On 11/02/16 18:14, Joel Stanley wrote:
> Looks like you didn't break it much. Congratulations on being the new
> memcons maintainer!
Argh, I forgot the golden rule of kernel development - if you touch it,
you own it... :)
> There's one issue with your error handling.
>
>> diff --git a/arch/powerpc/platforms/powernv/opal.c b/arch/powerpc/platforms/powernv/opal.c
>> index 4e0da5a..0256d07 100644
>> --- a/arch/powerpc/platforms/powernv/opal.c
>> +++ b/arch/powerpc/platforms/powernv/opal.c
>> @@ -724,6 +724,9 @@ static int __init opal_init(void)
>> of_node_put(leds);
>> }
>>
>> + /* Initialise OPAL message log interface */
>> + opal_msglog_init();
>
> If this fails due to eg. the magic number being wrong.
>
>> +
>> /* Create "opal" kobject under /sys/firmware */
>> rc = opal_sysfs_init();
>> if (rc == 0) {
>> @@ -739,8 +742,8 @@ static int __init opal_init(void)
>> opal_platform_dump_init();
>> /* Setup system parameters interface */
>> opal_sys_param_init();
>> - /* Setup message log interface. */
>> - opal_msglog_init();
>> + /* Setup message log sysfs interface. */
>> + opal_msglog_sysfs_init();
>
> This will succeed, leaving you with a sysfs file attached to a broken driver.
V3 on its way...
--
Andrew Donnellan Software Engineer, OzLabs
andrew.donnellan at au1.ibm.com Australia Development Lab, Canberra
+61 2 6201 8874 (work) IBM Australia Limited