On 06/20/18 11:23, Rob Herring wrote:
> On Sun, Jun 17, 2018 at 10:03 AM, <frowand.list@xxxxxxxxx> wrote:
>> From: Frank Rowand <frank.rowand@xxxxxxxx>
>>
>> A comment in the review of the patch adding the phandle cache said that
>> the cache would have to be updated when modules are applied and removed.
>> This patch implements the cache updates.
>>
>> Fixes: 0b3ce78e90fc ("of: cache phandle nodes to reduce cost of of_find_node_by_phandle()")
>> Reported-by: Alan Tull <atull@xxxxxxxxxx>
>> Suggested-by: Alan Tull <atull@xxxxxxxxxx>
>> Signed-off-by: Frank Rowand <frank.rowand@xxxxxxxx>
>> ---
>>
>> Compiles for one configuration.
>> NOT boot tested.
>> Not run through my normal process to check for new warnings, etc.
>
> I'm assuming you will resend a non-RFC version for me to apply.
Yes, I will.
>
> I think it would be a bit better if callers didn't have to do free and
> populate themselves, but just made an invalidate call (like a normal
> cache) and re-populating the cache could happen on demand. Or if it
> was done as a single call, you could just copy the old entries to the
> new larger array. But maybe there would be a race condition in doing
> that? In any case, all that could be a subsequent patch.
Yes, the unspoken, underlying issue is a race condition. I'll update
the commit comment to explain the race issues a little bit. And maybe
add a code comment if I can be concise enough.
-Frank
>
> Rob
>
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
