGenerator.html b/docs/CodeGenerator.html
index 2be31851e36..f26d0abfad5 100644
--- a/docs/CodeGenerator.html
+++ b/docs/CodeGenerator.html
@@ -86,6 +86,7 @@
   <li><a href="#targetimpls">Target-specific Implementation Notes</a>
     <ul>
     <li><a href="#tailcallopt">Tail call optimization</a></li>
+    <li><a href="#sibcallopt">Sibling call optimization</a></li>
     <li><a href="#x86">The X86 backend</a></li>
     <li><a href="#ppc">The PowerPC backend</a>
       <ul>
@@ -1731,6 +1732,50 @@ define fastcc i32 @tailcaller(i32 %in1, i32 %in2) {
    (because one or more of above constraints are not met) to be followed by a
    readjustment of the stack. So performance might be worse in such cases.</p>
 
+</div>
+<!-- ======================================================================= -->
+<div class="doc_subsection">
+  <a name="sibcallopt">Sibling call optimization</a>
+</div>
+
+<div class="doc_text">
+
+<p>Sibling call optimization is a restricted form of tail call optimization.
+   Unlike tail call optimization described in the previous section, it can be
+   performed automatically on any tail calls when <tt>-tailcallopt</tt> option
+   is not specified.</p>
+
+<p>Sibling call optimization is currently performed on x86/x86-64 when the
+   following constraints are met:</p>
+
+<ul>
+  <li>Caller and callee have the same calling convention. It can be either
+      <tt>c</tt> or <tt>fastcc</tt>.
+
+  <li>The call is a tail call - in tail position (ret immediately follows call
+      and ret uses value of call or is void).</li>
+
+  <li>Caller and callee have matching return type or the callee result is not
+      used.
+
+  <li>If any of the callee arguments are being passed in stack, they must be
+      available in caller's own incoming argument stack and the frame offsets
+      must be the same.
+</ul>
+
+<p>Example:</p>
+<div class="doc_code">
+<pre>
+declare i32 @bar(i32, i32)
+
+define i32 @foo(i32 %a, i32 %b, i32 %c) {
+entry:
+  %0 = tail call i32 @bar(i32 %a, i32 %b)
+  ret i32 %0
+}
+</pre>
+</div>
+
 </div>
 <!-- ======================================================================= -->
 <div class="doc_subsection">
diff --git a/docs/LangRef.html b/docs/LangRef.html
index 7cfa05205ce..eec06fcc023 100644
--- a/docs/LangRef.html
+++ b/docs/LangRef.html
@@ -5149,8 +5149,11 @@ Loop:       ; Infinite loop that counts from 0 on up...
       a <a href="#i_ret"><tt>ret</tt></a> instruction.  If the "tail" marker is
       present, the function call is eligible for tail call optimization,
       but <a href="CodeGenerator.html#tailcallopt">might not in fact be
-      optimized into a jump</a>.  As of this writing, the extra requirements for
-      a call to actually be optimized are:
+      optimized into a jump</a>.  The code generator may optimize calls marked
+      "tail" with either 1) automatic <a href="CodeGenerator.html#sibcallopt">
+      sibling call optimization</a> when the caller and callee have
+      matching signatures, or 2) forced tail call optimization when the
+      following extra requirements are met:
       <ul>
         <li>Caller and callee both have the calling
             convention <tt>fastcc</tt>.</li>
diff --git a/test/CodeGen/X86/sibcall.ll b/test/CodeGen/X86/sibcall.ll
new file mode 100644
index 00000000000..90315fd2f26
--- /dev/null
+++ b/test/CodeGen/X86/sibcall.ll
@@ -0,0 +1,218 @@
+; RUN: llc < %s -march=x86    -asm-verbose=false | FileCheck %s -check-prefix=32
+; RUN: llc < %s -march=x86-64 -asm-verbose=false | FileCheck %s -check-prefix=64
+
+define void @t1(i32 %x) nounwind ssp {
+entry:
+; 32: t1:
+; 32: jmp {{_?}}foo
+
+; 64: t1:
+; 64: jmp {{_?}}foo
+  tail call void @foo() nounwind
+  ret void
+}
+
+declare void @foo()
+
+define void @t2() nounwind ssp {
+entry:
+; 32: t2:
+; 32: jmp {{_?}}foo2
+
+; 64: t2:
+; 64: jmp {{_?}}foo2
+  %0 = tail call i32 @foo2() nounwind
+  ret void
+}
+
+declare i32 @foo2()
+
+define void @t3() nounwind ssp {
+entry:
+; 32: t3:
+; 32: jmp {{_?}}foo3
+
+; 64: t3:
+; 64: jmp {{_?}}foo3
+  %0 = tail call i32 @foo3() nounwind
+  ret void
+}
+
+declare i32 @foo3()
+
+define void @t4(void (i32)* nocapture %x) nounwind ssp {
+entry:
+; 32: t4:
+; 32: call *
+; FIXME: gcc can generate a tailcall for this. But it's tricky.
