From: Chris PeBenito <pebenito@ieee.org>
To: Lukas Vrabec <lvrabec@redhat.com>, selinux-refpolicy@vger.kernel.org
Subject: Re: [PATCH] Introduce allow_bluetooth boolean
Date: Sun, 5 May 2019 20:36:09 -0400 [thread overview]
Message-ID: <e6b8b03f-fff9-e247-2b4c-52b9e533ce3c@ieee.org> (raw)
In-Reply-To: <5a75d7e6-c56b-e4f5-653c-b846388cfbb4@redhat.com>
On 5/4/19 6:26 AM, Lukas Vrabec wrote:
> On 5/1/19 1:37 AM, Chris PeBenito wrote:
>> On 4/30/19 1:53 PM, Lukas Vrabec wrote:
>>> Because of new bluetooth_socket which is part of extended_socket_class
>>> policy capability, it's possible to control which SELinux domains can
>>> use bluetooth wireless technology. Default value of the boolean is
>>> turned off.
>>>
>>> Signed-off-by: Lukas Vrabec <lvrabec@redhat.com>
>>> ---
>>> policy/global_tunables | 7 +++++++
>>> policy/modules/admin/netutils.te | 4 ++++
>>> policy/modules/services/arpwatch.te | 4 ++++
>>> policy/modules/services/bluetooth.if | 4 ++++
>>> policy/modules/services/bluetooth.te | 4 ++++
>>> policy/modules/services/ntop.te | 4 ++++
>>> policy/modules/system/userdomain.if | 8 ++++++++
>>> 7 files changed, 35 insertions(+)
>>>
>>> diff --git a/policy/global_tunables b/policy/global_tunables
>>> index affc020f..69b356e2 100644
>>> --- a/policy/global_tunables
>>> +++ b/policy/global_tunables
>>> @@ -4,6 +4,13 @@
>>> # file should be used.
>>> #
>>> +## <desc>
>>> +## <p>
>>> +## Allow all system processes and Linux users to use bluetooth
>>> wireless technology.
>>> +## </p>
>>> +## </desc>
>>> +gen_tunable(allow_bluetooth,false)
>>> +
>>> ## <desc>
>>> ## <p>
>>> ## Allow unconfined executables to make their heap memory
>>> executable. Doing this is a really bad idea. Probably indicates a
>>> badly coded executable, but could indicate an attack. This executable
>>> should be reported in bugzilla
>>> diff --git a/policy/modules/admin/netutils.te
>>> b/policy/modules/admin/netutils.te
>>> index 46560a09..8821b108 100644
>>> --- a/policy/modules/admin/netutils.te
>>> +++ b/policy/modules/admin/netutils.te
>>> @@ -99,6 +99,10 @@ optional_policy(`
>>> xen_append_log(netutils_t)
>>> ')
>>> +tunable_policy(`allow_bluetooth',`
>>> + allow netutils_t self:bluetooth_socket
>>> create_stream_socket_perms;
>>> +')
>>> +
>>> ########################################
>>> #
>>> # Ping local policy
>>> diff --git a/policy/modules/services/arpwatch.te
>>> b/policy/modules/services/arpwatch.te
>>> index 87aed96f..6f05441a 100644
>>> --- a/policy/modules/services/arpwatch.te
>>> +++ b/policy/modules/services/arpwatch.te
>>> @@ -77,6 +77,10 @@ miscfiles_read_localization(arpwatch_t)
>>> userdom_dontaudit_search_user_home_dirs(arpwatch_t)
>>> userdom_dontaudit_use_unpriv_user_fds(arpwatch_t)
>>> +tunable_policy(`allow_bluetooth',`
>>> + allow arpwatch_t self:bluetooth_socket create_socket_perms;
>>> +')
>>> +
>>> optional_policy(`
>>> mta_send_mail(arpwatch_t)
>>> ')
>>> diff --git a/policy/modules/services/bluetooth.if
>>> b/policy/modules/services/bluetooth.if
>>> index dc61988c..9097803f 100644
>>> --- a/policy/modules/services/bluetooth.if
>>> +++ b/policy/modules/services/bluetooth.if
>>> @@ -68,6 +68,10 @@ interface(`bluetooth_stream_connect',`
>>> files_search_pids($1)
>>> allow $1 bluetooth_t:socket rw_socket_perms;
>>> stream_connect_pattern($1, bluetooth_var_run_t,
>>> bluetooth_var_run_t, bluetooth_t)
>>> +
>>> + tunable_policy(`deny_bluetooth',`',`
>>
>> Missed a deny_bluetooth.
>>
>> I'm also a little uneasy with adding this access here, as this would
>> seem to be a side effect to a simple stream socket connection. Perhaps
>> there should be a new interface like bluetooth_client(). It might also
>> make sense in bluetooth_role(), as there is a generic socket class rule
>> in there too.
