Fix yet another sed bug. The s/// command would copy the \ of substitutions before deciding what to do with them (generally overwriting the \ with the new data). When the substitution was A) at the very end of the new string, B) resolved to nothing, it could leave a trailing \ that didn't belong there and didn't get overwritten because the "copy trailing data" part that copies the original string's null terminator already happened before the \ overwrote it. The ghostwheel() function restarts regexes after embedded NUL bytes, but if the string it's passed is _longer_ than the length it's told then it gets confused (and it means we're off the end of our allocation so segfaults are likely). Fix: test for \ first and move the "copy byte" logic into an else case.

#!/bin/bash

[ -f testing.sh ] && . testing.sh

#testing "name" "command" "result" "infile" "stdin"

testing "mkfifo" "mkfifo one && [ -p one ] && echo yes" "yes\n" "" ""
rm one

touch existing
testing "mkfifo existing" \
	"mkfifo existing 2> /dev/null || [ -f existing ] && echo yes" "yes\n" "" ""
rm existing

testing "mkfifo one two" \
	"mkfifo one two && [ -p one ] && [ -p two ] && echo yes" "yes\n" "" ""
rm one two

umask 123
testing "mkfifo (default permissions)" \
	"mkfifo one && stat -c %a one" "644\n" "" ""
rm one

umask 000

testing "mkfifo -m 124" \
	"mkfifo -m 124 one && stat -c %a one" "124\n" "" ""
rm -f one