On 07/11/2014 06:35 AM, John Ferlan wrote:
Create a separate typedef for the hostdev union data describing PCI. Then adjust the code to use the new pointer
Signed-off-by: John Ferlan <jferlan@redhat.com> ---
Starting to get the rebase conflicts here; so yes, reposting the rest of the series would help review.
src/conf/domain_audit.c | 9 ++++---- src/conf/domain_conf.c | 25 ++++++++++++--------- src/conf/domain_conf.h | 12 ++++++---- src/libxl/libxl_conf.c | 9 ++++---- src/libxl/libxl_domain.c | 5 +++-- src/libxl/libxl_driver.c | 34 ++++++++++++++--------------- src/qemu/qemu_cgroup.c | 24 ++++++++++---------- src/qemu/qemu_command.c | 47 ++++++++++++++++++---------------------- src/qemu/qemu_hotplug.c | 11 +++++----- src/security/security_apparmor.c | 10 ++++----- src/security/security_dac.c | 20 +++++++---------- src/security/security_selinux.c | 20 +++++++---------- src/util/virhostdev.c | 34 ++++++++++++----------------- 13 files changed, 125 insertions(+), 135 deletions(-)
Based solely on visual review for this one:
@@ -10123,10 +10124,13 @@ static int virDomainHostdevMatchSubsysPCI(virDomainHostdevDefPtr a, virDomainHostdevDefPtr b) { - if (a->source.subsys.u.pci.addr.domain == b->source.subsys.u.pci.addr.domain && - a->source.subsys.u.pci.addr.bus == b->source.subsys.u.pci.addr.bus && - a->source.subsys.u.pci.addr.slot == b->source.subsys.u.pci.addr.slot && - a->source.subsys.u.pci.addr.function == b->source.subsys.u.pci.addr.function) + virDomainHostdevSubsysPCIPtr apcisrc = &a->source.subsys.u.pci; + virDomainHostdevSubsysPCIPtr bpcisrc = &b->source.subsys.u.pci;
Similar to 1/8, naming this 'a_pcisrc' or even 'first' might be easier to read. ACK, if the rebase is simple. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org