Hi Vincent,
On 11/18/19 7:28 AM, Vincent Whitchurch wrote:
> No changeset entries are created for #address-cells and #size-cells
> properties, but the duplicated properies are never freed. This results
> in a memory leak which is detected by kmemleak:
>
> unreferenced object 0x85887180 (size 64):
> backtrace:
> kmem_cache_alloc_trace+0x1fb/0x1fc
> __of_prop_dup+0x25/0x7c
> add_changeset_property+0x17f/0x370
> build_changeset_next_level+0x29/0x20c
> of_overlay_fdt_apply+0x32b/0x6b4
> ...
>
> Fixes: 6f75118800acf77f8 ("of: overlay: validate overlay properties #address-cells and #size-cells")
> Signed-off-by: Vincent Whitchurch <vincent.whitchurch@xxxxxxxx>
> ---
> drivers/of/overlay.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
> index c423e94baf0f..5f8869e2a8b3 100644
> --- a/drivers/of/overlay.c
> +++ b/drivers/of/overlay.c
> @@ -360,7 +360,7 @@ static int add_changeset_property(struct overlay_changeset *ovcs,
> pr_err("WARNING: memory leak will occur if overlay removed, property: %pOF/%s\n",
> target->np, new_prop->name);
>
> - if (ret) {
> + if (ret || !check_for_non_overlay_node) {
> kfree(new_prop->name);
> kfree(new_prop->value);
> kfree(new_prop);
>
Thanks for finding and proposing a fix for the memory leak.
The proposed patch conveniently uses check_for_non_overlay_node
which leads to a nice small patch. But ends up adding an
additional hidden meaning to the variable, resulting in more
fragile code.
I will propose a different solution and ask you to test it
to make sure it also solves the memory leak.
-Frank
