From owner-freebsd-current@FreeBSD.ORG Tue Aug 2 22:06:06 2011 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 7945B106566B; Tue, 2 Aug 2011 22:06:06 +0000 (UTC) (envelope-from jhb@freebsd.org) Received: from cyrus.watson.org (cyrus.watson.org [65.122.17.42]) by mx1.freebsd.org (Postfix) with ESMTP id 39DE18FC08; Tue, 2 Aug 2011 22:06:06 +0000 (UTC) Received: from bigwig.baldwin.cx (66.111.2.69.static.nyinternet.net [66.111.2.69]) by cyrus.watson.org (Postfix) with ESMTPSA id BA9F646B03; Tue, 2 Aug 2011 18:06:05 -0400 (EDT) Received: from jhbbsd.localnet (unknown [209.249.190.124]) by bigwig.baldwin.cx (Postfix) with ESMTPSA id 56B9F8A02F; Tue, 2 Aug 2011 18:06:05 -0400 (EDT) From: John Baldwin To: Andriy Gapon Date: Tue, 2 Aug 2011 18:06:04 -0400 User-Agent: KMail/1.13.5 (FreeBSD/8.2-CBSD-20110617; KDE/4.5.5; amd64; ; ) References: <4E23EE49.5040801@FreeBSD.org> <201107191116.07116.jhb@freebsd.org> <4E33A990.7040006@FreeBSD.org> In-Reply-To: <4E33A990.7040006@FreeBSD.org> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-15" Content-Transfer-Encoding: 7bit Message-Id: <201108021806.04759.jhb@freebsd.org> X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.2.6 (bigwig.baldwin.cx); Tue, 02 Aug 2011 18:06:05 -0400 (EDT) Cc: freebsd-current@freebsd.org Subject: Re: ichwd0: unable to reserve GCS registers X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 02 Aug 2011 22:06:06 -0000 On Saturday, July 30, 2011 2:49:52 am Andriy Gapon wrote: > on 19/07/2011 18:16 John Baldwin said the following: > > Hmm, can you get devinfo -r output from a working kernel with ichwd loaded? > > You might be able to just build the kernel with 'nooptions NEW_PCIB'. > > I believe that I've got a similar problem with amdsbwd(4). > It needs some resources (I/O ports) that belong to ACPI. > The problem is that the driver attaches to isa bus which is under > isab->pci->pcib and those particular resources are not assigned to the Host-PCI > bridge. > > I think that you already made a suggestion that perhaps isa bus should directly > attach to acpi bus when acpi is available. Not sure if there are any > alternative approaches. Can you try this: --- //depot/projects/pci/sys/dev/acpica/acpi.c 2011-06-25 12:05:19.000000000 0000 +++ //depot/projects/pci/sys/dev/acpica/acpi.c 2011-08-02 20:21:42.000000000 0000 @@ -1238,7 +1238,6 @@ struct resource_list_entry *rle; struct resource_list *rl; struct resource *res; - struct rman *rm; int isdefault = (start == 0UL && end == ~0UL); /* @@ -1291,15 +1290,29 @@ } else res = BUS_ALLOC_RESOURCE(device_get_parent(bus), child, type, rid, start, end, count, flags); - if (res != NULL || start + count - 1 != end) - return (res); /* * If the first attempt failed and this is an allocation of a * specific range, try to satisfy the request via a suballocation - * from our system resource regions. Note that we only handle - * memory and I/O port system resources. + * from our system resource regions. */ + if (res == NULL && start + count - 1 == end) + res = acpi_alloc_sysres(child, type, rid, start, end, count, flags); + return (res); +} + +/* + * Attempt to allocate a specific resource range from the system + * resource ranges. Note that we only handle memory and I/O port + * system resources. + */ +struct resource * +acpi_alloc_sysres(device_t child, int type, int *rid, u_long start, u_long end, + u_long count, u_int flags) +{ + struct rman *rm; + struct resource *res; + switch (type) { case SYS_RES_IOPORT: rm = &acpi_rman_io; @@ -1311,6 +1324,7 @@ return (NULL); } + KASSERT(start + count - 1 == end, ("wildcard resource range")); res = rman_reserve_resource(rm, start, end, count, flags & ~RF_ACTIVE, child); if (res == NULL) --- //depot/projects/pci/sys/dev/acpica/acpi_pcib_acpi.c 2011-07-22 18:19:55.000000000 0000 +++ //depot/projects/pci/sys/dev/acpica/acpi_pcib_acpi.c 2011-08-02 20:21:42.000000000 0000 @@ -541,6 +541,7 @@ { #ifdef NEW_PCIB struct acpi_hpcib_softc *sc; + struct resource *res; #endif #if defined(__i386__) || defined(__amd64__) @@ -549,8 +550,11 @@ #ifdef NEW_PCIB sc = device_get_softc(dev); - return (pcib_host_res_alloc(&sc->ap_host_res, child, type, rid, start, end, - count, flags)); + res = pcib_host_res_alloc(&sc->ap_host_res, child, type, rid, start, end, + count, flags); + if (res == NULL && start + count - 1 == end) + res = acpi_alloc_sysres(child, type, rid, start, end, count, flags); + return (res); #else return (bus_generic_alloc_resource(dev, child, type, rid, start, end, count, flags)); --- //depot/projects/pci/sys/dev/acpica/acpivar.h 2011-06-22 16:25:39.000000000 0000 +++ //depot/projects/pci/sys/dev/acpica/acpivar.h 2011-08-02 20:21:42.000000000 0000 @@ -382,6 +382,8 @@ struct resource *res, ACPI_RESOURCE *acpi_res); ACPI_STATUS acpi_parse_resources(device_t dev, ACPI_HANDLE handle, struct acpi_parse_resource_set *set, void *arg); +struct resource *acpi_alloc_sysres(device_t child, int type, int *rid, + u_long start, u_long end, u_long count, u_int flags); /* ACPI event handling */ UINT32 acpi_event_power_button_sleep(void *context); -- John Baldwin