[SRU,Xenial] UBUNTU: Add changelog entry for arm64 module loading issue

BugLink: https://bugs.launchpad.net/bugs/1695093

Ignore: yes
Signed-off-by: dann frazier <dann.frazier@canonical.com>
---
 debian.master/changelog | 3 +++
 1 file changed, 3 insertions(+)

On Wed, Oct 4, 2017 at 12:44 PM, dann frazier
<dann.frazier@canonical.com> wrote:
> BugLink: https://bugs.launchpad.net/bugs/1695093
>
> Ignore: yes
> Signed-off-by: dann frazier <dann.frazier@canonical.com>
> ---
>  debian.master/changelog | 3 +++
>  1 file changed, 3 insertions(+)

fyi, I'm not sure this is the right way to do a changelog-only commit,
but it seems to work with insertchanges. I'll wait for confirmation
before submitting a similar one for linux-hwe.

On 04.10.2017 20:44, dann frazier wrote:
> BugLink: https://bugs.launchpad.net/bugs/1695093
> 
> Ignore: yes
> Signed-off-by: dann frazier <dann.frazier@canonical.com>
> ---
>  debian.master/changelog | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/debian.master/changelog b/debian.master/changelog
> index e9228573d648..79a40c8bbab9 100644
> --- a/debian.master/changelog
> +++ b/debian.master/changelog
> @@ -1,5 +1,8 @@
>  linux (4.4.0-98.121) UNRELEASED; urgency=low
>  
> +  * Rebuild against gcc-5_5.4.0-6ubuntu1~16.04.5 to fix loading of
> +    certain modules on arm64 (LP: #1695093)
> +
>    CHANGELOG: Do not edit directly. Autogenerated at release.
>    CHANGELOG: Use the printchanges target to see the curent changes.
>    CHANGELOG: Use the insertchanges target to create the final log.
> 

I suppose it would not break the insertion of the additional changelog entries.
Though it could become a pain in case of unexpected changes of versions. And
then, how would this match with reality? The only trigger would be some rebuild
but whether that really was the expected gcc version this does not say.

As much pita it is, I think for such a case someone has to follow the cycle and
manually twiddle bug states.

-Stefan

On Fri, Oct 6, 2017 at 4:51 AM, Stefan Bader <stefan.bader@canonical.com> wrote:
> On 04.10.2017 20:44, dann frazier wrote:
>> BugLink: https://bugs.launchpad.net/bugs/1695093
>>
>> Ignore: yes
>> Signed-off-by: dann frazier <dann.frazier@canonical.com>
>> ---
>>  debian.master/changelog | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/debian.master/changelog b/debian.master/changelog
>> index e9228573d648..79a40c8bbab9 100644
>> --- a/debian.master/changelog
>> +++ b/debian.master/changelog
>> @@ -1,5 +1,8 @@
>>  linux (4.4.0-98.121) UNRELEASED; urgency=low
>>
>> +  * Rebuild against gcc-5_5.4.0-6ubuntu1~16.04.5 to fix loading of
>> +    certain modules on arm64 (LP: #1695093)
>> +
>>    CHANGELOG: Do not edit directly. Autogenerated at release.
>>    CHANGELOG: Use the printchanges target to see the curent changes.
>>    CHANGELOG: Use the insertchanges target to create the final log.
>>
>
> I suppose it would not break the insertion of the additional changelog entries.
> Though it could become a pain in case of unexpected changes of versions.

Do you mean unexpected gcc version changes? There are none in the SRU
queue, so I think that is low risk. However, I could change
"gcc-5_5.4.0-6ubuntu1~16.04.5" to "gcc-5 >= 5.4.0-6ubuntu1~16.04.5" to
future proof it.

> And
> then, how would this match with reality? The only trigger would be some rebuild
> but whether that really was the expected gcc version this does not say.
>
> As much pita it is, I think for such a case someone has to follow the cycle and
> manually twiddle bug states.

If that needs to be the case, c'est la vie, but if it's just the
wording, I'm happy to fix that up.

 -dann