Support for trampolines, except for X86 codegen which is

still under discussion.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@40549 91177308-0d34-0410-b5e6-96231b3b80d8

1 files changed, 105 insertions, 0 deletions

diff --git a/docs/LangRef.html b/docs/LangRef.html
index 6ebbb29..2cf74c2 100644
--- a/docs/LangRef.html
+++ b/docs/LangRef.html
@@ -200,6 +200,12 @@
           <li><a href="#int_memory_barrier">'<tt>llvm.memory.barrier</tt>' Intrinsic</a></li>
         </ol>
       </li>
+      <li><a href="#int_trampoline">Trampoline Intrinsics</a>
+        <ol>
+          <li><a href="#int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a></li>
+          <li><a href="#int_at">'<tt>llvm.adjust.trampoline</tt>' Intrinsic</a></li>
+        </ol>
+      </li>
       <li><a href="#int_general">General intrinsics</a>
         <ol>
           <li><a href="#int_var_annotation">
@@ -5123,6 +5129,105 @@ declare void @llvm.memory.barrier( i1 &lt;ll&gt;, i1 &lt;ls&gt;, i1 &lt;sl&gt;,
 
 <!-- ======================================================================= -->
 <div class="doc_subsection">
+  <a name="int_trampoline">Trampoline Intrinsics</a>
+</div>
+
+<div class="doc_text">
+<p>
+  These intrinsics make it possible to excise one parameter, marked with
+  the <tt>nest</tt> attribute, from a function.  The result is a callable
+  function pointer lacking the nest parameter - the caller does not need
+  to provide a value for it.  Instead, the value to use is stored in
+  advance in a "trampoline", a block of memory usually allocated
+  on the stack, which also contains code to splice the nest value into the
+  argument list.  This is used to implement the GCC nested function address
+  extension.
+</p>
+<p>
+  For example, if the function is
+  <tt>i32 f(i8* nest  %c, i32 %x, i32 %y)</tt> then the resulting function
+  pointer has signature <tt>i32 (i32, i32)*</tt>.  It can be created as follows:
+<pre>
+  %tramp1 = alloca [10 x i8], align 4 ; size and alignment only correct for X86
+  %tramp = getelementptr [10 x i8]* %tramp1, i32 0, i32 0
+  call void @llvm.init.trampoline( i8* %tramp, i8* bitcast (i32 (i8* nest , i32, i32)* @f to i8*), i8* %nval )
+  %adj = call i8* @llvm.adjust.trampoline( i8* %tramp )
+  %fp = bitcast i8* %adj to i32 (i32, i32)*
+</pre>
+  The call <tt>%val = call i32 %fp( i32 %x, i32 %y )</tt> is then equivalent to
+  <tt>%val = call i32 %f( i8* %nval, i32 %x, i32 %y )</tt>.
+</p>
+<p>
+  Trampolines are currently only supported on the X86 architecture.
+</p>
+</div>
+
+<!-- _______________________________________________________________________ -->
+<div class="doc_subsubsection">
+  <a name="int_it">'<tt>llvm.init.trampoline</tt>' Intrinsic</a>
+</div>
+<div class="doc_text">
+<h5>Syntax:</h5>
+<pre>
+declare void @llvm.init.trampoline(i8* &lt;tramp&gt;, i8* &lt;func&gt;, i8* &lt;nval&gt;)
+</pre>
+<h5>Overview:</h5>
+<p>
+  This initializes the memory pointed to by <tt>tramp</tt> as a trampoline.
+</p>
+<h5>Arguments:</h5>
+<p>
+  The <tt>llvm.init.trampoline</tt> intrinsic takes three arguments, all
+  pointers.  The <tt>tramp</tt> argument must point to a sufficiently large
+  and sufficiently aligned block of memory; this memory is written to by the
+  intrinsic.  Currently LLVM provides no help in determining just how big and
+  aligned the memory needs to be.  The <tt>func</tt> argument must hold a
+  function bitcast to an <tt>i8*</tt>.
+</p>
+<h5>Semantics:</h5>
+<p>
+  The block of memory pointed to by <tt>tramp</tt> is filled with target
+  dependent code, turning it into a function.
+  The new function's signature is the same as that of <tt>func</tt> with
+  any arguments marked with the <tt>nest</tt> attribute removed.  At most
+  one such <tt>nest</tt> argument is allowed, and it must be of pointer
+  type.  Calling the new function is equivalent to calling <tt>func</tt>
+  with the same argument list, but with <tt>nval</tt> used for the missing
+  <tt>nest</tt> argument.
+</p>
+</div>
+
+<!-- _______________________________________________________________________ -->
+<div class="doc_subsubsection">
+  <a name="int_at">'<tt>llvm.adjust.trampoline</tt>' Intrinsic</a>
+</div>
+<div class="doc_text">
+<h5>Syntax:</h5>
+<pre>
+declare i8* @llvm.adjust.trampoline(i8* &lt;tramp&gt;)
+</pre>
+<h5>Overview:</h5>
+<p>
+  This intrinsic returns a function pointer suitable for executing
+  the trampoline code pointed to by <tt>tramp</tt>.
+</p>
+<h5>Arguments:</h5>
+<p>
+  The <tt>llvm.adjust.trampoline</tt> takes one argument, a pointer to a
+  trampoline initialized by the
+  <a href="#int_it">'<tt>llvm.init.trampoline</tt>' intrinsic</a>.
+</p>
+<h5>Semantics:</h5>
+<p>
+  A function pointer that can be used to execute the trampoline code in
+  <tt>tramp</tt> is returned.  The returned value should be bitcast to an
+  <a href="#int_trampoline">appropriate function pointer type</a>
+  before being called.
+</p>
+</div>
+
+<!-- ======================================================================= -->
+<div class="doc_subsection">
   <a name="int_general">General Intrinsics</a>
 </div>