Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 03 Mar 2015 12:23:38 +0100
From:      Hans Petter Selasky <hselasky@freebsd.org>
To:        Marius Strobl <marius@alchemy.franken.de>
Cc:        svn-src-head@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org
Subject:   Re: svn commit: r279563 - head/sys/dev/usb/controller
Message-ID:  <54F599BA.5060708@freebsd.org>
In-Reply-To: <20150303110104.GA65391@alchemy.franken.de>
References:  <201503031021.t23ALtIK094082@svn.freebsd.org> <20150303110104.GA65391@alchemy.franken.de>

next in thread | previous in thread | raw e-mail | index | archive | help
On 03/03/15 12:01, Marius Strobl wrote:
> On Tue, Mar 03, 2015 at 10:21:55AM +0000, Hans Petter Selasky wrote:
>> Author: hselasky
>> Date: Tue Mar  3 10:21:54 2015
>> New Revision: 279563
>> URL: https://svnweb.freebsd.org/changeset/base/279563
>>
>> Log:
>>    Add quirk for USB 3.0 controllers which don't support 64-bit DMA.
>>
>
> Hrm, this is strange; I checked Linux sources when adding 64-bit DMA
> support and while they handle tons of quirks for USB controllers
> (which FreeBSD doesn't), they don't have a single such exception for
> 64-bit DMA. Are you sure using 64-bit DMA doesn't just reveal another
> problem? F. e., Linux issues an extra doorbell ring for XHCI_NEC_HOST
> and generally reads back the doorbell register in order to flush PCI
> writes.
>
> Marius
>
>

I'll check with Gary if he can test a patch for flushing posted PCI writes.

--HPS




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?54F599BA.5060708>