+
+; 64: t4:
+; 64-NOT: call
+; 64: jmpq *
+  tail call void %x(i32 0) nounwind
+  ret void
+}
+
+define void @t5(void ()* nocapture %x) nounwind ssp {
+entry:
+; 32: t5:
+; 32-NOT: call
+; 32: jmpl *
+
+; 64: t5:
+; 64-NOT: call
+; 64: jmpq *
+  tail call void %x() nounwind
+  ret void
+}
+
+define i32 @t6(i32 %x) nounwind ssp {
+entry:
+; 32: t6:
+; 32: call {{_?}}t6
+; 32: jmp {{_?}}bar
+
+; 64: t6:
+; 64: jmp {{_?}}t6
+; 64: jmp {{_?}}bar
+  %0 = icmp slt i32 %x, 10
+  br i1 %0, label %bb, label %bb1
+
+bb:
+  %1 = add nsw i32 %x, -1
+  %2 = tail call i32 @t6(i32 %1) nounwind ssp
+  ret i32 %2
+
+bb1:
+  %3 = tail call i32 @bar(i32 %x) nounwind
+  ret i32 %3
+}
+
+declare i32 @bar(i32)
+
+define i32 @t7(i32 %a, i32 %b, i32 %c) nounwind ssp {
+entry:
+; 32: t7:
+; 32: jmp {{_?}}bar2
+
+; 64: t7:
+; 64: jmp {{_?}}bar2
+  %0 = tail call i32 @bar2(i32 %a, i32 %b, i32 %c) nounwind
+  ret i32 %0
+}
+
+declare i32 @bar2(i32, i32, i32)
+
+define signext i16 @t8() nounwind ssp {
+entry:
+; 32: t8:
+; 32: call {{_?}}bar3
+
+; 64: t8:
+; 64: callq {{_?}}bar3
+  %0 = tail call signext i16 @bar3() nounwind      ; <i16> [#uses=1]
+  ret i16 %0
+}
+
+declare signext i16 @bar3()
+
+define signext i16 @t9(i32 (i32)* nocapture %x) nounwind ssp {
+entry:
+; 32: t9:
+; 32: call *
+
+; 64: t9:
+; 64: callq *
+  %0 = bitcast i32 (i32)* %x to i16 (i32)*
+  %1 = tail call signext i16 %0(i32 0) nounwind
+  ret i16 %1
+}
+
+define void @t10() nounwind ssp {
+entry:
+; 32: t10:
+; 32: call
+
+; 64: t10:
+; 64: callq
+  %0 = tail call i32 @foo4() noreturn nounwind
+  unreachable
+}
+
+declare i32 @foo4()
+
+define i32 @t11(i32 %x, i32 %y, i32 %z.0, i32 %z.1, i32 %z.2) nounwind ssp {
+; In 32-bit mode, it's emitting a bunch of dead loads that are not being
+; eliminated currently.
