[4/4] powerpc/powernv: Fix little endian issues OPAL error log code

Fix little endian issues with the OPAL error log code.

Signed-off-by: Anton Blanchard <anton@samba.org>
---

Anton Blanchard <anton@samba.org> writes:
> Fix little endian issues with the OPAL error log code.
>
> Signed-off-by: Anton Blanchard <anton@samba.org>

Reviewed-by: Stewart Smith <stewart@linux.vnet.ibm.com>

Do we also need any magic for the getting of error logs themselves?

You may want to check the md5 of the logs themselves on be and le.

> ---
>
> Index: b/arch/powerpc/platforms/powernv/opal-elog.c
> ===================================================================
> --- a/arch/powerpc/platforms/powernv/opal-elog.c
> +++ b/arch/powerpc/platforms/powernv/opal-elog.c
> @@ -243,18 +243,25 @@ static struct elog_obj *create_elog_obj(
>
>  static void elog_work_fn(struct work_struct *work)
>  {
> +	__be64 size;
> +	__be64 id;
> +	__be64 type;
>  	uint64_t elog_size;
>  	uint64_t log_id;
>  	uint64_t elog_type;
>  	int rc;
>  	char name[2+16+1];
>
> -	rc = opal_get_elog_size(&log_id, &elog_size, &elog_type);
> +	rc = opal_get_elog_size(&id, &size, &type);
>  	if (rc != OPAL_SUCCESS) {
>  		pr_err("ELOG: Opal log read failed\n");
>  		return;
>  	}
>
> +	elog_size = be64_to_cpu(size);
> +	log_id = be64_to_cpu(id);
> +	elog_type = be64_to_cpu(type);
> +
>  	BUG_ON(elog_size > OPAL_MAX_ERRLOG_SIZE);
>
>  	if (elog_size >= OPAL_MAX_ERRLOG_SIZE)
> Index: b/arch/powerpc/include/asm/opal.h
> ===================================================================
> --- a/arch/powerpc/include/asm/opal.h
> +++ b/arch/powerpc/include/asm/opal.h
> @@ -851,7 +851,7 @@ int64_t opal_lpc_read(uint32_t chip_id,
>  		      uint32_t addr, __be32 *data, uint32_t sz);
>
>  int64_t opal_read_elog(uint64_t buffer, uint64_t size, uint64_t log_id);
> -int64_t opal_get_elog_size(uint64_t *log_id, uint64_t *size, uint64_t *elog_type);
> +int64_t opal_get_elog_size(__be64 *log_id, __be64 *size, __be64 *elog_type);
>  int64_t opal_write_elog(uint64_t buffer, uint64_t size, uint64_t offset);
>  int64_t opal_send_ack_elog(uint64_t log_id);
>  void opal_resend_pending_logs(void);

Hi Stewart,

> Do we also need any magic for the getting of error logs themselves?
> 
> You may want to check the md5 of the logs themselves on be and le.

Good idea. An md5 of the logs in LE and BE on this box checks out, so I
think we are good.

Anton