Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Thu, 4 Oct 2001 22:15:48 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Thu, 4 Oct 2001 22:15:37 -0400 Received: from imo-d07.mx.aol.com ([205.188.157.39]:63977 "EHLO imo-d07.mx.aol.com") by vger.kernel.org with ESMTP id ; Thu, 4 Oct 2001 22:15:28 -0400 Date: Thu, 04 Oct 2001 22:14:29 EDT From: Telford002@aol.com Subject: PCI Device Setup Question To: Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit X-Mailer: Unknown (No Version) Message-ID: <66.154deb31.28ee7185@aol.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, I just joined the mailing list and have a quick question about the PCI Device Setup logic. I apologize if this is not the right forum. I am writing a multifunction driver for the Aurora serial cards. The driver supports asynchronous TTYs, synchronous TTYs and a sort of serial ethernet emulation as a network device. A port may act either as a serial TTY or as a serial network device, and in fact if the system is set up for dial up networking, the synchronous serial network interface is first brought up by sending commands to the corresponding asynchronous call out device. These cards have 4 or 8 synchronous/asynchronous ports on each card which hosts either 1 sab82538 or 2 sab82532 serial interface chips. These cards use the PLX9050 PCI interface chip to communicate with the PCI bus. Aurora also supplies a PCI extender chassis so that one slot can support 7 of either type of cards, i.e. up to 56 ports per slot. The PCI extender chassis has two DC21150 PCI-PCI bridges. The cards are never masters and do no support PCI PREFETCH and if PCI transactions are made to them that relate to PCI PREFETCH, they become confused and hang. DC21150s try to be really smart about PREFETCH and push the PREFETCH related transactions to the limit and therefore drive the cards insane if any of the cards memory resources is prefetchable (the cards use base address 0 for PLX9050 registers and base address 2 for the sab8253x registers). The PCI bus driver allocates the sab8253x registers as prefetchable memory resources. I see no comments in the changelogs indicating that anyone has run into the problem. I intend to fix the problem but I would like to chat with someone that is knowledgeable about the subsystem first. So I am curious if anyone on the list can point me toward someone who is "in charge" of this subsystem. As a sort of secondary almost unrelated question, I am not sure whether I should treat the Aurora card driver as a network device driver or as a TTY driver in the kernel build tree. I have created a special directory for it under the path drivers/net/wan/8253x. Thanks in advance for any assistance! Joachim Martillo - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/