[PATCH 1/2] Handle failures to instantiate storage devices when parsing kickstart.

Brian C. Lane bcl at redhat.com
Wed Sep 24 01:05:16 UTC 2014


On Tue, Sep 23, 2014 at 02:02:52PM -0400, Anne Mulhern wrote:
> > @@ -469,7 +468,7 @@ class BTRFSData(commands.btrfs.F17_BTRFSData):
> >                                         metaDataLevel=self.metaDataLevel,
> >                                         dataLevel=self.dataLevel,
> >                                         parents=members)
> > -            except blivet.errors.BTRFSValueError as e:
> > +            except (StorageError, ValueError) as e:
> >                  raise KickstartValueError(formatErrorMsg(self.lineno,
> >                  msg=e.message))
> I introduced BTRFSValueError in part so we could distinguish between
> us stupidly dividing by zero and the kickstart asking for unacceptable
> metadata.
> 
> With this change we lose the distinction, which feels like a step back.

Won't there still be that distinction? BTRFSValueError is a subclass of
StorageError so it will still be caught here.

-- 
Brian C. Lane | Anaconda Team | IRC: bcl #anaconda | Port Orchard, WA (PST8PDT)


More information about the anaconda-patches mailing list