From mboxrd@z Thu Jan 1 00:00:00 1970 From: Frank Rowand Subject: Re: [PATCH V2] preempt-rt: allow immediate Magic SysRq output for PREEMPT_RT_FULL Date: Fri, 23 Sep 2011 13:15:28 -0700 Message-ID: <4E7CE8E0.5070604@am.sony.com> References: <4E7BA334.2050003@am.sony.com> <7186.1316799377@turing-police.cc.vt.edu> Reply-To: Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit Cc: "Rowand, Frank" , "tglx@linutronix.de" , "linux-rt-users@vger.kernel.org" , "linux-kernel@vger.kernel.org" To: "Valdis.Kletnieks@vt.edu" Return-path: Received: from db3ehsobe002.messaging.microsoft.com ([213.199.154.140]:11235 "EHLO DB3EHSOBE002.bigfish.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751784Ab1IWUQG (ORCPT ); Fri, 23 Sep 2011 16:16:06 -0400 In-Reply-To: <7186.1316799377@turing-police.cc.vt.edu> Sender: linux-rt-users-owner@vger.kernel.org List-ID: On 09/23/11 10:36, Valdis.Kletnieks@vt.edu wrote: > On Thu, 22 Sep 2011 14:05:56 PDT, Frank Rowand said: >> Change from V1 to V2: add drivers/tty/serial/8250.c > >> drivers/tty/serial/cpm_uart/cpm_uart_core.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/mfd.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/nwpserial.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/omap-serial.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/sunhv.c | 4 2 + 2 - 0 ! >> drivers/tty/serial/sunsab.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/sunsu.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/sunzilog.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/uartlite.c | 2 1 + 1 - 0 ! >> drivers/tty/serial/xilinx_uartps.c | 2 1 + 1 - 0 ! > >> if (up->port.sysrq) >> locked = 0; >> - else if (oops_in_progress) { >> + else if (oops_in_progress || sysrq_in_progress) { >> locked = spin_trylock(&up->port.lock); >> } else >> spin_lock(&up->port.lock); > > Is it time to create a helper function for this chunk of code that appears > in almost identical form in a dozen drivers? Thanks for pointing that out. The "if (up->port.sysrq)" in this pattern already handles the case that I was trying to deal with in all of those drivers, except cpm_uart_core.c, so I was able to remove all the other drivers from V3 of the patch. -Frank