Re: [PATCH 1/4] acpi/ghes, cxl: Create a common CXL struct to handle different CXL CPER records

[Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com>]

On 1/2/2024 8:23 AM, Ira Weiny wrote:
> Smita Koralahalli wrote:
>> Currently defined cxl_cper_callback interface between CXL subsystem and
>> GHES module is just confined to handling CXL Component errors only.
>>
>> Extend this callback to process CXL Protocol errors as well. Achieve
>> by defining a new struct cxl_cper_rec_data to include cxl_cper_event_rec
>> and other fields of CXL protocol errors which will be defined in future
>> patches.
>>
>> Signed-off-by: Smita Koralahalli <Smita.KoralahalliChannabasappa@amd.com>
> 
> [snip]
> 
>>   
>>   static int __init cxl_pci_driver_init(void)
>> diff --git a/include/linux/cxl-event.h b/include/linux/cxl-event.h
>> index 17eadee819b6..afa71ee0437c 100644
>> --- a/include/linux/cxl-event.h
>> +++ b/include/linux/cxl-event.h
>> @@ -141,8 +141,12 @@ struct cxl_cper_event_rec {
>>   	union cxl_event event;
>>   } __packed;
>>   
>> +struct cxl_cper_rec_data {
>> +	struct cxl_cper_event_rec rec;
> 
> NIT: I would call this something like event to distinguish it from other
> record data.

What do you think of the below?

struct cxl_cper_event_info {
	struct cxl_cper_event_rec rec;
	struct cxl_cper_prot_err {
		struct cxl_ras_capability_regs cxl_ras;
		int severity;
	} p_err;
};

Addressed changing to sub-struct and copying the struct rather than 
pointer comments in patch 3..

> 
> Other than that this seems reasonable to me.
> 
> Reviewed-by: Ira Weiny <ira.weiny@intel.com>
> 
> [snip]
>