Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264092AbUDRANJ (ORCPT ); Sat, 17 Apr 2004 20:13:09 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264103AbUDRANJ (ORCPT ); Sat, 17 Apr 2004 20:13:09 -0400 Received: from sankara1.bol.com.br ([200.221.24.109]:8182 "EHLO sankara1.bol.com.br") by vger.kernel.org with ESMTP id S264092AbUDRANF (ORCPT ); Sat, 17 Apr 2004 20:13:05 -0400 Subject: local interrupt disabling From: Fabiano Ramos To: linux-kernel@vger.kernel.org Content-Type: text/plain Message-Id: <1082247507.1201.7.camel@slack.domain.invalid> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.5 Date: Sat, 17 Apr 2004 21:18:27 -0300 Content-Transfer-Encoding: 7bit X-Sender-IP: 200.165.174.198 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 914 Lines: 22 Hi all. I am doing some work on 2.4 to count the number of instructions executed for a given syscall.I set the trap bit on eflags at the syscall entry, causing a trap for each instruction executed until the iret at return_from_syscall. At the debug handler, I increment my counter. I know that the syscall code cannot be preempted by another exeception handler, only by interrupts or page faults it causes. I want to track the page faults, no problem, but I wouldn't like the syscall to be interrupted by a mouse interrupt, for examnple. Local interrupt disabling while servicing the syscall would solve my problem, or would it create other problems? TIA, Fabiano - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/