Date: Mon, 13 Sep 2004 15:36:11 -0700 From: Nate Lawson <nate@root.org> To: msch@snafu.de Cc: andre@freebsd.org Subject: Re: ISDN4BSD broken... Message-ID: <414620DB.9070509@root.org> In-Reply-To: <200409121046.02724.msch@snafu.de> References: <200408141854.38477.msch@snafu.de> <200408211927.36948.msch@snafu.de> <200409121046.02724.msch@snafu.de>
next in thread | previous in thread | raw e-mail | index | archive | help
Matthias Schuendehuette wrote: >>>Aug 14 18:19:46 current kernel: FreeBSD 5.2-CURRENT #0: \ >>> Sat Aug 14 17:43:56 CEST 2004 >>>[...] >>>Aug 14 18:19:46 current kernel: ACPI link \_SB_.PCI0.LNKA \ >>> has invalid initial irq 9, ignoring >>>(this is a new message, which doesn't show up with my Aug 11 >>>kernel) [...] >>>Aug 14 18:19:46 current kernel: isic0: [GIANT-LOCKED] >>>Aug 14 18:19:46 current kernel: isic0 at port \ >>> 0x580-0x59f,0x180-0x19f,0x980-0x99f,0xd80-0xd9f \ >>> irq 10 flags 0x3 on isa0 > >>>Any ideas? ACPI or IRQ-Routing related? It's probably IRQ routing related. Try changing the check in acpi_pci_link_is_valid_irq() to this: if (link->interrupts[i] == irq || AcpiGbl_FADT->SciInt == irq) I've suspected that it should always be ok to load a device onto the SCI since Windows does this for many systems. However, this change would break other systems that it's not valid for. This hack should only be done in the case where we're routing the initial IRQ. If the initial irq == the SCI then allow it, otherwise don't allow the SCI as valid. -- Nate
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?414620DB.9070509>