Date: Tue, 3 Feb 2009 17:59:58 -0800 From: Andrew Thompson <thompsa@FreeBSD.org> To: Mark Tinguely <tinguely@casselton.net> Cc: arm@freebsd.org, sam@freebsd.org Subject: Re: busdma problem Message-ID: <20090204015958.GA71913@citylink.fud.org.nz> In-Reply-To: <200902040009.n1409v7K033380@casselton.net> References: <20090203233028.GA68871@citylink.fud.org.nz> <200902040009.n1409v7K033380@casselton.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Feb 03, 2009 at 06:09:57PM -0600, Mark Tinguely wrote: > > On Fri, Jan 30, 2009 at 02:07:14PM -0800, Andrew Thompson wrote: > > As a test I removed the checking of KENTER_CACHE in > > arm/arm/pmap.c:pmap_kenter_internal() so all memory is uncached and now > > dma bounces to a stack variable work. > > > > busdma remaps the address nocache into vaddr_nocache pointer and uses > > that for the bcopy, obviously arm_remap_nocache() is not working > > correctly. > > > > Andrew > > I suspected this was another occurrance of the same cache problem that > has been mentioned starting about 3 weeks ago. I even was going to ask > you if it was using arm_remap_nocache(). > > I have a new concept patch for the kernel caching issue at: > > http://www.casselton.net/~tinguely/arm_pmap_unmanaged.diff > > I am still waiting for a drive to build a -current machine. I will also > look at any other caching ideas that others have. FWIW r188112 has fixed this issue for me. I will still be happy to help test your caching patch for any regressions, etc. Andrew
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20090204015958.GA71913>