[libvirt] [PATCH v2 1/3] virDomainGetDiskErrors public API and remote protocol

[Jiri Denemark]

On Mon, Jan 30, 2012 at 15:56:11 -0700, Eric Blake wrote:
> On 01/30/2012 09:00 AM, Jiri Denemark wrote:
> > We already provide ways to detect when a domain has been paused as a
> > result of I/O error, but there was no way of getting the exact error or
> > even the device that experienced it.  This new API may be used for both.
> > ---
> >  daemon/remote.c              |   40 ++++++++++++++++++++++++++++++++
> >  include/libvirt/libvirt.h.in |   32 +++++++++++++++++++++++++
> >  python/generator.py          |    3 +-
> >  src/driver.h                 |    7 +++++
> >  src/libvirt.c                |   52 ++++++++++++++++++++++++++++++++++++++++++
> >  src/libvirt_public.syms      |    1 +
> >  src/remote/remote_driver.c   |   34 +++++++++++++++++++++++++++
> >  src/remote/remote_protocol.x |   22 +++++++++++++++++-
> >  src/remote_protocol-structs  |   16 +++++++++++++
> >  src/rpc/gendispatch.pl       |   47 +++++++++++++++++++++++++++++++++++++
> >  10 files changed, 252 insertions(+), 2 deletions(-)
> 
> This is big enough that I might have split it into public API (include,
> python, src/{driver.h,libvirt.c,libvirt_public.syms}) and RPC
> implementation (daemon, src/remote, src/rpc).  But I'll go ahead and
> review this, whether or not you split it.

I did so and addressed the other comments in v3.

...
> >  /*----- Protocol. -----*/
> >  
> > @@ -2708,7 +2727,8 @@ enum remote_procedure {
> >      REMOTE_PROC_STORAGE_VOL_RESIZE = 260, /* autogen autogen */
> >  
> >      REMOTE_PROC_DOMAIN_PM_SUSPEND_FOR_DURATION = 261, /* autogen autogen */
> > -    REMOTE_PROC_DOMAIN_GET_CPU_STATS = 262 /* skipgen skipgen */
> > +    REMOTE_PROC_DOMAIN_GET_CPU_STATS = 262, /* skipgen skipgen */
> > +    REMOTE_PROC_DOMAIN_GET_DISK_ERRORS = 263 /* autogen autogen */
> 
> Nice that you figured out how to autogen the serialize/deserialize
> calls.

Yeah, although I didn't try so hard after making this a multi-return call, so
v3 now uses skipgen skipgen :-)

> We should probably do the same for a lot of the virTypedParameter clients
> someday
The generator already seems to know how to deal with virTypedParameter but
only for APIs with only one item in their *_ret structure.

Jirka