[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <2eb17b37-3bb2-12ac-2ea5-956537e45e45@intel.com>
Date: Wed, 19 Aug 2020 17:58:43 -0700
From: Jacob Keller <jacob.e.keller@...el.com>
To: David Miller <davem@...emloft.net>
Cc: netdev@...r.kernel.org, jiri@...lanox.com, kuba@...nel.org,
corbet@....net, michael.chan@...adcom.com, luobin9@...wei.com,
saeedm@...lanox.com, leon@...nel.org, idosch@...lanox.com,
danieller@...lanox.com
Subject: Re: [net-next v3 1/4] devlink: check flash_update parameter support
in net core
On 8/19/2020 4:36 PM, David Miller wrote:
> From: Jacob Keller <jacob.e.keller@...el.com>
> Date: Tue, 18 Aug 2020 17:28:15 -0700
>
>> @@ -991,6 +993,12 @@ enum devlink_trap_group_generic_id {
>> }
>>
>> struct devlink_ops {
>> + /**
>> + * @supported_flash_update_params:
>> + * mask of parameters supported by the driver's .flash_update
>> + * implemementation.
>> + */
>> + u32 supported_flash_update_params;
>> int (*reload_down)(struct devlink *devlink, bool netns_change,
>> struct netlink_ext_ack *extack);
>> int (*reload_up)(struct devlink *devlink,
>
> Jakub asked if this gave W=1 warnings. Then you responded that you didn't
> see any warnings with allmodconfig nor allyesconfig, but that isn't the
> question Jakub asked.
>
Ah, yes I should have been more specific:
> Are you building with W=1 explicitly added to the build?
>
I did
$ make allyesconfig
$ make W=1 &>allyes.txt
$ cat allyes.txt | grep warning | grep devlink
drivers/net/ethernet/huawei/hinic/hinic_hw_dev.c:917: warning: Function
parameter or member 'devlink' not described in 'hinic_init_hwdev'
and
$ make allmodconfig
$ make W=1 *>allmod.txt
$ cat allyes.txt | grep warning | grep devlink
drivers/net/ethernet/huawei/hinic/hinic_hw_dev.c:917: warning: Function
parameter or member 'devlink' not described in 'hinic_init_hwdev'
I also looked at those manually, there's about 15k warnings with W=1 on
allyesconfig, but none of them appear to be related to the changes in
this patch.
> The issue is this kerneldoc might not be formatted correctly, and such
> warnings won't be presented without doing a W=1 build.
>
> Thank you.
>
Sure.
I also manually ran:
$ ./scripts/kernel-doc -v -none include/uapi/linux/devlink.h
include/uapi/linux/devlink.h:232: info: Scanning doc for enum
devlink_trap_action
include/uapi/linux/devlink.h:246: info: Scanning doc for enum
devlink_trap_type
Hope this helps clarify.
Thanks,
Jake
Powered by blists - more mailing lists