Date: Mon, 17 Feb 2003 12:37:29 -0800 From: Marcel Moolenaar <marcel@xcllnt.net> To: Andrew Gallatin <gallatin@cs.duke.edu> Cc: Peter Wemm <peter@wemm.org>, cvs-src@FreeBSD.org Subject: Re: cvs commit: src/sys/alpha/alpha machdep.c Message-ID: <20030217203729.GA916@athlon.pn.xcllnt.net> In-Reply-To: <15953.15467.687489.517344@grasshopper.cs.duke.edu> References: <3E500717.65436EAF@imimic.com> <20030217072612.577382A89E@canning.wemm.org> <15952.57680.522607.835422@grasshopper.cs.duke.edu> <20030217193905.GB593@athlon.pn.xcllnt.net> <15953.15467.687489.517344@grasshopper.cs.duke.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Feb 17, 2003 at 02:47:55PM -0500, Andrew Gallatin wrote: > > Marcel Moolenaar writes: > > > > Not yet. We don't even get 2GB to work all the time due to sparseness :-) > > And ia64 is in the same position as alpha: busdma needs to be fixed. So > > we too have "collateral damage"... > > Does ia64 have any sort of scatter/gather facility? Or are you > planning to just bounce? I'd go for bounce. Make sure that's working and then see how we can avoid doing that when it's not necessary. For example, HP's zx1 chipset has I/O TLBs that can be used to translate DMA requests. That should allow us to place the data anywhere we like, and still program a 32-bit DMA address. Handy, but chipset specific. I'd rather get the foundation right and then hack on top of that... > BTW, most of the devices in this alpha are DAC capable (except the > nic, dammit). Does busdma currently have a way to handle that? I don't know. I guess it's more likely to work if bus_addr_t is 64-bit already. I don't know much about DAC though... -- Marcel Moolenaar USPA: A-39004 marcel@xcllnt.net To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-src" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030217203729.GA916>