Обсуждение: Remove more leftovers of AIX support

Поиск
Список
Период
Сортировка

Remove more leftovers of AIX support

От
Peter Eisentraut
Дата:
The make variables MKLDEXPORT and POSTGRES_IMP were only used for AIX,
so they should have been removed with commit 0b16bb8776b.
Вложения

Re: Remove more leftovers of AIX support

От
Heikki Linnakangas
Дата:
On 20/01/2026 08:08, Peter Eisentraut wrote:
> The make variables MKLDEXPORT and POSTGRES_IMP were only used for AIX,
> so they should have been removed with commit 0b16bb8776b.

+1, thanks! Even if we resurrect AIX support later, it's better to 
remove these now, so that we have all the AIX-related changes in the 
same commit that re-introduces the support.

- Heikki




Re: Remove more leftovers of AIX support

От
Tom Lane
Дата:
Heikki Linnakangas <hlinnaka@iki.fi> writes:
> On 20/01/2026 08:08, Peter Eisentraut wrote:
>> The make variables MKLDEXPORT and POSTGRES_IMP were only used for AIX,
>> so they should have been removed with commit 0b16bb8776b.

> +1, thanks! Even if we resurrect AIX support later, it's better to 
> remove these now, so that we have all the AIX-related changes in the 
> same commit that re-introduces the support.

Agreed.  Also, if this or similar code is still needed in AIX,
then presumably the meson infrastructure will need an equivalent.
It'll be confusing if the eventual re-introduction patch touches
only one of the two.

            regards, tom lane



Re: Remove more leftovers of AIX support

От
Peter Eisentraut
Дата:
On 20.01.26 15:57, Tom Lane wrote:
> Heikki Linnakangas <hlinnaka@iki.fi> writes:
>> On 20/01/2026 08:08, Peter Eisentraut wrote:
>>> The make variables MKLDEXPORT and POSTGRES_IMP were only used for AIX,
>>> so they should have been removed with commit 0b16bb8776b.
> 
>> +1, thanks! Even if we resurrect AIX support later, it's better to
>> remove these now, so that we have all the AIX-related changes in the
>> same commit that re-introduces the support.
> 
> Agreed.  Also, if this or similar code is still needed in AIX,
> then presumably the meson infrastructure will need an equivalent.
> It'll be confusing if the eventual re-introduction patch touches
> only one of the two.

committed