Re: [Xen-devel] [PATCH 2 of 3] blktap3/libxl: Check whether blktap3 is available

[Ian Jackson <Ian.Jackson@xxxxxxxxxxxxx>]

Thanos Makatos writes ("RE: [Xen-devel] [PATCH 2 of 3] blktap3/libxl: Check 
whether blktap3 is  available"):
> > Thanos Makatos writes ("[Xen-devel] [PATCH 2 of 3] blktap3/libxl: Check
> > whether blktap3 is available"):
> > > This patch implements function libxl__blktap3_enabled, the equivalent
> > > of the existing libxl__blktap_enabled for blktap2. The checks
> > > performed are rather simple and should be extended.
> > ...
> > > +/* FIXME get the tapback name from blktap3 instead of hard-coding */
> > > +#define TAPBACK_NAME "tapback"
> > > +#define CMD "pidof " TAPBACK_NAME
> > 
> > I'm afraid I'm fundamentally unhappy with this approach to detecting
> > the availability of blktap3.  Searching the process table for process
> > with particular names is not a good idea.
> > 
> > Can't you try to connect to its control socket ?
...
> There's no control socket yet. Could we leave this as a future improvement?

How do you control this thing then ?  Does it listen on xenstore ?

In any case, you need to do this test in some way which is reliable.

The reason why searching the process table for processes with
particular names is not a good idea is that there is no rule which
says that other subsystems aren't allowed to create processes with any
names they like.

Ian.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel