Date: Thu, 18 Nov 2004 00:43:36 +0200 From: Ruslan Ermilov <ru@FreeBSD.org> To: John-Mark Gurney <jmg@FreeBSD.org>, Bruce Evans <bde@FreeBSD.org> Cc: cvs-all@FreeBSD.org Subject: Re: cvs commit: src/sys/pci if_sk.c Message-ID: <20041117224336.GC1562@ip.net.ua> In-Reply-To: <200411172135.iAHLZNfR098453@repoman.freebsd.org> References: <200411172135.iAHLZNfR098453@repoman.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
[-- Attachment #1 --]
On Wed, Nov 17, 2004 at 09:35:23PM +0000, John-Mark Gurney wrote:
> jmg 2004-11-17 21:35:22 UTC
>
> FreeBSD src repository
>
> Modified files:
> sys/pci if_sk.c
> Log:
> only clear the IFF_OACTIVE flag when we have a chance of being able to
> queue a packet to the hardware... instead of when the hardware queue is
> empty..
>
> don't initalize cur_tx now that it doesn't need to be...
>
> Pointed out by: bde
>
> Revision Changes Path
> 1.92 +4 -4 src/sys/pci/if_sk.c
>
Wrong wording. The correct wording would be:
> clear the IFF_OACTIVE flag AS SOON AS we have a chance of being
> able to queue a packet to the hardware... instead of ONLY when
> the hardware TX queue is empty..
My old patch (that I didn't commit due to the lack of h/w),
resets IFF_OACTIVE inside the while() loop, like this:
sc_if->sk_cdata.sk_tx_cnt--;
+ ifp->if_flags &= ~IFF_OACTIVE;
SK_INC(idx, SK_TX_RING_CNT);
I believe this is more correct than what got committed,
but I cannot prove it. ;)
Cheers,
--
Ruslan Ermilov
ru@FreeBSD.org
FreeBSD committer
[-- Attachment #2 --]
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.6 (FreeBSD)
iD8DBQFBm9QYqRfpzJluFF4RAkgZAJ9GdwW0oDSyNwvU9ePO+XpdKyUCRACfZUYd
W2kzqba+Zlz1QrH+EXExDEI=
=4H9x
-----END PGP SIGNATURE-----
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20041117224336.GC1562>