From: Nicholas Piggin <npiggin@gmail.com>
To: Thiago Jung Bauermann <bauerman@linux.ibm.com>
Cc: Gautham R Shenoy <ego@linux.vnet.ibm.com>,
linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org,
Michael Bringmann <mwb@linux.vnet.ibm.com>,
Tyrel Datwyler <tyreld@linux.vnet.ibm.com>
Subject: Re: [PATCH v3] powerpc/pseries: Only wait for dying CPU after call to rtas_stop_self()
Date: Thu, 18 Apr 2019 11:55:06 +1000 [thread overview]
Message-ID: <1555552271.zy2w4ddjqe.astroid@bobo.none> (raw)
In-Reply-To: <87r2a0jffq.fsf@morokweng.localdomain>
Thiago Jung Bauermann's on April 18, 2019 11:00 am:
>
> Hello Nick,
>
> Thank you very much for reviewing this patch!
>
> Nicholas Piggin <npiggin@gmail.com> writes:
>
>> Thiago Jung Bauermann's on April 11, 2019 9:08 am:
>>>
>>> Thiago Jung Bauermann <bauerman@linux.ibm.com> writes:
>>>
>>>> diff --git a/arch/powerpc/platforms/pseries/hotplug-cpu.c b/arch/powerpc/platforms/pseries/hotplug-cpu.c
>>>> index 97feb6e79f1a..ac6dc35ab829 100644
>>>> --- a/arch/powerpc/platforms/pseries/hotplug-cpu.c
>>>> +++ b/arch/powerpc/platforms/pseries/hotplug-cpu.c
>>>> @@ -214,13 +214,22 @@ static void pseries_cpu_die(unsigned int cpu)
>>>> msleep(1);
>>>> }
>>>> } else if (get_preferred_offline_state(cpu) == CPU_STATE_OFFLINE) {
>>>> + /*
>>>> + * If the current state is not offline yet, it means that the
>>>> + * dying CPU (which is either in pseries_mach_cpu_die() or in
>>>> + * the process of getting there) didn't have a chance yet to
>>>> + * call rtas_stop_self() and therefore it's too early to query
>>>> + * if the CPU is stopped.
>>>> + */
>>>> + spin_event_timeout(get_cpu_current_state(cpu) == CPU_STATE_OFFLINE,
>>>> + 100000, 100);
>>
>> If the CPU state does not go to offline here, you should give up and
>> return online, right? Otherwise I think query-cpu-stopped-state can
>> get confused by CPUs in idle and you get a false positive.
>
> Can it get confused? My impression from reading the definition for
> query-cpu-stopped-state in the PAPR is that it will simply return a
> CPU_status value of 2 in that case, meaning that "the processor thread
> is not in the RTAS stopped state", but I don't know much about this.
In QEMU (non-KVM) mode, qcss I think may get confused between H_CEDE
and rtas-stop-self. KVM mode may be okay because H_CEDE is handled in
the kernel.
>> That race can still happen, we would really need a sequence count check
>> over current CPU state to ensure we got a race-free qcss value, but at
>> least a check here should make the race implausible to hit.
>
> Actually, since rtas_stop_self() panics if the processor fails to stop
> and also since callers of pseries_cpu_die()¹ already assume that it is
> going to succeed in stopping the CPU (given that the function returns
> void and can't signal an error), a more straightforward way of
> eliminating the race is to simply do this:
>
> diff --git a/arch/powerpc/platforms/pseries/hotplug-cpu.c b/arch/powerpc/platforms/pseries/hotplug-cpu.c
> index 97feb6e79f1a..2331a609f48f 100644
> --- a/arch/powerpc/platforms/pseries/hotplug-cpu.c
> +++ b/arch/powerpc/platforms/pseries/hotplug-cpu.c
> @@ -215,7 +215,7 @@ static void pseries_cpu_die(unsigned int cpu)
> }
> } else if (get_preferred_offline_state(cpu) == CPU_STATE_OFFLINE) {
>
> - for (tries = 0; tries < 25; tries++) {
> + while (true) {
> cpu_status = smp_query_cpu_stopped(pcpu);
> if (cpu_status == QCSS_STOPPED ||
> cpu_status == QCSS_HARDWARE_ERROR)
>
>
> What do you think?
Yeah I think that may be a good idea, just makes things much simpler.
Thanks,
Nick
prev parent reply other threads:[~2019-04-18 1:55 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-03-11 19:35 [PATCH v3] powerpc/pseries: Only wait for dying CPU after call to rtas_stop_self() Thiago Jung Bauermann
2019-03-12 9:32 ` Gautham R Shenoy
2019-03-12 16:30 ` Thiago Jung Bauermann
2019-04-10 23:08 ` Thiago Jung Bauermann
2019-04-16 4:36 ` Nicholas Piggin
2019-04-18 1:00 ` Thiago Jung Bauermann
2019-04-18 1:55 ` Nicholas Piggin [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1555552271.zy2w4ddjqe.astroid@bobo.none \
--to=npiggin@gmail.com \
--cc=bauerman@linux.ibm.com \
--cc=ego@linux.vnet.ibm.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mwb@linux.vnet.ibm.com \
--cc=tyreld@linux.vnet.ibm.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).