From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268105AbUH3O4m (ORCPT ); Mon, 30 Aug 2004 10:56:42 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268476AbUH3O4m (ORCPT ); Mon, 30 Aug 2004 10:56:42 -0400 Received: from 104.engsoc.carleton.ca ([134.117.69.104]:9893 "EHLO certainkey.com") by vger.kernel.org with ESMTP id S268105AbUH3OzJ (ORCPT ); Mon, 30 Aug 2004 10:55:09 -0400 Date: Mon, 30 Aug 2004 09:53:51 -0400 From: Jean-Luc Cooke To: James Morris Cc: linux-kernel@vger.kernel.org Subject: Re: CryptoAPI: schedual while atomic Message-ID: <20040830135351.GJ11307@certainkey.com> References: <20040830132453.GG11307@certainkey.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.6+20040722i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Aug 30, 2004 at 10:42:11AM -0400, James Morris wrote: > Where is the code which caused this? The transforms are safe to use (but > not allocate) in process and softirq contexts. In add_entropy_words of random.c my experimental code is calling crypto_digest_update(). In update() it calles crypto_yield. add_entropy_words() is being call ed directly from a keyboard_interrupt. I was hoping to tidy up the code a bit by not using batch_entropy_stores... but I guess that's unavoidable then? Last question: Would spin_lock_irqsave() spin_unlock_irqrestore() still needed to surround crypto_digest_update() calls if they're going to be scattered/gathered later? Thanks for your time. JLC