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

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.

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.