On 2/13/24 09:37, Adrian Moreno wrote:
>
>
> On 2/13/24 09:32, Adrian Moreno wrote:
>>
>>
>> On 2/9/24 17:17, Ilya Maximets wrote:
>>> For some reason annotation is made for a read-lock, while all the
>>> callers are correctly holding a write-lock.
>>>
>>> Fixes: 05df16238d43 ("ofproto/bond: Fix bond post recirc rule leak.")
>>> Signed-off-by: Ilya Maximets <[email protected]>
>>> ---
>>> ofproto/bond.c | 2 +-
>>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>>
>>> diff --git a/ofproto/bond.c b/ofproto/bond.c
>>> index cfdf44f85..c4b3a4a45 100644
>>> --- a/ofproto/bond.c
>>> +++ b/ofproto/bond.c
>>> @@ -423,7 +423,7 @@ update_recirc_rules__(struct bond *bond)
>>> static void
>>> update_recirc_rules(struct bond *bond)
>>> - OVS_REQ_RDLOCK(rwlock)
>>> + OVS_REQ_WRLOCK(rwlock)
>>> {
>>> update_recirc_rules__(bond);
>>> }
>>
>> Looks good to me.
>>
>> Acked-by: Adrian Moreno <[email protected]>
>
> Sorry, thought of something after sending the ack.
>
> If, as we're discussing in [1], we lock the global mutex in bond_unref(),
> there
> should not be any reason to keep the lock-less version of this function
> (update_recirc_rules__()) so we could fold both functions together.
>
> Unless you see a reason not to backport [1] to the same degree as this patch,
> I
> could merge this patch with my v2 of [1].
>
> WDYT?
Yeah, sure. If you can just incorporate that change into your patch
and get rid of the lock-less function entirely, that will be better.
Thanks!
Best regards, Ilya Maximets.
>
> [1]
> https://patchwork.ozlabs.org/project/openvswitch/patch/[email protected]/
>
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
