netdev - tc flower offload rule priority

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <CABJS2x7aQaJMC_B34dj+qeiCuAdHheZmSNDpaKpmwpYt4zzstg@mail.gmail.com>
Date:   Tue, 16 Mar 2021 00:19:58 +0530
From:   Naveen Mamindlapalli <naveen130617.lkml@...il.com>
To:     netdev@...r.kernel.org
Subject: tc flower offload rule priority

Hi,

I am new to the tc subsystem and trying to understand tc rule
priorities for hardware offloading. I can see that the tc flower
classifier rules are added with a default priority value of 49152 (low
priority) and then the priority is increased for subsequent rules.
When offloading tc flower rules to HW that supports minimal flow
entries (for example: 128), it will be difficult to maintain the
priorities in the same way tc subsystem expects because there is no
one-to-one mapping between SW and HW.

Is it possible for the NIC driver to modify the default behaviour of
the tc subsystem by using the priority value range specified by the
NIC driver to fit the number of HW entries supported such that
one-to-one mapping can be maintained between tc subsystem priorities
and HW entries?

Thanks,
Naveen