Re: Valgrind: was Bug: heap-buffer-overflow in function postprocess_terminfo

[Thomas Dickey]

On Sat, Oct 12, 2019 at 01:24:10AM -0400, Steve Litt wrote:
> On Fri, 11 Oct 2019 21:04:08 -0400
> Thomas Dickey <address@hidden> wrote:
> 
> > On Fri, Oct 11, 2019 at 08:49:12PM -0400, Thomas Dickey wrote:
>  
> > (I use asan occasionally, but valgrind frequently - it's slower but
> > usually more accurate).
> 
> Do you recompile to enable Valgrind without memory leak messages on
> valid programs, or has ncurses acquired the ability to specify valgrind
> compatability on the command line?

For the packages that I build for myself, I use the --with-trace and
--disable-leaks options (the rpm/deb scripts are in the packages
directory), and when testing use the private (non-api) calls to free
memory on exit.  (Making the library work properly after freeing all
of the memory work take some work; though it's doable - wasn't doable
before I (had to) get rid of the setbuf dependency).

My corresponding configure scripts (lynx, vile, xterm, dialog, etc.)
know about that in connection with their own debugging/no-leaks features.

Production code doesn't use the debugging features.
 
> If the former, how do you compile and store ncurses so it doesn't clash
> with the ncurses the rest of your computer uses?
> 
> Thanks,
> 
> Steve Litt
> Author: The Key to Everyday Excellence
> http://www.troubleshooters.com/key
> Twitter: http://www.twitter.com/stevelitt
> 
> 
> _______________________________________________
> Bug-ncurses mailing list
> address@hidden
> https://lists.gnu.org/mailman/listinfo/bug-ncurses

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

signature.asc
Description: PGP signature