+
+; 32: t11:
+; 32-NOT: subl ${{[0-9]+}}, %esp
+; 32: jne
+; 32-NOT: movl
+; 32-NOT: addl ${{[0-9]+}}, %esp
+; 32: jmp {{_?}}foo5
+
+; 64: t11:
+; 64-NOT: subq ${{[0-9]+}}, %esp
+; 64-NOT: addq ${{[0-9]+}}, %esp
+; 64: jmp {{_?}}foo5
+entry:
+  %0 = icmp eq i32 %x, 0
+  br i1 %0, label %bb6, label %bb
+
+bb:
+  %1 = tail call i32 @foo5(i32 %x, i32 %y, i32 %z.0, i32 %z.1, i32 %z.2) nounwind
+  ret i32 %1
+
+bb6:
+  ret i32 0
+}
+
+declare i32 @foo5(i32, i32, i32, i32, i32)
+
+%struct.t = type { i32, i32, i32, i32, i32 }
+
+define i32 @t12(i32 %x, i32 %y, %struct.t* byval align 4 %z) nounwind ssp {
+; 32: t12:
+; 32-NOT: subl ${{[0-9]+}}, %esp
+; 32-NOT: addl ${{[0-9]+}}, %esp
+; 32: jmp {{_?}}foo6
+
+; 64: t12:
+; 64-NOT: subq ${{[0-9]+}}, %esp
+; 64-NOT: addq ${{[0-9]+}}, %esp
+; 64: jmp {{_?}}foo6
+entry:
+  %0 = icmp eq i32 %x, 0
+  br i1 %0, label %bb2, label %bb
+
+bb:
+  %1 = tail call i32 @foo6(i32 %x, i32 %y, %struct.t* byval align 4 %z) nounwind
+  ret i32 %1
+
+bb2:
+  ret i32 0
+}
+
+declare i32 @foo6(i32, i32, %struct.t* byval align 4)
+
+; rdar://r7717598
+%struct.ns = type { i32, i32 }
+%struct.cp = type { float, float }
+
+define %struct.ns* @t13(%struct.cp* %yy) nounwind ssp {
+; 32: t13:
+; 32-NOT: jmp
+; 32: call
+; 32: ret
+
+; 64: t13:
+; 64-NOT: jmp
+; 64: call
+; 64: ret
+entry:
+  %0 = tail call fastcc %struct.ns* @foo7(%struct.cp* byval align 4 %yy, i8 signext 0) nounwind
+  ret %struct.ns* %0
+}
+
+declare fastcc %struct.ns* @foo7(%struct.cp* byval align 4, i8 signext) nounwind ssp
diff --git a/test/CodeGen/X86/tailcall2.ll b/test/CodeGen/X86/tailcall2.ll
deleted file mode 100644
index 90315fd2f26..00000000000
--- a/test/CodeGen/X86/tailcall2.ll
+++ /dev/null
@@ -1,218 +0,0 @@
-; RUN: llc < %s -march=x86    -asm-verbose=false | FileCheck %s -check-prefix=32
-; RUN: llc < %s -march=x86-64 -asm-verbose=false | FileCheck %s -check-prefix=64
-
-define void @t1(i32 %x) nounwind ssp {
-entry:
-; 32: t1:
-; 32: jmp {{_?}}foo
-
-; 64: t1:
-; 64: jmp {{_?}}foo
-  tail call void @foo() nounwind
-  ret void
-}
-
-declare void @foo()
-
-define void @t2() nounwind ssp {
-entry:
-; 32: t2:
-; 32: jmp {{_?}}foo2
-
-; 64: t2:
-; 64: jmp {{_?}}foo2
-  %0 = tail call i32 @foo2() nounwind
-  ret void
-}
-
-declare i32 @foo2()
-
-define void @t3() nounwind ssp {
-entry:
-; 32: t3:
-; 32: jmp {{_?}}foo3
-
-; 64: t3:
-; 64: jmp {{_?}}foo3
-  %0 = tail call i32 @foo3() nounwind
-  ret void
-}
-
-declare i32 @foo3()
-
-define void @t4(void (i32)* nocapture %x) nounwind ssp {
-entry:
-; 32: t4:
-; 32: call *
-; FIXME: gcc can generate a tailcall for this. But it's tricky.
