[PATCH v2] Obsolete support for Solaris < 10

[Jan Beulich]

On 27.11.2025 11:11, Rainer Orth wrote:
> Hi Jan,
> 
>> On 27.11.2025 10:27, Rainer Orth wrote:
>>>> On 27.11.2025 10:04, Rainer Orth wrote:
>>>>> Support for Solaris 8 and 9 has long been removed from GCC:
>>>>>
>>>>> * Solaris 8 support in GCC 4.8 (2013)
>>>>>
>>>>> * Solaris 9 support in GCC 5 (2015)
>>>>>
>>>>> so it's time to obsolete anything before Solaris 10 in binutils, too.
>>>>
>>>> But according to the actual code change and ...
>>>>
>>>>> Both haven't even built since binutils 2.41.
>>>>
>>>> ... (aiui) this related statement, they were already marked obsolete. What
>>>> instead you do is un-obsolete Solaris 2.10 and 2.11. That's fine with me,
>>>> but needs describing accordingly (including to make the subject unambiguous).
>>>
>>> I think you misread the patch and the situation: On current trunk, no
>>> Solaris version at all is obsoleted.
>>
>> Hmm, yes, I indeed misread it. But then what's the 2.41 related sentence about?
> 
> It just says that even though Solaris 8/9 weren't obsoleted, they didn't
> build for quite some time and no one noticed/complained.  This might
> even argue for immediate removal instead of obsoletion.

Oh, I see what you mean. Note that as a non-native speaker it was more
natural (in fact automatic until I started writing this reply) to insert
the seemingly missing "been" to make that sentence "Both haven't even
been built since binutils 2.41." May I suggest "Both wouldn't even build
since binutils 2.41"?

>> Also, please don't drop the null part. See the later case command about removed
>> targets, where the null is retained despite there being a long list. Keeping it
>> will make more obvious what it should go back to when the Solaris cases are
>> moved to that other case command.
> 
> Good point.  Here's the revised patch, retested as before, description
> and ChangeLog unchanged.

Okay preferably with the small wording adjustment above.

Jan