x86_64: Add a comment explaining the TASK_SIZE_MAX guard page

That guard page is absolutely necessary; explain why for
posterity.

Signed-off-by: Andy Lutomirski <luto@amacapital.net>
Link: http://lkml.kernel.org/r/23320cb5017c2da8475ec20fcde8089d82aa2699.1415144745.git.luto@amacapital.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>

arch/x86/include/asm/processor.h

@@ -893,7 +893,13 @@ extern unsigned long thread_saved_pc(struct task_struct *tsk);
 
 #else
 /*
- * User space process size. 47bits minus one guard page.
+ * User space process size. 47bits minus one guard page.  The guard
+ * page is necessary on Intel CPUs: if a SYSCALL instruction is at
+ * the highest possible canonical userspace address, then that
+ * syscall will enter the kernel with a non-canonical return
+ * address, and SYSRET will explode dangerously.  We avoid this
+ * particular problem by preventing anything from being mapped
+ * at the maximum canonical address.
  */
 #define TASK_SIZE_MAX	((1UL << 47) - PAGE_SIZE)