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

[Dave Anderson]

----- 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 at huawei.com>
> >>>> ---
> >>>>  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