On 2012-06-05 02:58, Anthony Liguori wrote: > On 06/05/2012 08:54 AM, Andreas Färber wrote: >> Am 05.06.2012 01:33, schrieb q...@buildbot.b1-systems.de: >>> The Buildbot has detected a new failure on builder default_ppc while >>> building qemu. >>> Full details are available at: >>> http://buildbot.b1-systems.de/qemu/builders/default_ppc/builds/417 >>> >>> Buildbot URL: http://buildbot.b1-systems.de/qemu/ >>> >>> Buildslave for this Build: qemu-ppc.opensuse.org >>> >>> Build Reason: The Nightly scheduler named 'nightly_default' triggered >>> this build >>> Build Source Stamp: [branch master] HEAD >>> Blamelist: >>> >>> BUILD FAILED: failed compile >> >> CC ppc-softmmu/kvm-all.o >> /home/build/qemu/default_ppc/build/kvm-all.c:66: error: field 'kroute' >> has incomplete type >> /home/build/qemu/default_ppc/build/kvm-all.c: In function 'kvm_init': >> /home/build/qemu/default_ppc/build/kvm-all.c:1289: error: 'KVMState' has >> no member named 'direct_msi' >> make[1]: *** [kvm-all.o] Error 1 >> make: *** [subdir-ppc-softmmu] Error 2 > > Aren't GSI and even MSI very x86-centric concepts? Why is this code in > kvm-all.c to begin with?
MSI is a PCI thing, and GSI in the context of KVM's in-kernel irqchip is mostly a historic misnomer. It should have been called VIRQ (virtual IRQ), and I'm using this term now wherever possible. Any architecture that will also support in-kernel irqchips with line-based IRQs or irqfd will have to support this concept as well, thus will most probably use the generic code in kvm-all.c. Jan
signature.asc
Description: OpenPGP digital signature