netdev - RE: [PATCH 4/4] asix: C1 DUB-E100 double rx_urb

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <AE90C24D6B3A694183C094C60CF0A2F6026B7481@saturn3.aculab.com>
Date:	Wed, 11 Dec 2013 16:28:50 -0000
From:	"David Laight" <David.Laight@...LAB.COM>
To:	"Eric Dumazet" <eric.dumazet@...il.com>
Cc:	"Dean Jenkins" <Dean_Jenkins@...tor.com>, <netdev@...r.kernel.org>,
	<davem@...emloft.net>
Subject: RE: [PATCH 4/4] asix: C1 DUB-E100 double rx_urb_size to 4096

> From: Eric Dumazet
...
> I fear that doubling rx_urb_size is artificially doubling the skb
> truesize (or not, and this would be a bug), and this has
> an impact on general performance of TCP stack.

Which means that the code in the ax179_178a driver is really horrid.
It causes 20kB URB to be passed to the xhci driver.

I know that hardware supports tcp transmit segmentation offload.
I don't know it if supports receive offload.
Best way to test is probably running at Ge speeds at usb2 speeds
- connect to an xhci port with a USB2 extension cable.

I'm waiting for my existing xhci patches to get processed before
I write any more.

	David