Re: 4.3.90 release candidate segfault on Mac OS 12.6

[John Graham-Cumming]

$ ls -al
total 608
drwxr-xr-x   4 jgc  staff     128 22 Sep 20:58 .
drwx------@ 14 jgc  staff     448 22 Sep 20:58 ..
-rw-r--r--   1 jgc  staff      24 22 Sep 20:58 Makefile
-rwxr-xr-x   1 jgc  staff  304384 22 Sep 20:58 make

$ cat Makefile
$(info $(wildcard foo))

$ ./make
zsh: segmentation fault  ./make

$ sudo gdb ./make

GNU gdb (GDB) 12.1
Copyright (C) 2022 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.
Type "show copying" and "show warranty" for details.
This GDB was configured as "x86_64-apple-darwin21.5.0".
Type "show configuration" for configuration details.
For bug reporting instructions, please see:
<https://www.gnu.org/software/gdb/bugs/>.
Find the GDB manual and other documentation resources online at:
    <http://www.gnu.org/software/gdb/documentation/>.

For help, type "help".
Type "apropos word" to search for commands related to "word"...
Reading symbols from ./make...
(gdb) run
Starting program: /Users/jgc/Downloads/empty/make
[New Thread 0x1903 of process 73263]
[New Thread 0x2503 of process 73263]
warning: unhandled dyld version (17)

Thread 2 received signal SIGSEGV, Segmentation fault.
0x000000010001e7fe in parse_file_seq (stringp=<optimized out>, stringp@entry=0x7ff7bfefe868, size=size@entry=16, stopmap=stopmap@entry=1, prefix=prefix@entry=0x0, flags=flags@entry=25) at src/read.c:3529
3529          NEWELT (concat (2, prefix, nlist[i]));
(gdb) p i
$1 = 0
(gdb) p nlist
$2 = <optimized out>

$ pwd
/Users/jgc/Downloads/make-4.3.90
$ ls ./lib/libgnu_a-*
./lib/libgnu_a-concat-filename.o ./lib/libgnu_a-fnmatch.o
./lib/libgnu_a-findprog-in.o ./lib/libgnu_a-glob.o

On Thu, Sep 22, 2022 at 8:39 PM Paul Smith <psmith@gnu.org> wrote:

On Thu, 2022-09-22 at 20:34 +0100, John Graham-Cumming wrote:
> Simply did ./configure. It’s an Intel Mac. Compiler seems to be
> clang. 
>
> Doing a clean build shows a few warnings when compiling.

Those are not a problem.

In addition to answers to the questions I asked below, can you send the
output of "ls -al" in the directory where you're running make?  Maybe
there's some other file in that directory which is causing the glob to
return something weird.

> > > Thread 2 received signal SIGSEGV, Segmentation fault.
> > > 0x000000010001e7fe in parse_file_seq (stringp=<optimized out>,
> > > stringp@entry=0x7ff7bfefe868, size=size@entry=16,
> > > stopmap=stopmap@entry=1, prefix=prefix@entry=0x0,
> > > flags=flags@entry=25) at src/read.c:3529
> > > 3529          NEWELT (concat (2, prefix, nlist[i]));
> >
> > Thanks.  Can you check the value of "i" and the contents of the
> > "nlist" array here?
> >
> > I'm also curious whether the configure step decides to use the
> > system glob and fnmatch.  It should not, because GNU make expects
> > the GNU version of glob, so it should build its own.  You can see
> > if these files exist in the build tree after make is built:
> >
> >   ./lib/libgnu_a-glob.o
> >   ./lib/libgnu_a-fnmatch.o