netdev - Re: [PATCH] vmxnet3: cap copy length at size of skb to prevent dropped frames on tx

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

Message-ID: <19987954.351014.1329422822704.JavaMail.root@zimbra-prod-mbox-5.vmware.com>
Date:	Thu, 16 Feb 2012 12:07:02 -0800 (PST)
From:	Shreyas Bhatewara <sbhatewara@...are.com>
To:	Neil Horman <nhorman@...driver.com>
Cc:	Max Matveev <mmatveev@...hat.com>,
	"David S. Miller" <davem@...emloft.net>,
	"VMware, Inc." <pv-drivers@...are.com>, netdev@...r.kernel.org
Subject: Re: [PATCH] vmxnet3: cap copy length at size of skb to prevent
 dropped frames on tx


> Also, since I'm unable to do this, it wuold be great if vmware could
> follow up
> this patch with some additional code commentary as to why 54 bytes is
> an optimal
> pull length for a virtual NIC driver.  The comment that introduced
> this was
> vague on that.  Thanks!


Neil, thanks for the patch. We fixed this bug in local repositories
recently and I was about to post a patch to LKML. As for the explanation,
the vNIC is optimized to work with most frequently occurring headers(TCP)
and hence needs particular number of bytes in first sg. Even in case of
UDP this pull is cheaper (CPU utilization wise) to do in guest than in
device emulation.

Signed-off-by: Shreyas N Bhatewara <sbhatewara@...are.com>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html