[PATCH] setfiles: clarify documented path resolution behaviour

[PATCH] setfiles: clarify documented path resolution behaviour

[Jonathan Lebon]

One thing that confused me when investigating
https://github.com/SELinuxProject/selinux/issues/248 (i.e.
https://github.com/coreos/fedora-coreos-tracker/issues/512) was that the
manual page for `setfiles` seemed to imply that paths were fully
resolved. This was consistent with the issues above where `setfiles` was
failing because the target of the symbolic link didn't exist.

But in fact, the wording around symbolic links in
`setfiles`/`restorecon` refers actually to whether the parent
directories are canonicalized via `realpath(3)` before labeling.

Clarify the man pages to explain this.
---
 policycoreutils/setfiles/restorecon.8 | 4 ++--
 policycoreutils/setfiles/setfiles.8   | 3 ++-
 2 files changed, 4 insertions(+), 3 deletions(-)

diff --git a/policycoreutils/setfiles/restorecon.8 b/policycoreutils/setfiles/restorecon.8
index bbfc83fe..06ec5a1d 100644
--- a/policycoreutils/setfiles/restorecon.8
+++ b/policycoreutils/setfiles/restorecon.8
@@ -159,8 +159,8 @@ The pathname for the file(s) to be relabeled.
 .SH "NOTES"
 .IP "1." 4
 .B restorecon
-does not follow symbolic links and by default it does not
-operate recursively on directories.
+by default does not operate recursively on directories. Parent directories
+are fully resolved before labeling.
 .IP "2." 4
 If the
 .I pathname
diff --git a/policycoreutils/setfiles/setfiles.8 b/policycoreutils/setfiles/setfiles.8
index 0188a75a..12f41967 100644
--- a/policycoreutils/setfiles/setfiles.8
+++ b/policycoreutils/setfiles/setfiles.8
@@ -214,7 +214,8 @@ option is used.
 .SH "NOTES"
 .IP "1." 4
 .B setfiles
-follows symbolic links and operates recursively on directories.
+operates recursively on directories. Parent directories are not fully
+resolved before labeling.
 .IP "2." 4
 If the
 .I pathname
-- 
2.26.2

Re: [PATCH] setfiles: clarify documented path resolution behaviour

[Stephen Smalley]

On Thu, Jun 4, 2020 at 3:26 PM Jonathan Lebon <jlebon@redhat.com> wrote:
>
> One thing that confused me when investigating
> https://github.com/SELinuxProject/selinux/issues/248 (i.e.
> https://github.com/coreos/fedora-coreos-tracker/issues/512) was that the
> manual page for `setfiles` seemed to imply that paths were fully
> resolved. This was consistent with the issues above where `setfiles` was
> failing because the target of the symbolic link didn't exist.
>
> But in fact, the wording around symbolic links in
> `setfiles`/`restorecon` refers actually to whether the parent
> directories are canonicalized via `realpath(3)` before labeling.
>
> Clarify the man pages to explain this.

Missing Signed-off-by line.  Otherwise I guess it is an improvement
although I think it is still somewhat unclear.
Maybe we should explicitly say that they are canonicalized via
realpath(3) in the man page.

> ---
>  policycoreutils/setfiles/restorecon.8 | 4 ++--
>  policycoreutils/setfiles/setfiles.8   | 3 ++-
>  2 files changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/policycoreutils/setfiles/restorecon.8 b/policycoreutils/setfiles/restorecon.8
> index bbfc83fe..06ec5a1d 100644
> --- a/policycoreutils/setfiles/restorecon.8
> +++ b/policycoreutils/setfiles/restorecon.8
> @@ -159,8 +159,8 @@ The pathname for the file(s) to be relabeled.
>  .SH "NOTES"
>  .IP "1." 4
>  .B restorecon
> -does not follow symbolic links and by default it does not
> -operate recursively on directories.
> +by default does not operate recursively on directories. Parent directories
> +are fully resolved before labeling.
>  .IP "2." 4
>  If the
>  .I pathname
> diff --git a/policycoreutils/setfiles/setfiles.8 b/policycoreutils/setfiles/setfiles.8
> index 0188a75a..12f41967 100644
> --- a/policycoreutils/setfiles/setfiles.8
> +++ b/policycoreutils/setfiles/setfiles.8
> @@ -214,7 +214,8 @@ option is used.
>  .SH "NOTES"
>  .IP "1." 4
>  .B setfiles
> -follows symbolic links and operates recursively on directories.
> +operates recursively on directories. Parent directories are not fully
> +resolved before labeling.
>  .IP "2." 4
>  If the
>  .I pathname
> --
> 2.26.2
>

Re: [PATCH] setfiles: clarify documented path resolution behaviour

[Jonathan Lebon]

On Mon, Jun 8, 2020 at 11:19 AM Stephen Smalley
<stephen.smalley.work@gmail.com> wrote:
> Missing Signed-off-by line.  Otherwise I guess it is an improvement
> although I think it is still somewhat unclear.
> Maybe we should explicitly say that they are canonicalized via
> realpath(3) in the man page.

Thanks, I spelled it out even more in v2.