Re: [PATCH] HWPOISON: add a pr_err message when forcibly send a sigbus

[Shuai Xue]

On 2023/8/31 17:06, Helge Deller wrote:
> On 8/31/23 05:29, Shuai Xue wrote:
>> On 2023/8/31 06:18, Will Deacon wrote:
>>> On Mon, Aug 28, 2023 at 09:41:55AM +0800, Shuai Xue wrote:
>>>> On 2023/8/22 09:15, Shuai Xue wrote:
>>>>> On 2023/8/21 18:50, Will Deacon wrote:
>>>>>>> diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
>>>>>>> index 3fe516b32577..38e2186882bd 100644
>>>>>>> --- a/arch/arm64/mm/fault.c
>>>>>>> +++ b/arch/arm64/mm/fault.c
>>>>>>> @@ -679,6 +679,8 @@ static int __kprobes do_page_fault(unsigned long far, unsigned long esr,
>>>>>>>       } else if (fault & (VM_FAULT_HWPOISON_LARGE | VM_FAULT_HWPOISON)) {
>>>>>>>           unsigned int lsb;
>>>>>>>
>>>>>>> +        pr_err("MCE: Killing %s:%d due to hardware memory corruption fault at %lx\n",
>>>>>>> +               current->comm, current->pid, far);
>>>>>>>           lsb = PAGE_SHIFT;
>>>>>>>           if (fault & VM_FAULT_HWPOISON_LARGE)
>>>>>>>               lsb = hstate_index_to_shift(VM_FAULT_GET_HINDEX(fault));
>>>>>>
>>>>>> Hmm, I'm not convinced by this. We have 'show_unhandled_signals' already,
>>>>>> and there's plenty of code in memory-failure.c for handling poisoned pages
>>>>>> reported by e.g. GHES. I don't think dumping extra messages in dmesg from
>>>>>> the arch code really adds anything.
>>>>>
>>>>> I see the show_unhandled_signals() will dump the stack but it rely on
>>>>> /proc/sys/debug/exception-trace be set.
>>>>>
>>>>> The memory failure is the top issue in our production cloud and also other hyperscalers.
>>>>> We have received complaints from our operations engineers and end users that processes
>>>>> are being inexplicably killed :(. Could you please consider add a message?
>>>
>>> I don't have any objection to logging this stuff somehow, I'm just not
>>> convinced that the console is the best place for that information in 2023.
>>> Is there really nothing better?
> 
>> I agree that console might not the better place, but it still plays an important role.
>> IMO the most direct idea for end user to check what happened is to check by viewing
>> the dmesg. In addition, we deployed some log store service collects all cluster dmesg
>> from /var/log/kern.
> 
> Right, pr_err() is not just console.
> It ends up in the syslog, which ends up in a lot of places, e.g. through syslog forwarding.
> Most monitoring tools monitor the syslog as well.
> 
> So, IMHO pr_err() is the right thing.
> 
> Helge
> 

Totally agreed.

Thank you.

Best Regards,
Shuai