On 07/23/2014 10:08 PM, Eric Blake wrote:
> On 07/21/2014 02:47 PM, John Ferlan wrote:
>> Create the structures and API's to hold and manage the iSCSI host device.
>> This extends the 'scsi_host' definitions added in commit id '5c811dce'.
>> A future patch will add the XML parsing, but that code requires some
>> infrastructure to be in place first in order to handle the differences
>> between a 'scsi_host' and an 'iSCSI host' device.
>>
>> Signed-off-by: John Ferlan <jferlan at redhat.com>
>> ---
>> src/conf/domain_audit.c | 20 +++-
>> src/conf/domain_conf.c | 47 ++++++++-
>> src/conf/domain_conf.h | 20 ++++
>> src/qemu/qemu_cgroup.c | 35 ++++---
>> src/qemu/qemu_command.c | 74 ++++++++++++---
>> src/qemu/qemu_hotplug.c | 36 +++++--
>> src/security/security_apparmor.c | 6 ++
>> src/security/security_dac.c | 12 +++
>> src/security/security_selinux.c | 12 +++
>> src/util/virhostdev.c | 200 +++++++++++++++++++++++++--------------
>> 10 files changed, 349 insertions(+), 113 deletions(-)
>>
>
>> static int
>> +virDomainHostdevMatchSubsysSCSIiSCSI(virDomainHostdevDefPtr first,
>> + virDomainHostdevDefPtr second)
>> +{
>> + virDomainHostdevSubsysSCSIiSCSIPtr first_iscsisrc =
>> + &first->source.subsys.u.scsi.u.iscsi;
>> + virDomainHostdevSubsysSCSIiSCSIPtr second_iscsisrc =
>> + &second->source.subsys.u.scsi.u.iscsi;
>> +
>> + if (STREQ(first_iscsisrc->hosts[0].name, second_iscsisrc->hosts[0].name) &&
>
> Do you need to match nhosts, or is nhosts always 1?
>
>
Only 1 host is supported (same as for a disk XML iSCSI resource). This
will be seen in patch 8 in virDomainHostdevSubsysSCSIiSCSIDefParseXML()
John
Going to push the first 6 patches shortly - thanks for the reviews so far.
>> +
>> + /* Only delete the devices which are marked as being used by @name,
>> + * because qemuProcessStart could fail on the half way. */
>
> Sounds funny; maybe "could fail half way through"
>
>
>> -
>> - /* Only delete the devices which are marked as being used by @name,
>> - * because qemuProcessStart could fail on the half way. */
>
> Then again, it's just code motion.
>
> Conditional ACK, if the match function doesn't need to track nhosts.
>
>