Re: [patch net-next 4/4] net: devlink: expose default flash update target

[Jiri Pirko <jiri@resnulli.us>]

Mon, Aug 22, 2022 at 07:09:35PM CEST, jacob.e.keller@intel.com wrote:
>
>
>> -----Original Message-----
>> From: Jiri Pirko <jiri@resnulli.us>
>> Sent: Friday, August 19, 2022 10:47 PM
>> To: Keller, Jacob E <jacob.e.keller@intel.com>
>> Cc: Jakub Kicinski <kuba@kernel.org>; netdev@vger.kernel.org;
>> davem@davemloft.net; idosch@nvidia.com; pabeni@redhat.com;
>> edumazet@google.com; saeedm@nvidia.com; vikas.gupta@broadcom.com;
>> gospo@broadcom.com
>> Subject: Re: [patch net-next 4/4] net: devlink: expose default flash update target
>> 
>> Sat, Aug 20, 2022 at 12:07:41AM CEST, jacob.e.keller@intel.com wrote:
>> >
>> >
>> >> -----Original Message-----
>> >> From: Jakub Kicinski <kuba@kernel.org>
>> >> Sent: Friday, August 19, 2022 2:46 PM
>> >> To: Keller, Jacob E <jacob.e.keller@intel.com>
>> >> Cc: Jiri Pirko <jiri@resnulli.us>; netdev@vger.kernel.org;
>> davem@davemloft.net;
>> >> idosch@nvidia.com; pabeni@redhat.com; edumazet@google.com;
>> >> saeedm@nvidia.com; vikas.gupta@broadcom.com; gospo@broadcom.com
>> >> Subject: Re: [patch net-next 4/4] net: devlink: expose default flash update
>> target
>> >>
>> >> On Fri, 19 Aug 2022 20:59:28 +0000 Keller, Jacob E wrote:
>> >> > > My intuition would be that if you specify no component you're flashing
>> >> > > the entire device. Is that insufficient? Can you explain the use case?
>> >> > >
>> >> > > Also Documentation/ needs to be updated.
>> >> >
>> >> > Some of the components in ice include the DDP which has an info
>> >> > version, but which is not part of the flash but is loaded by the
>> >> > driver during initialization.
>> >>
>> >> Right "entire device" as in "everything in 'stored'". Runtime loaded
>> >> stuff should not be listed in "stored" and therefore not be considered
>> >> "flashable". Correct?
>> >
>> >Yes I believe we don't list those as stored.
>> >
>> >We do have some extra version information that is reported through multiple
>> info lines, i.e. we report:
>> >
>> >fw.mgmt
>> >fw.mgmt.api
>> >fw.mgmt.build
>> >
>> >where the .api and .build are sub-version fields of the fw.mgmt and can
>> potentially give further information but are just a part of the fw.mgmt
>> component. They can't be flashed separately.
>> 
>> Yep, in my patchset, this is accounted for. The driver can say if the
>> "version" is flashable (passed as a compenent name) or not. In this case,
>> it is not and it only tells the user version of some fw part.
>> 
>
>I think we can just go with "is this flashable or not?" and then document that if no component is flashed, the driver should be flashing all marked components?
>
>Then we don't need a "default" since the default without component is to flash everything.

I dropped "default" from the patchset. We need to document the
semanticts for default at least. Is it always "everything"? Idk.

>
>> >
>> >Thanks,
>> >Jake