Re: [PATCH iproute2-next 1/4] man: get rid of doc/actions/mirred-usage

[Jiri Pirko <jiri@resnulli.us>]

Fri, Jan 12, 2024 at 03:55:46PM CET, jhs@mojatatu.com wrote:
>On Fri, Jan 12, 2024 at 7:42 AM Jiri Pirko <jiri@resnulli.us> wrote:
>>
>> Thu, Jan 11, 2024 at 07:44:08PM CET, stephen@networkplumber.org wrote:
>> >The only bit of information not already on the man page
>> >is some of the limitations.
>> >
>>
>> [...]
>>
>> >diff --git a/man/man8/tc-mirred.8 b/man/man8/tc-mirred.8
>> >index 38833b452d92..71f3c93df472 100644
>> >--- a/man/man8/tc-mirred.8
>> >+++ b/man/man8/tc-mirred.8
>> >@@ -94,6 +94,14 @@ interface, it is possible to send ingress traffic through an instance of
>> > .EE
>> > .RE
>> >
>> >+.SH LIMITIATIONS
>> >+It is possible to create loops which will cause the kernel to hang.
>>
>> Hmm, I think this is not true for many many years. Perhaps you can drop
>> it? Anyway, it was a kernel issue.
>
>Hmm back at you: why do you say it is not true anymore? It is still

Ah, I was falsely under impression this happens in reclassify loop.
Nevermind then.


>there - all in the marvelous name of saving 2 bits from the skb.
>If you want to be the hero, here's the last attempt to fix this issue:
>https://lore.kernel.org/netdev/20231215180827.3638838-1-victor@mojatatu.com/#t
>
>Stephen, please cc all the stakeholders when you make these changes.
>Some of us dont have the luxury to be able to scan every message in
>the list. I dont have time, at the moment, to review all the
>documentation you are removing - but if you had Cc me i would have
>made time.
>
>cheers,
>jamal