diff options
author | Max Rees <maxcrees@me.com> | 2019-06-02 17:58:55 -0400 |
---|---|---|
committer | Max Rees <maxcrees@me.com> | 2019-09-28 13:27:48 -0500 |
commit | 37dc8b13f40d85ee22938ab60bbc6f03d823f593 (patch) | |
tree | c375f6468f8a8f0b9d51b3c05302fe85d7b2cc3e /system/make | |
parent | 3b0612cdf8eef6408efe10cce2bc9e48d9c9d1fc (diff) | |
download | packages-37dc8b13f40d85ee22938ab60bbc6f03d823f593.tar.gz packages-37dc8b13f40d85ee22938ab60bbc6f03d823f593.tar.bz2 packages-37dc8b13f40d85ee22938ab60bbc6f03d823f593.tar.xz packages-37dc8b13f40d85ee22938ab60bbc6f03d823f593.zip |
system/coreutils: fix test failure with bind mounts
df-symlink tests whether df output is consistent in regards to where
devices are mounted. Specifically, it tries to ensure that if there is a
1:1 mapping of the device ("source") corresponding to the current
directory "." ("target"), then `df --output=source,target .` should be
the same as `df --output=source,target $(df --out=source . | tail -n1)`.
However, this is not necessarily the case in the presence of bind
mounts. For example:
$ df --out=source .
Filesystem
/dev/sda4
$ df --out=source,target /dev/sda4
Filesystem Mounted on
/dev/sda4 /
$ df --out=source,target .
Filesystem Mounted on
/dev/sda4 /git
This can be fixed by adding the --all flag to ensure that duplicate
devices are printed, so that the 1:1-checking if statement actually
checks it correctly.
Diffstat (limited to 'system/make')
0 files changed, 0 insertions, 0 deletions