diff --git a/Documentation/x86/intel_mpx.txt b/Documentation/x86/intel_mpx.txt
index 4472ed2ad921b74ce088643f2ca7c3cf494df30a..6ca6e2bd9ae9ac6c812ce6f6d5bf634b8557715d 100644
--- a/Documentation/x86/intel_mpx.txt
+++ b/Documentation/x86/intel_mpx.txt
@@ -30,9 +30,15 @@ is how we expect the compiler, application and kernel to work together.
    instrumentation as well as some setup code called early after the app
    starts. New instruction prefixes are noops for old CPUs.
 2) That setup code allocates (virtual) space for the "bounds directory",
-   points the "bndcfgu" register to the directory and notifies the kernel
-   (via the new prctl(PR_MPX_ENABLE_MANAGEMENT)) that the app will be using
-   MPX.
+   points the "bndcfgu" register to the directory (must also set the valid
+   bit) and notifies the kernel (via the new prctl(PR_MPX_ENABLE_MANAGEMENT))
+   that the app will be using MPX.  The app must be careful not to access
+   the bounds tables between the time when it populates "bndcfgu" and
+   when it calls the prctl().  This might be hard to guarantee if the app
+   is compiled with MPX.  You can add "__attribute__((bnd_legacy))" to
+   the function to disable MPX instrumentation to help guarantee this.
+   Also be careful not to call out to any other code which might be
+   MPX-instrumented.
 3) The kernel detects that the CPU has MPX, allows the new prctl() to
    succeed, and notes the location of the bounds directory. Userspace is
    expected to keep the bounds directory at that locationWe note it