Re: [PATCH 2/8] device tree: correctly ignore unit-address when matching nodes by name

[David Vrabel <david.vrabel@citrix.com>]

On 14/03/12 10:00, Ian Campbell wrote:
> On Tue, 2012-02-28 at 16:54 +0000, David Vrabel wrote:
>> From: David Vrabel <david.vrabel@citrix.com>
>>
>> When matching node by their name, correctly ignore the unit address
>> (@...) part of the name.  Previously, a "memory-controller" node would
>> be incorrectly matched as a "memory" node.
> 
> Are we reinventing this ourselves or is this derived from the Linux (or
> other) version of similar code?

Linux doesn't have an equivalent node_matches() function.  Linux could
do with one as it open-codes a bunch of these comparisons.

>> diff --git a/xen/common/device_tree.c b/xen/common/device_tree.c
>> index d50cb9c..e5df748 100644
>> --- a/xen/common/device_tree.c
>> +++ b/xen/common/device_tree.c
>> @@ -24,6 +24,20 @@
>>  struct dt_early_info __initdata early_info;
>>  void *device_tree_flattened;
>>  
>> +static bool_t __init node_matches(const void *fdt, int node, const char *match)
>> +{
>> +    const char *name;
>> +    size_t len;
>> +
>> +    name = fdt_get_name(fdt, node, NULL);
>> +    len = strlen(match);
>> +
>> +    /* Match both "match" and "match@..." patterns but not
>> +       "match-foo". */
>> +    return strncmp(name, match, len) == 0
>> +        && (name[len] == '@' || name[len] == '\0');
> 
> How can name[len] == '@' when len came from strlen? Wouldn't you need
> some sort of strchr construct?

I can rename len to match_len if you think it makes it clearer.

David