Re: [Xen-devel] [PATCH v4 2/2] x86: add accessors for scratch cpu mask

[Jan Beulich <jbeulich@xxxxxxxx>]

On 28.02.2020 12:41, Roger Pau Monné wrote:
> On Fri, Feb 28, 2020 at 12:15:21PM +0100, Jan Beulich wrote:
>> On 28.02.2020 11:31, Roger Pau Monné wrote:
>>> On Fri, Feb 28, 2020 at 11:16:55AM +0100, Jan Beulich wrote:
>>>> On 28.02.2020 10:33, Roger Pau Monne wrote:
>>>>> +    /*
>>>>> +     * Due to reentrancy scratch cpumask cannot be used in IRQ, #MC or 
>>>>> #NMI
>>>>> +     * context.
>>>>> +     */
>>>>> +    BUG_ON(in_irq() || in_mce_handler() || in_nmi_handler());
>>>>> +
>>>>> +    if ( use && unlikely(this_cpu(scratch_cpumask_use)) )
>>>>> +    {
>>>>> +        printk("scratch CPU mask already in use by %ps (%p)\n",
>>>>> +               this_cpu(scratch_cpumask_use), 
>>>>> this_cpu(scratch_cpumask_use));
>>>>
>>>> Why the raw %p as well? We don't do so elsewhere, I think. Yes,
>>>> it's debugging code only, but I wonder anyway.
>>>
>>> I use addr2line to find the offending line, and it's much easier to do
>>> so if you have the address directly, rather than having to use nm in
>>> order to figure out the address of the symbol and then add the offset.
>>>
>>> Maybe I'm missing some other way to do this more easily?
>>
>> In such a case we may want to consider making %ps (and %pS)
>> print a hex presentation next to the decoded one, in debug
>> builds at least. Andrew, thoughts? (There may be cases where
>> this is not wanted, bit if we made this a debug mode only
>> feature, I think it wouldn't do too much harm.)
> 
> If you agree to make %p[sS] print the address then I can drop this
> and send a patch to that effect (likely next week).

In principle I agree, but the effect in particular on stack
dumps needs to be looked at pretty closely.

Jan

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel