bug-ncurses
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Some symbols missed in llib-ltinfo and llib-ltinfot for 6.1?


From: Thomas Dickey
Subject: Re: Some symbols missed in llib-ltinfo and llib-ltinfot for 6.1?
Date: Tue, 27 Feb 2018 16:18:04 -0500
User-agent: Mutt/1.5.23 (2014-03-12)

On Tue, Feb 27, 2018 at 10:37:27AM +0100, Dr. Werner Fink wrote:
> On Tue, Feb 27, 2018 at 09:23:24AM +0000, Thomas Dickey wrote:
> > On Tue, Feb 27, 2018 at 08:22:46AM +0100, Dr. Werner Fink wrote:
> > > 
> > > OK ... the only problem seems to be that there are programs out there
> > > which links to both libtinfo.so.6.1 (threadsafe) as well as to
> > > libtinfow.so.6.1 (wide+threadsafe) and in the case od phython3 modules
> > > this could lead to a deadlock if the order is wrong, in other word
> > 
> > I recall that there was (perhaps still) a problem with python loading
> > different versions of ncurses for readline and the curses module.
> > Ultimately the fix for that was within python.
> 
> IMHO this is more complicated, please have a look at my previous followup

well - looks like there's only a few possible ways to improve:

a) supposing that if the symbol versions (not the symbols) for "ncurseswt"
   and "ncursest" version changed (by a configure option of course) to
   make them distinct from the symbol versions in "ncursesw"/"ncurses",
   then you could find out whether the linker is able to properly resolve
   two sets of symbol versions for the same symbols (I suspect that it is
   not).
b) link everything against "ncurseswt"

-- 
Thomas E. Dickey <address@hidden>
https://invisible-island.net
ftp://ftp.invisible-island.net

Attachment: signature.asc
Description: Digital signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]