On 9/25/19 10:28 AM, Randy MacLeod wrote:
On 9/24/19 9:41 PM, Robert Yang wrote:
Hi Randy,
On 9/25/19 4:46 AM, Randy MacLeod wrote:
On 9/24/19 4:15 AM, Robert Yang wrote:
It checked whether busybox is in pn-buildlist and task-depends.dot after run
"bitbake -g core-image-minimal", but busybox is not a must for
core-image-minimal,
You should explain why busybox is not a must since
usually busybox is the key package in c-i-m.
We have some images that deliberately exclude busybox and
I think this was found by using:
echo "BB_NO_NETWORK = \"0\"" >> conf/local.conf
echo "PNWHITELIST_LAYERS_remove = \"selftest\"" >> conf/local.conf
echo "SANITY_TESTED_DISTROS = \"\"" >> conf/local.conf
echo "DISTRO = \"wrlinux-graphics\"" >> conf/local.conf
echo "VIRTUAL-RUNTIME_no-busybox_forcevariable = \"\"" >> conf/local.conf
echo "PNWHITELIST_meta-layer0 += 'layerappendtest'" >> conf/local.conf
so mention that in general and VIRTUAL-RUNTIME_no-busybox_forcevariable
specifically.
Thanks for the explanation, I think that the maintainers know busybox
is not a must in oe-core, not matter for core-image-minimal or sato,
so I didn't add very detailed explanations.
commit logs are for maintainers, developers and users all
of varying ranges of familiarity with oe-core so
it's better to over-explain, without going over the top, IMO.
so the case may fail when busybox is not present, this case
is used for checking "bitbake -g", so check core-image-minimal itself is more
reliable.
It's true that busybox may not be present but we want to
test that some package is in the image, not just that the
task-depends.dot file exists and mentions c-i-m.
How about searching for 'base-files' ?
The base-files is my first thought, but it's not a must either, for testing
"bitbake -g recipe", I think that check the recipe itself is more reliable than
others.
If you had explained that in the commit log,
I wouldn't have replied, eh? ;-)
Sure base-files might not be part of the image but I
haven't seen an image that has done that. Your check will
always work so it really depends on the intent of the original
check.
Yes, make sense, I will send a V2 to check base-files.
// Robert
../Randy
// Robert
../Randy
Signed-off-by: Robert Yang <[email protected]>
---
meta/lib/oeqa/selftest/cases/bbtests.py | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/meta/lib/oeqa/selftest/cases/bbtests.py
b/meta/lib/oeqa/selftest/cases/bbtests.py
index 8e59baf..4103396 100644
--- a/meta/lib/oeqa/selftest/cases/bbtests.py
+++ b/meta/lib/oeqa/selftest/cases/bbtests.py
@@ -118,11 +118,12 @@ class BitbakeTests(OESelftestTestCase):
self.assertIn(task, result.output, msg="Couldn't find %s task.")
def test_bitbake_g(self):
- result = bitbake('-g core-image-minimal')
+ recipe = 'core-image-minimal'
+ result = bitbake('-g %s' % recipe)
for f in ['pn-buildlist', 'task-depends.dot']:
self.addCleanup(os.remove, f)
self.assertTrue('Task dependencies saved to \'task-depends.dot\''
in result.output, msg = "No task dependency \"task-depends.dot\" file was
generated for the given task target. bitbake output: %s" % result.output)
- self.assertTrue('busybox' in
ftools.read_file(os.path.join(self.builddir, 'task-depends.dot')), msg = "No
\"busybox\" dependency found in task-depends.dot file.")
+ self.assertTrue(recipe in
ftools.read_file(os.path.join(self.builddir, 'task-depends.dot')), msg = "No
\"%s\" dependency found in task-depends.dot file." % recipe)
def test_image_manifest(self):
bitbake('core-image-minimal')
--
_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.openembedded.org/mailman/listinfo/openembedded-core
