From mboxrd@z Thu Jan 1 00:00:00 1970 From: Takashi Iwai Subject: Re: Use less number of BDs in i810 Date: Wed, 07 Aug 2002 11:54:03 +0200 Sender: alsa-devel-admin@lists.sourceforge.net Message-ID: References: <3D4F9330.3070100@boosthardware.com> <20020806144804.A20615@thot.informatik.uni-kl.de> <3D50CB27.43CC7EA9@multitech.co.in> Mime-Version: 1.0 (generated by SEMI 1.14.3 - "Ushinoya") Content-Type: text/plain; charset=US-ASCII Return-path: In-Reply-To: <3D50CB27.43CC7EA9@multitech.co.in> Errors-To: alsa-devel-admin@lists.sourceforge.net List-Help: List-Post: List-Subscribe: , List-Unsubscribe: , List-Archive: To: Shaju Abraham Cc: alsa-devel@lists.sourceforge.net List-Id: alsa-devel@alsa-project.org At Wed, 07 Aug 2002 12:54:23 +0530, Shaju Abraham wrote: > > Hi all, > I suppose that present implementation of i810 driver uses > 32 BDs. no, it's maximal 32 BDs. and, in the current code, a period (as the alsa's terminology) corresponds to a BD, i.e. maximal 32 periods. > Can u pls direct me how to reduce this number to say only 16 or 8 BDs. > > Also, now if the number of frags I am getting is 8, size of each frag I am > getting is 4456 bytes. I have two questions regarding this. > > 1. How can I force the fragment size to be abt 1k or 2k bytes? via alsa-lib API. set the period size (or period time) explicitly. > 2. Is it strange that this number is not a multiple of 512 bytes? ( I > am getting physize of 35648 bytes. not 32K ? ) in the current code it's ok. the size of each BD is arbitrary. but if we use discontinuous kernel pages via scatter-gather mechanism, the period size must be aligned to the power of two. this is not implemented on intel8x0 driver yet, but once if it's there, the size will be restricted. Takashi ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf