From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from tulikuusama.dnainternet.net ([83.102.40.132]) by merlin.infradead.org with esmtp (Exim 4.80.1 #2 (Red Hat Linux)) id 1V6wea-0004KW-Am for linux-mtd@lists.infradead.org; Wed, 07 Aug 2013 05:51:25 +0000 Message-ID: <5201E03D.30306@iki.fi> Date: Wed, 07 Aug 2013 08:50:53 +0300 From: Jussi Kivilinna MIME-Version: 1.0 To: =?UTF-8?B?SsO2cm4gRW5nZWw=?= Subject: Re: [PATCH] alauda: do not use stack for URB transfer_buffers References: <20130806120329.8864.36301.stgit@localhost6.localdomain6> <20130806164913.GD10535@logfs.org> In-Reply-To: <20130806164913.GD10535@logfs.org> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Cc: David Woodhouse , linux-usb@vger.kernel.org, linux-mtd@lists.infradead.org List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On 06.08.2013 19:49, Jörn Engel wrote: > On Tue, 6 August 2013 15:03:29 +0300, Jussi Kivilinna wrote: >> >> Patch fixes alauda not to use stack as URB transfer_buffer. URB buffers need to >> be DMA-able, which stack is not. >> >> Patch is only compile tested. > > I have tested the driver back when I wrote it. Not sure why it worked > then, maybe the chip in my notebook back then didn't care about > alignment or I just got lucky with the memory allocations. You might not see problems with your hardware/architecture. "Documentation/DMA-API-HOWTO.txt" discusses about "What memory is DMA'able?". > > My old test hardware was likely thrown out in the last move, about two > years back. Performance sucked anyway, 400kB/s was the limit. Maybe > we should just remove the driver and not spend any more time on it? > > Jörn > > -- > Victory in war is not repetitious. > -- Sun Tzu >