From e1410f2d951d45aee3bdbcc05ecedaaa9db276e5 Mon Sep 17 00:00:00 2001
From: Jeff Garzik <jgarzik@pobox.com>
Date: Mon, 14 Nov 2005 14:06:26 -0500
Subject: [PATCH] [libata] fix bugs in ATAPI padding DMA mapping code

The ATAPI pad-to-next-32bit-boundary code modifies the scatterlist's
length variable, sometimes to zero.  x86-64 platform would oops if a
zero-length scatterlist entry was asked to be mapped.  Work around this
by ensuring that we never DMA-map a zero length buffer or SG entry.
---
Reading git-format-patch failed