Re: current bash.git.devel segfaults on my code

[Alex fxmbsw7 Ratchev]

could you give me maybe good urls about such to learn ( maybe not too long
ones .. ) ?

i can paste the two missing functions
or even paste the archive where they come from ( this same project )

it may also be non bash errors cause more newer sw used than usual.. no
idea..

to paste minimally additions, here are the two funcs
they are for code shortening there, not always to retype etc..

mkPATH is a function
mkPATH ()
{
    declare pre onlyargs IFS=: T=;
    declare -A t=();
    end='PATH=$T' per "${pre[@]}" ${onlyargs-$PATH} "$@"
'T=${T:+$T}${t[$per]-${t[1]+:}$per} t[$per]= t[1]='
}

per is a function
per ()
{
    declare per pre stu beg end;
    eval -- "$pre";
    for per in "${@:1:$#-1}";
    do
        eval -- "$beg
${!#}
$end";
    done;
    eval -- "$stu"
}

On Tue, Oct 12, 2021, 21:10 Chet Ramey <chet.ramey@case.edu> wrote:

> On 10/11/21 11:17 AM, Alex fxmbsw7 Ratchev wrote:
> > as the topic says
> > i fetched it earlier today, compiled, and tho it rather segfaults
> >
> > due to i dunno to debug this, this mail consists of 3 parts, 1 gdb + bt
> > ./bash (loading in bashrc a script which makesnit segfault)
> > and once with gdb ./bash -vx
> > third part is the code
> > forth part is around where in the code it happens
>
> I can't reproduce it. `mkPATH' isn't there, and there's probably a lot of
> your environment that needs to be reproduced for it to work. It looks like
> somewhere a free list block is getting clobbered, but I can't really tell.
> Since I can't reproduce it, you'll have to see if you can get more detail
> from something like valgrind or address sanitizer.
>
> Chet
>
> --
> ``The lyf so short, the craft so long to lerne.'' - Chaucer
>                  ``Ars longa, vita brevis'' - Hippocrates
> Chet Ramey, UTech, CWRU    chet@case.edu    http://tiswww.cwru.edu/~chet/
>