linux.git/drivers/of/dynamic.c, branch v6.3-rc2

of: dynamic: Fix spelling mistake "kojbect" -> "kobject"

2023-02-21T16:22:04+00:00

There is a spelling mistake in a pr_err message. Fix it.

Signed-off-by: Colin Ian King 
Reviewed-by: Frank Rowand 
Link: https://lore.kernel.org/r/20230220144422.873356-1-colin.i.king@gmail.com
Signed-off-by: Rob Herring

of: dynamic: add lifecycle docbook info to node creation functions

2023-02-20T21:37:19+00:00

The existing docbook comments for the functions related to creating
a devicetree node do not explain the reference count of a newly
created node, how decrementing the reference count to zero will
free the associated memory, and the caller's responsibility to
call of_node_put() on the node.  Explain what happens when the
reference count is decremented to zero.

Signed-off-by: Frank Rowand 
Link: https://lore.kernel.org/r/20230213185702.395776-8-frowand.list@gmail.com
Signed-off-by: Rob Herring

of: add consistency check to of_node_release()

2023-02-20T21:37:19+00:00

Add an additional consistency check to of_node_release(), which is
called when the reference count of a devicetree node is decremented
to zero.  The node's children should have been deleted before the
node is deleted so check that no children exist.

Signed-off-by: Frank Rowand 
Link: https://lore.kernel.org/r/20230213185702.395776-7-frowand.list@gmail.com
Signed-off-by: Rob Herring

of: do not use "%pOF" printk format on node with refcount of zero

2023-02-20T21:37:19+00:00

of_node_release() can not use the "%pOF" printk format to report
the node name of a node when the node reference count is zero.
This is because the formatter device_node_string() calls
fwnode_full_name_string() which indirectly calls of_node_get().
Calling of_node_get() on the node with a zero reference count
results in a WARNING and stack trace.

When the reference count has been decremented to zero, this function
is in the subsequent call path which frees memory related to the node.

This commit resolves the unittest EXPECT errors that were created in
the previous commmit.

Signed-off-by: Frank Rowand 
Link: https://lore.kernel.org/r/20230213185702.395776-6-frowand.list@gmail.com
Signed-off-by: Rob Herring

of: unittest: add node lifecycle tests

2023-02-20T21:36:51+00:00

Add tests to exercise the actions that occur when the reference count
of devicetree nodes decrement to zero and beyond.  Decrementing to
zero triggers freeing memory allocated for the node.

This commit will expose a pr_err() issue in of_node_release(), resulting
in some kernal warnings and stack traces.

When scripts/dtc/of_unittest_expect processes the console messages,
it will also report related problems for EXPECT messages due to the
pr_err() issue:
   **     missing EXPECT begin      :    5

Signed-off-by: Frank Rowand 
Link: https://lore.kernel.org/r/20230213185702.395776-5-frowand.list@gmail.com
[robh: Fix !CONFIG_OF_DYNAMIC build]
Signed-off-by: Rob Herring

of: Add missing 'Return' section in kerneldoc comments

2021-03-27T18:59:03+00:00

Many of the DT kerneldoc comments are lacking a 'Return' section. Let's
add the section in cases we have a description of return values. There's
still some cases where the return values are not documented.

Cc: Frank Rowand 
Cc: Mauro Carvalho Chehab 
Signed-off-by: Rob Herring 
Reviewed-by: Mauro Carvalho Chehab 
Link: https://lore.kernel.org/r/20210325164713.1296407-8-robh@kernel.org

of: dynamic: Fix incorrect parameter name and provide missing descriptions

2021-03-23T21:27:52+00:00

Fixes the following W=1 kernel build warning(s):

 drivers/of/dynamic.c:234: warning: Function parameter or member 'np' not described in 'of_attach_node'
 drivers/of/dynamic.c:286: warning: Function parameter or member 'np' not described in 'of_detach_node'
 drivers/of/dynamic.c:326: warning: Function parameter or member 'kobj' not described in 'of_node_release'
 drivers/of/dynamic.c:326: warning: Excess function parameter 'kref' description in 'of_node_release'

Cc: Rob Herring 
Cc: Frank Rowand 
Cc: devicetree@vger.kernel.org
Signed-off-by: Lee Jones 
Signed-off-by: Rob Herring 
Link: https://lore.kernel.org/r/20210318104036.3175910-3-lee.jones@linaro.org

driver core: Add fwnode link support

2020-12-09T18:10:21+00:00

Add support for creating supplier-consumer links between fwnodes.  It is
intended for internal use the driver core and generic firmware support
code (eg. Device Tree, ACPI), so it is simple by design and the API
provided is limited.

Acked-by: Rob Herring 
Signed-off-by: Saravana Kannan 
Link: https://lore.kernel.org/r/20201121020232.908850-9-saravanak@google.com
Signed-off-by: Greg Kroah-Hartman

of: of_detach_node() remove unneeded local return variable

2020-04-14T12:59:54+00:00

rc is unneeded, just return 0.

Signed-off-by: Ding Xiang 
Signed-off-by: Rob Herring

of: Rework and simplify phandle cache to use a fixed size

2019-12-24T21:17:52+00:00

The phandle cache was added to speed up of_find_node_by_phandle() by
avoiding walking the whole DT to find a matching phandle. The
implementation has several shortcomings:

  - The cache is designed to work on a linear set of phandle values.
    This is true for dtc generated DTs, but not for other cases such as
    Power.
  - The cache isn't enabled until of_core_init() and a typical system
    may see hundreds of calls to of_find_node_by_phandle() before that
    point.
  - The cache is freed and re-allocated when the number of phandles
    changes.
  - It takes a raw spinlock around a memory allocation which breaks on
    RT.

Change the implementation to a fixed size and use hash_32() as the
cache index. This greatly simplifies the implementation. It avoids
the need for any re-alloc of the cache and taking a reference on nodes
in the cache. We only have a single source of removing cache entries
which is of_detach_node().

Using hash_32() removes any assumption on phandle values improving
the hit rate for non-linear phandle values. The effect on linear values
using hash_32() is about a 10% collision. The chances of thrashing on
colliding values seems to be low.

To compare performance, I used a RK3399 board which is a pretty typical
system. I found that just measuring boot time as done previously is
noisy and may be impacted by other things. Also bringing up secondary
cores causes some issues with measuring, so I booted with 'nr_cpus=1'.
With no caching, calls to of_find_node_by_phandle() take about 20124 us
for 1248 calls. There's an additional 288 calls before time keeping is
up. Using the average time per hit/miss with the cache, we can calculate
these calls to take 690 us (277 hit / 11 miss) with a 128 entry cache
and 13319 us with no cache or an uninitialized cache.

Comparing the 3 implementations the time spent in
of_find_node_by_phandle() is:

no cache:        20124 us (+ 13319 us)
128 entry cache:  5134 us (+ 690 us)
current cache:     819 us (+ 13319 us)

We could move the allocation of the cache earlier to improve the
current cache, but that just further complicates the situation as it
needs to be after slab is up, so we can't do it when unflattening (which
uses memblock).

Reported-by: Sebastian Andrzej Siewior 
Cc: Michael Ellerman 
Cc: Segher Boessenkool 
Cc: Frank Rowand 
Acked-by: Sebastian Andrzej Siewior 
Reviewed-by: Frank Rowand 
Tested-by: Frank Rowand 
Signed-off-by: Rob Herring