Re: [PATCH v2 1/2] Replace if statement with WARN_ON_ONCE() in cmci_rediscover().

[Tang Chen <tangchen@cn.fujitsu.com>]

On 10/23/2012 05:52 PM, Borislav Petkov wrote:
> On Tue, Oct 23, 2012 at 10:55:13AM +0800, Tang Chen wrote:
>> So, how about warn once, and continue:
>> 	if (cpu == dying) {
>> 		WARN_ON_ONCE(cpu == dying);
>> 		continue;
>> 	}
>>
>> or, use BUG_ON() instead ?
>
> Let me ask you again, but I want you to think real hard this time:
>
> "Why do we need to warn? What good would that bring us?"

Hi,

First of all, I do think I was answering your question. As I said
before, if an online cpu == dying here, there must be something wrong.
Am I right here ?

If so, I think the "good" is obvious. If we don't output anything when
an online cpu == dying, nobody will know this happens. The kernel is in
wrong state, but nobody knows that, I don't see any good.

Actually, I used BUG_ON() in my v1 patch. So I dropped the if statement.
But Tejun asked me to use WARN_ON_ONCE(). And I forgot to add the if
statement.
Please refer to https://lkml.org/lkml/2012/10/16/528

And again, the "good" is inform user the kernel is in wrong state.

Thanks. :)


>