Re: [PATCH] Check for ncursesw first when searching for "tgetent"

[Sergio Durigan Junior <sergiodj at redhat dot com>]

On Tuesday, February 13 2018, I wrote:

> On Monday, February 12 2018, Simon Marchi wrote:
>
>> On 2018-01-31 04:04 PM, Sergio Durigan Junior wrote:
>>> Commit 5007d765ae09c10c7f3b18bb16841b9d2d59e181 ("Allow linking GDB
>>> with ncursesw") modified our configure.ac and included the check for
>>> "ncursesw" when searching for "waddstr".  However, there's one more
>>> place where we should check for "ncursesw" first:
>>> 
>>>   AC_SEARCH_LIBS(tgetent, [termcap tinfo curses ncursesw ncurses])
>>> 
>>> This patch changes the order of the libraries to be searched when
>>> looking for "tgetent", and puts "ncursesw" before "curses ...".
>>> 
>>> This is another patch we carry on Fedora GDB.
>>
>> I think it makes sense, but can you expand on why this is needed?
>
> The rationale for this patch was:
>
>   https://bugzilla.redhat.com/show_bug.cgi?id=1270534
>
> This is a bug that happened on Fedora GDB when linking against ncurses,
> but not ncursesw.  The bug has been fixed upstream by commit
> 5007d765ae09c10c7f3b18bb16841b9d2d59e181, but the AC_SEARCH_LIBS line
> for tgetent has not been modified, so, for the sake of completeness, I
> think it makes sense to leave configure.ac in a consistent state (i.e.,
> requiring ncursesw over ncurses whenever applicable).
>
> I guess that's the gist of it.  There's not much that can be said, the
> patch is really simple and its intention is to make things more uniform.
>
> I can include the link to the Red Hat bug in the commit message, if you
> want.

After talking to Pedro in private, I was convinced that this is not the
right explanation for the patch, either.  Therefore, I will write a
better commit message and submit a v2, hopefully addressing your
concerns.

Thanks,

-- 
Sergio
GPG key ID: 237A 54B1 0287 28BF 00EF  31F4 D0EB 7628 65FC 5E36
Please send encrypted e-mail if possible
http://sergiodj.net/