From deef325086c3897393b8f7d6bccd03405244fe18 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Mon, 12 May 2008 15:44:38 +0200 Subject: x86: disable preemption in native_smp_prepare_cpus Priit Laes reported the following warning: Call Trace: [] warn_on_slowpath+0x51/0x63 [] sys_ioctl+0x2d/0x5d [] _spin_lock+0xe/0x24 [] task_rq_lock+0x3d/0x73 [] set_cpu_sibling_map+0x336/0x350 [] read_apic_id+0x30/0x62 [] verify_local_APIC+0x90/0x138 [] native_smp_prepare_cpus+0x1f9/0x305 [] kernel_init+0x59/0x2d9 [] _spin_unlock_irq+0x11/0x2b [] child_rip+0xa/0x12 [] kernel_init+0x0/0x2d9 [] child_rip+0x0/0x12 fix this by generally disabling preemption in native_smp_prepare_cpus(). Reported-and-bisected-by: Priit Laes Signed-off-by: Ingo Molnar Signed-off-by: Thomas Gleixner --- arch/x86/kernel/smpboot.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'arch/x86/kernel/smpboot.c') diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 3898849..56078d6 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -1190,6 +1190,7 @@ static void __init smp_cpu_index_default(void) */ void __init native_smp_prepare_cpus(unsigned int max_cpus) { + preempt_disable(); nmi_watchdog_default(); smp_cpu_index_default(); current_cpu_data = boot_cpu_data; @@ -1206,7 +1207,7 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) if (smp_sanity_check(max_cpus) < 0) { printk(KERN_INFO "SMP disabled\n"); disable_smp(); - return; + goto out; } preempt_disable(); @@ -1246,6 +1247,8 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) printk(KERN_INFO "CPU%d: ", 0); print_cpu_info(&cpu_data(0)); setup_boot_clock(); +out: + preempt_enable(); } /* * Early setup to make printk work. -- cgit v1.1 From fcb43042ef55d2f46b0efa5d7746967cef38f056 Mon Sep 17 00:00:00 2001 From: "Zhang, Yanmin" Date: Tue, 24 Jun 2008 16:06:23 +0800 Subject: x86: fix cpu hotplug crash Vegard Nossum reported crashes during cpu hotplug tests: http://marc.info/?l=linux-kernel&m=121413950227884&w=4 In function _cpu_up, the panic happens when calling __raw_notifier_call_chain at the second time. Kernel doesn't panic when calling it at the first time. If just say because of nr_cpu_ids, that's not right. By checking the source code, I found that function do_boot_cpu is the culprit. Consider below call chain: _cpu_up=>__cpu_up=>smp_ops.cpu_up=>native_cpu_up=>do_boot_cpu. So do_boot_cpu is called in the end. In do_boot_cpu, if boot_error==true, cpu_clear(cpu, cpu_possible_map) is executed. So later on, when _cpu_up calls __raw_notifier_call_chain at the second time to report CPU_UP_CANCELED, because this cpu is already cleared from cpu_possible_map, get_cpu_sysdev returns NULL. Many resources are related to cpu_possible_map, so it's better not to change it. Below patch against 2.6.26-rc7 fixes it by removing the bit clearing in cpu_possible_map. Signed-off-by: Zhang Yanmin Tested-by: Vegard Nossum Acked-by: Rusty Russell Signed-off-by: Ingo Molnar --- arch/x86/kernel/smpboot.c | 1 - 1 file changed, 1 deletion(-) (limited to 'arch/x86/kernel/smpboot.c') diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 56078d6..3e1cece 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -996,7 +996,6 @@ do_rest: #endif cpu_clear(cpu, cpu_callout_map); /* was set by do_boot_cpu() */ cpu_clear(cpu, cpu_initialized); /* was set by cpu_init() */ - cpu_clear(cpu, cpu_possible_map); cpu_clear(cpu, cpu_present_map); per_cpu(x86_cpu_to_apicid, cpu) = BAD_APICID; } -- cgit v1.1