From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753048Ab0DBSd3 (ORCPT ); Fri, 2 Apr 2010 14:33:29 -0400 Received: from mail.windriver.com ([147.11.1.11]:57251 "EHLO mail.windriver.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752313Ab0DBSdB (ORCPT ); Fri, 2 Apr 2010 14:33:01 -0400 From: Jason Wessel To: torvalds@linux-foundation.org Cc: linux-kernel@vger.kernel.org, kgdb-bugreport@lists.sourceforge.net Subject: [PATCH 4/5] kgdb: Use atomic operators which use barriers Date: Fri, 2 Apr 2010 13:32:24 -0500 Message-Id: <1270233145-29335-5-git-send-email-jason.wessel@windriver.com> X-Mailer: git-send-email 1.6.3.1.9.g95405b In-Reply-To: <1270233145-29335-4-git-send-email-jason.wessel@windriver.com> References: <1270233145-29335-1-git-send-email-jason.wessel@windriver.com> <1270233145-29335-2-git-send-email-jason.wessel@windriver.com> <1270233145-29335-3-git-send-email-jason.wessel@windriver.com> <1270233145-29335-4-git-send-email-jason.wessel@windriver.com> X-OriginalArrivalTime: 02 Apr 2010 18:32:44.0974 (UTC) FILETIME=[E30BCCE0:01CAD292] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org A cpu_relax() does not mandate that there is an smp memory barrier. As a result on the arm smp architecture the kernel debugger can hang on entry from time to time, as shown by the kgdb regression tests. The solution is simply to use the atomic operators which include a proper smp memory barrier, instead of using atomic_set() and atomic_read(). Tested-by: Will Deacon Signed-off-by: Jason Wessel --- kernel/kgdb.c | 21 ++++++++++----------- 1 files changed, 10 insertions(+), 11 deletions(-) diff --git a/kernel/kgdb.c b/kernel/kgdb.c index 6882c04..7a56dc9 100644 --- a/kernel/kgdb.c +++ b/kernel/kgdb.c @@ -1379,8 +1379,7 @@ acquirelock: * Make sure the above info reaches the primary CPU before * our cpu_in_kgdb[] flag setting does: */ - smp_wmb(); - atomic_set(&cpu_in_kgdb[cpu], 1); + atomic_inc(&cpu_in_kgdb[cpu]); /* * CPU will loop if it is a slave or request to become a kgdb @@ -1391,7 +1390,7 @@ acquirelock: if (atomic_cmpxchg(&kgdb_active, -1, cpu) == cpu) break; } else if (kgdb_info[cpu].exception_state & DCPU_IS_SLAVE) { - if (!atomic_read(&passive_cpu_wait[cpu])) + if (!atomic_add_return(0, &passive_cpu_wait[cpu])) goto return_normal; } else { return_normal: @@ -1400,7 +1399,7 @@ return_normal: */ if (arch_kgdb_ops.correct_hw_break) arch_kgdb_ops.correct_hw_break(); - atomic_set(&cpu_in_kgdb[cpu], 0); + atomic_dec(&cpu_in_kgdb[cpu]); touch_softlockup_watchdog_sync(); clocksource_touch_watchdog(); local_irq_restore(flags); @@ -1449,7 +1448,7 @@ return_normal: */ if (!kgdb_single_step) { for (i = 0; i < NR_CPUS; i++) - atomic_set(&passive_cpu_wait[i], 1); + atomic_inc(&passive_cpu_wait[i]); } #ifdef CONFIG_SMP @@ -1462,7 +1461,7 @@ return_normal: * Wait for the other CPUs to be notified and be waiting for us: */ for_each_online_cpu(i) { - while (!atomic_read(&cpu_in_kgdb[i])) + while (!atomic_add_return(0, &cpu_in_kgdb[i])) cpu_relax(); } @@ -1483,17 +1482,17 @@ return_normal: if (kgdb_io_ops->post_exception) kgdb_io_ops->post_exception(); - atomic_set(&cpu_in_kgdb[ks->cpu], 0); + atomic_dec(&cpu_in_kgdb[ks->cpu]); if (!kgdb_single_step) { for (i = NR_CPUS-1; i >= 0; i--) - atomic_set(&passive_cpu_wait[i], 0); + atomic_dec(&passive_cpu_wait[i]); /* * Wait till all the CPUs have quit * from the debugger. */ for_each_online_cpu(i) { - while (atomic_read(&cpu_in_kgdb[i])) + while (atomic_add_return(0, &cpu_in_kgdb[i])) cpu_relax(); } } @@ -1736,11 +1735,11 @@ EXPORT_SYMBOL_GPL(kgdb_unregister_io_module); */ void kgdb_breakpoint(void) { - atomic_set(&kgdb_setting_breakpoint, 1); + atomic_inc(&kgdb_setting_breakpoint); wmb(); /* Sync point before breakpoint */ arch_kgdb_breakpoint(); wmb(); /* Sync point after breakpoint */ - atomic_set(&kgdb_setting_breakpoint, 0); + atomic_dec(&kgdb_setting_breakpoint); } EXPORT_SYMBOL_GPL(kgdb_breakpoint); -- 1.6.3.1.9.g95405b