Heya all, Doing a bit more testing with lhcs_regression on i386 this afternoon I found another issue. This is in the SAR test case (hotplug_06), which checks that sar is able to start up and run correctly when cpu's are offlined. If two or more of the system's cpu's are online when sar is started, it runs okay. Offlining all the offlineable cpu's while sar is running works as well - 'nan' is displayed for them, as expected. However, if all but one of the system's cpu's are turned off at the time sar is started, `sar -P ALL 1 0` fails to start, giving the message "Not an SMP machine..." I believe in this case, sar needs to recognize that the system *is* SMP, but just that there are some offlined CPUs. It should start up and run showing the offlined cpu's as 'nan'd. Bryce P.S., I'm about done with an updated version of lhcs_regression to correct for the non-offlineable cpu0 on i386. Just needed some extra sanity checks in there... On Tue, May 02, 2006 at 10:42:21AM -0700, Raj, Ashok wrote: > > > >> The result of the echo command should tell if it succeded or not. > > > >Yes, but from the perspective of a regression test, how do we know that > >it is OK for the echo command to fail? If the on-line file is there, > >the script thinks that it should off-line when we echo zero to it. If > >the echo comes back as failed, without any other information, the > scrips > >would report that as a FAILURE, when it isn't. > > > >So how do we know in user-space what the right response should be? > >> > > Its probably different for each architecture > > In general > > - if no online file, that cpu is not offlinable. > - if offline fails on a certain cpu, all you can do is just report it > CPUX, offline failed. > > So if a online file exist and offline on that cpu fails, unless its > known to fail (say its cpu0, and its known that cpu0 is not removable.. > just report success, since the behaviour expected is cpu0 shouldn't be > removed) > > Like: CPU0: offline failed: Expected: Fail >
