On Tue, Aug 22, 2023 at 05:15:14PM +0200, Thomas Haller wrote: > On Tue, 2023-08-22 at 13:39 +0200, Thomas Haller wrote: > > On Tue, 2023-08-22 at 10:54 +0200, Pablo Neira Ayuso wrote: > > > > > > nftables calls localtime_r() from print/parse functions. Presumably, > > we > > will print/parse several timestamps during a larger operation, it > > would > > be odd to change/reload the timezone in between or to meaningfully > > support that. > > > > > > I think it is all good, nothing to change. Just to be aware of. > > > > Thinking some more, the "problem" is that when we parse a larger data, > then multiple subfields are parsed. Thereby we call "time()" and > "localtime()" multiple times. The time() keeps ticking, and time and tz > can be reset at any moment -- so we see different time/tz, in the > middle of parsing the larger set of data. > > What IMO should happen, is that for one parse operation, we call such > operations at most once, and cache them in `struct netlink_parse_ctx`. > > Is that considered a problem to be solved? Seems simple. Would you > accept a patch for that? Caching this information in the context should be fine, and it might speed up things for a large batch? How complicate will the update look like?
