From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: AHCI driver preferring nr_ports over port map Date: Sat, 02 Feb 2008 17:16:59 +0900 Message-ID: <47A426FB.4090105@gmail.com> References: <47A344DF.76E4.0078.0@novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from rv-out-0910.google.com ([209.85.198.191]:45863 "EHLO rv-out-0910.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756872AbYBBIRH (ORCPT ); Sat, 2 Feb 2008 03:17:07 -0500 Received: by rv-out-0910.google.com with SMTP id k20so992914rvb.1 for ; Sat, 02 Feb 2008 00:17:06 -0800 (PST) In-Reply-To: <47A344DF.76E4.0078.0@novell.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Jan Beulich Cc: Jeff Garzik , linux-ide@vger.kernel.org Jan Beulich wrote: > Jeff, >=20 > while I realize that Intel's documentation may not be consistent with > anything more generic (which I don't know where to look for), this > current behavior seems to contradict what Intel documents for ESB2: >=20 > "23.3.1.4 PI =E2=80=93 Ports Implemented Register (D31:F2) > Address Offset: ABAR + 0Ch=E2=80=930Fh Attribute: R/WO, RO > Default Value: 00000000h Size: 32 bits >=20 > This register indicates which ports are exposed to the Intel=C2=AE > 631xESB/632xESB I/O Controller Hub. It is loaded by platform BIOS. It > indicates which ports that the device supports are available for > software to use. For ports that are not available, software must not > read or write to registers within that port." nr_ports is preferred over port_map when they disagree which shouldn't happen in the first place. On some earlier ahcis, PI was cleared to zero and lower nr_port number of ports should be used. The reason why nr_ports is preferred over PI comes from similar place. Hardware/BIOSe= n are more likely to get PI wrong than nr_ports, so... Do you have any real case where the above behavior causes problem? --=20 tejun