On Fri, 27 Sep 2024 13:59:35 +0200
"Thomas Schmitt" <scdbac...@gmx.net> wrote:
> Hi,
>
> Glenn Washburn wrote:
> > [...] grub-shell-luks-tester cleans up after
> > itself, if it returns success. grub_cmd_cryptomount has a test that
> > expects failure. But grub-shell-luks-tester doesn't know that this is
> > an expected failure and should cleanup and grub_cmd_cryptomount doesn't
> > ever cleanup after grub-shell-luks-tester. Perhaps
> > grub-shell-luks-tester should be passed a parameter to indicate
> > expected failure (eg. --xfail). Have any other ideas?
>
> I agree to the idea of an option to invert the effect of
> [ "$RET:-1" -eq 0 ] in cleanup() of tests/util/grub-shell-luks-tester.in .
>
> Assuming variable "xfail" is be set to a non-empty string exactly if
> argument "--xfail" is given, i'd replace:
>
> if [ -z "$debug" ] && [ "${RET:-1}" -eq 0 ]; then
> rm -rf "$lukstestdir" || :
> fi
>
> by:
>
> if [ -z "$debug" ]; then
> if [ -n "$xfail" ]; then
> if [ "${RET:-0}" -ne 0 ]; then
> rm -rf "$lukstestdir" || :
> fi
> else
> if [ "${RET:-1}" -eq 0 ]; then
> rm -rf "$lukstestdir" || :
> fi
> fi
> fi
>
> Equivalent, but heavily economizing on line count would be:
>
> if [ -z "$debug" ] && [ -n "$xfail" ] && [ "${RET:-0}" -ne 0 ]; then
> rm -rf "$lukstestdir" || :
> elif [ -z "$debug" ] && [ -z "$xfail" ] && [ "${RET:-1}" -eq 0 ]; then
> rm -rf "$lukstestdir" || :
> fi
>
> I tested both code pieces in a dry-run script with the 12 variations of
> debug={"", "1"} , xfail={"", "1"} , RET={undefined, "0", "1"} .
RET should never be undefined because $? always has a numerical value.
And in my change I default xfail to 0. So I change the if statement to:
if [ -z "$debug" ] && [ "$RET" -eq "$xfail" ]; then
This is a lot simpler, though perhaps a tad less clear. Do you see a
reason why this is less desirable?
Glenn
> Decision for removal happened only with:
> (debug="", xfail="", RET="0")
> (debug="", xfail="1", RET="1")
>
> The first code performs less []-expressions and seems clearer to me at
> the price of doubled line count.
>
>
> Have a nice day :)
>
> Thomas
>
_______________________________________________
Grub-devel mailing list
Grub-devel@gnu.org
https://lists.gnu.org/mailman/listinfo/grub-devel
