From: Simon Kagstrom Subject: [PATCH 0/2]: x86: Honor IO_DELAY IO port settings Date: Wed, 17 Mar 2010 13:28:49 +0100 Message-ID: <20100317132849.7d49939b@marrow.netinsight.se> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: Received: from ernst.netinsight.se ([194.16.221.21]:52008 "HELO ernst.netinsight.se" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with SMTP id S1754722Ab0CQM2z (ORCPT ); Wed, 17 Mar 2010 08:28:55 -0400 Sender: linux-serial-owner@vger.kernel.org List-Id: linux-serial@vger.kernel.org To: x86@kernel.org, linux-serial@vger.kernel.org, linux-kernel@vger.kernel.org Cc: mingo@redhat.com, tglx@linutronix.de, alan@linux.intel.com, akpm@linux-foundation.org, hpa@zytor.com Greetings! The kernel can be configured to use IO port 0xed instead of port 0x80 for IO delay writes and some boards also require this to function properly. These two patches fix two places where port 0x80 is hardcoded. * Patch 1: Use native_io_delay for the serial/8250 driver instead of always using 0x80. * Patch 2: Honor CONFIG_IO_DELAY_0XED if set for real-mode boot code The board we have works fine with using 0x80, but we're debugging a BIOS issue and are logging writes to port 0x80 (BIOS post codes). Avoiding the extra port 0x80 writes makes it easier to weed out the important information. // Simon