All of lore.kernel.org
 help / color / mirror / Atom feed
From: Phil Sutter <phil@nwl.cc>
To: lartc@vger.kernel.org
Subject: Various questions regarding action parameters
Date: Fri, 29 Jan 2016 17:07:51 +0000	[thread overview]
Message-ID: <20160129170751.GD13110@orbyte.nwl.cc> (raw)

Hi,

I'm in the process of writing man pages for tc actions. Sadly I wasn't
able to figure everything out by searching the web, grepping Git log or
reading kernel/iproute2 code. Therefore I present a number of open
questions here, hoping for you to help me work things out. Please feel
free to answer selected questions as well, any input is highly
appreciated.

connmark action
---------------
I struggle with the effect of the different branch targets. My
understanding so far is:

reclassify - restart classification by jumping back to the first filter
             on the same level.
pipe       - continue with the next action in line.
drop/shot  - drop packet without calling any further filters/actions

Is this correct so far? Could you please explain 'continue' and 'pass'
branches (and what distinguishes each from the others if questionable)?

Further, this action supports an 'index' option which seems to specify
the index at which this action is put into the list of actions attached
to the same filter. Is this correct? So is it just there as an
alternative to reordering the action list upon tc invocation? Or are
there other uses of that index?

police action
-------------
OK, this is lame: I failed to find out what distinguishes rate, peakrate
and avrate and what influence burst and mtu values have. That cell size
parameter to the latter was also completely obscure to me. I could
probably steal from tc-htb.8 for most of them, but writing things
without understanding them is usually not a good idea (actually
remembers me of some of my old professors at university ;).
Could you maybe point me to some documentation which applies to police
as well? Even though I finally understood what the rate precalculation
does, the respective kernel code looked pretty weird to me.

OK, that's it for now. (At least these are my 'XXX' entries right now,
review will probably arise more questions later.)

Thanks, Phil

             reply	other threads:[~2016-01-29 17:07 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-01-29 17:07 Phil Sutter [this message]
2016-01-29 19:14 ` Various questions regarding action parameters Andrew Beverley
2016-02-07 16:48 ` Martin A. Brown

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20160129170751.GD13110@orbyte.nwl.cc \
    --to=phil@nwl.cc \
    --cc=lartc@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.