The previous commit, "of/unittest: early return from test skips tests"
exposed broken tests for the of_platform_unpopulate() function. The
problem was the populate and depopulate calls were not symmetrical like
they were intended to be, and unpopulate depends on the parent device to
have it's of_node pointer pointing to the parent device node. Fix these
bugs so that the test case works correctly.
In the process, the test_bus used as a container for the test devices
has been changed from a statically allocated struct device (which is
bad) to a properly allocated device with a .release() method (which is
good). This stops the test code from being a bad example of abusing the
device model.
Signed-off-by: Grant Likely <grant.lik...@linaro.org>
Cc: Frank Rowand <frank.row...@sonymobile.com>
Cc: Pantelis Antoniou <pantelis.anton...@konsulko.com>
Cc: Pawel Moll <pawel.m...@arm.com>
---
(I've already got this commit in my tree queued up after the patch that
exposes the bug, which is why I've written the description the way I
did.)
drivers/of/unittest.c | 16 +++++++++-------
1 file changed, 9 insertions(+), 7 deletions(-)
diff --git a/drivers/of/unittest.c b/drivers/of/unittest.c
index 3b155774dfee..a4f90c238e78 100644
--- a/drivers/of/unittest.c
+++ b/drivers/of/unittest.c
@@ -751,14 +751,14 @@ static void __init of_selftest_match_node(void)
}
}
-struct device test_bus = {
- .init_name = "unittest-bus",
+static const struct platform_device_info test_bus_info = {
+ .name = "unittest-bus",
};
static void __init of_selftest_platform_populate(void)
{
int irq, rc;
struct device_node *np, *child, *grandchild;
- struct platform_device *pdev;
+ struct platform_device *pdev, *test_bus;
const struct of_device_id match[] = {
{ .compatible = "test-device", },
{}
@@ -787,20 +787,22 @@ static void __init of_selftest_platform_populate(void)
if (!np)
return;
- rc = device_register(&test_bus);
+ test_bus = platform_device_register_full(&test_bus_info);
+ rc = PTR_ERR_OR_ZERO(test_bus);
selftest(!rc, "testbus registration failed; rc=%i\n", rc);
if (rc)
return;
+ test_bus->dev.of_node = np;
+ of_platform_populate(np, match, NULL, &test_bus->dev);
for_each_child_of_node(np, child) {
- of_platform_populate(child, match, NULL, &test_bus);
for_each_child_of_node(child, grandchild)
selftest(of_find_device_by_node(grandchild),
"Could not create device for node '%s'\n",
grandchild->name);
}
- of_platform_depopulate(&test_bus);
+ of_platform_depopulate(&test_bus->dev);
for_each_child_of_node(np, child) {
for_each_child_of_node(child, grandchild)
selftest(!of_find_device_by_node(grandchild),
@@ -808,7 +810,7 @@ static void __init of_selftest_platform_populate(void)
grandchild->name);
}
- device_unregister(&test_bus);
+ platform_device_unregister(test_bus);
of_node_put(np);
}
--
2.1.0
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
