From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265473AbVBDUEV (ORCPT ); Fri, 4 Feb 2005 15:04:21 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266669AbVBDUDW (ORCPT ); Fri, 4 Feb 2005 15:03:22 -0500 Received: from www.ssc.unict.it ([151.97.230.9]:18180 "HELO ssc.unict.it") by vger.kernel.org with SMTP id S264284AbVBDUAh (ORCPT ); Fri, 4 Feb 2005 15:00:37 -0500 Subject: [patch 3/8] uml: Fix SKAS sig-handler reentrancy [before 2.6.11] To: akpm@osdl.org Cc: linux-kernel@vger.kernel.org, jdike@addtoit.com, user-mode-linux-devel@lists.sourceforge.net, blaisorblade@yahoo.it From: blaisorblade@yahoo.it Date: Fri, 04 Feb 2005 19:35:46 +0100 Message-Id: <20050204183546.4980F310B8@zion> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org From: Jeff Dike , Paolo 'Blaisorblade' Giarrusso CC: uml-devel This adds code which enables SIGSEGV reception to the SKAS sig_handler_common, which matches the tt code. I still need to figure out why the SA_NODEFER flag was backed out in favor of this. Signed-off-by: Paolo 'Blaisorblade' Giarrusso --- linux-2.6.11-paolo/arch/um/kernel/skas/trap_user.c | 8 ++++++++ 1 files changed, 8 insertions(+) diff -puN arch/um/kernel/skas/trap_user.c~uml-no-defer arch/um/kernel/skas/trap_user.c --- linux-2.6.11/arch/um/kernel/skas/trap_user.c~uml-no-defer 2005-02-04 06:14:11.966064760 +0100 +++ linux-2.6.11-paolo/arch/um/kernel/skas/trap_user.c 2005-02-04 06:14:11.969064304 +0100 @@ -20,6 +20,14 @@ void sig_handler_common_skas(int sig, vo int save_errno = errno; int save_user; + /* This is done because to allow SIGSEGV to be delivered inside a SEGV + * handler. This can happen in copy_user, and if SEGV is disabled, + * the process will die. + * XXX Figure out why this is better than SA_NODEFER + */ + if(sig == SIGSEGV) + change_sig(SIGSEGV, 1); + r = &TASK_REGS(get_current())->skas; save_user = r->is_user; r->is_user = 0; _