diff mbox series

[isar-cip-core,RFC] swupdate-handler: Add default config to fix u-boot build

Message ID 20230803102401.564807-1-Quirin.Gylstorff@siemens.com (mailing list archive)
State Rejected
Headers show
Series [isar-cip-core,RFC] swupdate-handler: Add default config to fix u-boot build | expand

Commit Message

Gylstorff Quirin Aug. 3, 2023, 10:24 a.m. UTC
From: Quirin Gylstorff <quirin.gylstorff@siemens.com>

This adds the missing configuration file to allow to build with u-boot
as the boot selector.

It uses the getroot mechanism from SWUpdate to select the update
partition.

Signed-off-by: Quirin Gylstorff <quirin.gylstorff@siemens.com>
---
CC: felix.moessbauer@siemens.com
TO: jan.kiszka@siemens.com

 .../files/swupdate.handler.u-boot.ini                  | 10 ++++++++++
 1 file changed, 10 insertions(+)
 create mode 100644 recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini

Comments

Jan Kiszka Aug. 10, 2023, 11:01 a.m. UTC | #1
On 03.08.23 12:24, Quirin Gylstorff wrote:
> From: Quirin Gylstorff <quirin.gylstorff@siemens.com>
> 
> This adds the missing configuration file to allow to build with u-boot
> as the boot selector.
> 
> It uses the getroot mechanism from SWUpdate to select the update
> partition.
> 
> Signed-off-by: Quirin Gylstorff <quirin.gylstorff@siemens.com>
> ---
> CC: felix.moessbauer@siemens.com
> TO: jan.kiszka@siemens.com
> 
>  .../files/swupdate.handler.u-boot.ini                  | 10 ++++++++++
>  1 file changed, 10 insertions(+)
>  create mode 100644 recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
> 
> diff --git a/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
> new file mode 100644
> index 0000000..0e6df08
> --- /dev/null
> +++ b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
> @@ -0,0 +1,10 @@
> +[image]
> +chainhandler=raw
> +
> +[image.selector]
> +method=getroot_rrmap
> +[kernel]
> +chainhandler=rawfile
> +
> +[kernel.selector]
> +method=getroot_rrmap

This is not being stressed by isar-cip-core, is it? Do we plan to add a
target that uses legacy u-boot script-based a/b switching?

Jan
Gylstorff Quirin Aug. 10, 2023, 11:26 a.m. UTC | #2
On 8/10/23 13:01, Jan Kiszka wrote:
> On 03.08.23 12:24, Quirin Gylstorff wrote:
>> From: Quirin Gylstorff <quirin.gylstorff@siemens.com>
>>
>> This adds the missing configuration file to allow to build with u-boot
>> as the boot selector.
>>
>> It uses the getroot mechanism from SWUpdate to select the update
>> partition.
>>
>> Signed-off-by: Quirin Gylstorff <quirin.gylstorff@siemens.com>
>> ---
>> CC: felix.moessbauer@siemens.com
>> TO: jan.kiszka@siemens.com
>>
>>   .../files/swupdate.handler.u-boot.ini                  | 10 ++++++++++
>>   1 file changed, 10 insertions(+)
>>   create mode 100644 recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>>
>> diff --git a/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>> new file mode 100644
>> index 0000000..0e6df08
>> --- /dev/null
>> +++ b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>> @@ -0,0 +1,10 @@
>> +[image]
>> +chainhandler=raw
>> +
>> +[image.selector]
>> +method=getroot_rrmap
>> +[kernel]
>> +chainhandler=rawfile
>> +
>> +[kernel.selector]
>> +method=getroot_rrmap
> 
> This is not being stressed by isar-cip-core, is it? Do we plan to add a
> target that uses legacy u-boot script-based a/b switching?

I had one request by downstream for supporting u-boot with suwpdate from 
cip-core.

Without this file the swupdate-helper will not build if U-Boot was 
selected as the bootloader.

If there is no plans for legacy u-boot a/b switching I would propose to 
remove the u-boot code from the swupdate FSTYPE.

