Date: Thu, 14 Apr 2005 13:11:02 -0400 From: Mathew Kanner <mat@cnd.mcgill.ca> To: Alexander Leidinger <Alexander@Leidinger.net> Cc: julian@elischer.org Subject: Re: de-dma uaudio Message-ID: <20050414171102.GJ2178@cnd.mcgill.ca> In-Reply-To: <20050414183511.4o2rsij4k0coksck@netchild.homeip.net> References: <20050410195645.GA2178@cnd.mcgill.ca> <20050414.021552.343134310.kazuhito@ph.noda.tus.ac.jp> <20050413172534.GF2178@cnd.mcgill.ca> <20050414161546.kwroviadwsw8k0w0@netchild.homeip.net> <20050414152146.GI2178@cnd.mcgill.ca> <20050414183511.4o2rsij4k0coksck@netchild.homeip.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Apr 14, Alexander Leidinger wrote: > Mathew Kanner <mat@cnd.mcgill.ca> wrote: > > >>Do I understand this patch right: it changes from DMA access to non-DMA > >>access? > >> > >>If yes: can someone please explain me why this is a good thing? > > > > Upper layer does not use DMA. It just isn't use so why > >allocate special memeory for it. > > And our USB stack ([eou]hci) can't ever grow the feature of doing DMA? That's a lower layer. I'm removing the DMA mem allocation in the PCM layer. The uaudio driver does it's own buffer allocation (though I don't think it's dma capable at that layer, but when it is...) It's bad that we have to do so many copies but we can't just easily rewrite the uaudio driver. --Mat --
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050414171102.GJ2178>