Re: [Qemu-devel] [PATCH] qemu-ga: stub out guest-suspend* for non-linux

[Michael Roth <mdroth@linux.vnet.ibm.com>]

On Fri, Mar 23, 2012 at 03:20:22PM -0300, Luiz Capitulino wrote:
> On Tue, 20 Mar 2012 19:54:09 -0500
> Michael Roth <mdroth@linux.vnet.ibm.com> wrote:
> 
> > This currently breaks the build for BSDs.
> > 
> > Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com>
> > ---
> >  qga/commands-posix.c |   22 ++++++++++++++++++++++
> >  1 files changed, 22 insertions(+), 0 deletions(-)
> > 
> > diff --git a/qga/commands-posix.c b/qga/commands-posix.c
> > index 89dde92..16737d7 100644
> > --- a/qga/commands-posix.c
> > +++ b/qga/commands-posix.c
> > @@ -24,10 +24,12 @@
> >  
> >  #include <sys/types.h>
> >  #include <sys/ioctl.h>
> > +#if defined(__linux__)
> >  #include <ifaddrs.h>
> >  #include <arpa/inet.h>
> >  #include <sys/socket.h>
> >  #include <net/if.h>
> > +#endif
> >  #include <sys/wait.h>
> >  #include "qga/guest-agent-core.h"
> >  #include "qga-qmp-commands.h"
> > @@ -542,6 +544,7 @@ int64_t qmp_guest_fsfreeze_thaw(Error **err)
> >  #define SUSPEND_SUPPORTED 0
> >  #define SUSPEND_NOT_SUPPORTED 1
> 
> Missing the macros and I think reopen_fd_to_null() is missing too, also
> doesn't apply to latest master.
> 

Shouldn't hurt, but yah, leaving them in is a bit sloppy.

Don't see why this wouldn't apply though... are you applying on top of
Michal's patch? This patch was meant to be the guest-suspend counterpart
rather than a standalone fix.

> >  
> > +#if defined(__linux__)
> >  /**
> >   * This function forks twice and the information about the mode support
> >   * status is passed to the qemu-ga process via a pipe.
> > @@ -729,6 +732,25 @@ void qmp_guest_suspend_hybrid(Error **err)
> >      guest_suspend("pm-suspend-hybrid", NULL, err);
> >  }
> >  
> > +#else /* defined(linux) */
> > +
> > +void qmp_guest_suspend_disk(Error **err)
> > +{
> > +    error_set(err, QERR_UNSUPPORTED);
> > +}
> > +
> > +void qmp_guest_suspend_ram(Error **err)
> > +{
> > +    error_set(err, QERR_UNSUPPORTED);
> > +}
> > +
> > +void qmp_guest_suspend_hybrid(Error **err)
> > +{
> > +    error_set(err, QERR_UNSUPPORTED);
> > +}
> > +
> > +#endif
> > +
> >  #if defined(__linux__)
> 
> I think it would be nice to reorganize the functions in this file so that
> we could have only one #if defined(__linux__) (or even split this into two
> files)...
> 

Yah..that's a good idea, since atm we have suspend, network-get-interface, and
fsfreeze. I'll post a standalone patch to move all the linux-only stuff
into a contiguous chunk. Seperate files I think is a bit extreme, since
AFAICT there's no major reason we shouldn't be able to get the first 2 ported,
and I'd prefer that we avoid doing linux-specific implementations in the
future. Script-deployment via guest-file-write/guest-exec is where we wanna go
long-term for stuff like that.

> >  static GuestNetworkInterfaceList *
> >  guest_find_interface(GuestNetworkInterfaceList *head,
>