From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751723AbbAKRKP (ORCPT ); Sun, 11 Jan 2015 12:10:15 -0500 Received: from nupi.pair.com ([209.68.5.149]:19970 "EHLO nupi.pair.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750860AbbAKRKM (ORCPT ); Sun, 11 Jan 2015 12:10:12 -0500 Message-ID: <54B2AE6E.3080909@flaterco.com> Date: Sun, 11 Jan 2015 12:10:06 -0500 From: David Flater User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.3.0 MIME-Version: 1.0 To: Takashi Iwai CC: Jaroslav Kysela , LKML , Alsa Devel Subject: Re: [PATCHv2 1/2] ALSA: fix emu8000 DRAM sizing for AWE64 Value References: <54B0775C.1080705@flaterco.com> In-Reply-To: Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 01/11/2015 05:52 AM, Takashi Iwai wrote: > At Fri, 09 Jan 2015 19:50:36 -0500, > David Flater wrote: >> CT4380 is commonly said to come with 1 MiB of DRAM, but Creative's AWE >> Control app agreed that mine has only 512 KiB. It has the same memory chip >> as the CT4520. > > And you couldn't actually use the 1MB, right? Confirmed with indirect evidence. Since the address space wraps around, the mode of failure (assuming that the load isn't blocked by checking against a correctly detected size) is silent corruption of the loaded sound font. Testing with a random sound font larger than 512 KiB but smaller than 1 MiB: Fender2.sf2, 545994 B On CT4390, the following test gives me a credible guitar in the left channel (and nothing else): asfxload -i -M Fender2.sf2 DRAM memory left = 3565 kB aplaymidi -p 17:0 E1M1.mid On CT4380 with DRAM size hard-coded to 1 MiB, asfxload catches no failure but aplaymidi produces a glitchy mixture of guitar and morse code beeping: asfxload -i -M Fender2.sf2 DRAM memory left = 493 kB aplaymidi -p 17:0 E1M1.mid DWF