Re: [Crash-utility] [PATCH 2/3] crash: ARM: get online CPUs from cpu_active_mask

----- Original Message ----- > 于 2014/4/27 7:43, Dave Anderson 写道: >> >> >> ----- Original Message ----- >>> On 2014/4/26 3:20, Dave Anderson wrote: >>>> >>>> >>>> ----- Original Message ----- >>>>> For ARM32 platfrom, The system will "offline" all CPUs except the >>>>> crashing one, by clear the cpu_online_mask. So we need to find >>>>> another way to get online-CPUs number for crash utility. >>>>> >>>>> This patch uses cpu_active_mask to get that value. >>>>> >>>>> Signed-off-by: Liu Hua <sdu.liu(a)huawei.com&gt; >>>>> --- >>>>> arm.c | 2 +- >>>>> 1 file changed, 1 insertion(+), 1 deletion(-) >>>>> >>>>> diff --git a/arm.c b/arm.c >>>>> index 3c38cd5..c94d7bf 100644 >>>>> --- a/arm.c >>>>> +++ b/arm.c >>>>> @@ -1518,7 +1518,7 @@ arm_display_machine_stats(void) >>>>> static int >>>>> arm_get_smp_cpus(void) >>>>> { >>>>> - return get_cpus_online(); >>>>> + return get_cpus_active(); >>>>> } >>>>> >>>>> /* >>>> >>>> When did this start happening? I ask because no other ARM users have >>>> reported this until now. I've only got one sample SMP vmcore, and >>>> clearly the non-crashing cpu was not offlined: >>>> >>> >>> >>> Hi Dave, >>> >>> commit 36a2e5010fe87 "kexec: offline non panic CPUs on Kdump panic " >>> introduces this change. <v3.10.27-1-g36a2e50> >>> >>> The following messges come from the ARM boards on my hand. The crash >>> utility has added my patchs. >>> >>> (1) ARM A15 16 CPUs >>> >>> crash> sys >>> >>> KERNEL: vmlinux >>> DUMPFILE: vmcore_my [PARTIAL DUMP] >>> CPUS: 16 >>> DATE: Fri Feb 28 10:07:55 2014 >>> UPTIME: 00:03:57 >>> LOAD AVERAGE: 0.03, 0.01, 0.01 >>> TASKS: 140 >>> NODENAME: arma15el >>> RELEASE: 3.10.37+ >>> VERSION: #2 SMP Thu Apr 24 21:13:28 CST 2014 >>> MACHINE: armv7l (unknown Mhz) >>> MEMORY: 3.1 GB >>> PANIC: "Internal error: Oops: a07 [#1] SMP ARM" (check log for >>> details >>> >>> crash> help -k >>> .... >>> cpu_possible_map: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >>> cpu_present_map: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >>> cpu_online_map: 0 >>> ... >>> >>> >>> (2) ARM A9 4 CPUs >>> >>> crash> sys >>> KERNEL: vmlinux >>> DUMPFILE: vmcore >>> CPUS: 2 >>> DATE: Mon Apr 14 11:09:10 2014 >>> UPTIME: 00:06:17 >>> LOAD AVERAGE: 0.16, 0.05, 0.02 >>> TASKS: 45 >>> NODENAME: arma9el >>> RELEASE: 3.10.37+ >>> VERSION: #6 SMP Tue Apr 22 13:23:30 CST 2014 >>> MACHINE: armv7l (unknown Mhz) >>> MEMORY: 1.5 GB >>> PANIC: "Internal error: Oops: 817 [#1] SMP ARM" (check log for >>> details) >>> >>> crash> help -k >>> .... >>> cpu_possible_map: 0 1 >>> cpu_present_map: 0 1 >>> cpu_online_map: 1 >>> >>> >>> crash> p per_cpu__runqueues >>> >>> PER-CPU DATA TYPE: >>> struct rq runqueues; >>> PER-CPU ADDRESSES: >>> [0]: c069a680 >>> [1]: c06a2680 >>> >>> With my patches, "sys" command is ok. And I can get per_cpu variables >>> correctly. >>> But something should be done for "help -k". May be I should add fully >>> support >>> for cpu_active_mask. If necessory, I will resend this patch series. >> >> It's not necessary, but it would be nice to have, and would be >> easy to add. >> >> Also, would it be possible that you can let me have a copy of your >> 16-cpu vmlinux/vmcore pair? That would be a great sample for >> my testing. >> >> Thanks, >> Dave > > Hi Dave, > > Sure, there seems no extra problems caused by the commit above. We can add it > when necessory. Will you be re-posting the original patch? (with or without the help -k support)

> > And, I am very sorry to tell you that I have no authority for sending binary files out of my > company, becasue of it's security rules. I have no ways to send that huge file out. Maybe > I can do some test for you on my platform. OK I understand. Thanks, Dave .