On 01/13/2011 06:04 AM, Jiri Denemark wrote:
--- src/qemu/qemu_capabilities.c | 58 +++++++++++++++++++---------------------- 1 files changed, 27 insertions(+), 31 deletions(-)
diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 3d10b42..0fb6ec9 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -1087,48 +1087,44 @@ fail: return -1; }
+ +static char * +qemuCapsExtractDeviceProps(const char *qemu, + const char *device) +{ + virCommandPtr cmd = virCommandNew(qemu); + char *output = NULL; + + virCommandAddArg(cmd, "-device"); + virCommandAddArgFormat(cmd, "%s,?", device);
Phooey - I have a patch in my smartcard series that conflicts with this. But I like mine better - I added testsuite coverage of device parsing, and have only 1 qemu run for all devices rather than 1 run per device. I guess I'd better clean it up and post right away, to compare our two approaches. -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org