>>
>
> Hi Chris,
>
> I'll update deny_bluetooth() to allow_bluetooth().
>
> So, you're suggesting create new interface bluetooth_client() but all
> bluetooth related stuff there, and use this interface in tunable_policy
> block instead of just allowing using bluetooth socket?
bluetooth_client() would have the rules of bluetooth_stream_connect()
and then the allow_bluetooth conditional rule. The stream socket
connection can remain unconditional, if the bluetooth_socket access is
sufficient to block bluetooth use.
> bluetooth_role() should be fixed to use only bluetooth_socket, not the
> general one. I'll create PR also for it.
>
> Thanks,
> Lukas.
>
>>
>>
>>> + allow $1 bluetooth_t:bluetooth_socket rw_socket_perms;
>>> + ')
>>> ')
>>> ########################################
>>> diff --git a/policy/modules/services/bluetooth.te
>>> b/policy/modules/services/bluetooth.te
>>> index 45e5a361..39af72d5 100644
>>> --- a/policy/modules/services/bluetooth.te
>>> +++ b/policy/modules/services/bluetooth.te
>>> @@ -133,6 +133,10 @@ userdom_dontaudit_use_unpriv_user_fds(bluetooth_t)
>>> userdom_dontaudit_use_user_terminals(bluetooth_t)
>>> userdom_dontaudit_search_user_home_dirs(bluetooth_t)
>>> +tunable_policy(`allow_bluetooth',`
>>> + allow bluetooth_t self:bluetooth_socket create_stream_socket_perms;
>>> +')
>>> +
>>> optional_policy(`
>>> dbus_system_bus_client(bluetooth_t)
>>> dbus_connect_system_bus(bluetooth_t)
>>> diff --git a/policy/modules/services/ntop.te
>>> b/policy/modules/services/ntop.te
>>> index 178bbb1d..537d9323 100644
>>> --- a/policy/modules/services/ntop.te
>>> +++ b/policy/modules/services/ntop.te
>>> @@ -96,6 +96,10 @@ miscfiles_read_localization(ntop_t)
>>> userdom_dontaudit_use_unpriv_user_fds(ntop_t)
>>> userdom_dontaudit_search_user_home_dirs(ntop_t)
>>> +tunable_policy(`allow_bluetooth',`
>>> + allow ntop_t self:bluetooth_socket create_socket_perms;
>>> +')
>>> +
>>> optional_policy(`
>>> apache_read_sys_content(ntop_t)
>>> ')
>>> diff --git a/policy/modules/system/userdomain.if
>>> b/policy/modules/system/userdomain.if
>>> index 5221bd13..5654de6e 100644
>>> --- a/policy/modules/system/userdomain.if
>>> +++ b/policy/modules/system/userdomain.if
>>> @@ -1189,6 +1189,10 @@ template(`userdom_unpriv_user_template', `
>>> corenet_udp_bind_generic_port($1_t)
>>> ')
>>> + tunable_policy(`allow_bluetooth',`
>>> + allow $1_t self:bluetooth_socket create_socket_perms;
>>> + ')
>>> +
>>> optional_policy(`
>>> netutils_run_ping_cond($1_t, $1_r)
>>> netutils_run_traceroute_cond($1_t, $1_r)
>>> @@ -1362,6 +1366,10 @@ template(`userdom_admin_user_template',`
>>> userdom_manage_user_home_content_sockets($1_t)
>>> userdom_user_home_dir_filetrans_user_home_content($1_t, { dir
>>> file lnk_file fifo_file sock_file })
>>> + tunable_policy(`allow_bluetooth',`
>>> + allow $1_t self:bluetooth_socket create_stream_socket_perms;
>>> + ')
>>> +
>>> optional_policy(`
>>> postgresql_unconfined($1_t)
>>> ')
>>>
>>
>>
>
>
--
Chris PeBenito
next prev parent reply other threads:[~2019-05-06 0:36 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-04-30 17:53 [PATCH] Introduce allow_bluetooth boolean Lukas Vrabec
2019-04-30 23:37 ` Chris PeBenito
2019-05-04 10:26 ` Lukas Vrabec
2019-05-06 0:36 ` Chris PeBenito [this message]
-- strict thread matches above, loose matches on Subject: below --
2019-04-26 9:27 Lukas Vrabec
2019-04-26 18:52 ` Dominick Grift
2019-04-27 17:16 ` Dominick Grift
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=e6b8b03f-fff9-e247-2b4c-52b9e533ce3c@ieee.org \
--to=pebenito@ieee.org \
--cc=lvrabec@redhat.com \
--cc=selinux-refpolicy@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).