env: unwrap panick when called program exited with SIGSEGV

[jhasse]

When calling a program with env that segfaults, I'm getting:

thread 'main' panicked at 'called `Option::unwrap()` on a `None` value', src/uu/env/src/env.rs:311:67

With GNU env the error message is more helpful:

“env ./foo” terminated by signal SIGSEGV (Address boundary error)

[tertsdiepraam]

I've been doing some testing and it seems like the message is not part of env, but env just fails with the signal that the subcommand does. Here's an example using GNU env:

Setup

$ cat signal.c
#include <signal.h>
#include <stdlib.h>

int main(int argc, char** argv) {
    raise(atoi(argv[1]));
}
$ clang signal.c -o signal

Bash

$ ./signal 11
Segmentation fault (core dumped)
$ env ./signal 11
Segmentation fault (core dumped)

Zsh

$ ./signal 11
[1]    20685 segmentation fault (core dumped)  ./signal 11
$ env ./signal 11
[1]    20633 segmentation fault (core dumped)  env ./signal 11

Fish

$ ./signal 11
fish: Job 1, './signal 11' terminated by signal SIGSEGV (Address boundary error)
$ env ./signal 11
fish: Job 1, 'env ./signal 11' terminated by signal SIGSEGV (Address boundary error)

This is different from what the linked PR does (but should be easy to fix)