Re: [PATCH bpf v2 0/2] bpf: fix wrong copied_seq calculation and add tests

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Sat, Dec 14, 2024 at 07:50:37PM +0100, Jakub Sitnicki wrote:
> On Mon, Dec 09, 2024 at 11:27 PM +08, Jiayuan Chen wrote:
> 
> [...]
> 
> > We added test cases for bpf + strparser and separated them from
> > sockmap_basic. This is because we need to add more test cases for
> > strparser in the future.
> >
> > Fixes: e5c6de5fa025 ("bpf, sockmap: Incorrectly handling copied_seq")
> >
> > ---
> 
> I have a question unrelated to the fix itself -
> 
> Are you an active strparser+verdict sockmap user?
> 
> I was wondering if we can deprecate strparser if/when KCM time comes

I am afraid not.

strparser is very different from skb verdict, upper layer (e.g. HTTP)
protocol messages may be splitted accross sendmsg() call's, strparser
is the only place where we can assemble the messages and parse them as a
whole.

And I don't think we have to use KCM together with strparser. Therefore,
even _if_ KCM can be deprecated, strparse still can't.

Regards.




[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux