Eli Schwartz authored Aug 05, 2019 and Christian Hesse committed Aug 06, 2019

Using popd at the very end of a shell script is unnecessary, because, as
the very last command, there is nothing to restore state for.
Immediately after, the shell subprocess is ended, and processes don't
control the cwd of the parent process. Changing the cwd for the last
microsecond of the shell process, during which no commands are run, is
a mildly expensive no-op.

By the same measure, if popd is never used, pushd is not needed to
record the old cwd. So simply use 'cd'.

Signed-off-by: Christian Hesse <mail@eworm.de>