From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1O1tOR-0006af-0F for qemu-devel@nongnu.org; Tue, 13 Apr 2010 23:35:59 -0400 Received: from [140.186.70.92] (port=51028 helo=eggs.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1O1tOM-0006aX-RK for qemu-devel@nongnu.org; Tue, 13 Apr 2010 23:35:57 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.69) (envelope-from ) id 1O1tOL-0001Uf-1Z for qemu-devel@nongnu.org; Tue, 13 Apr 2010 23:35:54 -0400 Received: from sj-iport-6.cisco.com ([171.71.176.117]:4035) by eggs.gnu.org with esmtp (Exim 4.69) (envelope-from ) id 1O1tOK-0001Mq-SB for qemu-devel@nongnu.org; Tue, 13 Apr 2010 23:35:52 -0400 Message-ID: <4BC53816.8060401@cisco.com> Date: Tue, 13 Apr 2010 21:35:50 -0600 From: "David S. Ahern" MIME-Version: 1.0 Subject: Re: [Qemu-devel] Re: ehci update References: <4BC3F3FC.6050809@cisco.com> <4BC4FFC8.3060702@web.de> <4BC50337.8040201@cisco.com> <73AB16E6-5B01-4848-B8B7-895AEC5DAF76@suse.de> In-Reply-To: <73AB16E6-5B01-4848-B8B7-895AEC5DAF76@suse.de> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Alexander Graf Cc: Jan Kiszka , qemu-devel@nongnu.org On 04/13/2010 07:20 PM, Alexander Graf wrote: >>>> It still needs a lot of love, but definitely an improvement from the >>>> last version. The biggest difference for the performance boost and >>>> stability is discovering that the usbfs in linux limits transactions to >>>> 16k versus the EHCI spec which allows 20k per qTD. I added a hack to >>>> submit which detects 20k requests from a guest and breaks it up into 2 >>>> requests through the host (a 16k and then a 4k). >>> >>> Did someone already bring this up on LKML or wherever usbfs is >>> discussed? Should be fixable, I naively guess. >> >> I submitted the patch to linux-usb and it was nack'ed. The response was >> that memory is allocated in powers of 2 so trying to up the limit from >> 16k to 20k means it will actually want to find 32k of contiguous memory. >> The suggestion was to handle it with multiple requests within qemu. I >> guess libusb does that. > > Any reason we're not using libusb? Good question. I was wondering the same. I was going to look at converting usb-linux to use libusb1 when I get some time. David > > > Alex >