Not-yet-finished ALTQ patch.

Matthew Dillon dillon at apollo.backplane.com
Sun Nov 30 12:24:31 PST 2003


:On Sun, Nov 30, 2003 at 02:20:16AM +0100, Max Laier wrote:
:> My crap newsreader won't let me quote your pgped post :(
:Blaim mutt or something... :)
:
:> Beware, your patchset seems to not only care about ALTQ but changes
:> the drivers very heavily in some cases (seems like a catch up with
:> 4.9, right?) Don't know if it's good to do that in one step as it will
:> make it hard to determine what causes new ill behaviour - ATLQ or the
:> new driver code.
:I have just modified the FreeBSD 4.8 patch to "work" with dragonfly, not
:added anything to the drivers.
:
:> Additionally I'd like to encourage you to use new altq headers. If you
:> take them off kame don't use older than 4 month. For pf I need the
:> mbuf tag stuff in place, which was introduced 4 month ago to kame by a
:> large commit from kjc@ Taking the sys/altq from OpenBSD is an option
:> as well, but at the expense of losing all "/dev/altq" glue and the
:> experimental classifiers such as jobs.
:I will take a look at this one. Just need to get altq to work first.
:
:--=20
:Eirik Nygaard

    Ok, Eirik, if you want this stuff to go in I can fast see it becoming
    a lot more complex... try to split the patch into less significant pieces
    that can be committed incrementally to 'get them out of the way'.  If
    you can submit smaller self contained incremental patch sets (e.g. small
    enough that I can review each one in ~10 minutes), then I will review them
    and you or I can commit them in pieces.  But this is just a bit too
    complex to just commit as is and cross our fingers :-)

						-Matt






More information about the Submit mailing list