Quirin

> 
> Jan
>
Jan Kiszka Aug. 10, 2023, 11:36 a.m. UTC | #3
On 10.08.23 13:26, Gylstorff Quirin wrote:
> 
> 
> On 8/10/23 13:01, Jan Kiszka wrote:
>> On 03.08.23 12:24, Quirin Gylstorff wrote:
>>> From: Quirin Gylstorff <quirin.gylstorff@siemens.com>
>>>
>>> This adds the missing configuration file to allow to build with u-boot
>>> as the boot selector.
>>>
>>> It uses the getroot mechanism from SWUpdate to select the update
>>> partition.
>>>
>>> Signed-off-by: Quirin Gylstorff <quirin.gylstorff@siemens.com>
>>> ---
>>> CC: felix.moessbauer@siemens.com
>>> TO: jan.kiszka@siemens.com
>>>
>>>   .../files/swupdate.handler.u-boot.ini                  | 10 ++++++++++
>>>   1 file changed, 10 insertions(+)
>>>   create mode 100644
>>> recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>>>
>>> diff --git
>>> a/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>>> new file mode 100644
>>> index 0000000..0e6df08
>>> --- /dev/null
>>> +++
>>> b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>>> @@ -0,0 +1,10 @@
>>> +[image]
>>> +chainhandler=raw
>>> +
>>> +[image.selector]
>>> +method=getroot_rrmap
>>> +[kernel]
>>> +chainhandler=rawfile
>>> +
>>> +[kernel.selector]
>>> +method=getroot_rrmap
>>
>> This is not being stressed by isar-cip-core, is it? Do we plan to add a
>> target that uses legacy u-boot script-based a/b switching?
> 
> I had one request by downstream for supporting u-boot with suwpdate from
> cip-core.
> 
> Without this file the swupdate-helper will not build if U-Boot was
> selected as the bootloader.
> 
> If there is no plans for legacy u-boot a/b switching I would propose to
> remove the u-boot code from the swupdate FSTYPE.

Well, if there is more than one request for that feature and adding this
file here would be the only reason for downstreams to override our
swupdate-handler-roundrobin recipe, we can take that (as "not tested").

How would the situation change if we get the handler merged into
swupdate upstream?

Jan
Gylstorff Quirin Aug. 10, 2023, 12:24 p.m. UTC | #4
On 8/10/23 13:36, Jan Kiszka wrote:
> On 10.08.23 13:26, Gylstorff Quirin wrote:
>>
>>
>> On 8/10/23 13:01, Jan Kiszka wrote:
>>> On 03.08.23 12:24, Quirin Gylstorff wrote:
>>>> From: Quirin Gylstorff <quirin.gylstorff@siemens.com>
>>>>
>>>> This adds the missing configuration file to allow to build with u-boot
>>>> as the boot selector.
>>>>
>>>> It uses the getroot mechanism from SWUpdate to select the update
>>>> partition.
>>>>
>>>> Signed-off-by: Quirin Gylstorff <quirin.gylstorff@siemens.com>
>>>> ---
>>>> CC: felix.moessbauer@siemens.com
>>>> TO: jan.kiszka@siemens.com
>>>>
>>>>    .../files/swupdate.handler.u-boot.ini                  | 10 ++++++++++
>>>>    1 file changed, 10 insertions(+)
>>>>    create mode 100644
>>>> recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>>>>
>>>> diff --git
>>>> a/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>>>> new file mode 100644
>>>> index 0000000..0e6df08
>>>> --- /dev/null
>>>> +++
>>>> b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
>>>> @@ -0,0 +1,10 @@
>>>> +[image]
>>>> +chainhandler=raw
>>>> +
>>>> +[image.selector]
>>>> +method=getroot_rrmap
>>>> +[kernel]
>>>> +chainhandler=rawfile
>>>> +
>>>> +[kernel.selector]
>>>> +method=getroot_rrmap
>>>
>>> This is not being stressed by isar-cip-core, is it? Do we plan to add a
>>> target that uses legacy u-boot script-based a/b switching?
>>
>> I had one request by downstream for supporting u-boot with suwpdate from
>> cip-core.
>>
>> Without this file the swupdate-helper will not build if U-Boot was
>> selected as the bootloader.
>>
>> If there is no plans for legacy u-boot a/b switching I would propose to
>> remove the u-boot code from the swupdate FSTYPE.
> 
> Well, if there is more than one request for that feature and adding this
> file here would be the only reason for downstreams to override our
> swupdate-handler-roundrobin recipe, we can take that (as "not tested").
> 
> How would the situation change if we get the handler merged into
> swupdate upstream?

The handler it self currently doesn't provide any default configuration.
So it could also fail.

Quirin
diff mbox series

Patch

diff --git a/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
new file mode 100644
index 0000000..0e6df08
--- /dev/null
+++ b/recipes-core/swupdate-handler-roundrobin/files/swupdate.handler.u-boot.ini
@@ -0,0 +1,10 @@ 
+[image]
+chainhandler=raw
+
+[image.selector]
+method=getroot_rrmap
+[kernel]
+chainhandler=rawfile
+
+[kernel.selector]
+method=getroot_rrmap