From: Herbert Xu Subject: Re: [PATCH 4/4] crypto: algif_skcipher - User-space interface for skcipher operations Date: Mon, 15 Nov 2010 19:51:20 +0800 Message-ID: <20101115115120.GA3797@gondor.apana.org.au> References: <20101107010813.GA20538@gondor.apana.org.au> <1289207420.2013.16.camel@martin> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: linux-crypto@vger.kernel.org To: Martin Willi Return-path: Received: from helcar.apana.org.au ([209.40.204.226]:60965 "EHLO fornost.hengli.com.au" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755242Ab0KOLvX (ORCPT ); Mon, 15 Nov 2010 06:51:23 -0500 Content-Disposition: inline In-Reply-To: <1289207420.2013.16.camel@martin> Sender: linux-crypto-owner@vger.kernel.org List-ID: On Mon, Nov 08, 2010 at 10:10:20AM +0100, Martin Willi wrote: > > The test program below runs 1000 encryptions: > > # grep nr_free /proc/vmstat > nr_free_pages 11031 > # ./test Thanks, Miroslav identified a bogosity where if we're not doing a whole page then the last sgl pointer is off by one which triggers the leak. With his fix I cannot reproduce this problem anymore. Other major problems identified by him include inadequage checks with respect to ALG_MAX_PAGES, and the use of generic SG chaining instead of crypto SG chaining. So here is a respin of the patches which I hope will be the last :) Cheers, -- Email: Herbert Xu Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt