On 10/2/23 14:56, Simon Glass wrote: > Hi Sean, > > On Mon, 2 Oct 2023 at 08:38, Sean Anderson <[email protected]> wrote: >> >> On 10/1/23 15:36, Simon Glass wrote: >> > Hi Sean, >> > >> > On Fri, 29 Sept 2023 at 10:12, Sean Anderson <[email protected]> >> > wrote: >> >> >> >> On 9/29/23 12:06, Sean Anderson wrote: >> >> > SPL doesn't have OF_LIVE enabled, so we can only run tests with a flat >> >> > tree. Don't skip them even if they don't use the devicetree. >> >> > >> >> > Fixes: 6ec5178c0ef ("test: Skip flat-tree tests if devicetree is not >> >> > used") >> >> > Signed-off-by: Sean Anderson <[email protected]> >> >> > --- >> >> > >> >> > test/test-main.c | 3 ++- >> >> > 1 file changed, 2 insertions(+), 1 deletion(-) >> >> > >> >> > diff --git a/test/test-main.c b/test/test-main.c >> >> > index 778bf0a18a0..edb20bc4b9c 100644 >> >> > --- a/test/test-main.c >> >> > +++ b/test/test-main.c >> >> > @@ -476,7 +476,8 @@ static int ut_run_test_live_flat(struct >> >> > unit_test_state *uts, >> >> > * (for sandbox we handle this by copying the tree, but not for >> >> > other >> >> > * boards) >> >> > */ >> >> > - if ((test->flags & UT_TESTF_SCAN_FDT) && >> >> > + if ((!CONFIG_IS_ENABLED(OF_LIVE) || >> >> > + (test->flags & UT_TESTF_SCAN_FDT)) && >> >> > !(test->flags & UT_TESTF_LIVE_TREE) && >> >> > (CONFIG_IS_ENABLED(OFNODE_MULTI_TREE) || >> >> > !(test->flags & UT_TESTF_OTHER_FDT)) && >> >> >> >> Upon further review, do we even need 6ec5178c0ef in the first place? >> >> ut_test_run_on_flattree looks like it's trying to do the same thing. >> > >> > Well one problem is that many tests are not run at all unless OF_LIVE >> > is enabled. The code as is is assuming that OF_LIVE is active. >> > >> > On boards where OF_LIVE is not active, many tests won't run at all >> > unless they are marked with UT_TESTF_SCAN_FDT. >> > >> > So I think that UT_TESTF_SCAN_FDT line needs to be removed. >> >> OK, so to clarify, since 6ec5178c0ef added that UT_TESTF_SCAN_FDT, you would >> like to >> revert that commit? > > Yes, I think that will work...but just check that tests without the > UT_TESTF_SCAN_FDT flag don't then run twice with sandbox. There was > perhaps something else wrong at the time.
Actually, upon further review, I think that the above patch is correct. A revert would cause tests with UT_TESTF_DM but without UT_TESTF_SCAN_FDT to run twice. --Sean
