On Fri, Jul 3, 2009 at 2:01 AM, Matthew Dillon<dillon@apollo.backplane.com> wrote:It shows ICU works. I think before bringing the ACPI code, we have to change our APIC interrupt handling code: FreeBSD uses ACPI GSI (ACPI3.0a, page120) concept directly (I think they apply it to the MP table too). So more than 24 idt vectors may be used, if, for example, there are two 24pin I/O APICs. As far as I understand our apic_abi.c, we only support 24 idt vectors; and the gd_[if]pending limit us to 32 vectors. Before we could sort that out or at least introduce another mapping layer, bringing in ACPI routing code from FreeBSD directly will cause trouble, IMHO. As about the new PCI code, I think we could bring it, but to compat with our APIC_IO, we probably need to put back the old APIC_IO stuffs (e.g. config I/O APIC pins during pci_cfgread(intline)) Best Regards, sephe -- Live Free or Die
| Hiten Pandya | Re: up? (emacs docbook xml ide) |
| Tarkan Erimer | Re: Dual-Licensing Linux Kernel with GPL V2 and GPL V3 |
| Greg Kroah-Hartman | [PATCH 005/196] Chinese: add translation of SubmittingDrivers |
| James Bottomley | Re: [Ksummit-2008-discuss] Fixing the Kernel Janitors project |
git: | |
| Gerrit Renker | [PATCH 27/37] dccp: Integration of dynamic feature activation - part 2 (server side) |
| David Miller | [GIT]: Networking |
| Jarek Poplawski | Re: [PATCH] pkt_sched: Destroy gen estimators under rtnl_lock(). |
