On 2/4/24 02:17, Hangbin Liu wrote:

On Fri, Feb 02, 2024 at 09:57:46AM -0800, Kui-Feng Lee wrote:

quoted

quoted

Hi Kui-Feng,

I may missed something. But I still could not get why we shouldn't use
expires for checking? If expires == 0, but RTF_EXPIRES is on,
shouldn't we call fib6_clean_expires()?

The case that expires == 0 and RTF_EXPIES is on never happens since
inet6_addr_modify() rejects valid_lft == 0 at the beginning. This
patch doesn't make difference logically, but make inet6_addr_modify()
and modify_prefix_route() consistent.

Does that make sense to you?

Thanks, this does make sense to me. If there will be a new version. It would
be good to add the following sentence in the description.

"""
This patch doesn't make difference logically, but make inet6_addr_modify()
and modify_prefix_route() consistent.
"""

Sure, I will add it to the commit message.

Reviewed-by: Hangbin Liu <redacted>

Regards
Hangbin