Re: [PATCH] misc services patches with changes Dominick wanted

Chris PeBenito <[email protected]>
Newsgroups org.kernel.vger.selinux-refpolicy
Message-ID <[email protected]>
On 1/21/21 8:46 AM, Russell Coker wrote:
> This patch has some changes Dominick wanted and some parts that he disliked
> removed.  The one place where I didn't make his change I gave less access than
> he recommended.
> 
> I think this is ready for merging.
> 
> Signed-off-by: Russell Coker <[email protected]>

I think I'm ok with the changes, though I have a couple questions/comments:


> Index: refpolicy-2.20210120/policy/modules/services/apache.fc
> ===================================================================
> --- refpolicy-2.20210120.orig/policy/modules/services/apache.fc
> +++ refpolicy-2.20210120/policy/modules/services/apache.fc
> @@ -83,6 +83,8 @@ HOME_DIR/((www)|(web)|(public_html))(/.*
>   /usr/sbin/rotatelogs					--	gen_context(system_u:object_r:httpd_rotatelogs_exec_t,s0)
>   /usr/sbin/suexec					--	gen_context(system_u:object_r:httpd_suexec_exec_t,s0)
>   /usr/sbin/wigwam					--	gen_context(system_u:object_r:httpd_exec_t,s0)
> +/usr/sbin/php.*-fpm					--	gen_context(system_u:object_r:httpd_exec_t,s0)
> +/usr/sbin/php-fpm[^/]+					--	gen_context(system_u:object_r:httpd_exec_t,s0)

I can fix this when merging, but please keep the fc entries in order.


> @@ -71,6 +71,7 @@ template(`apache_content_template',`
>   
>   	manage_dirs_pattern(httpd_$1_script_t, httpd_$1_rw_content_t, httpd_$1_rw_content_t)
>   	manage_files_pattern(httpd_$1_script_t, httpd_$1_rw_content_t, httpd_$1_rw_content_t)
> +	allow httpd_$1_script_t httpd_$1_rw_content_t:file map;
>   	manage_lnk_files_pattern(httpd_$1_script_t, httpd_$1_rw_content_t, httpd_$1_rw_content_t)
>   	manage_fifo_files_pattern(httpd_$1_script_t, httpd_$1_rw_content_t, httpd_$1_rw_content_t)
>   	manage_sock_files_pattern(httpd_$1_script_t, httpd_$1_rw_content_t, httpd_$1_rw_content_t)

There's a lot of mmapping being added.  Can you provide any additional context 
on this?  Is this induced by some config option? Is this apache only?


> @@ -63,3 +63,23 @@ interface(`aptcacher_stream_connect',`
>   	files_search_runtime($1)
>   	stream_connect_pattern($1, aptcacher_runtime_t, aptcacher_runtime_t, aptcacher_t)
>   ')
> +
> +######################################
> +## <summary>
> +##     read aptcacher config
> +## </summary>
> +## <param name="domain">
> +##     <summary>
> +##     Domain allowed to read it.
> +##     </summary>
> +## </param>
> +#
> +interface(`aptcacher_read_config',`
> +	gen_require(`
> +		type aptcacher_etc_t;
> +	')
> +
> +	files_search_etc($1)
> +	allow $1 aptcacher_etc_t:dir list_dir_perms;
> +	allow $1 aptcacher_etc_t:file mmap_read_file_perms;
> +')

Is this the only useful way to read these files?  There's no valid non-mmap 
access?  If regular read can be useful, then this should be 
aptcatch_mmap_read_config().



> @@ -254,6 +255,7 @@ auth_use_nsswitch(cupsd_t)
>   
>   libs_read_lib_files(cupsd_t)
>   libs_exec_lib_files(cupsd_t)
> +libs_legacy_use_ld_so(cupsd_t)

This seems broken and should probably be in a debian distro block.



-- 
Chris PeBenito
lmpx.com only provides a reader for public news (NNTP) servers. It is not affiliated with the servers or forums shown here and is not responsible for the content of articles, which is written by their respective authors.