Re: [PATCH 1/3] selftests/livepatch: Don't clear dmesg when running tests

[Joe Lawrence <joe.lawrence@xxxxxxxxxx>]

On 6/11/20 3:38 AM, Miroslav Benes wrote:
> Hi,
>
> On Wed, 10 Jun 2020, Joe Lawrence wrote:
>
> > diff --git a/tools/testing/selftests/livepatch/test-callbacks.sh b/tools/testing/selftests/livepatch/test-callbacks.sh
> > index 32b57ba07f4f..c3d949da5bb7 100755
> > --- a/tools/testing/selftests/livepatch/test-callbacks.sh
> > +++ b/tools/testing/selftests/livepatch/test-callbacks.sh
> > @@ -12,7 +12,7 @@ MOD_TARGET_BUSY=test_klp_callbacks_busy
> >   setup_config
> >   
> >   
> > -# TEST: target module before livepatch
> > +start_test "target module before livepatch"
> >   #
> >   # Test a combination of loading a kernel module and a livepatch that
> >   # patches a function in the first module.  Load the target module
> > @@ -28,9 +28,6 @@ setup_config
> >   #   unpatching transition starts.  klp_objects are reverted, post-patch
> >   #   callbacks execute and the transition completes.
> >   
> > -echo -n "TEST: target module before livepatch ... "
> > -dmesg -C
> > -
>
> A nit, but I think it would be better to place start_test here below the
> comment. The same for other occurrences in test-callbacks.sh.

The idea was to remove the duplicate # TEST: comment and then the same 
echo -n "TEST: ..." entries.

Would it still look okay if we move start_test to below the comment and 
omit that # TEST ... part?  (This might be what you're suggesting, but I 
wanted to make sure.)

> [...]
>
> > diff --git a/tools/testing/selftests/livepatch/test-state.sh b/tools/testing/selftests/livepatch/test-state.sh
> > index a08212708115..bf8db1613961 100755
> > --- a/tools/testing/selftests/livepatch/test-state.sh
> > +++ b/tools/testing/selftests/livepatch/test-state.sh
> > @@ -10,10 +10,7 @@ MOD_LIVEPATCH3=test_klp_state3
> >   
> >   setup_config
> >   
> > -# TEST: Loading and removing a module that modifies the system state
> > -
> > -echo -n "TEST: system state modification ... "
> > -dmesg -C
> > +start_test "Loading and removing a module that modifies the system state"
>
> start_test should get the message from the original echo command and not
> the comment above, I think...

Yup, good eye.

> >   load_lp $MOD_LIVEPATCH
> >   disable_lp $MOD_LIVEPATCH
> > @@ -41,10 +38,7 @@ livepatch: '$MOD_LIVEPATCH': unpatching complete
> >   % rmmod $MOD_LIVEPATCH"
> >   
> >   
> > -# TEST: Take over system state change by a cumulative patch
> > -
> > -echo -n "TEST: taking over system state modification ... "
> > -dmesg -C
> > +start_test "Take over system state change by a cumulative patch"
> >   
> >   load_lp $MOD_LIVEPATCH
> >   load_lp $MOD_LIVEPATCH2
> > @@ -85,10 +79,7 @@ livepatch: '$MOD_LIVEPATCH2': unpatching complete
> >   % rmmod $MOD_LIVEPATCH2"
> >   
> >   
> > -# TEST: Take over system state change by a cumulative patch
> > -
> > -echo -n "TEST: compatible cumulative livepatches ... "
> > -dmesg -C
> > +start_test "Take over system state change by a cumulative patch"
>
> ...because now we have two "Take over system state change by a cumulative
> patch" tests in the log.

Right, the start_test messages should be unique.  I'll fix those up for v2.

-- Joe