Fixes #96

tc39 · May 4, 2016 · c0800ce · c0800ce
1 parent 19df2c0
commit c0800ce
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 2 deletions.
diff --git a/tc39/shmem.html b/tc39/shmem.html
@@ -71,7 +71,7 @@ <h1><span class="secnum"></span>Introduction<span class="utils"><span class="anc
 
 <p>Changelog:</p>
 <ul>
-  <li> 2016-05-04 -- (Issue 94) Clarified slightly the meaning of lock-free.
+  <li> 2016-05-04 -- (Issue 94, Issue 96) Clarified slightly the meaning of lock-free.
 
   </li><li> 2016-04-06 -- Moved Agent Clusters and shared memory specific agent properties back into this spec; the agents spec is being split into  <a href="https://github.com/tc39/ecma262/pull/522">a PR</a> and ideas specific to the shared memory proposal.
 
@@ -276,6 +276,12 @@ <h1><span class="secnum">3.2</span>Agents (AMENDMENTS)<span class="utils"><span
         upper bound on how many machine steps may be required to
         complete a lock-free atomic operation.  </p>
 
+      <p>That an atomic access of size  <em>n</em> is lock-free does
+        not imply anything about the (perceived) atomicity of
+        non-atomic accesses of size  <em>n</em>, specifically,
+        non-atomic accesses may still be performed as a sequence of
+        several separate memory accesses.  </p>
+
     </emu-note>
 
   </emu-clause>

diff --git a/tc39/spec.html b/tc39/spec.html
@@ -68,7 +68,7 @@ <h1>Introduction</h1>
 
 <p>Changelog:</p>
 <ul>
-  <li> 2016-05-04 -- (Issue 94) Clarified slightly the meaning of lock-free.
+  <li> 2016-05-04 -- (Issue 94, Issue 96) Clarified slightly the meaning of lock-free.
   <li> 2016-04-06 -- Moved Agent Clusters and shared memory specific agent properties back into this spec; the agents spec is being split into <a href="https://github.com/tc39/ecma262/pull/522">a PR</a> and ideas specific to the shared memory proposal.
   <li> 2016-03-30 -- (Issue 95) Renamed Atomics.futexWait as Atomics.wait and Atomics.futexWake as Atomics.wake. No functional change.
   <li> 2016-03-30 -- (Issue 90) Factor AtomicFetchOp out as a semantic algorithm, and reference it formally where appropriate. No functional change.
@@ -240,6 +240,12 @@ <h1>Agents (AMENDMENTS)</h1>
         upper bound on how many machine steps may be required to
         complete a lock-free atomic operation. </p>
 
+      <p>That an atomic access of size <em>n</em> is lock-free does
+        not imply anything about the (perceived) atomicity of
+        non-atomic accesses of size <em>n</em>, specifically,
+        non-atomic accesses may still be performed as a sequence of
+        several separate memory accesses. </p>
+
     </emu-note>
 
   </emu-clause>