-
-; 64: t4:
-; 64-NOT: call
-; 64: jmpq *
-  tail call void %x(i32 0) nounwind
-  ret void
-}
-
-define void @t5(void ()* nocapture %x) nounwind ssp {
-entry:
-; 32: t5:
-; 32-NOT: call
-; 32: jmpl *
-
-; 64: t5:
-; 64-NOT: call
-; 64: jmpq *
-  tail call void %x() nounwind
-  ret void
-}
-
-define i32 @t6(i32 %x) nounwind ssp {
-entry:
-; 32: t6:
-; 32: call {{_?}}t6
-; 32: jmp {{_?}}bar
-
-; 64: t6:
-; 64: jmp {{_?}}t6
-; 64: jmp {{_?}}bar
-  %0 = icmp slt i32 %x, 10
-  br i1 %0, label %bb, label %bb1
-
-bb:
-  %1 = add nsw i32 %x, -1
-  %2 = tail call i32 @t6(i32 %1) nounwind ssp
-  ret i32 %2
-
-bb1:
-  %3 = tail call i32 @bar(i32 %x) nounwind
-  ret i32 %3
-}
-
-declare i32 @bar(i32)
-
-define i32 @t7(i32 %a, i32 %b, i32 %c) nounwind ssp {
-entry:
-; 32: t7:
-; 32: jmp {{_?}}bar2
-
-; 64: t7:
-; 64: jmp {{_?}}bar2
-  %0 = tail call i32 @bar2(i32 %a, i32 %b, i32 %c) nounwind
-  ret i32 %0
-}
-
-declare i32 @bar2(i32, i32, i32)
-
-define signext i16 @t8() nounwind ssp {
-entry:
-; 32: t8:
-; 32: call {{_?}}bar3
-
-; 64: t8:
-; 64: callq {{_?}}bar3
-  %0 = tail call signext i16 @bar3() nounwind      ; <i16> [#uses=1]
-  ret i16 %0
-}
-
-declare signext i16 @bar3()
-
-define signext i16 @t9(i32 (i32)* nocapture %x) nounwind ssp {
-entry:
-; 32: t9:
-; 32: call *
-
-; 64: t9:
-; 64: callq *
-  %0 = bitcast i32 (i32)* %x to i16 (i32)*
-  %1 = tail call signext i16 %0(i32 0) nounwind
-  ret i16 %1
-}
-
-define void @t10() nounwind ssp {
-entry:
-; 32: t10:
-; 32: call
-
-; 64: t10:
-; 64: callq
-  %0 = tail call i32 @foo4() noreturn nounwind
-  unreachable
-}
-
-declare i32 @foo4()
-
-define i32 @t11(i32 %x, i32 %y, i32 %z.0, i32 %z.1, i32 %z.2) nounwind ssp {
-; In 32-bit mode, it's emitting a bunch of dead loads that are not being
-; eliminated currently.
-
-; 32: t11:
-; 32-NOT: subl ${{[0-9]+}}, %esp
-; 32: jne
-; 32-NOT: movl
-; 32-NOT: addl ${{[0-9]+}}, %esp
-; 32: jmp {{_?}}foo5
-
-; 64: t11:
-; 64-NOT: subq ${{[0-9]+}}, %esp
-; 64-NOT: addq ${{[0-9]+}}, %esp
-; 64: jmp {{_?}}foo5
-entry:
-  %0 = icmp eq i32 %x, 0
-  br i1 %0, label %bb6, label %bb
-
-bb:
-  %1 = tail call i32 @foo5(i32 %x, i32 %y, i32 %z.0, i32 %z.1, i32 %z.2) nounwind
-  ret i32 %1
-
-bb6:
-  ret i32 0
-}
-
-declare i32 @foo5(i32, i32, i32, i32, i32)
-
-%struct.t = type { i32, i32, i32, i32, i32 }
-
-define i32 @t12(i32 %x, i32 %y, %struct.t* byval align 4 %z) nounwind ssp {
-; 32: t12:
-; 32-NOT: subl ${{[0-9]+}}, %esp
-; 32-NOT: addl ${{[0-9]+}}, %esp
-; 32: jmp {{_?}}foo6
-
-; 64: t12:
-; 64-NOT: subq ${{[0-9]+}}, %esp
-; 64-NOT: addq ${{[0-9]+}}, %esp
-; 64: jmp {{_?}}foo6
-entry:
-  %0 = icmp eq i32 %x, 0
-  br i1 %0, label %bb2, label %bb
-
-bb:
-  %1 = tail call i32 @foo6(i32 %x, i32 %y, %struct.t* byval align 4 %z) nounwind
-  ret i32 %1
-
-bb2:
-  ret i32 0
-}
-
-declare i32 @foo6(i32, i32, %struct.t* byval align 4)
-
-; rdar://r7717598
-%struct.ns = type { i32, i32 }
-%struct.cp = type { float, float }
-
-define %struct.ns* @t13(%struct.cp* %yy) nounwind ssp {
-; 32: t13:
-; 32-NOT: jmp
-; 32: call
-; 32: ret
-
-; 64: t13:
-; 64-NOT: jmp
-; 64: call
-; 64: ret
-entry:
-  %0 = tail call fastcc %struct.ns* @foo7(%struct.cp* byval align 4 %yy, i8 signext 0) nounwind
-  ret %struct.ns* %0
-}
-
-declare fastcc %struct.ns* @foo7(%struct.cp* byval align 4, i8 signext) nounwind ssp